112 FRESHWATER ANIMALS 



unglaciated portions. However, too little is known of the species 

 and their distribution to permit meaningful speculation about 

 the historical zoogeography of this genus. 



Svardson has found that the chief source of difficulty in the 

 recognition of Coregonus species is the extraordinary phenotypic 

 variability of each genetic stock (species). The coexistence of 

 several species in a lake (see Fig. HA), a circumstance which 

 often complicates the problem for the systematise has actually 

 provided Svardson with the major clues to the identity of the 

 species. In these ways both Coregonus and Daphnia present 

 somewhat similar problems and, as in Daphnia, there are evi- 

 dence of the formation of hybrids and some resemblances be- 

 tween the species of major drainage systems which suggest 

 introgressive hybridization. 



Phenotypic Variation. The primary reason for the failure of 

 orthodox taxonomic methods in Coregonus systematics has been 

 that the phenotypic characteristics are not constant for, and 

 therefore not indicative of, a given genotype. Quite the contrary 

 is true; the phenotype is highly variable, depending upon the 

 environmental conditions under which the genetic stock devel- 

 oped. These plastic characteristics include meristic ones (number 

 of vertebrae, fin rays, scales), size and proportions, and also 

 physiological ones ( growth rate, onset of sexual maturity, length 

 of life). The allometric growth relationships are complex because 

 the proportions within a stock differ not only in fish of different 

 size (length) but also in fish of the same length but of different 

 age (Svardson, 1949, 1950). The number of gill rakers is the only 

 phenotypic characteristic that is apparently unaffected by the 

 environmental circumstances during development. This is the 

 only phenotypic manifestation indicative of the genotype, and, 

 therefore, of taxonomic utility. However, the number of gill rakers 

 is under close selective control and may vary somewhat between 

 allopatric populations of a species. Svardson was able to modify 

 the g\\\ raker count in a population introduced into a formerly 

 Coregonus-hee lake by imposing an artificial selection for this 

 I actor at the time of introduction. It is little wonder that the 



