Source: UNIVERSITY OF MONTANA submitted to
DETERMINANTS OF GRASSLAND DYNAMICS IN TIBETAN HIGHLANDS: PASTORALISTS, LIVESTOCK, WILDLIFE AND SOCIAL INCENTIVES
Sponsoring Institution
Other Cooperating Institutions
Project Status
NEW
Funding Source
Reporting Frequency
Annual
Accession No.
0222035
Grant No.
(N/A)
Project No.
MONZ-65955
Proposal No.
(N/A)
Multistate No.
(N/A)
Program Code
(N/A)
Project Start Date
Sep 15, 2008
Project End Date
Aug 31, 2013
Grant Year
(N/A)
Project Director
Harris, R.
Recipient Organization
UNIVERSITY OF MONTANA
COLLEGE OF FORESTRY AND CONSERVATION
MISSOULA,MT 59812
Performing Department
College of Forestry and Conservation
Non Technical Summary
Although lacking clear documentation and differing in specifics, Chinese scientific papers and government policy statements are unanimous in viewing the Qinghai-Tibetan Plateau (QTP) as having become increasingly degraded in recent decades. In response, Chinese policy has eradicated small mammals, subsidized transition from semi-nomadic herding over large spatial scales to household-level ranching on much smaller spatial scales; and subsidized pastoralists to sell their herds and abandon livestock raising entirely. Even if these programs succeed in reducing grassland degradation, they are likely to carry enormous financial burdens and create social and cultural dislocation. Even where it can be safely assumed that grassland degradation is serious and that it has been caused by excessive livestock numbers, there is little rigorous research into the reasons for overgrazing. The synthetic approach that is critical to understanding the etiology of the problem and thus to proposing policies that are both ecologically and socially sustainable has been lacking. Sustainable grassland management and economic development will require more nuanced approaches which start with a more thorough understanding than currently exists of grassland vegetation dynamics, how pastoralists? actions affect them, and, in turn, how pastoralists respond to social, economic, and policy incentives. We will study the magnitude and causes of putative grassland degradation by multiple-year monitoring of permanent vegetation plots where we can simultaneously obtain information on hypothesized proximate drivers of vegetation state: 1) weather variables; 2) site variables; 3) small wild herbivores (mainly pikas); and 4) domestic livestock. In addition, we will seek investigate the complex historical, cultural, and economic, factors facing pastoralists as they make decisions affecting wild and domestic herbivores. Because the objective is to provide useful information that informs real policy in China and to develop useful models of local ecological dynamics within larger systems that are applicable more generally, we need detailed knowledge of what pastoralists are actually doing on these lands. Thus, our data gathering must occur in working pastoral allotments rather than experimental plots. We will use the natural variability in the system arising from different soils, slopes, and elevation, differing pastoral paddocks, and differing policy environments (subsidizing the transition from semi-nomadic to household scale vs. encouraging the complete elimination of pastoralism), variable annual weather, and varying pika densities (arising from differential success in the poisoning program) as a basis for interpreting causative mechanisms). With 4 years of data, we will first generate statistical models of grassland status as a function of these covariates to measure the strength, direction, and interaction of major factors. To generalize our results, we will produce an integrated model that links vegetation state through pastoralists to off-site socio-economic forces that will yield insights into sustainability and resilience of the coupled social and ecological system.
Animal Health Component
(N/A)
Research Effort Categories
Basic
(N/A)
Applied
(N/A)
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1210799107033%
1360830107033%
3060830107034%
Goals / Objectives
Grassland degradation is a global concern, affecting not only wild species and pastoralists who rely on healthy grasslands for their survival, but also non-local people. Livestock grazing is the dominant form of land use in Central Asia, and pastures of the Tibetan highlands are located upstream of 40% of the world's human population. Grasslands on the Tibetan plateau are described as increasingly degraded. Causes for this grassland degradation are attributed to over-stocking of livestock, poor livestock management, historical-cultural factors, alteration of land tenure arrangements, changes in socio-economic systems, climate change, and excessive herbivory and soil disturbance from wildlife. But studies have yet to provide clear support for any putative causative agents, and have not examined interactions and complexity among these factors. Thus policy choices to reduce or reverse grassland degradation are often made without a clear rationale and are based more on prejudice or convenience than evidence of their effectiveness. We will examine multiple correlates of grassland status and trends simultaneously, using replicated measurements at permanent plots in a multi-strata design, measuring the strength of evidence for various competing hypotheses. It will link ecological measurements directly to current and recent historical actions by pastoralists, which in turn are affected by cultural norms, economic incentives, and policies of central and provincial governments. In addition to biophysical attributes of each site, livestock density and pasture usage patterns will be quantified. Importantly, each site will also be described by the particular grazing strategy employed by the pastoralist managing it, and that strategy, in turn, will be related to the complex of economic and policy incentives and historical determinants that pastoralists face. This data will be used to motivate the development of models that link broad historical, policy, economic, and cultural factors to local grassland conditions as mediated by the agency of individual pastoralists which, in turn, may be used to evaluate the implications of different policy interventions. This work will deepen our understanding of the complex interactions involving geophysical, biological, social, and policy factors and feedback systems in determining grassland status. Because multiple factors affect grasslands simultaneously and interactions are critical, the multi-disciplinary, the systems-approach adopted by this project is fundamental. Our enhanced understanding of this socio-ecological system will provide important input for policies on grassland restoration, biodiversity, and economic development in arid ecosystems worldwide. In addition to direct training of students, the project will train numerous Tibetan field assistants, and will coordinate closely with local and provincial grassland and forestry officials. Direct collaboration with Chinese scientists and officials, as well as facilitated workshops, will enable research results to be understood by policy makers. Direct interactions with local pastoralists will allow immediate, practical applications of project results.
Project Methods
Using a systematic sampling (on a grid with of 250m), we will establish 30 0.5m2 vegetation plots within each of 16 winter allotments at each study site. All 960 plot locations will be permanent fixed plots to allow for repeat measurements. At each plot, fixed attributes to be recorded will include elevation, slope, aspect, soil type, soil structure, pastoralist responsible, and whether or not inside an existing winter pasture fence. Vegetation at each plot will be quantified two times/yr. We will record: 1) species present; 2) cover by species using canopy cover estimates; 3) total vegetation coverage; 4) an estimate of the amount of litter (g/m2); 5) a measurement of pedestalling and description of signs of erosion; and 6) whether or not there are evident signs of recent livestock use and an estimate of percent use for each vegetation species. Reference photos will be taken on a digital camera. Regressions of biomass on time from exclosure plots will be used to standardize all plots to a common day. Biomass will be estimated using reference specimens weighed on the day of sampling. Randomly selected plots will be selected for a companion, off-site clipping experiment, to calibrate that day's estimated weights. Species-specific dry weights will be estimated by clipping off-site plots and oven drying. Recorded fresh weights will then be corrected to dry weights using the nearest day's wet-to-dry weight ratios. Plant species will be categorized by their attractiveness to wildlife by 1) proportion found in fecal fragment analysis; and 2) use/availability proportion from fecal fragment analysis. A sub-sample of vegetation under various abiotic and herbivory conditions taken at the end of the growing season, i.e., prior to the return of wintering livestock, will be collected and subjected to laboratory analysis for protein and fiber content at the Chinese Academy of Agricultural Sciences in Beijing or similar institution. Herbivore exclosures wil provide a standardized measure of vegetation growth as affected by seasonal and annual weather patterns and a rough estimate of vegetation consumed by the excluded herbivores. Four types of exclosures will be established: 1) livestock excluded; 2) pikas excluded; 3) pikas removed and allowed to recolonize; and 4) both livestock and pikas excluded. The number of sheep, goats, yaks, and horses will be determined for each pastoralist, and the size of the winter allotment quantified via GPS delineation of the boundary. Livestock density for each species, sex, and age group will be calculated as number/area. For yaks, we will use GPS collars on a sample of animals; for smallstock we will use GlobalSat GPS dataloggers strapped to sheep (e.g., Kawamura et al. 2005) download data approximately every 2nd day, then place on another haphazardly selected animal. We will model utilization distribution (UD) by each herd (e.g., using well-explored kernel home range models), and application of herd size to the UD provides a measure of relative intensity of livestock use. Sheep will also be weighed when handled, allowing an estimation of mass dynamics among small-stock as related to all other independent variables.