16 U. S. BUREAU OF FISHERIES. 



ment. Certain lots have been separated for selective breeding. The 

 terrapins in these lots were selected for rapid growth, size, and 

 general healthy and vigorous appearance. The third generation of 

 domestic terrapins probably will result during the next breeding 

 season from this series of experiments. The results attained show 

 beyond a doubt that it is entirely practical to grow diamond-back 

 terrapins in conhnement for commercial purposes, and, indeed, this 

 appears to be the only hope of maintaining a supply of this animal, 

 which is nearing extinction. 



ECOLOGICAL AND OCEANOGRAPHIC STUDIES. 



CONTROL OF MOSQUITOES BY MEANS OF FISH. 



Investigations of fishes in relation to mosquito control were con- 

 ducted at Augusta, Ga., during the mosquito-breeding season of 1922 

 by Samuel F. Hildebrancl, ichthyologist, working in cooperation 

 with the United States Public Health Service. Particular attention 

 was devoted to a study of the relationship that various plants bear 

 to mosquito control by the use of fish, A single plant — the duckweed 

 Spirodela — was found to prevent mosquito production in a measure. 

 All other plants used in the experiments appeared to favor mos- 

 quito breeding. The extent to which plants favored mosquito breed- 

 ing and to what extent they formed " barriers " between the fish and 

 the wiggle-tails was arrived at through the creation of fishless areas 

 in ponds supporting various types of vegetation in which mosquitoes 

 were permitted to breed unmolested for a time. Later such areas 

 were restocked and the results were checked. 



Extensive studies of sex ratios in Gambusia (the most important 

 of the fishes used for mosquito control) were made, from which it 

 was determined that the males are proportionately fewest in August, 

 when among the specimens examined only 1 male occurred among 

 about 8 females. In October the males had become much more 

 numerous and the proportion was about 1 male to 2 females. The 

 conclusion is that the male is the weaker sex, and in August, just 

 before the early young of the season become sexually mature, so many 

 of the males have died or have been destroyed by enemies that the 

 disparity is very great. In October, when many of the young of 

 the previous spring and summer have become sexually mature, the 

 young males greatly increase the proportionate number of that sex. 

 Microscopic studies of the sexual organs of a limited number of 

 young fish showed a sex ratio of about 1 to 1. 



It is well known that in the northern part of its range Gambusia 

 spawns only during the summer, but it had previously been noticed 

 that gravid females may be taken at Key West at all seasons of the 

 year. Consequently the^ director of the Key West station, Mr. Gins- 

 berg, was requested to confine females in the aquarium and* make 

 careful observations relative to the period and frequency of spawn- 

 ing. These observations show that at Key West Gambusia spawn 

 the year round at more or less regular intervals of four weeks. 



Extensive experiments relative to the proper handling of Gam- 

 busia in confinement and shipment were conducted. It was shown 

 that a larger proportion of the fish survived when approximately 3 

 inches of water was placed in a vessel of a given diameter than if 

 « to 10 inches of water were used in the same container. 



