HANKIN: A MULTISTAGE RECRUITMENT PROCESS 



becomes large, shift to an adult-dominated age 

 structure, and reappearance of young individuals 

 when adult density declines through mortality to 

 levels at which fry survival once more occurs. In 

 those few instances in which refuge habitats have 

 been provided (Silliman 1948, 1968; Silliman and 

 Outsell 1958), the results have demonstrated that 

 roughly stable populations, of greater numerical 

 size, with finely graded age (or size) structure, 

 may be maintained for apparently indefinite peri- 

 ods. Presumably, the stability of these popula- 

 tions reflects the decreased period of time during 

 which fry are vulnerable to cannibalism. 



In previous guppy population studies the weight 

 of the mature female stock has not been recorded 

 and hence it is impossible to attempt an adequate 

 quantitative examination of stock-recruitment 

 relations which may have been responsible for 

 observed dynamics (e.g., Gulland 1962 based on 

 previous studies). The experiments described on 

 the following pages were specifically designed 

 to allow quantitative assessment of the stock- 

 recruitment relation. Refuge areas were provided 

 and fry successfully entering these areas were 

 considered as rough equals of recruited fish. 

 Although females do not release broods synchro- 

 nously (broods are delivered continuously with 

 respect to the entire population), data were col- 

 lected at discrete biweekly intervals. Net numeri- 

 cal change in a sampling interval could thus be 

 related to adult reproductive potential and adult 

 predator density at the beginning of an interval in 

 a fashion analogous to that which might be 

 attempted in analysis of the simple discrete-time 

 models. A simple difference in refuge design (as 

 compared with Silliman's earlier work) involving 

 spacing between glass rods in a refuge fence, 

 however, created unexpected patterns of numeri- 

 cal increase. These patterns were not anticipated 

 and could not be explained on the basis of simple 

 stock-recruitment theory. Analysis ultimately 

 showed the presence of a complex mechanism 

 involving both adult and immature population 

 components, an adult-juvenile stock-recruitment 

 relation. 



METHODS AND MATERIALS 



Experiments were performed in a 3 m x 3.7 m 

 room insulated on three walls, including two out- 

 side walls, from floor to ceiling. A small electric 

 floor heater maintained room temperature at 

 approximately 22° C, about 3° C above ambient 



winter temperature supplied by a propane heating 

 unit in an adjoining room. An air-conditioner in 

 the same adjoining room prevented summer tem- 

 peratures from exceeding 26° C. Experimental 

 aquaria were located along the three insulated 

 walls. 



Experimental Environments 



Twelve aquaria, of dimensions 31 cm x 62 cm x 

 41 cm, each holding about 80 1 of water, served as 

 experimental units. Each aquarium was equipped 

 with a 75 W thermostat-controlled aquarium 

 heater, about 1 cm deep layer of 3-5 mm gravel, a 

 large inside-type charcoal -glass wool filter, a full 

 hood reflector with two 15 W showcase bulbs, a 

 thermometer, and a refuge area (Figure 1). Refuge 

 areas were enclosed by a fence consisting of two 

 sheets of solid glass rods, 3 mm in diameter, 

 spaced (initially) 5 mm apart on centers, fitted in 

 Plexiglas^ frames glued at right angles. The 



^Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



-heater 



41 



cm 



crn inside 

 filter 



Figure l. — Schematic of the experimental environments for 

 guppies and southern platyfish. 



557 



