By Melissa Songer
Melissa Songer is an ecologist with the National Zoo; e-mail: SongerM@si.edu.
--------------------------------------------------------------------------------
The Przewalskis horse represents the last "true" wild horse left on planet Earth. Although there are horses that roam free in the American West and other regions, they all descend from horses that were once domesticated and later became feral.
Although they were able to resist domestication, the Przewalskis horses couldnt withstand other pressures from human activities and went extinct from the wild throughout their range during the 1960s. For decades, zoos supported the only surviving Przewalskis horse populations, thus saving the species from dying out completely.
The horses once ranged across the vast steppes, grasslands and deserts of Mongolia, into China, Kazakhstan and possibly beyond. For centuries, the Przewalskis horses shared their habitats with pastoral nomadic societies.
Due to their elusive nature, Przewalskis horses werent scientifically described until near the end of the 19th century, when they were named for a Polish naturalist, Nikolai Przewalski, who went to Central Asia to investigate rumors of the wild horses.
Przewalskis horses differ in appearance from their domestic cousins, with a shorter and stockier build, a tan coat and dark mane. Cave paintings of horses found in Western Europe look similar to Przewalskis horses and suggest they once roamed all the way to Spain some 20,000 years ago.
Melissa Songer prepares a satellite collar prior to capture of a Przewalskis horse in Kalameili Nature Reserve.
However, major changes to ecosystems brought about by the last ice age reduced their available habitat significantly. During the first half of the 20th century, poaching increased, and livestock management intensified dramatically, thus contributing to the decline of the wild populations.
After many generations of surviving only in captivity, the horses began to be reintroduced to some of their native habitats in Mongolia during the 1990s. In 2001, the Wild Horse Breeding Centre (WHBC) of Xinjiang, China, also began releasing Przewalskis horses in northwestern China, where the Przewalskis horse are called Yema, which, in Chinese, simply means "wild horse."
The release sites are within the desert ecosystem of the Kalameili Nature Reserve (KNR) in the Dzungar Basin, which has long been considered the cradle of the Przewalskis horses; its the place where the last surviving horses were sighted in the wild. The KNR also is home for a variety of other important wildlife, including the Asiatic wild ass (khulan), goitered gazelle, argali sheep and wolves.
Scientists and veterinarians from the Smithsonians National Zoo have joined the effort to provide technical support for monitoring the released horses. Theyre now using cutting-edge technology to follow these ancient animals.
Tracking Ancient Horses
Little information exists on the ecology and habitat requirements of Przewalskis horses, because they disappeared from the wild prior to any extensive scientific studies. National Zoo scientists Peter Leimgruber and I are working to implement satellite and radio tracking of released groups of horses by deploying tracking collars on individual horses. Partnering with WHBC staff and National Zoo veterinarians Nancy Boedeker and Suzan Murray, new horses have been collared each year since 2006.
Melissa Songer, National Zoo ecologist (left), and Ma Ke, Wild Horse Breeding Centre veterinarian, position a collar on a Przewalskis horse in the Kalameili Nature Reserve. Collars house a GPS receiver, a VHF radio beacon and satellite-transmission equipment.
Because the horses live in family or bachelor groups, many horses can be tracked by a single collar deployed on one of the group members. WHBC staff spend many hours tracking the horse groups to assess their health, monitor their interactions and assess risks. The staff can use radio telemetry on the collared horses, drastically cutting the time needed to find the groups.
The collars help staff check up on horses on the ground as well as record their daily movements via GPS. Regular positions of the horses are transmitted via satellite and received by e-mail at the National Zoos Conservation GIS Lab, where theyre combined with habitat, roads and other spatial data for analyses with ESRI ArcMap software. These analyses will deliver key information about the habitat and space requirements of the horses and provide a better understanding of their basic ecology and behavior.
Challenges to Reintroduction
The Kalameili is vast, covering more than 17,000 square kilometers (about 6,600 square miles), and there are many threats to the Przewalskis horses. During winter, nomadic Kazak families migrate through the area with more than 200,000 livestock, including domestic horses that they use for transportation.
Movement positions were recorded between August 2007 and September 2008 for a Przewalskis horse released in the Kalameili Nature Reserve, including Landsat ETM+ imagery from August 2002 (top), NDVI 16-day composite data from NASAs Terra Moderate Resolution Imaging Spectroradiometer (middle) and DEM data produced by NASAs Shuttle Radar Topography Mission (bottom).
Winters are harsh in the desert, and forage is scarce. The livestock bring competition for food and the potential for spreading domestic livestock diseases to the wild horses. Some domestic stallions compete with wild stallions, attempting to take over their harems. This puts stress on the wild stallions and brings risk of interbreeding among domestic stallions and wild mares. Hybrid foals born from domestic and wild horses are fertile and have the potential to dilute the Przewalskis horse species if they reproduce and mix with the wild horse population.
To add to the challenges, a major highway splits the reserve, and theres danger that the wild horses, drawn to water that collects in ditches by the highway, could be hit by the many large trucks and cars speeding through the reserve.
The Kazak families rely on the Kalameili for the survival of their herds when the snow and ice are too deep in the mountains"preventing their migration would threaten the livelihood of the Kazaks. Through community surveys and outreach, it has been learned that most of the families have no other source of income apart from raising livestock.
However, theyre enthusiastic about the efforts to return the Yema to the wild"they admire and depend on horses every day. Some of the elders remember seeing the Yema in the wild and are excited to see them return. Although some are wary of how it may impact their access to the Kalameili, theyre interested in the projects success.
For a successful reintroduction, all the moving pieces must be considered"humans, horses, other wild ungulates and livestock"within the matrix of the natural features of the ecosystem and man-made components. Scientists from the Xinjiang Forestry Bureau, the department that manages KNR, have been monitoring other wild ungulates in the area for several years.
Theyre also working with the Kazak herders to get key information needed to map the migration routes and livestock pastures in the Kalameili. By working together, all these components can be brought together for improved reintroduction strategies and management.
Mapping Primary Productivity
In addition to mapping these features, one of the keys to unlocking the puzzle is understanding the seasonal dynamics of vegetational changes throughout the reserve. This will help us understand how the horses can be sustained given the available resources and requirements of livestock and other wildlife as well as locate the best sites for future releases. Were now working to determine horses habitat preferences based on remote-sensing analyses of satellite imagery to assess grassland productivity in the reserve.
Wang Chen, monitoring team leader for the Wild Horse Breeding Centre, tracks Przewalskis horses in a valley using a radio telemetry receiver that detects VHF signals emitted from their collars.
Through maps of primary productivity, we can evaluate effects of winter grazing by livestock, assess the potential for competition with other wildlife and map the remaining good pasture for the Przewalskis horses. The project uses ERDAS Imagine to create seasonal primary productivity maps of the KNR using the Normalized Difference in Vegetation Index (NDVI), which is calculated by comparing vegetation reflection in the red and near-infrared spectrum.
NDVI data are continuously collected at a global scale by NASAs Terra satellite and its Moderate-Resolution Imaging Spectroradiometer. The data can be freely downloaded from the Web at edcimswww.cr.usgs.gov/pub/imswelcome. With the NDVI data, biweekly maps of vegetation productivity can be created, which can be linked to GPS tracking data from the Przewalskis horses as well as survey data for livestock, Asiatic wild ass and goitered gazelle.
Using animal tracking positions, predictive maps of Przewalskis horse habitats are being developed by determining NDVI as well as other environmental covariates such as distance to nearest water source and roads. With generalized linear mixed models, the project team will try to predict Przewalskis horse presence. All covariates will be extracted from map layers in ESRIs ArcMap, producing a Przewalskis horse habitat-prediction map for all of KNR for any given two-week time period.
As more information comes in about presence and abundance of wild ungulate species and livestock, such data can be built in as covariates in the Przewalskis horse habitat model. These analyses allow testing for competition by quantifying the extent to which selected species (e.g., livestock) displace Przewalskis horses.
Solutions for Horses and People
Unlocking the puzzle may lie in bringing all the components together via GIS and remote-sensing analysis. Geospatial analysis tools are providing the capacity to develop, implement and monitor conservation strategies by generating more and better information than traditional field surveys alone can provide.
By integrating maps of horse movements, winter migration of livestock, seasonal dynamics of natural resources and threat locations, the data can be analyzed in a spatially explicit environment to address key ecological and management questions. Based on such analyses, it will be possible to assess KNRs carrying capacity for Przewalskis horses, interspecific dynamics of wild ungulates, impacts of livestock, and how livestock management and stocking rates may need to change. The results will provide urgently needed information on the ecology of Przewalskis horses in the wild and inform selection of future release sites in China and elsewhere.
The ultimate goal is to enhance the chances for survival of the wild horses and other wildlife in ways that wont negatively impact the regions people. The maps generated from these analyses can be used to communicate effectively with decision makers as they link ecological processes (e.g., animal movement, biodiversity, population dynamics, etc.) to temporal and spatial patterns on the landscape in an easy-to-visualize format. This will be a key step in bringing key stakeholders and decision makers into the process and make it sustainable into the future.
In the near term, the results will provide essential information for developing a comprehensive land-use strategy beneficial for the Kalameili Nature Reserve, the people of the region and the wildlife.
| 
.jpg)
