| Abstract: | To estimate the role of history in variation of tree species richness in the forests of the Western Caucasus we analyzed correlation between their local richness (S--the mean number of species per 300 m2) and size of actual species pool (N--the number of species per 1 ha). If compared communities are differently distant from the point of evolutionary equilibrium one should expect a significant variation in correlation between S and N (determined with the greater sensitivity of N than S in respect of historical factors). The lower value of N/S corresponds to less saturated level of historically determined species richness. A mean N/S ratio in Japana temperate broadleaved forests (Masaki et al., 1999) provided the basis for analysis. The present tree species richness of the forest communities in the 1 ha plots seem essentially determined by the historical processes. The mountain forest communities of Western Caucasus are characterized on the average with lower saturation level of the actual tree species pool in comparison with the Japan temperate broad-leaved forest communities. On the Western Caucasus the middle mountain beech and coniferous-broadleaved communities (400-1600 m a.s.l.) are characterized with the higher saturation level of the actual species pool in comparison with communities located lower and higher. These results confirm published historical reconstructions, according to which the middle mountain forest communities in the Western Caucasus are older than forests located higher or lower. Present low mountain forests of the southern (to Black Sea) and the northern macroslopes of the Western Caucasus are characterized with similar saturation level of the actual species pool. These data agree with the assumption of Dolukhanov (1980) that low mountain zone of the southern macroslope was not a refuge for tree species in Pleistocene. |
