This investigation sought to develop robust ranger based monitoring strategies for the Northwest pre-Caspian population of the saiga antelope, Saiga tatarica, in two nature reserves created for its conservation. The saiga is a critically endangered species which suffered a 90% population decline due to poaching for horns to be used in traditional Chinese medicine and for its meat. Monitoring of this species is highly important as despite extensive conservation efforts, poaching is still occurring. The effect of poaching and the conservation work on the status of the species is vital information for effective future conservation planning. A GIS map was created of each reserve showing the location of vehicle tracks. It was found that there is a significant positive relationship between track density and saiga sightings in the current monitoring system suggesting this may be a significant source of bias. Two experiments into the accuracy of the rangers’ count estimates were conducted. A comparison of ranger estimates against photographs taken of saiga herds found the rangers tended to over estimate the number of saiga in a herd, although the data only covered a small range of herd sizes. An experiment using flocks of sheep as a proxy for saiga was conducted. The rangers tended to underestimate the number of sheep in a flock and in some cases there was a significant change in bias as the flock size increased. Finally, distance sampling was evaluated as a potential monitoring technique. However, due to the behaviour of the species it was found not to be suitable due to the model’s assumptions not being fulfilled.

The investigation provides recommendations for the future monitoring of saiga, based upon a strip transect sampling system and using stratification of data according to distributions of monitoring effort. This will increase the utility of the data collected by the rangers without having to introduce a complex monitoring system which could affect their ability to fulfil their main role of preventing saiga poaching.