105] STUDY OF SOUTHERN WISCONSIN FISHES-CAEN 105 



except Sayomia albipes larvae. By actual count 376 larvae were obtained, 

 of which 319 were still alive when removed from the stomach after the 

 fish had been out of water over three hours. While Sayomia forms a very 

 stable article of diet, the larvae seem to be more resistent to the digestive 

 juices of the cisco than any food consumed, as in a vast majority of the 

 stomachs examined the larvae have shown great activity after removal, 

 a condition found in no other item of food eaten by the fish. 



MIGRATIONS 



It has been said that the cisco is a bottom inhabitant and that it must 

 be regarded as a deep water fish. During that part of the year when the 

 oxygen conditions permit, Leucichthys artedi remains normally in the deep 

 water, spending most of the time within a meter or two of the bottom. 

 This fact has been demonstrated again and again, year after year by the 

 use of gill nets set at different depths. As I have said, this differs radically 

 from the habitat chosen by the Lake Michigan cisco as reported by Ward 

 (1897). However, with the formation of the thermocline and the accom- 

 panying increase in the area of water deficient in oxygen, the cisco is forced 

 to leave the bottom waters of the deeper parts of the lake, and assume a 

 position ever higher in a vertical scale. As the thermocline climbs upward 

 the cisco comes up with it and assumes a position in relation to the thermo- 

 cline which may be described as directly above it. Therefore, if one takes 

 the temperature of the water at different depths and thus obtains a fairly 

 accurate idea of the location of the thermocline, one can set gill nets just 

 above this depth and catch ciscos all summer. This was discovered in 1916 

 when the writer wanted the fish for an examination of the summer food. 

 Nets were set in the deep water where the fish were caught in abundance 

 during the previous winter, and caught nothing. After repeated failures a 

 series of eight nets was set, one above the other. The net which was at 10 

 meters yielded 11 ciscos; the one at 9 yielded none and the one at 11 

 yielded 2 in the upper foot of mesh; none was caught in any other net. 

 An analysis of the water at that time showed that the thermocline stood at 

 11 meters. This idea has been followed ever since, and has always yielded 

 fish. I have set 43 nets below the thermocline, in 9 different lakes, and have 

 never in a single instance caught a cisco. From these facts one must con- 

 clude that, while the cisco is essentially a deep water fish, it is driven from 

 its chosen habitat by uncongenial gas conditions, so that the species 

 really has a seasonal migratory rhythm in a vertical plane, correlated with 

 and determined by the shifting of the thermocline. 



This action of the thermocline in forcing the cisco up from the bottom 

 has an exceedingly important effect upon the fish concerned. Being an 

 inhabitant of the deep water, it is a cold water species. (Fig. 23.) The 

 early summer influence of the thermocline has little effect on the cisco as 



