O. Gy. Dely szerk.: Vertebrata Hungarica 22. (Budapest, 1984)
Korsós, Z.: Comparative niche analysis of two sympatric lizard species (Lacerta viridis and Lacerta agilis) 5-14. o.
crohabitat differences because of the close niche relations. Time as niche dimension may only crop up in connection with the animals' activity periods. Segregation of this regarded as occasional (SCHOENER 1974), but probably only little investigated. The relationship of the two lizard populations is presented according to these three main points of view. The study area, situated In the field site of the Game Biology Reserach Station of the University of Agriculture, Gödöllő, was characterized In detail in a previous paper (KORSÓS 1982). The methods of collecting data on the individually marked lizards were also described there. In addition to measurements of environmental variables (soil and air temperatures, Intensity of light exposure, plant density, soil type, wind velocity) the food of the lizards has also been studied.The basis of this was the material of stomach contents collected by I. SZABÓ. Some head proportions suspected to be correlated with food size were also measured. These are length and width of pileus, head depth and snout-ear distance; all of them expressed in percentage of total snout-vent length. The specimens from which the stomach contents examined are deposited in the Zoological Department of the Hungarian Natural History Museum. Niche overlap was measured according to SCHOENER (1968). That means graphically the common area under the curves of relative frequency distributions describing the demands of the two species. Mathematically: where p x j and Pyj are the relative frequencies of the two species (x and y) belonging to the i-th category of the resource set (s). SCHOENER' s formula has a great adventage over that of LEVINS (1968), namely that its extreme values are 0 and 1, and it is equal for both species (i.e. symmetrical). In the case of microclimatic variables the overlap was calculated on the basis of extreme, limiting values. This, being asymmetrical to the species is actually rather similar to the competition coefficients, but the three-dimensional representation of niche would have demanded a more difficult method to obtain a symmetrical formula. In some cases (calculating the overlap of the microclimatic variables, head measurements of the lizards, and the microhabitats and activity periods together) discriminant analyses were performed on an IBM 3031 computer using the BMDP statistical package (DIXON 1981). The option for reclassifying every case to one of the two groups according to the classification function obtained In the analysis results in a probability of misclassification, which may also be considered as a measure of niche overlap (STEINHORST 1979). Sample data are listed in Table 1. The data recorded in the study area include not only the handled but the sighted and identified lizards as well. The population sizes, estimated by the computer programme of OTIS et al. (1978) have been found as 48 Í 9 for L. viridis and 96 t 15 for L. agilis . The population densities (80/ha and 160/ha, respectively) are in accord with those of PETERS (1970) and TERTISHNIKOV (1970). It was possible to reduce the soil and air temperature and the measurement of light exposure out of the environmental variables on the basis of a principal component analysis detailed elsewhere (KORSÓS 1982). They affect niche segregation in a way similar to one another acting as microclimatic variables. The positive correlation between them also supports this relationship (X^ = 30.58; p<0.1 %). The volumes fit for survival of the populations are delimited by the extreme-limit values of the three variables as shown in Fig. 1. L. viridis has a restricted activity as regards air temperature and light exposure as well, but it definitely shows preference for warmer values of soil temperature. L. agilis has generally a greater tolerance, however it does not favour higher soil temperature. The overlaps calculated from the limiting values are 64.9 % for L. viridis and 31.2 % for L. agilis . Regarding not only the extreme positions of the survival volumes, but the relative frequency distributions of discriminant scores of the microclimatic variables, the overlap between the two species has been found as 63.5 %. MATERIALS AND METHODS s RESULTS
