Previous Page Table of Contents Next Page

Chapter III - Mongolia case study 1: Studies on long-distance transhumant grazing systems in Uvs and Khuvsgul aimags of Mongolia, 1999-2000 - B. Erdenebaatar


The entire study area is extensive grazing land, used by highly mobile production systems. The study looks at transhumance patterns that have re-appeared since privatization of the livestock industry in 1990, after forty years of collectivization. The study sites are in two western, mountainous provinces, both in great depressions surrounded by high mountains, where mountains and other physical features impose restrictions on the migration routes available. Both sites have a wide range of ecological conditions. Topographic features as well as weather and feed availability influence the itineraries chosen for migration, and in some cases the summer move is prompted by the arrival of biting insects. In both areas, all movements are longitudinal (west-east) to make better use of pasture resources and climatic differences. In Turgen, the movements cross five ecological belts, travelling twice yearly over two sets of passes: one for the spring journey and another for the return. The distance travelled annually is between 90 and 140 km. Rinchinlkumbe herders traverse only two ecological belts, making two long movements: from winter to summer pastures, and back. The distance travelled is between 90 and 180 km. There is a clear set of climatic and social indicators governing seasonal movements. The period of the study, 1999 and 2000, was a time of disastrous weather events. A severe drought, lasting several years, was followed by the severe winter of 2000, which caused much misery and economic loss. Herders were obliged to modify their travel patterns to save their stock as best they could.

Multispecies herding using hardy local landraces is the basis of the system. Horses, cattle, yaks, sheep and goats are the main stock, with some camels for transport. Success in herding lies in having stock fat enough in autumn to survive the winter-to-spring near-starvation period. Turgen herders move twice a year over the same passes between their summer and winter camps; those of Rinchinlkumbe make one movement to and from winter pastures. No herders stay at higher altitudes in winter. Seasonal, especially winter, camps are on one side of a mountain range that routes must cross. Herders of both sexes are highly literate, most to secondary level.

Since decollectivization, public services, including technical support to herders, have deteriorated, leading to problems of marketing and procurement of necessities. The major problem, however, is the lack of clear title to grazing and haymaking rights, which has led to trespass and squatting, and obliges many herders to restrict transhumance movements so that they can protect their winter camps. This exacerbates the deterioration of standards of pasture management, and some outlying areas are neglected while others are overgrazed and used at the wrong season.

FIGURE 3.1. Locations of the Uvs and Khuvsgul transhumant grazing systems studies and of the Arkhangai haymaking study.


Data subsetted from ESRI’s World Worldsat Color Shaded Relief Image. Based on 1996 NOAA weather satellite images, with enhanced shaded relief imagery and ocean floor relief data (bathymetry) to provide a land and undersea topographic view. ESRI Data and Maps 1999 Volume 1. Projection = Geographic (Lat/Long)

FAO Disclaimer

The designations employed and the presentation of the material in the maps do not imply the expression of any opinion whatsoever on the part of FAO concerning the legal or constitutional status of any country, territory or sea area, or concerning the delimitation of frontiers.

Selection criteria

The Mongolian livestock industry has undergone a great transformation since privatization in 1990. This study was designed to determine how herders now manage their livestock and to identify problems associated with their new situation. A major concern was to identify and select viable households representative of the community in terms of regularity of transhumant grazing movements and availability of necessary information. Criteria for household selection included:

Study sites and sample households

Thirteen herding households were chosen in Turgen sum, Uvs aimag, and fourteen (with six more selected in 2000) in Rinchinlkumbe sum, Khuvsgul aimag (Figure 3.1). Of those in Turgen, ten household heads were male and three were female; herding experience varied from 8 to 42 years. In Rinchinlkumbe, 12 heads were male and 2 were female; herding experience was from 5 to 38 years.

In addition to the selected households, studies in 2000 covered other families to cross-check data on livestock production and consumption at household level, and to gain more information on marketing and sale opportunities. Special attention was given to identifying alternative routes for long-distance movement; this necessitated interviewing more herding households and also local government officials.

Data were collected using participatory techniques, with group and individual interviews, semi-structured discussion, matrix ranking, etc. Special attention was paid to female informants. Secondary data covering the last three to five years were collected from the annual census and other valid statistical documents. The territory of the sums, the areas of the movement routes and camping sites of sample households were mapped, marking the seasonal movement routes of a group of households. Valuable forage plants in main seasonal pastures were identified and data on their natural productivity obtained by field observation and talks to herders. Data on herd performance of sums and sample households were gathered to estimate annual output of livestock products. Socio-economic and customary points directly and indirectly related to the transhumant movements, and factors determining the patterns of such movements, have been carefully considered and some indicators identified for cross-checking and verification.

Ecological and natural conditions of the study sites

Turgen is in the Great Lake Depressions along the Uvs and Khyargas lakes (Buyanorshikh, 1994). Most of Turgen is to the north of Uvs lake, stretching over the Kharkhiraa-Turgen mountains to the top edges of the Bairam-Yamaat ranges to the west. There are dry desert belts where, on the west side of swampy willow lowlands and dry desert, Uvs lake meets the forest belt of Kharkhiraa mountain and the basin of Uureg lake. Various authors (Buyanorshikh, 1994; Badarch, 1971) note that Turgen has an exceptionally harsh and extreme continental climate. Sum headquarters (Plate 14) is at roughly 49°15’ N, 92°04’ E, at the break of slope between the plains at around 1 100 m and the mountains, which rise to over 3 000 m. The highest areas have permanent snowfields and glaciers; the lowest areas are below 800 m.

Turgen is divided by a range of high, steep mountains running roughly northsouth. Spring pastures are on the plains to the east of the mountains, and winter grazing is on the other, western, side. At the lowest, easterly, extremity there are large areas of marshland, suitable for autumn and spring grazing and haymaking (Plates 15 and 16). It is arid to semi-arid, according to altitude (average annual rainfall at Ulaangom is 154 mm). Lowland vegetation is desert steppe, except in marshes. Part of western Turgen is a nature reserve. The climate is extreme, with very cold winters in the lowlands. Winter camps are in the hills.

Plate 14. Turgen sum headquarters. Herders and their herds cross these mountains four times each year.

The total area is 225.3 km2. Usable land covers 210.5 km2, of which pasture is 98.5 percent (Table 3.1). Turgen has vast winter-spring pastures, which provide plenty of room for seasonal camping and grazing for both transhumant and shortdistance migration. The sum is famous for its natural diversity, which has attracted herders and people from other sums to settle, making some seasonal pastures more crowded, leading to overconcentration of livestock and people in some parts of the territory. Geographically, the winter-spring pastures are unsuitable for grazing at other seasons; they are in a defined part of the sum and have strict boundaries and conditions for grazing.

Land resources in Turgen sum, Mongolia.

Land category


Aimag total





Land utilized



6 229.0


of which pasture land comprising



6 064.0


- summer-spring pasture



2 109.5


- summer-autumn pasture



2 661.1


- winter-spring pasture



1 195.6


Land not available





Total territory


6 985.6

Plate 15. Manual haymaking in the Uvs Lake basin.

Rinchinlkumbe sum headquarters is at 51°01’ N, 99°40’ E, at 1 583 m. It is also in a very mountainous area, but with much higher rainfall, 284 mm annually. It has an area of 1 400 km2, with some 18 percent covered by water and the rest is a state reserve. Pasture areas are as shown in Table 3.2. It has a large area of seasonal pastures on both sides of Bayanuul (Rich Mountain). In comparison with Turgen, it has a great network of surface water, including Khuvsgul Lake - the largest in Mongolia. Over 800 ha of land are reserved for haymaking.

Major land categories in Rinchinlkumbe sum, Mongolia.

Land category




Land utilized

1 288.1


of which pasture land(1) comprising

1 092.0


- spring-summer pasture



- summer-autumn pasture



- winter-spring pasture



Total territory

1 400.0

NOTE: (1) Includes sparsely distributed and lowland forest.

Plate 16. Measuring primary production in the marsh haylands, Uvs Lake basin.

Rinchinlkumbe is less varied in terms of landscape and altitude range. It has five major landscapes, all belonging to the Sayan-Khuvsgul region, namely:

Land-forms and altitude range

Turgen contains all the ecological zones of Mongolia: from desert steppe to high alpine, including steppe, forest steppe and Gobi-type landscapes. Based on literature sources (Erdenebaatar, 1996; Buyanorshikh, 1994) and information from herders, the topographical elements in Turgen and their use patterns are:

Rinchinlkumbe has interesting land forms, with elements of alpine, taiga, lowlands and high mountains. Lowlands (1 200-1 500 m) for late spring-summer and summer-autumn grazing are between mountain ranges over 2 500-3 000 m, whereas winter areas are in mediumaltitude forest steppe and steppe between 1 000-1 400 m. Some transhumant herders stay at 1 900-2 300 m over winter and early spring. No single area is used all the year round; there is a strict seasonal division according to ecological type and topographical conditions. Land such as high hills, low mountains and lowlands will be used several times in a year because of the linear location of resources and westeast direction of grazing movements: all upward movements are westwards and the return to lowlands eastwards.


Dominant soils are light brown in the steppe and desert light brown in desert-steppe. Brown and light brown soils are found along mountain sides and the slopes of low hills and hillocks. Small areas of mountain meadow, with alpine dark soils, are present in high mountains like Kharkhiraa, Turgen and Bairam. Saline soils are widespread in the proximity of Uvs Lake and the dry margins of swampy lowlands. There are no sands or sand dunes in Turgen, except small stony plain lands and alluvial fans at the base of mountains. Some mountain areas are not used for grazing due to excessive bare rocks, steep clefts between mountains and permafrost.

Soils in Rinchinlkumbe are fertile compared to Turgen, but a variety of permafrost, high mountain and lowland soils can be found. Most soils of the mountain edges are tundra, mountain meadow and permafrost meadow dark brown. Lowland areas form permafrost meadow, meadowswamp dark and meadow-swampy brown soils. Soils around Khuvsgul are lake meadow dark brown and mountain dark, as well as chestnut. Carbonate soils are present in lowland areas and riparian land of the lake.


Buyanorshikh (1994) reports annual temperature fluctuations of 80-90°C in the Uvs Lake Depression, which contains a third of the sum grazing land. This occurs as a result of specific geo-ecological patterns of the area: land-locked and heavily dissected by sequenced mountains and lowlands which hinder the exit of free flows of arctic winter and north summer Siberian cyclones/anticyclones. The Kharkhiraa-Turgen mountain ranges, which are less cold and snowy, are the winter camps of most of the herders. Badarch (1971) suggests that winter camping areas in Turgen have a cold Altai winter, a mediumcontinental climate.

These topographic peculiarities determine the annual temperature regime: mean annual temperature is +1.8 to 3.6°C. The annual range is some 70°C - from a maximum (in July) of +30°C to a minimum (in January) of -40°C. Annual precipitation in the Uvs Lake depression is 123-218 mm; 200-250 mm in the Kharkhiraa-Turgen mountains; and around 180-200 mm in other parts of Turgen. Rainfall is very unimodal: 85-90 percent falls in April-October, with over 30 percent in July alone. Snowfall is 4-30 mm, but heavy snow occurs around Uvs Lake. The prevailing wind is from the west, with an annual average speed of 1.3-3.0 m/second.

Records from Rinchinlkumbe indicate an annual average temperature of -2.1°C, with average isothermic values of -25°C in January and +23°C in July. The minimum temperature recorded was -43°C and the maximum +32°C. The climate of the sum is largely determined by east Siberian cyclones and microclimatic flows off Khuvsgul lake. Annual precipitation in the lowlands of Rinchinlkumbe is under 250 mm, about 300 mm in mountainous areas and over 400 mm in alpine and tundra. Rainfall is unimodal: much falls in June to August (over 60 percent of the total). Snow arrives from mid-November to January, with 12-20 cm (maximum 32 cm in a normal year). Summer areas of the sum are classified as dry cool, while most of winter camps of the sample herders are cold. Tundra and alpine zones are the coldest. The growing period is 75 to 115 days.

Vegetation forms and species composition


Turgen is in the Kharkhiraa-Turgen vegetation subregion of the Mongol Altai mountain steppe province, and in the Sagil-Davst and Ulaangom areas of Uvs lake semi-desert subregion of the North Gobi desert-steppe province of Eurasian Steppe region. Major ecological zones and belts were identified with the help of the overview of Buyanorshikh (1994). Samples were collected from permanent sites and randomly selected areas to determine major plant species composition and approximate condition and productivity of pasture types. Plants were grouped into forage, toxic and unpalatable. The following range types and major plant species were identified in Turgen.

Semi-shrub - tussock grass desert steppe Caragana pygmae, C. bungei, Stipa glareosa, Artemisia frigida, A. xerophytica, Kochia prostrata, Agropyron cristatum, A. nevskii, Stipa krylovii, S. gobica, Allium mongolicum, Oxytropis aciphylla, Anabasis brevifolia, Cleistogenes soongorica and C. squarrosa.

Grass - semi-scrub desert steppe Allium polyrrhizum, A. mongolicum, Reaumuria soongarica, Anabasis brevifolia, Ajania achilleoides, Caragana pygmae, C. bungei, Oxytropis aciphylla, Stipa glareosa, Hedysarum fruticosum, Artemisia frigida, Cleistogenes squarrosa and C. soongorica.

Riparian meadow - swampy lakeside Salix-grass meadow Salix ledbouriana, Betula microphylla, Caragana spinosa, C. bungei, Leymus angustus, Phragmites communis, Achnatherum splendens, Iris lactea, Agropyron cristatum, Oxytropis salina, Halerpestes ruthenica, Saussurea salsa, Kalidium gracile, Nitraria sibirica, Stipa glareosa, Oxytropis aciphylla, Plantago salsa, Poa pratensis, Carex lithophila, Juncus salsuginosus, Sium suave, Inula linariaefolia and Hordeum brevisubutum.

Tussock grass dry steppe Caragana pygmae, C. bungei, Agropyron cristatum, Stipa krylovii, S. orientalis, Poa attenuata, Koeleria macrantha, Veronica pinnata, Cleistogenes squarrosa, Festuca lenensis, F. valesiaca, Arenaria meyerii, Leymus chinensis, Bupleurum bicaule, Youngia tenuifolia, Aster alpinus, Astragalus mongholicus, A. brevifolius, Ephedra sinica, Artemisia monostachya, A. frigida, A. rutifolia, A. santolinifolia, Allium prostata, A. mongolicum, A. eduardii, Thalictrum foetidum, Elymus sibiricus and Orostachys spinosa.

High mountain meadow Festuca lenensis, Poa attenuata, Koeleria altaica, Koeleria sp., Potentilla sericea, Aster alpinus, Festuca krylovii, Leymus chinensis, Stellaria pulvinata, Arenaria formosa, Potentilla nivea, Carex rupestris, C. amgunensis, Silene repens, Kobresia smirnovii, Saxifraga ceruna, Rosa acicularis, Geranium pseudosibiricum, Galium boreale, Vicia cracca and Bromus korotkiji.

Tundra alpine meadow Kobresia sp., Carex rupestris, C. pediformis, Poa siberiana, Festuca altaica, Polygonum viviparum, Hedysarum alpinum, Astragalus frigidus, Vaccinium vitis-idaea, Launaea sp. And Pyrola incarnatus.

Coniferous (Larix sibirica) forest Larix sibirica, Pinus sibirica, Lonicera altaica, Vaccinium vitis-idaea, Calamagrostis obtusata, Cicerbita azurea, Geranium albiflorum, Trollius asiaticus, Festuca altaica, Polygonum viviparum, Hyloconium splendens, Aulaconnium palustre, Climatium dendroides and Rhytidium rugosum.

High mountain herbs - tussock grass meadow steppe Helictotrichon desertorum, Elymus aegilopoides, Polygonum alpinum, Potentilla evestita, P. biflora, Artemisia monostachya, A. laciniata, Helictotrichon schellianum, Carex pediformis, Dasiphora fruticosa, Leontopodium ochroleucum and Vicia multicaulis.


The sum territory contains the following vegetation types.

Polytridrum rocky mountain tundra and high alpine Polytridrum sp., Carex bigelowii, Waldheimia tridatylites, Sourea coracephola, Logatis altaica, Lloydia serrotina and Saxifraga sibirica.

Polytridrum, Carex-Polytridrum and Polytridrum-semi-bush treeless high alpine Carex sp., Betula rotundifolia, B. humilis and Cladonia cetraria.

High mountain sparse rocky and grass Caragana jubata, Carex sp., Carex microglochin, C. capitata, Eriophorum brachyantherium, Lloydia serrotina and Logotis altaica.

Abundant Carex-Koeleria-Festuca grass Larix forest Larix sibirica, Polytrichum cladonia, Kobresia bellardii, Festuca lenensis, Polygonum viviparum, Ptilagrostis mongolica, Salix berberifolia, Cetraria sausurea, Claytonia joanneana, Ledum palustre, Potentilla nivea and P. gelida.

Koeleria-Festuca mountain steppe Kobresia bellardii, Festuca lenensis, Polygonum viviparum, P. angustifolia, Ptilagrostis mongolica, Potentilla nivea, P. gelida, Koeleria macrantha, Poa attenuata, Aster alpinus, Artemisia comutata, Bupleurum bicaule, Potentilla sericea and Pulsatilla turzcaninovii.

Picea-Betula-Caragana jubata-Carex meadow Picea obovata, Betula rotundifolia, Caragana jubata, C. tenuifolia, C. capitata, C. microphylla, Carex melanantha, Salix reticulata, S. saxatilis and many other herbs.

Pasture condition

In order to assess the condition of major types of natural pasture in Turgen sum, Uvs, and Rinchinlkumbe sum, Khuvsgul, seasonal sampling was done in May, August and November 1999 and 2000.

An overview by Buyanorshikh (1994) was used as reference to cross-check findings.

Data from field surveys (1999-2000) require several observations on the natural productivity of pasture: precipitation was extremely poor, with little snow in winter, and few, late, short-lived heavy showers in summer; so that precipitation was 70-80 percent lower than normal. The spring, especially, was very hot and dry, which greatly delayed regrowth. Surveys in Turgen in 2000 showed a very different situation; with the exception of the early spring months until the thaw, it was pleasant due to abundant snow in winter and more rain in summer. Pasture growth, which had been considerably affected by drought, recovered to a reasonable extent and forage was readily available (some 30-40 percent more compared to 1999 in all areas in all seasons).

In 1999-2000, the weather in Rinchinlkumbe was relatively pleasant in terms of temperature fluctuations, precipitation and pasture grass growth. The data in Tables 3.3 and 3.4 show natural forage available to livestock by season, and earlier data from the literature. The 1999 data were collected from areas where few animals and people concentrated. No fencing was used but some attempts were made to prevent continued and heavy grazing, so that the real availability of natural forage could be approximated. Herders were interviewed to find out the initial productivity of each pasture type; their simple calculations indicated that yields were 10-30 percent below what non-grazed pasture can produce.

The comparable usable pasture in summer and autumn of 2000 (real winter begins from late November), is seen by experienced herders of Turgen to be the result of two major factors:

Meteorologists suggest that the normal range of snow depth in Turgen is between 7 and 15 cm and snowfall is mostly in November-December. This is completely different from that reported for winter 1999-2000 and spring 2000, which caused a number of unusual changes in the transhumant movements of local herders.

Areas with some soil moisture reserves (rich underground water, permanent swamps and other non-rainfed sources) grew some young grass, but were grazed immediately by all graziers in a very competitive manner. This led to localized overconcentration of livestock, making forage availability even lower at all seasons. Neither herders nor meteorologists expected so severe a drought in 1999. In the project studies, for example, a 75- year-old, who has been herding all his life, could not remember a drought like the summer of 1999, nor a severe winter like 2000. According to his assessment, the 1999 drought and 2000 snowfall covered a large area and affected all neighbouring sums (Sagil, Tarialan, Chandmana and Davst); and covered all ecological areas, leaving herders with no chance to use grazing elsewhere.

The weather of 1999 and part of 2000 is stressed because of its effect on studies of pasture productivity and range condition. Due to the atypical weather, past experience had to be relied on to provide details of movements, seasonal grazing patterns, pasture condition and other important details.

Forage productivity of types of natural pasture in Turgen sum, Mongolia (kg/ha).

Range types

Field survey (1999)

Buyanorshikh (1994)




Semi-scrub - tussock grass desert steppe


1 050


1 600-3 500

Grass - semi-scrub desert steppe




1 500

Riparian and swampy Salix - grass meadow

2 460

4 530

2 180

10 000

High mountain meadow

1 020

2 100


4 800

Tundra alpine meadow


1 000


800-1 000

Coniferous (Larix sibirica) forest


2 800

1 000

3 500

High mountain herb - tussock grass meadow steppe

1 200

2 700


4 700-6 300

TABLE 3.4 Approximate forage production of natural pastures in Rinchinlkumbe sum (kg/ha).

Pasture type






Lowland Salix meadow

1 000-1 200

3 300-4 300*

2 000

1 200-1 300*

Woodland and mountain foot

800-1 000

3 900-4 700*

3 100-3 600

700-2 100

Mountain valley and woodlands


5 500-8 800*

5 200-7 800*

3 300-3 500

Low-hill pasture

1 500-1 700*

1 700-1 900

2 100-3 400

1 300-2 100*

NOTES: * data presenting production of forage in grazing-free seasons

Conversations and interviews with herders suggest that they sense a deterioration in the condition of the sum’s natural grazing, because of ecological and social factors:

These changes in the rainfall distribution have a strong influence on grass growth, and on grazing patterns as well, as some areas of Turgen have suffered large-scale trespass by herders and stock from other sums. This was partially described by Erdenebaatar (1996). Serious cases of trespass and unauthorized (informal) residence have two consequences:

No reasonable solution has been found for this situation. Although summer and winter in 2000 were good, local herders still worry seriously about trespass. Since the first arrival of trespassers in Turgen, herders have had, each year, to cede parts of the area where they have been grazing since time immemorial; they are worried that this will continue unless there are emergency measures to stop trespass (some herders argue that they would be pushed out of the sum territory). A conclusion from the above is that herders who are displaced from their home areas are forced to change their seasonal camping patterns, and a knock-on effect causes others to do the same. Herders who formerly made long movements had either to shorten the distance or move even farther. Some herders have changed seasonal movement routes completely and stay two seasons in the same area, thus violating the informal land tenure system and legal land rights.

A visual assessment and a partial transect over crowded grazing areas suggest that crucial changes in the vegetation cover, species composition and overall productivity have taken place due to several factors, including:

Despite obvious pasture degradation in parts of the sum, herders share a common perception that the pasture is in good condition. This is indefensible for two clear reasons:

Data on pasture degradation in Turgen are consistent with the latest information on the pasture condition in many parts of the country. Tserendash (1999) identified the condition of seasonal pastures in Uvs as shown in Table 3.5.

High mountains (the highest in Uvs) and some swampy willow forest areas of Turgen remain intact because they are inaccessible to stock. Some areas at the outlets of mountain passes are heavily degraded. In late May, there were signs in Rinchinlkumbe of the result of regular movement of large herds at least twice yearly: in early November to winter camps and in March-April on the way back to warm areas.

Proportion of degraded pasture, Turgen sum, Uvs aimag.


Degree of degradation (percent)






Uvs aimag(1)






Turgen sum(2)






SOURCES: (1) Tserendash, 1999. (2) Approximate estimates, 1999.

Conversation with herders showed that they do not really believe that these areas are degraded. They think that they are in a biological rest, beyond the grazing frontier. This is hard to believe if the movement patterns followed by herders in Turgen are studied. The above-outlined pasture degradation has taken place in all grazing areas used by transhumance systems.

Plate 17. Temporary camps on the autumn migration. The ranges have to be crossed yet again. Tarialan, Uvs, Mongolia.

Long distance movement in the study site involves several clear patterns:

Any seasonal or long-distance migration has, therefore, to be organized on the basis of routes defined by passes (there are no routes they can use all the time); the need for suitable pasture; and available transport. The lack of permanent facilities where they can store items not required on migration increases the cost and labour requirements for moving. Long movement in the study area has strong traditional roots, as well as practices inherited from the collective period. To determine how households secured, or are securing, grazing rights and camp sites, all family heads were interviewed. The results are summarized in Table 3.6.

Criteria for security of grazing rights.


Turgen sum (n=14)

Rinchinlkumbe (n=8)

Grazing rights secured because:

- grazed since collective period



- selected voluntarily (mostly by joining others)



- trespassed (i.e. without official permit)

3 (winter camps in Sagil sum)


Herders secured their grazing rights in several ways: some trespassed on a neighbouring sum; others moved to areas where they had had access in the collective period. A case of trespass in Turgen began as a result of a dispute in early 1992: some moved into Turgen from neighbouring sums and pushed out some local herders from their lowland and winter areas. In search of new space, these displaced herders entered Sagil and were accepted by the local community, but the “new” residence area straddles the border of two sums, which is why the local community had to allow their settlement. According to the environment protection officer of Turgen, trespassers are “plotting” to occupy more land by clandestine means, building camps in the heart of winter and summer areas, as they start their first settlements in distant lowlands.

In 1999, many herders (four from sample households) from Turgen had no choice but to enter part of the winter pasture of Sagil sum to share existing resources, due to hard drought and short grass in summer 1999 and deep snow in winter 2000. Neither local nor visiting herders consider this trespassing, saying it is a part of reciprocal access to key or available resources; it is an important assistance that herders always expect at a time of disaster in their own area. They strongly believe that the only way to cope with similar shocks is to seek reciprocal permission from neighbours.

In February-March 2000, many herds were crossing the border of Turgen and Sagil sums. Many of the large animals, however, were left behind in the mountains, unattended, because, first there was not enough grazing on the way to support big herds, and, second, large animals are usually not friendly to small ones in sharing grass. Large animals push ahead of small ruminants and graze throughout the night, while small animals suffer heavy predation by wolves and snow leopards, but herders were not greatly concerned, saying that it is cleaner and more acceptable than having their animals dead at the edge of the khot - the herders’ camp area.

Human population

Turgen has three bags: two rural and one at the sum centre. In 1998, it had a population of 2 483, of which 1 839 belonged to 365 households that were purely herding. The sum had 62 104 head of livestock in 1998: 765 camels, 4 579 horses, 7 435 cattle, 32 726 sheep and 16 599 goats. In the survey, family size varied from household to household. Average family size was 4.7 (range 2-9), of which 2.92 were male and 1.78 female. According to national age classification, the number of working age per family would be 2.6 (1.7 males and 0.9 females). Each family had 1.14 children under 16, a source of labour in the peak period outside the school term.

In late 1998, sample households in Rinchinlkumbe had a family size of 5.2: 23.8 percent under 16 years old; 59.5 percent of active working age; and 16.7 percent pensioners and members of early pension age. The sample households in both sums owned five species of animals, following the common patterns of multispecies herding. The species composition and age and sex groups of livestock of the sample households are shown in Tables 3.7 and 3.8.

The number of livestock per household in Turgen varied widely: from none to 800 or more. Average herd size was 267 head in 1993, and increased to 358.2 in 1998. Livestock per family member was 76.2, and 135.7 per member of working age. There was no increase in the number of people engaged in herding, so the herd-upgrading strategy common in the Mongolian pastoral economy has been taking place here too.

Adult females in herds were 22.3 percent for mares, 32.1 percent for cattle, 39.0 percent for ewes and 38.3 percent for does. There was enough young stock for herd replacement: 11.7-31.0 percent. The species and age composition of herds depends on personal preference and the herding goals of individual families. It is not possible to give standard figures to describe herd growth strategies among sample households. Rinchinlkumbe herders own more yaks than those in Turgen, perhaps because they need male pack yaks for transport and have a lower herd offtake. Horses are important for riding and herding.

Sample herd composition statistics, Turgen sum, 1998.



of which

















































2 953




1 150




1 073



1 345











5 015




1 865




1 637


NOTES: (1) Including stud males. (2) Male and female are aggregated. (3) As of December 1998.

SOURCE: 1998 Livestock Census Report, Governor’s Office, Turgen sum.

Sample household herd composition, Rinchinlkumbe sum, 1999.



of which






































1 118











1 858











1 092











4 726




1 781




1 320


NOTES: (1) Including reproductive males. (2) Male and female aggregated. (3) As of December 1998.

SOURCE: 1999 Livestock Census Report, Office of the Sum Governor.

In early 2000, livestock per household in Rinchinlkumbe were 315: 63.8 per family member or 98.4 per member of working age. No remarkable increase in the number of people of sample households engaged in herding was identified, explained on the basis that the wealth of the herders of Turgen selected for the study was relatively high, since long distance movement is mostly made by rich herders; and, in Rinchinlkumbe, all herders must move long distances because of specific climatic and geographical conditions; sample herders were selected randomly, including ones with fewer or more animals.

The low proportion of adult female and young camels is due to the high number of castrated males. Turgen and Rinchinlkumbe are not traditional camelbreeding areas. Some camels are used as transport during migrations and for short-distance transport of living needs. Turgen herders use male camels for riding more in winter months than herders from Rinchinlkumbe. This can be easily explained as camels are exceptionally hardy and endure chronic feed shortage; and the geographical and topographical characteristics of winter ranges in Turgen are flatter than in Rinchinlkumbe, and camels are more popular for riding, whereas Rinchinlkumbe winter areas are heavily dissected and dangerous for humans and animals to walk over.

Stud males, breeding females and male:female ratio, Turgen sum, 1998.


Stud males

Breeding females

Ratio (1)















1 150






NOTE: (1) Number of female stock per stud male.

Herders believe that overall herd growth depends on the number of stud males and the number of females per male. The male:female ratio of small ruminants of the Turgen sample households (Table 3.9) does not match national averages; there is very high mating pressure on rams and bucks. The pattern appeared similar in Rinchinlkumbe. Nevertheless breeding pressure on rams and bulls seems to be less than in Turgen as a whole. From individual household statistics, many households have no male camels, bulls or stallions, so there is a risk of low conception, or they borrow males. Some herders have to delay mating until others loan males, or are told to keep their females away until the end of the season, especially camels and mares.

The low numbers of breeding male stock reflects the small herd size of most households, which does not support a large number of mating males nor provide broader opportunities to select good sires. As some herders have to keep the breeding males with their herds all year around, this makes keeping many male stock difficult because of a high probability of out-of-season mating.

Current management

Mongolian pastoralism is based on nomadic grazing of seasonally available pasture and other grazing resources, such as naturallyavailable minerals, water and other deposits, to cope with a harsh environment and frequent natural disasters.

Some official sources suggest that natural disasters (known as zud, an aggregation of hardships) have become more frequent. Previously, the expected zud occurrence interval was 7-10 years, but recently it has been 3-8 years. At the same, the size of the area affected has been expanding. The 1999-2000 zud covered 13 aimags out of 21, and 157 sums out of 300. So serious a disaster has been recorded only once since 1944-1945. In 2000-2001, zud conditions were also widespread and devastating.

In such circumstances, better herd survival and improved returns from stock raising depend on open grazing of natural forage. So pastoral livestock management, in terms of long-distance transhumance, focuses purely on careful and effective use of available sources of natural pasture. Grazing efficiency under transhumance is determined by many internal and external factors, affecting the timing, distance and other arrangements.

Timing and pattern of herd movement and household

It is known (Erdenebaatar, 1996) that the fundamental principles of traditional grazing management are based on cyclic grazing in the four seasons of the year. The long movements in the study area are strongly seasonal, so some criteria have been identified for the timing of migrations between grazing areas, including macrolevel indicators and natural processes.

Change in pasture quality or productivity of grazing areas determines the timing of seasonal movements, and this productivity is determined by two primary factors:

The comments of herders (see Box 3.1) show that herding movements are not spontaneous, but a clearly determined process. Seasonal movements allow herds to graze fresh forage and protect them from exposure to extreme climatic shocks.

Differences in relief and topographic conditions in different areas are important if long-distance movements have to be made. This is largely a case of the selection of seasonal camping areas and pastures. This shows how herders recognize macro- and microclimatic differences that are significant for them. Selection of seasonal pastures is based on several factors that assist herders in deciding on herd and family movements. In Box 3.2, a herder from IV bag in Rinchinlkumbe comments about local knowledge when selecting seasonal pastures.

The herder’s comments show that each seasonal pasture in Rinchinlkumbe has unique conditions that the herders can exploit. Very similar information can be found for Turgen, but herders do not stay in lowlands in summer and woodlands in winter.

BOX 3.1
Seasonal movement timing and changes in pasture productivity

Mr A., a herder in Bayankhairkhan bag, commented on how changes in pasture productivity determine the timing of long-distance movements. From experience over many years, he says that in late August, when all stock must move to lowland autumn pasture, the average yield of pasture is 500-800 kg/ha, falling to 100-200 kg/ha by mid-September. This serious drop in available herbage spurs herders to move to areas between autumn and winter camps because, if they linger, their stock will not be sufficiently fattened and their ability to survive the winter will be reduced. When the end of a season has less forage than usual then immediate moves are necessary. When there are sharp changes of weather herders cannot stay and must move. There is, therefore, a set of weather factors and norms which tell herders when the time to move to another seasonal area is approaching.

BOX 3.2
Herder’s perception of selection criteria for seasonal camping area and pastures

Seasonal movements aim to make the best use of positive differences in the overall climatic conditions between grazing sites.

- Spring camping (khavarjaa) areas and pastures are in areas where grass regrowth and production starts early, so that animals can get fresh grass to recover their body condition. It is also important for young animals to get more milk. These pastures have to be capable of supporting herds for a month or more until early summer pastures are well grown. Spring pastures must be sheltered from cross winds and not exposed to winds coming over high mountains (locally known as “mountain cold daily winds”). For Rinchinlkhumbe herders, spring camps are in the bottom of stream canyons and valleys.

- Summer (zuslan - from mid-April to late August) areas have to be purely in lowlands in the Darkhadyn Khotgor (a 200 km-long lowland along the west slopes of the Khordil Sardig Range), where more open, mosquito-free pastures are available. These areas are not suitable for use in cold months due to continued low temperatures and deep permanent snow. High and open hill tops and open hollows are most suitable for summer residences. Abundant natural soda, open water and high grass are also important conditions for summer pastures and camp sites.

- Autumn pastures (namarjaa) are chosen in low mountain-foot woodlands and mid-valley slopes facing the sun. Autumn pastures are a staging area between summer and winter pastures, where animals find fresh and slowly drying grasses to put more weight on to obtain better consolidated body condition.

- Winter camping sites (uvuljuu) and pastures are on the eastern side of the Khordil Sardig Range, mostly in open areas in narrow valleys and woodlands. The most important conditions for selecting winter areas are high standing forage and less cold. The informant confirmed that their winter area has temperatures higher by 5-10°C than other parts of the sum, and hopefully protected from cold and snow winds.

Changes in pasture quality are due to weather and seasonal forage consumption by livestock. This means that if weather is far worse than normal, some areas have to support more animals for longer in a grazing season, and important grasses are quickly and heavily grazed, leaving less palatable species. Thus pasture quality changes, necessitating movement to other areas. In such cases, herders, including some who usually move short distances, have to go in search of land with available forage. When weather conditions are inclement in one place, an overconcentration of households and herds in a comparatively small area may occur, which in turn leads to quicker depletion of existing resources, and all herders have to move - both resident herders and newcomers.

Seasonal availability of other resources (water, natural salt licks, etc.)

Herders of the study sites argue that water availability at cold times of the year and naturally-available mineral sources make them move long distances. Herders say that “natural minerals stop worsening animal body condition and help consolidate animal fatness, and a mouthful of stream water in winter is equal to a khormoi [a traditional measurement unit equal to 3-5 kg of dry and frozen materials] of snow”. In winter, spring and late autumn - the hard times of the year - animals suffer if they have to walk unnecessarily to find water and minerals.

TABLE 3.10
Weather- and resource-based indicators for seasonal migration timing in Turgen sum.

Seasonal movement

Typical timing

Criteria and indicators

Leave winter camp for spring area

20 March - 10 April

Melting of snow cover; cross-winds; poor forage availability.

Move to summer pasture

10-20 June - early July

Hot; arrival of mosquitoes; rare rainfall.

Return movement to autumn pasture

10-20 August

Cool; early frost; heavy wind and rain (more frequent and heavy hail); poor pasture.

Move to winter camp

Early November

Cold cross-winds and heavy breezes; early snow in lowlands; poor forage availability.

Other important conditions looked at and carefully evaluated when planning immediate and future movements are natural indicators, including:

Centuries-old herding practices and modern scientific information are the basis of how herders select the timing and organization of migrations. These are widely referred to as movements between seasonal camps; some criteria for deciding on the timing of movements are summarized in Table 3.10. There are, however, some camp-changing moves in particular seasons which are do not have strictly fixed timing.

The number of intermediate camps depends on the distance between the area of departure, the destination area and the season. Sample herders in Turgen make 5-6 temporary camps of 1-3 days to reach winter camp, covering a distance of 20-25 km daily. Herders of Rinchinlkumbe say that the distance between two consequent temporary camps for small ruminants is 12-18 km per day, sometimes less. The duration of long migrations is limited by how far small stock can walk daily. Topography also determines daily movement. Herders in the study area have to cross huge mountain ranges, so intermediate moves are short. Rinchinlkumbe herders need to make 6-8 intermediate camps to get to winter areas and move back to spring areas. The daily distance is shorter than in Turgen because of difficult topography: high passes, narrow and steep ravines, and single-track trails along which to walk herds and pack animals.

Animals are moved almost together. Small stock walk with the caravan or are trekked as a single flock; large animals may be moved separately. Herder interviews were indecisive as to whether there was any noticeable difference if personal belongings, including ger and furniture, are moved by motor transport or by pack animals.

In areas with very difficult relief, people prefer pack animals; more open, less broken, areas allow use of motor vehicles. Use of motor transport is common for spring moves, when many young stock have to be moved. Moving the dwelling and other facilities by motor transport does not change any of the important components of migrations since transhumance itself is the action of moving herds from one area to another. Around 90 percent of sample households in Rinchinlkumbe use male yaks, cattle and a few camels for transporting gers and furniture; over 60 percent of herders in Turgen move using camels.

In winter-spring 1999-2000, sample herders in Turgen organized wintering of their animals in a specific way: large animals, like horses, bulls and dry cows and some camels, were sent for free range, unattended grazing on distant pastures in high mountains and flat valleys, while small stock remained at the herders’ camps.

Community decision-making

Formal decisions about long seasonal migrations are not made communally, but herders who are to move to the same area (within certain locations) consult each other. Moving as a single family is not desirable, but neither is many households moving together encouraged. Two to three families prefer to start together. Sometimes herders may agree to hire motor transport and make a common accord on timing and payment. They did not comment on whether they need community decision making procedures to make migration more organized and effectively arranged.

An important issue in decision-making could be delegating somebody to visit the destination area to see if all previous camps are ready to receive traditional residents and any more households. If they found insufficient resources to accommodate all herds and households, then a limited range of consultation might decide who moves and who does not, or how to arrange sites for the extra animals.

Movement patterns

Turgen belongs to a society where herding mobility is the only way to cope with the harsh climatic and ecological conditions. This mobility helps make better use of natural resources, and the success of herding depends on flexible, mobile herding.

Discussions with sample herders identified two main types of grazing movement: short-distance and transhumant. The study thereafter focused on the transhumant systems (known as kholyn nuudel [where nuudel is a general term for any kind of moving of things, including livestock], sometimes alsyn nuudel). According to local herders, alsyn nuudel is migration between autumn pasture and winter camping. The local definition of transhumant movements can be summarized as: moving livestock to distant seasonal pastures or camps to ensure better productivity and survival through mobilizing all material means and social linkages.

In an attempt to determine the reality of long transhumance, households without close kinship relations were selected. This aimed to clarify whether distant movements by individual households are made purposely and to check how mutual help mechanisms work in areas where a typical khot-ail [a group of a few families that work and camp together - with or without family links] structure does not exist (Erdenebaatar, 1996). Small-scale interviews to check if this definition was acceptable to our respondents also involved six non-sample families. An impressive set of information was provided (summarized in Table 3.11) confirming that herders in the study area have ideas on seasonal long migrations that differ from those elsewhere (such as Arkhangai and central areas (Erdenebaatar, 1995)).

Sample respondents put better productivity and survival of herds in first place because they are totally dependent on animal production in an extremely harsh and risky natural and social environment. Social linkages are not thought important, while access to markets and services worry them less. Herders do not expect services for marketing and other needs to improve in the near future. They believe that they can no longer expect such services because the government does not do what it ought on behalf of the public.

While distance was a basis for determining whether a movement is called alsyn nuudel, some herders specified another type of transhumance, namely summer movement. Most (over 90 percent) of herding families and the majority of livestock keepers migrate to escape the mosquitoes that are abundant in lowland areas from late June to mid-August in Turgen (Plate 18). To get away from early summer areas, herders start moving livestock and gers to the highlands at the start of the mosquito season, ca. 15-25 June. This is an upward movement. When mosquitoes have disappeared, herders migrate downwards, ca. 15-20 August. Many herders also make long moves between autumn-winter and winter-spring camping areas.

Movements by Rinchinlkumbe people differ in that herders move to autumn pastures, in the woodland at the foot of the mountains in late August-early September, and stay there until late October, whereas Turgen herders migrate from their winter pasture to spring pastures around mid-March or early April, stay six weeks, then move again over the mountains to reach summer areas at higher altitudes, where they remain until mid-August. Their next movements are towards winter camps or areas, in late October or early November, where herders select to stay over winter in bad years.

The specific patterns of seasonal migrations and long-distance movements are strict and depend on the location of seasonal pastures. There are very few opportunities for herders to act in any other way, and this holds generally for both Turgen and Rinchinlkumbe.

TABLE 3.11
Cross-checking of local definitions of transhumant migration between the sample and voluntary respondents, Rinchinlkumbe, August 1999.

Interview points

Male respondents (n=11)

Female respondents (n=8)

Average scores

Social linkages




Better productivity




Better survival




Residence area




Access to market




Access to services




NOTES: 2 = lowest score; 5 = highest score.

FIGURE 3.2. Major long-distance transhumant routes, Turgen sum, Uvs aimag. Map based on a sketch map drawn by Mr S. Orshikh, a 63-year old herder.

Plate 18. Turgen summer pastures. These are at higher altitude to avoid the insects of the lake basin. Winter camps are also in the hills, to avoid the cold air that collects in the lake basin.

In Rinchinlkumbe, movements are strongly seasonal. A long migration is made between summer grazing and winter camping areas. Herders leave summer areas in late August towards winter camps, and en route they stay for some months elsewhere in the mountains (western side of Rinchinlkumbe and eastern end of Khoridol Sardig range) till it gets colder and snows. Then they move closer to their winter camps, which are the final destination of one of two major long-distance migrations.

The sample households were followed on their pastoral movements, and sites of intermediate and final camps were identified and total duration and distance noted in both areas. Respondents in both sums were questioned about routes taken in the last five years. In 1999, the routes over which Turgen herders moved were carefully tracked as a case study, and mapped. Overall, the Turgen make a circular transhumance (Figure 3.2):

The topographic location of Turgen seasonal pastures and the seasonal division of grazing areas is shown in Figure 3.4.

FIGURE 3.3. Major transhumant routes in Rinchinlkumbe sum. Map based on a sketch map drawn by Mr D. Chinbat, local Environment Officer.

Due to common spring and summer areas, where herders stay from mid-March to early to mid-September for summer grazing and late October to early November for autumn grazing, herders in Rinchinlkumbe have two long movements (see Figure 3.3):

The Rinchinlkumbe pastures and their seasonal division are shown in Figure 3.5.

Turgen sum

Interviews with herders and the survey identified two major long-distance movements made by the sample households: via Khamar davaa (Khamar pass), and via Ulaan davaa (Red pass).

The first route is for moving from spring-early-spring areas to summer areas, and back to autumn areas, but a second route is followed to reach specific areas. The sample households formed two migration groups based on their transhumance routes. They follow similar routes to reach and return from the final areas and camps. The first group use the Khamar davaa to summer and spring areas, but take different routes to return from winter camps to spring areas and return to winter camp from autumn areas.

FIGURE 3.4. Topographic location of seasonal pastures and seasonal division of grazing areas, Turgen sum, Uvs aimag.

FIGURE 3.5. Topographic location of seasonal pastures and seasonal division of grazing areas, Rinchinlkumbe sum, Khuvsgul aimag

TABLE 3.12
Distance, duration and number of intermediate camps of long transhumance, Turgen sum.

Migration route

Total distance (km)

Duration (days(1))

No. of intermediate stops(2)

group I

group II

group I

group II

group I

group II

Route A







Route B







Route C







Route D







NOTES: (1) Includes first day of migration. (2) Excluding initial and final camps.

Differences in duration and number of intermediate stops between migration routes (cf. Table 3.12) depend on two factors: season and distance per day. In warm weather, animals walk over 40-50 km daily; in cold months, when animals are exhausted and some lambs and kids have been born, migratory groups walk only 25-30 km daily. The range was 15 to 35 km. In warm seasons, herds move rather quicker than in winter and spring. Herders attributed this to:

Rinchinlkumbe sum

To reach winter camps Rinchinlkumbe herders migrate over four major mountain passes of the Khordil Sardig range to enter areas west of Khuvsgul lake: (i) moving up Sharga gol [river] to winter grazing, and (ii) moving over Uliin davaa, (iii) moving up Arsain gol, and (iv) moving over Khoroin gol. Distances and duration are shown in Table 3.13.

The routes herders follow are largely dictated by the timing of long movements: routes W and X lead herders and livestock from the winter camps to spring areas, whereas routes Y and Z direct the migration from summer areas to winter camps. There is a difference in duration and number of intermediate stays between Turgen and Rinchinlkumbe because:

TABLE 3.13
Distance, duration and number of intermediate camps of long transhumance movements, Rinchinlkumbe sum.

Migration route

Total distance (km)

Duration (days(1))

No. of intermediate stays(2)

Route W




Route X




Route Y




Route Z




NOTES: (1) Includes first day of migration. (2) Excluding initial and final camps.

Herds need some rest on the way to their final destination because prolonged walking of well-fattened animals without rest leads to unproductive loss of body condition, which inevitably accelerates further decline in body condition. Spring is also when animals tire easily, so a stop to allow grazing and rest is essential when about to walk over long distances in the next days. Herders argue that one, two or even more days are needed between stages in the spring movements, so that herds graze more and newborn offspring get more milk. Such stays take place in the middle of the journey, unless there are other reasons for stopping or good areas are found in other parts of the route. A decision on which routes to take depends on the location of summer areas and winter camps. Those with summer areas in the southwest corners of the sum and winter camps in the south end near Khuvsgul lowlands take route Y both ways. Some exceptional cases were observed: in 1998: five herders who moved along route W to winter camps took route W to return to summer areas because of exceptionally late melt of snow drifts in Sharga river valley, which forced them to leave already exhausted winter pastures before the main lambing.

Exceptional 1999 summer and 2000 winter-spring movements in Turgen

It became obvious that both general and special cases had to be studied, instead of focusing all activities on a single year, so some details regarding the overall patterns of long-distance migrations in both sums are described. In 1999-2000, there were strong variations in the seasonal transhumant movements of herders of Turgen due to severe summer drought. This demonstrates that seasonal movements in Mongolia are complex and multipatterned. There was insufficient forage in all seasonal pastures in 1999-2000, and lack of haymaking accompanied by weather disasters and continued cold. The local meteorological stations recorded 28 days with temperatures of -35°C and colder at night, and -25°C during the day. Very deep snow, 20-30 cm on the plain and over 35-50 cm in the mountains, altered the movement patterns for a large number of herders. Many had to take a variety of manoeuvres to seek space for temporary accommodation, and to share available grazing resources. In the course of these movements, all habitual practices of long transhumance were modified to some extent, as the numbers of intermediate stays were increased and distances between two consequent camps also varied significantly.

There are very specific forms of transhumance in Mongolia and the variations mentioned above are not usual. Herders believe that transhumance, as an impor- tant component of pastoralism, is subject to natural spirits, as in other pastoral societies.

On a trip to Uvs aimag in early November 2000, the author met a widow and discussed how she survived the 1999-2000 winter and 2000 spring. The substance of that interview is summarized in Box 3.3. The decision-making process, errors and realities show that 1999-2000 movement patterns were completely abnormal for some sample herders.

BOX 3.3
An interview with a Turgen family head

Ms S. made exceptional movements in the first three quarters of 1999. On the basis of instinct and personal observations she (a family head) recognized that the winter camp area could not support her animals any longer, so decided to move to spring lowlands 20 days earlier than normal, which is mid-March. Before deciding, she consulted other herders but found no support. However, having confidence in her decision, she moved her herds and reached a spring shelter after five intermediate camps. She expected the spring area to provide forage for her animals. She was partially correct and animals grazed some standing dead grass. Some experienced herders play on the specific behaviour of animals, whereby changing the grazing area motivates animals to forage more actively than before. Taking advantage of this natural instinct, they use prolonged moves, spending two, three or more days at an intermediate site. When the rest of the herders reached the spring shelters, Ms S. had already grazed what her areas had to offer, so she looked for a place to keep animals until the appearance of young grass. There was some forage for other herders who arrived later than hers but she chose not to trespass and instead moved to an unknown area lying across the borders of Turgen and Tarialan sum, where she had never stayed before. Young grass did not appear, although long expected, but instead came a drought. Herders competed for access to better areas before others and Ms S. found no room. She did some short moves within the area where she was. When the time came to move up to summer areas she could select one of two routes: route B or take a new one. She chose the latter option, and moved her herds and ger over totally unfamiliar land, spending eight days before reaching the traditional summer area. Dry and stunted pasture met her, promising inadequate support for her animals. Following other herders, she moved back to the spring area. As of early October 1999, she had made over 23 moves, including short-term camp changing. She estimated that for the rest of the year she might need to make at least 10 more moves. Based on heavy summer drought, which led herders to adopt non-standard decisions, her strategy for the 1999-2000 winter-spring season looked to be:

- separate the herd into two parts: small animals (sheep and goats) and large animals (horses, cattle and yaks);

- send large animals to her winter camp in Baruun salaa valley; and

- set up a temporary camp in Abilga valley for a month while the other herders find room for them, and move to areas near the Russian border.

In eleven months, Ms S. and her family had moved over about 180 km, with 18 intermediate (overnight and camp changing) stays. There were some changes in terms of date of major herd movements:

Ms S. said that the survival rate of her herd was much higher than that of other herders due to circumstances and correct decision-making. Her strategy worked very well for:

In conclusion, overall success of pastoral mobility, including long-distance transhumance, largely depends on the strategy herders choose and implement. Some herders prefer to move a long distance, while others prefer better herd and grazing management.


Trees are not considered fodder, either in Turgen or in Rinchinlkumbe. Larix forest is an area where cattle and horses can be herded in warm months, with regular supervision. Herders identified Salix trees and Caragana bushes as useful browse for camels and goats in hard times. Trees provide protection from periodic high wind, rain and snow storms. They provide fuelwood at all seasons and shade on hot, sunny days. No trees have been cultivated for fodder, and nor did herders entertain such a proposal. A family uses about 6-8 m3 of fuelwood annually in Turgen and 12 m3/yr in Rinchinlkumbe, at a cost of tugrik 800-1000 (US$ 0.7-0.95)/m3.

Community and households

All herders and their family members were literate and had completed primary (fourth class) and secondary school (8 to 10th classes). Some children study at Universities and Colleges in Ulan Bator and Darkhan. Young people have a higher education level than the older generations thanks to Government policy on secondary education. No herders had training in matters related to livestock production and herd management. A young nursing graduate had returned to herding to help her mother; she provided health assistance to herders voluntarily. Four young men in both sums were trained professional drivers but were not driving. Twelve herders’ children of Turgen and eight of Rinchinlkumbe went to secondary school. Four (Turgen) and five (Rinchinlkumbe) children had left school for various reasons, mostly a parent’s decision to have more labour for herding.

TABLE 3.14
Annual labour profile (percentage basis).





















Collecting of fodder




Livestock herding




Fuelwood collecting




Work outside household




NOTE: (1) Two herders contributed some money and labour to a grain grower and helped with harvesting.

All households are wholly engaged in herding and all work patterns are determined by it and related activities, see Table 3.14. On average, over 80 percent of annual labour is devoted to herding; there are almost no crops. Three Turgen herders sowed vegetables in 1999, but did not tend them so the harvest was small. Two herders assisted in grain growing in 2000, and shared a small harvest - three sacks (120 kg) of unprocessed wheat.

Work outside the household

Labour division between men and women is very similar to other herding societies in Mongolia. The respondents shared a common perception that herding needs good cooperation between households. Major forms of cooperation in the study areas are similar to those elsewhere in the country: labour pooling and resource sharing. Men and boys are responsible for outdoor work, whereas women perform household activities. Herding in summer is done by children of pre-school and school age, while men and grown boys have to herd animals in winter and spring, because of extreme cold and snow.

A similar profile of outside household work was described by respondents(Table 3.15), as all herding households are entirely pastoralists. Herders provide labour (people come to help) and some items (tools, pack animals, pack harness). No monetary assistance is given as loans or charity. During the 1999-2000 natural disaster, herders of Turgen received some assistance from donors and NGOs, and from the state.

TABLE 3.15
Profile of work outside the household (percentage - the total comprises 3 percent of annual labour).

Major activities






Wool shearing



Goat cashmere combing



Felt making



Help in migration



Cutting and collection of fuelwood



Collecting mineral additives and salt






TABLE 3.16
Typical average animal production values.



Liveweight (autumn weight; kg)

Cattle adult bulls



Cattle adult cow



Sheep - ram



Sheep - ewe



Goat - billygoat



Goat - nannygoat



Milk production (litre)

Cow (per 160-day lactation)



Mare (annual estimate)



Ewe (annual estimate)



Goat (annual estimate)




Cashmere fibre (gram)



Sheep wool (kg)



Production estimates

Production estimates of sample households’ herds are all low, and little information on herd productivity was available (see Table 3.16). Herders never measure individual productivity of animals; sometimes information provided by local livestock inspectors was used to estimate production. Some indexes were measured by the survey team. The yield of mares may seem very low, but this is because, in the area studied, they are not milked much. Sheep in the study area produced 0.1-0.3 kg more wool and weighed 2.8-3.5 percent more than the Khalkh breed, due to a greater proportion of the Bayad and Darkhad mutton strains.

Sale of livestock and livestock products are the main source of household income (Table 3.17). The amount sold varies depending on herd size and household needs. Most income comes from sale of meat, animals (live and bartered for food and goods), skin, hides and cashmere. Income per family member was US$ 173.9. The figures in Table 3.17 do not include items given as gifts or assistance.

Fodder balance

Natural forage comprises over 90 percent of total animal feed. This was the case in the study area. Interviews revealed that they cut a little hay in 1998, but only 0.2 kg per animal fed. No crop residues or trees were used. Only a very little (240-400 kg) manufactured fodder was bought for exhausted animals and riding horses. Visits in 2000 showed that the impact of lessons herders had learned from the 1999-2000 natural disaster was uneven amongst sample households. Herders showed high interest in making more hay in early summer, but a visit to Turgen in mid-September showed some drop in enthusiasm. Nevertheless, the amount of hay prepared had increased. A household cut hay in the range of 350 to 900 kg, which was said to be 10-35 percent higher than that made in 1998 (almost no hay was mown in 1999). Much more natural mineral lick was collected, as herders thought it was abundant because of the dry summer. Wild hay comprised 90 percent and more of the feed balance, and provided opportunities for supplementary feeding to some livestock. No other materials were available to improve structure and nutrient value of animal feed.

TABLE 3.17
Annual sales of livestock produce (total annual income per household).


Average amount sold

Income (US$)





Milk (frozen)
























Meat (all categories)






Dairy products











304.2-1 209.6

Sheep skin






Goat skin






Cattle hide






Horse hide






Sheep intestines








Income per family member


NOTE: (1) Maximum and minimum amounts

The only supplementation was to some (on average 3-5 percent of total herd) animals in winter and spring; supplementary feeding is short, depending on needs. There are no planned feeding schemes. Year-round grazing was the only way of feeding animals. As a rule, small stock and cattle grazed during the day and returned to camp overnight, whereas horses and camels are left at pasture far from main camps and seasonal pastures for small livestock. Checking on animals at pasture varied by season. In summer they are checked weekly, and every 2-3 days in winter to offer minerals and see if all animals are well.

There are different ways of providing minerals (mostly natural soda licks). In spring and autumn, Turgen herds have free access to bulk licks. In late autumn, herders go to natural deposits near Uvs lake to collect soda and store it for winter and summer. In Rinchinlkumbe, herders from the southern part of the sum walk their animals to a source in Khurgany khooloi - small ponds full of salty water. Herders in the northern pastures have open access to swampy areas with minerals.

Rinchinlkumbe herders make more hay than Turgen people. In September 2000, five households were interviewed on this topic to find out what they do to make feed. Usually herders travel to their winter camping areas where there are fields for cutting wild hay. All respondents said that they went to make hay in mid-August, when the hayfield grass growth was at its peak, and stayed for 25-30 days. Their description shows that the haymaking areas are productive enough: yielding 1 240-2 000 kg/ha. Herders measure the amount of hay by the numbers of stacks they make: each weighs 50-100 kg. Total haymaking by sample herders ranged from 56 to 100 stacks.

BOX 3.4
Private haymaking and supplementary feeding strategy, Rinchinlkumbe sum
From an interview with Mr B., fourth bag, Rinchinlkumbe.

Three households agreed to stay together over winter. Three men and a boy went to make hay and stayed 26 days, including four travelling by horse. Total hay prepared was 4.5 tonne. These households own 350 livestock and expect to give supplementary feed to about 8-10 percent of them (30-35 head) if winter and spring are normal. Duration of winter-spring supplementation is about 50-60 days. Gross daily consumption of these animals is 50-65 kg of hay, and the hay made is enough for another 10-20 animals. He is happy and sure that his hay will cover all his needs. As a reserve for newborn stock, he cut another tonne of hay at his late spring camp in the lowlands where herders arrive in early April direct from winter camps.

Turgen sum has some good areas for hay which could provide enough for local herders, but, because they are close to the aimag centre, there is a high level of competition from trespassers (Plate 19). Hay is divided into two parts: that for winter feeding, and hay reserved for young stock. Herders of study sites have two distinct plots for haymaking: one in the lowlands and one in the mountains. Hay for newborn animals is made near spring camps, cutting small areas in the lowlands. Hay prepared from the mountain fields at the winter area is used in winter.

Community participation

To enhance community participation in the studies, a range of activities were undertaken:

Plate 19. Hayfields must be protected and respected. Camels from migrating camps destroying lucerne hay fields, Turgen, Mongolia.


· Two major types of long-distance transhumance movements have been identified:

- seasonal movement across five different ecological belts (in Turgen sum); and

- seasonal movement across two ecological zones (in Rinchinlkumbe sum).

· All movements are longitudinal (west-east) to make better use of the variety of forage resources and climatic differences. There are a clear set of climatic and social indicators governing seasonal transhumance movements.

· The study shows that transhumance movements are not of the same direction and frequency: Turgen herders move twice a year over the same passes between their summer and winter camps; those of Rinchinlkumbe make one movement to and from winter pastures. No herders stay at the higher altitudes in winter. Seasonal, especially winter, camps are on one side of a mountain range that routes must cross.

· Various routes were identified by tracking the migration of households and groups of households. Due to the limited options for seasonal pastures and camping areas, mostly involving traversing mountain passes, herders in both sums must follow the same routes to reach particular areas and to return. This is an extremely important pattern that distinguishes the complexity of the system and its integrity.

· The major herding strategy is a multispecies herd with upgrading for subsistence, with marketing of surplus produce. The average family size is 4.7, comprising people of all ages.

· Pasture condition varies in different areas of the sums, from seriously degraded to underused. Increase in the proportion of negative indicator plants indicates the degree of damage to natural pastures.

· Long-distance transhumance movements consist of a number of intermediate camps and the final destinations. Geotopographic conditions of localities along the routes determine how many intermediate camps are required. The number of intermediate camps also depends on herd ability to walk long distances each day: shorter and slower in spring and longer and faster in summer and late autumn.

· Moving from autumn camps to winter areas is one or two days shorter than the return to spring areas because newborn stock walk slowly and all animals are exhausted, having lost 10-15 percent or more of their autumn weight.

· Herders sell animal products, using various marketing options, to earn income for buying food and goods. In 1999, average sale income per household was US$ 817.5. Household consumption was 60-70 percent of the annual sale income.

· Natural grazing is the main source of animal feed for both areas. Haymaking is a common form of fodder supply. Sown forage and cropping are practically absent.


There are a range of factors contributing to the adoption of long-distance transhumance as a single strategy, but the primary need is for all involved to agree.

Local conditions in both sums differ markedly from other areas of Mongolia, where short distances and fewer seasonal and grazing movements are made. For instance:

Marked and strong seasonal changes in the weather dictate when to move, because early and hard frost transforms green forage grass into standing litter; herders always worry that early snow in the mountains may make movement over the passes difficult or even impossible: at a sign of early heavy snow they move 15-20 days earlier than normal.

Herders in some areas move in a very short time to another area. This is a needbased strategy that ensures the temporal and spatial dimensions of pastoral grazing. Herders call it “moving as frightened birds take off”, leaving behind them all the pastures totally free of grazing.

Herders do not believe that catastrophic damages to traditional movement systems was wrought by pastoral collectives (negdel) under the planned economy. They provide a lot of their own arguments to show that collectives did much to strengthen pastoral production in all aspects of its functioning. They noted some development statistics for the livestock sector: in 30 years of existence, collectives provided water to over 75 percent of natural pastures at national level; after 10 years of post-liberalization, the water supply to pastures had fallen by 40 percent.

Previous Page Top of Page Next Page