SECTION 1 - INTRODUCTION:
THE PROJECT, PILOT AREAS AND THE CONCEPT AND STATUS OF PASTORAL RISK MANAGEMENT IN QINGHAI PROVINCE

Chapter 1 - INTRODUCTION AND BACKGROUND

1.1 Introduction

Qinghai province is a typical upland pastoral region of Northwest China. The total area of the province is 720,000 km²,accounting for 7.5 percent of the territory of China. The high plateau of Qinghai is covered by mountains with about 80 percent above 3,000 meters asl. Farmland suitable for agriculture in eastern Qinghai makes up only 1.5% of the area of the province. Alpine pasture is the major land resource for the extensive pastoral livestock industry in Huangnan, Guoluo, Yushu, Hainan and Haibei Prefectures. The total pasture area in Qinghai is 36.45 million ha, making up 50.5% of the total area of the province, of which 31.61 million ha is available for grazing.

According to 2004 statistics, the population of Qinghai is 5.38 million, with 55% of Han nationality, and 46% of 33 minority nationalities. The Tibetan population in Qinghai is 1.09 million, accounting for 21% of the total provincial population. The rural population is 3.89 million people, of whom 85 percent are in the cropping area and 15 percent in the pastoral area. Total agricultural production value (APV) is 5,692 million Yuan, of which 3,061 million Yuan (54 percent of total APV) is from the livestock sector.

Qinghai is one of the five largest pastoral regions in China with extensive rangeland resources and a long history of pastoral livelihoods. Livestock production plays a central role in the economy of the province, especially for minority ethnic groups. However, harsh natural conditions and environment limit rangeland productivity. The alpine pastoral ecosystem is now in a fragile condition. Poverty and environmental deterioration are pronounced, and natural disasters (such as severe snow disaster and spring drought) are frequent. They severely jeopardise the livelihoods and production activities of local herders, and at same time undermine the sustainable development of the local economy and threaten the stability of the society. This combination of high exposure to natural disasters, poverty of local herders, steadily increasing pressure from livestock diseases and decline of pastoral productivity, caused by and resulting in poverty and a deteriorating environment, in turn inhibit the herders’ and government’s capacity to prevent risk, and manage and protect the environment.

Family-based pastoral livestock production units in the province are further constrained, and their capacity for disaster prevention is undermined, by several factors:

• A lack of forage production and storage for the cold seasons, i.e. winter and early spring.

• Poor infrastructure: villages and herders cannot afford to invest in infrastructure to improve production or to build up a risk prevention capability.

• Inadequate social services. The technical service capacity of governmental line agencies for livestock and pastoral management is limited, so that the service can not meet the demand of herders for preparedness for, and resilience to, risk. An effective disaster early warning system is not yet established to predict disaster. Lack of animal insurance and credit for small producers makes herder households vulnerable to natural disasters.

• Poor educational level of herders in the high plateau is a constraint to preventing losses during disasters and achieving self-reliance afterwards.

• Poverty and lack of opportunities for economic diversification and income-generating activities in the high plateau are economic constraints on pastoral risk management. The vicious cycle built between pastoral poverty and frequent natural disaster seriously affects the livelihoods of herders.

• Degradation of vegetation and changes in the ecosystem result in a more fragile environment for livestock production, worsen the situation and creates another vicious cycle. Variable and sometimes insufficient pasture and inadequate management capacities of herder households, hamper the development of livestock production into a more intensive system.

• Management and technical services at provincial, prefecture and county level, especially in townships, are at a low level. The departments concerned lack risk management ability. This leads to a situation where herders usually do not maintain close links with these departments and technical institutions, which further reduces the effectiveness of efforts made by the government in poverty alleviation, risk resistance and disaster mitigation.

As a result of these factors vulnerability steadily worsens, and drastically accelerates due to sudden natural calamities. In recent years, natural disasters have occurred more frequently. Since the 1950s, there have been 14 major snow disasters. Ten were in the 40 years between the 1950s and 1980s, but four occurred in the 1990s alone. This is almost a doubling in frequency during the last decade. Due to these disasters, 22 million head of livestock were lost, and direct economic losses are estimated at RMB 1.1 billion (USD 137.5 million). It is the highest political and economic priority for the North-western Provinces of China to reduce these economic losses.

In order to break the vicious cycle of disaster-poverty-resource degradation, the Chinese Government at various levels has given highest priority to risk management and disaster prevention or reduction as key elements in the national and as well as the Western-Region development strategies or programmes.

However, the concept of pastoral risk management is new in China and the Government lacks experience of how to design and implement comprehensive risk prevention and management plans in pastoral communities. Therefore, the Chinese Government sought FAO’s technical assistance through the TCP programme (project number FAO/TCP/CPR/2902 - ‘Pastoral Risk Management in Qinghai province’) to strengthen the capability for risk management by the relevant departments and institutions, and to strengthen the ability of herders to resist risk in the project area. The product of this work - a package of strategies and techniques for pastoral risk management - will be extended to other areas of China with similar natural conditions and livestock production practices. The risk management strategy and action plan tested by project TC/2902 in the project area can be adopted by other Northwest China Provinces with similar ecosystems.

1.2 The TCP Project

(1) The project objective

The long-term development objective of the TCP project CPR/2902 was to reduce the regular animal losses of pastoral herders in Qinghai due to recurrent natural disasters, and to build up their own capacities to prevent damage from natural calamities. The project should also contribute to the overall improvement of herders’ future livelihoods, reduced environmental degradation and an increased capacity for risk prevention and management at local and provincial levels.

Immediate objectives were:

• To develop a comprehensive pastoral risk management strategy, including annual risk management contingency plans for two counties in Qinghai Province as pilot areas for further replication;

• To develop and field test in selected villages, together with herders, innovative risk management techniques and improved livestock production options to improve rational utilization of the family-based ranches so as to reduce losses when disasters occur.

(2) Project outputs

Specific project outputs were:

1.3 The Project Pilot Areas

The project counties of Henan and Zeku in Huangnan prefecture are contiguous and lie between 100⁰ 34’ and 102⁰ 15’ N and 33⁰ 04’ and 35⁰ 32’ E. The pilot villages in both counties are of easy access from all-weather roads and less than one hour’s drive in each case from the county towns. The climate is extreme and markedly continental as would be expected at such an altitude in the centre of a continent. Winters are long and harsh and the growing season short. There is no frost-free season: the dates of the first and last killing frosts are variable and frost can occur any day of the year, and ground frosts can occur at any time in summer. The growing season is from May to September. The annual precipitation is relatively high comparing with the more northerly and westerly parts of Qinghai: about 600 millimetres annually in Henan and around 500 mm in Zeku, mainly falling in summer and winter. The average annual temperature in Henan is -1.3 ⁰C - 1.6 ⁰C and that of Zeku -3 ⁰C - 2.8 ⁰C.

The grasslands of the project area are at an altitude of 3,500 - 3,700 metres on gently undulating plains with occasional low hills, typical of much of the southern grasslands of the Qinghai plateau. The vegetation is mainly low-growing with Kobresia and other Cyperaceae. Elymus nutans and Poa sp. are important grasses and many broad-leaved plants of which Ranunculus spp. are very evident. Although pastures are hard grazed, ground cover is generally good to very good. The entire area is above the tree-line.

Both counties entirely depend on a pastoral economy but differ in degree of development and relative wealth. Demonstration villages were selected in each county. Nanqi village in Henan is relatively prosperous, permanent dwellings are in brick and have glazed winter shelters. Jilong in Zeku has winter quarters built in rammed earth and mud-walled pens for winter livestock but no glazed shelters.

By using the tools and methods of participatory rural appraisal-PRA, a socio economic baseline study was carried out in collaboration with the county officials working in grassland management and livestock production in 2004, which lead to the following overall picture.

The herders were formerly fully transhumant and tent-based but now have permanent winter quarters and only live in tents on the summer pastures. Herders in Zeku are Tibetan whereas those of Henan are of distant Mongolian origin, although they are now fully assimilated into Tibetan culture. The level of literacy of older herders is low but, now that they have a fixed base, children receive education. Dairy products and meat are important in herders’ diets and cereals are also widely consumed. Animal products are bartered for household necessities as well as being marketed.

The livestock are mainly yak and sheep of local breeds. Fine-wool sheep were introduced in the collective period but are no longer raised; Oula sheep have been used in breed improvement. Yak are essential to subsistence for dairy production, as a source of down and hair, for meat as well as transport; they are also marketed. Goats and ponies are kept in relatively small numbers. The motorcycle is replacing the pony for personal transport on the grasslands. Stock, especially yak, are very small and it is said that the weight of yak has decreased by about 40 percent in the past thirty years and sheep somewhat less; the reasons for the decrease are not defined and may be due to a combination of nutrition, management and genetics.

The average pasture area for each person in Jilong village is about 7 hectares. With some land in the summer pastures seriously damaged by rodents and black beach soils, herdsmen have insufficient pasture for present stock numbers, which leads them to rent pasture from herders who have moved to nearby Zequ town. The average pasture area for each person in Nanqi village is about 14 hectares with a higher quality than that of Jilong village. The investigations also showed that almost all households in the pilot villages plant some fodder in their sheep pens for supplementary feeding in winter. Average area planted per household is 0.29 ha for Nanqi and 0.33 ha for Jilong.

Before the announcement of the project, no herdsmen’s association had been set up in Jilong, and there was no joint purchase of livestock or fodder. However there was some joint renting of winter pasture. With the approval of the project a herders’ association has been set up, supported by the local authorities in Nanqi village, and 23 households have joined. The association concentrates on the reproduction of Oula sheep, selecting breeding males, fattening lambs, and fattening sheep and yak for sale.

No important livestock epidemic disease have occurred in the pilot areas during recent years.

Following the nation-wide rural reform (the family responsibility system), livestock were distributed to families in the early 1980s and pasture allocation came later. According to the national land tenure policy, the Qinghai government declared in 1993 that the pasture allocation policy would be a long-term part of the livestock household responsibility system. Regulations of the same year required that pasture be allocated to households for at least fifty years. This was a final allocation: when household size or animal numbers change, the allocated pasture stays the same.

In the project area grassland has been allocated under the family responsibility system into family lots for winter-spring pastures and to groups for summer pastures; this limits the scope for mobility and radical adjustments to grazing systems. The allocation of winter-spring pastures is complete in both counties: allocation of summer pastures is complete in Henan and being finalised in Zeku. Winter pasture blocks are individually fenced with link-mesh; the fencing is not stock-proof but is important to herders since it symbolises their permanent grazing rights. According to the Grassland Station the recommended fence is currently zinc plated steel wire. Herders usually use seven wires in their fence. The unit price for fencing in 2004 was 7 RMB/metre.

Herbivorous rodents occur in great numbers on pastures. They consume large quantities of herbage and cause further damage by eating roots and disturbing the surface layer by burrowing. The main species are the plateau pika Ochotona curzoniae and the zokor Myospalax baileyi. Marmots, Marmota himalayana, are also present. Pika are by far the most numerous; they inhabit open land with short vegetation, live in family groups and can be a cause of soil erosion and pasture degradation (locally called "black beach"). Rodents frequently reach plague proportions and can have a serious effect on winter risk so monitoring is necessary. They are sometimes controlled by poison baits but controlling one species may make way for expansion of another. Pikas are said to be animals of very short grassland but it is not obvious how the close-grazed steppe could be made to grow taller to reduce pika infestation while maintaining livestock production and family incomes. Pikas are the food of foxes and many raptors, however, due to the radical change of the climate and degradation the pasture vegetation, the raptors have almost disappeared in the seriously deteriorated pastoral areas.

More specific community and household basic data on household situation, grassland condition, livestock, fodder, purchased feed, income, household consumption and occurrence of disasters are available in separate consultant reports

1.4 Grazing resource management

Natural pastures The grasslands of Henan are mainly alpine meadow and mountain meadow, those of Zeku are mainly alpine meadow. These pastures are dominated by Cyperaceae with Kobresia spp. very important and Kobresia humilis especially so.

Water. The project area is well supplied with water so this should not be a constraint in livestock production nor a serious component of winter risk. There are numerous streams and watercourses throughout the area. The villages visited could reach ground water with shallow wells and extract it with either hand pumps or submersible electrical ones.

(1) Grazing systems

The grazing management system in the project counties is simple. There are usually only two seasonal units of grazing land:

- autumn-winter-spring pastures are combined and herders’ permanent quarters are built there; these pastures are all allocated to individual families and fenced; the livestock stay on them for eight to nine months of the year. Some herders, in some areas, do have separate autumn pastures; per capita (early 1980s allocation) pasture in Henan is less than 100 mu (around 7 ha).

- summer pastures which are allocated, or are about to be allocated to family groups and where herders camp and graze their stock for three to four months; summer pastures in the project counties are not noticeably higher than the winter ones and transhumance is more lateral than vertical; the distance between summer and winter pastures is relatively short - ten to twenty kilometres.

Allocation of winter pastures on 50 year leases gives herding families security of tenure so they can invest in housing and infrastructure; this is a vast advantage over several other herding countries which have de-collectivised.^[1] Fencing is expensive and much of it does not appear to be very solid or stock-proof, but it is important symbolically since it demarcates a family’s pasture and indicates their grazing rights. Grazing areas were allocated according to family size at the time of de-collectivisation of pasture; this has caused some anomalies and is hard on those who were young bachelors at that time and now have families.

Government policy encourages the protection and reserving of winter grazing but, with fencing, there has been a tendency to let stock graze at will within the three-season enclosure rather than to herd them to specific areas at specific seasons. Better spatial and seasonal distribution of grazing is desirable for efficient pasture use but the means of encouraging herders to tend and direct their stock daily has not been found.

There is a market in grazing and some herders rent pasture, often from former herders who have settled in town; one herder interviewed had rented 300 - 400 mu (20-30 ha) for two months for RMB 1,000.

(2) Grassland management and its problems in pilot villages

In general the grasslands in the project area seem hard grazed but basically in good condition with good ground cover. There is a common degradation problem on the Tibet-Qinghai Plateau known as "black beach" where the natural vegetation has been destroyed leaving a black surface without plants. In the baseline survey 20 families had no black beach and 11 had a total of 3,220 mu. In Jilong village of Zeku county 170 households had 52,207 mu of winter-spring pasture and 40,262 mu of summer pasture; 27,740 mu, some 30 percent of the total, was "black beach" but this is much higher than average.

(3) Grassland monitoring

A grassland monitoring system has been in place for many years and is well established and regularly executed. Monitoring of primary production is by the traditional method of periodic clipping of quadrates, recording fresh and air-dry weight and dividing the harvested material into palatable and unpalatable species. This method gives precise measurements of yield at the sampling sites but it is labour and time-consuming so only a limited number of quadrates can be handled and geographic coverage is sparse.

It is highly desirable that much more rapid techniques of herbage measurement be found to supplement the clipping system, for two purposes: to get better geographic cover so as to have a picture of the overall grassland situation, and to provide a rapid method of estimating the standing biomass at the end of the growing season (mid- to late August) to assist in formulation of the winter risk management plan, and the forecasting of risk.

1.5 Pastoral Risk and Approaches to Pastoral Risk Management (PRM)

(1) Pastoral risk

Hazards (destructive natural or human events) in pastoral communities such as those of the Qinghai-Tibetan plateau usually take a different form from hazards in agricultural or urban communities. Examples of such pastoral hazard sequences are:

• #183;drought or snow disaster >> weakened animals>>higher animal death rate>>loss of income>>increased poverty in households for whom animals are main capital stock

• animal disease epidemic>>nobody wants to buy animals>>collapse in livestock prices>>increased poverty in pastoral households which depend on income from animal sales.

The main sources of hazard in pastoral areas include:

• #183;Weather (also important in agricultural communities);

• Animal disease (problem mainly in pastoral communities);

• Predators (eg wolves) and pests (eg rodents) - (problems largely specific to pastoral communities);

• Market instability (special conditions in pastoral communities).

A future hazard of great potential impact is global climate change which will exacerbate other types of hazard.

Pastoral risk management (PRM) is the way pastoralists and government organise and prepare to reduce the impact of hazards. Pastoral risk management programmes set out what individual herders, communities and governments should do:

- every year, to prepare for the inevitable hardships of winter;
- over the medium and long term, to reduce herder vulnerability to severe hazards

Pastoral risk management is a set of four related activities:

- Risk reduction is the long term group of activities which reduce vulnerability: for example a land tenure framework that encourages sustainable pasture use, genetic improvement of animals to create better resistance to disease or extreme cold for example, or a livestock insurance scheme.

- Risk planning is the set of activities designed to prepare the herding economy for stress periods such as summer drought and winter snow disaster. Risk planning includes, for example, winter preparation of animals, or better risk forecasting.

- Risk reaction is the set of activities triggered by early warning of an impending hazard, or by the hazard itself.

- Risk related recovery refers to the proactive integration of risk mitigating measures into the recovery processes while households are re-establishing their normal or new livelihood strategies after the event, focussing on opportunities for beneficial change.

The chief actors in PRM are not just government. It is especially important to bring all key stakeholders into the programme through joint planning: involving pastoral communities themselves including experienced herders, community-based organisations, national and international NGOs.

PRM should not be treated as an isolated activity, separated from other government and community programmes. It is especially important to link it to poverty alleviation, and natural resource management.

(2) The Government disaster prevention programme in Qinghai

Qinghai Provincial government has paid particular attention to the problems of herders and especially to the mitigation of winter disasters and has recognised that it is both more efficient and cheaper to reduce the occurence of disasters rather than to deal with their aftermath. The programme (‘4CM’) has four main points:

• #183;Allocation of winter pastures to households and subsidised fencing;

• Construction of winter shelters for livestock;

• Construction of a permanent dwelling for the herders, who continue, however, to travel to summer pastures using tents;

• Sowing of 5 mu (one third of an hectare) of fodder, for hay.

The implementation of the programme is well advanced but differs in degree between counties, probably reflecting the degree of poverty of their herders. The 4CM interventions are heavily subsidised but herders still have to make some investment. In the project area autumn, winter and spring pastures are all in the same unit and have been allocated. Winter shelters, brick-built with glazed roofs and half-walls are installed throughout Henan but are absent or rare in Zeku: mud-walled sheep pens have been built throughout. The quality of permanent dwellings differs greatly between counties: those in Henan are built in brick and are generally large; in Zeku houses are of rammed earth and often small. The sowing of oats is far behind other interventions but has begun in both districts on a very modest scale: walled sheep pens are small and the target of five mu (a mu = one fifteenth of an hectare and is the common measure of farmland) per household is far from being attained.

Field investigations carried out in the two pilot villages on herders’ attitudes to the four counter-measures programme (4CM), found that the overarching attitude among herders towards the programme was positive in principle, but that full implementation of the 4CM however has not been achieved yet for various reasons.

• #183;Many households in Jilong village had not enrolled in the 4CM programme. They understood the advantages of 4CM, and would like to join the programme, but did not meet the selection standards, being too poor to pay the necessary share of the total cost. The rate of 4CM enrolment in Nanqi village was around 70 percent.

• All six randomly selected households specifically interviewed interviewed on the issue in Jilong village had taken loans from the credit cooperative with yak and sheep as the collateral. The loans were mainly used to buy forage in winter. The implementation and adoption rate of 4 CM in Zeku was lower than in Jilong,

(3) Integrating PRM and the existing government programme for disaster prevention: Elements for an enhanced PRM agenda in Qinghai

A successful programme to reduce pastoral risk and vulnerability in Qinghai means to build on and complement the structures and approaches already successfully in place, such as the 4CM, and to create additional new strategies to further enhance the resilience and capacities of herders and herder’s communities for proactive natural risk management. The policies of the provincial government must encourage such strategies.

During implementation the project identified five key components and twelve complementary strategies directly relevant to better addressing pastoral risk and vulnerability in the pilot areas and Qinghai province as a whole. The range of these issues goes beyond the scope of the original TCP design, which had identified during project preparation a more limited number of key issues. As a result the TCP project - due to its limited scope and financial volume - had neither a mandate to advise, nor the resources to conduct pilot research or to intervene on all of them. However, since these strategies are mutually reinforcing, and are all part of the recommended policy agenda for pastoral risk management in Qinghai, they are all introduced and flagged here, even though they will be discussed later on in the report at different levels of depth. The policy agenda is likely to be most successful if it implements all strategy elements in a well integrated way.

Within the agenda three concepts are important:

• Strategies are the complexes of linked activities promoted by government and the herders themselves in order to reach economic and social goals;

• Policies are the sets of rules and frameworks adopted by government to encourage certain strategies, and discourage others: policies include for example changes to pasture land rules to encourage more conservative grazing;

• Policy instruments are the specific means used to implement policies and strategies: policy instruments include for example laws, economic subsidies or new institutions.

The key components and strategies of the recommended PRM agenda are:

I. GRAZING AND PASTURE RESOURCE MANAGEMENT

• #183;Strategy 1 Measuring and monitoring grazing resources

• Strategy 2 Managing grasslands

II. WINTER FEED

• #183;Strategy 3 Feed and fodder production

• Strategy 4 Fodder markets, fodder banks and emergency fodder funds

III. LIVESTOCK PRODUCTION, BREEDING AND HEALTH

• #183;Strategy 5 Animal production and breeding

• Strategy 6 Improvement of animal health

IV. ENCOURAGING HERDER RESILIENCE TO PASTORAL RISK

• #183;Strategy 7 Early warning and rapid reaction system for PRM

• Strategy 8 Improved housing and infrastructure

• Strategy 9 Promoting herder cooperation

V. POLICY AND INSTITUTIONAL FRAMEWORKS FOR PRM

• #183;Strategy 10 Mainstreaming PRM within a reshaped pastoral economy

• Strategy 11 Financing PRM

• Strategy 12 Improving governance for PRM

Section two of this report presents and discusses in-depth the findings and recommendations derived from the project, related to those strategies for which the project had the mandate and resources to work on. Section three presents an integrated policy agenda which establishes, in spite of the need for further investigations on some aspects, guidelines for a comprehensive pastoral risk management strategy for ecosystems and production systems in Qinghai similar to those in Jilong and Zeku counties.