From 1 - 4 / 4
  • This dataset describes environmental conditions at 135 Saiga antelope calving sites (from a total of 214) in Kazakhstan where the predictor variables required for the modelling were available at sufficient resolution. Data collected included climatic variables associated with haemorrhagic septicaemia in the literature, including humidity, temperature and precipitation. Indicators of vegetation biomass, phenology and length of the winter preceding calving were represented using the Normalised Difference Vegetation Index (NDVI), snow depth and snow presence data. Saiga antelope are susceptible to mass mortality events (MME), the most severe of which are caused by haemorrhagic septicaemia following infection by the bacteria Pasteurella multocida. These die-off events tend to occur in May during calving, when saigas gather in dense aggregations. As the bacteria is a commensal organism, which may live harmlessly in the respiratory tract of the saiga, it is believed that an environmental trigger is involved in a shift to virulence in the pathogen or reduction in immune-competence in the host. The attached data show environmental conditions at a set of calving sites of the Betpak-dala population of saigas. This population, one of three in Kazakhstan, is located in the central provinces of the country and is the only one in which massive haemorrhagic septicaemia outbreaks have been recorded. At most of the recorded sites, calving progressed normally, whilst at others mass mortality events occurred during calving or just afterwards, namely in 1981, 1988 and 2015. A set of environmental predictor variables was used to model the probability of an MME at calving aggregations. The dataset, modelling process and results are described in Kock et al. (2018): A related shapefile of the full set of 214 sites, and metadata concerning site characteristics and the provenance of the location data is available at: The attached dataset and site metadata in the above-mentioned Shapefile attribute table can be combined using the variable ID in order to merge the environmental data with information on the calving and MME sites. Full details about this dataset can be found at

  • The database of chemical composition of Central Asian forage plants contains just under 1000 desert and steppe species with information such as Latin and Russian names and family and related records of chemical composition from various sources including percentages by weight of protein, ash, cellulose and fat. Where available, it also includes data on digestible protein content, metabolisable energy and Soviet Feed Units (SFU). Records also include information on the country, location, season or month and phenological phase at time of collection of each sample. As one of the original uses of the database was for modelling food and energy intake by the saiga antelope, it also includes information identifying saiga food plant species along with sources of this information. Data on the edibility of many species for livestock in different seasons are also available. See the detailed documentation available here for more information on the data types, definitions and sources. NB The database is in text format and must be imported e.g. into relational database software, as Unicode (UTF-8) in order to convert the Cyrillic characters in Russian names. Full details about this dataset can be found at

  • This dataset records the Saiga antelope die-off and calving sites in Kazakhstan. It represents the locations (and where available dates) of (i) die-offs and (ii) normal calving events in the Betpak-dala population of the saiga antelope, in which three major mass mortality events have been recorded since 1988. In total, the data contains 214 saiga die-off and calving sites obtained from field visits, aerial surveys, telemetry and literature. Locations derived from field data, aerial surveys or telemetry are polygons representing the actual size and shape of the die-off or calving sites; locations sourced from the literature are point data around which buffers of 6km were created, representing the average size of calving aggregations. Of the 214 locations listed, 135 sites for which environmental data were available were used to model the probability of a die-off event. The collection and use of these data are written up in more detail in papers which are currently under review (when published links will be added to this record). Saiga antelope are susceptible to mass mortality events, the most severe of which tend to be caused by haemorrhagic septicaemia following infection by the bacteria Pasteurella multocida. These die-off events tend to occur in May during calving, when saigas gather in dense aggregations which can be represented spatially as relatively small sites. The Betpak-dala population is one of three in Kazakhstan, located in the central provinces of the country (see map). Full details about this dataset can be found at

  • Field-pathological findings of 33 saiga antelope carcasses (adults and new-born) found in two sites (Tengiz and Turgai, Kazakhstan) during a mass die-off event in May 2015. In Kazakhstan May 2015, approximately 200,000 saiga antelopes died within a month-period causing a loss of two-thirds of the global population. The dramatic event occurred during calving season when females and young males stop migrating and form massive aggregations for calving purposes. With 100% morbidity and 100% mortality of affected herds observed, the 2015 die-off left the largest saiga population, Betpak-Dala, with approximately 30,000 survivors based on post mortality census, highlighting the imminent extinction threats to this critically endangered species. The lack of pathological investigations during historical mass mortality events has limited our understanding of disease-related mortalities in saiga antelope. Generally, aetiological agents were isolated from dead saiga, but the disease course and a full necropsy were not performed nor present in the records. However, for the first time, a full pathology report was possible during 2015. Full details about this dataset can be found at