Ihe Problem 



[he term "richness ot fauna" may be understood in two ways: first, as a great number of 

 species found in any one place, and second, as a large number of individual animals. For the 

 first case, it is better to use the term "qualitative richness ot fauna" and for the second, 

 "quant i ta t i ve". 



The significance of topographic variation on qualitative richness ot fauna is generally known. 

 The majority of zoogeographers explain diversity by historical causes, although no one denies the 

 influence ot present-day ecological conditions. The relative importance of these two groups of 

 factors is still uncertain. 



Introduced to science by Fisher (Fisher, 1^25, 1928) and rapidly developing now (Sheffe', 

 1959; Plokhinski, I960), the method of "dispersion analysis" or otherwise called "analysis of 

 variance" affords the opportunity to determine the relative significance of the influence of 

 separate factors on the total variability of a phenomenon. Statistically eliminating contemporary 

 ecological factors (in the first place, abiotic), it is possible to approach the determination of 

 the magnitude of influence of historical factors. Will this be the same for different groups of 

 organ i sms? 



M e t ho ds 



Inventories of the terrestrial vertebrates of the USSR have reached a stage wh ' ch permits the 

 application of quantitative methods. The icthyofauna has also been well studied, but the complex 

 of basic ecological factors on land and in the water is so vaned (Hesse, Allee, Schmidt, 1950 

 that it would be more reasonable to devote a special paper to the analysis of the icthyofauna. 



was divided into 66 equal-sized quad 

 number of species found in each quad 

 rid. The raw data tor these compila 

 ribution of terrestrial vertebrates 

 44; Denent'ev and Glavkov, 1951-195*1 

 es from the data taken out of the af< 

 ng to wore recent information. The < 

 into quartrat' was completed by S. \ 

 ces of abiotic (physical) conditions 

 of maps from the Bol'shoi Sovietskoi 

 rid (1937), and the book by A. A. Bor 



■at! 



of our 

 and 



of 



for 



Fncy 



nep 

 the 

 clop 

 (19 



10" 

 det er 



obta 

 un t ry 

 ent 'e 

 ned s 

 bor-c 

 and 

 separ 

 edia 

 48). 



longiti 



Timed 

 ined fr 



v and 

 ummar i < 



de 



ate 

 (1957), 



For the sake of standardization and generalization of present statistical treatment, the 

 systematized scheme and designations of the book of N. A. Plakhensk (l96l) were uted, with the 

 exception of the term "deviation" tor which the more common term "variance" was substituted. 

 Variance for each quadrat is the square of the deviation of that quadrat's value from Ihe near 

 value for all quadrats; r-1 is the number of degrees of freedom of non-random factors; n-r is 

 the number of degrees of random factors; \* is the correlation ratio for non-random factors; 

 07* is the factor's variance; Og is the error's variance; F is the variance ratio (the con- 

 fidence ievel of 95? is used). For condensation in the tables, the following abbreviations are 

 made: M = mammals, A = birds, R = rept ; les, and Am = amphibians. 



Formal I ndices 



It stands to reason that latitude and longitude, 

 consideration "ecological factors", but when these are 

 causal factors it is possible to consider several exee 

 noted that "an'mal and vegetable species become more n 

 to the equator", and th's is now known to be generally 

 views are contained in many other reports (Eccart, 191 

 Romer devoted an entire speech on the 1st of December 

 Naturalists especially to the question of the d'munuti 

 the increase in geographic latitude (Romer, 1907)- ln 

 the quantitative expression of the stated dependence f 

 19*6; Kusenov, 1957i and others). By a priori conside 

 the changes in dependence on longitude will be in grea 

 not ecological, principles. 



