PART TWO
ECONOMIC AND ENVIRONMENTAL CHANGE

6
Four Centuries of Agropastoral Change

It has widely been assumed that Khumbu agriculture and pastoralism were relatively static, traditional practices for many generations before they were transformed by a variety of changes after 1960. The crop- and stock-raising practices that Western visitors first observed in the 1950s were believed to be a traditional set of practices that had been practiced since well before 1900. As far as was known or suspected the history of Khumbu agriculture consisted of three major phases. The first of these was an early period about which little was known, but during which Sherpas presumably practiced a form of mixed agropastoralism that they had brought with them from Tibet. Following this was a period that began in the middle of the nineteenth century and lasted until the 1960s and which was characterized by the distinctive reliance on the potato as the main staple crop of the region and the nak as the mainstay of pastoralism. This period, finally, was superseded during the past thirty years by an era in which traditional crop-growing and stock-raising practices have been undermined and dramatically transformed by the far-reaching effects of tourism on regional economy and land use (Bjønness 1983; J. Fisher 1990; Fürer-Haimendorf 1984).

Sherpas do not support this view of their past. Khumbu oral traditions and oral history contradict the idea that the pre-1960, potato-based agriculture and nak-based pastoralism were a relatively static way of life for many generations. They tell a story instead of a more dynamic history of innovation and adaptation. Sherpas suggest that since the late nineteenth century few aspects of crop production or pastoralism have remained static. The crop repertoire, cropping patterns, total amount of

land under cultivation, technology, agricultural knowledge and belief, and community regulation of agricultural practices have all changed significantly.^[1] Khumbu pastoralism historically has been even more dynamic than crop production, with major changes during the past century and a half in the types of stock raised, herd size, the goals and operation of local pastoral management systems, and seasonal herd-movement patterns. Yet at the same time Sherpas also testify to considerable continuity underlying these economic transformations. They dispute the contention that the changes of the past thirty years constitute a fundamental break from their long-characteristic subsistence strategies and practices. Khumbu agriculture and pastoralism are changing, but Sherpas are changing them within a context of continuing affirmation of local knowledge, cultural traditions, and valued land-use practices and lifestyles.

This chapter surveys Khumbu agricultural and pastoral history from the Sherpa perspective from the early days of settlement until the recent past. The oral traditions and history relating to changes in the subsistence economy, like those concerning the settlement history which I discussed in chapter 1, become richer in detail in the nineteenth and twentieth centuries. Although there are suggestions of what life was like and how it changed in earlier periods, it is only possible to analyze economic change in depth over the past hundred years. This, though, is still plenty of ground to work with—sufficient to establish both the dynamism of change in the "traditional" subsistence economy before 1960 and the processes involved. This in turn also establishes a better baseline for evaluating the changes of the recent past, some of which reflect continuing processes such as agricultural intensification, crop diffusion, and deemphasis of commercial stockbreeding which have been at work in Khumbu for a long time. I leave for chapter 10 the reassessment of what has and has not been transformed during the past thirty years by the impact of tourism.

Early Khumbu Agriculture

Khumbu Sherpas have probably practiced settled, mixed agropastoralism since the early days of their settlement of the region. Oral traditions about the first settlers and early life in the Bhote Kosi valley mention the growing of crops, the keeping of yak and sheep, and the use of manure for fertilizer. Not much more can be said of what was grown or how in those early years, or of when the pattern familiar for the past century and more of movement among main villages and herding settlements was established. In the early days it seems very likely that herders lived in black yak-hair tents in the summer rather than in the now-characteristic stone herding huts. Such tents figure in early stories from

Dzongnangpa's era and according to oral history were used in both the upper Dudh Kosi and the upper Bhote Kosi valley as recently as the late nineteenth century by some families. The idea that Sherpas were nomadic prior to the introduction of the potato in the mid-nineteenth century (Bjønness 1980a :61; Hardie 1957) may have come from stories of black-tent pastoralists. But abundant oral traditions testify that Sherpas have been a village-based people since the early days of their settlement of the region, and all evidence indicates that the main settlements familiar today had all been developed by 1830.

Oral traditions and oral history begin to give a firmer portrait of regional crop-growing practices in the late nineteenth century. By then the pattern of multialtitudinal crop production in the gunsas, main villages, secondary high-altitude crop-production sites, and high-altitude herding areas had already been established. All of the current main and secondary herding settlements had been settled and the crops familiar since then as the staples of Khumbu agriculture—potatoes, barley, and buckwheat—were all being grown.^[2] But crop-production emphases were rather different. Potatoes were just one of several tuber crops rather than the dominant crop of the region and were not being grown at all in some of the high-altitude herding settlements or at Dingboche. Radish and turnip were being grown as major field crops, suggesting that they may have been the staple pre-potato tuber crops.^[3] Buckwheat then as today was the most widely grown grain, but it was grown on a much larger scale than in the twentieth century including in places such as Tarnga, Thamicho, and Nauje where it is not grown today. Barley too was more widely cultivated and was being grown at several places in the Bhote Kosi valley.

Barley was apparently grown more widely in the nineteenth century than it has been in this century, although the need to irrigate it must have always made it of less regional significance than rain-watered buckwheat. In the late nineteenth century barley was cultivated not only at Dingboche but also at Tarnga in the Bhote Kosi valley and perhaps other places as well.^[4] An elderly Thami Teng couple recalled seeing barley at Tarnga before 1920 while herding yak there. Other Sherpas recall that gembu Tsepal of Nauje was supposed to have owned barley land at Tarnga as well as at Dingboche in the early years of the century.^[5]

Tarnga may have been the major agricultural center of Khumbu at one time and in the nineteenth century may have equaled or exceeded Dingboche as a barley center. There are large numbers of abandoned terraces above the settlement which were probably irrigated in an earlier time, and much of the land that is currently in potato terraces may have once been irrigated with a system that was far more complex than that employed at Dingboche. Elderly Tarnga residents have heard that for-

merly there was an irrigation ditch leading from the Chu Nasa creek north of the settlement.^[6] Higher up the creek there is evidence of an old diversion and ditch that lead more than a kilometer from the creek to the head of the slope above the settlement. From here the irrigation network appears to have branched. One channel contoured along the slope to the west whereas the other descended through a flight of now-abandoned, upper-slope terraces. This system could conceivably have watered much of the field area of Tarnga.

The decline and disappearance of barley cultivation in the Bhote Kosi valley remains a mystery. There are no oral traditions describing the decline of barley growing in Tarnga. People have heard that formerly there was a prohibition on growing buckwheat there for fear that it would offend the barley. Barley, they note, has indeed disappeared from the area, and suggest that it might have declined on account of the introduction of buckwheat. Buckwheat is remembered to have flourished at Tarnga early in the century when it was a noted center of buckwheat production.^[7] Beyond this they have no explanation.

Several other factors could conceivably have caused a decline of barley production at Tarnga and the abandonment there of many terraces. Crop disease and declining soil fertility might have figured in farmers' decisions; so too might the increasing scarcity of available land as the population rose, making higher-yielding buckwheat a more attractive option. Buckwheat could also have produced better crops on less fertile soils. Yet it is difficult to envision the total abandonment of barley for these reasons. Barley is still cultivated on a large scale at Dingboche despite the apparently poorer soils there and greater land scarcity than there would presumably have been in the nineteenth century at Tarnga. The cultural value and social status involved with growing barley remain strong. Yet even wealthy Sherpas with land at Tarnga have not grown even small patches of barley for many decades.

The most likely cause for the decline of barley cultivation at Tarnga may have been a failure of the irrigation network. The creek flow during May when the irrigation of barley is most critical is minor, barely enough to send a trickle of water through the ditch that supplies drinking water to the settlement. At one time more water was presumably available to feed the far more extensive earlier irrigation network. A declining flow of irrigation water could well have led to disputes over access to water and to the gradual abandonment of barley cultivation on lower-slope terraces. Once the available water declined significantly it is possible that there might not have been enough community enthusiasm to keep the out-take channel and ditches clear, and barley cultivation even on a limited scale would then not have been possible. Terraces would have been converted to a rotation of rain-watered buckwheat and

tubers, and the upper-slope terraces that had formerly been valued for their ready access to irrigation may have been considered to be too marginal in terms of soil fertility to be worth planting to unirrigated crops.

The Introduction of the Potato

It has been widely assumed that the potato reached Khumbu about 1850, but nothing is known of the original variety introduced, its source, the exact date and place of its introduction, or the pace of its diffusion within Khumbu.^[8] Fürer-Haimendorf believed that Khumbu Sherpas began planting potatoes "about the middle of the nineteenth century". According to his interpretation of two oral-history accounts the new crop had diffused even to such a "conservative" village as Phurtse "soon after 1860" (Fürer-Haimendorf 1964:9). Although his analysis of these oral-history accounts may have been incorrect, as I discuss below, an introduction before the last quarter of the nineteenth century is still likely. Presumably potatoes were well established in Khumbu before 1866-1876 when Rolwaling oral traditions reported by Sacherer testify that they were brought to Rolwaling by settlers from the Bhote Kosi valley (J. Sacherer, "The Sherpas of Rolwaling: A Hundred Years of Economic Change," 1977, in Seddon 1987).

Potatoes could have reached Khumbu in the nineteenth century from several sources including Kathmandu, Darjeeling, and Tibet. William Kirkpatrick (1975:180), an early English envoy to Kathmandu, reported potatoes there in 1793 and noted that the seed potatoes had to be brought each year from the Patna region of India. They were probably introduced to Darjeeling sometime soon after the British established a hill-station settlement there in 1835 and were certainly being cultivated there by 1848 when they were noticed by Sir Joseph Dalton Hooker (Hooker 1969:230). By the mid-nineteenth century potatoes could also have been grown in southern Tibet. George Bogle, who journeyed to Shigatse in 1774, had been instructed by the governor general of India, Warren Hastings, to introduce potatoes to regions through which he passed (Sandburg 1987:103). It is not known whether Bogle did establish potato growing in Tibet, but he is thought to have successfully introduced the crop to Bhutan (ibid.:103) and it might well have reached Tibet from there. Grenard (1974:247) reported potatoes in Lhasa in 1893.

The case for a Kathmandu origin for potatoes in eastern Nepal dates back to Hooker's explorations of northeasternmost Nepal in 1848. He noted that potatoes were being grown in the high-altitude fields at Yangma and at Kambachen in the Gunsa area northeast of Taplejung

where he was given "some red potatos, about as big as walnuts" (1969:247). At Yangma, a place very near the Tibet border and inhabited by "Tibetans" (possibly Sherpas or Bhotias), he remarked that

there was no food to be procured except a little thin milk, and a few watery potatos [sic ]. The latter have only recently been introduced amongst the Tibetans, from the English garden at the Nepalese capital, I believe; and their culture has not spread in these regions further east than Kinchinjunga [Kachenjunga], but they will very soon penetrate into Tibet from Dorjiling [Darjeeling], or eastward from Nepal. (ibid.:230)

Hooker apparently rejected the idea that the Yangma potatoes had come from Darjeeling, despite the proximity of the two regions, because he had not found them cultivated in the intervening area.

Fører-Haimendorf (1979:8-9) noted that Kathmandu and Darjeeling were the two most likely sources for the diffusion of potatoes into eastern Nepal. He raised the possibility that Hooker was wrong and that Darjeeling was the origin for both Khumbu and Yangma potatoes. Oral traditions he heard in Khumbu in 1953 seemed to indicate an introduction there after 1850, a date that would correlate well with earlier cultivation at Yangma if potatoes had reached eastern Nepal from Darjeeling (ibid.:9).

Both Fürer-Haimendorf's and Hooker's speculations about the origin of potatoes in eastern Nepal are based on assumptions that diffusion takes place in an orderly progression and constant pace across space. Such assumptions do not seem to be very well supported by what is known of the history of the diffusion of crop varieties in the region in the twentieth century. Diffusion is a complex process and is often difficult to predict. Other factors besides location and distance are clearly important. Peoples, villages, families, and individuals differ in their degree of receptivity to new crops. Regional patterns of trade and travel may bring one people into contact with a new crop rather than another people who may inhabit a region closer to the source. It only takes a single accidental encounter for someone to recognize a new crop and introduce it to his or her home region. Twentieth-century introductions of other potato varieties into Khumbu illustrate how important a factor chance can be. No analysis of typical Sherpa trade patterns would have concluded, for instance, that during the 1970s a potato found in a teashop on a trail near Darjeeling would come to dominate Bhote Kosi valley agriculture or that an apparently identical variety would be found just a few years later in a monastery kitchen north of Kathmandu and would rapidly diffuse from Khumjung to become the major variety grown in eastern Khumbu. Nor would it have seemed likely that in at least three twentieth-century

cases Khumbu Sherpas would have acquired new potato varieties from Sikkim, Darjeeling, and central Nepal before other Sherpa groups and Rais who lived considerably closer to those source areas, or that potatoes brought from these remote areas to Khumbu would have subsequently diffused from Khumbu into areas that might have been expected to have had them long before.

In the mid-nineteenth century Khumbu Sherpas did not frequent either Kathmandu or Darjeeling. Khumbu was politically and culturally more linked to the Tibetan world than to the Kathmandu valley and very few Khumbu Sherpas went to Kathmandu for trade or pilgrimage. Before 1850 they also probably had little to do with Darjeeling, although later in the century Sherpas went to the British hill station to trade and to work as porters and at other occupations and in the twentieth century many Khumbu Sherpas went to Darjeeling to work for mountaineering expeditions. The lack of major contact with Kathmandu or Darjeeling does not mean, however, that a single Sherpa trader or a pilgrim visiting either area could not have brought back a few potatoes to Khumbu, or that Sherpas might not have encountered the crop in Shorung or other areas where Khumbu Sherpas traded and traveled frequently and which were in closer contact with Kathmandu.^[9]

As important as the introduction of the potato is to Khumbu economic history there are unfortunately no surviving oral traditions today in Khumbu which describe the event. Fürer-Haimendorf (1979:9), however, had the opportunity to hear two accounts of the introduction of the potato when he began fieldwork in Khumbu in 1953. One of these testimonies was given by an eighty-three-year-old woman of Thami who is now deceased. The other was offered by a Phurtse resident, Sun Tenzing. From these two testimonies Fürer-Haimendorf (ibid.:9) arrived at a hypothesis that the original introduction of the potato probably took place not long before the 1860s. It is worth looking more closely at both the accounts and their interpretation.

It is immediately evident that the two testimonies differ considerably in detail and that in both cases it would be useful to know more about the varieties involved and the time referred to. Whereas the Phurtse account specified the location of the first fields planted and gave a fairly specific time frame, the Thami account was extremely general. For Thami we learn only that an eighty-three-year-old woman of that village (Thami Og?) "told me in 1953 that potatoes were brought to her village by people of her father's generation" (ibid.). By itself this account does not establish very much, for the language used makes it difficult to narrow the suggested time frame. A considerable span of years could fall within the era of "her father's generation." The introduction might have occurred before her birth, during her childhood, or conceivably even

during the time before she married and left her parent's household. In the latter case, the time being discussed could be the very end of the nineteenth century.

The Phurtse account was rather different. Fürer-Haimendorf records that:

In 1953 Sun Tensing of Phortse, then a man in his middle forties, told me that as a young boy he knew an old man of over ninety of whom it was said that he had first planted potatoes on Phortse land and I was shown the plots of land on the right bank of the Imja Khola, roughly opposite Milingo, where these first potato fields were supposed to have been. (ibid. :90)

Here again, however, crucial facts are missing. The account does not, for instance, indicate how old the man referred to was at the time he planted these potatoes, which variety of potatoes he planted, where he obtained them, or how long it was until potatoes were planted in the village of Phurtse itself.

In 1985 I had the opportunity to seek these details from Sun Tenzing, with whom I had a number of discussions. According to Sun Tenzing a man named Zoa Dolma did indeed introduce a new variety of potatoes at Tsadorji, a gunsa settlement east of Phurtse on the northern, right bank of the Imja Khola and across the river from Milingo. Zoa Dolma had brought nine small, long, white potatoes to Khumbu which he said that he had obtained from Darjeeling and which he believed had come from Belait (England). This variety, which is today most often referred to as a type of kyuma, was therefore also called riki belati (English potato) by many people. He brought this potato to Khumbu, however, half a century later than Fürer-Haimendorf had supposed. The potato variety had not been introduced long before Sun Tenzing's birth but when he was a boy of about five years of age, around 1914. At that time Zoa Dolma had been in his fifties, for in his earlier account Sun Tenzing had meant to convey that Zoa Dolma would have been in his nineties had he been alive in 1953 when he had told his story to Fürer-Haimendorf. Equally as startling, in the context of further questioning he also pointed out that while Zoa Dolma had brought the first kyuma or Belati to the Phurtse area he had not introduced the first potatoes there. Before the long white potato was introduced by Zoa Dolma, he noted, Phurtse villagers were already planting a small, round potato known as riki koru (white potato).^[10] . The introduction of this potato took place before his birth and he knew nothing of the circumstances or date. I was not able to uncover any of these details from other Khumbu Sherpas either.^[11]

There were thus apparently several early introductions of potato varieties to Khumbu (table 20), at least one of which came from Darjeeling. The earliest variety was introduced before 1900 and was being cultivated at least several decades earlier, if Sacherer is correct in her hypothesis that potatoes reached Rolwaling from Khumbu by 1876 ("The Sherpas of Rolwaling: A Hundred Years of Economic Change," 1977, in Seddon 1957). But the original date and source of introduction is more uncertain today than ever.

Early Twentieth-Century Agriculture

By the beginning of the twentieth century the potato was apparently being grown in all Khumbu villages and in most of the secondary sites where it is grown today.^[12] By the 1920s, and perhaps for quite a while earlier, kyuma was planted in the Bhote Kosi valley throughout the altitudinal range from the low gunsa to high phu. In the Dudh Kosi valley it was grown not only in the main village of Phurtse but also in Na and other high-altitude areas east of the river (although apparently it was not cultivated at west-bank sites such as Luza and Dole where today it is grown on a small scale). Other potatoes were also being grown widely. But although widespread and important to regional agriculture and household subsistence, the potato had not yet come to dominate Khumbu crop production in the way that it would a few years later and

does today. At this time potatoes were being grown on less than 50 percent of the cultivated area and indeed were only one of several tuber crops rather than the sole tuber grown. Elderly Sherpas maintain that during their youths (ca. 1910-1925) four crops were important in Khumbu: barley, buckwheat, potato, and radish. A fifth crop, turnip, was then being grown on a small scale. In most of Khumbu potatoes were less important a crop in the first part of the century than they have been in recent decades. In the Bhote Kosi valley there are reports that radish and turnip were grown on as large a scale as potatoes were in Thami Og and Thami Teng and that unlike today large areas of land were also in buckwheat. Estimates of the amount of buckwheat cultivation vary from half of all land in Thami Teng and Tarnga to perhaps only a quarter of the cultivated area of Thami Og. Radish was also being grown at Tarnga as a major crop, and the area was then considered to be excellent for radish growing just as it is for potato cultivation today. From Phurtse there are also reports that in the early decades of the century the potato was just one of several tubers, not the most important one. In Khumjung, by contrast, several elderly Sherpas remember that even during the 1920s the potato was the main tuber, and by then it was the major crop of Nauje and was being monocropped by many families.

Not one but rather three potato varieties were important in this era: the long kyuma whose introduction to the Phurtse area was discussed by Sun Tenzing, the early, round, white potato grown there (henceforth referred to as koru ), and a second variety of white, round potato (koru 2 ) which was apparently introduced after kyuma in the 1920s.^[13] These varieties seem to have diffused at different paces in different valleys. Kyuma, for example, was evidently grown decades earlier in the Bhote Kosi valley than at Phurtse, for Thamicho Sherpas older than Sun Tenzing do not remember the introduction of kyuma during their lifetimes and have not even heard stories about it from their parents. The same is true in Khumjung and in Nauje.

Many elderly people remember both kyuma and koru 2 as the potatoes of their youths.^[14] Kyuma was higher yielding and more reliable, although it was also notorious for producing extremely poor crops in bad years and for being highly susceptible to devastating blight infections. Many farmers recall times when a full day's work at harvesting kyuma would not yield enough potatoes to serve the work crew for lunch. Accounts about the productivity of kyuma, however, are inconsistent, for the variety did supplant the earlier koru and was still grown on at least a small scale in much of Khumbu until about twenty-five years ago. Good harvests are also reported, and clearly some families produced surpluses of kyuma. Some dried kyuma tubers were even exported to Tibet on a small scale.^[15] That farmers remember the productivity of

kyuma differently may reflect differences in experience as well as in memory. Harvests could have been better in some areas and fields than in others due to differences in local soil and microclimatic conditions, altitude, disease problems, seed stock, or planting and manuring practices. Some Sherpas may tend to best remember the disastrous years and others the good ones. Some may best remember the variety from more recent years when it may have produced more poorly than it formerly did, possibly reflecting a loss of disease resistance or a regional decline in soil fertility. And others may base their evaluation not on productivity per field but on recollections of harvest size relative to family requirements, an equation that would have been different in earlier periods when so many families spent the winter outside of Khumbu on trade and pilgrimage journeys to lower regions of Nepal. Certainly on a Khumbu-wide level a surplus would have been achieved earlier in the century at a much lower level of regional production than at present, for the population of Khumbu was much smaller then and mostly only resident in the region for part of the year, and locally grown potatoes were not also consumed by thousands of tourists as they are today.

The Potato Revolution Reexamined

It has been commonly assumed that the potato was widely and quickly adopted throughout the region and that this radically transformed local agriculture in the nineteenth century in ways that also had major implications for regional demographic and sociocultural change. As already mentioned Hardie has suggested that before the potato was introduced Sherpas were nomadic pastoralists, a view that has been recently echoed by BjØnness (BjØnness 1980a ; Hardie 1957) whereas Fürer-Haimendorf remarked that "it is difficult to imagine conditions before potatoes found their way into Khumbu" (1979:8). Fürer-Haimendorf ascribed to the adoption of potato cultivation both a population boom and a new level of prosperity that made possible a flowering of Sherpa culture in the late nineteenth and early twentieth centuries which included the building of an unprecedented number of temples, shrines, and religious monuments and the founding of the region's first monasteries (ibid.:10-11).

Links between the Asian and European adoption of New World food crops and major population gain have been noted in a number of parts of the world, including the adoption of potatoes in Switzerland (Netting 1981), and maize in Nepal (MacFarlane 1976). Fürer-Haimendorf attributed a similar growth of population in Khumbu to potatoes, noting that

the population of Khumbu was a fraction of its present size until the middle of the nineteenth century and there can be no doubt that the great

increase of the last hundred years coincided with the introduction and spread of the potato. . . . No great imagination is required to realize that the introduction of a new crop and the spectacular increase in population must have been connected. (1979:10)

Even leaving aside for the moment the basic question of the degree to which potato cultivation transformed Khumbu agriculture in the nineteenth century, it seems that this issue deserves further exploration for the data on which to base such conclusions are rather slender. Indeed, the only available data on regional population size during the entire period from 1800 to 1957 are Fürer-Haimendorf's count of Khumbu households in 1957 and a count that he derived from 1836 tax documents (1979:5, 11,118). The 1836 document lists only 169 households in Khumbu whereas in 1957 the count was 597. The regional difference of 428 households established over this 121-year period represents the supposed major population growth that has been presumed to have been precipitated by the introduction of the potato in-the mid-nineteenth century. This tripling of the population (assuming that the average regional household size was similar throughout this period and that the number of households was tallied equally well on both occasions) is roughly comparable to some estimates of the population growth during the same period in Nepal as a whole, where demographic change is also believed to have been related to agricultural innovation, the introduction of maize, and the diffusion of the practice of growing irrigated rice in permanent terraced fields (MacFarlane 1976).^[16]

Although the potato (along with increased trade for grain) undoubtedly provided the means to support the higher regional population density that developed after the early nineteenth century it remains premature to assert that the adoption of the crop caused this population growth. It is not clear how much of the increase in regional population can be attributed to natural population growth and how much is due to immigration. Natural rates of population increase may indeed have been raised by higher food availability and better nutrition, which in turn may have lowered mortality rates, an argument that Netting (1981) has made for the introduction of the potato in Switzerland. But it is also certain that a great deal of the regional population growth in the late nineteenth and twentieth centuries has reflected migration from Tibet, as is indicated by the high percentage of Khamba households in the region in 1957. According to Fürer-Haimendorf's 1957 data (1979:102-104) (and assuming that at least half of Nauje's households then were Khamba) at least 45 percent of the households in five of the eight main Khumbu villages were Khamba immigrants from Tibet. This means that the number of Sherpa households in Khumbu less than doubled over the 121

years from 1836 to 1957. This does not seem like a major population boom. It may, however, still represent a significant increase in population growth rates over earlier centuries. And it may also be that greater potato cultivation was a factor in the increased immigration during this period. Surplus Khumbu food production and the ability to pay wage laborers in food, for example, would have helped create a regional economic climate attractive to immigrants. But a great number of other factors were also very likely involved and the role of the potato in fostering immigration should probably not be overly stressed.

Beyond the issue of the degree of natural population increase there is the more fundamental question of whether the potato was at the center of a revolution in agricultural production in nineteenth-century Khumbu. I have suggested that the adoption and diffusion of potatoes was not as rapid as had been assumed. The potato-based agriculture familiar to foreign visitors since the 1950s was not universal in Khumbu even earlier in this century, much less in the nineteenth. The potato's rise to preeminence in Khumbu agriculture was a much more complex and lengthy process than has been thought. Not only was the area in potato cultivation during the late nineteenth century and early twentieth centuries apparently considerably less than that planted to the tuber during the past fifty or sixty years, but yields were also apparently lower and poor harvests due to disease more common.

The initial diffusion of the potato was still in progress as late as the 1920s and in at least one important crop-producing settlement there remained powerful resistance to the adoption of the new crop. At Dingboche at that time the planting of potatoes was forbidden out of fear that the crop might offend the barley.^[17] The ban was upheld into the twentieth century by community understanding (a community which then comprised individuals from at least four Khumbu villages). Gembu Tsepal of Nauje, Khumbu's major political leader of the time and a large landowner at Dingboche, and his son and successor gembu Pasang are both said to have been particularly concerned with upholding this ban.^[18] This attitude toward potatoes was not universally shared by Dingboche residents, and the lama Kuyung Rinpoche is said to have moved from the area after a conflict with gembu Tsepal over potato planting. Eventually the degree of community agreement on the ban deteriorated to the point that it was no longer enforceable. But it was only in about 1925 that Dilu, the mother of the powerful Kunde pembu Ang Chumbi, precipitated the abandonment of the ban by advising other women at Dingboche to plant whatever they wanted to, insisting that gembu Pasang had no authority to tell them what they could or could not plant in their own fields.^[19]

The process of the adoption and diffusion of the potato thus did not

take place overnight in Khumbu, but involved incorporating the new tuber in an earlier cropping complex where it supplemented but did not immediately replace tubers cultivated earlier. Adoption of the new crop involved the development of new knowledge and new beliefs, including the abandonment of a belief that potatoes threatened barley yields and the invention of community measures to guard against devastating outbreaks of late blight. Not one but several different introductions of different varieties of potatoes were involved, and although the process of adapting the new cultigen to local agricultural patterns had advanced considerably even as early as the beginning of the twentieth century, the potato did not become the mainstay of Khumbu agriculture until the widespread cultivation of a newer, higher-yielding variety, the red potato, in the 1930s.

This reevaluation of the historical process of the adoption of the potatoes has significance for the second part of Fürer-Haimendorf's potato-revolution thesis as well, the idea that there was a connection between the adoption of the potato and the flourishing of religion in Khumbu. According to him:

the foundation of monasteries and nunneries as well as the construction of new village temples and many religious monuments have taken place within the last fifty to eighty years. This points to economic events which favoured a sudden spurt of non-productive activities and in my opinion there can be little doubt that these events were brought about by the introduction of the potato and the resulting increase in agricultural production. (1979:10-11)

Here too, however, a more complex process appears to have been involved, and the role of higher-agricultural yields may have been relatively minor. Closer examination is necessary of the circumstances of the construction of religious buildings and monuments, and especially of the sources of capital and labor which made this possible. Labor for these projects seems to have been, then as now, primarily volunteer labor from nearby communities.^[20] The ability of communities to mobilize this labor is based on whether local leaders can inspire participation or demand it by virtue of their offices. There is no direct link beteen local food supplies and the building of religious monuments. Even the availability of leisure time is not really an issue, for in the Khumbu agropastoral system there has long been a substantial part of the year in which there are no pressing demands. Khumbu life is built on the basis of a considerable amount of leisure, and the development of the local festival cycle can only be understood in this light. The establishment and maintenance of monasteries, temples, and monuments, however, also

requires substantial capital, for institutions must be endowed, specialists employed in their construction and sometimes in their maintenance, and many precious materials obtained for building and blessing the shrines. Yet here a larger role was often played by a few wealthy individuals rather than by the community as a whole. These sponsors were not men who had become wealthy as a result of potato growing, but for the most part were large-scale traders in other merchandise. In some cases they were not even Khumbu Sherpas: Shorung Sherpas were especially important in the original establishment of the Tengboche monastery (see Ortner 1989)^[21]

There is a problem too with the chronology of the development of monumental religious architecture in the region. It is not entirely accurate to emphasize the post-1880 period only, for most temples, chortens, and smaller religious monuments were actually established earlier. All of the village temples other than that at Nauje date to 1830 or earlier. Almost all the large chortens were built before 1880 (the main exceptions are two smaller chortens at Khumjung, one of which dates to the 1920s and the other to 1984, and three chortens in the Bhote Kosi valley). Most major mani walls (religious monuments composed of slabs of stone into which have been chiseled prayers and sacred texts) also predate 1880, although parts of the Khumjung mani wall and the mani wall at Tengboche were built after 1915. Most of the prominent expressions of religion faith in the Khumbu landscape thus seem to have been there well before the potato came to be a central focus of local agriculture in the early twentieth century. It may be that both before 1880 and since then the galvanizing force for the construction of religious monuments has not been the potato but rather the leadership of charismatic individuals who were able to inspire the financial support of well-to-do Sherpas and mobilize the volunteer labor of entire communities. And rather than a single period of major elaboration of monumental religious architecture there appear to have been a number of such periods stretching far back into Khumbu history.

Although the introduction of the potato may not have immediately been the seminal event in Sherpa history as has been assumed, it was nevertheless an important agricultural innovation that ultimately had far-reaching impacts on land use. As potatoes began to be more widely cultivated in the late nineteenth century they must have begun to increase regional agricultural production while at the same time lowering the amount of land required to meet household subsistence requirements. They provided a means of agricultural intensification, and the Sherpa response to this possibility may account not only for historical shifts in cropping emphases but also for changes in the area in crops and the abandonment of some fields in many parts of Khumbu.

Abandoned Terraces

By the late nineteenth century a great deal of land was under cultivation in Khumbu despite the relatively light (even by Khumbu standards) population density. In the Bhote Kosi valley more land seems to have been cultivated during the nineteenth century than has been in production since, a phenomenon reflected in the considerable amount of terraced cropland there that has been abandoned for many decades. Although some of these old fields were returned to cultivation later, a great many have never been reclaimed.^[22] Abandoned terraces are also found today in widely scattered sites throughout much of Khumbu, as indicated in map 14. Table 21 identifies sites with abandoned terraces and attempts a rough estimate of the periods when different sites were taken out of cultivation.^[23] The most extensive abandoned areas are located in the Bhote Kosi valley. It is clear from this map that there is an extraordinary number of sites with abandoned land in this part of the region and that these areas are situated through much of the altitudinal spectrum of Khumbu farming. More than forty terraces were abandoned at Tarnga. At Samde much terraced land has been neglected and there are the ruins

Map 14.
Abandoned Terraces

of eleven abandoned houses as well. Five houses and associated crop-land have been abandoned at Samshing where only two houses remain. And all of the more than twenty houses at Leve were abandoned before 1930 and not a single one of the scores of terraces there is now farmed. There is also a considerable number of abandoned fields at Thami Og as well as some in the other Thamicho main settlements

The immediate cause of abandonment can be determined in some cases, especially for those areas that were cultivated as recently as 1930. Diverse factors were involved. At Samshing and Samde emigration and the death of childless couples were factors. In Phurtse some terraces were also abandoned by families who emigrated to Darjeeling. Some fields abandoned near Thami Teng in the early years of the century are said to have been neglected because they had been producing poor crops. Terraces near Kenzuma in the Dudh Kosi valley were abandoned by Nauje settlers, it is said, because of repeated crop losses due to the depredations of langur monkeys more than half a century ago. Problems with Himalayan tahr were a factor in the gradual abandonment of crop production at Tashilung in the lower Bhote Kosi valley near Nauje over the past fifty years. Tahr were also a factor in some families' decisions to give up farming at nearby Nyeshe.^[24]

The reasons for the abandonment of terraces before 1930 are harder to evaluate. The abandonment of the settlement of Leve, situated only a few minutes' walk from the major villages of the Bhote Kosi valley, for example, remains somewhat mysterious. There are legends describing bad luck at the place, but few details of how the houses and terraces came to be neglected. Some traditions regarding a few of the families who abandoned the settlement, however, suggest that emigration was a factor in some cases. Two families left Leve for Rolwaling and other families simply gave up growing crops at the site and concentrated instead on their nearby fields at Thami Og and other valley locations.^[25] It is conceivable that the underlying factor in all of these cases was a perception that Leve had become marginal for crop production. This would also explain why no one has taken up growing crops there since. But there is no insight into this in the oral traditions. It is possible that disease may also have played some role. When traveling through Khumbu in 1885 Hari Ram reported that there had been an outbreak of smallpox in the 1850s (Ortner 1989:208, n. 4). Again, however, there is no hint in surviving oral traditions about an epidemic in Khumbu, much less a link between this and the subsequent abandonment of any settlements.

Agricultural intensification may also have been a factor. Sacherer cites this as the explanation for old abandoned terraces in Rolwaling, which she attributes to the adoption of potato cultivation. The new high

yields possible with the new crop, she suggests, rendered cultivation of some marginal land superfluous.

There can be no doubt at all that the introduction of the potato brought about an economic revolution in Rolwaling where today one can easily observe the presence of abandoned fields in the more rocky and inaccessible high altitude areas despite the fact that the valley population has increased since the conversion from barley to potatoes. (J. Sacherer, "The Sherpas of Rolwaling: A Hundred Years of Economic Change," 1977, in Seddon 1987)

In Khumbu a similar phenomenon could well have occurred in the late nineteenth and early twentieth century. Even if the adoption of the potato in the region was a more gradual process than has previously been assumed and if harvests were relatively small and variable by the standards of recent decades, the new crop may still have been more productive than the other available tuber crops of that era.

The location of abandoned terraces in the Bhote Kosi valley may lend some support to an intensification hypothesis. Two kinds of sites are most common: small numbers of old, unfarmed terraces located either at the edges of current settlements or at a slight distance beyond them and more extensive, abandoned terraces in some gunsa settlements. Abandoned terraces at the edges of main villages and secondary high-altitude sites may well have been considered marginal land. The abandonment of so much gunsa land requires more detailed consideration. The degree to which fields have been carved out in the steep slopes of the lower Bhote Kosi valley for gunsa fields is unparalleled elsewhere in Khumbu. The greater historical intensity of land use in the Bhote Kosi valley may well reflect greater nineteenth- and early-twentieth-century population pressure there than in Khumbu's other valleys, which led people to cultivate even quite minor patches of relatively easily terraced land. The demand for land also would have been greater in this period because more land would have been required per household for subsistence in the time before the introduction of higher-yielding varieties of potatoes. In this era it was also more common to cultivate grain as well as potatoes and to try to harvest enough to fulfill most family grain requirements.^[26] Families may thus have needed to farm up to twice the amount of land that they do today. Gunsa land would have been highly appealing in this situation due to the advantages that it offers for labor scheduling. The attempt to cultivate substantially more land in the main village could have strained family labor resources to carry out planting during the relatively brief spring planting period even if the field area

had been available. Making use of gunsa lands would have expanded the agricultural season by nearly two months. Families from Thami Og, Thami Teng, and Yulajung could have planted large amounts of land as early as the beginning of March rather than waiting until late April.

The later adoption of higher-yielding potato varieties and the ability to grow more food on less land might well have led to a consolidation process as families devoted less time and energy to farming and relied on the larger harvests now possible on their best fields in main villages and high-altitude secondary agricultural sites. Presumably the further introduction of even higher-yielding potato varieties meant that many terraces were never returned to cultivation, although some were certainly sold to new immigrants.

Agricultural Change 1930-1973

Agricultural change can be reconstructed in much more detail for eras within living memory, and far more careful cross-checking of sources is possible for the period after 1930 than for earlier eras. This oral history testimony suggests that the middle of the twentieth century was characterized by an increasing emphasis on potato cultivation. This was associated with the adoption of another variety of potato, riki rnoru (red potato). This round, red-skinned, pink-flowered potato was both better yielding and more consistent in yield than earlier Khumbu varieties had been. It was adopted throughout the region and supplanted almost all other tuber cultivation. Turnip and radish ceased to be grown as field crops, kyuma was grown on a smaller scale than it had been, and other potato varieties began to disappear from the region (maps 15a and 15b). With the adoption of the red potato thus began the era of potato-dominated Khumbu agriculture so familiar to foreign visitors since 1950. There were also other less spectacular changes. The area in crops expanded on a small scale in some areas such as Phurtse, Nauje and lower Pangboche where terraces were constructed on previously unfarmed slopes.^[27] Draft plowing was also more adopted for grain farming during this period, especially after the 1950s.

The Introduction of the Red Potato (Riki Moru)

There are several accounts of the introduction of the red potato. These may reflect several different introductions to different parts of Khumbu over the course of two decades. According to one Khumjung account the red potato was originally brought to Pharak from Lazhen (a valley in Sikkim) by Karke Tikpe and Ang Pasang, who planted it in Tsermading

(half a day's walk south of Nauje) more than sixty years ago. A second Khumjung resident recalls that Ou Sungnyu of Kunde first brought the red potato to Khumjung and Kunde close to fifty-five years ago. One Thami Teng resident who was seventy-seven years old in 1985 recalled getting his first red potato seed directly from Pharak when he was about twenty-four years old (ca. 1932). He also remembers hearing that Pharak Sherpas had gotten the variety from Lazhen. This Thamicho man recalls that from his first small pot of seed potatoes he harvested a single basket load and from this small beginning (and a few potatoes brought from Pharak by other Thamicho Sherpas) the new variety spread throughout the Bhote Kosi valley. The red potato may have been introduced to eastern Khumbu slightly later.

When moru first appeared it had a very high yield, comparable in some people's memories to that of the yellow potato when it was first grown in the 1970s. In good years some large landholding families harvested so many red potatoes that they were unable to sell their surplus even at a quarter-rupee per eleven kilograms (versus a 1991 price of fifty rupees/11kg) and had to throw many tubers out even after they had used all they could as livestock fodder. The new crop was susceptible to frost and to blight, but represented a much more dependable food source than had earlier potato varieties.

The red potato did not totally supplant kyuma in the region, which was still cultivated on a small scale in main villages until the 1960s and even more recently in some high-altitude settlements. Many Sherpas maintain that kyuma was abandoned not simply because a higher-yielding variety became available, but also because its yields began to decline. This is widely believed to have been related to the introduction of the red potato. Many people believe that the introduction of a new potato variety into a settlement adversely affects the characteristics of varieties already being grown there. Yields decline because plants feel offended or neglected by the attention being transferred to the new variety. The resulting poor performance may hasten the old variety's abandonment. The same explanation is also given to account for the later decline of red potato yields and the recent perceived decline of yellow potato yields.

The red potato was not the only potato introduced between 1930 and 1975, but it was the only variety that was widely adopted. Some families experimented briefly with linke , but this white, watery variety developed a reputation for causing stomach problems which led to its total rejection in Khumbu. The brown potato (mukpu ), which is regarded so highly in the Bhote Kosi valley, may also have been introduced during this period. According to some poeple it was originally obtained from Pharak Sherpas.

Map 15a.
Agricultural Change: Bhote Kosi Valley Agriculture, circa 1920

Technological Change

Prior to the 1950s plowing in Khumbu was usually done by teams of men rather than by draft animals. It is unclear why this practice persisted as long as it did throughout the region. Draft plowing was the usual prac-

Map 15b.
Agricultural Change: Bhote Kosi Valley Agriculture, 1987

tice in Tibet, and indeed the dimzo raised in Khumbu for sale there were primarily in demand as plow and pack animals. Yak are also used as plow animals in Tibet and today some Sherpas plow behind yak. Yet earlier in the century neither yak nor zopkio seem to have been widely used in this way in Khumbu. It was only during the 1950s that draft

plowing began to supplant plowing by human teams throughout the region. Men still pulled plows in Phurtse until only a few years ago. In Phurtse the increasing adoption of draft traction was the cause for some tension within the community. Men who had made good wages by pulling plows resented the loss of this opportunity. This had been the best paying day labor in the area, for each of the three or four men pulling a plow was paid double to triple the normal wage for agricultural day labor and was given better quality food and beer than normal laborers. Some families who began hiring yak- and zopkio-pulled plow teams to prepare their buckwheat fields are said to have been told by men who formerly had plowed their fields that they could have the livestock do their weeding and harvesting as well, for their families would not work for them again.

The adoption of draft plowing may still be underway in Khumbu today. During the 1980s a few families in Nauje and in Khumjung began having large potato terraces plowed, breaking with the custom of only plowing grain fields. This was judged to be cheaper than hiring agricultural laborers to perform the same work. It is not yet, however, a very widespread practice. Virtually all families continue to rely on either reciprocal labor arrangements or hired labor to dig and prepare their potato fields with hoes. A cultural factor may also be involved, for some people feel that digging potato fields gives more energy to those fields than plowing does, producing better crops. This impression could conceivably be related to a difference in the depth to which the soil is worked. The Khumbu scratch plow probably does not work the soil as deeply as hoe digging does.^[28]

Post-1960 Agricultural Changes

In the past thirty years there have been a number of changes in Khumbu agriculture. These include changes in values that have affected community agricultural management practices, the introduction of additional new potato varieties, the decline of buckwheat production, an increase in the attention given to growing fodder crops, and continuing experimentation with new techniques and crops.

Blight-prevention Practices

During the 1960s the enforcement of regulations aimed at preventing blight declined across much of Khumbu except Phurtse, Pangboche, and Dingboche. In Nauje few of the prohibitions had been enforced with any stringency for some years and even in the 1960s the only bans carefully enforced were the exclusion of livestock from the village after Dumje

and the prohibition on bringing freshly cut wood into the village after the fourth day of the sixth month, Dawa Tukpa, two weeks later. The ban on bringing freshly cut wood into the village during the time of maximum blight danger ceased to be enforced around 1965 and after that the Nauje nawa only enforced the livestock-control measures. During the late 1960s and early 1970s many blight-protection provisions were also abandoned in Thamicho, Khumjung, and Kunde. By the mid-1970s only Phurtse, Pangboche, and Dingboche enforced prohibitions against bringing in freshly cut wood after the fourth of Dawa Tukpa. All settlements, however, continued to enforce the mid-summer bans on grazing near the villages.

The decline in the enforcement of some blight-related rules in several villages appears to go back to well before 1960 and reflects changing beliefs. But the striking decrease in enforcement of restrictions on freshly cut wood in the late 1960s and early 1970s in Nauje, Khumjung, and Kunde may be related to changes in local resource management after 1965. The implementation of new forest regulations by a government office in Nauje led to the abandonment of local enforcement of tree-felling rules in those three villages. Perhaps it was decided that all other forest-related controls had also been made unenforceable by the government's announcement that it would be responsible for forest management.

The Introduction and Diffusion of the Yellow Potato (Riki Seru)

In the middle of the 1970s yet another new potato variety was separately introduced to Khumbu by two men from different parts of the region who encountered it in very different parts of the Himalaya. Both had recognized the variety as something new, a potato of a different color and texture which had a reasonably good taste and produced a high yield of large tubers. Both brought back a few seed potatoes to grow at home, one planting these in Thamicho and one in Khumjung. Following these introductions the tuber became the most important variety throughout the main villages of Khumbu within ten years and diffused from Khumbu to Pharak, Shorung, Salpa, and even the Rai regions of the Hinku and Hongu Khola and the lower Dudh Kosi.

The yellow potato was brought to Khumjung by Pemba Tenzing about 1976.^[29] He encountered the new variety at Singh Gompa north of Kathmandu on the trail to Langtang National Park. While employed by a tourist trekking group he made an overnight stop at the monastery and met an old Tamang friend who was then working in the monastery kitchen. His friend served him some boiled potatoes which Pemba

Tenzing noticed were unusually large in size. He asked if he could have half a load (15-20kg) of them to take back to plant in Khumbu.^[30] The Tamang was unwilling to part with that many, but did give him five potatoes. From the five potatoes that Pemba Tenzing and his wife planted they harvested five tins (ca. 50kg) of the yellow-tinted, slightly watery, large, oval tubers. From these few potatoes, he notes, the yellow potato spread all over Khumbu.

That they did spread throughout Khumbu and far beyond in remarkably little time had a good deal to do with Pemba Tenzing's neighbors, Konchok Chombi and his wife Ang Puli. Konchok Chombi and Ang Puli obtained two yellow potatoes from Pemba Tenzing's wife. She cut eighteen eyes (mik ) from these and planted them, harvesting nearly a tin (11kg) of potatoes. These were planted the next spring and yielded four loads. During the next few years they obtained successively greater amounts of seed potato and before long were able to convert almost their entire Khumjung production to the new variety. Soon they were harvesting more than 200 loads per year.

Konchok Chombi realized the importance of the new variety and spread the word widely about the discovery. Equally as important, he was able to provide seed potatoes for others to buy and introduce into their own fields and villages. By Khumbu standards Konchok Chombi is a large landowner with four large fields at Khumjung which can produce harvests of up to 400 loads of yellow potatoes in the best years, far above his household requirements. Not only could he provide seed potatoes, but in order to meet the widest possible demand and diffuse the new variety as rapidly as possible he limited purchases to a small amount of seed potatoes per customer. He would sell no more than three tins (33kg) to any one family, enough for them to harvest sufficient seed potatoes the following year to plant much of their fields in the variety. Beyond this he even promoted the new potato in other regions, sending girls of seed potatoes to the Sherpa villages of Golila, Gepchua, Mure, and Tingla west of Shorung, to Chaunrikharka in Pharak, and to Sher-pas in the Katanga and Kulung regions. Villagers from Pharak began to seek him out at home for the new tuber. Within a few years the potato became known in much of the region as "Au (uncle, a respectful term of address) Chombi's potato." Pemba Tenzing's discovery of the variety has been nearly totally forgotten.^[31]

Also largely unknown is the fact that there were not one but two introductions of the yellow potato to Khumbu. Outside of Thamicho most Sherpas are totally unaware of the separate introduction of the variety to Pare and Yulajung. About 1973 Dorje Tingda, a Thamicho man who has houses in Pare and Yulajung, discovered a potato in Dar-jeeling which is now considered to be identical to the Khumjung tuber.

He brought back a small quantity of potatoes that he had first noticed while having lunch in a tea shop near Darjeeling while guiding tourists on a local pony trail. It is said that for three or four years he kept knowledge of the new variety to himself because he was worried that if other families also began producing the new, yellow variety it might affect the business he was then doing selling his crop to the Japanese-operated Everest View Hotel near Nauje. According to some Thamicho residents he only began selling villagers potatoes after three local Sherpa panchayat officials asked him to sell them a few. One of these officials, a Yulajung Sherpa, noted that from the single tin of potatoes he obtained he harvested four large loads (more than 160kg) the first year. He had thought that this was an unparalleled harvest until he learned that one of the other officials had harvested five loads from his single tin. After that, he remarked, the new potato spread rapidly through Thamicho.

The pace of the diffusion of the yellow potato through Khumbu can be roughly charted. Thamicho acquired the variety mainly through its introduction to Pare, but it was not until the early 1980s that there was very wide diffusion of the tuber because of the difficulties in acquiring seed potatoes.^[32] By 1981 the potato was becoming well established in Nauje as well as Khumjung and Kunde. In these places it was grown on an approximately equal basis with the red potato by 1985, and by 1987 dominated crop growing. In eastern Khumbu the yellow potato became established slightly later. It was first planted in Phurtse in 1981, but many families there did not begin planting it until 1983 or 1984. It has, however, become the major variety grown there during the past few years. The yellow potato was first planted at Dingboche in 1982 and may have been tried at Pangboche a few years earlier.

The rapidity with which the yellow potato was adopted and at which it supplanted the red potato in the main villages reflects Sherpas' interest in experimentation with new varieties and with adopting high-yielding varieties. In this case the degree of interest in greater harvests outweighed a number of early reservations that many people had about the variety. A considerable number of people initially disliked its taste and many farmers initially refused to grow it. It was also noticed very early on that the variety had a different growing pace and different characteristics than the red potato, and some farmers were concerned that this would diminish the yield of intercropped radish. Some families for whom radish production for use as fodder was important were reluctant to cultivate the new variety as a result. There were also questions about its performance at higher altitudes. It was felt that the yellow potato's longer growing season put it at greater risk than the red potato in the short summer season of the high-altitude settlements. Some people who experimented with it in the Dudh Kosi and Imja Khola valleys

at altitudes above 4,000 meters also found that the tuber seemed to become still more watery when grown at that altitude. This reputation for poor taste at high altitude led a number of families in eastern Khumbu to decide not to grow the variety at sites above the main villages.

Many of these early evaluations, however, were modified within a few years as the process of testing and evaluating the yellow potato continued. Families experimented with the variety at different sites and people noted with interest the crop experiments of the pioneer growers in each locality. Word of good harvests spread quickly. Farmers who said in 1984 that they would never plant the yellow potato were planting it by 1987. The reservations about its taste became less pronounced, although the red potato continued to be proclaimed the finest-tasting potato in Khumbu.^[33] It even began to be grown more widely in high-altitude settlements. By 1987 the variety was being planted by many families at Dingboche and also by several families at Na despite earlier reports that yellow potatoes that had been planted there had had poor taste. By 1987 it had also become the major variety planted at Tarnga. The yellow potato, however, has still not been universally accepted as fit for high-altitude planting. Some people refuse to plant it at Dingboche and in the upper Dudh Kosi and others are reluctant to plant it at Bhote Kosi valley sites higher than Tarnga. The increased wateriness of the tuber at these altitudes is still the factor most often noted as the main factor in this decision.

The Introduction of Other New Potato Varieties

Following the introduction of the yellow potato at least two and possibly three more varieties have been introduced to Khumbu by Sherpas who encountered them elsewhere. The most important across Khumbu is development potato (riki bikasi ), which is said to have been introduced about 1981 from Phaphlu in Shorung by a Nauje man who brought back some from the agricultural extension office there.^[34] Thus far the development potato has been grown primarily in lower Khumbu, especially in Nauje, Khumjung, Kunde, Thami Og and Thami Teng. In the Bhote Kosi valley it has been experimented with as high as Tarnga. Very little was being grown in 1987 in Yulajung (although families there have been experimenting with it for at least two years), but it is being grown fairly widely in Thami Og and Thami Teng. The adoption of the variety was slowed down in Yulajung by concerns over its storage qualities and in eastern Khumbu both by doubts about its hardiness at higher altitudes given its long growing season and by a lack of seed potatoes.^[35] There

have been reservations also about its taste and its high degree of wateriness. Some people suspect it causes stomach problems and others are convinced that it is lower in food energy than other potato varieties. But the yields of development potatoes are usually so outstanding even in comparison to the yellow potato that the new variety continues to be experimented with by many farmers and to be adopted increasingly widely. By the late 1980s there were fewer reservations about both its altitudinal fitness and taste, although many people still consider it far inferior in taste to the red potato, the yellow potato, and kyuma.

It is possible that the brown potato (riki mukpu ) was also introduced in this period, although some people put the date of its introduction earlier. The brown potato is a round, dark-skinned variety of uncertain origins, which some Sherpas maintain was introduced in the late 1970s from Pharak.^[36] It has developed a reputation for yielding well at high altitudes and is much sought after by people who want to experiment with a few plants in their fields. The secondary high-altitude agricultural site of Goma in the upper Bhote Kosi valley is very well known as a center of brown-potato cultivation, but thus far the variety is little grown outside of the upper Bhote Kosi valley.

The latest addition to the Khumbu potato repertoire was introduced by Nauje families (including Sonam Hishi's family) from Pharak in 1987. It has no commonly accepted name. Sonata Hishi's wife Chin Dikki is calling it, with much amusement, long tail (ngamaringbu ) after a hairlike protrusion from the base of the tuber.

The Decline of Buckwheat Cultivation

Until the 1970s substantial amounts of buckwheat were being grown in all Khumbu villages other than Nauje.^[37] In Khumjung, Kunde, Pangboche, and Phurtse buckwheat continues to be grown on a great deal of land and in Phurtse and Pangboche it is planted on nearly 50 percent of the cropland. Formerly considerable buckwheat was grown in the villages of the Bhote Kosi valley as well as in some of the gunsa and at Tarnga. But buckwheat has long been less emphasized there than in the other buckwheat-growing areas. According to elderly residents even in the early decades of the century buckwheat was grown on as little as a quarter of the land in Thami Og. During the past ten years interest in planting buckwheat in the Thamicho villages has plummeted. In 1986 it was being grown in Thami Teng in only three fields and the following year was only cultivated in three small patches. In 1987 only a single, small field was planted in Yulajung and no buckwheat whatsoever was grown in Thami Og. The Bhote Kosi valley has become monocropped with potatoes from the gunsa to the high-altitude herding settlements.

There may be several factors in this increased emphasis on potato cultivation in the Bhote Kosi valley. One may be population pressure and a response to it by intensifying production. As food demand in the Bhote Kosi valley increased during the twentieth century it might be expected that potato production would be further emphasized, for it produces much more food per hectare than any other Khumbu crop and can form the bulk of a household's diet if necessary. For at least half a century families with little land have tended to put their limited land to potatoes and to grow less buckwheat. One of Khumbu's oldest residents, a man of Khumjung, noted that when he was young if a family had a good deal of land it would plant half in kyuma and half in buckwheat and rotate them annually, but that if it was poor it would primarily plant potatoes. A concern with intensification probably also led families in Nauje, where land is in very short supply relative to the size of the population, to emphasize potato monoculture. There grain cultivation was abandoned very early in the century.

Yet there remain some questions about the role of response to population pressure in the conversion of land from buckwheat to potatoes. It is unclear how great recent population growth has been in the Bhote Kosi valley and to what degree fragmentation has affected land ownings per household. It does not seem likely that land shortages there are significantly greater than in eastern Khumbu where buckwheat has not been abandoned. Even if the average size of household land holding has declined valleywide there are still problems with a simple intensification explanation, for not all families are equally land-poor, and there is no doubt that many Thamicho families do own amounts of land which are substantial by Khumbu standards. That even these households have chosen to specialize in potato production suggests that factors other than population growth are involved in the current monoculture of potatoes in the valley.

Commercialization might be a factor, although this is difficult to substantiate given the lack of pertinent household and land data. Thamicho has long exported small amounts of dried potatoes to Tibet and since the early 1970s it has played an increasingly important role in the small-scale regional exchange of potatoes within Khumbu itself. Many Nauje families have depended for generations on purchasing some potatoes to augment their own production. Before the mid-1970s much of the Nauje demand was met by Khumjung production. During the mid-1970s more attention was turned towards the Bhote Kosi valley when Khumjung suffered a series of disastrous harvests. During the last years before the widespread adoption of the yellow potato Bhote Kosi valley surplus potato production became important not only for Nauje families but also for many Khumjung and Kunde households. Demand was so high that

tension broke out between Nauje villagers and those of Khumjung because some Khumjung residents were intercepting Thamicho farmers on their way to the Saturday market at Nauje and buying out their entire supply of potatoes.^[38]

The adoption of the yellow potato eased the shortage of potatoes in Khumbu, but demand for tubers continued to increase as a result of growth in the numbers of Nepali residents in Nauje and dramatic increases in the scale of tourism. Tourists consume large quantities of locally grown potatoes that are an important component of both the food offered by local lodges and that cooked by commercial camping tours. Potatoes are sold to families, lodgekeepers, and trekking groups at the Saturday market at Nauje as well as continuing to be bartered in direct family-to-family exchanges. Some Nauje households, for example, make deals with Thamicho farmers for potatoes months before the harvest, offering cash, tea, kerosene, and other commodities bought at bulk prices at the Nauje market or in Kathmandu.

The possible relationship between greater opportunities to sell potatoes locally and increased Thamicho emphasis on their production requires further study. As of now I cannot evaluate how common it is for Thamicho farmers to sell surplus potatoes or the role that interest in producing a surplus plays in crop decisions. When I attempted to pursue this line of investigation in 1987 I found that Thamicho farmers unanimously denied that an interest in selling potatoes had anything to do with their decisions to discontinue the cultivation of buckwheat. They also did not cite land shortage as a factor. This does not mean that commercial motives and interest in intensification are not factors in land-use decisions, for certainly an interest in higher yields has driven the recent widespread adoption in Thamicho of the yellow potato and the development potato, and some families must make substantial income from the sale of potatoes. But it does suggest that these were not the most immediate reasons for the relatively rapid and extensive recent decline in buckwheat growing.

According to Thamicho farmers the decline of buckwheat in their region was caused by the increase in the number of crossbreeds herded in the valley and the breakdown of local pastoral managment regulation. This has made it increasingly risky to cultivate buckwheat. Yak and zopkio have always been considered threats to buckwheat crops and urang zopkio especially are considered very apt at slipping down valley and getting into fields. The great increase in the 1980s in the number of crossbreeds kept by Thamicho villagers magnified this risk. Thamicho people complain that whereas yak and nak seem content to graze in the high pastures of the upper valleys in late summer that some urang zopkio move down valley during the night and break into buckwheat

fields. By the time the damage is discovered in the morning entire fields can be ruined. This risk is further increased in late August and early September when many Thamicho herders take pack stock to Nauje to meet mountaineering expeditions. According to local regulations they are only allowed to spend a single night in the zones closed to grazing around the main villages as they move their stock through to Nauje. But many herders abuse this custom and keep their stock based in the villages for days. These violations and the nawa's inability to control them have not only seriously threatened buckwheat crops but have also been an important factor in undermining the effectiveness of the entire system of herding regulation in the Bhote Kosi valley during the mid-1980s.

Fodder Crop Production

Another trend during the past two decades has been an increase in fodder-crop and hay production. This has been especially marked in the Bhote Kosi valley and in Nauje. In Nauje more terraces are being planted to fodder crops and a few fields in Thami Og and Khumjung have also been planted in fodder crops recently rather than in food crops. Many new hayfields have also been established, especially in the Bhote Kosi valley. Some were created from pastureland in the upper valley, whereas others in the main villages were converted from cropland. Large numbers of such converted fields can be seen at Yulajung where they represent a major hay-growing resource for settlement families. There are also examples in Thami Teng and Thami Og as well as in Tarnga, Marulung, and some lesser settlements.^[39] The interest in producing more hay and fodder in Thamicho probably reflects both the end of winter herding in Tibet and the recent decline of the nawa-enforced regulations that formerly protected some winter pasture and areas where wild grass was collected for hay from year-round grazing.

In Nauje some potato fields have been converted to production of barley and wheat grown for fodder. These fields are planted late, some as late as the first of August when grain is put into fields from which the earliest potatoes in Khumbu have already been harvested. When planted this late there is no opportunity for the grain to ripen, and the early harvest of potato also diminishes the yield of the potato crop. Barley and wheat stalks, however, can be dried and make fine hay. The second cropping of grain in potato fields is a new phenomenon and is mainly restricted to Nauje where the first barley fields were planted in 1984. It is not widespread even there, and less than 5 percent of the village field area is involved. There are only two examples elsewhere in Khumbu of growing grain for fodder, a single wheat field in Thami Og and a barley field at Khumjung. But in Nauje there is increasing interest in growing barley. In

1984 only four fields were planted in barley, all of them on terraces that were used in the spring and early summer as trekking campsites and hence could not be planted in potatoes. In 1987 twelve fields were planted to barley, six of them as second crops following potatoes. One field was planted to wheat. Since then the trend has continued and both barley and wheat are being grown on a small scale by still more Nauje families. Interest in Nauje in fodder cultivation reflects the increasing scale of livestockkeeping there, the greater fodder requirements of urang zopkio as compared with yak, the impact on local pastoral resources of the post-1979 pattern of year-round grazing in the Nauje area, increasing shortages of wild grass to dry as fodder, and escalating hay prices. Villagers' lack of hay land in the higher valleys is also undoubtedly a factor. Few Nauje families own hayfields, although one family has recently begun to grow hay at Tarnga and others at Samde and Tashilung.

Agricultural Experimentation

Sherpa experimentation with new crops, crop varieties, and agricultural techniques is a continuing process. Women constantly trade seed potatoes and take keen interest in the productivity and other characteristics of these potatoes when they are grown in the microenvironmental conditions of their own fields. A few women and men also take an interest in other forms of agricultural experimentation and try out new cultigens, varieties, and techniques, the results of which their neighbors observe with great interest.

Although potatoes have been the major focus of Sherpa agricultural experimentation in the twentieth century, some Sherpas have also experimented with maize at Jangdingma (a gunsa site below Nauje on the Bhote Kosi), with wheat in Nauje, Thami Og, and Dingboche, with barley at Nauje and Khumjung, and with wheat, buckwheat, peas, and white barley at Dingboche. Families in Nauje have recently adopted a variety of new household garden plants, including cabbage, cauliflower, carrots, and spinach. They obtain seeds for these plants from Kathmandu and from foreign friends. A few families in Thami Og and Pangboche are experimenting with these and other vegetables and one family has unsuccessfully attempted to raise cabbage and cauliflower at Dingboche. There has also been much interest in fruit. Apple trees have been planted Nauje and in the lower Bhote Kosi valley for several decades, although thus far without any success.

There seems to be more interest in experimenting with crops and crop varieties than with new agricultural techniques. Two exceptions are the current efforts at Samde and at Phulungkarpo to irrigate hayfields. The Samde experiment is the first Khumbu use of polyvinylchloride pipe for

delivering irrigation water. There have been no experiments thus far with mechanized tools or pumps, nor with insecticides, herbicides, or fungicides. A couple of farmers have brought small amounts of chemical fertilizer from the agricultural development office in Shorung and tried them out, but the transport involved and the expense of the fertilizer have limited the use of such fertilizers to a few experiments.

Pastoral Change in Traditional Times

Khumbu pastoralism has commonly been portrayed as having recently undergone a major transformation from the traditional yak-herding practices of the 1950s. There has indeed been very significant change in pastoralism since 1960, and the economic, social, and environmental implications of this are only beginning to be understood. Many of these recent changes are related to tourism development and will be discussed in the final chapter of this book. But before taking up the processes and implications of recent change it is necessary to examine more closely the notion that as recently as 1960 Sherpas had longstanding herding practices that were then suddenly transformed by national and international political and economic developments. In the remainder of this chapter I reexamine historical change in Khumbu pastoralism from the early days of Sherpa settlement until the 1960s.

Early Herding in Khumbu

It is very likely that Sherpas arrived in Khumbu already familiar with the herding of yak, sheep, and goats. These animals are raised throughout the Tibetan world including the Kham region. Oral traditions about the Sherpa migration from Kham to Khumbu speak of the immigrants bringing their yak with them to Nepal. There are accounts, for example, of conflicts that these yak precipitated with both the inhabitants of Tibet and the Rais who then lived in the country around the Pi Ke peak in the Golila region. Another oral tradition relates how one of the first Sherpa settlers used to bring sheep manure each spring from his winter gunsa in a cave near Phurte to his Tarnga fields.^[40] The tending of goats is not mentioned in oral traditions about the early decades of Sherpa life in Khumbu, but it may nevertheless also be a very old Khumbu practice. Goats, along with yak and sheep, are considered to be under the special protection of Khumbu Yul Lha. Paintings of the god in temples and the one on the huge boulder above Nauje which figures centrally in Dumje ceremonies also depict a goat, a yak, and a sheep. And at the summer Yerchang ceremonies barley flour representations of all three animals

are ritually offered to Khumbu Yul Lha in asking him to guarantee the welfare of the herds.^[41]

During the early centuries of Khumbu pastoralism the present system of using stone-walled huts as gunsa and high-altitude herding bases had apparently not yet evolved. Several legends mention the use of black tents (ri bu ), a practice associated today with Tibetan pastoralists and some Tibetan-culture Himalayan groups (e.g., the people of Dolpo), but not with Sherpas.^[42] These were still in use in parts of Khumbu until the turn of this century, and in some high summer herding settlements such as Dole, Marulung, and Tarnak huts may have first been built only in the late nineteenth century. Some families in the late nineteenth and early twentieth centuries set up elaborate camps in the highest pastures and rich families erected as many as ten or twelve tents, each of them requiring a yak to transport.^[43]

It is not clear when the first herding huts were built in the high country and at what pace they replaced the use of tents in the different valleys. It could well be that tents were in common use for many generations in the highest pastures and were replaced by huts only where hay growing developed and a need arose for some permanent, roofed structure to serve as a barn. Today herders who want to base in an area where they have no hayfields are often satisfied with using resa for a few weeks and are content with a simple bamboo mat or a tarpaulin thrown over a low structure of rough rock walls. But at places where they grow hay the same herders have more permanent, roofed huts where they can store hay through the winter and which they can use as a base in the winter and early spring when they take their stock up to the often still-snowy high country to feed them hay and fertilize the hayfields. The building of herding huts in the phu may thus very well be closely connected with the development of hay cultivation in various high-altitude herding areas.

The origins of the Khumbu practice of growing hay in stone-walled fields, however, is also unknown. Oral traditions are silent about either early hay cultivation in Khumbu or its later introduction. The growing of hay could be a tradition dating to Kham or it could have been developed as recently as the nineteenth century. Without archaeological evidence there is no way to know. Neither Tibetans from the Tingri region nor the Rais and other peoples of the Dudh Kosi region grow hay in walled fields, and this practice also seems to be rare elsewhere in Nepal.

Some Sherpas speculate that hay growing was probably much less significant to Khumbu pastoralism before the late nineteenth century. Before then, they suggest, the total number of livestock would have been much less in the region and there would have been abundant grazing in both summer and winter without the need to story hay. This is only speculation, for there is no way to be certain that early Khumbu house-

hold herds were not larger than herds have been during the past few generations. But elderly people believe on the basis of what they have heard from parents and grandparents that the number of stock in the second half of the nineteenth century was less than that in the twentieth century, and also note that the practice of taking yak and nak to Tibet for winter grazing meant less stress on Khumbu pastoral resources.

In the high-altitude settlements where hay is grown today there were hayfields as early as the late nineteenth century. Sherpas believe that their development was related to two factors. One was a harsher winter and spring climate with heavier snowfalls. In the late nineteenth and early twentieth centuries there were also very heavy snowfalls in some years, events that Sherpas call kaurnuche . In such times so much snow fell that people climbed out of their houses through upstairs windows, and there were avalanches that destroyed houses and took human and livestock lives.^[44] These snows made grazing impossible for weeks at a time even in lowermost Khumbu and Pharak and made it vital to have hay supplies. The second factor that people believe led to increasing cultivation of hay was a regional increase in the number of livestock. This increased the competition for winter and spring grazing and made it advisable to grow hay as well as to dry and store other types of fodder such as wild grasses.

The Origins of the Nawa System

The nawa system has generally been assumed to be an ancient, traditional Sherpa system. It, like so many other facets of Khumbu land use, could date to premigration practices in Kham. But a Kham origin may be unlikely, for if the practice was that old it might be expected to be typically Sherpa rather than an institution that is unique to Khumbu. Sherpas in other areas have some seasonal regulations regarding the movement of livestock, but with the possible exception of the Rolwaling region none of them has a system of opening and closing a series of zones to different land-use activities or of coordinating these in valleys that are the homes and grazing grounds of people from several different main villages. In Khumbu multivillage zonal systems may be no older than the middle of the nineteenth century. Before then grazing regulation appears to have been the province of individual villages that zealously restricted the use of their village pastures to stock belonging to village residents and which may have had regulations regarding times of the year when grazing was allowed or forbidden in the vicinity of the village and surrounding winter pastures.

Konchok Chombi believes that the distinctive nawa system familiar in twentieth-century Khumbu must have been developed no earlier than

about the middle of the nineteenth century. He thinks that it originated after an intervillage grazing dispute when Kunga Hishi of Thami Og was the gembu of Khumbu. According to an oral tradition Kunga Hishi married a Phurtse woman of the Sherwa clan. She was given a nak by her parents at the time of the wedding and took this nak with her to her new home in Thami Og. The nak, however, soon tried to return to its familiar Dudh Kosi valley grazing grounds. On its way to Phurtse it passed through Kunde-Khumjung village lands and at Zarulungbuk, the ridge between the Bhote Kosi valley and Kunde, it was killed by Khumjung villagers. A major court case ensued. The infuriated gembu demanded compensation for the nak. Khumjung villagers maintained that they had been within their rights, for at that time grazing areas were controlled on the basis of village boundaries rather than by the open range and zone system familiar today.

The dispute worked its way up through a series of Nepali courts. In that era the government court at Olkadunga had not yet been established and the case went first to Those and then to Charikot before reaching Kathmandu itself. Here a decision is said to have been passed down by the king himself. This edict informed the people of Khumbu that the decision should be made by the region's leading local political authority—gembu Kunga Hishi! Gembu Hishi thereupon declared a verdict in his own favor and further ruled that villages no longer had the right to limit the use of their village lands to village residents. Henceforth all of Khumbu was to be open range and the rangelands of Khumbu were to belong to all Sherpas and their stock with the proviso only that all herders were responsible for keeping their charges out of fields and hayfields.

This verdict threatened the very basis of the communal regulation of resources of that era. Villages did not give up the old system quickly, as several subsequent court cases attest. Later in the nineteenth century, for example, Pangboche villagers made a formal appeal to the government to halt Kunde-Khumjung livestock from being herded in their area. This appeal, it is said, was presented to a visiting Nepali official for judgment. He is said to have appeared to favor the Pangboche cause, for he told Pangboche villagers that he had decided in their favor and would present them with documents to verify their rights to restrict grazing on their village lands. He did give them a document, written apparently in a bureaucratic Nepali that no Pangboche villager was able to read. The villagers presented the official with a number of expensive, woven, yak-hair and wool tarpaulins. Indeed, according to one tradition they gave him these before he had ruled in their favor and at his request or demand. They thought the matter over until the dispute flared up again a few years later and they took the case to court. Here, when they confi-

dently presented the document that established their case, they found that it did not win the day, for Kunde-Khumjung villagers had been given an identical document. Neither document upheld the right of villages to exclusive grazing lands. Pangboche people had to acquiesce to herders from other parts of Khumbu making use of their rich pastures and ever since Kunde herders in particular have relied on "Pangboche" land.

There have also been some cases in the twentieth century of attempts to limit village lands to grazing by residents' stock. In the 1940s, for example, Phurtse villagers attempted to keep Nauje stock from being herded on their lands. One wealthy Nauje trader had sent his large herd of zhum to the upper Dudh Kosi Phurtse pastures in the care of a hired Phurtse herder. Phurtse people technically could not ban Nauje stock from their area, but they could decide that henceforth no villager could act as a herder for stock owned by non-Phurtse people. This put an end to the Nauje zhum grazing in that area. But in this century villagers have otherwise had to accept the open range policies established more than a century ago. In recent years many Thamicho, Khumjung, and Kunde residents have been unhappy with increased grazing by Nauje stock on their lands, but they have not had any legal ground to ban Nauje urang zopkio from the Mende, Phurte, Khumjung-Kunde, Langmoche, Gokyo, and Tengbo areas. Bans on outside stock were imposed successfully, however, against non-Sherpa herders. During the early twentieth century a number of Sherpas strongly resented Gurung grazing in Khumbu and there were efforts to stop it and to put pressure on the Khumbu pembu who had authorized it. In one case a Khumjung pembu was pelted with stones by fellow villagers for having allowed Gurungs to bring their flocks into the region. Yet, while for much of the early twentieth century Gurungs as well as Sherpas benefitted from the open range tradition, their grazing access was ultimately ended. Thirty years ago Nauje residents banned Gurung sheep from entering Nauje village lands. Villagers posted a sign on the bridge below the village which forbade the Gurungs to continue with their flocks, and some residents stole into the Gurung camp at night to show them that they were serious about closing Nauje to Gurung grazing by carrying off some of their sheep.

The development of the nawa system may represent an ingenious response to the regional undermining of village control of village lands. Konchok Chombi speculates that the development of the system of zonal regulations was a reply to the impact of Kunga Hishi's ruling. Something had to be done to decrease the risks of crop losses from livestock and the need to preserve winter fodder and grazing areas, and villages were interested in setting their own regulations for their own

areas. The establishment of the nawa system would have enabled communities to establish strict controls on herding and the cutting of wild grass without contradicting the new requirement that access to an area could not be limited solely to local residents. The nawa system affirms that access is open equally to all, but it also makes everyone subject to the same regulations that may restrict certain activities in certain areas for specified periods of time. Villages could thus largely continue to decide which activities they would tolerate on the lands immediately around them. The opening and closing of village areas to grazing had to be coordinated with the operation of zonal restrictions by other communities in the valley, and where several villages shared grazing grounds it might be necessary to jointly administer zones. Crops, hay-lands, and winter pastures, however, could be protected from grazing. If such a sequence of institutional change did take place in the nineteenth century it represented a superb example of an adaptive response to both environmental and social conditions. This is true even if the concept of village officials charged with enforcing community relations such as seasonal exclusions of stock from villages predated Kunga Hishi's decision. The killing of the gembu's wife's nak may have re-suited precisely because it had violated such a village summer ban on stock, perhaps damaging crops in the process. In this case the development of the zonal systems now in use would have been a creative elaboration of new institutions from an older institutional base in response to changing social conditions.

The establishment of the current form of the nawa system certainly predated 1900 and very probably took place at least a generation before that. This did not, however, end Khumbu concerns with the boundaries of village lands and the control of pasture and forest areas within them. Since the late nineteenth century there have been a number of major disputes over boundaries, for control of areas determined which villages decided on which resource-use rules would be enforced there. The shift of a boundary line could in effect change a forest from being a carefully protected, sacred forest to one in which there were no prohibitions of any kind on tree felling, or a pasture area from one closed to grazing all summer to one open year round. A late-nineteenth-century dispute between Khumjung and Phurtse over the Mong area resulted in the end of strict protection of the nearby sacred forest. And during the twentieth century Nauje villagers have been involved in several long-standing disputes with Khumjung-Kunde and Thamicho villagers over village land boundaries that would make an enormous difference in how some pasturelands are regulated. Nauje villagers have been trying to gain control over the Shyangboche area from Khumjung and Kunde for over half a century so that they could remove the summer grazing restrictions there.

They have similarly tried to extend their control north in the Bhote Kosi valley from the present Nauje village boundary south of the gunsa of Phurte. In the Shyangboche case Nauje villagers have been arguing for at least two generations that the traditional boundary between their village and Khumjung-Kunde is unfairly located just above Nauje. This, they say, reflects a decision reached a century-and-a-half ago when Nauje was a very small place and not the second largest village in Khumbu. They would like to move the boundary about a kilometer from the outskirts of the settlement to the watershed between their village and Khumjung and Kunde. Both pasture access and forest control were issues here, for the disputed Shyangboche area was within the boundaries of the Khumjung-Kunde-administered rani ban. This dispute has on several occasions very nearly sparked violent conflict and once, a few decades ago, almost led to the collapse of both the Nauje and the Khumjung-Kunde nawa systems before villagers were persuaded by a Nepali official to make peace with one another and resume nawa regulation. But though the boundary remains, the Khumjung-Kunde ban on summer grazing in this area is often ignored by Nauje herders and their stock. This continues to be a cause of great tension between the villages and threatens to ultimately undermine the continued operation of the Khumjung-Kunde nawa system.

Yak and Nak Herding in the Early Twentieth Century

In the early decades of this century Sherpas kept herds of a size that has never been equaled since. A number of families had herds of more than sixty head of yak and nak and a few had herds of eighty head—more than twice the largest Khumbu herds of the past forty years. These large, early-century herds were owned by men who were interested in breeding and selling crossbreeds and thus were primarily composed of nak. But some families also kept relatively large numbers of yak to use as pack stock on trading expeditions to Tibet. Many of these large herds of nak and yak were tended by hired professional herders. Usually these men were recent Tibetan immigrants, some of whom also kept sizeable herds of their own. There were so many head of large stock in Khumbu in that era that the winter range and wild fodder resources of the region were inadequate to support them. Rather than go to the trouble to produce vast amounts of hay to winter their herds, many big stock-owners sent their herds to Tibet each winter to graze on the vast grasslands beyond the Himalaya.

The great herds are well remembered today in Khumbu by people old enough to have seen them or who heard about them from their parents

and grandparents. One of the greatest herds was kept by the Mendoa family of Khumjung, who at one time owned more than eighty nak and yak. People say that in the autumn when Mendoa sent livestock north to Tibet he usually sent more than sixty nak and yak, and that he had five yakherders (yakpa ) to drive them. His fellow villager, U Kunggu, sometimes sent sixty-one yak and nak north to Tibet and he liked to boast that he had one more head of stock there than Mendoa. Sundokpa, Pemba Kitar, Ang Chumbi, Thaktoa, and Yulha Tarkia all also had large numbers of yak early in the century. Some Thamicho families also kept big herds. Men in their seventies remember that early in their herding careers and in their fathers' time there were so many nak and yak in the Bhote Kosi valley that it was difficult to feed all the livestock. Here too many people sent their stock to Tibet for the winter with professional herders. Samshing Kitar and Gardza were especially famous for their large herds. People today still talk of the long line of stock that Samshing Kitar sent north each year to Tibet.

Nauje families also owned large nak herds in those days. Today the village has a reputation as the Khumbu village least involved in nak herding, but early in the century some of the greatest herds of all were kept by Nauje families. The greatest of the Nauje herds belonged to gembu Tsepal. He owned more than eighty head of stock, including yak, nak, dimzo, and zhum, and kept two lang (Tibetan bulls) for breeding purposes. These animals were cared for by several hired specialists. A Pinjo herded the nak and Yakpa Tundu the yak (as well as herding fifteen or twenty head of his own yak). A woman named A Droma was responsible for the milking. At that time Nauje families made much use of the rich grazing on Pangboche lands in the upper Imja Khola valley. Tsepal had a house and four fields at Dingboche and herding huts at Pheriche, Bibre and Chukkung. His herd was so large that when it was driven from Pheriche to Dingboche the line of stock stretched over a kilometer from the moraine above Pheriche all the way to the eastern edge of Dingboche. People used to come out of their houses to see the spectacle and excited children would shout "the gembu is coming, the gembu is coming." Other Nauje families also kept large herds. Ang Dorje kept thirty to forty nak. He herded them in the summer at Tsolo (near Dzongla) and at Lobuche. Other village families took their nak and yak to the upper Imja Khola valley. Oungu (Ang Dawa's father) had ten to fifteen nak and herding huts at Chukkung, Dzongla, and Bibre as well as fields at Dingboche. Nauje Urken's parents and Ang Gelgen's grandfather also had nak-herding bases in Chukkung. Another Nauje resident, Guru Nima, who immigrated to Khumbu from the famous yak-raising Chang region of Tibet, also had a large number of yak and nak. Unlike other Nauje villagers he kept his stock most of the year in Tibet

in the Rongshar valley. The nak that comprised the greater part of the herd were based there year round and cared for by hired herders who in exchange kept a half-share of the milk and calves. Even the yak, which numbered more than thirty, were mainly kept in Tibet and were only brought to Khumbu when Guru Nima needed them as pack animals on his trading trips.

Until the 1930s it was typical for the really large herds of yak and nak to be taken to Tibet each winter. The yak and nak of smaller-scale herders also often went with them, for the professional herders who worked for the big stockowners would take other stock for a fee. Yakpa Kasare, who worked for the grandfather of the present head of the Khumjung Mendoa family, for example, not only took the large Mendoa herd north but also the stock of other herders. He charged half a rupee per head for herding yak and nak throughout the winter, a fee then equivalent to about a day's pay for unskilled labor, and his service was popular with families who wanted to devote these months to trading trips into southern Nepal and southern India rather than herding on the Tibetan plateau. On his annual herding trip north Kasare took hundreds of head of nak and yak belonging to families from Thamicho, Kunde, Khumjung, and Nauje. Kasare set out for Tibet each year in late October or November (Dawa Chuwa, the tenth month) and returned five or six months later in March (Dawa Sumba, the third month). In those days there were no grazing fees in Tibet, since the Tibetans did not resent sharing the grasslands as long as the Sherpa herds were not kept longer than a week in one place. Different Khumbu herders used different pasture regions. Kasare preferred to use Chakpakok, Surcho, and sometimes Lungar, whereas other Khumbu herds focused on Shalung and other nearby regions. For three months Kasare kept the stock on the move through the vast Tibetan grasslands north of the Nangpa La, shifting his base every two or three days. Then in mid-winter he turned and retraced his route back towards the Nangpa La. In early spring he would send a message back to Khumbu alerting the stockowners that it was time for them to reclaim their animals, and they would meet him in Tibet. Many of them then put their yak and nak to good use, buying Tibetan salt in Ganggar and Kaprak and hauling this with the pack stock back to Khumbu.

The big herds had mostly been developed as commercial enterprises, especially for breeding of crossbreed calves. The male crossbreeds, dimzo, were in great demand in Tibet as pack stock and there was a market for the female crossbreeds, dim zhum, in Shorung where they were esteemed as milch stock. Khumbu breeders had no trouble selling their calves to the Shorung livestock dealers who annually came to Khumbu on buying trips. Some Sherpas had long-term business arrange-

ments with these dealers and made advance contracts for the delivery of a particular number of calves. The Shorung dealers then took both the dim zhum and dimzo south to Shorung where they were grazed for one or several years on the good grass of that region before being sold again. The dim zhum remained in Shorung and nearby areas, but the dimzo were mostly sold back to Khumbu men who then traded them in Tibet. Business was good enough that Khumbu herders could focus exclusively on breeding crossbreeds rather than also trying to raise nak to keep up their herd size. Nak were readily and cheaply available from Tibet and could be obtained in Khumbu from both Sherpa and Tibetan traders if one was not interested in taking the trouble to scour Tibet for good stock.

The Decline of Large-scale Nak and Yak Herding

During the first decades of the twentieth century some Khumbu stock-owners thus had herds of as many as eighty yak and nak, while several others had herds of more than fifty head and a number had herds of more than thirty. The big herds were broken up, however, by the 1930s and since then the ownership of even thirty head of stock has been very rare. A number of different factors could have precipitated this change. The trade in crossbreeds, for example, could have become less lucrative. There was excellent money in raising crossbreeds when the complex trading circuit that it was a part of was operating smoothly and the ultimate prices paid for dim zhum in Shorung and dimzo in Tibet were good enough. But a change in any one of a number of elements could have made it much less profitable. A rise in the price or a decline in the availability of Tibetan nak would have a long-term impact on Khumbu nak herding. A more immediate impact on the profitability of keeping large nak herds could have come from such diverse changes as a decline in the price offered for crossbreed calves, a decline in access to Tibetan pastures or the imposition there of grazing fees, difficulty in recruiting hired herders, herd taxes in Nepal, trade fees, new Nepalese regulations against exporting crossbreeds or importing nak, Tibetan regulations against nak sales or crossbreed imports, or closure of the border to trade. A number of these factors were involved in a further decline of nak herding in the 1960s, but none of them seem to have been the cause of the breakup of the big herds of the early century. Sherpas remained free to use Tibetan pastures for herding until the 1970s. Tibetan immigrants who were interested in herding for low wages continued to arrive in Khumbu during subsequent decades. Nak continued to be imported from Tibet on a substantial scale until new Chinese regulations were

implemented in the 1960s and trade in crossbreeds in Tibet remained lucrative until then as well. Nak herding remained important in Khumbu and owning them still conveyed great status. But the great herds nonetheless disappeared.

It may be that the business of breeding crossbreeds gradually became less attractive than other forms of commercial enterprises. By the 1930s and 1940s wealthy Sherpas were heavily involved in the trade of butter, paper, and other goods to Tibet and the return trade of salt, wool, and Tibetan luxury goods to Nepal and were moving beyond trade merely with Ganggar to dealing also in Shigatse, Lhasa, and as far afield as Calcutta. The owners of the great herds may have gradually been attracted by the opportunity to make greater profits by shifting their capital away from pastoralism and into trade. There may have been a generational factor at work as well, for it may be that the sons of these larger herders were less inclined to perservere in large-scale pastoralism and more inclined to trade. When sons inherited their share of the family herd some may have been content with herding this smaller number of stock on their own without hired herders. Others may have sold off much of this inheritance and kept only a few pack or milk animals. By the 1940s those lineages that continued to raise livestock raised much smaller herds and the wealthiest households among them were not those that had the largest herds but those that most successfully devoted their energies and capital to trade. Nak herding was left to families who were interested in it as a lifestyle and as a source of minor income from calf sales. Families who kept herds of more than twenty or twenty-five head of nak were still regarded with great respect and some envy by many Sherpas, but some of the truly wealthy families now kept no nak at all.

There was also a decline before 1950 in the keeping of yak as pack stock. A trend away from yak keeping can be traced back quite far in parts of Khumbu. In the Bhote Kosi valley elderly herders recall that they heard from their parents and grandparents that there had been a shift in their lifetimes away from the keeping of yak and towards a greater emphasis on nak. Au Puta of Thami Og, for example, a man in his seventies, commented that in his grandfather's time many more yak were kept and that these were useful as pack animals in the Tibet trade. In his father's time they cut back on yak to emphasize nak for crossbreeding. This same process appears to have taken place in Khumjung and in Nauje as well, although in Nauje some traders such as Guru Nima kept large numbers of yak for pack stock well into the 1930s. A decline in the scale of the bulk trade in iron and salt during this period may have been one factor in the declining emphasis on yak.

The Nepal government also played a role in the decline of Khumbu yak numbers early in the century. In July 1903 a British military force

under the command of Francis Younghusband set out from Sikkim for Tibet. He was charged with discussing trade terms and international political issues with the Dalai Lama and ultimately, after several military encounters with Tibetan troops, reached Lhasa. The government of Nepal had pledged to support the Younghusband expedition and one of the ways in which it did so was by providing large numbers of pack stock. Khumbu was ordered to contribute yak. Large numbers of yak, for which the Nepal government is said to have only paid a pittance in compensation, were dispatched for Kalimpong. They were sent under the care of professional Khumbu yak herders, but the low-altitude route through the Nepal midlands proved to be fatal to most of the stock. Almost none reached Kalimpong alive.

Another important factor in the declining interest in keeping large numbers of yak during the early twentieth century was the availability of Tibetan pack stock. Until 1959 Tibetan traders came to Nauje each year with large numbers of pack yak to haul salt. On the return trip to Tibet there were insufficient loads for this many animals and the Tibetans accordingly offered to transport loads for Khumbu traders. This worked out very well for many Khumbu traders whose trade revolved around taking bulk goods to Tibet and importing higher-value commodities back to Nepal. Khumbu traders could also hire Thamicho yak for transport and there were also always Khumbu Sherpas willing to work as porters on the trade route for low wages, at least until the 1950s when higher-paying work for mountaineering expeditions became available in Khumbu. These opportunities for arranging hired transport for goods made it unnecessary to bother with the labor and expense of maintaining one's own pack stock. During the 1940s and 1950s Nauje traders relied entirely on hiring either pack stock or porters.

The breaking up of the large nak and yak herds may not have meant that the total number of head of large stock in Khumbu declined, for the stock may simply have been divided among heirs or purchased by other smaller-scale Khumbu herders. Regional grazing pressure nonetheless increased, for with the end of the large herds the practice of winter herding in Tibet declined dramatically. It was no longer easy for any herder to send even a few head of stock north by making arrangements with the hired herders of the rich stockowners, for there were no more big herds and no more professional hired nak and yak herders. Only Thamicho herders still took their stock to Tibet to winter. Herders throughout the rest of Khumbu began keeping their livestock year round in Khumbu instead. This meant that each head of stock now required more grazing than before. The number of stock that could be supported by regional range resources would have been lowered, and there would especially have been increased risk that the carrying capacity of the

winter and spring pastures of lower Khumbu would be exceeded unless much greater amounts of hay and other fodder were stored for this time of the year. In the absence of a regional decline in stock numbers herders had no option other than to devote more effort to hay cultivation and run greater risks that the survival of their stock would be endangered by overgrazing, a poor haying season, or a snowy winter in which their fodder supplies might become exhausted too early.

The Declining Importance of Sheep in Khumbu

Although the keeping of sheep and goats goes far back in Khumbu Sherpa traditions, in the twentieth century these animals have primarily been kept by poorer Sherpa households. For such families both sheep and goats offered a critical source of manure and small amounts of wool or hair and meat. They were also sources of income. During the past twenty-five years the number of sheep in the region has probably never exceeded 1,000 and today is not much over 500. There have been even fewer goats, and at the time they were banned in 1983 there were only about 300 of them in all of Khumbu, many of them kept by non-Sherpa, blacksmith-caste families. Yet historically sheep were much more important in regional pastoralism. Some Sherpas kept sheep in larger numbers than today and large numbers of sheep were brought into Khumbu each summer from lower-altitude regions.

Today a flock of twenty sheep is large. In Pangboche in the 1930s and 1940s, however, there were three or four families who kept quite large flocks, one of which had more than eighty sheep and goats. In the era before 1950 there were also several Nauje families that kept large flocks, some of them as large as forty head. A few Nauje families used to use sheep as pack animals in the way that is also common in Tibet and in many Bhotia-inhabited, high Himalayan areas. No sheep are used for this purpose today, but as recently as the early 1960s a Nauje family used sheep to carry rice back to Khumbu from places as far afield as Namdu and Kabre, villages more than a week's walk toward Kathmandu.

The decline of Sherpa sheep and goat raising appears to have been directly associated with increasing affluence. Some Nauje families who kept sheep and goats during the 1950s and 1960s abandoned the practice in favor of cattle keeping once they could afford dairy animals. A similar process may well have taken place in other settlements. Sheep are currently kept only by very poor families. Some of the largest flocks, which at present number only as many as thirty head, are kept by non-Sherpa, Nauje blacksmith families who took up shepherding after the recent local government ban on goat keeping.

The relative lack of recent Sherpa emphasis on sheepherding might be thought to reflect a local perception that Khumbu is not well suited for sheep. Several observers have indeed concluded that Khumbu is poor sheep country (Brower 1987:171; Fürer-Haimendorf 1979:13). Brower (ibid.) has suggested that this is related to precipitation, that Khumbu pastures are less well suited for sheep than the rain-shadow Tibetan pastures. Yet Sherpas, Gurungs, and Rais all raise sheep in wetter areas and at equally high altitudes in both the Annapuma range and the upper Arun region. Along most of the span of the Himalaya pastures such as Khumbu's are highly valued by middle-altitude shepherds who see in them a promise of rich, high-country summer grazing and escape from the heavy rains, treacherous ground, disease, and leeches of the monsoon season in the lower mountains. Areas as high in altitude are regular summer herding destinations for Gurung and Rai shepherds in both the Annapurna-Lamjung ranges and in eastern Nepal. In the Modi Khola valley Gurung shepherds have made a two-week spring migration for generations up through a heavily forested gorge in order to reach 4,000-6,000-meter pastures far less extensive than Khumbu's. Arun region Rai shepherds similarly use the high Kempalung pastures, and those in the Hongu Khola, a Dudh Kosi tributary southeast of Khumbu, take flocks of more than 300 sheep each as far as Pharak as well as to the Naulekh area near Mera. Gurungs from the lower Dudh Kosi valley and adjacent areas today send sheep to Pharak, the Mera region, and northern Shorung, all of which are wetter in the summer than Khumbu. Their sheep in recent years have grazed on the pastures on the shoulder of Kwangde and even today graze high on Tamserku in view of Khumbu.

Oral traditions suggest that Khumbu may indeed have been formerly used as summer pasture by Rais and was certainly used by Gurung shepherds from at least the late nineteenth century until less than thirty years ago. The oral traditions about early Rai herding are few and sparse in detail. There is little to go on other than the idea that the Rais once came to Khumbu in the summer and that certain apparently long-abandoned ruins in the upper Dudh Kosi valley are said to be early Rai herding huts. Some Sherpas speculate that Rai summer herding could have been established before the Sherpas arrived in Khumbu and that it might have continued during the following centuries. In some other nearby areas such as the upper Hinku Khola valley and the Arun-Barun river areas Sherpas today share summer herding grounds with non-Sherpas, and in the Pharak, Kulung, Salpa, and Arun regions they have made arrangements to herd on Rai lands.^[45]

Although the legends of early Rai pastoralism in Khumbu may be few and vague, the same cannot be said for nineteenth- and twentieth-century Gurung transhumance to Khumbu. Sherpas alive today remem-

ber the Gurung herds in Khumbu and have heard that in their parents' and grandparents' time still-larger Gurung herds grazed even more extensive areas of Khumbu. Some point to the ruins of old high-altitude herding huts or corrals in the upper Dudh Kosi valley, sites high on the slopes above the highest yak-herding huts as possibly old Gurung or Rai sheepherding bases. Others note that many of the placenames in the upper Imja Khola valley and the valley of the Lobuche Khola seem to be associated with Gurung sheepherding in that region, Tugla, Dusa, and Ralha among them. Dusa, for example, is said to be from dumsa and to refer to a place where rams and ewes are allowed to graze together to breed. A number of elderly Sherpas who have herding bases in the nearby region remember hearing stories of previous Gurung use of that area for just that purpose, the main herds of ewes being kept between Pheriche and Dingboche in the Chajung area and the rams being taken up to Tugla-Palung until late summer when the herds would meet at Dusa at mating season. This seems to have been a nineteenth- and perhaps very early twentieth-century practice, for during the lifetimes of eighty-year-olds the Gurungs have not summered at Tugla or Dusa, but instead have been restricted to the area south of the Imja Khola.

In the twentieth century Gurungs primarily herded sheep in the Imja Khola valley, although there was some use also of pasture in the Dudh Kosi valley and in the Bhote Kosi valley. Phurtse people recall Gurung sheep moving from Pangboche above Phurtse and across to Dole on several occasions forty or fifty years ago, and have heard that more than a hundred years ago sheep used to be taken up the east side of the Dudh Kosi into the high country.^[46] In the Bhote Kosi valley Gurungs herded sheep for several years in the Bhotego area on the west side of the river in the 1940s and long after that took sheep up onto the southern shoulder of Kwangde.

The main center of Gurung grazing during this century, however, was the south side of the Imja Khola valley, particularly the Mingbo and Ralha areas (map 16). Each spring several Gurung herders came up the Dudh Kosi valley and into Khumbu, each with as many as 200 or 300 sheep. In the late 1940s and early 1950s as many as 1,000 to 1,500 Gurung sheep reportly grazed in the Mingbo and Ralha areas each summer. Their migration path did not take them through the main Sherpa villages. Instead they traversed high above the Dudh Kosi to near the confluence of the Dudh Kosi and the Imja Khola and then continued on to the southern Imja Khola valley pastures.^[47] Sherpas gained little from the Gurungs. They did not demand any grazing tax, although it is possible that Gurungs may have made gifts to Sherpa pembu in order to obtain their blessings on grazing in the region. Local villagers were not interested in purchasing wool from the Gurungs, for

Map 16.
Gurung Twentieth-Century Herding in Khumbu

Sherpas consider the wool from the Gurung sheep to be far inferior to that of the Tibetan breeds of sheep that Sherpas raise. Sherpas imported wool from Tibet rather than from the Gurungs. The only real gain most Sherpas realized from sharing their pasture areas with the Gurungs was manure. In the spring the sheep dung was valued for manuring hayfields in the high-herding settlements and Sherpas often asked Gurung shepherds to pen their sheep at night on their hayfields. In return they fed both the shepherds and their dogs.

Sherpas resisted several Gurung attempts to enlarge their grazing areas. It is remembered that Dimal Gurung, a powerful and rich Gurung sheep owner, brought sheep up across from Orsho in the upper Imja Khola for three years to the great resentment of local Sherpas and that once he tried to bring his sheep into Dingboche. Some Sherpas had apparently invited him to bring his flock to Dingboche so that they could benefit from the manure, but others were strongly opposed to the idea. A community meeting was held and as the level of hostility rose Dimal is said to have fled to avoid a beating.^[48] Dimal subsequently attempted to purchase grazing rights in the upper Imja Khola valley from the government. He is said to have offered a fortune—five loads of coins, a thousand coins per load—for these rights, but was unsuccessful. In the Bhote Kosi valley

Gurungs attempted to move up into the Konyak area, but were turned back by Sherpa villagers. In about 1960 Nauje Sherpas also closed their region to Gurung grazing and posted a notice on the bridge into Khumbu below their village advising that Gurung sheep were banned.

Gurung sheepherding in Khumbu ended about 1960.^[49] A number of Nauje Sherpas raided the Gurung camp near the confluence of the Dudh Kosi and Bhote Kosi rivers and carried off their sheep. The Gurungs called in the police from the post at Nauje, who searched village houses, arrested four Sherpas, and returned the sheep. The Gurung shepherds then turned back to Pharak and no Gurung flocks have come to Khumbu since.

Although Gurung sheep are no longer taken to Khumbu they do still make a summer migration as far as Pharak and the Naulekh and Mera areas on the nearby upper Hinku Khola (map 17). Sheep from Jubu and Duwe (south of Aislalukarka) are brought up to Naulekh and Mera each summer in large numbers. Each shepherd is in charge of several hundred sheep and the total number involved exceeds 1,000, more than double the number of Rai sheep in the area. These pastures are also used by Sherpa herds. Both Gurungs and Sherpas must pay grazing taxes to the Rais. Gurungs are charged sixty rupees for the right to establish a herding base in one of the high pastures and an additional seven rupees per sheep for three months of summer grazing. Gurung sheep from the Rumjatar area west of the Dudh Kosi are taken up into Sherpa-inhabited northern Shorung to areas north and east of Junbesi.

Pastoral Change in the 1960s

There have been several major changes in Khumbu pastoralism during the past thirty years. One of these was a further regional decline in the importance of nak herding that began in the early 1960s and still continues today. A second major trend has been an unprecedented increase since the mid-1970s in the keeping of crossbreeds, and particularly in the ownership of male urang zopkio. This is directly linked to the increasing involvement of Sherpas in the tourism industry which will be discussed further in chapters 9 and 10. A third change that has already been mentioned was the end of goat herding in the region in 1983.

The decline in nak and the increased emphasis on crossbreeds show up very clearly in counts of regional livestock taken in 1957, 1971, 1978 (table 24), and 1984 (table 10). It is clear from these statistics that there was a pronounced decline in the number of nak herded in the region between 1957 and 1971 and that this was particularly important outside of the Bhote Kosi valley. In Thamicho there was also a decline in nak keeping, but this was much less sizeable and occurred primarily after

Map 17
Gurung and Rai Sheep, Summer Migration

1971. Despite this regional variation, the Khumbu-wide trend was a massive loss of nak with the total number of stock in 1984 (1,121) barely half that of 1957 (2,061). The second clear pattern is that after 1971 the increase in crossbreeds virtually balances the continuing decline in nak.

The dramatic decline in nak herding in the 1960s was the result of Sherpa responses to changing conditions in Tibet and in particular to changes in Tibetan economy and trade policies after Chinese administration of the area north of Khumbu was established in 1959. During the next few years the Chinese introduced new regulations that seriously undermined the international trade in dimzo on which Khumbu nak

herding was largely based. On the one hand the Chinese hampered the export of Tibetan nak to Nepal and this affected Sherpa herders' ability to restock and build their herds, but the most serious impact on Khumbu herding was a decline in the profitability of dimzo trading. Khumbu traders continued to take dimzo to Tibet for many years, but it became much less lucrative a venture than it had been. By the 1970s the Chinese had developed new policies that required that dimzo could be sold only to a government office established in Ganggar, and at controlled prices lower than what Sherpas had been accustomed to receiving. They had also begun to encourage Tibetan herders to breed dimzo themselves despite their religious scruples against such breeding practices. With the decline in the demand and price for dimzo in Tibet the Shorung demand for Khumbu crossbreed dimzo calves evaporated, although there continued to be interest in zhum calves. But an important source of income from keeping large nak herds had been lost and it was probably this above all other factors that led many herding families to abandon nak herding altogether or scale down the size of their herds.

While nak keeping was in general decline during the 1960s there was also a minor counter trend for a few years in some places, especially Nauje. Again the economic and political changes in Tibet after 1959 were responsible. With the arrival of large numbers of refugees in Khumbu in 1959-1960 it suddenly became possible to build up a herd of yak and nak very inexpensively. Many refugees had brought their herds with them and selling stock was one of the few sources of support they had in Nepal. Many were also eager to sell because the great increase in livestock quickly exhausted grass in many areas of Khumbu and herders tried to sell stock before the animals starved. Sherpas talk of how in those days the forest was thickly littered with rotting sheep carcasses. One Sherpa witness thought that 99 percent of the Tibetan's stock ultimately starved to death, with the sheep dying first, then the goats, and finally the nak and yak.

Yak and nak were cheaper than they had ever been.^[50] A number of families built up herds at this time, especially by purchasing yak. Yak remained useful as pack animals as well as conferring prestige. Some Nauje families that had never been able to afford yak before now built up small herds, turning them loose to graze in summer in the nearby Gyajo valley and keeping them in the winter in the Nauje area. During the middle 1960s these families kept a total of more than 100 head of yak. Within a few years, however, most of these families decided that yak herding was more trouble than it was worth and sold off their stock.

While grazing became difficult in Khumbu due to the influx of Tibetan stock it became better in the border areas of Tibet. There were several Thamicho families who had continued to go north each winter with their herds and who were able to benefit from the now richer Tibetan pastures. These families did not rely on professional herders, but instead either cared for their own stock or worked out a cooperative arrangement with other local families. During the 1960s several Tibetan areas including Shalung, Jalung, and Melungjang continued to be popular Thamicho winter herding grounds. It now became necessary, however, to pay a small fee to Tibetan villagers for grazing privileges. Eight or nine families continued to go annually to Tibet despite the new fees. These families also took the livestock of their relatives and even that of a few non-Thamicho people. Maila, a Nauje man, for example, used to send his many yak to Tibet with Thamicho herders. In the late 1970s, after many Tibetans had returned to Tibet and livestock numbers there increased again, the grass was no longer so exceptional north of the border. Grazing fees climbed and with administrative changes it became difficult to determine the right people to whom to pay the fees. In some areas where good grazing was in short supply Tibetans were reluctant to allow the Sherpas to herd. Thamicho herders grew discouraged by these

conditions and by 1980 the old Khumbu tradition of herding on Tibetan winter pastures was a thing of the past. As herders turned to grazing year round in Khumbu they found that they had to put in much larger stocks of fodder to winter their yak and nak. Some of them had not bothered with cultivating much hay before and had put so little care into their hayfields that they had not even bothered to manure them. In Thamicho today herders pay as much attention to such details as they do everywhere else in Khumbu.

There have been other changes in the past twenty years in Khumbu agriculture and pastoralism. Tourism especially has been a new catalyst of change. I discuss tourism and its impact on land use in chapters 9 and 10. Before turning to tourism, however, there is another dimension of historical Khumbu land use in which the relative roles of tradition and change deserve more careful attention. The next two chapters look in detail at historical change in Khumbu resource management and the role of Sherpa subsistence practices in environmental change.

The village of Nauje (3400m) in the Bhote Kosi valley, and Kwangde peak.

Sonam Hishi of Nauje.

Konchok Chombi of Khumjung.

Pangboche village (3985m) and Tawache peak. Sacred juniper trees flank the temple in the center of the settlement.

Phurtse village (3840m), high above the confluence of the Dudh Kosi and Imja Khola. The forest which extends from the left side of

Phurtse in mid-summer. The dark-hued fields are potatoes and the lighter-hued ones are buckwheat.

Potato fields in the village of Kunde (3840m).

Potato varieties. Upper row, from left to right: brown (mukpu), red (moru), and
yellow (seru) potatoes. Lower row, left to right: kyuma (from the Salpa region,
southeast of Khumbu), "development" (bikasi), and "English" (Belati) potatoes.

Buckwheat.

Dingboche (4358m), in the upper Imja Khola valley. Lhotse is at the left, Imja Tse
or Island Peak in the valley center, and Cho Polu to the right.

Yak in the upper Imja Khola valley (4700m) in spring.

Pheriche (4272m), a high-altitude herding settlement in the upper Lobuche Khola
valley. On the far side of the river are the herding settlements of Tsamdrang (left)
and Naongma.

The forest at Yarin across the Imja Khola from Pangboche. No tree could be felled
which could be seen from the Pangboche temple at the lower right.

The lama's forest at Phurtse. In the foreground is the bridge forest where tree
felling has increased since the early 1960s.

Looking west across southern Khumbu from near the Tengboche monastery. The lower Dudh Kosi valley is in the foreground,
and the Bhote Kosi valley lies at the foot of the skyline ridge. Khumjung village is in the center of the photograph, and the
gunsa settlement of Tashinga is at the lower right. Note the contrast in forest cover along the ridge crest above the Dudh Kosi
valley which forms the southern border of the Khumjung-Kunde rani ban.

Abandoned terraces near Nauje at Nyershe (3400m) in the lower Bhote Kosi valley.

Fir forest (Abies spectabalis) at Namdakserwa (3400m) near Nauje. This area
was formerly part of the Nauje rani ban. A major woodcutters' trail can be seen
descending the slope, which before the mid-1960s is said to have been densely
forested.

The Nauje lodge district. The first Sherpa lodge in Khumbu is the small wood-roofed structure just to the left of the new-style lodge in
the center of the photograph.

The national park-built lodge at Lobuche (4928m) and its woodpile of high-altitude
shrub juniper.

Matted shrub juniper near Tugla at 4600m. Juniper has been removed from the area
in the center of the photograph and is being dried at the left for use as fuel wood.

Urang zopkio loaded for trekking, Nauje.

Cattle-trail terracettes in the lower Dudh Kosi valley near Nauje.

7
Subsistence, Adaptation, and Environmental Change

One of the achievements of cultural ecology has been the documentation of the adaptiveness, ingenuity, and creativity with which indigenous peoples have developed ways of life that are based on profound knowledge of the local environment and ecosystemic relationships and processes.^[1] This perspective on indigenous peoples contrasts sharply with older views that denigrated the land management of "traditional" peoples in many parts of the world as unscientific, superstitious, ignorant, and even destructive. At the same time cultural ecology research has cautioned against a counter tendency to overromanticize indigenous ways of life and resource management. There was a period in cultural ecology research during the 1960s and early 1970s when much fieldwork and analysis reflected possibly premature assumptions about indigenous peoples' harmony with nature or, put another way, their development of a way of life that established a dynamic homeostasis between traditional land use and environment (Netting 1984; Rappaport 1968). Anthropological work especially tended to deemphasize human impact on environment and overemphasize stability and homeostasis, partly due to a lack of concern with historical change (Moran 1984:15-16). In recent decades, however, there has been increasing recognition that many indigenous peoples had profound impacts on the ecosystems of their homelands and that this took place in pre-capitalist, so-called traditional societies as well as in those whose former ways of life and social organization had been transformed by colonialism, frontier dynamics, increasing integration into national and global economies, and other processes. Cultural ecology has largely moved away

from assumptions about the harmony or homeostasis of indigenous peoples with their physical habitats towards a more analytic consideration of the interaction of specific peoples and their ways of life with the environments of specific places.^[2] This has been true in geography perhaps more than anthropology. The recognition of human impact on the earth throughout history has been important in geography for decades (Blaikie and Brookfield 1987; Sauer 1925, 1956; Thomas 1956).^[3]

Khumbu Sherpas, as might be expected of a people who have lived for more than four hundred years in a few small valleys, possess considerable local environmental knowledge of their homeland. This has certainly influenced their land-use practices in many ways and has enabled them to develop relatively sophisticated local resource-management systems through which they have buffered to some degree the environmental impact of their forest use and pastoralism. Local spiritual and other cultural values and beliefs also contributed to channel resource use and moderate some environmental impact, especially through protecting spirit trees and sacred forests.

Sherpas' Buddhist ethics influence daily life and land use also in a number of other ways. Buddhism gives Sherpas a strong belief in nonviolence towards other forms of life. Sherpas value the consciousness within all life, which they believe extends even to the smallest blades of grass. All Sherpas consider the killing of animals a sin and some believe that cutting trees or uprooting plants is sinful (Fürer-Haimendorf 1979:276). Many people go to great efforts to avoid even the killing of insects. When unwelcome insects are discovered inside home they are carefully shepherded outdoors rather than exterminated.

There are limits, however, to the degree to which Sherpas adhere to these Buddhists beliefs in everyday life. It was not true, for example, that "green wood was never cut, mainly due to the maintenance of Buddhist practices" (Bjønness 1983:170). Villagers have long felled trees for timber and for fuel wood. And Sherpas have been know to exterminate wildlife when they believe that it endangers their crops and livestock.

Villagers, for instance, try to prevent pheasants from uprooting their crops of potatoes and grain by constructing scarecrows, mock bamboo traps, and other devices to attempt to frighten off the birds. Before the national park implemented strict wildlife-protection regulations, however, Sherpas sometimes as a last resort sent for the non-Buddhist blacksmiths of Nauje and asked them to shoot the pests. The blacksmiths were also called in to kill snow leopards when these were discovered near the main settlements of lower Khumbu. And Sherpas have so tenaciously trapped and beaten wolves that no more are alive in Khumbu.^[4]

While the conservation qualities of the Sherpas' Buddhist beliefs are important influences on their lifestyles, they must not be over-

romanticized. It would similarly be a mistake to simply ascribe to Sherpas, or any other indigenous people, a set of attitudes and practices that amount to an ethic of living lightly on the land in terms of conservation beliefs that limit household consumption of certain resources or provide ethical constraints on overexploiting resources from particular sites. Some indigenous societies have held such beliefs, or at least certain enlightened individuals have lived and taught these views among their people. But the degree to which these kinds of conservation ethics have influenced land-use practices and continue to shape them today must be carefully evaluated for each indigenous people and community. It would be easy to assume that people living in difficult environments would be very likely to have developed such adaptive sets of ethics, particularly toward key subsistence resources that are most limited, but this need not be the case. I have not found that Khumbu Sherpa cultural attitudes toward resource use are tempered by such scruples. Instead household decisions about the amount of resources to use and where to obtain them are largely made on the basis of lifestyle preferences, individual desires, and convenience. For all the role of religion in daily life, Khumbu economic life nonetheless seems to have operated primarily according to materialistic values which were themselves also culturally sanctioned and socially rewarded. The constant tension between the implications of these private goals and community efforts to enforce certain restrictions on seasonal resource use and particular resource-use practices has long been a significant feature of Khumbu Sherpa life and affected the relationships among individuals and families within villages, levels of support for pembu, and the regional environment.^[5] And the goals and regulations of the community management institutions were themselves also shaped by cultural attitudes about respect for individual household economic freedom, cultural definitions of the good life, and the limits of local environmental ethics.

This view of Khumbu Sherpa resource use and management contrasts sharply with the more romantic "conventional wisdom" about Sherpa life during the pretourism era. Khumbu traditional resource-management institutions in particular have been widely depicted as a fine example of a local system based on communitywide participation and decision making which relatively successfully regulated the use of common property resources (and particularly forest and pasture use) to levels that conserved them. This depiction draws primarily from the description of the nawa system and shinggi nawa institutions as they existed in 1957 given by Fürer-Haimendorf (1964, 1975). Fürer-Haimendorf concluded that:

Compared with the forests of lower and climatically more favoured regions where peasants of Chetri, Brahman, and Newar stock have in

recent generations wrought enormous devastation, the forests of Khumbu are on the whole in good condition. This is mainly due to an efficient system of checks and controls developed and administered by a society which combines strong civic sense with a system of investing individuals with authority without enabling them to tyrannize their fellow-villagers. (1964:112-113)

Many others have echoed Fürer-Haimendorf's evaluation of the shinggi nawa forest management system (eg. Bjønness 1980a , 1980b ; Byers 1987b ; Fisher 1990; Hagen 1980; McNeeley 1985; Rhoades and Thompson 1975; Schweinfurth 1983; Thompson and Warburton 1985). For twenty-five years it has been the standard interpretation of Sherpa forest management, and indeed the most commonly cited depiction of a Himalayan resource-management system. Yet these assertions about the effectiveness of traditional local resource use and management have been reached without much attention to historical change in forest (or pasture) use and management, analysis of the relationships among resources, resource-use patterns, and resource-use regulations, or attention to the Sherpas' own observations of historical change in their local environment.^[6]

This chapter examines evidence of historical environmental change in Khumbu and the ways in which this change reflected both the particular types of "traditional" Sherpa land use and "traditional" local resource management. I begin with a discussion of the possible role of early Sherpa and even pre-Sherpa Khumbu inhabitants in shaping the basic patterns of Khumbu distribution of forests, woodlands, grasslands, and shrublands. I then consider how the historical patterns of Khumbu Sherpa forest use and forest management influenced change in the extent and composition of Khumbu forest between the mid-nineteenth century and the 1960s. I focus here on the period before forest nationalization in the mid-1960s and the subsequent establishment of Sagarmatha National Park, both of which had profound impact on forest use. This chapter thus explores forest use and change during the time when Khumbu resource use and management was entirely in local hands. This is also the period before tourism began to have a major role in the regional economy and resource use. The second part of the chapter looks at pastoral regulation. I analyze the nawa system as a conservation institution and survey some of the historical impact that Sherpa herding practices and local management institutions may have had on Khumbu vegetation.

I take up the environmental impacts that resulted from the decline of some traditional local resource-management institutions after the mid-1960s in the next chapter and examine the environmental impact of tourism in chapter 10.

Fire, Forests, and Grasslands

Forests and woodlands now cover only approximately half of the area that may once have supported forest cover (Hardie et al. 1987:21, 49; Naylor 1970). Virtually all of the remainder of the area below 4,000 meters is grass and shrubland, in which occasional isolated trees are found.^[7] These vast grass and shrub-covered slopes are especially prominent in the vicinity of settlements, each of which is at least partly encircled by a zone of grasslands. The proximity of the main villages to the most expansive, lower-altitude (below 4,000m) grasslands, the importance of these areas as rangelands, and the probability that local climate and soils can support forests and woodlands raise questions about the possible human role in the creation and maintenance of these grasslands. It is possible that Sherpas may have chosen to settle at the margins of already existing grasslands, preferring a settlement site that did not require clearing forest to establish fields and building sites and that was close to both good grazing and readily available timber and fuel-wood resources. Alternatively they may have played a part in creating these grasslands. Whichever the case may be, there is no doubt that they have played a role in maintaining them (and in places expanding them) in recent centuries through their herding activities, forest use, and perhaps through the use of fire.

Oral history and oral traditions suggest that virtually all of the area that is now grassland has been so at least since the late nineteenth century. There are a few localities where small-scale forest and woodland clearing has occurred since the mid-nineteenth century, as will be discussed below. But there is no oral-history evidence to suggest that Sherpas during the past century or more had any part in the creation of the extensive grasslands on the eastern side of the lower Bhote Kosi river, the slopes immediately above Nauje, or the slopes above the Dudh Kosi river east of Khumjung and Nauje. Soil and pollen analyses also suggest that the grasslands are of considerable antiquity. Some of the grassland areas were established long enough ago to have developed a distinctive soil horizon above an earlier forest soil. Analysis of pollen from soil samples (unfortunately no pollen cores from lake sites are available) suggests that open fir/birch/alder forests were converted to grasslands on sunny, south-facing slopes at least 400-800 years ago (Byers 1987b :199-204). Charcoal found in this fifteen-centimeter-deep soil layer suggests that increased frequency and severity of fires could have been a factor in this vegetation change. People could have played a role in this transformation through deliberately setting fires and also by clearing forest.

Other pollen findings suggest that people may have played a role in

the conversion of forest to grassland even if this took place eight-hundred years ago or earlier, well before the arrival of the Sherpas. Byers found cereal pollen (unfortunately of an undetermined variety) dating back perhaps as far as two thousand years (ibid.:199).^[8] If correct, these dates suggest that people have inhabited Khumbu, or seasonally visited it, for a very long time indeed. The idea of ancient Khumbu settlement or resource use, while startling to those, including many Sherpas, who have believed that Khumbu must have been uninhabited when Sherpas arrived, is not entirely alien to Sherpa oral traditions. There are Khumbu oral traditions of pre-Sherpa settlement, such as the tradition mentioned in chapter 1 that the ancestors of today's Rais attempted to settle in Khumbu for some years before ultimately abandoning it as a permanent habitation site (although they may have continued to use Khumbu for summer grazing ground). It is also possible that earlier groups had crossed into Khumbu from Tibet just as the ancestors of the Sherpas ultimately did, later either moving on or being incorporated into the Sherpa population. Of this, however, there is no oral tradition except a possible reinterpretation of the story of a Sherpa who is said to have come to Khumbu at the very beginning of the settlement of the region via Rongshar and Rolwaling, rather than over the Nangpa La, and to have lived in the Bhote Kosi valley of Khumbu. It may be that this tradition in fact recognizes that there were migrations from Tibet into Khumbu even before the Sherpas arrived, perhaps long before they arrived, and that the time frame has simply been compressed in the oral tradition to become incorporated with the oral traditions of early Sherpa settlement of the Bhote Kosi valley.

The charcoal found in soil-sample layers, which corresponded to the early era of major Khumbu forest change and the establishment of the grasslands, may be evidence of the early use of fire in farming and pastoralism. According to Markgraf, who conducted the pollen analysis of the soil samples Byers collected, the pollen evidence and the associated charcoal in the samples suggest both fire and possibly logging as factors in forest change (V. Markgraf 1987, cited in Byers 1987b :199). The use of fire to improve grazing conditions and perhaps also for swidden cultivation may have had an impact on much more extensive areas of forest than would have the felling of trees to meet the timber and fuel-wood needs of a presumably rather small number of people. If the Rais or more ancient groups from the country south of Khumbu had settled long ago in the high valley or grown supplementary crops there while using the high alpine summer pastures it is probable that they, like most of the peoples of the hill region of eastern Nepal before 1850, practiced swidden cultivation.^[9] Buckwheat and possibly other crops could have been successfully swidden cultivated in lower Khumbu.

Gurungs haved historically carried out swidden cultivation at comparable altitudes in central Nepal and as recently as the 1960s Pharak Sherpas used swidden fields to grow buckwheat. Early Khumbu settlers may also have used fire in order to improve grazing. It is a very common technique in much of the Himalaya to fire the forest and woodland floors in early spring in order to improve the growth of new grass. This is done by Pharak Sherpas.

It is impossible at this time to say whether or not the Sherpas themselves were involved in the early creation of the Khumbu grasslands. If Byers's estimate is correct that the key final phase of vegetation change took place between eight hundred and four hundred years ago (1987b :199), it is conceivable that Sherpa land use may have played a late role. Khumbu oral traditions, however, offer no insights into possible early Sherpa forest clearance or early Khumbu Sherpa use of swidden agriculture. In the past hundred years Khumbu Sherpas have had no tradition of setting fire to forests and woodlands to improve grazing or clear forest. But this does not rule out their earlier use of fire. The early Khumbu settlers might have already been familiar with swidden techniques from Kham, for several peoples of present-day Sichuan and adjacent Yunnan certainly make use of them. Or they could have learned them from their new neighbors in Nepal, as Sherpas have presumably done in Pharak, Shorung, and the Arun region. Sherpas could also once have used fire to improve pasture in a way that they have long since abandoned.^[10] As of now, however, we simply do not know. Further pollen interpretation will be needed to refine our understanding of the dynamics and timing of early Khumbu vegetation change, and archaeological study is needed to clarify the early settlement history of Khumbu.

Traditional Forest Management: Strengths and Limitations

It is likely that the most extensive conversion of Khumbu forest to open woodland, shrubland, and grassland took place centuries ago, conceivably long before Sherpas arrived on the scene. Yet Sherpas have also had environmental impacts on forests over the past four centuries. The nature of these impacts was related not only to historical patterns of local forest use but also to the effectiveness of local forest management. Change in Khumbu forest during the period of Sherpa settlement reflects the goals and effectiveness of local Sherpa managment of different types of forests and the types of vegetation change associated with their use both of these protected forests and other unprotected Khumbu forests.

The goals set for the various types of protected forests generally appear to have been fairly well achieved. Sacred groves were kept from

desecration. Administration of bridge forests, avalanche-protection areas, and rani ban generally met the objectives set for them. Residents of Nauje, Khumjung, Kunde, and Pare were able to continue to find beams in adjacent rani ban, Bachangchang continued to supply critical soluk.

Elderly Sherpas maintain that throughout their lifetimes both sacred forests and some rani ban were very well protected. The strength of local respect for sacred groves remains notable to the present day. Phurtse villagers, renowned in Khumbu for their strict protection of the local lama's forest, often observe that there was little for shinggi nawa to do because no one would think of violating the customs protecting the sacred forest. They say that in the especially sacred forest area north of the settlement people kept an eye even on their own relatives. The sacred forest at Yarin, the sacred junipers at Pangboche, and temple groves throughout Khumbu have also been relatively well protected from cutting and lopping throughout this century.

The regulations established for several of the rani ban were also well maintained. The Bachangchang rani ban, for example, was long exceptionally well administered during the early part of the century when Yulha Tarkia was pembu. Villagers declare that the rules were nearly as carefully observed during the years when his son Konchok Chombi had power. The fact that Bachangchang had been considered a sacred forest by local villagers before it was designated a rani ban must have made the enforcement of protective regulations there relatively easy. The Khumjung-Kunde rani ban and the rani ban near Nauje were also relatively well maintained during the course of several changes in pembu administration, and regulations were carefully enforced in both until the mid-1960s. The Bhotego rani ban, which was administered by a rotation of the shinggi nawa office among the residents of Pare, is also known for having long strictly enforced use regulations.

Control of some other protected forests, however, was less successful. Some of the rani ban were not strictly protected for long, as pembu or their successors seemed to lose concern with ensuring that regulations were enforced. This seems to have been the case at Samshing, Tesho, Gupchua, and Nakdingog, where any formal enforcement of regulations appears to have lapsed well before the 1950s. Even when pembu or local communities did maintain fairly good control over forest use it was not always possible to enforce the rules. Fürer-Haimendorf observed in Khumjung in 1957 that shinggi nawa had had to issue several fines that year for unauthorized tree felling (1979:112). In some formerly protected areas near Nauje people note that for years there were small-scale violations of the rules. Clandestine tree felling took place over a period of decades before the 1960s especially when dense summer mist or low clouds made detection difficult. One elderly villager, when describing

what had happened to the trees in the Chorkem area near Nauje earlier in the century, said, "when the clouds came in the trees went away."

For the most part, however, Sherpas continued to respect the customs that placed sacred forests off limits to cutting and restricted the use of rani ban and some other protected forests. Violations of the rules, other than the cutting of much juniper at Shyangboche in the early and mid-1960s, were small scale and seem to have been relatively uncommon. Community cooperation, social pressure, and the continuing practice of appointing shinggi nawa and other forest-management officials appear to have kept forest-management systems functioning relatively effectively until the 1960s.

Yet local management had significant limitations in terms of how well it regulated local forest use to regionally sustainable levels. These shortcomings had nothing to do with how effectively regulations were enforced. They were instead the result of the rules themselves, or more accurately, of the forest management objectives they were devised to meet. These goals did not include regulating regional forest use to sustainable levels or maintaining, protected forests in a given state of composition or density. In all of the protected forests uses were allowed which were capable of transforming the density, composition, age-structure, and extent of those forests. Nowhere in Khumbu, for example, was any forest or woodland ever closed to grazing throughout the year, much less for a period of years. At most, protected forest areas in lower Khumbu were closed to grazing for three months during the summer, and during the rest of the year many of them endured considerable grazing pressure. This nearly unregulated livestock browsing and trampling may have had an adverse effect on forest regeneration, and the lack of more than brief seasonal restrictions on the collection of dead wood, leaves, and needles from the forest floor may have further contributed to this by removing valuable nutrients. In a number of protected forests, including nearly all of the rani ban and the bridge forests, trees could be felled with proper authorization, and it was up to the user to decide which trees to cut. This selection was often made with convenience in mind rather than any view to minimizing environmental impact and forest change. There were no cultural conservation ethics or community customs or regulations that limited the size of one's home or required that one fell trees only in areas where similar trees were plentiful or other Sherpas were not cutting. Each household chose which trees it wanted to fell in a rani ban and the key concern in this selection seems usually to have been to obtain appropriate trees as close as possible to the building site. There were thus several significant loopholes in protected forest management which could have affected both the long-term sustainability of use and the integrity and continuity of the forests themselves.

There were thus several serious limitations to local forest management as a system of conservation. There were even greater limitations to local forest management when it is viewed on a regional scale and all Khumbu forests are considered rather than just the protected forests. Only about half of the Khumbu's forested area was under one of the several local management systems. The remaining forests were affected by the patterns of use that the protected forest regulations established, however, because the relatively strict protection of sacred forests and, in the twentieth century at least, of rani ban, tended to shift the main demand for timber and fuel wood away from the protected forests and onto forest and woodland areas outside their boundaries (map 18). In these areas there were no rules whatsoever. Here villagers cut as much as they wished from wherever they chose. Time, energy, and demand governed the process which followed. Shinggi nawa or other administrators exercised no authority here, nor were there any cultural constraints on the scale of individual household fuel wood or timber use. At the regional level Khumbu forest management thus created a situation in which some forests were protected relatively well and pressure was instead focused on others in which a lack of management prepared the way for the classic "tragedy of the commons" scenario described by Garrett Hardin (1968) in which unregulated individual access to commons ultimately leads to environmental degradation.

Map 18 illustrates patterns of forest use by Nauje, Khumjung, and Kunde villagers after the establishment of the local rani ban and before Sagarmatha National Park regulations banned tree felling in Khumbu other than for house beams.^[11] As can be seen in this map, dead wood and beams could be obtained from rani ban areas immediately adjacent to the settlements, and prior to the establishment of the rani ban timber was also obtained from these areas. People recall, for example, that the forest near the Khumjung village entry arch (kani ) was a very good source of boards at the beginning of this century. Even at this time, however, more remote areas were also being exploited for timber, including forest on the far side of the Dudh Kosi. After the establishment of the rani ban, tree felling for timber, rafters, and fuel wood was channeled as shown into other nearby, unregulated areas. The resulting impact on those forests and woodlands is recorded in the landscape, in oral traditions, and in living memory.

Early Sherpa Impact on Forests

The uncertainty about the extent of forest cover in Khumbu at the time of the first Sherpas' arrival there and the lack of oral traditions about forest use during the early centuries of Sherpa settlement make it impos-

Map 18
Village Forest Use Patterns, 1915-1976

sible to evaluate fully the scope of Sherpa impact on forests over the past four centuries. Oral traditions and oral history, however, do make it possible to say something of the scale and processes of forest change during the last hundred to hundred and fifty years.^[12] Sherpa interpretations of place-names, family traditions about the sources from which trees were obtained generations ago for house beams, and elderly Sherpas' recollections about the forests of their youths all testify to historical deforestation.

Place-names contain several possible clues to the former location of forests. One widely cited example is Namche Bazar, the Nepali name for Nauje, which has been said to derive from the Sherpa for "big forest" (Byers 1987b :201, n. 9) or "dense forest" (Bjønness 1983:270). Sherpas are uncertain about the origin of the name Namche Bazar, but some speculate that the name Nauje may be derived from a phrase which would translate literally as "big corner forest," possibly a reference to the natural amphitheater shape of the place and its past vegetation. The upper part of the basin in which the village is set is now conspicuously bare of trees and only a few scattered lu-inhabited trees are found within the settlement area of Nauje itself. Elderly residents indicate that the slopes above Nauje have not been forested during their lifetimes and that they were not told of major deforestation there during the lifetimes of their parents and grandparents.^[13] There are oral traditions, however, about smaller-scale forest change. Some people have heard that there was once fir on the south side of Nauje near the present site of the weekly market and that villagers formerly (probably before 1900) felled trees there for rafters. The name of one field in the central pair of Nauje, Tongbajen, is thought to possibly refer to large rhododendron, and here and in several other areas in the village, as well as on the slope immediately above it, many roots were found when establishing terraces half a century or more ago. It is also remembered that until the 1960s there were more juniper in one small area of the upper slopes near the crest of the basin where today several abandoned terraces can be seen. Formerly there was a hermitage at that site, and the juniper were cut down after it was abandoned as a retreat.

The stories told about house beams recall forest where there is none today. In Thami Teng and Thami Og, for example, there are houses that have been rebuilt several times and have immense juniper beams of a size unobtainable today in Khumbu. The prized standing beams often have a history, for the work of cutting and hauling them is long remembered, and beyond that they are regarded as one of the house's more sacred features. Stories about them are sometimes passed through generations. In Thami Og and Thami Teng such stories tell of juniper cut at sites which are now adjacent to the village or within the settlement area by the grandparents of villagers who are now in their seventies. In those places today there is not a tree standing. In Phurtse similar stories are told of the building of the village's early houses with juniper from now bare slopes, and in Nauje at the end of the nineteenth century it was possible to fell large juniper from an area now in the middle of the village.

Oral histories of forest change and deforestation are much richer, of course, from the experiences of living Sherpas. Here there is abundant recollection of the disappearance of trees from subsistence use areas. In

the 1930s, for example, a primary wood-collecting region for Pangboche was a juniper woodland along the Imja Khola to the west of the village. A short distance downstream Phurtse villagers cut juniper for fuel wood on the slopes above the Imja Khola. Both these areas ceased to be major fuel-wood gathering areas forty or fifty years ago after wood had become scarce. Several areas along the Dudh Kosi between Nauje and Teshinga were cut by fuel-wood collectors over the past half-century. Woodland was thinned and cleared at Shyangboche and from several sites on the slopes above Khumjung. Tree felling for timber and fuel wood also affected birch forest near Thamo, birch and fir forest north of Samshing, birch and rhododendron forest above Thami Teng and near Kerok, and birch forest further north along the Bhote Kosi at Dokyo. Elderly Sherpas also note more gradual changes in the composition of some forests which took place over the course of many decades. The forests at Bachangchang and Tesho and the small grove south of Phurtse are all said, for example, to have gradually decreased in density and to have changed in composition, with fewer large birch and more young rhododendron today than before. Table 25 lists sites where pre-1965 forest change was reported, and these are indicated on map 19. Note the correlation here between unregulated forest and reported areas of forest change. Those protected forests where pre-1965 degradation occurred are usually either considered to have been laxly administered or were places that had lost their protected status some time earlier as a result of village boundary disputes.

The total amount of forest area lost during the twentieth century has been small, for each of the sites mentioned above as having experienced deforestation or other change is at most only a few hectares in size. Yet the cumulative effects are noticeable to those who remember the forests of early in the century. In 1985 an eighty-three-year-old Sherpa from Thami Teng told me that in his youth it was not possible to look from the western side of the Bhote Kosi valley and see people moving across the river along the forest paths between Thamo and Nauje because the trail was hidden within the forest. Now, he pointed out, there are long stretches of clearly visible trail in that area and people can be seen moving up and down valley.

When examined from local perspective and in historical context Khumbu forest management was a diverse set of institutions and practices that had a variety of economic, environmental, and spiritual concerns. Khumbu forest management was not a static, "traditional" institution. Instead forest management reflected the dynamic development of a number of different approaches to using and conserving forests which embodied cultural beliefs and assumptions about the natural and supernatural environment, natural resources, and the proper sphere of social

intervention in household economic decisions. Some aspects of forest use and management in Khumbu reflect an awareness of environmental change and local institutional responses, to it, yet the pattern of resource use and the nature of institutional and cultural regulation suggest that regional sustainability of forest use was not the primary orientation that outsiders have assumed it to be. Sherpas have had impacts on forests within the remembered past, and had not developed a regionally sustainable system of forest management before the establishment of Sagarmatha National Park.

Map 19.
Reported Forest Change before 1965

The continuing preservation of fine forest in the immediate vicinity of villages, however, does reflect an extraordinary continuing commitment in Khumbu to the protection of sacred places and the timely development in some areas of new institutions to protect key sources of certain forest products. The diversity of forest management systems in Khumbu and the relatively large area regulated by them is unequalled among the five peoples who inhabit the Dudh Kosi valley and unparalleled among the Sherpa groups with which I am familiar. Although earlier assumptions about local resource use and management in the Mount Everest region certainly require major revision, Sherpa forest management in Khumbu nevertheless may well remain an example of local resource regulation that is exceptional by Himalayan standards.

Traditional Pastoral Management: Strengths and Limitations

The evaluation of the degree to which past and present Sherpa pastoralism has represented an effective adaptation to local environmental conditions hinges on two questions. One is the degree to which Sherpas have been able to develop pastoral strategies, techniques, and social arrangements based on local knowledge of Khumbu environment which enable them to limit risks to levels they consider acceptable. The second is the degree to which regional livestock-keeping practices and grazing levels have been maintained within the carrying capacity of Khumbu rangelands. Here the issue is whether or not pastoralism has been carried out and regulated in a manner that is environmentally sustainable in that it has not diminished the productivity of regional rangelands. These questions are interlinked, for ultimately even small-scale and gradual diminishment of the productivity and carrying capacity of regional rangeland may lead to increased risks for stockkeepers unless new patterns of pastoralism are adopted or grazing intensity is reduced through changes in herd size, stock types, procurement of feed from outside the region, or increases in the productivity of hay cultivation in Khumbu itself.

In examining the success with which a people has coped with the challenge of adjusting grazing levels and intensity to local pasture conditions there are several basic factors to consider. Do cultural values, community institutions or outside authorities act to control grazing pressure through approaches such as controlling herd size and composition at household, community, and regional levels, rotational use of pastures, or limits on the number of stock grazing particular pastures? If communal systems regulate pasture use precisely how do these function and how effective are they? And is there evidence in the landscape and in

local perceptions of changes in vegetation, pasture quality, fodder availability, or slope stability?

Khumbu Sherpas depend more on institutional regulation of grazing than do some peoples. Few cultural beliefs affect decisions about herd size and structure other than to encourage the keeping of cattle and yak rather than smaller stock and to equate large herds with high prestige.^[14] There are no beliefs that it is unlucky to graze in a particular place more than a certain number of days, or that to have more than a particular number of animals in a given site is inauspicious. Decisions about herd size are entirely left up to the individual family as are decisions about the movement of herds during most of the year. It is only during the summer and early autumn, when the nawa system is in operation, that the community intervenes in any way in household pastoralism.

The operation of the nawa system thus becomes the pivotal place at which to assess the adaptiveness of Khumbu pastoralism in terms of its effectiveness in regulating pasture use to sustainable levels. Yet it is precisely this detailed knowledge of the functioning of the institution relative to environment which has been lacking in earlier work on Khumbu pastoralism. Earlier treatments of pastoralism have described the nawa system without mentioning the existence of zones of pasture management (Fürer-Haimendorf 1979), ignored community pastoral management altogether (Bjønness 1980a), or described the nawa system as a rotation system of grazing without critically assessing it (Brower 1987). Here I will examine Khumbu Sherpa pastoral management in more detail and then look at evidence for grazing-related environmental change.

Evaluating Sherpa pastoral management systems, or indeed any system of common-property resource management, requires attention to several different major features. First the effectiveness of the system must be evaluated in terms of its achievement of its own goals. How well, for example, are rules enforced that limit the group of resource users, such as by restricting use of commons to the members of a particular village? How well are they enforced within the community itself, and particularly against attempts to circumvent them by politically powerful families? On a second level the orientation of the rules themselves must be considered. What are the goals of the system? Is environmental sustainability one of them? What means are taken to achieve this? Are they adequate, or are there shortcomings inherent in the system itself? At a third level the flexibility of the system must be evaluated. How well can it be adapted to perceived changes in environmental conditions or to newly evolving patterns of land use and economic orientation? How well can it be maintained in the face of commercialization, redefinition of the membership of the resource-users' group, or challenges to its existence

by government through the nationalization of land, the establishment of other systems of resource management, rejection of local efforts to limit resource use to a particular group, or challenges the legality of specific local use regulations?

In the first of these dimensions, the enforcement of regulations against both Sherpas and outsiders, the Khumbu nawa system has been relatively successful as a form of pastoral management. Throughout the region nawa continue to effectively administer the opening and closing of zones to livestock and hay cutting except in the Bhote Kosi valley and Nauje. Where nawa management has been maintained the system has successfully protected crops from livestock depredation and from fear that village activities might cause catastrophic blights. It has continued to reduce the risk from year-round pastoralism in Khumbu by protecting expansive areas of lower Khumbu from summer grazing, thus ensuring both that there will be considerable wild grass available for autumn hay making and that there will be good winter grazing in the main village areas. In eastern Khumbu the operation of the more complex zonal system of the upper Imja Khola and Lobuche Khola valleys has established a limited rotational grazing system in the late-summer high pastures. Until the undermining of the system in Nauje in 1979 and Thamicho by 1984 the regulations regarding pastoralism were consistently implemented throughout Khumbu despite a number of challenges by Sherpa villagers, non-Sherpa Gurung herders, and Tibetan refugees. Attempts by Gurung shepherds in the late ninteenth and early twentieth centuries to obtain exemptions from observing the zone closures were resisted. Until the late 1970s no major challenges by Sherpas had succeeded in eroding the authority of communities to administer nawa regulations or seriously compromised the enforcement of regulations. Several attempts by villagers from one community to deregulate another community's restrictive grazing zones by challenging village land boundaries came to naught, although to preserve the grazing controls in one case the men of Khumjung had to march to Nauje to back up their nawa's enforcement of the herding regulations with the threat of violence. A number of attempts by wealthy and powerful herders of several different villages to ignore the rules were thwarted by social pressure and in one case, in Phurtse, by a change in rules governing the selection of nawa to prevent political influence from undermining equitable enforcement. Even the widespread disobedience of nawa regulations at Phurtse during the 1940s, when many yak herders moved their stock into the village before it was authorized, did not lead to the collapse of the institution there. Instead the enforcement of the regulations was tightened successfully the following year and has been maintained ever since. The undermining of the Nauje nawa system in 1979 by massive

civil disobedience and the parallel collapse of pastoral regulation in Thamicho in the 1980s are recent anomalies in what has otherwise been a long record of successful operation.

The Khumbu nawa system has a more mixed record at the second level of evaluation of communal management systems, where goals rather than merely enforcement are considered. The system clearly has a number of different goals: protecting crops from livestock, protecting winter grazing, protecting fodder through keeping stock out of hayfields and areas where wild grass is cut for hay, restraining the collection of wild grass until it has matured, and rotating grazing through a series of zones. Some of these are immediate, subsistance-related goals: protecting the harvest, ensuring that stock can survive the winter and spring, and maintaining good grazing through the summer through shifting the concentration of grazing. Some of these goals deliberately, or perhaps inadvertently, also support environmental sustainability. Rotational grazing limits the intensity of grazing of particular areas and decreases the risk of overgrazing. It is important, however, to recognize that the nawa system has systemic limitations in terms of conservation objectives. Although some dispersion of grazing is certainly accomplished through the opening and closing of various zones to livestock, in four of the five systems this rotational system only operates in limited lower-valley areas, leaving most Khumbu rangeland as unregulated, open-access commons.^[15] There is no mechanism to control the intensity of grazing within a given zone by placing limits on the number of stock that can be grazed there or to set a ceiling on the number of stock herded in a village, valley, or in Khumbu as whole.^[16] Herders may decide that poor grazing in one area is a reason to move elsewhere, but this is solely up to the household itself as are all decisions about household herd size and composition. The system does not include any mechanism for incorporating environmental monitoring and response. Nawa do not evaluate pasture conditions and have no authority to respond to observed increases in grazing intensity or deterioration in pasture by ordering or recommending special changes in zonal boundaries or the period when areas are open to grazing. Such changes could be made by a consensus of residents and herders, but there are no traditions of such issues having been brought up for consideration and acted on. The dates of zone closures and openings in some cases are relatively fixed and embedded in a socioreligious calendar and the border zones have not been adjusted for generations.^[17] There is also no precedent for closing a degraded grassland or woodland area to grazing in order to allow it to recover from intensive use.

There thus remains within the nawa system itself the possibility that overgrazing can take place and cause the degradation of pastures and forests without there being any compensatory mechanism. To control

localized overgrazing would require expanding the powers and concerns of the system beyond any historical precedent. The lack of strong conservation goals suggests that pasture deterioration might be allowed to proceed for some time without any institutional response, for to redefine zone boundaries or take more radical steps such as limiting the number of stock requires fundamental changes in the institution of a sort that may be socially difficult to arrive at. Recent Nauje and Thamicho history suggests that strong demands on resource use and increasingly scarce resources may lead to the abandonment of the system rather than attempts to modify and strengthen it, with grass cutters and herders choosing to compete with each other without community intervention or long-term environmental concerns.

The nawa system thus represents relatively constrained communal intervention in individual household herding decisions. In most areas Sherpas enforced little more than the minimal level of stock exclosure necessary to protect crops from livestock depredation and blight and to protect winter pasture and fodder resources. The continuing institutional capability to achieve these goals is now under severe testing in the region. For a century or more the nawa system has operated on the basis of community consensus and social pressure. This base has deteriorated over the past ten years in much of Khumbu. Communal management of livestock in the Bhote Kosi valley has been discontinued, is under pressure in Khumjung-Kunde and in the upper Imja Khola valley, and was abandoned in Nauje in 1979. This situation and what it suggests about the limits of the institutional adaptability and flexibility of the system are discussed in the context of the decline of local resource-management institutions in the next chapter.

Grazing and Historical Vegetation Change

Grazing has probably had a significant role in shaping Khumbu vegetation and landscape. Centuries of herding by Sherpas, Gurungs, and possibly Rais or other lower Dudh Kosi valley peoples have brought thousands of yak, cattle, crossbreeds, sheep, and goats to graze and browse in Khumbu grasslands, shrublands, woodlands, and forests. Historical changes in livestock numbers and types, patterns of transhumance, the regulation of community grazing, and the relative importance of hay and other fodder, all contributed to place different degrees of stress on different Khumbu rangeland areas. Centuries of intensive use of certain areas could be expected to substantially pattern vegetation. So too would herders' efforts to improve grazing by the use of fire which, as already mentioned, could even have shaped the relative distri-

bution of forests and grasslands familiar from the past century. The vast expanses of grasslands below 4,000 meters that have provided such fine grazing as far back as oral history can illuminate could very well be a legacy of pastoralism.

Yet it is difficult as of now to analyze the historical extent of vegetation change in the region due to grazing. Oral traditions and oral history have little to say about declining pasture conditions before the past few decades. There is no testimony that any particular herding area has dramatically declined in productivity during the nineteenth and early twentieth centuries, much less that grazing has been abandoned in any part of Khumbu because of gullying, landslides, or heavy erosion related to pastoralism.

It seems probable that the composition of much of the rangeland vegetation in Khumbu reflects the historical effects of cattle grazing. In particular the relative role in Khumbu grassland and shrubland vegetation of a number of relatively unpalatable plants such as barberry, rhododendron, rose, ephedra, iris, and potentilla probably reflects the cumulative and continuing effects of pastoralism (Brower 1987:277). The great changes in Khumbu pastoralism in the past fifteen years would be expected to have had a particular role in shaping the vegetation in some areas, especially the lower Bhote Kosi valley and the Nauje area where large numbers of crossbreeds have been grazed year round for the first time in known Khumbu history. But the relatively high incidence in the vicinities of main villages and heavily used, higher-altitude herding areas of certain temperate and subalpine species that are poisonous or unpalatable to yak, cattle, and crossbreeds could be expected to be a general pattern that goes back long into the past. In some areas there may be more unpalatable, thorny rose and barberry shrubs today than decades ago, but Sherpas recall that concentrations of such shrubs near villages are not new in the Khumbu landscape.

The spectacular terracetting so ubiquitous in lower Khumbu and certain other areas such as Dingboche also has historical origins. Vast slopes are tightly patterned with tiny terraces two or three feet deep and several feet high, apparently the result of many generations of livestock crisscrossing the landscape in search of grazing. Some observers have taken these terracettes to be evidence of recent overgrazing and they undoubtedly are a reflection of continuing grazing pressure. Areas recently closed to grazing by the national park to establish forest plantations have begun to recover within a few years as grasses and annual plants recolonize and grow over the bare ground of the cattle trails. Yet while extensive networks of cattle trails reflect continuing patterns of high grazing intensity, they are a long-standing feature of the landscape. Sherpas report that the slopes near Nauje, Phurtse, Dingboche, Pang-

boche, and other sites have been terracetted in this fashion throughout their lifetimes.

The composition of forests, woodlands, and shrublands has also very likely been affected by grazing, particularly in areas near main settlements where seasonal herding patterns have long focused considerable grazing pressure. Browsing and trampling have probably inhibited seedling survival and maturation and may account for forest areas near some villages that have aging stands with little evidence of regeneration. In other areas there seems to be little evidence of regeneration other than by rhododendron, which is unpalatable to local livestock. Over time this has probably contributed to the twentieth-century transformation of birch and fir forests near Phurtse and Thami Og to rhododendron thickets.^[18] Grazing is also very probably a major factor in the noticeable lack of tree colonization in shrubland and grasslands adjacent to forested areas.^[19]

During the past half century the decline in the use of winter pastures in Tibet, the increasing emphasis on crossbreed and cow keeping, and the breakdown in effectiveness of communal pastoral management have all affected the use of Khumbu pastoral resources. There is some evidence that increased competition for grazing and grass for hay has led to a growing scarcity of critical winter pastoral resources in some areas and that the intensity of grazing may also be decreasing long-term range productivity. This is not entirely only a phenomenon of the late 1980s. Some herders in the Bhote Kosi valley, for example, are convinced that grass was much better in that area in the 1950s and 1960s before more Thamicho stock began wintering in Khumbu rather than in Tibet. As one large nak-herd owner noted, during the era when the herds went north to Tibet in winter, the spring in Khumba was a time of fattening for stock. Since the 1970s herders have had to feed nak hay all winter and when they really need it in spring none remains. A number of Khumjung herders also note that earlier in their lives both hay and wild-grass fodder was much more available than today even though the number of livestock then herded in the upper Dudh Kosi valley was much greater and the number of hayfields, on the western side of the valley at least, was about the same as today. They also relate this to the decline in the practice of herding nak and yak in Tibet during the winter as well as to the increased Khumjung herding of zopkio, zhum, and cows.

Herders also note that there was a major decline in the quality of grazing in some parts of Khumbu during the early 1960s as a result of the large numbers of Tibetan refugees who brought their yak, nak, and sheep with them. For a few years Khumbu summer pastures were especially intensely grazed and there were severe shortages of winter and spring grazing. Herders who did not have considerable fodder stored ran the risk of having their stock starve. The herds of the Tibetans were

particularly at risk and the numbers of these stock that perished in the early 1960s were tremendous. After this, however, grazing pressure decreased and Sherpas do not note any long-term, adverse impact from this period on Khumbu pasture quality.

Adaptation, Land Management, and Environmental Change

Khumbu Sherpas are an excellent example of a people who during the course of many generations have come to know intimately the resources and environment of their homeland and who have developed strategies for making use of its opportunities and avoiding its risks. Their development of local resource-management institutions has been one component of this adaptive way of life. In Khumbu this institutional development has been the central way in which local demand on natural resources has been regulated and has been much more important than the development of cultural conservation ethics in regulating individual and household land use and their environmental impact. Yet the power of the community to intrude into household economic behavior was also limited by cultural and social beliefs about the limits of communal management and the community's right to influence household economic decisions.

There were also other limits on communities' powers of resource regulation and management. In many parts of Nepal, as in other regions of the world, villagers have limited use of common property resources to their own residents or even to particular groups of residents. In Khumbu, however, community resource management could not draw on such power. Since the mid-nineteenth century villages have not been able to restrict the use of common lands to their residents even though there were defined village land boundaries that were legally upheld and over which court battles were fought. The declaration of a particular area as the property of one village or another only influenced which community might decide if the area was to be treated as a protected forest or as a zone in which grazing and grass cutting might be regulated by nawa. But no village had the right to limit access to that land to its own residents or to enforce the resource-use rules differently for outsiders. Nauje people could cut trees in the Khumjung-Kunde rani ban as long as these were to be used for house beams and the shinggi nawa were consulted or could take their stock to graze anywhere in Khumbu as long as they respected the local nawa-enforced regulations. For many years even Gurung shepherds were welcome to use Khumbu resources provided they adhered to local community management regulations.

Khumbu resource-management institutions have also historically been unable to limit levels of resource use through enforcing household or

community ceilings on the use of forest products, the amounts of wild grass cut for hay, or the number of stock pastured in particular areas. And while nawa, shinggi nawa, and pembu had the authority to issue fines, there have never been any use fees which could have been used to manage the level of resource demand and the ways it was met. Communal institutions thus were not a mechanism that limited the scale of household resource demand or influenced local standards of living.

No Khumbu resource-management institution was charged with monitoring the scale and intensity of use in particular places and associated environmental change, much less was given the power to respond to these observations by adjusting resource-use policy and regulations. Nor should it be assumed that the goals of Khumbu Sherpa traditional resource-management institutions were foremost and centrally concerned with environmental conservation. These institutions addressed a broad range of local concerns, not all of them by any means concerned with moderating the environmental impact of subsistence. Some central goals were concerned only with safeguarding subsistence resources in as short-term and specific a fashion as keeping livestock out of fields. Other goals concerned broader kinds of subsistence protection such as defending against disease and bad luck by protecting forests, trees, and springs which were the domains of powerful spirits. Such goals and the rotational grazing regulations and restrictions on tree felling and pollution through which they were achieved certainly had important ramifications for environmental quality and resource conservation. But these were by-products of economic and spiritual concerns, not the result of either core beliefs in preserving biological diversity or native vegetation communities or a desire to place regional resource use in general on an environmentally sustainable basis. Such cultural values were not translated into Khumbu-wide environmental and land-use policies nor coordinated into community or regional planning as explicit goals of local resource-management institutions.

Traditional pastoral, agricultural, and forest-management systems, however, were concerned in some ways with the adverse environmental effects of unmanaged, individual household resource use. This included concern with change in specific forests, sometimes reflecting a concern for protecting trees in order to spiritually protect the community and in other cases to ensure a sustained supply of timber and forest litter. Despite the shortcomings of local resource management from the standpoint of regional protection of native vegetation and biological diversity, the development of the nawa system of pastoral management and the rani ban and bridge-forest systems of forest use are evidence of local concern with moderating demand on certain resources to ensure their availability for at least the near future.

In Khumbu, as in many cases elsewhere in Nepal (Zurick 1990:32), perception of a real need (such as averting an environmental crisis or natural resource shortages that would affect local lifestyles) may have been the crucial factor in mobilizing community support for implementing such changes in resource-use patterns. In the case of Khumbu forest management it seems to have often been the case, as Gilmour (1989:3) suggests for other parts of Nepal, that "a perceived shortage of forest products is the stimulus that has been most important in causing individuals and communities to change their patterns of behavior toward their use of forests."^[20] A perceived scarcity of important forest products does seem to have been a factor in the selection of some rani ban sites and their subsequent protection through community support. The pembu who selected forests near Nauje, Khumjung-Kunde, Pare, and Samshing (and perhaps also Nakdingog and Gupchua), and who instituted rules allowing only the cutting of trees for beams in those places, seem to have been reacting to concern about a threat to these resources.^[21] Yet it should also be noted that the first Khumbu protected forests—and those which have remained the best conserved—were created not out of concern with the supply of forest products or avoiding immediate environmental calamities but out of a community dedication to honoring sacred places and trees.^[22]

The history of the Khumbu protected forests does not fit neatly into a model such as the one Gilmour develops in which a direct relationship exists between the level of accessibility to forest resources and the degree of community resource management (ibid.:4).^[23] It might be expected that there would be relatively little interest in protecting forests when forest resources are abundant near a community and that a community's concern with regulating individual resource use would rise as forests were depleted and their resources became scarce—especially as the gathering of fuel wood and other products began to require a full day's trip from home (ibid.). But the first and best-protected forests in Khumbu, the sacred forests of the Phurtse and Pangboche areas, were established adjacent to villages where there was not yet any shortage of forest products, whereas there was no community response in other areas such as Thami Teng, Thami Og, and Yulajung where nearby forest resources were depleted even in the nineteenth century. In this part of Khumbu, when the rani ban were established in the early twentieth century, there were no suitable forests left to be protected near the villages other than the sacred grove at Bachangchang. Indeed, rather than being an example of a place where there has been a simple relationship between increasing resource scarcity and local institutional response, Khumbu seems to be a much more complex set of different, special cases. No local community action had been taken anywhere in

the region to establish protected forests in which only limited cutting was authorized until orders came from Kathmandu to establish the rani ban. Only this intrusion into local politics permitted several pembu to mobilize community support to conserve resources. Since then, however, perceived threats to resources have inspired other community efforts at forest protection, including efforts to revive local protection of some forests during the past decade.

Sherpa resource-management institutions thus both reflected adaptative responses to environmental degradation and increasing resource scarcity and illustrated a lack of a holistic, historical concern with regionally sustainable resource use and environmental protection in the sense it has now come to be regarded by ecologists and conservation planners. This should not detract, however, from recognition of the significance of the Khumbu Sherpa achievement in developing their systems of forest and pastoral management. These institutions were for the most part highly effective in meeting the goals for which Sherpas designed and maintained them. Such attempts to control and channel natural resource use through community and even valleywide systems are not universal in the Himalaya, and the Khumbu institutions are certainly more complex, sophisticated, and environmentally protective than those developed by their neighbors. These institutions reflect the application of considerable creativity, foresight, and social and political skill by both individual Sherpa leaders and their communities across several generations.

The preliminary reevaluation of Khumbu traditional resource-management institutions I have attempted here suggests the complexity of resource-management goals and means and the historical dynamism of Khumbu resource management. It underscores the caution with which the subject of the historical relationship between Sherpas and their environment must be approached. Sherpas have had an impact on their environment. They have fashioned a productive landscape of terraces and rangelands from what may have been a rather different environment at the time of their initial settlement of the region. In the past century and a half they have in some places rolled back the edges of forests and woodlands and altered their composition and density, have inadvertently encouraged subtle changes in grassland vegetation, and perhaps have increased the degree of livestock trailing on the heavily grazed slopes near villages. Their resource-management institutions probably have not regulated the use of regional forest resources within long-term, sustainable levels. Neither resource use nor resource management was explicitly geared as much toward minimizing environmental change as to achieving subsistence goals and minimizing risks. Short-term environmental change was not necessarily a concern if it did not threaten these goals and long-

term environmental changes that might ultimately threaten standards of living (such as declining forest regeneration) were not necessarily perceived or responded to with alterations in land use or institutions. When communities did act to manage local resources it was either to protect sacred places or to meet immediate challenges such as preserving fodder and grass for getting the herds through the winter, maintaining sources of leaf litter and large beams, or guarding against crop disease and wildlife and livestock depredation. Protection of the wild diversity of flora and fauna and native vegetation communities, regional forest and woodland cover, or the environmental quality of rivers, lakes, and streams were not direct concerns except in those spheres of daily life where local religious belief touched land-use practice. This was most notably the Buddhist respect for life that extended to all species not perceived as threats to crops and livestock and a religious respect for springs, trees, and forests believed to be inhabited by spirits. At the same time Khumbu Sherpas have not, during the past century at least, precipitated any known environmental crisis. They have not despoiled Khumbu resources to the point that a deteriorating land base has created regional poverty and spurred large-scale out-migration. Nor have they created conditions fostering unstable slopes and massive erosion.

Khumbu Sherpa traditional resource-management institutions thus had both strengths and limitations even before 1960 when Khumbu land use and the economy had not yet been affected by major tourism development, government nationalization of forests, and the establishment of Sagarmatha National Park. How Sherpas responded to these new pressures and what became of local resource management and the Khumbu environment is the subject of the following chapters.

8
Local Resource Management: Decline and Persistence

One of the greatest changes since 1960 in Sherpa land use and management has been the decline in several parts of the region of traditional resource-management institutions. The abandonment of local forest regulation in some areas during the 1960s has been widely reported. Less well known is the abandonment of the nawa system of pastoral management in Nauje in 1979 and in the Bhote Kosi valley during the 1980s. These events have had great significance for Sherpa subsistence, the Khumbu environment, and current conservation efforts. This chapter examines the change in Khumbu resource-management institutions since 1960. I focus first on the processes and events that led Sherpas to abandon local management institutions in some places and to adapt and maintain them elsewhere, then explore the environmental changes that have accompanied the resulting changes in natural resource use.

Khumbu Forest Nationalization and Its Aftermath

Füer-Haimendorf, revisiting Khumbu in 1971 after fourteen years away, was the first to report that there had been a major change in Khumbu forest use and management. He observed considerable tree felling near Nauje and Khumjung in areas where this had been tightly regulated when he had lived in the region in 1957. Here, he learned, the traditional shinggi nawa regulation of forests had been abandoned because of the government's nationalization of forests. Fürer-Haimendorf subsequently reported that

all forests which are not on privately owned land have been declared state forests and the villagers have no more control over them. Persons who require timber for house building have now to apply to the office of the district panchayat which is at Salleri in Solu, at least four days' walk from Khumbu. The permits issued by that office specify the quantity of timber to be felled. The procedure is cumbersome, and as there are no officials of the forest department in Khumbu, unauthorized fellings cannot be controlled. The forests in the vicinity of villages have already been seriously depleted, and particularly near Namche Bazar [Nauje] whole hill-slopes which were densely forested in 1957 are now bare of tree growth and villagers have to go further and further even to collect dry firewood. In this case the replacement of an efficient and well-tried system of local control by a bureaucratic machinery has not been successful, and the Sherpas are conscious of the diminishment of local timber resources without being able to stop the inroads into forests which, being claimed by government as state property, are no longer under their control. (1975:97-98)

This analysis has become the standard description of the demise of Sherpa forest management in Khumbu and one of the classic Himalayan examples of how national government resource-management efforts can undermine previously effective local resource-management institutions to the detriment of local environment (Byers 1987b; McNeely 1985; Thompson and Warburton 1985:122). Thompson and Warburton, for example, used the Khumbu case as their sole example in a discussion of the shortcomings of Nepal's national bureaucratic resource-management efforts. Their summary of the events in Khumbu drew strongly on Fürer-Haimendorf's observations and analysis in concluding that

in the 1950s the forests were nationalized. Management was taken out of local hands and transferred to central government—a reform which, we know, destroyed the indigenous management system (based on a rotating village office of forest guardian who managed the traditional village forest as a renewable resource) that had worked successfully for centuries and then did not work itself. (1985:122)

Khumbu has in this manner become a classic example of what has been supposed by a number of authors to be a Nepal-wide process through which post-1957 forest nationalization undermined traditional customs of local forest management (Acharya 1984; Arnold and Campbell 1986; Bajracharya 1983; Messerschmidt 1986, 1987:375).

In the past few years this thesis has been increasingly questioned. Foresters and anthropologists working in several parts of Nepal have found that greater familiarity with local oral history suggests that the stereotypical view of the impact of 1950s forest nationalization on local management institutions and Nepal's forests does not hold true in some

areas and that it may be an exaggeration even in national terms (Archarya 1989; Gilmour and R. Fisher 1991:12-13). Gilmour and Fisher, who are very familiar with forest use and management in hill regions near Kathmandu write that:

It is generally believed that widespread indiscriminate cutting of forests occurred during the years following 1957 because village people felt that their forest had been taken away by the government . . . . While this unintended effect of nationalisation has been widely reported, we believe that it has been exaggerated. In the course of fieldwork in Nepal we have obtained a great deal of oral evidence which calls this phenomenon into doubt. . . . We have never heard local claims that there was a crisis in forest management following the nationalisation of forests . . . . Contrary to the view that rural people reacted to the 1957 legislation by destroying forests is the observation that a great number of indigenous forest management systems (which were commonly set up by villagers to protect degrading forests) had their origins about 1960. (1991:13)

Closer oral history examination of Sherpas' response to forest nationalization also raises questions about early reports about the impact of the 1957 legislation on Khumbu forest management and use. On the one hand Sherpas present testimony that supports Fürer-Haimendorf's observations. Forest nationalization did indeed undermine some local Khumbu forest-management institutions and had an adverse impact on forests. Yet only some Khumbu forest-management systems were affected and then only in certain localities: Elsewhere local forest management was maintained. Sherpa institutions were not as fragile as Fürer-Haimendorf and Thompson and Warburton had assumed. From the Khumbu perspective the issue of nationalization and its impact is thus much more complex than either Fürer-Haimendorf and Thompson and Warburton, on the one hand, or Gilmour and R. Fisher, on the other, suggest.

Sherpas today verify that local forest management in several communities was indeed undermined by forest nationalization, although they disagree with some of the details of Fürer-Haimendorf's account and put the date of the abandonment of local management in the 1960s and 1970s rather than the 1950s. They also testify that increased tree felling took place as a result in some forest and woodland areas (particularly near the villages of Nauje and Khumjung) and note that the impact of this was visible in the landscape. They disagree, however, that villagers lost all control over the forests they had previously administered or that widespread deforestation took place throughout the region. Instead they testify that in much of the region they continued to maintain local forest management despite the nationalization of forests. They speak of a far

longer, and still continuing, process of conflict and negotiation between local residents and government officials over the management of forests rather than of a simple, regionwide collapse in the 1950s of Sherpa efforts to maintain their regulation of protected forests. This difference in perception is not just a matter of historical interest, for it has enormous implications for national park management today and in the future. It is therefore worth reviewing the process of the nationalization of Khumbu forests and its effects on local institutions and the environment, paying particular attention to variation within the region.

The Nepal government has long taken an interest in the management of forests, including issuing occasional orders for forest conservation. In the nineteenth century both the Shah and Rana administrations sent out edicts to the hill regions of the country which specified restrictions on tree felling near water sources, along roadsides, and in religious places, and both governments authorized village headmen through numerous directives to regulate local forest and pasture use according to their local customs of management (Acharya 1989: 16, 17; 1990). We have already seen one example of this which had direct significance for Khumbu forest conservation, for Rana regime policies in the early twentieth century resulted in the creation of locally managed protected forests in eastern Nepal, including the eight Khumbu rani ban. After the fall of the Ranas with the 1950-1951 revolution, which restored the monarchy to greater power, the national government began to become much more involved with the nationwide administration of forests as national resources. One of the early measures in this process was the enactment of the Forest Nationalization Act of 1957, which returned to state ownership and management large-scale private forestlands and those areas of common forestlands that had not been already legally recognized as communal lands (kipat ).^[1] In the Khumbu context the Forest Nationalization Act of 1957 reasserted the government's legal authority over virtually all forest and woodlands in the region. Only a few, very small, private forests near Thami Teng were exempt from the act. Yet the passage of this legislation in Kathmandu in 1957 had no immediate impact on Khumbu. No government officials arrived in Khumbu ordering Sherpas to give up their former forms of forest management, and local residents continued to administer rani ban, lama's forests, and other protected forests exactly as they had before.

Khumbu forest use was more greatly affected by the government's continuing efforts to establish nationwide forest management based on a permit system of tree felling. This was to be the responsibility of district Forest Department offices.^[2] Decisions about where tree felling would be allowed, which species could be cut, and on what scale felling could be carried out were to be made by professional foresters and the administra-

tion of these policies in the forests was to be the responsibility of trained government forest rangers. These rangers were to select the particular trees to be felled, collect timber fees, and monitor the operation. The first efforts to. implement this style of government forest management reached Khumbu about 1965, when a branch (sahayak anchyaladhish karyalaya ) of the Sagarmatha zone office, which coordinates the administration of Solu-Khumbu and neighboring districts, was opened in Nauje. The officials of this office were given the responsibility of administering tree-felling permits along with their other duties for several years until district forest offices could be established at Salleri and Lukla. Thus for the first five years of forest nationalization in Khumbu, until about 1970, officials based in Nauje itself rather than in distant Salleri issued permits.^[3] This meant that the officials responsible for administering the new forest rules in Khumbu were not so far removed from the region as Fürer-Haimendorf had thought, and it made the procedure of obtaining permits for tree felling, before 1970 at least, somewhat less cumbersome than he had portrayed.^[4] The permits issued in Nauje specified the area from which the trees were to be cut and the total volume of timber authorized. An important part of the permit process, one hitherto overlooked in the literature, was the requirement that Sherpas first obtain the approval of the pradhan pancha , the leading locally elected Sherpa official.^[5] Only after this had been granted could one apply to the branch of the zone office for a felling permit. The system thus had some degree of local control built into it, a fact unrecognized by Fürer-Haimendorf.^[6]

The Nauje branch office is said by Sherpas to have granted them permission to obtain any amount of timber which they wished, provided that the appropriate fee was paid. And it authorized the felling of trees for timber anywhere in the region, irrespective of earlier Sherpa protected forest boundary lines and regulations.^[7] According to one of the pembu of that time, the Nepali official who implemented the system overrode the objections of Sherpa pembu who confronted him with the documents that had authorized the establishment of the rani ban and demanded recognition that they had legal jurisdiction over Khumbu's forests. The official's reply was that a paper issued by the Rana regime was not binding on the present government.^[8] Some Sherpas say that the government's early interest did not seem to be forest management as much as collecting forest revenue, and they call the Nauje office the "wood-selling office" and say that it "gave away the forests."

Sherpa forest regulation did not automatically collapse throughout the region despite the refusal of national government officials in Nauje to recognize any local rights to forest management. Local forest protection might not have had any legal basis from Kathmandu's perspective, but the local forest-use regulations continued to meet local concerns. In a number

of areas Sherpa villagers decided that even though they could now obtain permits to cut within protected forests they would simply not apply for them. Instead they would continue to administer their own regulations in their own forests. Phurtse villagers continued to choose shinggi nawa to administer sacred forests and Pare residents continued to select shinggi nawa for the Bhotego rani ban. In Pangboche the Yarin forest continued to be watched over by the temple caretaker and the nawa continued to enforce the bans on bringing freshly cut wood into Pangboche or Dingboche during the summer. Protection was maintained especially well for religiously safeguarded forests, and temple and lama's forests were very seldom transgressed even in places such as Khumjung which soon did abandon its other traditions of local forest protection. Local forest management continued to be enforced in all of these areas other than Yarin until the establishment of Sagarmatha National Park, when some communities reconsidered the relevancy of their forest management.

In Nauje, Khumjung, and Kunde, however, a different process took place. Here rani ban regulation was undermined. In these three settlements it became increasingly difficult to administer the old rules after the permit system was inaugurated. Villagers from these settlements applied for permits to cut in formerly protected forests and defied shinggi nawa and nawa by maintaining that they had obtained both the permission of the locally elected pradhan pancha and the representative of the Kathmandu government. Here people began to cut wood where it was easiest to get, which was often in the rani ban where timber was relatively plentiful close to the villages precisely because it had been so carefully safeguarded before.^[9] Within a year Nauje nawa gave up efforts to enforce any of the old regulations, feeling that it was useless since, as Sonam Hishi observed, few people were paying any attention to the nawa and instead took advantage of the new system to obtain timber more easily than before by felling trees that the nawa would not allow them to cut. In Khumjung and Kunde there was an attempt to keep the system going, and locally elected village panchayat officials were given the shinggi nawa positions. But those villagers who had permits to fell trees were nevertheless allowed to fell them, and by the early 1970s both Khumjung and Kunde had ceased choosing shinggi nawa.

Clearly in some villages a process of abandonment of local regulation took place which did not occur elsewhere. Several factors might have been important. One could have been that in these three villages, all within a short walk of the Nauje branch office, the new forest-use regulations might have been more vigorously promoted and enforced. In Khumjung the fact that powerful and wealthy villagers, including pembu, took an early lead in felling trees in the former rani ban was probably a more important factor. Here the pradhan pancha himself

encouraged such tree felling and approved permit applications in the rani ban, with special dispatch it is said, for friends and relatives.

The role of the Khumbu pradhan pancha of the 1960s and early 1970s in undermining forest management in their home villages of Nauje and Khumjung has not previously been explored. These men were certainly placed in an awkward position and it is understandable that they did not continue to confront central government officials in Nauje over forest ownership and management. Yet they could have nevertheless supported local management by simply refusing to forward permit applications that requested permission to fell trees inside sacred forests or asked for trees from inside rani ban for purposes other than house beams. In these cases, where Sherpa rules were stricter than the new government ones, Sherpa rules could have been supported. Shinggi nawa and other Sherpa officials could have continued to enforce Sherpa rules in Sherpa protected forests. At the same time government objectives would have been met. Tree felling would have been put on a permit basis and the central government would have received revenues on all timber cut in the area.

Khumbu pradhan panchas, however, apparently did not make any use of their role in the permit system to uphold traditional rules in rani ban. At least one pradhan pancha, however, made use of these new responsibilities to play out old contests of power in new arenas through denying permission to cut trees in certain areas near settlements (including former rani bans) to enemies while promoting the applications of relatives and friends. In such struggles pride could win out over a sense of community and environmental responsibility. One former pembu who had been noted for his zealous enforcement of Khumbu rani ban regulations recalls triumphantly a victory over an old rival who had been elected pradhan pancha which came at the expense of the formerly protected forest at Khumjung. After he had been denied permission for his permit to cut timber to rebuild his house he waited until the pradhan was away from the area and then obtained the permit from a subordinate official. He cut his timber from the slopes just above the village, an area in which trees have long been scarce and that had been within the rani ban.

The undermining of local management regulations and institutions could, of course, have been easily avoided if the new approach to national forest management had been implemented differently. A central government directive that preexisting local regulations were to be supported when they did not conflict with the goals of Forest Department regulations would, for example, have made a considerable difference in Khumbu. But Kathmandu did not send out such instructions and the local officials in Nauje who implemented the new system showed little interest in initiating such an approach themselves. Their efforts seem instead to

have been primarily aimed at the establishment of the government's claim to forest land and its sole power over forest regulation. There were occasional exceptions. One case is remembered, for example, when the branch office used its veto power over logging-permit applications to protect trees in a previously locally protected area. In this instance government officials denied a Nauje villager permission to cut trees in one area of the former rani ban on the grounds that it was necessary to conserve these trees for future use in bridge repairs. Such action by a central government officer, however, seems to have been highly unusual.

In other communities Sherpas integrated traditional, local institutions with the new permit procedures. In these places both local and national institutions were in place simultaneously for several years. Villagers obeyed the stricter local regulations maintained by local values, social pressure, and fines. At the same time they followed the new procedures of the national system, despite their redundancy and costs in time, trouble, and cash which the government enforced with the threat of arrest, prosecution, fines, and possible imprisonment. It was possible to continue to obey the local rules and to comply with the new Kathmandu bureaucratic regulations as well. In effect this established a two-tier system of forest regulation. A particular forest use first had to be locally legitimized through the local institutional procedures. Only then could one take the locally sanctioned request on to the government for it to certify to its satisfaction. In other communities no such two-tier system developed, and instead some individuals used the existence of the national institution to circumvent the local one. That the two-tier system did not develop in the 1960s in Nauje, Khumjung, and Kunde could be related to the fact that powerful local individuals, including pradhan pancha in some cases, took the lead in subverting initial community efforts to establish this kind of parallel set of institutions. The location of the national government office in Nauje itself—close at hand also to Kunde and Khumjung—moreover enabled villagers to appeal to the primacy of central government regulations as a justification for violating community customs. The prospect of ready support from Nauje government officials must have enhanced the ability of Sherpas who wanted to violate local customs to negotiate their actions with shinggi nawa, nawa, and their fellow villagers.

Forest Nationalization and Deforestation

It has been widely reported that the nationalization of Khumbu forests led to rapid and widespread deforestation prior to the establishment of Sagarmatha National Park. Fürer-Haimendorf's 1972 observations that

"forests in the vicinity of villages have already been seriously depleted" and that "near Namche Bazar whole hill-slopes which were densely forested in 1957 are now bare of tree growth" (1975:98) often have been cited and assumed to have been an accurate accessment of conditions.^[10] Ecologists, foresters, and park planners who visited the area in the early 1970s also reported a deforestation problem (Lucas, Hardie, and Hodder 1974; Mishra 1973) and this perception was one of the major factors that lent urgency to the proposal of national park status for the region and shaped early national park resource-management policy.

Sherpas do not agree that widespread deforestation took place between the time the government forest-use regulations were implemented in 1965 and the establishment of Sagarmatha National Park in 1976. They contend that descriptions of entire recently deforested slopes are exaggerated. They suggest instead that forest clearing was very localized. In contrast to the vague general impressions of extensive forest change offered by outside visitors their oral history accounts provide detailed insights into perceived change in particular forests and woodlands. Oral history also offers insights into the processes that Sherpas believe precipitated these forest and woodland changes and their assessment of the degree to which forest nationalization was the cause.

Sherpa testimony about forest change since 1960 indicates that in some areas small-scale clearing occurred prior to the establishment of the national park. In several other areas changes in forest density and composition became increasingly evident even though there was no major change in forest or woodland cover. Most of the areas where thinning or clearing took place are said to reflect tree felling for timber for house building, although vegetation change at a few sites (including some areas near Nauje on the slopes north of the Dudh Kosi) are thought to have primarily reflected the cutting of trees for fuel wood. The table below identifies sites of reported forest change and the types of forest use reported in those areas; these are shown in map 20.^[11]

Most of the sites of forest change between 1960 and 1976 are in areas which were not traditionally protected forests. A few, however, reflect the abandonment after 1965 of earlier local restrictions on timber cutting. The most outstanding examples of forest change that were directly related to forest nationalization were the result of tree cutting in parts of what had been the Nauje and Khumjung-Kunde rani ban. This is presumably the logging that led Fürer-Haimendorf (1975:98) to remark, perhaps with some exaggeration, that considerable deforestation had taken place adjacent to Nauje between 1957 and 1971 and that forest had also been depleted near other villages. Although forest clearing near Nauje was not nearly as extensive as his account suggests, Sherpas testify to localized change after 1965 in two areas adjacent to the village:

Map 20.
Reported Forest Change, 1965-1979.

the slopes just to the south of the settlement at Namdakserwa Nasa and the slope immediately below and west of the village. Both of these areas were previously within the Nauje rani ban. Villagers began applying for permits to cut house timber as well as beams in these areas soon after the Nauje branch of the Sagarmatha zone office implemented the new government forest-management system, even though there was far from unanimity in Nauje that such tree felling was a good idea. As far as I have heard no requests for cutting in these areas were turned down by central government officials or by the local pradhan pancha, and several Nauje families who built or rebuilt houses between 1965 and 1976 can trace their timber to trees cut in these two areas. The small platforms that were dug in the slope to support board-sawing stations can still be seen. The region below Namdakserwa Nasa, which is said to have been so densely forested that it was difficult to gather leaves there while carrying a basket, is now considerably more open, and today three soluk gatherers moving abreast would have no trouble negotiating this area. Sherpas testify, nonetheless, that there was no major change in tree cover on the slopes above Nauje during this period. Trees here, they say, had already long since been cut except for a few juniper that surrounded the site of a meditation retreat high above the village. These juniper were cut during the 1960s after the resident of the hermitage passed away.

In the late 1960s the slopes above Khumjung and Kunde also became an important timber source following the relaxation of rani ban regulation. Trees here had previously been protected against felling other than for use as beams. A large number of junipers were felled above Khumjung where today few trees of any stature are left.^[12] The Khumjung slope had not been densely forested before 1965,^[13] but people do recall some stands of denser, taller juniper thirty years ago as well as more scattered juniper. One of these denser stands was high on the slope in the vicinity of a hermitage. Another was situated lower on the slope and closer to Kunde village near where local Tibetan immigrants now maintain a prominent shrine. There are no large junipers in the latter site today and the trees around the hermitage are said to be fewer. Sherpas note that a great number of trees were cut in a long-regulated area and they no longer refer to the slopes above Khumjung as being forested.^[14]

Other forest change during the 1960s and 1970s reflected continued cutting in areas that were outside the boundaries of Sherpa protected forests. Nauje and Khumjung residents, for example, recall that during this period there was much fuel-wood collection from several areas along the north slope of the lower Dudh Kosi above its confluence with the Bhote Khosi, including the Kenzuma area and slopes closer to Nauje. This is said to have led to a decrease in the tree cover of the open woodland there. Nauje villagers mention another area near Tesho and Samshing on the east side of the Bhote Kosi where the felling of fir for house rafters led to some small-scale clearing in the 1960s.

Some of the vegetation changes reported by Sherpas in the 1960s and 1970s can be attributed neither to the nationalization of forests nor to long-established Sherpa forest-use and management patterns. At Shyangboche, for example, considerable change in juniper woodland cover took place within a rani ban regulated area under extraordinary circumstances unrelated to the operation of the Nauje branch of the Sagarmatha zone office. Before the early 1960s, according to local memory, parts of Shyangboche were places of fifteen-foot-high junipers which were dense enough that it was easy for herders to loose track of yak there on foggy days. Now it is very open woodland and large areas have very few tree-sized junipers. Some of the juniper was cut by Tibetan refugees who camped in the area between 1959 and 1962 and some was cut during the construction in the early 1970s of the Everest View Hotel and its airstrip. Most, however, was cut by Sherpas during the 1960s following a land dispute. The area had been coveted by Nauje villagers for some time on account of its many junipers and rich pasture and, as mentioned previously, there had been arguments over it for years. In the 1960s Nauje villagers increased their timber and fuel-wood gathering in the area and

one Nauje family tried to file a land claim there with district officials. This led to more widespread cutting in response by Khumjung and Kunde families, which in turn spurred more cutting as Nauje families who were determined to claim a share of any remaining Shyangboche juniper vied for resources.

Other special cases of forest change resulted from outsiders' violations of protected-forest regulations or from their being granted exemptions from the usual rules. Phurtse villagers, for example, say that they began felling trees for house building from the area that had been protected as a bridge forest after the village had decided to cut in that area for timber with which to build the Hillary school in the early 1960s. Thamicho villagers insist that they only began lopping and felling trees in the Bachangchang rani ban after Nepali members of the local police post began cutting trees there for fuel wood.

New Demands on Forests

Khumbu demand for forest products was changing at the same time as some Sherpa communities were loosening their control of forest management. During the 1960s and 1970s several new demands were placed on regional forest resources as a result of changes in subsistence forest use, the requirements of the growing tourist trade and the increasing numbers of Nepali officials and personnel stationed in the region, and the establishment of Sagarmatha National Park.

Increased Sherpa use of fuel wood, possibly the most widespread and significant of these changes, was related to a change in lifestyle. Until the mid-1960s many Sherpa families spent much of the winter away from Khumbu on trading expeditions and pilgrimages to lower-altitude regions as distant as northern India. It was common for families to set out from Khumbu in November and to be away for most of the winter, returning in March to begin preparations for spring planting. Families who had livestock either sent these to Tibet or left them in Khumbu in the care of a family member or a relative. A substantial part of the regional population thus spent four or five months outside of Khumbu, and some men were away as much as ten months out of the year on more extended trading journeys in Tibet and India.^[15] Food was available more cheaply in the south, and the more time families spent subsisting there on grain grown at lower altitudes the less grain they needed to carry back on foot. This changed when Sherpas ceased barter trading salt for grain in the lower-altitude regions of Nepal. The declining availability of Tibetan salt in the early 1960s and the establishment a few years later of the weekly periodic market system in Solu-Khumbu rapidly undermined the former practice of winter trading in the south. More Sherpas began

to spend the winter at home in Khumbu unless trekking work took them elsewhere.^[16]

This change in life-style had a significant effect on forest use as well as prompting adjustments in agricultural practices.^[17] A side benefit of the old way of life was that families were away from home during precisely the part of the year when their fuel-wood requirements would have been the highest. After 1965 when families began spending the entire year in Khumbu their fuel-wood requirements approximately doubled. A family that burned half a load of fuel wood per day in the winter would have found it necessary to obtain as many as seventy-five more loads (2.25 metric tons at 30kg per load) of fuel wood per year. If this family lived in an area where a half a day's work was required to collect a single load of wood, not an uncommon situation at that time, gathering a winter's supply of fuel wood would have required thirty-seven person days of hard labor.

The change in Khumbu seasonal settlement patterns in the 1960s could thus have created a huge new demand for fuel wood which was met from the forests and woodlands closest to the main villages. To this increased local demand for fuel wood were added several new sources of demand. Fuel wood was required by the increasing number of lowland Nepalis who began to be stationed in Khumbu during the 1960s and 1970s to staff army, police, and medical posts, schools, the post office and bank, government offices, the government yak farm, the airstrip at Shyangboche, and the national park. Tibetan refugees had significantly increased fuel-wood demands in the region for a few years during the early 1960s. Increasing numbers of mountaineering expeditions during the first half of the 1960s and the beginning of commercial trekking in 1964 also began to put new pressure on the fuel-wood resources of some areas, particularly along the main route to Mount Everest. Tourism has been widely reported to have had an especially severe impact on Khumbu forests. This issue deserves separate treatment, for many of these reports have considerably exaggerated and overgeneralized recent deforestation while overlooking tourism's role in more subtle vegetation change in the region. This will be discussed in detail in chapter 10 along with other impacts of tourism on Khumbu land use and environment.

One effect of these changes in fuel-wood demand and the presumed increasing scarcity of dead wood near main villages that accompanied it, may have been a new Sherpa concern with using fuel wood more frugally. Faced with a choice either of devoting considerably more time to fuel-wood gathering or economizing on wood use, it appears that many households became interested in the conservation of fuel wood.^[18] It is precisely at this time that more fuel-efficient means of cooking began to be widely adopted. Until the mid-1960s virtually all families cooked over open fires,

either setting the cooking pot on a tripod of standing stones or on a simple iron grate. Now many families began to install cheap and efficient stone and mud wood-burning stoves, copying an old Tibetan design. According to one estimate this type of stove increased the efficiency of fuel wood by five to thirteen times over simply cooking on an open fire (M.N. Sherpa "Conservation for Survival: A Conservation Strategy for Resource Self-Sufficiency in the Khumbu Region of Nepal," 1985, cited by Fisher 1990:64). Sherpas must have long been aware of this type of stove from their trading and pilgrimages in Tibet, and the several Khumbu families who owned houses in Ganggar may well have had them in their homes there.^[19] During the 1970s some families also began experimenting with other fuel-saving measures including the use of pressure cookers, which many families now use routinely. In the 1960s large Chinese thermoses also became popular as a way of avoiding having to rekindle fires frequently during the day to brew more tea—a significant task in a region where tea is the main nonalcoholic beverage consumed and where it is common for adults to drink dozens of cups per day. More recently there has been considerable interest in a new stove design that adds a simple water-circulation and reservoir system (backboiler heater) to the wood-burning stove so that it heats water at the same time as it supplies energy for cooking. These conservation measures have been widely adopted despite the many beliefs about the sancity of the hearth and fire, both of which are inhabited by spirits who must be treated with special care since they are believed to affect the health and welfare of the household. There have also been recent changes in vernacular architecture that reflect a concern with the conservation of fuel. Until recently the interior design of Khumbu Sherpa houses had changed little for a number of generations, but today there is a marked trend in several villages towards remodeling older houses or building new ones in which the old style single, large, open-beamed room in which the hearth occupies a central place along one wall has been replaced with smaller, separate rooms, a wood stove in a separate kitchen, and a ceiling.^[20]

Sagarmatha National Park and Forest Regulation

In the late 1960s and early 1970s there were increasing pressures on Khumbu forests from both new levels of Sherpa demand for fuel wood and timber and from tourism. In some areas, at least, local forest management had been abandoned. Highly visible tree felling was taking place in the Khumjung and especially in the Nauje area on a scale very different from previous decades. These changes troubled a number of visitors to the Khumbu region in the early 1970s, among them a series of

government officials and advisors who had come to the area to evaluate its potential as a national park. These observers reported that deforestation was a serious regional environmental threat (Blower 1971; Lucas, Hardie, and Hodder 1974; Mishra 1973).

As early as 1971 John Blower, an official from the United Nation's Food and Agricultural Organization involved in the development of conservation and park programs in Nepal, urged the establishment of a national park in the Mount Everest area. Subsequent park-planning visits were made by Hemanta Mishra, a Nepalese ecologist, in 1972 and in 1974 by a New Zealand team sent to evaluate New Zealand's participation in the possible establishment of a national park. New Zealand, the homeland of Sir Edmund Hillary, ultimately signed a bilateral agreement with Nepal in which it took responsibility for training several Sherpas to be park administrators and offered to provide park-management personnel as advisors during the early years of the park's operation. Sagarmatha National Park was formally gazetted by the government of Nepal in 1976. For the first five years of its operation a series of New Zealand park managers, assisted by Nepalis from the Department of National Parks and Wildlife Conservation, oversaw the establishment of the national park headquarters complex on Mendelphu hill adjacent to Nauje, prepared the first park-management plan, and implemented new resource-use regulations.

The New Zealanders endeavored to make Sagarmatha a world-class national park and launched an ambitious program in which both the protection of the environment and concern for Sherpa welfare were central. Real efforts were made to consult with Sherpas about the design and management of the park. Village meetings, for example, were held as early as 1974 when the New Zealand planning mission drew up its recommendations for the establishment of the park. A local advisory committee was set up. And through New Zealand's efforts the first Khnmbu Sherpas joined the Department of National Parks and Wildlife Conservation, completed university training in conservation and park management, and were stationed in national parks as administrators. This far-reaching concern for local involvement in future park management resulted in several Sherpas acting as chief or assistant administrators in Sagarmatha National Park for most of the 1980s.

Forests were an early concern. Nurseries were begun in 1979 to supply seedlings of indigenous conifers for a reforestation program. Plantations were established at Mendelphu, Khumjung, and Kunde, each protected either by stone walls or by barbed wire fencing. Furthermore, forest use throughout the region was placed under a system of regulation intended to curtail the adverse impact on forests of both local subsistence use and tourism. Regulations were implemented which banned fuel-wood use by

mountaineering and trekking groups and new rules were enforced which outlawed much traditional Sherpa use of Khumbu forests.

Sherpas were suspicious of the idea of their home becoming a national park, as they made clear in the 1974 village meetings held by Nepali and New Zealand park planners to broach the concept of the park and solicit local input. Sherpas accused Hillary, who had been one of the main proponents of the park, of betraying them. Many Sherpas did not share the New Zealanders' vision of the national park as a way to buffer and ameliorate the adverse impact of tourism on both the local environment and culture, and were afraid that the establishment of the park would be a disaster for their way of life. Many feared that the entire population of the region would ultimately be evicted and resettled elsewhere in Nepal in order to make Khumbu a wilderness area. This fear was not entirely groundless, for this is what had already been done in Nepal's only previously established national park, Chitwan, and it was soon to occur again at Rara National Park. There were fears that even if Sherpas were allowed to remain in their homeland that park authorities would implement new regulations that would prevent them from continuing with the subsistence practices on which their settlement of the region had relied. Many Sherpas were worried, for example, that national park administrators would introduce new concerns about overgrazing into regional pastoral management and with them restrict the size of livestock herds and regulate access to pastures. Here too their concerns were not entirely misplaced, for the first national park management plan was soon to advocate these measures. And there was a widespread fear that the park would ban cutting trees even for subsistence purposes. Anticipating the establishment of the park, many Sherpas put up large stockpiles of fuel wood and timber while there was still the chance. Tree felling for both timber and fuel wood may thus have been at extraordinary high levels from the time the news of the coming establishment of the national park reached Khumbu in 1973 or 1974 until the late 1970s when the new forest regulations Sherpas had feared were indeed imposed.

The early management policies developed at Sagarmatha National Park were undoubtedly intended to be responsive to local Sherpa concerns as well as international environmental ones. An attempt was made to incorporate local consultation in park management in a way uncommon in that era in developing countries. By legally excluding the villages and herding settlements from the park boundaries, national park planners acknowledged Khumbu Sherpas' rights to remain in their homeland and choose for themselves their own paths of village development, a gesture that reflected a degree of sensitivity to the rights of indigenous peoples which was then extremely rare in the world national park movement.^[21] Park administrators also recognized the importance of support

for Sherpa subsistence lifestyles and culture even when these required use of natural resources located inside the park—allowing grazing, for example, even though domestic stock may have competed with the Himalayan tahr for forage and allowing trees and shrubs to be cut for fuel for cremations and ceremonial chotar and gotar poles. Ultimately park authorities decided against imposing any new grazing regulations in the region and did not even step into the management vacuum created by the abandonment of the nawa-system regulations in Nauje in 1979. But the national park did take a stand on forest use and it is this stand that so greatly colored Sherpa attitudes about the park in general for many years. Things might have gone differently if new forest regulations had been implemented more gradually and with a stronger foundation established first through programs designed to build good relations with local communities, develop local conservation education, win support from local political and religious leaders, evaluate current patterns of forest use and management, and find ways to integrate new conservation concerns with older, local forest-management institutions. But there was a perception that immediate action was vital.

After the establishment of Sagarmatha National Park, New Zealand and Nepali administrators instituted new forest-use regulations that were the strictest in Khumbu history and represented a fundamental change in resource-management goals. These were enforced by patrolling Khumbu with a Nepalese army "protection unit."^[22] All tree felling was prohibited and lopping branches for fuel wood or fodder was banned. Only dead wood from the forest floor was to be gathered for fuel wood and all timber was to be brought in from outside the national park. In Nauje a permit system to regulate dead wood gathering was introduced on a trial basis. Free passes were given to gather wood in particular areas on particular days. This system of regulating dead wood collection was intended to rotate the gathering of dead wood and leaves from the forest floor in order to avoid overuse of the most accessible areas. Some Sherpas, however, feared that national park authorities intended to begin charging a fee for the passes and thus make gathering even dead wood costly as well as more difficult.

The approach that Sagarmatha National Park adopted for Khumbu forest management, unlike that followed for the management of grazing on park lands, thus adhered to the strict nature-protection philosophy that goes back to the establishment of the world's first national park and which is often referred to today as the "Yellowstone model." Here the idea is preservation uncompromised by human use, or at least uncompromised by commercial and subsistence land use other than tourism development, for the Yellowstone legacy also gave world national park planning a precedent that tourist access must be safeguarded and enhanced

(Stevens 1986a). This approach to protected area management led to the resettlement of indigenous peoples from newly created national parks in many parts of the world and greatly influenced early national park planning throughout Nepal. Sagarmatha, like Langtang National Park, was on one level an exception in that efforts were made to devise park planning that protected traditional settlement and land use. But in forest-use policy a rather different line was drawn. Here the national park directly confronted Sherpas over their traditional access to an important subsistence resource and assaulted their sense of land ownership and homeland. The unfortunate result of this clash of perspectives was continuing resentment and fear and a local lack of support for the national park which has remained widespread ever since.

The new park forest policies evoked anger and local protest. Park authorities, perhaps taken aback by the strength of this reaction, responded with efforts to meet local concerns at least part way. The permit system for dead wood collection was abandoned and there was some relaxation of the tree-felling prohibition. Under the new rules families were to be allowed to fell three trees for use as beams when building new houses. This was not entirely satisfactory to many Sherpas, some of whom noted that three trees were not sufficient even to provide all the necessary house beams, much less to compensate them for the expense and trouble to which they were now faced to go in order to have all other timber cut in Pharak and transported to Khumbu.^[23] Some saw this as no better than declaring all of Khumbu a rani ban in which one was entitled to fell even fewer trees than the rani ban system had allowed.

Over the years an uneasy truce evolved. Sherpas came to accept the existence of the park, but many bitterly complained that all of Khumbu had been turned into a protected forest at the cost to Sherpas of making their lives more difficult. Some could see no reason for this or for the establishment of the Sagarmatha National Park forest plantations other than creating more forested scenery for the enjoyment of foreign tourists. Anger and bitterness were further fueled with fear when the army unit began to patrol forests and arrested Sherpas for "illegally" obtaining fuel wood and construction timber even though they were obtaining them in traditional ways for traditional uses from areas where they were traditionally procured. This kind of confrontation continued through the 1980s. The distrust and hostility created during the first few years of national park administration over forest use have handicapped other park efforts including its reforestation program. Many Sherpas suspect that the land walled in for reforestation is land permanently taken away from Sherpas, and some people breach walls or barbed wire fences in order to allow their stock to graze in the former rangeland.^[24] Few Sherpas seem to believe that they will ever obtain much benefit from the

reforested areas. They cannot be blamed for having reached this conclusion, since ten years after the plantations' establishment they have not yet obtained any natural resources from them and have not yet even been allowed to cut the grass for fodder. There has been no indication from the national park that Sherpas will ever be allowed to obtain fuel wood or timber from the mature plantations.

National park forest regulations and particularly their enforcement by soldiers were widely regarded by Sherpas as having "taken away" their forests and their responsibility for their management. Forests began to be referred to as "the national park's forests" and villagers felt that park planners and administrators were oblivious to their continuing efforts to maintain local forest-management institutions. Some park officials may indeed have believed that new approaches to Khumbu resource management were necessary and that existing Sherpa institutions were inadequate to meet changing conditions and new challenges. But early New Zealand park planners (Lucas, Hardie, and Hodder 1974:15) and administrators as well as the New Zealand-trained Sherpa administrators who succeeded them after 1981 have also continued to affirm that Sherpas must have a real role in national park management and that this is vital to its success. Many Sherpas have overlooked these good intentions, however, just as many have misunderstood the intent of conservation policies. A greater dialog between the park and its people is sorely needed.^[25]

After the establishment of Sagarmatha National Park some of the communities that had managed to maintain effective local forest administration despite forest nationalization began to lose their ability to continue to enforce their own regulations. Soon after the establishment of the national park the residents of Pare decided to stop shinggi nawa administration of the Bhotego rani ban. Shinggi nawa continued to be chosen in Phurtse, but a village controversy erupted over whether national park regulations or traditional rules should be enforced. Some people began to lop branches for fuel wood from the lama's forest to the north of the village.^[26] When confronted by angry fellow villagers they maintained that this was permitted under national park regulations. The controversy over the protection of the sacred forest was further increased by a respected Sherpa from Kunde who, when consulted about the situation, mocked the tight traditional control at Phurtse and asked villagers if they were just saving the forest to provide wood for their cremations. In other areas Sherpas complained that the national park was insensitive to local protected forest regulations, since the authorities permitted trees for beams to be cut even in areas such as the Yarin lama's forest where this was not allowed under local forest regulations. There were also charges that the national park gave permits to people

who had not first obtained approval from shinggi nawa and other local officials.^[27]

Sagarmatha National Park forest-use regulations instituted in the late 1970s remain in force today and have been enforced since 1981 by a series of three Khumbu Sherpa administrators as well as several Nepalis. These Sherpas were trained in New Zealand in national park management and have come to support many "Western" attitudes about resource management. They have, however, introduced significant new forest-use management approaches that are based on Sherpa shinggi nawa traditions. Such measures represent creative attempts to promote conservation and a synthesis of Western and Sherpa concepts of resource management by working with local residents and traditions. This may in time help to defuse tensions, but issues of resource and land rights will undoubtedly long remain a source of distrust between Sherpas and the outside authority of the national government and Sagarmatha National Park administrators.

Recent Forest Use and Management

Since the end of the 1970s Sagarmatha National Park regulations have greatly decreased the felling of trees in Khumbu for timber and other purposes and have also shifted patterns of fuel-wood procurement away from forests and woodlands adjacent to villages by requiring that only dead wood be used and by banning trees from being felled. For more than a decade all Khumbu timber other than a limited number of beams has come from outside of the region and new areas for village fuel-wood collection have developed at sites which in some cases require a full day's journey to obtain a single load of fuel wood. Nauje, Khumjung, and Kunde villagers have experienced the greatest change in their fuel-wood gathering. Before Sagarmatha National Park forest regulations were implemented the residents of these three settlements obtained fuel wood from relatively near forest and woodland and were able to fell trees for fuel wood (outside of the rani ban) rather than depend solely on dead wood. Most families obtained adequate household supplies from areas that were no more than half an hour from their houses. This changed in the 1970s with the introduction of new attitudes about forest use and preservation. With tree felling (and even lopping branches) banned, demand increased for dead wood to a point where nearby areas were picked clean of fallen branches. It became necessary to go further and further afield in order to find adequate supplies of fuel wood.^[28] The main fuel-wood gathering areas of Nauje villagers shifted from forest and woodland north of the confluence of the Bhote Kosi and Dudh Kosi rivers to the

western shore of the Bhote Kosi (especially the Satarma area) and the slopes of Tamserku on the far side of the Dudh Kosi. Khumjung and Kunde villagers began regularly making expeditions across the Dudh Kosi (using the bridge at Phunkitenga) to the Nakdingog and Chuar areas. These areas are shown in map 18. Most Nauje villagers today obtain fuel wood by hiking through the fir and pine forests near the village, descending three hundred meters to the Dudh Kosi, crossing the river on flimsy "woodcutter's bridges," hiking roughly another three hundred meters up the far slope, gathering fuel wood, and then carrying thirty- to forty-kilogram loads back down to the river and up the steep climb to the village. Many hours are spent getting to and from the fuel-wood gathering area whereas only a few minutes are actually taken to gather a full load at the site. Khumjung and Kunde villagers leave very early in the morning for the long trip to Nakdingog and Chuar. Travel time has also been increased for residents of the other villages.^[29]

Most Sherpas obey national park regulations despite the inconvenience required to respect them. These regulations have without question radically altered local forest-use patterns and in general established a new level of forest protection in the region. Some people, however, revert to more traditional resource-use patterns when these are not likely to be observed. A good deal of fuel wood is burned which is not dead wood. Much of this is obtained by lopping branches and occasionally felling small trees in areas far from the villages and national park headquarters. But at dusk and at dawn one also comes across people carrying loads of green wood from the vicinity of settlements, most often elderly people or young children for whom the expedition to the now-distant sources of dead wood would be difficult or impossible. Trees are sometimes felled surreptitiously for building material. Other people abuse permits. I have known a case where a man who claimed that he had permission to fell three trees for beams had instead cut at least ten times that number in the Dole area from one of the highest-altitude forest regions, and a Nauje case where a man who had been authorized to fell three trees for beams had ended up with fourteen—the result, it was remarked wryly in the village, of these three having conveniently toppled the others when they fell. Still more ironically, in both cases much of the lumber thus obtained was not used for main beams at all but rather for joists and rafters. In one case the operation was carried out on a foggy day in a remote corner of the region, but the other took place quite openly with the trees hauled into the center of the village and sawn into joists on the main street on the day of the weekly market. These incidents as well as another much-discussed local incident in which a national park ranger apprehended a man in the act of illegally felling a large juniper only to have his superior release the unfined offender and

return to him his axe and the firewood suggested to some Sherpas that the same national park rules did not apply to all Khumbu residents equally.

In some cases villagers have become concerned that the national park rules have not worked well enough. This is especially so when formerly tight local control of sacred forests or rani ban has eroded. In Kunde dissatisfaction with national park protection of the rani ban adjacent to the village led villagers to resume choosing shinggi nawa in 1979 and to double the number of guardians to four. In 1984 shinggi nawa were chosen once again in Khumjung and Pare for the first time in many years to watch over areas that had formerly been rani ban, and residents of the settlements of Mingbo and Chosero chose shinggi nawa to oversee the nearby lama's forest. In all of these cases the impetus to community action was a perception that tree felling had reached unacceptable levels and that self-policing by villagers would be more effective than national park regulations and infrequent army patrols.

During the past decade several Sagarmatha National Park administrators reached a similar conclusion. In 1983 Mingma Norbu introduced a national park-sponsored, Khumbu-wide system in which two village-elected shinggi nawa would be responsible in each settlement for keeping an eye on local forests. These guardians would be paid a small salary (100 rupees per month) by the national park and would enforce Sagarmatha National Park regulations, turning offenders over to the national park authorities for judgement. The park would also keep any fines collected. This new system built on the traditions of forest management of Khumjung, Kunde, Phurtse, and Pare, extending the shinggi nawa approach to forest management to all Khumbu communities for the first time. Yet while drawing on indigenous forms of management it did not in any way reverse the shift in decision making about how forests were to be administered which had passed from local communities to the central government during the 1960s. The new shinggi nawa were in effect nothing more than employees of Sagarmatha National Park. Perhaps because of this the approach did not have universal support and its early effectiveness in many communities was limited. It represented, however, the first attempt by government officials since forest nationalization to work with local people and to build on local institutions in developing regional resource management. In some places there were notable successes in building greater local support for forest protection. During 1986 and 1987, for example, Mingma Norbu's successor Lhakpa Norbu appointed young monks from the Thami monastery to be shinggi nawa in the Thamicho area. These new shinggi nawa took up their responsibilities with considerable enthusiasm and several times even apprehended police-post woodcutters at work in the former rani ban of

Bachangchang near Thami Og. Lhakpa Norbu also responded to local appeals to grant shinggi nawa more independent authority and to give their forest-protection role more standing in the community. The forest guardians were given the power to issue and collect fines, and the money from these fines could now be put to village projects rather than being turned over to national park authorities. Lhakpa Norbu also called a major meeting of Sherpa leaders and shinggi nawa from all over Khumbu which did much to enhance the pride and stature of the shinggi nawa.

In the past few years there have been more far-reaching changes in relations between Sherpas and the national park which have had ramifications for local resource management issues. During the late 1980s long-standing local unhappiness with national park goals and policies and dissatisfaction with the personal administrative styles of two consecutive Sagarmatha chief administrators (the first a locally born Sherpa and the second a non-Khumbu Nepali) precipitated local demands for their removal. This culminated in a 1989 petition to the Kathmandu-based director general of the Department of National Parks and Wildlife Conservation to transfer the Nepali administrator from Sagarmatha National Park. This petition was the first attempt by Khumbu Sherpas to mobilize public pressure in this way to influence park policy and administration. The document contained a long list of grievances, some of them complaints about the stringent enforcement of national park regulations, others complaints about unjust treatment. Perhaps the most compelling charge was that the army and national park rangers had harassed village woodcutters who were not guilty of having cut green wood and had confiscated their fuel wood and cutting implements. The petition demanded action within a month. When the park administrator was subsequently transferred to another assignment this was interpreted as a victory for local people.

A new chief warden, Surya Bahadur Pandey, arrived in Khumbu in the spring of 1990. That spring was a momentous time in Nepal. Mass demonstrations in Kathmandu and many other parts of the country were being staged which resulted in the end of the panchayat form of government and redefined the role of the king in Nepal's politics. A significant politicization of Nepali society took place through widespread grassroots political organization and street demonstrations, and with the April legalization of political parties and the acceptance by King Birendra of demands for a new constitution, a general election, and a constitutional monarchy, expectations were raised about a new level of local participation in decision making and local authority. In Khumbu, as in the rest of the country, the local panchayat form of government was dissolved in the spring, and two multisettlement village development committees

(one for each of the two former "village" panchayats) were established to handle vital administrative functions until a new national form of local government could be adopted. The new atmosphere also led some Khumbu leaders to believe that the time was right to press again for recognition of Sherpa rights to land and resource management that had been lost since the mid-1960s. When the new warden of Sagarmatha National Park called a meeting of Khumbu leaders in the summer of 1990 to discuss forest management and the establishment of a new local forest-management advisory committee, Konchok Chombi of Khumjung took the podium and delivered an angry demand that the Khumbu forests be returned to Sherpa control and that greater access to forest resources be given to villagers.

Pandey was startled at first by the degree of outrage in Khumbu over forest issues, but proved to be sympathetic to Sherpa concerns. He believed that a new relationship between the local population and the park had to be developed and was shocked at the level of distrust that existed after fifteen years of park operation. He believed that conservation must be based on local development and participation in resource management in order to be effective and advocated a greater role for the park as an agent of development. He also had considerable respect for traditional Sherpa land-use knowledge and resource-management institutions. Out of the new dialogue that ensued between the park administration and local leaders a number of basic changes in policy came about. These increased local communities' roles in regulating local forest use and established a regional forest management committee with a representative from each village which would consult with Sagarmatha National Park officials on forest-use policy. The national park would continue to subsidize shinggi nawa, of which there are now twenty-two in Khumbu, but these local officials would now enforce regulations that were at least in part decided by their own communities. The villages would decide how many trees they would authorize to be cut for beams and where these could be cut, although final authorization would require approval from national park authorities. Each proposal now had to pass through a multilayer review. It first had to gain the approval of the elected village representative to the forest management committee. Only then could it be reviewed by the national park. This established throughout the region the kind of two-tier system that had been in existence in some parts of Khumbu in the 1960s. Villagers now met to decide which regulations they should require their forest management committee member to uphold in his permit review and policed these with their own shinggi nawa. Communities could now much more effectively ban tree felling in sacred forests and other areas as they saw fit, and the fees formerly paid as royalities to the park for trees authorized

to be cut as beams (130 rupees per tree) were now to go to the villages themselves, which would keep these on deposit in Nauje bank accounts and use them for projects approved by the village development committees.^[30] Responsibility for all regulations concerning forest use and wildlife protection and their enforcement in the area around the Tengboche monastery, from the Milingo Chu to the Phunki Chu, were to be returned to the monastery.

These moves were hailed by local Sherpas, including some men such as Konchok Chombi, who had been among the most outspoken critics of the national park. Sherpas felt that they had regained some measure of control over their forests. At least for a short time a new spirit of cooperation with the national park developed. Nauje leaders, for example, accepted a national park-imposed ban on all fuel-wood gathering in the Satarma area, a place on the western shore of the Bhote Kosi where there had been heavy demand for fuel wood by Nauje families for some years. This was the first time any total ban on local fuel-wood gathering had been proposed by the national park, much less agreed to by Sherpas.^[31] The national park administration, for its part, agreed for the first time to build a bridge across the Dudh Kosi river to a previously remote forest area (Tougekok) which would serve as a source of fuel wood to replace Satarma, and Pandey also agreed that some tree felling for purposes other than house beams could now be done in certain areas. One remote area south of the Dudh Kosi, for example, will henceforth serve as a limited source of fir shakes for roofing. Here there will be a limit of two trees per household.^[32] In the spring of 1991 some Sherpas believed that other changes might also be ahead and that limited felling of trees for rafters and other uses might eventually be allowed, with the national park maintaining the authority to designate sites and select the particular trees to be cut.

The proposed development of a new national park management plan may provide the occasion to formealize these new approaches to park resource use and management, although they remain very much debated within the Department of National Parks and Wildlife Management. Whether or not these new regulations will be a brief experiment or a lasting new order remains to be seen.

These changes raise several important issues. One is the parameters within which local management can be exercised within a national park. National parks in Nepal, as elsewhere in the world, are established because of concerns for the protection of environmental quality which may include the preservation of some areas in order to safeguard biological diversity and protect endangered species and habitats. Sustainable use is a goal that is compatible with national park management only if there is zoning to buffer core areas where stricter nature protection is a goal and

where there is agreement that some traditional, local practices may have to be regulated in nontraditional ways such as imposing hunting seasons and limits or bans on the killing or collecting of endangered species of plants and animals. If a national park or protected area is to have zones in which local settlement and land-use activities are supported, the introduction of sustainable use principles may require local people to modify long-established land-use practices or resource-management institutions. There thus remains the likelihood of disagreement about resource management, for local goals may not always be compatible with park goals.

At Sagarmatha National Park there are already examples of such conflicts between resource-use regulations sanctioned by the community and those upheld by the national park authorities. Most Khumbu villages have proposed their own conservation-oriented regulations for forest use. Typically they have raised the number of trees their residents can fell in local forests for house beams from the national park's past policy of three to five to seven. Most of the villages are very serious about enforcing these limits and have selected leading residents to the offices of shinggi nawa and forest management committee representative. Konchok Chombi, the Khumjung representative on the forest management committee, has even advocated that his fellow villagers ban the cutting of trees for beams everywhere in the village vicinity and require that trees for beams be cut only on the far side of the Dudh Kosi river. Yet it is still true that the new limits represent regulations that make sense in terms of the economic goals of land use—enough trees to build a house—but which do not necessarily restrict demand to sustainable levels in terms of the ability of local forests to continue to supply trees suitable for house beams in sufficient numbers over the long term. Sherpas have discussed what they need from forests and the importance of restricting their use to only certain forest products and certain sites. They recognize that heavy demands from local forests will change those forests, but the villages have had to make decisions about how much tree felling should be allowed without the benefit of an analysis of the condition of particular areas of forest. Ecological study of the forests on which villagers rely and a program to monitor their condition would probably be welcomed by many local leaders as being in accord with their own conservation concerns. But this is not necessarily the case in all villages. In Nauje, for example, there has been less concern for conservation, and one villager gained approval at the village level for felling forty-five trees to use in building a large lodge adjacent to national park headquarters. In this case the national park intervened and blocked permission. After negotiation permission was ultimately granted for twenty-one trees. But if Nauje continues to pressure for large amounts of timber the concept of local sustainability of use could be called into question.

A master plan is sorely needed that will provide a mechanism for establishing local guidelines for the annual number of trees that can be felled in specific forest areas. To determine this will require forestry research of a type that has not yet been done in Khumbu, and to enforce it will require considerable organization, effort, and coordination between the national park and villagers. The issue of control of forests is thus very likely still far from settled, and the process of developing a new master plan will very probably involve much negotiation over the permissible range and intensity of subsistence uses of forests. The new master plan, if it does indeed support the new notion of sustainable forest use rather than preservation, will have to propose mechanisms to establish how this level of use will be determined and enforced, and it may have to consider developing different types of use regulations for different Khumbu forests. The establishment of special protection areas for the preservation of biological diversity or undisturbed habitats, for example, would require that little or no subsistence uses be allowed. Some areas may have to be closed to some uses, even grazing, for a period of years to allow forest regeneration. Other areas may be found to be suitable for limited and carefully supervised tree felling for local use or for creating fuel-wood plantations. Planning for local forest use, moreover, will have to be reconsidered if the Nepal-Austria hydroelectric project now under construction near Thami Og is successfully completed and has the considerable impact on fuel-wood use in Thamicho, Nauje, Kunde, and Khumjung that it is intended to have.

The current changes in forest management and national park policies in Khumbu, moreover, raise the prospect that the purposes and structure of protected area administration in the region may need to be reconsidered. It may be that all of Khumbu is not appropriate for national park management. This was recognized at the time that Sagarmatha National Park was founded and all of the villages, gunsa, secondary high-altitude agricultural sites, and high-altitude herding sites were left as islands outside the legal jurisdiction of the national park. This in effect established a rudimentary zoning system. In this arrangement, the park controlled forest, rangelands, and the high peaks with regulations that allowed grazing and some forest use whereas the Sherpas managed their settlements as crop- and hay-production places.^[33] In the early 1970s, when Sagarmatha National Park was planned, this was an innovative step and there existed no other model in Nepal for multi-zoned, protected area management. Today other approaches are being tried and the lessons being learned may well be applicable in Khumbu. In both the Annapurna range and the Makulu-Barun area a major distinction has been made between the more strictly protected core area and the areas where villagers live and make use of adjacent forest and

rangeland resources. In the Makalu-Barun case the high country, including summer herding areas, will be zoned as a national park in which strict nature-preservation principles will apply, whereas the lower-altitude areas where the main Rai, Sherpa, and Bhotia villages are located are to be administered as a conservation area. This conservation area will place much more emphasis on local management of development and conservation and on sustainable use of natural resources. The same approach would be appropriate in Khumbu. Below 4,000 meters the valleys could be declared a conservation area. The area above that as well as some representative and relatively undisturbed lower-altitude areas could remain a national park. This would mean greater Sherpa management authority in the lower valleys, especially over the forests, woodlands, and grasslands they use most intensively. In time such measures may become necessary, for Khumbu Sherpas are already becoming aware of the differences between national parks and the new conservation areas and the significance that this has for the way they can live their lives. The establishment of a Khumbu conservation area to complement Sagarmatha National Park would formalize government commitment to Khumbu conservation (in the sense of the wise use of resources) and local development as well as promoting tourism and preserving nature in a stricter sense. This complex set of concerns and goals has underlain the policies of Sagarmatha National Park since its establishment, but recognizing them in this formal fashion would greatly inspire Sherpa confidence and involvement in Khumbu protected area management, and in so doing greatly further the successful achievement of these goals.

Forest Nationalization and Impacts Reconsidered

The common view that forest nationalization caused the rapid abandonment of traditional Khumbu forest-management institutions and precipitated extensive deforestation is obviously in need of revision. I have suggested that the introduction of new forest-use regulations and Sherpa responses to them were more complex than has been allowed for, unfolding in different ways and at different paces in different parts of Khumbu and affecting different forests to varying degrees and in varying ways. Existing local institutions were not abandoned everywhere simply because Kathmandu declared the forests nationalized or an official in Nauje began to issue logging permits. In some parts of Khumbu local management has continued to be effective to the present, and although in other places local management was indeed undermined, this did not take place for some years after 1957, had less far-reaching environmental impact than had been thought, and was not necessarily permanent. In

some places Sherpas revived local management after a few years and they are in the process today of reinvigorating it throughout the region. Yet at the same time the history of forest change since 1965 also suggests that there was a basis for the concern with which Fürer-Haimendorf and the various national park planning teams reported that local forest-management institutions had been abandoned and that deforestation was taking place. Although deforestation was far less extensive and there may not have been a regional crisis underway, there were particular forest areas where logging for timber had increased and Sherpas had noted a sudden change in forest density in several areas where rani ban regulations had previously been in effect. There has perhaps been too much concern among outsiders with deforestation in the sense of a decline in forest cover and too little with the environmental implications of the new, post-nationalization patterns of resource use in terms of changes in the density, composition, and stand age of forests, woodlands, and shrublands and changes in the supply and accessibility of specific species and sizes of trees and shrubs in local demand for particular uses. In respect to both changes in vegetation and resource availability, moreover, it is important to place more emphasis on evaluating specific sites and variation across the region than simply to offer Khumbu-wide generalizations. The same attention to regional variation is also necessary, of course, in considering institutional change. Localized forest change in Khumbu after 1965 was significant in social as well as environmental terms in the places where it occurred, for it testified to conflicts within and among Sherpa villages over how forests should be used and managed and to the ultimate decision to suspend any local community responsibility for intervening in individual households forest-use decisions.

Khumbu forest management in the 1960s and early 1970s thus has to be seen as a highly variable, local, institutional accommodation to forest nationalization which had varying regional impacts on the environment and reflected regionally varying social dynamics. In some parts of the region the new, outsiders' concepts of forest management were wholly adopted and local institutions were abandoned. But elsewhere Sherpas acquiesced in the government's demand for new procedures and payments without surrendering their own authority to choose and enforce local concepts of how forests should and should not be used. In these places villagers observed when necessary the rules of the new system while continuing to use and regulate forests according to their own needs, concerns, and goals. And even in communities that ultimately abandoned their institutional regulation of local forests there were often cultural understandings that significantly shaped local patterns of resource use. Khumjung, Kunde, and Nauje villagers, for example, all

ceased to enforce community controls on the felling of rani ban trees without as individuals abandoning the belief that trees in the sacred groves around village temples and nearby hermitages must not be felled or lopped. Here no community meetings were held to reiterate regulations and no community officals were selected. Cultural values and conscience alone protected the groves. Thus even in these villages some woodland remained under traditional forest protection. While in some places the relatively recently adopted rani ban regulations were abandoned, far older customs of carefully protecting sacred trees and forests continued to be respected.

The literature on recent change in Khumbu forest management and forests has perhaps focused too much on the assumed impact of forest nationalization and not enough on the impact of the establishment of Sagarmatha National Park. The fundamental contradiction between long-standing Sherpa forest-use practices and Sherpa forest-management goals and early Sagarmatha National Park policies has had a far more significant effect on Sherpa forest use and management than did forest nationalization. Khumbu forest-use history from 1965 through 1976 demonstrates that forest nationalization is not necessarily incompatible with continued local resource use or even continuing local resource management. But the strict nature protection-oriented policies of Sagarmatha National Park created a very different relationship between national conservation policy and local resource use and management. The conflict between Sherpas and the park over the use of natural resources in Khumbu reflected apparently irreconcilable differences between Sherpa traditions of local resource use and management, which are based on a principle of strict protection of some areas and unrestricted local use of others, and the national park policy of strict protection of all forest. But the national park attitude has been tempered by some appreciation of the importance of forest use for Sherpa subsistence, and although regulations that have forced Sherpas to turn to Pharak for timber remain in place, the national park has continued to allow forest grazing, the collection of dead wood, and some limited tree-felling for beams. The process of developing protected area management policies that meet both national park goals and Sherpa concerns is still continuing.

It is now twenty-five years since forest nationalization was first implemented in Khumbu. Two different central government-directed systems of forest regulation have been tried in the region and several important lessons have been learned about the effectiveness of different approaches to forest management. One lesson has been that neither the permit system of tree felling in use from 1965 through 1976 nor the national park ban on tree felling has been as fully effective as Sherpa religious beliefs, self-restraint, and community vigilence in protecting

sacred forests. Another lesson has been that it may have been wiser to build on local management institutions to begin with rather than to undermine them for nearly twenty years and then attempt to reverse direction. A third lesson has been that while a ban on tree felling enforced by the army may noticeably lessen the impact of local subsistence use on forests in the short run, it may not create a climate in which a national park and its conservation ideals are likely to win support from the indigenous people whose sovereignty in their homeland has been compromised. This support for new conservation ideals and the new kinds of resource-management measures needed to achieve them, however, can be critical in the long term, especially if future park management is to be carried out with the participation of local people. These lessons are not unique to Khumbu and similar conclusions drawn on the basis of experience in many parts of the world have led to a new global movement in the design and management of protected areas (Stevens 1986a). Yet increased understanding of the strengths and limitations of pre-national park Sherpa forest management also suggests that future resource management in a place that is now both a Sherpa homeland and a national park of international significance cannot be based simply on a return to "traditional" patterns of forest use and regulation.

The Decline of Communal Regulation of Herding

There have been a number of challenges to the Khumbu nawa system of local pastoral management since Sherpas developed it in the nineteenth century. These include dissension within communities of users (both due to disregard of regulations by wealthy and powerful families and as a result of more widespread civil disobedience), conflicts among communities over the boundaries of village lands and the question of who makes the regulations for particular areas, and attempts by outside herders to circumvent local pasture-use regulations. Until recently, however, none of these efforts to undermine local traditions of seasonally excluding livestock from the nawa-administered areas was successful. Early in the century in the Imja Khola valley, for example, Pangboche and Dingboche residents were successful in forcing Gurung shepherds to honor local grazing regulations. In the 1940s Phurtse villagers were able to stop the widespread violation of grazing bans precipitated by a few wealthy nak herders by implementing a system of rotating the office of nawa among all village households so that it became more difficult for any household to use its power and influence to subvert community regulations.^[34] For decades Khumjung-Kunde and Thamicho villagers were able to persuade Nauje herders to comply with their nawa regula-

tions in areas near Nauje despite Nauje residents' disagreement over the ownership and control of those areas. And in recent decades community pressure in Khumjung and Kunde has forced obedience from powerful families who had tried to ignore the rules.

In recent years, however, Sherpas have been less successful in maintaining the effectiveness of communal resource management against challenges from individuals seeking private gain. Since the late 1970s the nawa system has come under enormous pressure in most of Khumbu, and in some regions the former resiliency and strength of the institution has been lost. Nawa ceased to be chosen in Nauje after 1979, and after 1983 the system went into decline as well in Thamicho. It was abandoned in some areas in 1984 and 1985 and throughout the Bhote Kosi valley after 1988. Khumjung and Kunde villagers have maintained grazing restrictions in Teshinga and the immediate surroundings of the village, but not without some contention. There has been much disagreement on whether to continue to choose nawa and on how to deal with the frequent violations of Khumjung-Kunde herding regulations by Nauje livestock. Pangboche continued to administer livestock bans effectively through 1986, but in 1987 regulations in the Dingboche area were widely violated when livestock were moved early into some areas closer to Pangboche. Since then some villagers have been quite concerned that the institution may ultimately be undermined. Only in Phurtse has there been relatively little sign of recent challenge to nawa enforcement of regulations.

The events that precipitated the crisis in Nauje in 1979 are well remembered. According to one of the three nawa of that year, trouble began immediately after livestock had been excluded from the village after the Dumje celebration. A number of families were tardy in removing their stock from the village and had to be fined five rupees per head of stock. Offenders were given two to three days to remove their livestock and were warned that if they refused to do so they would be fined more heavily. In 1979, however, even the levying of increasingly severe fines did not induce all households to comply with the village rules. Several households still had not moved their livestock out of the closed area after several visits from the nawa and being fined first ten rupees and then twenty rupees per animal. At this point matters were made far worse when a number of other families brought their stock back to the village, announcing to the nawa that if some herders were going to refuse to adhere to the rules, they would also refuse to obey the nawa.^[35] According to one of the nawa that year soon "everybody came down." The nawa gave up trying to enforce regulations in the face of mass civil disobedience. "This is not one person's job, it is everybody's job," one of the last nawa noted. "If they [the other villagers] will not listen, forget

it." The next summer no new nawa were appointed for the first time in memory and none have been chosen since.^[36]

Since 1979 Nauje decisions about where to herd livestock have been left entirely to each household's discretion. Some families still move their livestock outside the old restricted area each summer to higher pastures, but many families keep zopkio in Nauje for at least part of the summer and it is now common for zhum and cows to be grazed in the village environs year round.^[37] The presence of livestock in the village during the summer months has led to increasing problems with crop losses, despite the efforts by many families to build more secure walls around their fields, and is perceived to have degraded the quality of winter grazing around the village.

The decline of nawa management in Nauje has also affected adjacent areas. During the summer Nauje stock commonly stray onto nearby Khumjung, Kunde, and Thamicho pastures that have been closed to grazing. The continuing violation of the Khumjung-Kunde herding restrictions in the Shyangboche area by Nauje livestock has been especially prominent.^[38] Khumjung and Kunde nawa have come in for considerable censure from their fellow villagers for not better handling Nauje violations in this area. Individual nawa have made real efforts to control Nauje stock, trying to move dozens of animals back outside the restricted area, but have been ineffective. Several dozen head of Nauje stock at a time are often on Khumjung-Kunde land and as one Khumjung nawa put it, "daytime I push them [the livestock] out, but at night they come back. I cannot do duty all night." Some Khumjung villagers note that in the old days there would have been more of a confrontation and recall times when the entire village had marched to Nauje to demand compliance and threatened to fight for it if necessary. But today there is a lack of consensus over what kind of response should be made to Nauje violations, and this and uncertainty over legalities have stymied the effectiveness of the Khumjung and Kunde nawa in dealing with Nauje stockowners. To make matters worse some Khumjung-Kunde villagers are now losing their own dedication to observing the rules. As early as 1984 some Khumjung herders were complaining about Nauje violations and proclaiming that if Nauje stock were going to violate the regulations and graze Khumjung winter grass then they were going to bring their herds down early too. Thus far, however, only a few herders have begun to ignore the rules and these have experienced considerable social pressure from fellow villagers to comply with them. Prominent Khumjung villagers note, however, that people are beginning to tire of arguing and fighting over the herding regulations and that the community may well decide one year soon to abandon the effort of maintaining nawa regulation. Thamicho families also found Nauje stock violations of

the grazing regulations in the lower Bhote Kosi valley extremely annoying and some of them unsuccessfully asked the national park to enforce these regulations against Nauje families.

In the Bhote Kosi valley nawa administration went into decline during the 1980s and by the end of the decade the system had been abandoned throughout Thamicho. The first signs of difficulty came in 1984 when no nawa were selected in Pare. In 1985 none were selected anywhere in Thamicho and it appeared that the institution had been abandoned. In 1986, however, Lhakpa Norbu, a Thami Teng Sherpa who was at that time chief administrator of Sagarmatha National Park, inspired Bhote Kosi residents to resume selecting nawa. That year nawa were again chosen, but they had a great deal of difficulty enforcing the regulations against fellow villagers and Nauje livestock. The following year, with little interest by the general public in accepting nawa duties, the six Thamicho adekshe (ward officials in the panchayat government) were given the responsibility by their fellow villagers.^[39] It was further decided that stiffer penalties should be enforced and the fine was set at fifty rupees per head of stock. Violations, however, were again widespread and the adekshe had trouble addressing them. Attempts by the adekshe, for example, to force an old woman to move a cow out of the closed area in midsummer were defied to general amusement, and a villager noted that "six adekshe cannot control one old lady with a cow." By the end of summer the grazing regulations were being widely violated. The following year it was decided to go back to electing nawa rather than simply assigning the job to adekshe, but it remained difficult to enforce the regulations. No nawa were chosen in 1989 or in 1990.

The nawa systems of eastern Khumbu (Phurtse and the Pangboche-Dingboche areas) have not been challenged to the same degree, but friction in the Imja Khola valley appears to be increasing. In 1987 the nawa of Dingboche had trouble enforcing their decisions about the timing of the beginning of the barley harvest and the opening of the area a few weeks later to grazing. After two unsuccessful attempts to halt early harvesters the nawa gave up the effort and many head of stock were brought into the area before it had been declared open for grazing. Some Pangboche families who have land at Dingboche openly wondered whether nawa regulation would be maintained there in the future. There remains, however, a very strong conviction that blight prevention and livestock-control regulations should be enforced, and the nawa were much criticized for not acting effectively in 1987.

Phurtse is the only place where livestock exclosures currently seem to be maintained with little village debate. One factor here may be that the system is administered by a single village to regulate its own residents' livestock. It may also make a difference that this is a relatively small

village where many families herd nak. Perhaps there is a greater consensus about resource regulation here, or more effective social control or local leadership.^[40] We have already seen, however, that even in Phurtse the enforcement of grazing restrictions has not always been straightforward and that the system was under considerable pressure even forty years ago. Whether the community can continue to respond so successfully to its own internal challenges is by no means certain. The 1990 decision to change very old regulations regarding the collection of fuel wood in the local lama's forest is a sign that differences of opinion on how to manage local resources may be increasing. In this case middle-aged and older Phurtse residents found themselves for the first time in a major debate with young householders on an important issue that involved land use and religion, and they ultimately acquiesced to the changes demanded by young villagers.^[41] Now dead wood will be collected in the sacred forest next to the village, and in local interpretation this also allows the lopping of dead branches.

It is difficult to pinpoint the underlying causes for the breakdown in nawa management at Nauje and in Thamicho. Nationalization was not the factor here that it was in the undermining of some local forest-management institutions. Although the national park has considered the implementation of rotational grazing schemes and the possible establishment of ceilings on livestock numbers (Bjønness 1980a; Garratt 1981), the central government and the national park have not yet moved to intervene in any way in either pasture use or herd composition other than in encouraging the banning of goats. Indeed, as already mentioned, there have instead been at least two cases in which officials have supported the nawa system or urged its reinstatement. The continuing tension between private interests and communal attempts to regulate the use of rangelands has, of course, been a common theme in all the areas where there have been difficulties in maintaining the nawa system. But this is a very long-standing conflict between cultural values and social institutions and does not go far in explaining what has tipped the balance against community regulation so strikingly in recent years in Nauje and Thamicho. In the Thamicho case competition for resources due to increased summer grazing on Thamicho-administered lands by Nauje stock was one important factor, as was the Nauje precedent of abandoning nawa enforcement. But Thamicho villagers themselves emphasize the increased number of zopkio and cows that have been kept in recent years by Thamicho herders, the irresponsibility of some Thamicho herders who have disregarded the communal regulations by bringing their pack stock to graze for days at a time in the lower Bhote Kosi valley while waiting for the arrival of mountaineering expeditions, the increasing loss of crops to zopkio, and the inability or unwillingness of the nawa

to successfully police the area. There have been so many different points of pressure and loss of control that people have lost confidence in the system. They no longer feel that they can rely on it to protect their crops or to ensure that they have equal access to lower-valley pastoral resources. There is also the possibility that some new herders who only keep a few head of dairy stock or zopkio would prefer to remain based in the main village in the summer and not have to bother with the lifestyle demands of shifting to high-altitude resa and herding huts. This last factor may have been important in the erosion of community commitment in Nauje to maintain the nawa system. Many of the Nauje families who began keeping zopkio, cows, and zhum in the 1960s and 1970s had not previously practiced the annual routine of shifting base to a summer herding station outside of Nauje. They may not have been happy with the traditional regulations that required cows and zhum to be herded far away from Nauje in places such as the Pulubuk valley east of Tengboche in the summer and may have relished the opportunity to forgo the cost of hiring herders to watch the stock in these areas or the trouble of doing it themselves. The opportunity to remain all summer in Nauje, with stock grazing adjacent slopes, thus may have been welcomed by many families even though it meant poorer winter grazing. And once some families began defying the system and keeping their stock in Nauje there would have been a great temptation for other families to likewise opt for personal convenience and the opportunity to give their stock the same chance at grazing the rich summer grass near Nauje that their neighbors' stock had.

Environmental Consequences of the Abandonment of Local Pastoral Management

Nauje's abandonment of communal enforcement of summer livestock herding regulations and autumn grass-cutting rules and the abandonment of herding regulations in the Bhote Kosi valley have had repercussions on the Khumbu environment and pastoral resources. Grazing pressure has increased in areas where seasonal grazing restrictions were once enforced, especially in the vicinity of Nauje and the lower Bhote Kosi valley but also in the Khumjung-Kunde area where Nauje stock have increasingly transgressed areas within the zone that Khumjung-Kunde closes to summer grazing. Nauje's abandonment of the nawa system has also affected pasture conditions by ending the former practice of restricting wild grass harvest until the nawa declared in the autumn that this could begin. Villagers feel that the amount and quality of hay made from wild grass have both declined. These pressures have been further accen-

tuated by the increasing trend in these settlements towards the keeping of greater numbers of zopkio, zhum, and cows, which have different grazing preferences and fodder requirements than yak and nak.

Nauje herders have observed that winter grazing near the village and in easily accessible areas of the lower Bhote Kosi and Dudh Kosi has declined since 1979, making it more important than ever to store large amounts of fodder for use in the winter and spring. In the 1980s, for example, some Nauje families began sending stock (dimzo and urang zopkio) across the Bhote Kosi river to the Satarma area in the late spring because the grass had become so poor nearer to Nauje.^[42] This, they noted, had not been necessary in the 1970s. Yet at the same time that spring grazing becomes more difficult it has also become harder to store large amounts of fodder. Nauje herders have noticed a major decline in their ability to obtain wild grass for autumn hay making. Before the 1980s a day's work cutting grass on the slopes near the village supplied a large, forty-kilogram load of wild grass. No one bothered to go far afield to cut grass and it was rare for grass cutters to cross to the far side of the Dudh Kosi or the Bhote Kosi. Now grass cutters work the other side of the rivers and in a day people are content with a small load of ten to fifteen kilograms. Some Nauje wild-grass cutters are now even working in more remote areas of Khumbu, including Tengboche and the upper Imj a Khola valley, where they store hay for use by their yak and zopkio along the mountaineering and trekking route to Mount Everest.^[43]

Villagers give a number of reasons for the increasing scarcity of wild grass for hay making. These include the lack of enforcement of the summer livestock exclusion and the resulting heavier grazing of areas that formerly supplied wild grass, the increase in the number of head of village stock caused by the increased herding of zopkio and dairy stock, and premature cutting.^[44] Since 1979 Nauje stock owners have begun cutting wild grass earlier than was previously normal in order to gain an edge on their competitors. In 1987 grass cutting began on August 18, several weeks before the nawa would have opened the area to cutting in the years before 1979. The result is said to be that grass is cut before its seed fully matures, making the resulting wild grass less nourishing fodder. This practice may also lead to a longer-term loss of range productivity, for as Brower notes, "an early ritsa harvest means wild hay grasses are cut before they have finished their annual cycle, presumably undermining their chances for persistence" (1987:299). Heavier summer and early-autumn grazing must accentuate this process. The decline in the enforcement of summer livestock grazing bans has also undoubtedly increased the amount of grazing pressure in woodland, forest, and shrub areas in the lower Bhote Kosi valley, lower Dudh Kosi valley areas near Nauje, and Shyangboche. Although it seems likely that this has contrib-

uted to poorer forest regeneration, studies necessary to establish this clearly have not yet been carried out.

It seems likely that continued year-round grazing by large numbers of livestock, and particularly by large numbers of zopkio, may gradually lead to a decrease in pasture carrying capacity. Declining grazing resources and availability of wild grass for hay may lead to increased risks to herders of winter stock losses, particularly in years of high snowfall when greater amounts of fodder than usual are required to save livestock from starvation. Herders who own insufficient private hayfields and those who lack the financial resources to purchase large amounts of fodder would be especially at risk since they would be less likely to set aside enough of a stock of fodder for such conditions and would instead hope for normal weather.^[45] It is tempting to ascribe the major winter livestock losses that took place in Khumbu in 1985-1986 to this process of declining range carrying capacity. But the fact that this mainly affected yak and nak—whose upper valley grazing and fodder resources have been least affected by the increase in the number of zopkio and the decline in nawa regulation—suggests that weather and not grazing impacts was the key factor in the scarcity of grass and subsequent starvation of stock. This view is also supported by the fact that these losses did not take place only in Thamicho but also in the upper Imja Khola valley among Pangboche stock.^[46] No such large-scale winter die-offs have taken place since, even though autumn snowfall was early and heavy in 1987 and this might have been expected to create a crisis. Nor have disastrous winters been simply a recent phenomenon. Extensive loss of stock in winter and spring was reported for at least one winter in the 1950s and another in the 1970s (Bjønness 1980a; Fürer-Haimendorf 1975).

Whether the situation will worsen is not clear. Fodder shortages and increasingly high fodder prices may begin to discourage families from investing in more zopkio or even lead to a decline in their total numbers. A change in the policies of trekking companies and expeditions that favored the use of porters rather than pack stock would certainly also dramatically effect regional stock numbers and grazing and fodder requirements. It is also likely that Sherpas will respond to the regional shortages of fodder by attempting to increase hay yields in their walled fields and through more cultivation of fodder crops. Experiments in Nauje with cultivating wheat and barley as a second crop following potato production, experiments with irrigating hay in Samde, and continuing interest in the introduction of new hayfield grasses all suggest that Sherpas may be able to raise regional stock-carrying capacity. There are also more efforts today to import hay from Pharak, even though this is considered to be less nourishing than Khumbu-grown and gathered

hay. And some Sherpas remain interested in reviving nawa regulation in Nauje and Thamicho, although there is not very much optimism that this can be done. There also remains the possibility that concern over livestock-related environmental damage will lead to government intervention in the changing relationships between pastoral use and Khumbu range resources. Sagarmatha National Park administrators have shown an interest in encouraging the continuing operation of the nawa system, and over the years there have been a number of proposals that the national park should take action to control Khumbu herding. These proposals have included a ceiling on regional livestock numbers (Bjøn-ness 1980a ) and the implementation of a new rotational grazing system (Bjønness 1980a ; Garratt 1981). The issue of limitations on the number of zopkio has also been raised and debated (Brower 1987:310). Park managers and advisors and Kathmandu officials agree, however, that government intervention in Sherpa livestock ownership or management would be highly resented by Sherpas and very difficult to implement. For the time being, at least, Sherpas continue to work out the dynamics of their pastoralism within their own economy, culture, and environment. Whether they will successfully cope with new and increasing pressures on pastoral resources and institutions remains to be seen.

9
From Tibet Trading to the Tourist Trade

Khumbu Sherpas, like many Himalayan high-altitude peoples and many mountain peoples all over the world, have long relied on trade as a basic and vital component of economic adaptation to the limitations of their environmental conditions.^[1] Exchange provides access to the resources of other altitudinal zones making possible a broader lifestyle than the limited local repertoire of crops and livestock allows. The possibility of obtaining lower-altitude grain in particular decreases dependency on highly variable high-altitude crop yields. This has several important implications for household and regional economy and lifestyles. It lessens the risk of food shortfalls, permits a larger population to be supported in the higher reaches of the mountains than local resources alone would allow, and makes possible a greater specialization in the crops and livestock most suited to local environmental conditions. In places like Khumbu, where there is relatively little interest in commercial agriculture and only a few families are involved in commercial pastoralism, transit trade based on the products of other regions has enabled families to keep decisions about crop production and some kinds of pastoralism free of market considerations. This has had important implications for land use and lifestyles.

Khumbu is situated at the edge of Tibet, at the entry to the final, high-altitude crux of a major trans-Himalayan trade route linking Nepal and northern India with Tibet. Although the Nangpa La is one of the highest of the many passes that cross the Himalaya to Tibet, it is not a difficult passage for yak and for people who are properly equipped for snow and cold and knowledgeable about mountain weather and glacier

crossings.^[2] Khumbu Sherpas thus have been ideally situated to be middlemen on an important long-distance trade route, and since 1828 the Nepal government has granted them a monopoloy on part of this trade by banning Tibetan traders from proceeding any further south than Nauje and non-Khumbu Nepalis (including Pharak and Shorung Sherpas) from trading across the Nangpa La.^[3]

Khumbu became a major exchange and transshipment point. Goods from the south carried by Rais and by Khumbu, Shorung, and Pharak Sherpas were exchanged here for goods from Tibet carried south by Tibetan and Khumbu traders. Khumbu Sherpas carried some of the southern goods on into Tibet and some of the products of Tibet on south into Nepal. In Tibet they traded in the villages of the adjacent Tingri region, in the regional centers of Ganggat and Shekar and in more distant Shigatse and Lhasa. A few Khumbu traders continued on further across Tibet, crossing west to the trans-Himalayan regions of Nupri, Manang, Thak Khola, and Mustang in north-central Nepal and east to Sikkim, Kalimpong, and Calcutta. Many Khumbu people traded southwards, bartering Tibetan salt and wool in lower-altitude Nepal for grain to be consumed back in Khumbu. Some continued as far as the Tarai town of Rajbiraj and the north Indian city of Jaynagar to obtain trade goods for resale in Tibet.

These routes were part of much vaster networks that moved goods across large portions of Nepal, Tibet, India, and western China from central Nepal to Calcutta and from the Ganges plain to Lhasa and the tea-producing regions of Yunnan and the Chang Jiang (Yangzi River). Khumbu Sherpas were one of a number of Himalayan peoples who played a small role in this trade, benefitting from the movement of goods but powerless to do more than respond to the trade opportunities created or terminated by political and economic actions taken by other peoples in distant places. Throughout the late nineteenth and early twentieth centuries Sherpas prospered through their increasing participation in international trade and moved freely among Nepal, Tibet, Sikkim, and India. But during the 1960s, when both political and economic change in Tibet, China, and India took place, this vast international trade network was severely disrupted by war, civil unrest, border policies, new trade and transit regulations, and development policies. Sherpas continued to participate in the new, limited forms of trans-Himalayan trade, but their attention began to shift in two new directions: their new role as consumers in a new intra-Nepal periodic market system and their quest for new opportunities for earning the money to play that role. Their ability to participate in the new interregional economy and ultimately to maintain the adaptive way of life that trade had supported, came in a very short time to depend on their ability to earn cash through sources other than

farming and trading. The foremost new source of income proved to be tourism, and the phenomenonal speed with which Khumbu Sherpas embraced new opportunities in the tourist trade led by the mid-1970s to a new regional and household economy in which the former role of trade had been virtually entirely transferred to tourism development.

This chapter surveys historical change in Sherpa trade with Tibet and lowland regions, the decline of long-established trade networks in the 1960s, and the rise of tourism as a basic component of the Khumbu economy and Sherpa subsistence.

Trade Routes, Trade Goods, and Trade Systems

Khumbu Sherpas have conducted several different types of trade, each of which has varied historically in terms of scale, products exchanged, routes, and markets. Families from different villages and with different capital and labor resources, trading experience, connections, and lifestyle preferences were involved in various types of trade. Seven types of Sherpa trade can be identified as well as three types of trade conducted in Khumbu by non-Sherpas (table 27, and figs. 15 and 16).

The most important trade over the Nangpa La for most Sherpas was the salt trade. Salt was an immensely valuable commodity in Nepal as it has been in many parts of the world, and for centuries the major supply came from the great interior saline lakes of the Tibetan plateau. Nomads evaporated salt along the lakeshores and brought it by pack stock toward the settled heartland valleys of Tibet where they bartered it for barley tsampa from villagers who came there to meet them. These villagers in turn traded the salt to merchants in Ganggat and Kaprak. Sherpas traded in both of these nearby places for salt offering barter goods from Nepal and India in exchange. Both Sherpa and Tibetan traders then carried salt south across the Nangpa La and into Khumbu. From there it was traded across a vast region of eastern Nepal that reached as far as the Arun valley in the east, the confluence of the Dudh Kosi and Sun Kosi in the south, and the Tamba Kosi valley in the west. Salt bound from Tibet to eastern Nepal was traded south over only a few passes, and the Nangpa La was one of the most important in terms of the area it furnished with salt.

In most other Himalayan border areas similar trade for Tibetan salt was based on a barter exchange for Nepal-grown grain. This has also been assumed to have been the case for the Nangpa La trade. Khumbu Sherpas are usually described as hauling Nepal-grown rice and other grains north to Tibet to barter on good terms for salt and then profitably bartering this salt in lower-altitude areas of Nepal for grain (J. Fisher 1990; Fürer-Haimendorf 1964, 1975, 1984). This view of the basis of

Sherpa trans-Himalayan trade is partly mistaken in that a salt for grain barter was important only for the Nepal portion of the trans-Himalayan trade and was not the key to the entire trade both in Tibet and Nepal. Sherpas exported very little grain north over the Nangpa La, and the substantial stocks of rice and other grains that passed through Ganggar came instead from the Kuti and Rongshar routes farther to the west. The Sherpa barter trade in Tibet for salt was based not on grain but on butter, paper, and other goods obtained from Shorung and the lower Dudh Kosi valley. In the late nineteenth and early twentieth centuries Sherpas also exported a good deal of iron from the Newar mines at Those. Some Sherpa traders also purchased salt in Tibet using Nepalese, Indian, Tibetan, and even Chinese currency. It was only once Sherpas had trans-

Figure 15.
Trade to Khumbu. All arrows indicate trade conducted by Khumbu
Sherpas; other groups that traded to Khumbu are indicated in large type.

ported the Tibetan salt into Nepal that the grain for salt barter exchange became important. Sherpas traded salt for grain on a large scale with Rais, lower-altitude Sherpas, and other peoples both in Khumbu itself and in more southerly regions. Very little of the grain that Khumbu Sherpas thus obtained, however, was ever taken over the Nangpa La and into Tibet. It was bound instead for Khumbu where it became a fundamental component of the Sherpa diet, supplementing locally grown barley, buckwheat, and tubers. This trade freed families from relying entirely for subsistence on production from their own fields, and it was the demand in Khumbu for rice, maize, millet, wheat, and buckwheat from lower-altitude regions that spurred the widespread Khumbu Sherpa involvement in the trans-Himalayan salt trade. For most Sherpas participation in this trade was their only means to acquire grain.

Three very different types of traders were involved in the salt trade. Most of this business was in the hands of a very few big Sherpa and Tibetan traders (tsongba ). For decades in the middle of the century four

Figure 16.
Sherpa Trade from Khumbu

or five Nauje men dominated the trade who dealt in entire yak caravans of salt and imported tons each year from Ganggar (ten days north by yak) and Kaprak (six days north). One of the great Nauje traders of the 1940s and 1950s is said to have brought fifteen tons of salt into Khumbu per year (6,000 pathi), an amount that required 375 fully laden yak to transport.^[4] These big traders did not bother transporting salt any further south than their Khumbu homes, from which they sold it to other Sherpas and Rais. More Sherpas imported salt on a smaller scale making use of their own few yaks or hiring a small number of pack stock or porters to bring it south over the Nangpa La to Khumbu. Still more families did not go to Tibet at all, but instead obtained small quantities of salt in Khumbu from larger traders. They then transported this salt south on foot by the basketload to trade with Rais and other residents of the lower altitudes. This form of trade, known as nangzum , required laboriously hauling loads of salt by tumpline in a series of shuttle trips to Rai, Sherpa, Hindu, and Newar communities where it would bring a better exchange than back in Khumbu. Families of moderate means and poor households participated

only in this form of trade and often obtained their salt on loan from the big Sherpa traders. In the spring, after they had carried back to Khumbu the grain they had received in barter exchanges in the south, they repaid the loan of salt with a share of the grain. Such petty traders were the Sherpa equivalents of the many Rais who came by foot to Khumbu carrying baskets of grain to barter in Nauje and other places for salt. The better exchange rate they got from Sherpa traders in Khumbu compared to that which the Sherpas offered in their home communities made their time and physical effort worthwhile.

A very different type of trade revolved around the import of wool from Tibet. This was carried out by a number of Sherpa small traders who obtained wool (and also other products of the country such as dried sheep meat and fat) in Ganggar in the late spring and early summer in exchange for sugar, snuff, cloth, and other products of the Tarai and northern India. They then carried loads of wool back to Khumbu either on foot or with the help of a few yaks or porters. Back in their home village their families (and sometimes also hired helpers) worked the wool during the summer months and after the crops had been harvested in the autumn they went south with their entire families to trade these woolen goods to Rais and Shorung Sherpas. Woolen blankets and light rugs were traded to the Rais (along with heavy yak-hair tarps), but perhaps the most valuable wool products were matils , the striped aprons universally worn by adult Sherpa women. In Shorung they sold aprons for cash and also sometimes for other goods valued in the Tarai, such as a certain type of dog bred in Shorung. Then they moved on to the Tarai, and usually continued on all the way onto the Ganges plain to the town of Jaynagar, twelve to thirteen days on foot from Nauje.^[5] Here they bought a variety of high-value, low-bulk goods ranging from Indian manufactured goods, dyes, and snuff to material for making prayer flags. Some traders also purchased quantities of brown and white sugar in large multikilogram cakes. They they moved these goods back toward Khumbu, sometimes hiring Tamang porters to help. On the way they often spent a few weeks in January and February in Shorung and the lower Dudh Kosi valley buying grain with cash, trade goods, or salt they had cached there on their way south. The rest of the late winter and early spring was devoted to shuttling grain back to Khumbu before family efforts had to be focused on preparing fields and planting crops. Once the crops were in it was time for another round of trading in Tibet. Here the sugar in particular turned a fine profit. In Jaynagar in the early 1950s two twenty-kilogram loaves of brown sugar could be purchased for five Indian rupees. Back in Nauje they would be worth five times that, whereas in Ganggar they brought enough wool for thirty-five aprons, each of which would bring seven or eight rupees back in Shorung.

There were also several different types of livestock trading. Most of these were linked to large-scale, international trade in dimzo. Khumbu was a major base for the crossbreeding of nak and Tibetan varieties of cattle, and virtually all male dimzo calves born in Khumbu were ultimately exported to Tibet. Female crossbreed dim zhum were in demand as dairy cattle in Shorung and most of these calves were also sold rather than being raised as part of Khumbu herds. Most of the dimzo calves also initially were dispatched to Shorung. Shorung traders purchased these calves from Khumbu stockbreeders and dealt them to Shorung herders who raised the calves for several years and then resold them to Khumbu men who traded the two-, three-, and four-year-old zopkio to Tibet. The Khumbu zopkio traders were primarily Nauje men, although a few men from other villages also engaged in this business. They either journeyed to Shorung to buy dimzo or else contracted in Khumbu for their purchase and delivery. In the latter case the young zopkio were delivered to them in Khumbu during the monsoon months and the traders herded them there until the autumn when they drove them over the Nangpa La to the Tingri area. Some of the larger stock traders took as many as a 140 head of stock north. In the 1950s a three-year-old zopkio worth 200-300 rupees in Nauje could be traded for two nak, forty-five kilograms of wool, or about eight loads of salt. Occasionally Khumbu traders drove zopkio beyond Ganggar, for there was also demand for them in the trans-Himalayan areas of central Nepal north of the Annapurna Range, particularly in Manang and Mustang. Groups of Khumbu herders sometimes drove scores of young zopkio across the grasslands of Tibet to this region.

There was also a trade in importing nak from Tibet to Khumbu, where they were sold to herders who were involved in the crossbreeding business. The nak trade also sometimes involved long-distance drives, for the best nak were obtained not in the Ganggar region but far beyond it on the high plateau country of Chang. Nak could be readily purchased for cash or obtained in exchange for zopkio calves in Ganggar or in the Tibetan villages closer to Khumbu, but sometimes Sherpas went as far afield as Chang to obtain better-quality nak at lower prices. This journey began with a six- or seven-day trip on foot from Ganggar to Lhartse (four days by horse). Here traders crossed the Tsangpo by boat (or took a different route to a suspension bridge) and then continued two more days on foot to Dokshum. Here nak could be bought for seven to nine Indian rupees in the 1940s. Mostly only small-scale herders made the effort to go this far, and in the spring or autumn groups of such Sherpas would set out for Chang to buy three nak or so each to build up their personal herds and others went each autumn (during the tenth and eleventh lunar months) with young dimzo to bargain for nak. Seven nak

for two dimzo was the usual rate, and this included the cost of having the nak herded all winter by the Tibetans before the Sherpas returned to pick them up in the spring (in the fourth month). Sherpas note that for one's own use the best nak come from Pakrukok (beyond Lhartse) for that area has less grass than Khumbu and when the nak are brought to Khumbu they thrive on the richer pasture and breed better. It was considered best to import nak in the spring, for if they did not feed on Khumbu grass during the summer they were not considered likely to be healthy and strong for the winter ahead. If one was looking for nak to sell to other Khumbu herders it was considered best to buy them in the Sangsang area of Chang. These nak graze on excellent grass and their fine appearance brings a good price in Nepal. Knowledgeable stockmen, however, believe that such stock grow thin and weak on Khumbu grass and as a result breed poorly. Nak were also brought back from parts of Tibet closer to Khumbu, including the Tongnak, Gupchichutang, and Nyiring Gompa areas.

For twenty years one Nauje Sherpa conducted a rather different nak import business, keeping his own large herd of nak and yak in the Rongshar valley of Tibet, where hired Tibetan herders cared for them in exchange for a share of their milk and calves. He is said to have brought eighty or ninety nak per year to Khumbu for sale until 1962, when he discontinued his Tibetan operation for fear that it would be halted by the Chinese.

Horse trading was another special pursuit. The Nangpa La was considered to be too high in altitude and too cold for horses to cross, although there are stories of a few that survived the trip. Khumbu traders instead took horses that they purchased in Ganggar southwest over a low pass (the Pusi La) into Rongshar and then followed the valley of the Rongshar Chu south into Nepal. Some of these were sold in Shorung, others were brought all the way to Khumbu. Most were sold in the annual trade fair in southern Nepal at Maini, and subsequently became tonga horses pulling the carriages popular in the towns of the Tarai and northern India.

Some people also dealt in herbs. These families collected hugling and other medicinal mountain herbs in Khumbu and adjacent regions and took them south to the Tarai or India. Profits from their sale could be invested in grains for Khumbu consumption and also put into trade goods for sale in Tibet or bartered there for salt or wool.

Finally, a relatively small number of traders conducted long-distance trade in luxury goods. Most of them were Nauje men, although some Kunde and Khumjung men also were involved in this kind of trade that revolved around the export of luxury items from Tibet to Nepal, most of them intended for sale to Sherpas either in Khumbu or other nearby

regions. Among the items that were important in various periods were "Tibetan" brick tea (actually tea grown in Yunnan, Sichuan or the middle Yangzi valley and specially prepared for the Tibetan market), carpets, silver goods such as teacup lids and stands and belt buckles, Chinese silk and silk brocade, jewelry and precious stones (turquoise, coral, and the prized zi agates), and a range of religious artifacts including paintings, statuary, prayer wheels and books. Such goods could be obtained in major Tibetan urban centers and Khumbu Sherpas sought them especially in Shigatse and Lhasa.^[6]

Sherpa traders employed several different strategies for obtaining Tibetan luxury export goods in Sikkim, Darjeeling, Kalimpong, and Calcutta. The most common tactic involved a three-step process. First bulk goods from lower-altitude areas of Nepal or higher-value, less bulky items from northern Indian bazaar towns would be obtained during winter trading trips south. These would then be sold for cash in Ganggar and Shigatse. With this money traders would then purchase luxury goods in Shigatse or Lhasa to import to Nepal. Some traders with greater cash reserves chose instead to make the journey directly to Shigatse or Lhasa to purchase goods for import to Nepal. Others chose yet another tactic and began by traveling to Kalimpong or Calcutta. There they purchased goods which they resold in Lhasa or Gyantse for a profit which could then be used to buy Tibetan luxury goods for export home. Some traders left Khumbu with a caravan of yaks and porters carrying butter, paper, iron, and other bulk goods, sold these in the nearest Tibetan market centers, traveled across Tibet to Kalimpong on foot and then went by train to Calcutta, returned to Tibet and traveled north to Lhasa, Gyantse, and Shigatse to trade goods obtained in Kalimpong and Calcutta, and finally arrived back in Khumbu ten months or more later with a cargo of highly valuable Tibetan goods. Other traders spent years shuttling back and forth between Kalimpong and Tibet before ultimately calling an end to their trade activities and heading home with a few loads of immensely valuable trade goods.

Historical Changes in Trade

Although there are indications that Khumbu Sherpas have been trading with Tibet since at least the early nineteenth century, there is little evidence with which to analyze the origins of the trade, its scale, or its emphases prior to the late nineteenth century. It is clear that the Nangpa La was used as a trade route as early as the 1820s and that at that time Khumbu Sherpas had been given special trading privileges over the pass. Fürer-Haimendorf described an 1828 Nepal government document that he was shown in 1957 in Khumbu which granted Khumbu Sherpas alone

the right among Nepal's citizens to trade across the Nangpa La and which restricted Tibetans from trading any further south than Nauje. I have seen later documents that refer to the 1828 edict and which indicate that Khumbu Sherpas repeatedly had to appeal to the government to continue to enforce their privileges.^[7] The original decree probably only recognized and regulated already existing trade, but as of now there is no further evidence to establish how long this may have been underway. Khumbu oral traditions suggest that it predates the establishment of Nauje in the early nineteenth century, for there are accounts that the Chorkem area near Nauje was formerly a place where Sherpas and Rais met to exchange Tibetan salt and grain grown at lower altitudes. Ortner relates an oral tradition that describes an encounter by the early Sherpa settlers with Rais in Pharak in which they exchanged Tibetan salt for flour (1989:37-38). Detailed insights into early trade, however, are only possible after the mid-nineteenth century. Oral traditions and oral history offer a wealth of information about late-nineteenth- and early-twentieth-century trading practices.

Khumbu trade in the late nineteenth and early twentieth centuries revolved primarily around the Khumbu crossbreed trade and the exchange of Nepal iron, paper, and butter in Tibet for the Tibetan salt and wool which Sherpas needed in order to obtain lower-altitude-grown grain. In this era Khumbu traders apparently seldom ventured far into Tibet. Most primarily traded in Ganggar (fig. 17). Early in the century much iron was exported over the Nangpa La to Tibet and Khumbu traders obtained this either directly at the Those mines or from Shorung Sherpa middlemen who transported it as far as Nauje. Butter was primarily obtained from Shorung Sherpas who carried it to Khumbu in order to barter it for salt. These Shorung Sherpas used a now virtually abandoned summer trail that provides a direct route from the Junbesi area to Pharak via high summer pasture areas. Another important trade good was shere , a root gathered in Shorung that was used in the manufacture of incense. Paper came from the Hongu valley and other southern areas and was manufactured from the bark of the daphne shrub. Butter, paper, and shere were all largely exported for use in Tibet's monasteries.^[8]

The degree to which Khumbu Sherpas were involved in trade to Kalimpong or Darjeeling in the nineteenth century is still unclear, but was very probably relatively minor. Sherpas began seeking income opportunities in Darjeeling by the 1880s and may have been frequenting the place still earlier. According to the 1901 Darjeeling district census, the first one taken, there were already 3,450 Sherpas living in Darjeeling (A. J. Dash, Darjeeling , 1947, cited in Ortner 1989:160), and they may have begun emigrating there any time after the British took control of the former Sikkimese village in 1835 and began developing it into a hill

Figure 17.
Changes in Sherpa Trade, 1920-1970

station, a tea-producing region, and the summer capital of the government of India. How many of these people were Khumbu Sherpas is not known.^[9] It appears likely that most Sherpas who made the move worked as laborers on road- and rail-building crews and in construction, porter, and rickshaw work (Mason 1955:157).

The iron trade declined early in the century (Fürer-Haimendorf 1975:62), presumably eclipsed by British export of iron over the new road from Kalimpong into southeastern Tibet. Although a small amount of iron continued to be traded over the Nangpa La even into the 1950s, primarily in the form of agricultural implements, trade had declined considerably by the 1930s.^[10] The livestock trade may also have declined slightly by the 1930s, as is suggested by the end of the keeping of large nak herds in Khumbu. The business of driving dimzo to Tibet, however, was still lucrative, and a few of the really prosperous Nauje traders of the 1940s and 1950s made their fortunes on dimzo. Salt also remained an important commodity until the early 1960s. Here too, however, a decline in the scale of the business was apparent as early as the 1930s

(ibid.). Increased salt imports from India had apparently begun to cut some of the demand for Nangpa La salt in the southern edge of its former range. There remained considerable demand for Tibetan salt, however, and it was the main source of prosperity for the wealthiest Khumbu traders through the 1940s and into the 1950s. A few big salt traders, primarily Nauje men, still made fortunes based on salt during the 1940s.

Throughout this period and until the mid-1960s the small-scale barter trade of salt for grain continued to be the most important trade of all for most Khumbu Sherpas. These families lacked the resources to transport salt on the scale of entire caravans and often dealt in just one or two loads. Many earned the cash to purchase a load or two of Kaprak or Ganggar salt by working for the big salt traders as porters. Others obtained their salt in Khumbu itself on loan, either getting the money to buy it from the Tengpoche monastery or being advanced salt by the big Nauje traders who would take their repayment later in grain. Most Sherpa families spent the winter months transporting this salt south into lower-altitude areas of Nepal as far east as Dingla (ten days from Nauje), as far west as Those (six days west of Nauje) and as far south as the confluence of the Dudh Kosi with the Sun Kosi (ten days from Nauje). These lower-altitude areas were unsuitable for yak, and few Khumbu Sherpas had any means to move salt or grain other than by foot, hauling thirty- to sixty-kilogram loads up and down valleys in baskets. Entire families joined in the work.^[11] Families made a number of shuttle trips to barter salt in the villages of Pharak, Katanga, Shorung, and the Rai country.^[12] Usually salt was bartered for grain, especially the cheaper maize, wheat, and millet. Those who could afford rice bartered for it rather than for the less-preferred grains. Families often went to the same villages year after year to trade. There they stayed with friends and benefited from their help to trade with their fellow villagers. Often traders established bonds of ceremonial friendship with these southern families, and such bonds were sometimes made with non-Sherpa Rais as well as with Sherpas. The exchange rate in the south varied considerably during the past century and in general the amount of grain offered for salt declined through this period. Yet the relatively small amount of salt required to obtain a given amount of grain in Kharikhola or Aislalukharka as opposed to what had to be given up for that much grain in Nauje made all the difference for families who had very few resources. In the south in the 1950s one might get three pathi of rice for a forty-kilogram load of salt when the rate in Nauje was only two and a half pathi. This difference in comparative value also led many Rais to transport grain to Khumbu to sell to families who were well-off and who had no interest in the back-breaking work of hauling salt and grain on foot

all winter.^[13] Rais have been coming to Khumbu for a long time to trade grain for salt. There are many oral traditions from the late nineteenth century that attest to this and to the way in which particular Rai families dealt with the same Khumbu family every year. Some of these Rais went to Thamicho, Khumjung, and Kunde to trade their salt, but most dealt in Nauje. The Indian explorer Hari Ram noticed a large number of Rais headed to Khumbu when he was traveling up the Dudh Kosi valley in 1885 and commented that much rice from the rich rice-producing Ra Khola and lower Dudh Kosi valleys ended up in Khumbu (Rawat 1973:162, 164). Elderly Thamicho residents remember that early in the twentieth century the Rais were so interested in Tibetan salt that if a sufficient supply was unavailable in Thamicho the Rais would wait for Sherpas to go to Tibet for more, camping out in Khumbu in the meantime. Many Rais came to Khumbu for salt every autumn and it was not uncommon for them to establish a trading relationship with a particular Sherpa family that lasted for several generations.

It appears as if the overall volume of trade to Tibet over the Nangpa La may have been greater during the first forty years of the twentieth century than it was during the late nineteenth century, although there are no reliable records to support this impression gained from oral history accounts. It seems that during the twentieth century more Khumbu Sherpas began trading at the scale that involved transporting bulk goods north with caravans of yak. Ortner has also suggested that an increase in trade from Solu-Khumbu in general took place during this period and has linked it to economic growth in Nepal and Tibet sparked by British colonial development and trade activities and by independent economic growth in the Ganggar (Tingri district) region of Tibet (1989:105-108). I am not so certain that the impact of the British on Sherpa trade fortunes was significant. I believe that it would be extremely difficult to document that the British Raj's economic and political policies had an impact on the Tibetan demand in Ganggar and Shigatse for butter, paper, shere, dimzo, and other goods traded by Sherpas (although they may have had with iron) or that they affected the supply and cost of these items in the regions of Nepal where Khumbu Sherpas obtained them. It may be that increases in trade, if there were indeed increases, were not the result of sudden new Khumbu wealth from Darjeeling connections or a vast new demand in Tibet for Sherpa trade goods, but rather the result of a general, moderate growth in the population and economy of both Nepal and Tibet, Khumbu Sherpas' increasing interest in different types of trans-Himalayan trade, and the gradual (often multigenerational) family development of the requisite experience, knowledge, contacts, and capital. Certainly from the 1940s to 1960s there were many examples of traders who over the course of their careers went from very

small-scale, relatively local trade to operating on the largest scale in the most lucrative and highly capitalized long-distance forms of trade.

In the 1950s a new generation of traders began to rise to prominence, many of them the sons of the great traders of the previous two decades. More families than ever seem to have begun trading over the Nangpa La, and in particular a high percentage of Nauje families who previously had just dealt in salt and grain between Khumbu and the lower regions of Nepal entered into the international trans-Himalayan trade. More traders seem to have been attracted to the import of luxury goods from Tibet and the large-scale export of paper, butter, and other goods not only to Ganggar but beyond to Shigatse. More traders went still farther afield as well to trade in Lhasa, Kalimpong, and Calcutta. The iron trade was over by this era and salt was not the business that it had been. There was still enough profit in salt and zopkio, however, that some traders who were familiar with these products and the necessary routes, markets, sources, buyers, and organization of transport remained content to continue with them. There was bigger profit, though, in trading butter, paper, and shere, particularly if one was willing to take it beyond Ganggar. The big traders of the 1950s set out from Khumbu with caravans of up to ninety yak loaded with these goods bound for Ganggar and Shigatse. Shigatse, second only to Lhasa as a Tibetan urban center, was reached either by a journey with pack yaks to Lhartse and then a five-day river journey by yak-skin and willow coracles or by a twelve-day mule-train journey by a more direct but slower land route. Both had their dangers: brigands, heat, wind, and poor grazing were problems on the overland route, whereas the river route ran the risk of a raft wreck. One Nauje trader remembers too well a cargo of twenty-six yak loads of paper lost on the river. But prices were better in Shigatse, double what were paid in Nauje for butter, paper, and shere delivered there on contract from the south.^[14] Some traders who were unable to sell out in Shigatse or were as interested in pilgrimage as in business carried on to Lhasa, by the late 1950s only a day away by truck on the new Chinese road from Shigatse.

During the 1950s a few traders, mainly from Nauje and Khumjung, extended trade further still, some leaving Khumbu on trading trips that only brought them back to Khumbu many months later. This trade mainly focused on Kalimpong. Traders would go north out of Khumbu with yaks and porters, trade in Ganggar, continue on by mule or boat and sell out their Nepal goods in Shigatse, take their cash and travel by foot to Kalimpong, buy trade goods there and return for more business in Shigatse, and finally head back for Khumbu with a few loads of jewelry, carpets, porcelain, silver, Chinese brocades, tea, and other precious things. In Kalimpong they bought Bhutanese silk (bure ), Indian

red and green wool for Tibetan boots (gonam ), incense, tobacco, turquoise, amber, zi stones, dyes, and cotton for trade in Tibet. Some traders went from Kalimpong to Calcutta by train in order to get lower prices and a better selection of some of these goods. A few Sherpas worked the Shigatse to Kalimpong trade for years at a time, returning only very infrequently to Khumbu. They took wool, horses, and other goods from Tibet south to sell in Sikkim and Kalimpong for cash to buy trade goods in demand in Tibet.^[15]

The Transformation and Decline of the Tibet Trade The place of trade in Khumbu economy declined dramatically during the 1960s. Several factors were involved. Probably the most important was the major political and economic change in Tibet following the flight of the Dalai Lama to India and the establishment of full-fledged Chinese administration. This affected the supply of trade goods and transport in Tibet, Tibetan demand for imports, Sherpas' freedom of movement within Tibet, and Sherpa traders' ability to maintain contacts and contracts with their Tibetan trade partners. But the war between India and China was also significant, as was the less dramatic but crucial undermining in Nepal of the demand for Tibetan salt due to the increasing availability of Indian salt and the gradual erosion of cultural biases against it. By the end of the 1960s trans-Himalayan trade was still being carried out by only a few small-scale traders. Today fewer than fifty men still go to Tibet each year, most of them either recent Tibetan immigrants, Bhote Kosi valley residents, or aging Nauje traders. Only one of the Nauje Sherpas who traded during the early 1960s was still going to Tibet in 1991 and he has not been able to interest his son in carrying on the business.

The transformation began in 1959 when the Dalai Lama and more than 80,000 Tibetans fled to India and Nepal. There was economic chaos in Tibet. Some Sherpa traders suffered considerable business losses when their goods were confiscated by Chinese soldiers or their Tibetan trading partners fled without fulfilling their contracts. As many as 5,000 Tibetans poured into Khumbu with their livestock, and for the next two years salt, sheep, and wool were so cheap in Khumbu that there was no need to make a journey to Tibet to obtain them. Many Sherpa built up substantial herds of yak overnight, so cheaply were these available, and gorged on one-rupee sheep. By 1962, however, the Tibetans had moved on to new refugee camps in Shorung, and Khumbu Sherpas were discovering that trade conditions in Tibet were considerably changed. Salt was scarce in Tibet for several years. The trade system that had brought salt from the Changtang lakes to the border region had apparently been

weakened by the massive population movements and socioeconomic reorganization in Tibet. Even when small amounts of salt could be obtained it had become more expensive and difficult to arrange transport south from Ganggar, for Tibetan yaks were not available for hire as readily as they had once been. Meanwhile villagers in the country to the south of Khumbu were overcoming their initial resistance to using the Indian salt that was then becoming more easily and cheaply available.^[16] By the time that Tibetan salt became more available there was no longer the market for it that there had been, although in Khumbu itself it remains preferred for many uses to this day. By the mid-1960s the salt trade to the Nepal midlands was finished.

At the same time long-established trade patterns were also being undermined in other ways. In Tibet the demand for some of the main Sherpa exports was also declining. During the 1960s and early 1970s thousands of Tibetan monasteries and temples were destroyed, and even before that many of them had ceased to function on their former scale after the mass exodus of 1959-1960. With the decline of monasticism there was a much diminished market for butter, paper, and shere. The war between China and India in 1962 led the Indian government to seal the border with Tibet and this abruptly blocked trade between Tibet and Darjeeling, Kalimpong, and Calcutta. And the terms of trade were beginning to change even as close to home as Ganggar. Sherpas were still free to come and go in Tibet, but the market for Khumbu livestock declined as collectivization spread and the Chinese encouraged local breeding of zopkio, banned the export of nak from Tibet, and limited wool exports.

Contrary to popular portrayals, however, Sherpa trade did not totally cease as a result of these difficulties. Some of the most prosperous Sherpa traders of recent times began their careers after 1959, whereas others who had been trading for many years continued to do so. Traders adjusted their scale of trade, product emphasis, and routes to new conditions. The eastern sphere of Sherpa trade, including Lhasa and Kalimpong, was abandoned, but Sherpas continued to trade in Shigatse until 1967. In 1962 a group of Sherpas drove more than 200 dimzo west across Tibet and into the Manang, Mustang, and Thak Khola regions north of the Annapurna range in central Nepal. Most trade in these days shifted from bulk goods to objects of great value and light weight. Some traders specialized in carrying artwork and religious objects out of Tibet. And in Tibet there was demand for certain luxury items, some of which may have been coveted by Chinese military personnel. Some Sherpas did a good business in Swiss watches.

The spread of the cultural revolution in Tibet, with the consequent suspension of normal law and order, made it dangerous as well as diffi-

cult to trade by the late 1960s. A few men tried to continue trading in Shigatse, traveling only under the cover of darkness. But they soon gave it up as too much of a risk. Many gave up trading altogether and few went beyond Ganggar. Some traders also experienced problems with the army and there were cases when goods were confiscated by border patrols. The Chinese had by now officially limited travel with a regulation that prevented Sherpas from going beyond the Ganggar area. Ganggar itself became off-limits once the Chinese had established a small trading post at Lungtang a few kilometers south of Ganggar. Sherpas who wanted to conduct business in the Tingri region had to spend their days and nights in this compound while they conducted business and to pay in the meantime for fodder for their pack stock. Business had to be carried out at fixed prices set by the Chinese, whose policies limited the goods available and those that were accepted in exchange. It was possible to trade with Tibetans, but only under Chinese supervision. If one wanted to sell a dimzo it was necessary to obtain permission first from an official in Ganggat. Prices were not what they had been nor was the quantity or quality of Tibetan goods. Yet still some money could be made by importing wool, Chinese tennis shoes, thermoses, flashlights, and other goods to Khumbu. Buffalo hides and some Indian goods were in demand in Tibet, and there was still profit in selling dimzo even when the only approved buyer was a government office. Only a few Khumbu men continued trading, but there was never a time when trade halted altogether.

China began easing trading restrictions between Tibet and Nepal in the early 1980s. Tibetan traders reappeared in Khumbu in 1983 for the first time in twenty years, and in December 1984 a group composed of Shato, Nejung, Penak and Kura men arrived in Khumbu with seventy-two yak laden with wool. Sherpa traders reported being able to barter now at the trade depot rather than having to accept fixed prices there, and they found that wool had become more available. Some had success in trading in the villages, although this was still technically not allowed. A few Khumbu Tibetan refugee traders began trading as far afield as Shigatse and a few Sherpas and Khumbu Tibetan traders began bringing a few more dimzo north. The refugee traders found it possible to sell crossbreeds again directly in the villages, although they were not clear on whether this was entirely lawful. One Nauje Sherpa family twice took more than 100 head of zopkio into Tibet for sale during the 1980s, but the total number of zopkio traded has remained far below 1950s' levels. Most of the current trade is in the hands of Tibetan families who settled in Nauje after 1959 or Bhote Kosi valley Sherpas who take advantage of their proximity to the border and large yak herds to import salt to Khumbu.

The scale of the Tibet trade today is very small. There is some demand still for Tibetan salt in Khumbu itself, but little is carried any further south. Wool is still good business and there is continuing demand in Tibet for zopkio, water buffalo skins, and some Indian and Kathmandu products including polyester, dye and snuff. The days of the big traders seem to be over, however. Few traders now are being succeeded by their sons, who find the money better and the life easier in trekking and mountaineering. Only Thamicho men still cross the Nangpa La in any numbers—fewer than thirty of them make the journey—and only one of the Nauje Sherpa traders is still going north. Unless new trade policies change current conditions in favor of trade over the remote passes such as the Nangpa La rather than via the main road between Kathmandu and Tibet, the long-vital Tibet trade seems destined to be nothing more than a dwindling relic in the Khumbu economy.

The New Regional Market System As the old international system of trade declined during the 1960s a new form of economic interaction was developed in Nepal as a result of government initiative. This was based on the tradition of periodic markets. In some parts of eastern Nepal there were long-standing periodic market systems, in which a once-a-week, open-air market was held at certain designated sites in a number of localities on different days of the week. Traders circulated through a region selling their wares at a series of different periodic markets, whereas farmers and herders now had a convenient outlet at which to locally market their surplus and entrepreneurs found that shops and teashops could profitably cater to the market crowds.

There had not previously been a periodic market system in the Dudh Kosi valley or anywhere in the Solu-Khumbu district where traditionally trade had entirely been conducted by individual traders. In the mid-1960s, however, weekly markets were established in Solu-Khumbu at government instigation. These included weekly Saturday markets at Dorphu (later Naya Bazar), near the district center of Salleri, and at Nauje as well as additional weekly markets at Aislalukharka in nearby Khotang district and at Olkadunga in the Olkadunga district. The Nauje market was initiated in 1965 and was soon established at its present site at the southwest edge of the village. The operation of this market soon totally replaced the earlier salt for grain barter trade that had been such a fundamental part of the household economy of most Khumbu families. By the late 1960s no Sherpas were going south in the winter any longer to trade salt for grain. Rai farmers and traders, however, continued to carry grain to Nauje as they had for more than a century, and in larger quantities than ever before. But after the establishment of the market

they now no longer bartered rice and other grains in Nauje for salt, but instead exchanged them for the rupees Sherpas had begun to earn in large amounts from tourism. The high prices at Nauje, where grain can be sold for triple what it would bring in Kathmandu, continue to bring many Rais and much rice into Khumbu. Today hundreds of hawkers converge on Nauje each Saturday with their wares, primarily rice and other grains, but also dairy products, eggs, fruit, vegetables, kerosene, cooking oil, and large amounts of candy, cookies, soft drinks, mineral water, and other goods destined for the tourist market. Goods are carried to Nauje from the road-ends at Jiri and Katari, respectively a week and ten days distance for a heavily laden porter, and rice is brought from places as far as ten days away.

Trade and Land Use Trade long made a few Khumbu Sherpas rich and provided families throughout the region with important supplies of grain that freed them from any necessity to achieve self-sufficiency in crop production. This had important implications for Khumbu agriculture which could be specialized to a degree otherwise impossible. The ongoing Sherpa involvement in extraregional trade networks made possible the increasing reliance on monoculture potato cropping, which supported more people than any other use of Khumbu arable land could. The ability to gain grain through trade removed the necessity to devote large areas of cropland to buckwheat. It also rendered unnecessary subsistence strategies based on the cultivation of a combination of high-altitude and middle- or low-altitude fields. Khumbu Sherpa trade activities in lower-altitude regions made it unnecessary for them to acquire land there and grow winter crops.

Very little Khumbu-grown produce was ever traded outside the region. A few dried potatos (shakpa ) were exported to Tibet and a few seed potatoes and dried radishes were sold to Rais in Nauje, but this was incidental. Lack of participation in a wider market economy has meant that the selection of crops and crop varieties has remained based on local preferences rather than on what Rais or Tibetans would buy. Khumbu crop production remained linked to local assessment of what is best to grow in terms of taste, yield, and risk in local environmental conditions.

Trade influenced pastoralism more profoundly. Pastoralism in Khumbu (or at least cattle raising) was not a matter of subsistence production for many owners. Certainly livestock products were valued for home use, and there was status and pleasure in keeping yak and nak. But the size and structure of the nak herds revolved around the sale of crossbreed calves to a significant degree, and changes in Khumbu pastoralism—including the major decline in nak keeping in the 1960s—may often have

been related to changes in zopkio trade opportunities. The high degree of emphasis on nak raising and the characteristic herd size associated with it reflect the requirements of crossbreed production for trade. The striking historical lack of emphasis on zhum as compared to cows may also be linked to trade conditions, in this case to lucrative opportunities to sell zhum calves to Shorung herders. The importance of yak keeping was also historically linked to trade, for the main use of the yak as a pack animal was for Tibet trading. When owning large herds of yak for transport became unnecessary for trading activities early-century Nauje traders ceased to keep them. These trade-related changes in the scale of nak and yak herding have also affected use of pastures. Regional decline in the numbers of yak and nak probably decreased grazing pressure on some of Khumbu's high-altitude summer pasture areas.

Tourism in Nepal

Nepal was once one of the world's most remote and most difficult to visit countries. Until 1949 the Nepal government rigorously excluded foreigners from traveling outside of Kathmandu and other than by the most direct route to the capital from India. Beginning in 1949 a few mountaineering expeditions were allowed to visit remote parts of the country to attempt ascents of the high peaks, but until 1955 Nepal continued to adhere to a policy of excluding foreigners which dated back to the establishment of the kingdom nearly two centuries before.

With the coronation of King Mahendra in 1955 there came the beginning of a new attitude. In that year the first tourist visas were issued, the first hotel was opened, and the first visitors began to arrive. But tourism was slow to develop. In 1962, the first year in which a count was kept, only 6,179 tourists entered the country. Since then mass tourism has found Nepal and the Nepal government has responded with increasing interest in tourism development. During the 1970s tourism surpassed first rice exports and then Gurkha remittances to become the country's leading source of foreign exchange. Kathmandu's Tribhuvan airport was enlarged to accommodate jets, the government supported hotel development in the capital, and Nepal became a major tour destination. The pace of tourism development increased markedly during the 1980s with a profusion of new hotels opened in Kathmandu at every level from budget to luxury. The number of visitors climbed from 162,897 in 1980 to 265,943 in 1988 (Nepal. Ministry of Tourism, Department of Tourism [1989]:17), before political friction with India in 1989 and political unrest in Nepal itself in 1990 led to a slight decline in visitors. Twenty-seven percent of all the international travelers who visited Nepal in 1988 came from India, while the United States, the United Kingdom, West Ger-

many, Japan, France, and Australia accounted for another forty-one percent (ibid.:13). The spectacular architecture, art, and living cultures of the Kathmandu valley continue to draw most visitors, but Nepal has also become a major adventure-travel destination during the past twenty years, offering mountain climbing and hiking, white-water rafting and kayaking, and wildlife viewing. Today the country is renowned especially for trekking, multiday hiking journeys on the extensive trail system that threads through what are still largely roadless mountains. Trekking offers a combination of adventure, closeup views of some of the most spectacular high country in the world, and direct contact with the land's diverse mountain peoples. These attractions have been skillfully marketed in Europe, North America, Japan, Australia, and New Zealand, and scores of trekking companies based in Kathmandu have prospered by catering to the international demand.

Tourism is today one of the leading sectors of Nepal's economy and a major focus of its development planning. Nepal continues to have one of the most pro-tourism policies of any country in Asia. The Ministry of Tourism favors further growth and to aid in this the government has recently relaxed its policies regarding foreign investment in hotels and transport. The official goal of current policy is to quadruple tourist arrivals to one million per year within the coming decade.

As Nepal tourism has increased in scale and diversified in scope, however, a number of potential problems have become visible. Thus far the economic gains from tourism have been highly localized, and in a large part investment and profit have remained largely restricted to the Kathmandu valley. The rise of mass tourism in Kathmandu has created small-scale "tourist ghettoes" of the "Freak Street" and Thamel areas and has supported the development of a smaller-scale "tourist trap" along scenic Phewa Tal lake in Pokhara. New social problems have developed in some areas, associated in part at least with tourism, including drug use, child labor, and professional begging. And a variety of economic, sociocultural and environmental problems have been reported in the scenic mountain hinterlands visited by trekkers. Only a relatively small number of Nepal's tourists take to the trails for multiday trekking holidays; in 1984 41,206 trekking permits were issued and the government estimated that about 15,000 tourists participated in trekking (Nepal. Ministry of Tourism 1985), and whereas by 1988 the number of trekking permits issued had climbed to 61,273, trekkers nonetheless constituted only 23 percent of the tourists who visited Nepal (Nepal. Ministry of Tourism [1989]:13, 50).^[17] Even relatively small numbers of tourists, however, have had significant local impact on remote mountain areas and their inhabitants.

Khumbu has been one of Nepal's premier mountaineering and trek-

king centers ever since these forms of tourism developed, and today the Mount Everest region and its famous Sherpas attract about 20 percent of all trekkers and nearly half of all mountaineering expeditions.^[18] The impact of tourism here has received more worldwide publicity than that in any other part of the country. At the same time Khumbu is the most striking Himalayan example of small-scale, locally controlled tourism development based on trekking tourism and mountaineering. Sherpas have responded to new opportunities in tourism with alacrity and adeptness and have prospered not only from earnings made from work as sirdars (responsible for managing porters, packstock, and camps), camping crews, cooks, and porters but also from their operation of strings of pack stock and owner-managed lodges, teashops, and shops. Sherpas have even begun establishing their own Kathmandu-based trekking agencies.^[19] They have become one of the most affluent of all high Himalayan peoples as a result. The remainder of this chapter traces the increasing importance of tourism in the Khumbu Sherpa economy and describes the way in which it ultimately has replaced trans-Himalayan trade as one of the fundamental components of the regional economy and household subsistence strategies.

Early Sherpa Involvement in Mountaineering

Sherpas have been associated with tourism since 1907, when several were hired in Darjeeling as porters by mountaineers who were making attempts on peaks in Sikkim.^[20] Within twenty years they had become world-renowned for their exploits on high mountains and particularly on Mount Everest. Sherpa high-altitude mountaineering porters soon developed a reputation for toughness, endurance, courage, and loyalty which made them preferred above all other Himalayan peoples as companions on the great peaks. They also began to earn considerable wealth as well as fame climbing on behalf of foreigners. Foreign mountaineering was forbidden within Nepal itself, but young Khumbu Sherpas could reach Darjeeling, the British hill station that was the major mountaineering center in the Himalaya until World War II, in only ten days of walking. It was from here that the British expeditions set out for Mount Everest, for in those days the only permissible approach to the mountain was via Tibet. Sometimes Khumbu Sherpas were able to make arrangements to sign on for these expeditions without going all the way to Darjeeling to do so. Several dozen Khumbu Sherpas, for example, simply crossed the Nangpa La and joined the 1933 Everest expedition at its Rongbuk (Rumbu) base camp. Some Sherpas lived this expedition life for only a few years, but others remained in Darjeeling for decades and settled

there permanently.^[21] Among these emigrants was former Bhote Kosi valley resident Tenzing Norgay, the Tibetan-born Sherpa who climbed Mount Everest with Sir Edmund Hillary in 1953. Those Darjeeling Sherpas who maintained ties with Khumbu sometimes returned money earned through mountaineering to Khumbu as remittances, and those mountaineers who ultimately returned to their native villages often did so as relatively wealthy men. This new money was used in old ways, to purchase land and livestock and to finance trade ventures. It appears to have had relatively little impact, however, on the regional economy before 1950. Trade remained the greatest source of Khumbu wealth.

Mountaineering Comes to Khumbu

In 1949 the Rana government allowed small teams of foreign mountaineers to explore several parts of Nepal and in 1950, the same year that a large French expedition climbed Annapurna, a small U.S.-British exploratory party received permission to travel to the Mount Everest area. Among the members of this group were two renowned mountaineers, Charles Houston, of K2 fame, and William Tilman. They reached Khumbu on foot from India, arriving in Nauje on October 1, and during the next two weeks camped their way across Khumbu to the Tengboche monastery and on toward the foot of Mount Everest. Tilman and Houston ascended a ridge of Pumori to a point at 5,545 meters which later became known as Kala Patar and thus became the first foreign visitors to enjoy what is now one of the most famous tourist vistas in the Himalaya (Tilman 1952). The route they had taken through Khumbu would one day become one of the major trekking thoroughfares in Nepal.

The following year a British team led by Eric Shipton (which included New Zealander Edmund Hillary) arrived for a detailed mountaineering reconnaissance of Mount Everest and made the first ascent of the dangerous Khumbu icefall. In 1952 Shipton returned, this time for an attempt on Cho Oyu, which was considered a training mission for the 1953 Everest expedition. The Swiss also were in Khumbu in 1952, sending large mountaineering teams for both a premonsoon and a postmonsoon attempt on Everest. The Swiss reached the south Col with the assistance of hundreds of Sherpa porters and a team of talented Sherpa high-altitude porters. Tenzing Norgay and Raymond Lambert very nearly reached the summit before abandoning the attempt. The British tried again in the spring of 1953 and, as all the world soon learned, Edmund Hillary and Tenzing Norgay reached the summit on May 29, 1953. This achievement by no means ended foreign mountaineering interest in Khumbu, and for the next eleven years there were generally one or two

European, American, or Indian expeditions each year that ascended the valleys en route to Everest, Lhotse, Cho Oyu, Ama Dablam, or other peaks. The worldwide publicity these mountaineering expeditions focused on the area and its people brought other foreigners as well. Special permission had to be obtained to visit the area from the government of Nepal, but nonetheless a few anthropologists and journalists began finding their way to Khumbu along with several expeditions searching for the yeti. Tourism, however, in the sense of a stream of individual or group sightseers, had not yet reached Khumbu.

Most of the early expeditions were massive affairs that had dozens of climbers and employed hundreds of porters. The 1952 Swiss team hired 163 porters, the 1953 British Everest expedition 450 porters, and the 1963 U.S. expedition to Everest a full 900 (Dittert, Chevally, and Lambert; 1954; Hunt 1954; Unsworth 1981). Wages were high compared to the money offered for agricultural day labor. Base camp porters made as much as seven times the daily wage of field workers and the high-altitude porters who carried loads up onto the mountain itself were still better paid. Work, however, was limited. Expeditions were few, and for years they only came to the region for premonsoon spring climbing. Each required help only for a few weeks to establish and later abandon its base camp. Only the high-altitude porters who stayed with the expedition climbers throughout their time on the mountain had the opportunity to make a good deal of money. For other Sherpas the arrival of an expedition was simply a chance to make a little money more easily than could otherwise be done locally. Many families turned out every family member old enough and fit enough to carry a load. From Nauje to Everest base camp took four days, and if one walked all the way to Kathmandu to meet the expedition there was perhaps another twelve days' work. It was not enough money to live on or to finance a major trading venture, but the income might mean that one need not work as an agricultural day laborer that year or as a porter for a Tibet trader at low wages in order to be able to buy salt for the next winter's grain trading. Pangboche villagers note that during these years they ceased to serve as porters across the Nangpa La for Nauje traders as they had formerly done (Stevens 1983:25-28, 1988b :72-73).

The Trekking Era

The government of Nepal banned foreign mountaineering expeditions from 1965 until the spring of 1969. Ever since then, however, mountaineering has flourished. The Ministry of Tourism handles permits for 104 peaks that are open for summit attempts by foreign mountaineering expeditions and collects peak fees.^[22] Between the autumn of 1987 and

the spring of 1990 there were an average of 101 expeditions per year, of which 43 percent climbed in Khumbu according to records kept by longtime Reuter's Kathmandu correspondant Elizabeth Hawley (personal communication). In some years more than fifty expeditions have attempted Khumbu peaks. In the spring of 1990 alone there were twenty-eight expeditions in Khumbu, including six Mount Everest expeditions and four Pumori expeditions, with a total of 163 foreign mountaineers and 136 high-altitude porters. Some of the expeditions of recent years have been large-scale affairs like many of the expeditions of the 1950s and 1960s; for example, the autumn 1988 U.S.A.-U.K. Everest expedition had eleven mountaineers and twenty-eight high-altitude porters. Many expeditions, however, are considerably smaller and employ only a few high-altitude porters, many of whom are non-Khumbu Sherpas and Tamangs. Over the years many Khumbu Sherpas have died in mountaineering accidents, and many young men and their families have decided that less dangerous careers in trekking are preferable to the glory and financial rewards that can come from climbing. Most of the Khumbu Sherpas who climb today as high-altitude porters are from Phurtse and Pangboche. There has also been a major decline during the 1980s in the numbers of Khumbu Sherpas who work as porters for mountaineering expeditions. Most porters today are Rais or other lowlanders. There is also a general trend away from moving gear by porters to using pack stock. Khumbu Sherpas have profited considerably from this.

During the mid-1960s, when mountaineering was banned, the Nepal government opened several areas of the country's mountains, including Khumbu, to visits by ordinary tourists. In 1964 it became possible to obtain a permit to walk along certain designated routes into areas far from Kathmandu. A new tourism industry began to develop to offer hiking tourists multiday camping trips in the mountains, and Himalayan trekking was born. Trekking companies such as the Kathmandu-based Mountain Travel began to offer guided camping tours of the highlands to adventurous, middle-class Europeans and Americans, smoothing their way with trained staff who catered to them in a style long-since developed on Indian Army marches and on mountaineering expeditions in Tibet and the Himalaya. These trekking groups traveled about the land like small, self-contained armies, with their porters to carry tents and provisions and a local crew to cook and handle camp chores. Among the authorized routes was that leading from the Kathmandu valley to Mount Everest, a journey of two weeks or more in each direction across more than 200 kilometers of rugged hill country and a series of seven passes. The 1964 building of a small, STOL airstrip at Lukla by a team associated with Sir Edmund Hillary's Khumbu aid projects, however, had already provided an easier way to approach Khumbu. With Lukla only a thirty-

five-minute flight from Kathmandu, one could be in Nauje the next morning. The first small commercial trekking group arrived in Khumbu in February 1965.

Trekking did not, however, immediately make the fortunes of its pioneering entrepreneurs. In the late 1960s it was an idea still a few years ahead of its time, and fewer than 300 of the nearly 35,000 tourists who visited Nepal in 1969 cited trekking as their purpose for visiting the country. But skillful promotion and a boom in "adventure travel" in the U.S.A. and Europe made trekking by the mid-1970s the big business by Nepalese standards which it has been ever since. By 1979 more than 25,000 trekking permits per year were being issued. Many of these tourists had the villages of Khumbu and a glimpse of Mount Everest as their goal. In 1964 only twenty tourists visited Khumbu (J. Fisher 1990:148). Seven years later tourism had increased to the point where during the fall season of 1971 and the spring season of 1972 a record 1,406 visitors arrived. The following year more than double that number came to Khumbu. Table 28 below illustrates the dramatic subsequent increase in the number of tourists.^[23] By the end of the 1970s more than 4,000 visitors per year were touring the area and in 1980 the total surpassed 5,000 for the first time. Since then the number of tourists entering Sagarmatha National Park has risen to more than 8,000 per year. This is a small number relative to tourist levels in mountain regions in California, Colorado, the Alps, or Peru, but it is substantial relative to Khumbu's population. Most tourists, moreover, travel along a narrow corridor to Mount Everest passing through an area that is inhabited by fewer than 1,000 Sherpas.

In contrast to earlier mountaineering tourism the increasing number of trekkers made possible a fundamental change in the Khumbu economy. Khumbu Sherpas became very involved in the trekking industry. For a number of years they secured a virtual monopoly on jobs throughout Nepal as camp crews, cooks, and sirdar, while many more worked in Khumbu as porters. The work was very nearly as well paid as expedition work and was not only considerably safer than climbing but was more widely available. Although it was possible to work for most of the year if one was interested in doing so, most Sherpas worked only during the four or five peak months of the tourist year, especially during October-November and April-May. During the late 1960s and early 1970s many families took advantage of the new opportunities for income, and prestige began to become associated with employment with particular companies as well as with particular jobs. By the late 1970s a study of household involvement in tourism in four Khumbu villages (Nauje, Kunde, Khumjung, and Phurtse) found that in three of the villages three-quarters or more of all families were involved in tourism, and that even

in relatively uninvolved Phurtse nearly half of all households had tourism income (J. Fisher 1990:115).^[24] During the next few years still more families became involved until by 1985 65 percent of all Khumbu families had income from trekking, and in all the settlement areas other than Phurtse and Pangboche the percentage of families with direct tourism

income (from all sources including pack stock) was above 80 percent. This degree of involvement in tourism is unparalleled in highland Nepal where it is more typical for the villagers of areas that are trekking destinations to secure very few trekking jobs.

These jobs, it should be noted, were almost entirely limited to work for men. For many years there were very few opportunities for women in trekking and even today there are only an extremely small number of Sherpa women who work as sirdar, camp staff (often called "sherpas" regardless of ethnicity), or cooks in the trekking business. More work occasionally as porters. These are mostly women in their late teens or early twenties, and usually they come from households in which there are no other family members who can make wage incomes from tourism. Women tend to avoid portering work whenever possible, except for a few young women for whom it is a welcome source of cash and adventure. During the past few years more young unmarried women have begun driving pack stock on treks, often with groups for which their father or elder brothers are employed as sirdar or camp staff.

With the increase in the scale of trekking in Khumbu came new Sherpa interest in establishing tourism businesses. In Pangboche it became big business to supply expeditions and trekking groups with fuel wood. Villagers from Nauje, Khumjung, and Kunde began establishing small lodges along the route to Mount Everest. A number of Nauje families (especially Tibetan refugees) opened shops. Many families in the southern Khumbu villages, those closest to the entrance and approaches to the region, began to invest in pack animals, for a single yak or crossbreed could carry double the load of a porter, and several animals could earn a single herder a considerable daily wage (Stevens 1983).

The growth in shops was limited primarily to Nauje. The first tourist shop was established in 1967.^[25] Today twenty-one of the twenty-nine shops in Khumbu are located in Nauje, flanking the two main streets of the village.^[26] Three-fourths of the Nauje shops mainly stock goods aimed at the tourist market, although many also sell some staple goods and clothing to Sherpas and a few specialize in sales to local customers. The goods offered to tourists consist of a very similar array of mountaineering clothing, food bought from expeditions or imported from Kathmandu, and Tibetan and Sherpa curios. A few shops have specialized in selling Tibetan handicrafts or renting and selling mountaineering gear. Both men and women operate shops and a number of Nauje shops are owned and operated by Tibetan refugees.

More significant for local development has been the proliferation of small inns. Sherpas, unlike some other Himalayan peoples such as the Thakalis, had no tradition of innkeeping. The first lodges of the early

1970s were nothing more than Sherpa homes with a signboard out front welcoming guests to a bed on the floor and the chance to join the family for Sherpa-style meals. The first Sherpa lodge, the Sherpa Hotel, was opened in 1971 in Nauje. In 1974 one of the more successful of Nauje's traders opened the Tawa Lodge, the first lodge built to be an inn. By 1979 a network of lodges was developing along the trails that lead to Everest, with concentrations occurring at the spots that had become popular overnight stops from the time of the first expeditions.^[27]

The total number of Sherpa lodges in Khumbu was relatively minor even in the late 1970s. In 1978, for example, there were nineteen in the region.^[28] During the following five years the number of lodges more than doubled to forty-five. Most of these were built along the route to Kala Patar, and many at already popular sites on that path. Nine new lodges were opened, for example, in Nauje and three in adjacent Chorkem. Some lodges, however, began to be developed in new sites between the main overnight stopping places, and secondary development along the Gokyo route became more significant with six new lodges.^[29] In the past few years both these trends have continued. The current regional pattern is indicated by map 21, which illustrates the location of the lodges opened by 1991 relative to the main trekking trails. Note the continuing small number in the Bhote Kosi valley north of Nauje and in the villages of Khumjung, Kunde, and Phurtse. Between 1986 and 1991 the number of lodges increased from sixty-two to eighty-one, many of them relatively simple lodges opened by Pangboche families in Pangboche, Dingboche, and Chukkung. The first Phurtse lodges were also opened. There were by this time also thirty teashops above Nauje which catered primarily to porters and trekking staff. In Nauje there was a score of such tea and beer shops, but many of them primarily did business with the Saturday market crowd and a number were only open for that event. Six more large Nauje lodges were under construction, two newly built ones were being operated, perhaps temporarily, as teashops, and two lodges were for the moment not being run due to extraordinary family circumstances. These ten establishments are not included in table 30 on page 365.

The growth of the local lodge system was supported by a new kind of tourism, independent trekking. The early lodges had a clientele of hardy souls interested in closer interaction with Sherpa families, a more authentic village experience, and freedom from organized group travel. Word of the possibility and delights of lodge-hopping through Khumbu (including the fact that even in 1990 it could be done for five dollars per day) spread by word of mouth, guidebook descriptions, and travel articles. A greater number of travelers came to the region on their own (among them many people who had originally toured the area with a trekking group) and virtually all of these independent trekkers traveled without

Map 21.
Sherpa Lodges, 1991

tents. By 1988 more trekking permits were issued to individual travelers to visit Khumbu (5,698) than were handled by Kathmandu trekking agencies (5,668) (Nepal. Ministry of Tourism ca. 1989:52-53).

Increasing numbers of guests, income, familiarity with hotel-keeping practices in Kathmandu and elsewhere, and money and advice from

foreign friends enabled Sherpas to begin developing more sophisticated lodges during the 1980s.^[30] These have primarily been developed in Nauje where they have virtually captured all business from those earlier lodges that have not upgraded. The new lodges bear little resemblance to Sherpa houses—the main Sherpa element in the architecture, indeed, is merely the use of stone wall construction. Lodges now are multistoried, and three of the new Nauje lodges have four stories. Dormitory sleeping and dining rooms are standard and some lodges are beginning to add private double rooms. In 1983 the first lodge opened in Nauje with a top story dining room with large windows surveying the village and peaks. By 1986 there were seven, and one had been opened at the Tengboche monastery (Stevens 1983, 1988b ). In the spring of 1991 there were eleven such lodges in Nauje, with two more currently being operated as local teashops and another two under construction. There were also two in Pangboche and one in Kunde as well as the earlier one at Tengboche. Some of these lodges were primarily financed with money saved from years of trekking work and smaller lodge operations, but most of them enjoyed substantial support from foreign benefactors in the form of gifts and loans. U.S., Japanese, and European money has helped build lodges in several Khumbu localities.

Another outstanding feature of the changing economy of Khumbu during the tourism era has been a tremendous increase in the number of livestock kept as pack animals. Until the 1970s yak were the most popular pack animals. In recent years, however, urang zopkio have become the principal means of moving expedition and trekking loads. Urang zopkio are preferred over yak for trekking and mountaineering work because they are able to haul gear on the Lukla to Nauje route as well as in the higher valleys. Sherpas believe that yak are not suited for the trip to Lukla because this requires them to be taken to altitudes below 3,000 meters and it is feared that the animals are susceptible to serious disease at such low elevations. Urang zopkio labor under no such handicaps and they are enormously profitable. A single urang zopkio carries two loads and earns more for his owner than a Sherpa earns per day as a porter. At 3,000 rupees (more than a months' salary for many sherpa), purchasing an urang zopkio is a substantial investment, but is a far more affordable one for most Khumbu families than establishing a shop or lodge. Urang zopkio keeping has become one of the most popular forms of investment in Nauje, Khumjung, Kunde, and the Thamicho villages. Families that have one or more members working as sirdar on Khumbu treks are especially likely to invest in urang zopkio. Since sirdar make the arrangements for pack animals and porters their own animals are assured of considerable work. By 1986 70 percent of Nauje families with sirdar positions owned urang zopkio.

The total increase in urang zopkio numbers in Nauje is astounding. In 1963 only five families kept urang zopkio pack stock totaling fewer than twenty head.^[31] By 1986 there were more than 140 head owned by fifty-two households.

The Transition From Trade to Tourism

After visiting Khumbu in 1971 Fürer-Haimendorf came away with the impression that trade had more or less ceased after 1959 and that the resulting void in the Khumbu economy had fortunately been promptly filled by the rise of tourism. He suggested that:

Had the Chinese stranglehold on the Sherpas' trade occurred even twenty years earlier the effect on their standard of living would have been catastrophic. Indeed it might well have caused a depopulation of the region of high altitude where farmers and pastoralists can only subsist if their income is supplemented by outside earnings. Fortunately for the Sherpas the checks on their trading activities imposed by political events in a neighbouring country coincided with the opening of Nepal to foreigners and the subsequent development of mountaineering and tourism which soon became major sources of income benefiting the inhabitants. (1975:3)

Yet, although this general process did occur, the transition was not nearly so timely and smooth. When the decline in trade culminated about 1967 there were a few years of regional doubt and concern. Mountaineering had not shown the capacity to be a major foundation of local economic growth during the previous seventeen years and was in any event at that time banned altogether with no assurances of when, if ever, it might be resumed. Trekking tourism had begun, but in the first years it was not the substantial phenomenon that it became by 1971. In the early days it must have seemed primarily a young man's field, work fit for the few Sherpas who had previously worked for the expeditions but hardly appealing to men who had spent their lives trading in Tibet and India. During the 1960s some men indeed began to abandon Khumbu to move to Kathmandu or to Darjeeling where they tried their luck at road contracting and other ventures. The rise of trekking tourism during the early 1970s slowed this emigration, although there remained a steady flow outward of many of the brightest and best educated to Kathmandu, where a number of Sherpas ended up working in trekking agency offices. With the increasing scale of Khumbu tourism in the 1970s and, more importantly, with the rise of both commercial group and individual trekking there were abundant jobs in the new service sector as well as opportunities for entrepreneu-

rial activity that involved, in the early days at least, relatively minor commitments of cash. A new era in the local economy began as former traders opened lodges and shops, sirdar began to prosper, and most Khumbu households began making enough money from tourism to pay for the grain they had once earned through trade.

10
Tourism, Local Economy, and Environment

The rise of tourism as a major global phenomenon has made it one of the most important sources of international economic development and cultural exchange. It has also been associated in both the developed and the developing world with a wide range of adverse economic, sociocultural, and environmental impacts. Even extremely small-scale tourism can precipitate adverse impacts, especially in areas that have not been previously integrated into the global economy, have had little interaction with Western cultures, or are located in sensitive environments.^[1] During the past few years international organizations, governments, citizens' groups, and local communities (as well as members of the tourism industry) have become increasingly concerned with the adverse effects of recreational, nature, and cultural tourism on the very ecosystems, peoples, and places whose special character has made them tourist attractions.

This chapter explores the role of tourism in recent change in Khumbu's economy and environment. I examine the past and present role of tourism in the Mount Everest region in changing local land and resource use and the concomitant environmental impact which this has had to date.^[2] It is beyond the scope of this book to consider the wider cultural and social impacts of tourism, as significant as these are, to analyze the effectiveness of tourism in Khumbu thus far as a path for local economic development, or to present recommendations for more culturally and environmentally sensitive tourism and tourism planning.

Tourism and Subsistence

In the preceding chapter I suggested that for many Khumbu families tourism has taken the place of trade in their household economy. Like trade it has offered villagers the means to diversify their economic activities and to decrease their dependence on Khumbu resources for their basic sustenance. Tourism has provided opportunities for a range of different types of involvement. The equivalent in tourism to the common pre-1965 small-scale salt for grain barter trade is the widespread practice of working in trekking and mountaineering for a only a few months each year in order to make wages that are then primarily spent on grain and other food purchased at the periodic market.^[3] But, like trade, tourism also offers more profitable avenues of activity to those interested in investing capital and running a more complex business operation. Here tourism has provided a range of options from the ownership of a few head of pack stock to the operation of shops and lodges, and even the establishment of Kathmandu-based trekking companies.^[4]

Tourism is also similar to trade in its role in the Khumbu economy in that tourism employment and entrepreneurial enterprises can be integrated into lifestyles in which much time and attention are also given to subsistence household agropastoral production and small-scale, commercially oriented stockraising. Tourism has been fully integrated into longstanding Khumbu subsistence strategies. During these first decades of tourism development, at least, even families making considerable income from lodge operations have all continued to farm and to devote most of their land, as before, to achieving self-sufficiency in tubers and, if possible, in buckwheat and barley. This has also remained the practice of the many families who are involved in tourism through wage making only. Tourism has become the major source of regional income and probably directly or indirectly accounts for more than 90 percent of all the money earned in wages and business profits in the region.^[5] Yet it remains, like trade before it, for nearly all families a supplementary activity which is integrated into a basic subsistence economy and which is less important in terms of household labor and time than crop production and pastoralism for household consumption. That this remains true reflects cultural values as well as economic conditions and is probably related to concepts of identity as well as to local perspectives on diversification, gender roles, and social interaction.

Tourism, Affluence, and Differentiation of Wealth

Differentiation of wealth is nothing new in Khumbu, as I discussed in chapter 2, and the region has long been relatively more wealthy than

many other nearby areas of Nepal. Differences in regional, village, and household wealth made for differences in standards of living (shelter, diet, clothing, conspicuous consumption, leisure time, cultural pursuits), status, political power, and land use. The rise of tourism in the regional economy, however, has increased the overall affluence of Khumbu society and widened the gap between it and other regions. It has also accentuated disparities in wealth among Khumbu villages and among households within villages. This has contributed to new patterns of land and resource use and these in turn have also placed new pressures on some communal resource-management institutions.^[6]

A variety of employment opportunities in tourism are possible in Khumbu today. Some are strictly day-labor wage work. Others are careers for which men receive year-round monthly salaries and benefits. There is a definite hierarchy of jobs in terms of both pay and status. The range of employment also often represents a hierarchy of experience with young men moving up a ladder beginning as porters or kitchen boys and ultimately rising to the better paying and more prestigious jobs of camp-crew sherpa, H. A. (high-altitude mountaineering porter or sherpa), cook, or sirdar. Today most men specialize in either mountaineering or trekking work, although a few pursue opportunities in both.

Mountaineering provides short-term day labor for large numbers of local porters who carry loads only as far as base camp. This was once an important source of income for both Khumbu men and women, but today it is very rare for Khumbu Sherpas to work as base-camp porters. Mountaineering also offers a relatively smaller number of other jobs, which range from the base-camp and climbing sirdar and the team of H. A.s who carry loads high on the peaks to cooks, cooks' assistants, and mail runners. Sirdar are the elite of mountaineering crews responsible during the approach for the logistics of the camps and the organization of the porters, for the efficient performance of the H. A.s on the mountain, and for the operation of base camp. On large expeditions there are often two sirdar, one in charge of base-camp operations and the other in charge of the high-altitude porters on the peak itself.

Trekking work provides a similar set of jobs with the exception that there is no need for mail runners or H. A.s whereas there are more opportunities for camp staff. Camp staff are today usually referred to as sherpas in the trekking industry, even though increasingly men of other ethnicities are filling this position. They are responsible for making and breaking camp and often accompany trekking group members as personal guides and companions. Typically there is one sherpa for every four trekking clients.

Pay varies with position both in terms of wages and other benefits. Porters are the least well paid in this larger sense, although in Khumbu they can make relatively good wages. In 1990 porters received from

120-250 rupees per day for work in Khumbu. This was without any additional allowance for obtaining food and shelter. The large range in pay rate reflects differences in supply and demand, tourists' itineraries, and the bargaining skill of the parties involved.^[7] Individual trekkers in particular often pay quite high wages for porters since they tend to have little experience with hiring them, are unfamiliar with local rates, and often are given little choice by porters well aware of how short the supply of people willing to perform such work often is at the Lukla airstrip, the Jiri roadhead, and in Khumbu itself. The local rate Sherpas paid in autumn 1990 for porters was 120 rupees per day (up from one hundred rupees per day the previous year). Trekking groups paid 150-160 rupees per day without food. At these wage rates a porter can make a higher daily wage than men in the more prestigious trekking and mountaineering positions. They seldom, however, receive the substantial amount in tips and gifts that sherpas, cooks, and sirdar typically are given.

High-altitude mountaineering porters make between seventy and one hundred rupees per day whereas mountaineering sirdar probably average one hundred rupees per day. Both are also given an issue of equipment that can have a resale value which is much higher than the wages themselves.^[8] Beyond this direct income many high-altitude porters and mountaineering sirdar who have attained the summits of major peaks have been awarded with long-term trips abroad. During these foreign trips they are often offered opportunities to work at wages which allow them to return to Nepal with considerable savings by local standards.

Trekking cooks typically made seventy to ninety rupees per day in wages in 1990 and cook-crew members fifty to seventy rupees per day. Trekking sherpas usually received slightly more than cook-crew members and slightly less than cooks. Pay varied with different trekking companies, the lowest-paying companies offering sixty rupees per day for sherpas in 1990 whereas the highest paying paid seventy to eighty rupees per day. Food and shelter were provided during the trek free of charge to all of the cook crew and sherpas as well as to the sirdar.

Trekking sirdar make different wage depending on whether they are "company sirdar" on contract to a particular trekking agency or freelance guides who work for different companies for short periods or even attempt to recruit trekkers more directly in Kathmandu or Khumbu lodges. Sirdar's pay also varies with their degree of experience, connections, and command of foreign languages. If they can perform as a group leader as well as sirdar, which requires considerable interaction with tourists as guide and counselor, they make considerably higher wages. Sirdar hired for an individual trip may be paid as little as seventy-five rupees per day plus food, whereas others on company contracts receive

a monthly stipend of 1,700 to 3,000 rupees per month year round for four to six months of work per year.^[9] A few of the finest group leaders may be paid by foreign salary standards and may make as much as fifty dollars per day, thus earning in two days the monthly salary of the highest-paid company sirdar.

Both trekking and mountaineering sirdar are also able to greatly augment their salary through money made from the recruitment of porters and pack stock and negotiations for campsites, fuel, and food. On all of these there are profits to be made from overreporting expenses to Kathmandu trekking companies and often also from kickbacks and other offerings. A sirdar might, for example, report to Kathmandu that he had hired porters for 130 rupees per day when he had actually obtained them for 100 rupees. This can very quickly add up to a substantial amount of money. Porters may also be expected to secure their jobs with a gift of some sort to the sirdar. All kinds of charges may be over-reported, from fuel wood and camping fees to food expenses. These practices are widespread and can easily enable one to triple one's salary. When a trek or expedition takes place in Khumbu itself much more money can be made. One may choose to camp on one's own lands and those of relatives. Family members, household servants, and relatives and friends may be hired and one may use one's own pack stock rather than recruit porters. At 1990 rates of eighty rupees per load per day (160 rupees per zopkio), a man who could mobilize five or six of his own yak or zopkio could make as much money from this sideline in a month of trekking as an average sirdar's salary would bring him in a year.^[10] Mountaineering sirdar who are in charge of large expeditions have the same opportunities on a far vaster scale, for they are responsible for arranging for hundreds of loads to be carried, for hiring dozens of high-altitude porters and a sizeable cooking staff, and purchasing enormous amounts of food and fuel. Some mountaineering sirdar have greatly increased their incomes by making advance preparations for expeditions. It is then possible to use their own livestock to shuttle tons of equipment and food to base camp. In one case a few years ago a Nauje sirdar who owned more than ten yak and zopkio transported more than four hundred loads to Mount Everest base camp from Lukla with the help of his son. At today's rates this would bring them over 200,000 rupees. Even in the mid-1980s, when prices were much lower, it was said that a sirdar on a major mountaineering expedition could make 100,000 rupees or more if he was skillful at the business side of his craft.^[11]

A number of trekking sirdar and group leaders, like mountaineering sirdar and H. A.s, have been awarded with trips abroad by grateful clients and foreign friends. Some Sherpas have made repeated trips to Europe, Japan, and the United States in this way. The contacts devel-

oped over a trekking and mountaineering career also bring other benefits such as scholarships for one's children and low-interest or no-interest loans for building houses or starting businesses. Sherpas are very aware that the relationships they develop with tourists may change their lives and some carefully court possible patrons and benefactors.

There is also money to be made in tourist businesses from selling stocking hats at 100-120 rupees apiece to operating shops, lodges, and pack trains. The most lucrative business of all is owning and managing a lodge. All Khumbu lodges are run as family operations. The price of a bed is kept extremely low out of fear that anything more than minimal prices will scare away business. A bed in a dormitory runs from five to ten rupees, fifteen to thirty cents a night. Even a double room seldom exceeds fifty or sixty rupees, less than two dollars a night. Lodges do not make much money on rooms even by local standards of income. There are considerable profits in selling meals, however, since there are no restaurants in Khumbu and guests normally take all meals in their lodge (or pay a penalty of a higher lodging rate if they do not). It is common for guests to spend five dollars a day on food, and those with a beer or soft-drink habit may easily spend a great deal more at three or four dollars a bottle for beer and one to two dollars for a bottled soft drink. Prices are generally fairly similar among lodges in a given area, but increase with distance up valley from Nauje. Five dollars per day in basic food costs seems rather little, but it is nearly twice the daily wage of most trekking Sherpas. Many lodges serve at least ten tourists per day during the trekking season and more popular places average twenty tourists per night and still more for lunch.

All the cooking and other lodgekeeping tasks are generally done by family members if the business is of modest scale. If business booms families hire low-paid servants to help with gathering fuel wood and water, cooking, serving, and cleaning. Some of the largest lodges now have four servants working for them and in Nauje even small lodges have two. With these low labor costs it is not hard to understand why so many Khumbu families have been interested in going into the lodge-keeping business. Shopkeeping has been much less attractive. Most of the region's shops are operated by Tibetan refugees or as side businesses by families who keep lodges.

Since tourism arrived in Khumbu in 1950 wage labor rather than entrepreneurial activities has been the main source of Sherpa income from tourism. Over the years the nature of Khumbu Sherpa participation in tourism work has changed dramatically. Until the 1970s the majority of Khumbu Sherpas who worked in tourism held jobs as porters, either as local porters who carried loads to mountaineering base camps and accompanied trekking groups or as high-altitude porters on expeditions.

Only a few men had jobs as mountaineering sirdar and there were very few trekking sirdar before the trekking boom of the 1970s. During this era the most coveted position in tourism other than sirdar was undoubtedly that of the high-altitude porter. Mountaineering Sherpas risked death and debilitating injury, but the rewards were substantial. Until the 1980s Khumbu Sherpas filled almost all these positions on expeditions in Khumbu as well as other parts of Nepal. On expeditions and treks within Khumbu, moreover, nearly all mountaineering and trekking sirdar, cooks, and camp staff were Khumbu Sherpas. With increasing affluence and contacts in the mountaineering and trekking businesses, however, Khumbu Sherpas have begun to concentrate on better paying and less dangerous jobs, leaving the others for non-Khumbu people. Very few Khumbu people work as local porters anymore, and in Khumbu what is not transported by zopkio or yak is today mainly carried by Rais.^[12] Fewer young men have been drawn to high-altitude-porter work as well in recent years. On the major U.S. Everest expedition of 1963 all thirty-two high-altitude porters were Sherpas (twenty-five of them from Khumbu, four from Darjeeling, and three from Solu). The 1988 French Everest expedition, by contrast, employed twenty-eight high-altitude porters, of whom twenty-six were Sherpas but only seven were Khumbu Sherpas (J. Fisher 1990:123). Most of the young Khumbu men who work today as high-altitude porters come from Phurtse.

The annual income Khumbu families make from tourism varies considerably as a result of differences in wages and other income made on treks and mountaineering expeditions and from tourism businesses. It also varies with the number of family members who work in tourism and the number of months in which wages are earned. The amount of time men commit to trekking and mountaineering work varies a great deal, and in general is rather less than the total length of the tourist season.^[13] In Nauje it is typical to work only four or five months per year at trekking or mountaineering, although a few men work eight or even ten months per year. Some trekking sirdar who own no livestock or who mainly trek in areas outside of Khumbu make as much as 7,000 rupees per month between their salaries and other money made through their position when they are trekking, and 1,700 to 3,000 rupees per month for the rest of the year from their salary. This gives them a minimum annual income of more than 41,600 rupees (more than $1,400), a fortune by Nepal's national standards. Those who own zopkio or yak can make much more than this. A man who works as a sherpa, however, can expect to make as little as 7,200 rupees a year for four months of work. Income from lodgekeeping can also obviously vary tremendously. One Nauje lodge that today does a relatively small business takes in about 70,000 rupees per year. A lodge that averaged twenty guests per night

over the entire eight months of the tourist season would very likely have gross receipts well over ten times that amount.

Khumbu Sherpas' relative affluence compared to other groups in the Solu-Khumbu region and adjacent districts has had important effects on regional labor migration and on the development of regional trade patterns. As already discussed, many young men and a few young women from other parts of the Solu-Khumbu district and neighboring areas have been drawn to Khumbu by the wages offered to household servants. And during the past fifteen years increasing number of Rai men and women have also come to the region for seasonal field and forest work. The money these migrant workers and servants can make in Khumbu is minor by Khumbu Sherpa trekking, mountaineering, and even agricultural day-labor standards, but this income is substantial in comparison to what can be made in their home regions. Many of the young men, moreover, are drawn by the possibility of beginning a trekking career and ultimately making larger incomes. They take positions in the households of Khumbu sirdar at low wages, or in some cases for no wages at all, in order to be able to accompany the head of the household on his trekking trips as a porter, pack-stock driver, or a member of the cook or camp crew. It is possible to earn substantial money by the standards of the Nepal midlands from such work even if one must share a percentage of it with one's benefactor and work the rest of the year as a household servant for nothing but food and a place to sleep.

Considerable differences in relative affluence exist within Khumbu itself. Some villages are getting steadily richer while others share more modestly in the new wealth from tourism. Nauje, Kunde, and Khumjung have far higher standards of living than the other villages. A half a day's walk in other directions leads to villages where rice is still only eaten on very special occasions and polyester dresses, stereos, and corrugated metal house roofs are only seen on market day trips to Nauje. And even within villages the contrast in lifestyles is tremendous between sirdar and lodgekeeping families and the quarter of Khumbu's Sherpa families who have little or no tourism income.

This regional differentiation in wealth is related to differences in the percentage of village households involved in tourism, the types of jobs they hold, the number of multicareer families, and the percentage of village families involved in keeping lodges and shops. Table 29 illustrated the contrast among villages in terms of the percentage of families who have income from tourism, including lodges, shops, and pack stock. The differences between Nauje, Khumjung, Kunde, and the other villages are striking. Further grounds for differences in village wealth become apparent from table 31 which illustrates the ways households from

different villages are involved in trekking tourism. In Nauje, Kunde, and Khumjung a high percentage of those families who have trekking work have the highest-paying jobs available. Note especially the number of families having at least one member working as a sirdar and contrast this to the situation in Phurtse, Pangboche, and the Thamicho settlements. Here few families have a sirdar's income and many simply have what cash they make by hiring out some of their livestock for a few weeks per year as pack animals. By 1991 the contrast among villages had narrowed slightly as more Pangboche and Phurtse men improved their career positions to sherpa and sirdar. By 1991, though, there had also developed a contrast among Nauje, Khumjung, Kunde, and the other villages in the number of families in which two or more members were working in trekking.

The extent of ownership and operation of lodges has also long varied among villages. This was especially true in the 1970s and early 1980s, as can be seen in the figures for 1983 in table 32. Nauje families have for many years taken advantage of their village's site on the main route to Mount Everest and have also opened lodges further along the route in places such as Lobuche and Dingboche. The development of lodges in Kunde and Khumjung has been greatly inhibited by their location less than two kilometers from the major tourist center of Nauje and off the main trail to Mount Everest. As of spring 1991 there were only two lodges in Khumjung and one in Kunde, all opened only in the past few

years. Families from those two villages, however, have converted many herding huts on the high routes to Gokyo and Mount Everest into simple lodges. Villagers in Thamicho and Phurtse who own no land on the main tourist thoroughfares have had fewer opportunities. The two lodges in Thami Og have been opened relatively recently and do a light business. For many years there was no lodge in Phurtse. Today there are only two very basic ones in the village and Phurtse families have now opened a few humble teashop-lodges elsewhere in the upper Dudh Kosi valley. There were also few lodges in Pangboche until the late 1980s when several large, new lodges were built in the village and more local people also began to convert their herding huts in the upper Imja Khola and Lobuche Khola into lodges. By 1991 the relative involvement of the different villages had thus narrowed considerably.

The late involvement by Phurtse, Pangboche, and Thamicho villagers in the lodge business may be due to several factors. Location on the main route to Mount Everest may have been of especially great importance in the 1970s when far fewer trekkers visited the upper Dudh Kosi valley. The lack of a strong entrepreneurial tradition in Phurtse and Pangboche, where only a very few families traded with Tibet on even a small scale, may have been a factor. The relative lack of available capital in the poorer villages may well also have been important. It is also possible that villages in which few families had worked in trekking could also have been at a disadvantage because fewer families would have had the well-developed language, social, and culinary skills that are important for running a lodge. Few families in these villages, moreover, could have had the foreign contacts so often useful for obtaining loans to establish lodges. Village differences in the extent to which families set up lodges, like differences in types of tourism employment, could also reflect different levels of land and livestock wealth and contentment with agropastoral lifestyles. Pangboche, Phurtse, and Thamicho are all regions where a great deal of nak pastoralism is still practiced and where many families have relatively large land holdings. Families from these areas were not only slower than those from Nauje, Khumjung, and Kunde in opening lodges but were also somewhat slower in taking up work as sherpas and sirdar in the trekking trade. Villagers preferred, it seems, the more short-term time commitments of working as porters and H. A.s.

This regional differentiation in wealth seems likely to continue to deepen. It already has reached a level of inequality that probably surpasses the differences among villages during the trading era. In the pre-1960 era Nauje was noted as the wealthiest of the villages, for it was there that most of the big traders lived. There were many poor families in Nauje as well, however, and the overall contrasts between that village

and the others were not as pronounced as they are today. Within Nauje the differences in wealth that once existed between the very few major trading families and the rest of the village have perhaps narrowed in the tourism era when a higher percentage village families have a major source of income. Families who own popular lodges, however, are becoming wealthy to an extraordinary degree in comparison to their fellow villagers. It remains to be seen what long-term repercussions intra and intervillage economic differentiation will have on Sherpa society. There is a strong possibility that a new elite will develop in Khumbu, composed largely of families wealthy from sirdar work and lodge operations. There is not room for everyone to operate an lodge, and the costs of entering the business are becoming prohibitive for most families now that standards are rising everywhere and being competitive in Nauje requires a multistory lodge having a restaurant with a view. It is the sirdar and the lodgekeepers who have the money for new business enterprises and for sending their children to the upper classes of Khumjung's school or to boarding schools in Kathmandu or Darjeeling. These are the families, moreover, who gather most of the gifts and scholarships donated by philanthropic tourists who tend to be most generous toward the families of their guides and hosts. All of this will probably continue to translate into status and power as well as increasingly different lifestyles.

Monetarization, Social Relations, and Inflation

The rise of the tourist economy in Khumbu has brought a great deal more cash into the region than there had ever been in the past. To some extent this has fostered a shift toward cash-based economic exchanges. Grain and other goods from lower regions of Nepal that were once obtained through barter trade are now paid for in cash. And while labor in Khumbu was formerly paid primarily in butter, potatoes, and grain this has now given way to cash wages, although the custom of payment in food has not entirely disappeared. This increasing monetarization of Khumbu life, however, has not entirely permeated the local economy and social interaction, much less fundamentally altered cultural values. Reciprocal work groups are still, for example, very important in all the villages. Mutual aid groups continue to work alongside hired laborers in building houses and communities still maintain trails, temples, and ceremonies through the contributions of unpaid labor from all village families.

The degree of monetarization of the economy that has occurred has brought negative impacts, including inflation.^[14] The price of rice has increased tremendously, from nine rupees per pathi in 1964 to twenty-six rupees ten years later, and thereafter rising more precipitously to thirty-

five rupees in 1978 and ninety rupees in 1988 (J. Fisher 1990:116). In autumn 1990 rice cost ninety to one hundred rupees per pathi depending on quality. This is more than three times the price of rice in lower-altitude areas of the Solu-Khumbu district or Kathmandu. These prices are not the result of direct buying by mountaineering expeditions, trekking groups, or lodges, for tourists consume a relatively minor amount of rice, maize, millet, or buckwheat. They are an indirect impact of tourism, however, for they reflect Rai traders' perceptions that the now-affluent Sherpas can be made to pay high prices.^[15]

Tourism has probably had some role in increasing the price of Khumbu-grown potatoes as well, although these have not inflated in price quite as much over the years as rice has. Tourists do eat a considerable amount of potatoes, and a relatively popular lodge may require as many potatoes during the eight-month tourist season for its guests as the family itself consumes in a year. Lodges that take in large numbers of guests may require double the family's subsistence requirements of potatoes in order to feed them. Across Khumbu, with its more than eighty lodges, the added demand for potatoes may thus be substantial. It is possible that tourists now account for over 10 percent of the total potato consumption in the region. The increasing quantities of potatoes required as tourist numbers rose over the past twenty years may well have contributed to the rise in the price of Khumbu-grown potatoes. The cost of potatoes rose from two rupees per eleven-kilogram tin in 1964 to fourteen rupees per tin by 1974 and twenty rupees per tin in 1978 (ibid.). By 1986-1987 potatoes sold in Nauje for twenty-five to thirty rupees per tin and in 1991 were selling there for thirty-five rupees per tin. Potato prices have been pushed much higher by tourism in some parts of Khumbu, particularly in the upper Imja Khola and Lobuche Khola valley. Here some lodgekeepers prefer to buy potatoes from Dingboche and Pangboche farmers as this makes transport much easier than from Nauje. Prices here reached fifty-five rupees per tin after the harvest in 1990 even though it had been a relatively good harvest in Dingboche, and by spring 1991 a tin cost seventy rupees.^[16] Increases in potato prices over the past fifteen years would undoubtedly have been much greater but for the great increases in regional production due to the widespread cultivation of high-yielding varieties. Without this reliance on new varieties it is doubtful whether regional production would be sufficient for both feeding the tourist population and meeting Sherpa needs.

Tourist demand has had a much greater and more direct impact on the market price of important regional commodities such as kerosene, eggs, cooking oil, fruit, vegetables, and powdered milk, as well as goods of importance only in the tourist trade such as candy, cookies, beer, soft drinks, and bottled water. These products are bought by mountaineering

groups, trekking parties, and lodges on a large scale at the weekly market in Nauje. Many Sherpas believe that these purchases not only greatly inflate prices but also create shortages, and this is one of the most widely cited local complaints about the adverse impacts of tourism.

Increases in the rates of pay for trekking and mountaineering work, however, have more than compensated for these commodity price increases. In 1971 rice cost less than twenty-five rupees per pathi (Fürer-Haimendorf 1975:77) and potatoes were fourteen rupees per tin, which means that rice quadrupled and potato prices increased by two-and-a-half times over the past twenty years. In 1971-1972 sirdar were paid twenty-five rupees per day while H. A.s and cooks received fifteen rupees and porters ten. In 1991 a sirdar's wages were seventy-five to one hundred rupees per day while cooks made seventy to ninety rupees, H. A.s seventy to one hundred rupees, and porters one hundred and twenty rupees. Thus while rice prices quadrupled over twenty years the pay of porters had gone up twelve times and that of cooks and H. A.s by five to six times. Sirdar, it is true, were being paid by the day at a rate only three or four times greater than twenty years earlier, but the relatively low pay given to them reflects the expectation that they would make much more money through the use of their position. Many individuals, moreover, have increased their incomes during this period through career advancement, moving from cook-crew work to positions as camp-crew members or cooks, or from camp-crew work to positions as sirdar and group leader. In many families total income has also risen as sons have taken up trekking and mountaineering work. The number of such multi-income families greatly increased during the 1980s.

Inflation thus has not been a serious problem in Khumbu. During the 1980s most Khumbu families who have been involved in tourism have probably become relatively beter off despite inflation. So too have families who perform skilled labor such as carpentry, for carpenters today make triple the wages that they did even ten years ago, and command day wages of eighty to one hundred rupees. Even Khumbu families who depend for their cash income entirely on wages from agricultural day labor have seen their earnings rise even faster than rice prices. Twenty years ago in Khumjung women were paid a maximum of three rupees per day when a local porter received ten rupees per day from mountaineering expeditions (Fürer-Haimendorf 1975:42, 89). Whereas rice prices have quadrupled and potato prices have increased by two-and-a-half times since 1971 agricultural day labor rates have gone up ten times to thirty rupees. The only households that have really suffered from inflation are those that have next to no cash income, such as elderly couples who are maintaining their own households. Sherpas who have been left out of the opportunity to make money from tourism work have become

increasingly disadvantaged in comparison with the rising affluence of their neighbors.

Tourism and Agricultural Change

Tourism has been reported to have had a number of adverse impacts on local subsistence agriculture in Khumbu despite the fact that little Khumbu-grown food other than potatoes is served to tourists.^[17] Tourism can have major indirect impacts on land use, labor, and lifestyles, and these are the factors which are thought to have transformed Khumbu agriculture. Among the impacts that have been reported are the development of a generational gap in farming (BjØnness 1983:268; J. Fisher 1990), a decline in the amount of land in crop production (Fürer-Haimendorf 1984:15; BjØnness 1983:269), a decline in crop yields (BjØnness 1983:269; J. Fisher 1990:122; Fürer-Haimendorf 1984:8), inflation of agricultural day-labor rates and shortages of agricultural labor (Fürer-Haimendorf 1984:8, 15, 17), and increased use of non-Khumbu migrant labor (J. Fisher 1990:122). According to one report, "tourism has replaced agriculture as the mainstay of the economy" (Karan and Mather 1985:93).

Whereas some of these reported impacts are indeed significant, others are rather less important than has been imagined. Inflation in agricultural day-labor rates has taken place, and there is indeed a greater influx of agricultural workers from outside the region than there was in the past,^[18] but reports of labor shortages, declining yields, abandoned land, a generation gap in farming, and the eclipse of agriculture in the Khumbu economy may have been exaggerated. The production of food for household consumption continues to be the basis of Khumbu agriculture, and the role of crop growing in local economy, culture, and lifestyles has not yet eroded.

Tourism has not commercialized Khumbu crop growing to any significant degree. The need for larger amounts of potatoes has led some lodgekeeping families to acquire additional cropland, and the high regional demand for potatoes and the money to be made from the sale of surplus tubers may have been a factor both in the recent widespread adoption of high-yielding varieties and in the conversion of land from a two-year rotation of potatoes and buckwheat to continuous potato cropping. Yet an interest in production for sale to lodges probably has been a far less important factor in the recent regional intensification of potato production than continuing dynamics that go back a century in Khumbu. Recent intensification, like older intensification, probably primarily reflects a response to population growth and fragmentation of land holdings. Everyone has adopted the new potato varieties, not just families

who are interested in selling surplus potatoes. Likewise, buckwheat cultivation has been abandoned by the entire population of the Bhote Kosi valley and not just by large landowners who have a surplus to market. There has also been little evidence of local interest in the commercial production of other crops for tourist consumption. Fields are not being taken out of potatoes and put into cabbage, spinach, cauliflower, or carrots, even though these bring exceptional prices at the weekly market. Some Nauje lodgekeeping families have shown great interest in producing more vegetables and spices for use in their lodges and have developed the finest household gardens in Khumbu.^[19] But no one has yet ventured to plant even a single field in vegetables for sale in the market. In Pharak, however, the commercialization of farming in response to the new tourism market and increasing Khumbu Sherpa affluence has had an important effect on local land use and many Pharak Sherpa families sell produce in the Nauje market.

Most families have not found that the time demands of trekking and mountaineering work have compromised their ability to conduct field work. It is true that the height of the spring trekking season overlaps with the weeks when fields must be prepared and planted, and that the autumn mountaineering season and the beginning of the autumn trekking season overlap with potato harvesting and with buckwheat and barley harvesting and threshing. It has often been assumed by outsiders that the absence of men during parts of the agricultural season must have a severe effect on crop growing, yet this is not the problem it would be in a culture in which men were responsible for most farming work. In Khumbu women farm, and until now very few women have worked in mountaineering and trekking. Women still place a priority on farming and tailor their other activities around its rhythms. Those who operate lodges, for example, find relatives, neighbors, friends, servants, or children to run things when they need to be in the fields or else arrange for servants, hired help, or daughters to work the potatoes and buckwheat. If all else fails they simply close their lodges for a few days. The preoccupation of men with their trekking and mountaineering assignments may have some small impact on those limited aspects of crop production in which they lend a hand: plowing buckwheat and barley fields, carrying manure to the fields and potatoes in from the fields, harvesting grain, and cutting hay and wild grass. Their presence is missed in some of these tasks, and as a result women do have to do more of the work than would otherwise be the case. Probably the greatest additional pressure is put on women at grain harvesting and threshing time, for in the pretrekking era in villages like Pangboche men and women handled this work equally or men performed even a greater share of the work than did women. Now women in most families are on their own, and for these tasks there is a element of

urgency in the harvest because it has to be completed before bad weather sets in and damages the crop. Yet while women's agricultural responsibilities are thus greater now than ever, the fact that they have long carried out most of the work of producing crops means that the loss of help from men is not as critical as it would otherwise be. The men's absence is also more than compensated for in field work by the increasing use of wage laborers from outside of Khumbu. Rather than a regional labor shortage tourism has created a situation of regional labor abundance through giving many more families the ability to hire agricultural day laborers. Such laborers, as we have seen, are paid less than half the daily wage of the lowest-paid trekking workers. Families who a decade ago either relied entirely on their own efforts or on their participation in reciprocal work groups today hire help for field preparation, planting, weeding, and harvesting. This has enabled Sherpas to cope with the increased labor involved in harvesting the large crops of yellow potatoes and development potatoes, made it possible for some families to expand the area they farm, and enabled more women to give their places in reciprocal work groups to hired helpers and to reduce the amount of work they do in their own fields.

Some Sherpas complain that the increasing reliance on help from non-Khumbu agricultural workers lowers harvests. They remark that outsiders do not have the knowledge and experience that local people have in farming and that hired people do not take the care in their work that they themselves do.^[20] It is not clear how much credence should be attached to such remarks. In Shorung one hears similar comments about Khumbu Sherpa women being less skillful in farming than Shorung women. It may be that non-Khumbu people are less experienced in the planting and harvesting of potatoes, but digging fields and digging up potatoes are not tasks involving especially great skills and all the decisions about crop selection and the timing of agricultural tasks continue to be made by the families themselves. It is hard to imagine that field laborers' work could be too poor given that they are usually closely supervised by household members or work alongside Sherpa women as part of reciprocal labor groups.

There is no evidence that there has been a general decline in Khumbu agricultural productivity in the sense of lower regional crop yields. It is true that specific varieties of potatoes, especially kyuma and the red potato, are perceived to produce less than they formerly did—a change which could be related to a subtle decline in field fertility or to changes in characteristics such as disease resistance of the plants themselves.^[21] Yet productivity per field and across the region has nevertheless increased since the 1970s due to the adoption of high-yield varieties. The universal perception among Khumbu villagers is that regional potato

production is today at an all-time high due to the replacement of the red potato by the yellow potato and the development potato.

It is also not true that less land is in production than a few decades ago. As was discussed in chapter 6, the large numbers of abandoned fields that can be seen in some parts of Khumbu probably reflect emigration and intensification in the late nineteenth and early twentieth centuries, not the recent impacts of tourism. In the few cases where terraces have been taken out of production during the past twenty years these have mostly been associated with places where wildlife damage to crops has been particularly high, such as Tashilung and Nyeshe in the lower Bhote Kosi valley.^[22] Fürer-Haimendorf appears to have been mistaken about recent abandonment of cropland at the gunsa of Teshinga (1984:15). The decline in nak pastoralism, however, may have led some families to deemphasize their production of hay in the high-altitude herding areas and even to abandon small potato fields in places where they no longer herd. There are a few examples of this in the upper Dudh Kosi valley. Neither this nor the general decline in nak herding itself, however, can be attributed to tourism.

The recent trend in Khumbu has been toward an expansion rather than retraction of the area under cultivation. During the late 1980s some long-abandoned terraces were clandestinely put back into production despite attempts by Sagarmatha National Park officials to discourage this, and since national park opposition declined in 1989 some Sherpas have begun registering and reclaiming substantial areas that have been out of crop production for half a century and more. Many old terraces in the Samde area, for example, are being put into hay and potato production. Other families are using income from tourism to buy cropland. There are a number of cases, for example, of land-poor Nauje families (many of them second or third generation immigrants from Tibet) who have made money in mountaineering and trekking and have used it to purchase crop and hay land in Nauje, Mishilung, Nyeshe, Langmoche, Chosero, and Tarnga. Some families who have opened lodges have also bought more crop land in these areas.

Many of the adverse impacts ascribed to tourism in Khumbu thus do not yet seem to have occurred. Land is not being abandoned for lack of workers with the time to cultivate it or declining dramatically in productivity. Crop production remains oriented to subsistence production rather than to market sale and tourism. Yet the issue of whether or not a generational gap in farming (BjØnness 1983; J. Fisher 1991, personal communication) is developing is nevertheless worthy of further analysis, for this could occur even without a decline in regional agricultural productivity.

I take generation gap in this context to be a change in attitudes towards lifestyle, customs, beliefs, and values between parents and their

children which is so widespread a phenomenon as to seem commonplace or even characteristic of a region and an era. In some of the world's mountain regions, and particularly in the Alps, such a generational gap may indeed have been an important factor in a post-World War II decline in farming. The movement here of large numbers of young people to wage labor in the towns and cities has left the family farm in the care of aging parents and relatives. In the Alps this does seem to reflect value changes as well as responses to new economic opportunities in the lowlands. The Khumbu situation, however, is different. Here there has also been some out-migration to Kathmandu (and to other countries) and again this has primarily been by the young. Yet several points must be made about this. The scale of this migration, to begin with, is very different. Only a small percentage of Khumbu youth are emigrating from Khumbu. Many of those who do move to the city, moreover, are men who return to Khumbu a few years later to marry and establish a household there. Some young families have shifted their residence to Kathmandu for a few years, often so that their children can attend the good schools there as well as to enjoy a period of urban living and take advantage of job opportunities. Many of these families also return to Khumbu, although there is an increasing community of Khumbu families who appear to be permanently settling in Kathmandu. There are cases where Khumbu land is tended by aged parents whose children have all emigrated, and houses are occupied by renters or relatives of people who have gone to Kathmandu or abroad. But this remains the exception rather than the rule. Much research remains to be done to establish the scale of migration that has taken place and whether it is now increasing as it seems to be. I do not believe, however, that even the present level suggests that urban migration is characteristic of a generation of Khumbu Sherpas.

There also does not seem to be a decline in interest in farming as a way of life among Sherpas who remain in Khumbu. There is no generational discontinuity in lifestyles and household economy among Khumbu residents. Young Khumbu families living in the region have not declined their inheritance of main cropfields or sold, rented out, or abandoned the land. I do not believe that there is a single example today in Khumbu of any of these phenomena. Every Sherpa family in Khumbu today farms. No matter how well-to-do many Khumbu Sherpa families have become from tourism they continue to produce as much of their household food supplies as they can. This is as true of households that have just been established as of those that have farmed for decades.

Finally, in Khumbu a generational gap in farming means above all a change in attitudes and lifestyles between mothers and their daughters, for it is women who grow crops. I do not know of any cases in Khumbu

today of daughters who have refused to work in the family fields or who on marrying and establishing their own households have chosen not to farm. There are probably no more than a few cases even of young women who rely entirely on hired agricultural labor or household servants to farm their land. Some may take advantage of their relative affluence to hire more assistance than their parents could afford to. And some certainly aspire to other careers, such as shopkeeping and running their own lodge. But for Khumbu women today as in the past nonfarm occupations supplement rather than replace crop production.

It is certainly conceivable that in future years Khumbu agriculture may change considerably and that tourism may be an important factor in this. It is very possible that agriculture may become more commercialized than it now is, with more land put to vegetables for sale to the lodges. It is possible that some households may begin to rely more on their income from tourism and less on crop cultivation and that as a result more land may be worked by hired labor, rented out, or even sold or abandoned. A generational gap may yet develop. That none of this has occurred thus far seems to be grounded in cultural attitudes more than regional conditions of land tenure, access to markets, or the particular forms tourism has taken. The high degree of land fragmentation, the inability of Khumbu families to be self-sufficient in food production from their own lands, and the resulting increasing reliance on purchasing food would seem to make the cultivation of high-value crops for market sale an appealing opportunity. Families throughout the region have ready access to markets, both to the weekly market in Nauje and directly to the lodges that are so numerous in much of Khumbu. There is considerable demand for fresh fruits and vegetables from both increasingly affluent Sherpas and from mountaineering and trekking groups and lodges. Yet Sherpas appear to be reluctant to give up a way of life that has proven reliable for so long. For many generations Sherpas have thrived on a balance of subsistence agropastoralism and nonfarm income and resources, and tourism may not yet seem a stable enough base to give up that diversity. Beyond that it may also be that many women continue to value crop growing as a way of life bound up with their sense of themselves as women and as Khumbu Sherpas. It is here that a generational gap in attitudes about farming could greatly affect Khumbu lifestyles and the economy. If greater opportunities for women in nonagricultural work should ever develop there may be more interest in scaling down farm operations.

Tourism and Pastoral Change

A number of tourism-related pastoral changes have taken place in recent years in Khumbu even though tourism has not yet resulted in the

commercialization of either locally produced dairy products or meat.^[23] Tourism has played a greater role in recent changes in pastoralism than it has in agricultural change and has been a factor not only in changing household economic practices but also in community resource management. These changes have primarily been related to the increased keeping of pack stock and particularly the keeping of urang zopkio. The zest with which Sherpas have purchased these crossbreeds from Pharak and Shorung breeders in recent years has, along with the less spectacular but also significant increase in numbers of cows and zhum, almost balanced the major decline since the 1960s in the number of nak being kept in Khumbu.

According to Sherpas zopkio were kept only in small numbers by a very small number of families before the mid-1970s. Despite their common use as plow animals in Tibet they were not widely used in this fashion in Khumbu before the 1950s (and even today only a very small number of teams are trained for this purpose). Yak were more valued as pack animals for the Tibet trade. Low altitude and poor bridges led Sherpas to pack loads south from Khumbu on foot rather than risk yak or dimzo, although a few families used sheep and urang zopkio to ferry loads.^[24] In Nauje, where today zopkio are the predominate form of stock and more than 50 percent of all households own at least one, there are said to have been only five families who had zopkio in the early 1960s.

Since the 1960s, and particularly since 1975, urang zopkio have become increasingly important in Nauje, Khumjung, Kunde, and Thamicho.^[25] By 1978 the total number of adult zopkio (urang and dimzo) had reached 80 (BjØnness 1980a: 66). During the next few years this increased sharply, reaching 482 in 1984 (Brower 1987:189) when urang zopkio constituted 18 percent of all large stock in the region. Figures differentiating Nauje-owned urang zopkio and dimzo are not available for 1978 and 1984, but in 1987 there were 148 urang zopkio and 8 dimzo. These were kept by fifty-four families who owned an average of just under 3 head of stock per household. Fifty-two households, 58 percent of the Sherpa households in the village, owned urang zopkio and these constituted 65 percent of the village's 226 total head of large livestock (excluding calves). This is a substantial increase from just three years earlier when they were 48 percent of village stock.^[26] By the spring of 1991 still more urang zopkio were being kept in Nauje. The village total had climbed another 15 percent to 171 urang zopkio, kept by fifty-one families. Nearly as many were kept by Khumjung (158) and in the Thamicho settlements (149), and the Thamicho count may well be incomplete. Across Khumbu there were at least 580 urang zopkio in the spring of 1991, more than any other form of cattle other than nak.

Urang zopkio are neither cheap to purchase (at approximately 3,000 rupees, an adult zopkio costs more than a month's wages for most trekking Sherpas) nor cheap to keep, since they require substantial amounts of fodder and are often owned by families who have no hayfields. But they can very quickly return the investment made in them. At 1990 pack rates of 80 rupees per load, 160 per zopkio, a pack animal can make back its purchase price in less than three weeks. Owning zopkio is particularly convenient for sirdar who are in charge of procuring porters and pack animals for their groups or expeditions and can ensure that their own livestock are kept fairly continuously in work. As mentioned earlier, 70 percent of those Nauje households where a family member works as a sirdar owned urang zopkio in the late 1980s. In the spring of 1991 twenty-five Nauje households having one or more members employed as sirdar had urang zopkio, representing 64 percent of all village households that had sirdar and 96 percent of sirdar households that owned pack stock. This link between sirdar and zopkio was not limited to Nauje. In Khumjung, for example, twenty-four of the village's thirty-eight sirdar households (63 percent) had urang zopkio.

The increase in the number of urang zopkio has brought more income for some of the already most well-to-do families and has increased the relative wealth of Nauje, Khumjung, Kunde, and Thamicho. This change in regional pastoralism has also put new livestock pressure on buckwheat crops in the Bhote Kosi valley and precipitated increasing conflict over summer range regulations in the villages and lower valleys. As I will discuss later in this chapter, zopkio have also placed new stresses on pasture and forestlands in certain areas with their very different grazing and browsing habits and different seasonal patterns of transhumance.

The other recent change in regional livestock keeping has been an increased emphasis on zhum and cows. Since the 1950s, and particularly during the past twenty years, the numbers of zhum and cows have increased considerably. Both are considered to be more effective milk producers than nak and are practical for households that want dairy products for their own use and are only interested in keeping a few head of stock without the commitment of time, energy, and lifestyle required to herd nak. Usually families keep only one or two zhum or cows each, enough to supply some milk for family use. Milk from such stock is only rarely sold and then only to fellow villagers in quantities of a few pints per week. Such arrangements are made directly among neighbors and friends when they are made at all; and zhum's and cow's milk and dairy products are not offered for sale at the weekly market in Nauje or sold to lodges.^[27] Commercial dairy production for the tourist market thus is not a factor in this regional herding change, as BjØnness thought

(1980a: 67). The increasingly widespread practice of keeping a cow or two, however, may well be indirectly related to tourism. It is the new income and affluence from tourism that now enables families who previously did not own cattle to purchase and keep cows and zhum.

The increasing involvement of so many Khumbu families in tourism, and particularly the employment of so many Khumbu men in trekking work, has also had an impact on herding. It is impossible today to find men, even those from the poorest households in the region, who are willing to work as hired herders. When men are often away from home for four or five months per year on treks and expeditions it becomes difficult for many households to maintain their former herding patterns using their own labor resources. Some families discover that it is difficult to maintain the size of herd they formerly cared for, and this is especially the case with nak herds. Families may lack a family member who has the free time and aptitude to take nak up into the high valleys in early spring or to cut and store the hay harvest and wild grass in the autumn when the men are away on treks. It is difficult for the women of the household to undertake these tasks in addition to their crop-production responsibilities, although there have been some cases when teenage daughters have taken on the job of nak herding. There is less of a problem with the care of cows, zhum, and zopkio, which can be allowed to graze freely in the village vicinity and are easily tended in the early morning and evening by women and by children who are still able to devote their main attention to crop-production tasks and school.

Families who have long herded nak have several options. They can forgo opportunities in tourism, arrange short-term, reciprocal labor exchanges with relatives and friends who may be able to watch stock while a herder is away on a trek, hire more day laborers to help with hay cutting and fodder gathering, ensure that a son or daughter can be entrusted with herding responsibilities, or decrease the scale of herding. It is very common to make use of cash income from tourism in order to hire workers to compensate for the lack of sufficient household labor at hay-cutting and fodder-collecting time in the autumn. Usually hired helpers are not entrusted with the care of nak, however, for many Khumbu herders are loathe to put their prized stock in the charge of people in whom they do not have full confidence, and it is widely believed that the lower-altitude migrant workers who make up so much of the wage-labor pool do not have the requisite experience and knowledge. There are households where either a father or a son deliberately forgoes tourism work when there is a scheduling conflict with herding. There are also cases where families have scaled down the size of their nak herd or even given it up entirely. This is particularly common when none of the sons shows any interest in continuing with nak herding and

either their father's career in tourism or his increasing age lead him to decide that he cannot maintain the family herd in the same way any longer. Sometimes such families retain a few head of cows, zhum, or zopkio, but sell off their nak, high-altitude herding huts, and many of their hayfields. This is the one area of Khumbu agropastoralism where a generational decline of interest in "traditional" lifestyles has contributed to a shift in land use. It must be noted, however, that these situations were not responsible for the large-scale decline in nak keeping since the 1950s as has sometimes been suggested. The great decline in nak keeping occurred in the 1960s before tourism had a major impact on Khumbu pastoralism and, as discussed in chapters 6 and 9, was instead related to changes in international trade conditions.

Tourism and Changing Forest Use

Tourism has placed several types of new demands on Khumbu forests. For decades, until new Sagarmatha National Park regulations were implemented in 1979, mountaineering and trekking groups routinely cooked their meals and warmed themselves at night with campfires. Early teashops and lodges were built with timber from local forests until the national park put an end to most tree felling in Khumbu. And nearly all of today's lodges continue to rely solely on fuel wood.

Until the 1970s the major new demand for Khumbu forest products came from the mountaineering expeditions. Expeditions made use of kerosene and gas stoves at their camps on the high peaks, but made wood cookfires at base camp and during their approach marches through Khumbu. So did the hundreds of porters who carried their supplies. As expeditions made their way through Khumbu on the way to Mount Everest and nearby peaks they usually camped for a few days near Nauje, Tengboche, and Pheriche, scouring nearby forest for fuel wood. Once at base camp they established fuel-wood stockpiles large enough to supply the base-camp crew for six weeks or more.^[28] It is difficult to estimate what the total fuel-wood requirement of an average expedition might have been. Nima Wangchu Sherpa (1979) has put this at 960 porter loads of fuel wood, or a minimum of about 28,000 kilograms of fuel wood over a two to three month period in Khumbu.

Much of a Mount Everest expedition's fuel-wood use came at base camp and thus was obtained from the region's highest-altitude forest and shrublands. The Swiss began the custom in 1951 of hiring local Sherpas (usually from Pangboche) for day wages to supply their base camp with fuel wood. Much of this was shrub juniper from the Tugla and Tsolo regions near the terminus of the Khumbu glacier. Some additional fuel wood was also carried up to base camps from the highest forest, the

birch and rhododendron woodland northeast of Pangboche along the Imja Khola. During the early 1960s expeditions began to buy fuel wood by the load and a few years later it became customary to pay by the kilogram, a practice that encouraged the sale of freshly cut green wood rather than lighter dead wood. Since all of the areas that supplied fuel wood to the base camps for Mount Everest, Lhotse, Nuptse, Pumori, and Cho Oyu lay outside of the borders of locally protected forests, this new commercial activity did not compromise local values or practices. The sale of fuel wood became an important part of the Pangboche economy for a few years. Pangboche residents also benefited from the early expeditions' use of large numbers of tree trunks to bridge ice-fall crevasses. According to one report a 1974 Mount Everest expedition used 200-250 mature trees for ladders and crevasse bridges (Lucas, Hardie, and Hodder 1974).^[29] These would also have come from the upper Imja Khola valley above Pangboche and good wages would have been made to cut and transport them to base camp.

Early trekking groups also required fuel wood, and with the rise in popularity of trekking during the 1970s this probably came to be a more significant demand on local forests than mountaineering. The large commercial groups almost invariably depended entirely on local wood supplies for their cooking fuel, and it was a popular custom to have an evening bonfire as well. BjØnness found in 1978, two years after the area had been declared a national park and just before tourists' campfires were made illegal, that only 7 percent of trekking groups carried fuel for cookstoves (1980b: 126). Nima Wangchu Sherpa (1979) estimated that trekking groups used about three loads (approximately ninety or more kilograms) of fuel wood per day.

The Sherpa lodges that began to be opened in the 1970s also relied on fuel wood for cooking and in cold weather often kept fires going beyond mealtimes to warm guests. The amount of wood required by lodges varied both seasonally and with the number of guests they accommodated. In 1983 Nauje lodge owners estimated that their fuel wood use during the tourist season ranged from two to four loads per day and fell to half a load per day during the tourist off-season when it was only necessary to cook for their own household. Since then lodge fuel-wood use has increased in many cases due to the spread of the practice of selling hot showers to tourists.^[30]

As new lodges were built beginning in 1974 there was also demand for construction timber. The first Sherpa lodge which was built for that purpose (rather than remodeled from a house), for instance, was constructed at Nauje from timber cut just below the village in an area that had been within the rani ban. Other lodges built or remodeled in the early 1970s also utilized Khumbu timber. This was a factor only in the

Nauje, Tengboche, and Pangboche (for timber for Pheriche lodges) areas, however. Most of the lodges opened elsewhere during the 1970s were simply remodeled from houses and herding huts and thus did not require the cutting of large amounts of new timber.

These early patterns of tourism-related forest use were greatly affected by the establishment of Sagarmatha National Park. Both the Nepali ecologist who scouted the region in 1973 (Mishra 1973) and the 1974 New Zealand mission that recommended New Zealand aid to establish Sagarmatha National Park (Lucas, Hardie, and Hodder 1974) reported that deforestation in the region was a critical problem and were concerned about new demands on forests because of tourism as well as uncontrolled Sherpa traditional forest use. The New Zealand team concluded that the impact of tourism on forests was so severe that, "There is no doubt that, in the present situation, the economic benefits tourism brings to the Khumbu [region] are largely off-set by attendant environmental pressures" (ibid.:11). A few years later the first draft national park management plan placed the blame for regional forest degradation primarily on tourism, noting that:

The major problems stem from inroads made into the forests of Khumbu for structural timber for Sherpa "hotels" and other buildings to serve the growth of tourism and the increasing use of fuelwood by these tourist establishments and by the many trekking and other tourist parties which camp and cook their way up and down the valleys. (Croft 1976:22)

The measures that were implemented in 1979 in response to this perceived crisis were at first relatively effective. The regional ban on felling trees for timber (other than beams) averted further damage from the use of local forests for building material for lodges.^[31] The regulations prohibiting the sale of fuel wood to foreigners and the use of fuel wood by mountaineering expeditions and trekking groups were effectively enforced and significantly lowered tourist fuel-wood use—although they did not entirely eliminate it since porters and staff were still permitted to obtain and use fuel wood for themselves.^[32] A fuel depot was set up near the entrance to the park at Jotsale with assistance from the German Alpine Club to make kerosene and other fuels available to trekkers and mountaineers.

These early efforts to eliminate the impact of tourism on forests had one major loophole: no regulations were issued that prohibited or limited the use of fuel wood by teashops and lodges. During the 1970s when there were relatively few lodges and most tourists camped in tents this was not a critical failing. But during the 1980s the growing popularity of independent trekking and the rapid growth in the importance of local

lodges made the early national park regulations increasingly inadequate. It is possible that that more fuel wood is now being burned in Khumbu due to tourist demand than ever before. Although well aware of the increasing lodge use of fuel wood for cooking and hot showers, Sa-garmatha National Park administrators have felt that prohibiting fuel-wood use by lodges would be too politically explosive a measure given the already-tenuous relationship between the national park and local residents. Such a measure, however, has been successfully implemented in the much-touristed upper Modi Khola valley of the Annapurna Conservation Area in central Nepal. Here a ban on fuel-wood use for cooking was enacted in 1987 by a local committee of lodge owners on the advice of conservation area administrators, and a kerosene depot was established (Stevens 1988a ). Upper Modi Khola valley lodges have been cooking entirely on kerosene stoves for the past five years. The cost of the kerosene is passed along to tourists through slightly higher meal prices.

Across Khumbu, fuel-wood use by lodges and trekking-group and mountaineering-expedition porters may account for more than 20 percent of regional fuel-wood use (Hardie et al. 1987:38). But this demand for fuel wood is not spread evenly across the region, and tourism has increased fuel-wood demand on particular forests, woodlands, and alpine areas far above the level already exerted by Sherpas for subsistence purposes. The many lodges in Nauje and adjacent Chorkem, for example, probably require as much or more fuel wood during trekking season than the entire Sherpa population of the village.^[33] Fuel wood for the lodges is obtained from the same areas where villagers collect fuel wood for household use, the slopes west of the Bhote Kosi river in the Satarma area, the area above the Dudh Kosi river just below Khumbu near Monjo, and the lower slopes of Tamserku. The pressure for fuel wood being put on the high-altitude juniper in the Tugla-Tsolo area in the Lobuche Khola valley region must exceed by many times the seasonal fuel requirements of local herders. The impact of this focused demand on local vegetation will be explored later in this chapter.

Tourism, Resource Use, and Land Use Change

Tourism has thus created several new demands on Khumbu natural resources, some of them through the direct requirements of tourists and their porters and pack stock and some of them through the indirect impacts of changing Sherpa subsistence practices as a result of new patterns of lifestyle and affluence. The direct demand tourism has had on resources in Khumbu thus far has largely added to preexisting patterns of

Sherpa resource use rather than introducing entirely new requirements. The relatively small scale of regional tourism to date and its basis in Sherpa-operated small lodges and camping trekking have kept new resource-use demands to relatively moderate levels. Tourism has not yet brought the major demands on local resources and land familiar in so many other major, international mountain resorts from roads, parking lots, and major hotels (even the Japanese Everest View Hotel has only twelve rooms) to ski facilities, swimming pools, tennis courts, and golf courses. Khumbu is not yet the Swiss, French, Italian, or Japanese Alps, nor is it the Yosemite valley or even an Indian Himalayan resort. Yet even the small scale of tourism development thus far in Khumbu has meant some increase in the use of fuel wood, timber, and building stone as well as more demand on local pastures, greater fodder production, and increased demand for water for drinking, cooking, and bathing.

The new regional and household income that tourism has created in Khumbu through opportunities for employment and business operations has also affected local resource demand and land-use practices by effecting changes in Sherpa standards of living, lifestyle goals, household labor allocation, and community resource management. These changes reflect not only the types and scales of Khumbu tourism but also Sherpas' responses to new opportunities and their decisions about how to use new wealth.

Affluence has led to new types and levels of resource demand, including desires to participate more fully in the "good life" as defined both in traditional terms and in new ones, which include the adoption of some Western tastes in clothing, diet, and electronics as well as new local fashions in architecture such as multiroom houses, wood paneling, ceilings instead of open-beam construction, glass windows rather than translucent paper ones, and corrugated iron roofing. Increasing aspirations for a higher quality of living in traditional terms and the wealth to meet them has meant some changes in land use. Land-poor families have displayed interest in acquiring new land, but have mostly been frustrated in this by the lack of sellers. More important has been the purchase of livestock. Here I would distinguish between the purchase of pack stock which is an investment in tourism and purchase of dairy stock which is intended solely to better family diets. Meeting new tastes and needs has almost entirely involved increasing the importation of goods from outside the region rather than changing resource demands and land use inside Khumbu itself.^[34] The trend towards larger houses has had some implications for local forest use, but Sagarmatha National Park restrictions have diverted most of this impact to Pharak. The impacts on regional vegetation and pasture productivity of the new regional herd composition, however, may be more serious. In terms of their way of life

affluence has not yet led people to decide that the effort to cultivate crops is no longer necessary, much less to conclude that it is incompatible with changing lifestyles and ideas of status or identity. With pastoral-ism the issue is again more complex and there has been some decline in interest in nak herding as a way of life.

The question remains of how these changing land uses and attitudes toward resources have affected the operation of local resource-management institutions. There is no doubt that tourism can have major impact on the organization, goals, and effectiveness of local resource-management institutions through introducing social, cultural, and environmental changes. Tourism can undermine the ability of communities to continue to enforce former resource-management regulations by changing the local household economy, patterns of labor allocation, and subsistence land use, commercializing natural resources, introducing new values, and creating new community patterns of wealth, class, and status. I have already noted that new stock-keeping practices, and specifically the keeping of larger numbers of urang zopkio, zhum, and cows, may have played some part in the undermining of the nawa system of livestock management in Nauje and Thamicho. Some observers have also argued that the commercialization of fuel wood played a similar role in undermining the shinggi nawa management of forests: "The traditional forest wardens (shing nawas) had ceased functioning by the early 1970s (although they were still active in Phortse) as the astronomical sums tourists paid for firewood had led to massive cutting that systematically undermined the forest wardens' authority" (J. Fisher 1990:142; see also Nima Wangchu Sherpa 1979).

Yet too much should not be made of the link between tourism development and the decay of local Sherpa resource-management institutions. It is clear that the nawa system of pastoral management has not collapsed throughout the region and that it is still maintained today even in areas such as Kunde and Khumjung which are heavily involved in tourism and where a shift in herd composition has taken place along the same lines of that in Nauje and Thamicho. It also seems incorrect to draw a connection between tourism development and the breakdown of traditional Khumbu forest-management institutions. I argued previously that the most striking example of the abandonment of local forest management, the decline in management of Nauje and Khumjung-Kunde rani ban, resulted from local responses to forest nationalization. The later abandonment of rani ban management at Pare was similarly related to changes in the regional control of forests, in this case by Sagarmatha National Park, and not to tourism. And even at Pangboche, where local use of forests and other fuel-wood resources was indeed greatly affected by tourists, it is not correct to ascribe a decline in local community

management to tourism. Prior to the advent of tourism Pangboche villagers managed forest resources in two ways, by banning the importation of freshly cut wood to the village while crops were maturing and by banning all tree felling in the nearby lama's forest of Yarin. The village has never ceased to enforce its regulations against bringing fuel wood into the settlement during the late summer and village-chosen nawa have continued to fine violators. Community protection of the sacred forest at Yarin did indeed diminish for a few years in the 1980s, although it was ultimately restored, but this had nothing to do with the sale of fuel wood to mountaineering expeditions and trekking groups. The main sources of that fuel wood, as I have already remarked, were far beyond the borders of the Yarin forest. The traditional restraints on tree felling were instead relaxed because of national park intervention in granting permits to outsiders (Kunde and Khumjung villagers) to fell trees for beams there and the subsequent Pangboche reaction to this.^[35] I do not believe that any Khumbu local forest-management systems ceased to function because the profitability of selling fuel wood to trekking groups or mountaineering expeditions led Sherpas to violate community resource-use regulations and defy shinggi nawa.

Tourism and Environmental Change

Nowhere in the Great Himalaya is concern for the environment more intense than in the Khumbu area. Processes of change have brought a plethora of environmental disruption to this formerly remote, unspoiled region. A major factor is tourism and the hordes of overseas tourists and trekkers. (Karan and Mather 1985:93)

Tourism has both accentuated old pressures on Khumbu's natural resources and environment and introduced new ones. The impact of tourism on Khumbu forests and the increasing accumulation of trash along the trails and at camping and lodge sites have been among the most widely reported Himalayan environmental problems. Less heralded, but potentially perhaps even more serious, is the possibility that tourism development may be contributing to localized overgrazing.

Litter, Pollution, and the Garbage Trail to Mount Everest

The increasing amount of litter along Khumbu trails and the large dumps that have developed at some expedition base camps have

shocked and offended many Khumbu visitors and generated some concern in Nepal that this may ultimately diminish the attractiveness of the region to tourists. During the 1980s the accumulations of garbage became a much more visible eyesore at such renowned tourist stops as Nauje, the Tengboche monastery, and Lobuche, and the mountains of discarded trash left behind at base camps by climbing expeditions became notorious. Cleanups organized by concerned foreigners, trekking companies, Nepalese organizations, the Sagarmatha Club (a Khumbu youth organization based in Nauje), and the recently founded Sherpa Pollution Control Committee have periodically rid the main routes through the region of litter and have consolidated, burned, buried, and carried out expedition rubbish from high-altitude base camps. One 1984 expedition collected sixteen tons of trash from the Khumbu glacier Mount Everest base camps alone. But these efforts have only temporarily alleviated conditions, despite Sagarmatha National Park regulations requiring mountaineering expeditions and trekking groups to remove their garbage. Expeditions and groups continue to flaunt this regulation and the national park has not yet developed an effective system of administering fines to violators. The situation has been compounded by the fact that Sherpas themselves contribute to the solid-waste problem by scattering discarded clothes, cans, bottles, and other refuse into village environs, including streams. There are no village dumps anywhere in the region. Much of the trash in the village vicinities is local rubbish, not tourist waste. But there is also a tourist component. Even those lodgekeepers who carefully place rubbish containers in their dining and sleeping rooms usually seem to dispose of their waste in the same haphazard way that villagers handle household garbage. There was also previously little concern with establishing facilities to deal with human waste in upper Khumbu. At such important tourist overnight sites as Gokyo, Gorak Shep, and Lobuche there were no pit toilets or other facilities until the late 1980s, and the construction of a number of new pit toilets at Lobuche may not have entirely solved the problem there, for one of these has been built immediately above the local spring.

Trekkers, trekking agencies, and mountaineering teams have begun to show some signs of taking increased responsibility for policing their own littering and trash disposal, but increased national park and local Sherpa enforcement activities are clearly required. The problem of expedition garbage could be dramatically reduced by requiring all expeditions to post a substantial bond that would not be returned until their camp had been certified clean. If it was not, the forfeited money could be used to fund Sagarmatha National Park cleanup campaigns. Nepal's association of trekking agencies (TAAN) could be required to organize a once- or twice-yearly cleanup of the trails. More trash receptacles could

be set out for use by individual trekkers and a campaign could be organized to make them more aware of the importance of using them. Lodges could be required to collect their own trash and keep a waste pit for disposing of it. But it will require other types of programs to address the wider problems of tourist waste. Until a village-based program is put in place to deal with solid waste and to recycle glass, plastic, and metal no number of effective measures to force mountaineering and trekking groups to clean up their camps, or lodges to collect their trash, will entirely solve the current problem. New values must be promoted among Sherpas as well as tourists, for only this will make it possible to effectively establish and maintain a Khumbu-wide waste-disposal and recycling system. Dealing with trash in Khumbu will require procedures for regularly disposing of garbage in local dumps, packing recyclables out of the region, and burning combustible trash. These practices must be carefully developed both to avoid pollution and to avoid offending local religious beliefs. Burning bad-smelling things at expedition base camps, for example, would be considered dangerous by many Sherpas, for it might offend the spirits of the mountains on whose slopes they risk their lives.

A new effort initiated in Khumbu during the summer of 1991 offers some hope that the garbage situation may soon be alleviated. Mingma Norbu Sherpa (former chief administrator of Sagarmatha National Park and now director of World Wildlife Fund USA's South Asian and Himalayan program) approached Khumbu leaders with a proposal to establish a Sherpa Pollution Control Project with the assistance of funds provided by World Wildlife USA. Mingma Norbu suggested that an ongoing, regionwide, community-organized cleanup and recycling program be initiated based on a combination of volunteer efforts and hired labor. The idea won considerable support from influential Sherpas and at a public meeting held in Nauje. A local committee was formed to oversee the project and several different operations were initiated. The first Sherpa cleanup of Mount Everest base camp took place a few days later. More than eighty Sherpas participated and much use was made of hired yak to pack several tons of garbage out of the base-camp area to a site where it was burned and buried. Permanent rubbish dumps have been established at a number of places. The project also opened an office at Sagarmatha National Park headquarters near Nauje where all expeditions are required to leave a deposit based on the number of loads they carry into the region. Thus far there is no provision to inspect base camps to ensure that cleanups have been carried out, but expeditions have to show they are packing out garbage in order to reclaim the deposit left in Nauje. The project has enormous potential, but it remains to be seen whether or not it can be maintained over a period of years.

Tourism, Pastoralism, and the Environment

During the early years of international concern over environmental degradation in Khumbu there were a number of reports of overgrazing. The original United Nation's Food and Agricultural Organization's recommendation for the establishment of a national park cited overgrazing and severe, localized gullying among the environmental problems which made a park necessary (Blower 1971).^[36] The New Zealand mission that laid the groundwork a few years later for the establishment of Sagarmatha National Park also reported overgrazing and gully formation (Lucas, Hardie, and Hodder 1974).

These early observations of overgrazing were supported by the first research on Khumbu herding patterns, grazing intensity, and pasture conditions. Bjønness concluded that overgrazing was a significant regional problem and reached this assessment on the basis of a livestock census, a study of the degree of grazing that took place in specific herding sites, and an analysis of pasture conditions on the basis of Landsat satellite imagery. She identified a number of overgrazed areas, including the immediate surroundings of the villages of Pangboche, Khumjung, Thami Og, and Thami Teng and the high-altitude settlements of Tarnga, Dingboche, and Gokyo and concluded:

Today overgrazing has led to heavy depletion of vegetation in many areas, leaving a remnant cover less capable of holding the soil. Early stages of erosion are much in evidence; the result is not only due to overgrazing, but also to forest depletion partly through firewood collection and browsing. (Bjønness 1980a :63, map and 74)

These findings suggested to her that national park action would be needed to regulate grazing, and she proposed both the development of a rotational grazing system based on grazing permits and regulations to limit the size of household herds.

Subsequent researchers, however, have questioned whether Khumbu is indeed overgrazed (Brower 1987; Byers 1987b ). In particular they have cast doubt on whether vegetative ground cover is as poor as has been reported or erosion as serious. Both Brower and Byers suggest that alarm over poor vegetative cover and presumably high erosive rates may have reflected an overreaction to the deceptively sparse vegetative condition of Khumbu pasturelands between late autumn and early spring by observers who were unfamiliar with the way in which Khumbu ground cover improves markedly before the onset of the summer monsoon rains in mid-June. They also note that early observers apparently failed to

appreciate that the total summer precipitation in Khumbu is considerably less than in the central and eastern Nepal midlands and that there are few occasions when rainfall is heavy. These conditions of climate and vegetation make for lower erosion rates than had been expected (Brower 1987:278, 280-282; Byers 1987b :229-231, 245).^[37] Byers' careful studies of soil erosion uncovered no evidence that serious erosion was taking place even in heavily grazed grassland areas in lower Khumbu. This led him to question whether even the extensive cattle-track terracettes on slopes in the vicinity of villages should be regarded as proof of overgrazing, for the shrubs that grow on terracette risers may act to stabilize the slope by impeding the down-slope progress of eroded material (ibid.:132,134). Byers suggests that "the hypothesis of over-grazed and eroded Khumbu pastures . . . is clearly in need of some revision" and that "the contemporary stability of the slopes and soil solum . . . indicates that they are and have been in balance with land use practices as observed in 1984" (ibid.:232). Brower, who conducted studies of Khumbu grasslands during the summers of 1984 and 1986, concurs, noting that although "Khumbu is a place that has been subject to human manipulation for a long time" and that "the result is a probably radical conversion of natural vegetation," this human impact on local environment "has enhanced its utility to people with livestock, without creating serious environmental degradation" (1987:303).

Yet while present levels of Khumbu grazing do not appear to be causing severe or extensive regionwide erosion, there still remain questions about whether or not localized overgrazing is taking place. Byers, for example, found that erosion from the slopes above Dingboche was much higher than in the sites he monitored in lower Khumbu (1987b ). He suggested that the sparse groundcover here was due to heavy grazing and the cutting of juniper and other shrubs for fuel wood.^[38] And if vegetation change rather than erosion alone is used as a measure of overgrazing, one might evaluate the impact of current grazing levels in lower Khumbu very differently. Here the issue, for example, might be whether or not current grazing practices are changing the composition of grassland and shrubland vegetation or lowering their productivity in terms of supporting cattle. Sherpa perceptions of changes in the availability and quality of grazing suggest that the productivity of some heavily used areas may indeed be declining. There remains a dearth of data, however, on the nature, pace, and severity of either pasture degradation or grazing-related inhibition of forest regeneration. In the absence of such data it is too early yet to judge whether or not subtle vegetation change is underway and what the role in it may be of the tourism-related practices of keeping larger numbers of urang zopkio, zhum, and cows. The issue of whether or not terracettes are an indication of overgrazing

also requires further consideration, for whereas erosion levels may not be high even from these areas, the productivity of the heavily trailed pastures is very probably quite lower than other grassland areas and they may well have a reduced carrying capacity. Heavily terracetted areas appear to have a much higher percentage of bare ground than other grassland areas and a far higher percentage of shrub species not palatable to Khumbu cattle. Only multiyear monitoring of pasture and woodland areas and the establishment of permanent cattle-exclosure study plots is likely to yield data to evaluate the current environmental impact of Sherpa grazing practices and determine the direct and indirect roles of tourism in pasture degradation and poorer forest regeneration.

Tourism and Deforestation

There have been many reports of widespread and severe deforestation in Khumbu due to tourism development. Several observers who have been important in developing Sagarmatha National Park policies, for example, noted in an article titled "Saving Sagarmatha" that

this mecca is fast becoming an environmental mess. The success of the Park over the past two decades brought more people (over 5000 a year) and with the masses has come the scourge of deforestation because of excessive cutting of forests for fuelwood. Since the watershed is now nearly destroyed, this in turn causes severe soil erosion and devastating spring floods in downstream areas from the Park.

Of course, deforestation is a general malaise in the Nepalese Himalayas, but it is particularly acute in the Park. (Hinrichsen et al. 1983:203)

They concluded that:

Sagarmatha has suffered more deforestation during the past two decades than in the preceding 200 years. Visitors to the Park are confronted with stark and denuded slopes in many areas, especially around Namche Bazar [Nauje]. Twenty years ago, foreign visitors described the area as rich in forest cover with lush stands of juniper. Today, these forests have been leveled, leaving only isolated clumps of shrub juniper. . . . Deforestation has, of course, been aggravated by tourism. (ibid.:204)

Another outsider familiar with the region, Sir Edmund Hillary, similarly contrasted the forests and high-altitude juniper shrub he saw in Khumbu in 1951 and 1953 with the much changed landscapes of the 1980's. He observed that

the valley of the Dudh Kosi river was still very beautiful, but the forest was woefully thinned by the axes and saws of the Sherpas, cutting timber for

buildings. . . . The forests around Thyangboche [Tengboche] had lost many of their mighty trees, and the Pangboche area was almost bare. Up the Khumbu Glacier Valley [upper Lobuche Khola valley] there was hardly a juniper to be seen. (1982:698)

Amplifying this testimony, Newsweek not long ago reported that:

Dense alpine forests once covered the snowy slopes of Sagarmatha, as the Nepalese call Mount Everest, and the steep sides of its Khumbu valley were lush with a dark green carpet of junipers. But that was the Everest of 1953, when Sir Edumund Hillary and his Sherpa guide Tenzing Norgay, became the first men to conquer the highest peak on earth. Today the dense forest at Everest's base is 75 percent destroyed, replaced by a tumble of rocks interspersed with lonesome trees. (Begley and Moreau 1987:104)^[39]

Bjønness reported that the "cutting of firewood is not only the most obvious [impact of tourism] but is also causing the most critical impact on the natural environment" and linked large cleared areas along the main trail to Mount Everest to trekking tourism (1980b :124,126). She and others have also suggested (incorrectly, as discussed in chapter 7) that the prominent, nearly treeless slopes above Nauje are another example of the destructive impact of tourism and forest nationalization (1983:270).

Such depictions of widespread Khumbu deforestation have been challenged in recent years by Charles Houston (1987) on the basis of his recollections and photographs of forest cover in 1950 and by several researchers (Brower 1987; Byers 1987a , 1987b ; Pawson et al. 1984; Stevens 1983). After visiting Khumbu in 1981 he announced that he had observed little change in forest cover since his 1950 visit there as a member of the first group of Westerners ever to enter the region, claiming that there is "as much or more forest cover [today] than there was in 1950 and I have the pictures to prove it" (Charles Houston, "Return to Everest . . . A Sentimental Journey," 1982; cited in in Thompson and Warburton 1985:120). Byers has questioned earlier depictions of wide-scale deforestation on the basis of a comparison of 1984 conditions and those shown in photographs taken in the 1950s and early 1960s (1987a :77-80, 1987b :205-221), as has James Fisher (1990:144-145).^[40]

It is obvious to anyone who visits Khumbu today that accounts of ongoing, extensive deforestation in Khumbu are considerably exaggerated. The new understanding of historical forest change in Khumbu that is emerging from the analysis of oral traditions, oral history, and pollen samples also questions whether tourism can be blamed for what forest change has taken place. This does not mean that tourism has not had any

impact on Khumbu forests, but these have been considerably more minor and highly localized than has been depicted. To assess these impacts it is necessary to change scales from looking for highly visible, regionwide deforestation to exploring possible recent change at the particular sites where the demand for fuel wood by mountaineering expeditions, trekking groups, and lodges has added to local demands on specific areas of forest, woodland, shrubland, and the scattered shrubs of alpine regions. This focuses attention on the specific sites that supply fuel wood to popular tourist overnight or lunch stops on the major tourist routes through Khumbu. On the route to Mount Everest these places are Nauje, the Tengboche monastery, Pangboche, Pheriche, Dingboche, and Lobuche, whereas on the upper Dudh Kosi valley trekking route they are Dole, Machermo, and Gokyo. The fuel-wood gathering sites where the greatest amount of added pressure has been focused are shown in map 22, including the areas near Tugla and Tsola Tso where large amounts of high-altitude juniper are gathered for the lodges of Tugla, Lobuche, and Gorak Shep, and where early Mount Everest expeditions also obtained much of their fuel wood. At these places tourism has indeed added significantly to the demand of local seasonal residents, sometimes dramatically raising the total amount of fuel wood gathered. The sites vary in their capacity to absorb this pressure, for they differ considerably in type of vegetation cover, altitude, and microclimate. No study has yet been done to monitor change in these specific sites, but Sherpas insist that although no extensive areas of forest have been cleared there may have been some thinning of woodlands near Dole (a major source for fuel wood and construction material for lodges at Dole, Machermo, and Gokyo) and that there has been a decrease in the juniper shrub cover in the Tugla and Tsola Tso areas. It is this change in high-altitude juniper cover that Hillary alluded to in his account and which has been exaggerated by the overly dramatic Newsweek report. These and other post-1979 changes that Sherpas have noticed in forests and woodlands are shown in the table below and in map 23. Those areas where Sherpas associate vegetation change with tourism (as well as with local use) are indicated with an asterisk. They believe that forest change in the other sites is related solely to local subsistence use.

Yet while Sherpas suggest that in some places tourism has accentuated local use of fuel-wood resources and helped diminish the density of certain small areas of forest and woodland they do not believe that any widespread loss of cover has taken place in the region. They find puzzling the reports that tourism has caused the stripping of the slopes above Nauje and Khumjung, for while they have witnessed localized change in the vegetative cover on those slopes they attribute this to very different causes. They are also puzzled by reports that the Mount

Map 22.
Tourism and Forest Use

Everest base-camp area has been deforested during the past few years, for throughout their lifetimes that area has been located far above the highest trees. Local perspective suggests that many accounts of deforestation have been exaggerated by people either too ready to find a crisis or too quick to attribute to tourism what are really historical changes in the landscape. This assessment has also been reached by preliminary photographic research (Byers 1987a ). It seems clear that lack of familiarity with the history of Sherpa forest use and the historical extent of Khumbu forest cover has led some people to confuse old and new change and link to tourism deforestation that took place long ago and as a result of other processes altogether. The impacts of local subsistence use on forests both before and after the nationalization of forests have also been underestimated or misattributed to tourism.

Yet it must also be kept in mind that even small-scale forest change may be highly significant for people who depend on forests for critical subsistence resources and have cultural concerns for the protection of sacred places. This seems especially important given that there has been little progress during the 1980s in slowing the demand for fuel wood by lodges. Even the Austrian-Nepal hydroelectric power project under construction near Thami Og in 1991 will alleviate the impact of tourism on forest use only in the Bhote Kosi valley and the Nauje, Khumjung, and Kunde areas, leaving current patterns of forest use unaffected along most of the routes to Everest and Gokyo. In the coming years tourism demand for fuel wood may well yet have an increasingly visible environmental impact, particularly if the government of Nepal succeeds in its plans to increase the scale of tourism.

Map 23.
Reported Forest Change after 1979

Tourism, Land Use, and Environmental Planning

Tourism development in Khumbu has thus had a variety of impacts on Sherpa land use and the Khumbu environment. Thes have, however, mostly been less dramatic than has been widely assumed by outsiders unfamiliar with Khumbu history, culture, economy, and environment. The subtler environmental impacts of increased urang zopkio herding or fuel-wood use by Khumbu tourist lodges, however, should not be underestimated.

If further study does indeed establish that current Khumbu herding practices are leading to a poor forest regeneration in certain areas and a decline of pasture productivity in heavily grazed rangelands, there may be a need for new regional livestock-management planning. This will clearly be difficult for either national park officials or communities to address. The lack of local customs that would impose ceilings on livestock numbers either in terms of household ownership or the number of head of stock permitted to graze particular areas and the strong value placed on the freedom of individual households to decide which type of stock to herd and where to herd it make any intervention highly controversial and all too likely to misfire. Perhaps the place to start would be with an effort to convince Nauje and Thamicho villagers to reestablish their former nawa regulation of summer pasture use and autumn wild-grass cutting. This in itself would relieve stress on some of the most heavily grazed lower Khumbu areas as well as improve the quantity and quality of hay made from autumn wild grass. This might also have some effect on some families' decisions about the size and structure of their herds. Grazing pressure can also be relieved by developing more supplementary fodder. Many Sherpas are interested in increasing the productivity of their hayfields through better fertilization and the introduction of new grasses and some are closely following the results of pioneering efforts to increase hay production through irrigation. The introduction of new types of grasses should be very carefully considered, for it could well have a long-term impact on regional vegetation if new species become established outside of the walled fields. If this is indeed a threat then a policy about it will have to be developed by Sherpa communities, for national park officials lack the authority to issue orders about land use within settlements. Another possibility for increasing the fodder supply is to import more hay from Pharak, if this can be done without adverse impacts on that region. An increased fodder supply will be especially important if the number of stock grazing in lower Khumbu and the village vicinities continues to rise. Pack stock have become critical to supplying Khumbu's lodges and trekking groups and their

numbers may well increase if the scale of tourism keeps growing. The recent arrival of a team of mules from the Annapuma region of central Nepal, owned and managed by people from that area, both suggests that the money to be made in Khumbu transport is enormously inviting and that this may continue to introduce new factors that will complicate regional pastoral management still further.^[41] The restoration of communal management institutions in the Bhote Kosi valley should thus be a high priority for both the regional Sherpa government and for Sagarmatha National Park.

Decreasing or even eliminating the use of local fuel wood in the tourism trade seems much more politically feasible. Sherpas have shown much interest in adopting alternative cooking and heating technologies, and the main obstacle to alleviating fuel-wood use for tourists is the cost of developing alternative energy programs. Little real savings of fuel wood have been accomplished by the two small-scale hydroelectric facilities constructed in the 1980s at Nauje and the Tengboche monastery. The 600-kilowatt hydroelectric project now being built in the Bhote Kosi valley may, however, have a great impact on fuel-wood use in western Khumbu, and substantially lower the need for it not only in the lodges of Nauje, Khumjung, and Kunde but also in ordinary village households there and in Thamicho. It does not seem likely, however, to totally fill the role of fuel wood in these communities. All the lodges of eastern Khumbu, including most of those on the trail to Everest and more than two-thirds of the lodges in the national park, will continue to rely solely on fuel wood and dung for cooking. More thought and funding must be given to decreasing fuel-wood use by high-altitude lodges and to alternative energy for Pangboche and Phurtse. The use of more fuel-efficient wood stoves and cooking techniques, kerosene stoves, and small-scale hydroelectric and wind power should be explored, as should such a simple measure as banning the selling of wood-heated showers. Regional alternative energy planning requires careful reconsideration, beginning with a continuing discussion among Sherpas, Sagarmatha National Park officials, and donor agencies to clarify the types and scales of projects needed in the region. A regional energy development plan to coordinate the activities of the diverse organizations interested in Khumbu alternative energy and forestry projects should be another early step.

It would be unwise to relax concern over managing the diverse environmental impacts of toursim simply because these have not yet been catastrophic. The discovery that there is less of a crisis than has often been proclaimed does not mean that the measures taken thus far have been unnecessary. Clearly they have not yet been sufficient to cope with even the relatively small-scale environmental pressures experienced to date. There are no guarantees that future changes in the nature and

scale of tourism will not have major new impacts on the regional economy, land use, and the environment. The development of more effective air transport or a road to the region would very likely rapidly swell tourist numbers, with major ramifications for the commercialization of regional natural resources, increased pressure on forests, local shortages of drinking water, and more serious pollution and solid waste problems. Even with existing access, Khumbu tourism shows no signs of leveling off and it can only be expected that if the government goal of quadrupling tourism within the next few years is realized this will have significant impacts on the Mount Everest region. One unfortunate corollary effect of the increasing scale of Khumbu tourism could well be greater interest by non-Sherpa Nepali and foreign entrepreneurs in the operation of Khumbu hotels and other tourist enterprises, a process all too likely to lead to the alienation of land and the loss of current opportunities for both Sherpa entrepreneurs and those employed in the trekking industry. This may already be underway. In the autumn of 1990 the Japanese Everest View Hotel reopened to great fanfare, including the helicoptering to Khumbu of the Japanese ambassador to Nepal and high-ranking officials from the Nepalese Ministry of Tourism. The airstrip at Shyangboche is back in operation, with several flights per day from Kathmandu of a Pilatus Porter aircraft that is leased to the hotel and also frequent helicopter service. Beyond renovating the earlier building the hotel operation has also begun expanding its facilities and may add more rooms. The autumn of 1990 also saw the arrival in Khumbu of the first hotel chain, Sherpa Guide Lodges, owned by a group of California investors in association with Kathmandu businessmen. No Khumbu Sherpas appear to be among the owners. The chain will operate thirteen lodges between the road end at Jiri and Lobuche, near Mount Everest. In the autumn of 1990 lodges were in operation as far as Nauje, and in the summer of 1991 construction was rumored to be about to begin at Dingboche and Lobuche. Reservations for rooms can be made in Kathmandu or booked through one of the foremost U.S. trekking companies. This chain may well begin to compete directly with Khumbu lodges for independent trekkers' business as well as shift some trekking groups from their tents into the chain's lodges. These two events raise major questions about the future of tourism in Khumbu as an avenue of local development, and they may preface an era of larger-scale tourism that may have have very different sociocultural, economic, and environmental impacts on Khumbu.

Sherpas and Sagarmatha National Park authorities have clearly not yet achieved an optimal integration of tourism with local land use and the environment. The challenge to do so only promises to become more difficult in the future.