16 UNIVERSITY OF WISCONSIN STUDIES 



The latter conclusion is clearly indicated by the fact that 

 two-thirds of all the species with wide geographic ranges 

 invade every habitat in the lake, while none of the species 

 with limited geographic ranges are found in all habitats. The 

 species confined to rivers are not restricted in habitat, except 

 that they are confined more or less to turbulent water forma- 

 tions. The cula is a mud eater and hence has an ubiquitous 

 food but is limited in range and in ecological distribution, 

 ranging geographically through Trinidad and the Valencia 

 basin and being limited ecologically to the rushes along the 

 lake shore and to the streams. The coroncho is still more 

 limited in geographic range and is restricted ecologically to 

 swift streams which contain large boulders. It clings to the 

 under side of rocks and subsists entirely on the fine muddy 

 deposits which collect in the eddies. 



Perhaps enough has been said to bring out the point the 

 writer wishes to make; a wide-ranging species usually feeds 

 upon an ubiquitous food and is often able to invade a number 

 of habitats; species with a limited geographic range are limited 

 ecologically in food, habitat, or both. The fishes discussed 

 in this paper illustrate this principle very well. Shelford (1911) 

 relates the activities of fishes to the law of minimum and 

 says, "the activity which determines the range of conditions 

 under which a species will be successful is the activity which 

 takes place within the narrowest limits." He also points out 

 that "centers of origin" are sometimes only places where 

 optimum conditions for a number of species exist. The writer 

 would emphasize the view that, without other evidence, the 

 fact that a species had a wide range would not necessarily be 

 taken as indication that it was older than an associated species 

 with a restricted range, as geographers have sometimes asserted. 

 The ecology of each species would have to be considered with 

 its range before making generalizations. 



Tables III and IV show the catches made in gill nets. 

 These are arranged so as to compare the four chief habitats in 

 the lake: (1) the shores of the islands, which were rocky, 

 pebbly, or sandy, and at the heads of bays sometimes a little 

 marshy; (2) the marshy, rush grown shores around the margin 

 of the lake; (3) the mouths of streams; (4) the deep water. 

 The catch per hour in gill nets is a good means of determining 



