Species Distribution Model of <i>Varanus griseus caspius</i> (Eichwald, 1831) in Central Asia: an Insight to the Species’ Biology

Abstract

A species distribution model has been developed for one of the most well-studied varanid lizards — Varanus griseus caspius . The model was developed based on several hundred documented localities of the lizard and a number of abiotic variables using statistical and GIS-based approaches. The results of evaluation using the model demonstrated the appropriate spatial accuracy of the model. What is more important, the analysis of key variables as revealed by the model and its comparison to the published information on the biology of the Desert Monitor demonstrated the value of the species distribution model as a tool for providing additional and reliable information on certain aspects of species biology. Known data on V. griseus confirms the validity of the selection of key variables by statistical modeling of living species in suitable habitat conditions. The method of Species Distribution Modeling may be of use when studying poorly-known species, as it is able to draw the researcher’s attention to certain environmental factors previously unknown or unavailable for direct observation. The map of potential distribution of the Desert Monitor and the list of key abiotic variables are discussed in detail, emphasizing the impact of temperature and moisture on different stages of the lizard’s annual cycle. The particular impact of some key variables remains questionable. This problem may be explained in two ways. First, the species distribution model is quite imperfect as it deals with a limited set of variables which does not include biotic factors such as food availability; the pattern of vegetation; the presence of competing species; the existence of natural barriers; and others. Another explanation is more speculative and sophisticated as it assumes a lack of knowledge concerning certain aspects of the organism’s life-cycle and the complex nature of the interaction of abiotic factors. However, the value of Species Distribution Modeling should not be underestimated, as the modeling procedure is unaffected by the «human factor» and the subjectivity of an individual researcher.