Improved management of small mammals in Tibetan grasslands

• Chinese Academy of Sciences, China
• Department of Agriculture of Tibet Autonomous Region, China
• Tibet Bureau of Agriculture and Animal Husbandry, China
• Tibet Academy of Agricultural and Animal Sciences, China

Grasslands in western China, particularly Tibet Autonomous Region, are important for livestock farming systems. These systems are important for maintaining traditional farming approaches which, based on competing for scarce resources, form the basis of the Tibetan farming economy. Amongst the competitors for forages are rodents and other small mammals that, through burrowing and consumption of vegetation, contribute to soil erosion and plant losses and consequent grassland degradation. Activities of these small mammals have degraded up to 15 per cent of the valuable grassland ecosystems on which Tibetan farmers depend.
Control operations against these small mammals currently use non-specific toxins across an area that covers 1.6 million hectares. Species biology and habits are not incorporated into this broad-scale approach. In some places these measures have effectively eradicated up to 95 per cent of rodents, but in most areas the results have been far less impressive.

The main animal studied was the plateau pika, the keystone species. Determining effective control techniques for the pika focused on areas used for summer grazing. The effects and role of pasture management in these was also assessed. The impact of these measures was assessed for non-target species, such as other pika species, zokors and marmots and birds. An immuno-contraceptive vaccine, administered through appropriate oral delivery systems, was also examined.

Pika population numbers were monitored in Naqu, Bange and Anduo counties, using techniques such as burrow counts. This was done at the start and end of the breeding season, enabling population estimates from three seasons - two breeding and one non-breeding. Capture, mark and recapture activities were carried out in four areas to test pika abundance over time. Plant cover, height and bio-mass, plus some species composition data were collected, along with climatic data. All the information gathered was assessed as part of developing the modelling strategies to study the factors affecting population numbers.
At rodent control areas, fences were tested in various designs. Limited grazing to allow higher levels of plant coverage was also implemented. Some severely degraded grasslands were replanted. Levels of abundance of small mammals were recorded in all areas where pika population data were collected. Changes in numbers over time were used in assessing impacts.
The scientists sought to develop a broad-scale fertility-control vaccine for use in non-toxic baits, beginning with a model system for oral delivery to laboratory mice.
The project also included studies of ecologically-based options for farmers in south-eastern Australia's cereal production areas to manage mouse plagues. In Australia experiments were conducted at the Mallee Research Station, Victoria. The scientists studied the processes most likely involved in breeding onset and breeding cessation.

Throughout the project it became clear that overgrazing was a major issue for the plateau, and that if the serious degradation taking place was not stopped the productivity of the system would be lost - for many decades at least and possibly indefinitely. The alpine meadow system is characterised by a hard turf layer, and when this is broken down it exposes a much looser and less productive soil which the locals call 'black soil'. It is extremely difficult to restore the system when it gets to this state and considerable effort should be made to ensure that areas not already affected remain this way.
Livestock numbers on the plateau have more than doubled in the past 50 years, but livestock carcase weight has declined in this time, suggesting that the system has been pushed so that animal productivity is declining. At the same time plateau pika numbers have increased, due to some factor in the system that has changed to their benefit. This is likely to be either a change in pasture composition and/or structure, and/or improved burrowing conditions due to increased erosion. The project team recommended a reduction in livestock density, to prevent the system degrading further and moving to the highly degraded 'black soil' state. More work is required to ensure this is accomplished whilst maintaining or improving the livelihoods of the local people. A focus on animal productivity will likely be one part of the solution.
What is not clear is the importance of concurrent pika control in moving back from the boundary where the system transitions to the black soil state. Hence, at the same time work is required to determine whether the pika problem will disappear with a change in livestock grazing practice (reduced grazing), or whether pika control will be required, at least in the initial stages, to help the system return to a state where pikas are not benefited.
If pika control is required, then the results from this study indicate that the current technique of 'once-off' control with Botulin toxin C on wheat bait is not appropriate in the short alpine meadow. While immediate control of pikas was rapid (ca. 90% reduction), populations were able to recover rapidly over the following breeding season. The team also found no strong evidence that controlling pikas resulted in increased pasture production over summer, so the control practice provided little benefit to local people.
One way to prevent rapid population recovery, which is driven primarily by reproduction, is to use fertility control. This component of the study was carried out in Beijing on Brandt's vole, a species that is not a pest in Tibet, but one that is a pest in other parts of China, to build on past research funded by ACIAR. Results suggest that oral delivery of immuno-contraceptive control will be technically difficult and chemical control may be a more viable option.
Tibetans conserve forage for winter by the use of fencing on a subset of land and this appears to benefit pikas, with higher populations inside fenced areas after winter than in adjacent areas outside fences. This result is not surprising, given the extremely low biomass of vegetation that exists outside fenced areas at the start of winter under current grazing practices. This result means that the pika density remains relatively high throughout the winter season inside fenced areas, potentially providing significant competition to the stock. However, the team gathered no data that directly measured this impact.
The issue of biodiversity impacts following pika control is a complex one. The current diversity and abundance of non-target species that exist around Naqu could, at least in part, be considered a response to current grazing practice and pika abundance and hence not necessarily representative of the natural diversity and/or abundance of biodiversity for the area. Data were collected on raptors, which eat pikas, and on some of the small birds that nest in pika burrows. The information will enable people to determine the impact on these species of long-term, large-scale changes in management of the grassland around Naqu, but thought needs to be given to what the target densities should be. It is also important to note that benefits to biodiversity will probably occur from a reduction in both livestock grazing and pika populations.
The skills required for grassland management and for scientific investigation by local people in Tibet increased significantly in this project. Already there is evidence the techniques have been adopted and used in studies outside this project. Another good linkage with a social scientist from Tibet provided the project with valuable data.