ICARDA is helping small-scale farmers in CWANA to cope with the challenges of their traditional production systems. These include a significant decline in rangeland productivity in Central Asia, which has been exacerbated by droughts, and reduced market opportunities for traditional products. At the same time, opportunities are offered by the development of potential new markets for non-traditional products. To improve productivity, research has focused upon both plausible marketing options and the use of alternative fodder sources to cope with feed shortages in the rangelands. In 2002, studies in Uzbekistan and Turkmenistan demonstrated the potential for milk production in sheep breeds traditionally used for wool and pelts. Farmers found that cheese production was a valuable source of additional income. In addition, to address dry-season feed constraints in Turkmenistan, highly saline irrigation drainage water, otherwise dumped in the desert, was used innovatively to irrigate cultivated salt-tolerant plants. Even in a drought year, resulting forage yields were up to 60 times greater than those from natural rangelands.

Diversifying into sheep milk production in the steppes of Uzbekistan and Turkmenistan

The dissolution of the large Soviet market, after the breakdown of the Soviet Union, plunged Central Asian sheep production into crisis. Previously, production was centered on wool and pelts, meat was a by-product and milk production was not even considered an option. Now there is little local and regional demand for wool and pelts, prices are low, and the scale of production has shifted; huge landholdings and flocks have been replaced by thousands of small privatized units, each having a handful of animals. Returns per head and per flock are low. Thus, there is a need for increased diversification and the reorientation of production systems, to match current production to demand and to the scale of operations.
     ICARDA and Central Asian scientists, with financial support from the International Fund for Agricultural Development (through the Integrated Feed and Livestock Production Project) investigated production opportunities and market niches that could be exploited by small-scale producers. Some systems, such as those based on indigenous sheep (e.g. the Karakul sheep of Turkmenistan and Uzbekistan), require diversification, in view of the difficulties involved in marketing pelts and wool. Simultaneous studies on consumption of, and demand for, milk products suggested that production of milk derivatives could be an alternative. Therefore, to obtain much-needed baseline data, scientists studied milk production in native sheep breeds.
     The first study considered Karakul sheep in Uzbekistan, a breed native to the steppes of Central Asia. New-born Karakul lambs are slaughtered for their high-quality pelts. After they lamb, the so-called 'Mary' ewes are not milked until the next lambing season, as milk production is not usually the aim of the system.
     During spring, in 2000 and 2001, the milk production of 50 Karakul Mary ewes of two color genotypes (gray and black) was investigated in collaboration with the Karakul Sheep Breeding and Desert Ecology Research Institute in Nurata, Uzbekistan. Milk yields were monitored every 10 days for 70 days during the period March to May. This is the characteristic suckling period, before the start of the summer. Total milk yield was 54.5 kg per ewe (Table 22); production peaked in the third 10-day period and then decreased. By the fifth 10-day period, some 15% of the total milk produced overall was still to be collected; in the final 10-day period, at the end of lactation, only 5.5% of the total milk yield was produced. As in other breeds, milk was richest in the first 10 days (Table 22). Fat and total solid contents then decreased (being lowest at the point of maximum milk production in the third period), and then increased as milk production dropped.

     Both age and genotype affected milk production. Total milk production, and its fat and solid contents, decreased with age: 3- to 4-year-old ewes produced the largest yield (65 kg) while older animals produced 20-30% less. The gray and the black Karakul sheep demonstrate many genetic differences. Gray lambs are usually heavier than black, and their pelage (fur) is usually over-developed; but, they are less healthy. Black ewes trailed gray ewes in total milk yield (50.65 vs. 57.67 kg). However, the milk of the black ewes was richer in fat and total solids. Gray ewes were depleted more rapidly during lactation than black ewes

(whose milk contained high levels of fat and solids until the end of lactation).
Though low compared with specialized milk-sheep breeds, the production of 54.5 kg in 70 days, under rangeland conditions-that are sometimes marginal-is encouraging. Thus, there exist possibilities for increasing income from the present pelt-production system and, most promisingly, for increasing income-generating activities available to women.

strategy—as early as 60 days after lambing—in addition to some level of feeding with hay and concentrates (250-300 g per head), could stimulate milk production to the extent that some milk could be recovered. Under these conditions, a total of over 20 kg of milk per ewe would be possible. This could provide income in addition to the production of lambs.
     Finally, the stories of two innovative Sarajin sheep farmers in Turkmenistan illustrate just how strong the desire of farmers to diversify production can be. The two farmers (Sapargulyev and Imamgulyev) collaborate with ICARDA and Turkmen scientists in an on-farm research network in Turkmenistan. Sarajin is a breed native to Turkmenistan, its meat is in high demand and its wool is used in the renowned Turkmen carpets. The breed has not previously been used for milk production.
     In mid-2000, the farmers attended a traveling workshop in West Asia that exposed them to production strategies, including milk production and processing, which are already being practiced by their counterparts in Syria and Turkey. The farmers did not wait to formally assess their flocks' milk production, but, using the knowledge they had gained, immediately milked their animals and turned that milk into cheese. In 2001, farmer Sapargulyev milked 40 of his lactating ewes from January to March, obtaining 1,148 kg of milk (28.7 kg per ewe). With techniques he had learned, he processed the milk into 230 kg of cheese that he sold for around US$380. In 2002 he milked more ewes, for a longer period (January to April), and produced 386 kg of cheese that sold for US$675. Also in 2002, farmer Imamgulyev decided to milk 30 of his ewes from January to April. This yielded 654 kg of milk (21.8 kg per ewe); the resulting cheese sold for US$150.
     Although Karakul and Sarajin are not specialized dairy breeds, the milk obtained under targeted management strategies offers possibilities for diversification and income generation in times when frustrated farmers face difficulties in marketing their pelts and wool. In 2002, ICARDA conducted a successful workshop on milk processing technologies designed for women farmers in Uzbekistan. It is expected that these efforts will help farmers to sustain their livelihoods during the critical period of their transition to a market-oriented economy.

New opportunities to enhance the feed base: using waste drainage water in the deserts of Turkmenistan

The breakdown of the former Soviet Union has had a major impact on both the feed production chain in Central Asia and the provision of fodder during critical periods of feed shortage. Previously large landholdings were broken up, and now the vast majority of producers farm on a small scale, with small numbers of animals, while others are landless. They have neither the resources nor sufficiently large flocks to justify seasonal grazing, nor can they produce fodder for the months of scarcity in drought and winter periods. This results in low productivity levels and serious degradation around villages. The crisis is severe in countries with harsh winters, where animals have to remain housed for several months due to extreme cold and snow. But the situation is also critical in countries (such as Turkmenistan and Uzbekistan) with less severe winters and extended periods of drought and hot weather.

     By September, the most productive species (Climacoptera lanata and annual Atriplex) were able to produce 7.8 and 8.8 tonnes of dry matter (DM)/ha, respectively. When compared with the small amount of biomass that is naturally available on the range at this time (less than 0.15 t/ha, Fig. 12), it is not surprising that local farmers are very interested in the results.
The trial was repeated in 2002. Seeds collected in 2001 were cleaned, dried and sown again in January and February, at a rate of 12-14 kg/ha in the case of C. lanata, 8-10 kg/ha in annual Atriplex, 6 kg/ha in perennial Atriplex and 8 kg/ha in S. altissima. The total area in this second year amounted to 3.75 ha, because part of the plot occupied in 2001 by S. altissima and A. heterosperma was left unplowed to study reseeding potentials.