196

 commencement of the wet season by several months,

 suggesting that reproduction in this species so well

 adapted to south Florida's wet season-dry season regime may

 be keyed to some factor other than rain--perhaps

 photoperiod, temperature, or crowding. Recruitment, or

 effective survival of fingerlings, must be a function of

 water surface area. Reduced population density reduces the

 probability of encounter between fingerlings and potential

 predators, which, in the case of Mosquitofish, includes

 their parents.

 In the Allen curve (Figure 47) the sum of the areas

 under the curve from July 9 to August 6, 1974, and from

September 17 to October 22 provides a crude estimate of

production for the period from July 9 through October 22

(105 days, or 3.45 mos). A second recruitment is evident

at the point where the number of fish begins to increase

and the average size of fish declines. The increase in the

average number of fish that occurred simultaneously with

the continued increase in the average size of fish

beginning around October 22 suggested that concentration as

well as production had been taking place in October;

therefore, a correction was made in the curve from October

1 through October 22, shown by a dashed line. Since

production is probably minimal during the dry season, no

attempt was made to extend the line beyond October.

 According to the Allen curve, Mosquitofish production

for the 1974 wet season was approximately 0.4 g dry wt/sq-