Previous Page Table of Contents Next Page

Chapter IV - Mongolia case study 2: Haymaking from natural pasture in Arkhangai, Mongolia - V.I. Lkhagvajaw and B. Erdenebaatar


Stock raising in Mongolia relies on natural pasture. The thermal growing season is around one hundred days. Conservation of natural herbage for winter and spring supplementation is highly desirable. Since the privatization of the livestock sector in the early 1990s, collective systems of haymaking have collapsed and small-scale haymaking has yet to replace them. Herders have little experience of haymaking. Trials and demonstrations carried out from 1996 to 1998 at Ikh Tamir in Arkhangai Aimag are described.

Spreading of animal manure and irrigating by forming ice-sheets with spring water over potential hay land were used as possible ways of increasing yields. Drought prevailed throughout the period, limiting the availability of water to form ice, as well as restricting growth of rainfed herbage. Nevertheless, all treatments had a positive effect on hay production. Mineral fertilizer had the greatest effect, but is unlikely to be economically interesting. Simple haymaking techniques, using animal-drawn equipment, were satisfactory, and the demonstration effect of the work has encouraged herders to increase haymaking on their own initiative.


The pastoral situation and the effect of decollectivization on Mongolian herding have been described in Chapter II. Centralized hay production was discontinued soon after the change of system, and nothing came in its place. This chapter describes studies carried out in Arkhangai on methods of haymaking suitable for herding families, the possibility of increasing hay yield by water (ice) spreading, and the use of dung.

Traditional livestock are all, of necessity, well adapted to the harsh climate; they can regain condition and build up fat reserves rapidly during the short growing season. The hump of the camel and the fat rump of local sheep breeds provide energy reserves to help tide them over winter and spring. Yaks, camels and cashmere goats develop winter down in their coats, which helps reduce heat loss. All can survive outdoors throughout the long, cold winter with little or no shelter nor supplementary feed. The young are generally born in spring and their dams benefit from the fresh grass. The livestock are generally small. The herders’ strategy is therefore to get the stock into as good condition as possible in autumn so that they survive on dried-off or frozen pasture until the grass regrows in late spring; non-breeding stock are sold or slaughtered in late autumn and frozen naturally.

While it is not possible in a growing season of a little over three months to conserve feed for the rest of the year, it is highly desirable that some reserves of conserved fodder be available to assist the survival of young, breeding and weak stock. Hay from natural pasture is probably the only source of conserved feed available to herders, and then only in areas where sufficient moisture is available. Historically, each herder was entitled to possess defined areas of common land for many years, where hay was cut. Following the nationwide privatization of the early 1990s, however, every hayfield became the focus of local disputes between individual herders and within herding groups, as well as with herders from neighbouring communities. Apart from this, the repeated cutting which had become common in the later decades of the planned economy led to a serious decline in the natural productivity of haymaking areas, and there is no sign as yet that herders have invested in improvements, or will do so. There are several reasons for this:

In 1997, at national level, only 667 000 tonne of hay was made, or about 4.8 kg/animal - one fifteenth of the official supplementary feed norm.

The High Altitude Research Station of the Mongolian Research Institute for Animal Husbandry has its headquarters at Ikh Tamir sum in Arkhangai aimag. It was the base for this work because it is in a suitable agro-ecological zone and has high calibre pasture specialists. According to geographic zoning, Ikh Tamir belongs to the high mountain area, where the size of natural hay fields is limited by relief and geomorphological conditions.

Because of its altitude, Ikh Tamir has a harsh, continental climate. The warmest months are June-August, with mean temperatures between 13.8°C and 16.6°C; the coldest season is December-February, with average temperatures between -12.9°C and -16.0°C. The absolute maximum is 34.5°C, recorded in August, and the lowest is -36.5°C, recorded in January. The average mean temperature measured at the soil surface is 4.1°C, with a maximum of 65.6°C in June and a minimum of -41.0° to -42.5°C in December-January. The annual precipitation at Ikh Tamir sum is 363 mm, with about 80 percent falling in May-August. The thaw takes place about 20 May, and the first frosts come in early September. There are an average of 100 frost-free days, but with a wide range of 70-131 days.

Trials extended over three growing seasons, studying the conversion of natural pasture into hay land to assess the impact of simple techniques, mainly based on local materials; on the productivity of previously constantly grazed areas; and to study the feasibility of making hay from natural herbage. Specific activities were:

Experimental treatments and techniques

The treatments were:

(i) Provision of moisture to the soil in late spring through forming an ice sheet (ice cover) over the fields during winter.

(ii) Spreading animal dung at 50 tonne/ha.

(iii) Ice irrigation + dung.

(iv) Mineral (N+P) fertilizer at 60 kg/ha N and 90 kg/ha P.

(v) Ice irrigation + mineral fertilizer.

Ice irrigation has been known in the area for a long time, and was used in the pre-collective period; it involves spreading water from streams or springs over hay land so that a sheet of ice is formed. When this thaws in spring it supplies moisture at a time when the vegetation has great need of it. The water from the thawing ice is distributed over the pasture by digging ditches. This can, of course, only be done where water supply and topography permit. The technique is seemingly also used in Russia and FAO (1959) quotes Larin (1953) as reporting on the “Liman” irrigation of fodder crops with ice.

The dung used was the dry excreta of cattle, horses and small ruminants, taken from night shelters; no bedding is used but there may be some soil admixture. This dung (which is also used as fuel) is hard and compacted by trampling and drying, and has to be broken up before spreading. The herders grow no crops so there is no competition with other agricultural enterprises for the dung, which was hand-spread over the plots. Spreading is done in spring and has to be carried out relatively quickly, before plant growth is well started.

With such treatments, especially ice irrigation, and broken terrain, formal experimentation was not possible. In any case, large differences are being sought at this exploratory stage. Large observation plots were therefore used. The areas of the various treatments were: dung - 2.76 ha; mineral fertilizer - 0.9 ha; and fertilizer + dung - 2.67 ha. In addition, investigations were continued on the plots treated in 1996: 1.5 ha of ice irrigation plus dung; 0.25 ha dung plus mineral fertilizer; and 0.38 ha mineral fertilizer. In addition, areas of pasture were fenced as exclosures to serve as controls.

Geobotanical description was used to determine botanical composition; bare ground was estimated visually as a percentage. Height of the dominant species was measured on thirty plants selected at random. The above-ground phytomass was measured by clipping 5 plots of 0.5 m2 at a height of 0.5-1.0 cm. To determine botanical composition and changes in the proportion of the different botanical groups, 0.25 m2 plots were cut and separated into grasses, legumes, sedges and other forbs.

The experimental sites

Eight plots at five sites in different areas of Ikh Tamir (Plates 20 and 21) were selected, taking into account botanical composition and distance to winter camps. All sites belonged to private herders and had been used for both grazing and haymaking. Some sites were where an ice cover could be formed in winter through spreading spring water. Three sites were on southfacing slopes, two in open valleys, two in the lower parts of forested valleys and one on an open plain (see Table 4.1). In general, the sites represented most of the grazing and haymaking areas of the high mountain zone, and reflected their previous use. Some areas of the sum have been damaged by poor management (uncontrolled cutting and grazing, water and wind erosion, and damage by vehicle tracks).

Plate 20. Haymaking study. Protected plots at Ikh Tamir, Arkhangai, Mongolia.

Plate 21. Haymaking study. Plots about to be sampled, Ikh Tamir, Arkhangai, Mongolia.

Site 1 is mountain steppe with sandy brown soil on a 3-5° slope. In years of adequate rainfall, an ice sheet forms from a spring at the top of the slope. It is a degraded Festuca lenensis-Stipa krylovii- forbs pasture, which has been grazed at all seasons. Ice irrigation began here in 1996.

Site 2 is riparian land in a north-facing narrow valley; its slope is 3-5° and the surface is even. Ice sheets may form if summer rainfall is adequate to feed the river. The soil is dark brown alluvium. The vegetation is a Carex-Leymus chinensis-forbs community. This site had been mown until 1993 and thereafter used as summer grazing. Considerable surface damage has been caused by burrowing rodents. A partial trial with ice irrigation began in 1996.

Site 3 is in a short, narrow mountain valley facing northwest; the slope is 10-15°. The ground water is near the surface. The plant communities and use are similar to those of Site 1.

Summary description of project sites.


Altitude (m)


Plant communities

Major species

Site 1 Bulagt


Riparian meadow area in open valley on southern slope of mountains

Herbaceous-Carex duriuscula-Leymus chinensis mountain steppe

Leymus chinensis, Stipa kyrlovii, Festuca lenensis, Koeleria cristata, Carex duriuscula, Artemisia laciniata, A. glauca, A. comutata, Plantago adpressa

Site 2 Burgast

~ 1700

River side in narrow, north-facing valley

Herbaceous-Leymus chinensis-Carex pediformis steppe-like meadow hayfield

Leymus chinensis, Koeleria cristata, Agropyron cristatum, Carex pediformis, Artemisia laciniata, A. mongholica, Potentilla tanacetifolia, P. anserina, Galium verum, Bromus inermis, Vicia amoena, Silene repens, Sanguisorba officinalis

Site 3 Khusluurt

~ 1740

Basin of northwest high valley

Herbaceous-Leymus chinensis-Carex pediformis meadow-steppe type pasture

Geranium simplex plus species as at Burgast

Site 4 Chono


Closed depression in southeast-facing valley

Meadow-type steppe Herbaceous-Carex pediformis-Poaceae grazing area

Agropyron cristatum, Poa subfastigata, Festuca sp., Carex pediformis, Artemisia laciniata, A. dracunculus, A. glauca, Thalictrum simplex, Galium verum, Potentilla tanacetifolia

Site 5 Mukhar

~ 1760

10-15° slope on northern face of mountain

Hay and grazing area in Herbaceous- Poaceae-Carex pediformis steppe meadow

Bromus inermis, Calamagrostis epiodis, Elymus turczaninovii, Festuca sp., Stipa baicalensis, Carex pediformis, Artemisia laciniata, Geranium pratensis, Galium boreale, Vicia amoena, Hedysarum alpinus, Thalictrum simplex, Silene repens, Anemone crinata, Scabiosa comosa

Site 4 is a rainfed mountain meadow on dark brown clay on an east-facing slope at the foot of a mountain forest. The plant community is forbs-Carex pediformis-Poaceae. In years of adequate growth it had been used for hay; otherwise it was grazed. Work began in 1997.

Site 5 is on the north-facing slope of a valley with mountain meadows on dark brown clay. The slope is 10-15°. The plant community is forbs-Carex pediformis-Poaceae. The site had been used for haymaking and for winter-spring grazing. Work began in 1997.


During the period of the work, the weather was far from ideal for haymaking; no rain fell from April to late July. Heavy dust storms in spring and hot, sunny days in summer led to serious damage to the vegetation. Runoff from heavy downpours in late July did not provide enough moisture for the plants to recover, and drought continued into August. At a meteorological station 27 km from the sites, total rainfall in 1996 was 94.7 mm and the annual average temperature was 1°C above average. Snow cover during the 1996-1998 winters was also low and the lowlands were “brown” - snow free. The first rain in the springs of 1997 and 1998 fell in mid-May.

Effect of ice irrigation

Two sites at head-points of small springs had winter ice cover, beginning around 20 November and thawing by 10-20 April. The average ice depth varied from over 18 cm in the centre to 3-7 cm at the periphery. Two key moments are noted when more moisture is absorbed by the soil: in the initial days as the soil profile becomes filled before all is frozen over; and at the thaw. Soil covered by ice throughout winter contains more moisture than bare ground. This is probably important for early growth so ice-irrigated plots showed better performance. Although ice cover was insufficient to provide a reserve of moisture for after the thaw, it assured a better intensity of grass growth for the rest of the growing period. Ice irrigation, therefore, gives far better early growth than that which occurs in open areas. The moisture percentage of soils under the different irrigation regimes during the growing season are shown in Figure 4.1. In early spring, the ice-irrigated land contained over a third more moisture than the control, and the difference persisted through the season.

FIGURE 4.1. Effect of ice irrigation on soil moisture during the growing season.

Effects on total phytomass

Since the nutrient treatments were applied to different plant communities at different sites, the aggregated results are shown (Table 4.2) by site. Annual variations, largely influenced by precipitation, are illustrated in Figure 4.2.

Application of dung and ice irrigation (although irrigation was inadequate due to drought) had a positive effect on yield; mineral fertilizer, as expected, gave a large, positive, response, but is unlikely to be economically attractive to herders.

Despite severe drought in the three consecutive years of the trial, a number of changes in the growth of the major plants and their abundance and role in the formation of plant communities were observed. These changes were attributed to the influence of the type of manure, as well as the basic characteristics of the plant communities involved, the microclimate and other ecoclimatic conditions of the mountain zone and the sites.

Dung gave a reasonable acceleration of plant growth, despite the drought, and brought about some changes in the development and appearance of young plants through the better distribution of stolons and rhizomes of valuable short and creeping tussock grasses (Leymus chinensis, Bromus inermis, Festuca rubra, Hordeum brevisububatum and Poa subfastigata) and some forbs. No loss of species in the plant communities was noted. An annual species, Chenopodium album, increased greatly in the first two years but was greatly reduced thereafter as it could not compete with well-established perennials. C. album was easily suppressed by more valuable species on the plots where mineral fertilizer was used.

With mineral fertilizer and the combination of fertilizer and dung, short tussock and creeping plants (Festuca lenensis, Koeleria cristata, Agropyron cristatum, Plantago sp., P. adspera, Leontopodium ochroleucum, Heteropappus hispidus and Potentilla anserina) decreased, as they were dominated by taller species. In ice-irrigated plots, creeping herbs increased; this was especially obvious in 1996 and 1997. Grasses that increased under mineral and mixed fertilization were Leymus chinensis, Bromus inermis and Poa subfastigata. Forbs that increased were Artemisia mongholica, A. vulgaris, A. laciniata and Silene repens.

FIGURE 4.2. Annual variations in yields of treatments (kg/ha air-dry material).

During 1996, shoot length of Leymus chinensis increased greatly in comparison to the control: by 37.4 percent under ice irrigation, 64.8 percent under ice irrigation + dung; and 141.8 percent with ice irrigation + dung + fertilizer. The same was noted in 1997/98 but real values were lower because of drought and insufficient ice cover. The results are shown in Tables 4.2 and 4.3.

In order to clarify the real effects of treatments on the vegetation, detailed observations were made at two sites, Burgast and Khusluurt.

Burgast (Site 2)

This was a forbs-Leymus chinensis-Carex hayfield. Major changes in the proportion of different plant groups in relation to treatments are shown in Table 4.4.

Yields of hay trials by site (kg/ha air-dry herbage).

Site and Type


Ice irrigation

Ice irrig. + Dung

Ice irrig. + Dung + NP

Ice irrig. + NP

Tsagaankhad, mountain steppe




1 170


Mukhar, mountain steppe meadow

1 450

2 060

2 060

3 570

4 340

Bulagt, mountain steppe


1 000


2 160


Khusluurt, mountain meadow

2 010

2 500

2 680

4 560

2 710

Burgast, steppe meadow

1 360

1 070

2 530

3 340

2 150


1 196

1 394

1 923

2 960

2 480

Assessment of the effect of treatments on the bioproductivity of Carex-Leymus chinensis-forbs pasture community at Site 2.

Absolute height of leaves (cm)

Plants per m2

Average fresh weight of single plant (gram)





Ice Irrigation




Ice irrig. + dung




Ice irrig. + dung + NP




Ice irrig. + NP




Change of plant groups in Carex-Leymus-forbs mountain-steppe hayfield (Burgast - Site 2) (average for 1996-1998; percentage).






of which

of which





Artemisia mongholica











Ice irrigation









Ice irrig. + Dung









Ice irrig. + Dung + NP









Ice irrig. + NP









Change in plant groups in a Carex-Leymus-forbs mountain-steppe hayfield, average for 1996-1998 at Site 3 (Khusluurt) (proportion of plant groups in whole community; percent).






of which

of which





Artemisia mongholica




















Dung + NP


















Khusluurt (Site 3)

This was a Leymus chinensis-Artemisia-Carex steppe meadow pasture. Here the proportions of different plant groups changed in a way different from Site 2, sometimes in an undesirable manner. The proportion of creeping grasses and forbs remained the same as the control under dung and dung + mineral fertilizer, while pure mineral fertilizer gave an enormous increase in forbs, including Artemisia mongholica. This increase in forbs caused a proportional decrease in grasses in the total yield (see Table 4.5). In contrast to the decrease of Carex in plots receiving mineral fertilizer, dung fields had a great increase of sedges. Shoot length of Leymus chinensis and Bromus inermis increased greatly in comparison with the control (see Table 4.6).

The strong increase in forbs, especially Artemisia mongholica, reflects their great ability to react to additional, easily available nutrients, and clearly limits growth of other plant forms by dominating in both above-ground growth and roots. Bright sunshine and continued moisture stress in June and early July badly affected the growth of Leymus chinensis, the major dominant plant of this community. This was very evident in 1997, when the summer was exceptionally hot.

Assessment of effect of treatments on the bioproductivity of individual plants at Site 3 (Khusluurt).


Leymus chinensis

Bromus inermis

Length of vegetative shoots (cm)

Plants per m2

Fresh weight of individual plants (gram)

Length of vegetative shoots (cm)

Plants per m2

Fresh weight of individual plants (gram)















Dung + NP







Ice Irrigation + NP







Cart design

Transporting and spreading dung is one of the hardest tasks. Dung is not always available close to suitable hay sites, and has to be transported from the shelters. For this, simple mechanization is necessary. The compacted dung has to be broken up to allow even spreading. Wooden ox-carts of traditional design, with wooden wheels, are widely used in the area; wood is relatively plentiful in the mountain zone, mainly from larch (Larix dahurica) forests, but the carts are easily damaged when hard dung is broken in them. An improved version was built and tested wherein the parts where dung is crushed for spreading were made of steel. Initial tests showed that metal is much more resistant and more suitable for heavy work. Two carts were built and used in 1998. Their use greatly simplified work and reduced by 47 minutes the time required to spread dung on one hectare, although much slower than spreading from lorries. Ox-carts have the great advantage that no fuel is required and the herders are well used to handling them. Spreading from carts was estimated to be cheaper by tugrik 690/tonne of dung (¡ÖUS$ 0.8 at 1998 prices) compared with spreading using a motor lorry.

Three training sessions were held in 1997 for local administrative staff, researchers and herders. Field training sessions were followed by a broad discussion of what had been seen and observed; this took most of one day. Based on the major findings and preliminary conclusions of the two years of field work, a small booklet was printed, describing the technical aspects of hay improvement using locally available materials - Practical recommendations on hay and pasture in Khangai High Mountain Area - and distributed to those who had participated in training, as well as to other interested herders.


The yield of the control plots was generally very low and would have led to difficulties in handling so thin a hay crop. Even under drought conditions, ice irrigation along with dung can raise hay yields, and the costs of so doing are low. No doubt in “normal” years the effect would be greater. The technique of developing ice-sheets and then spreading by ditches the water from their melting in spring has proved to be relatively simple and costs little other than labour. Mineral fertilizer, of course, gives large responses, as is usual on grassland, but price and non-availability are likely to rule out its use by subsistence herders.

There are indications that the treatments, along with closing sites to grazing to allow haymaking, does lead to changes in the botanical composition of the sward. Longer study would be necessary to determine the full effects of treatments on plant communities.

The sites have had a valuable demonstration effect, familiarizing herders with simple animal-drawn equipment and manual techniques. Some have taken to producing hay on their own initiative; this may spread if better weather improves the availability of herbage suitable for mowing.

Previous Page Top of Page Next Page