FISHERY BULLETIN: VOL. 73. NO. 4 



Week 

 3-week period 



Figure 5.— Course of competing populations, test pair. Solid line, guppy; broken line, swordtail. Numbers indicate 



target exploitation rates. 



(week 30), under the mistaken impression that the 

 greater biomass then achieved by the swordtail 

 indicated a greater productive capability. Exploi- 

 tation produced a rapid decline in the swordtail to 

 a low population level (Figure 5). Cessation of 

 swordtail exploitation at week 59 and initiation of 

 guppy exploitation at week 62 did not lead to 

 recovery of the swordtail, in spite of a drastic 

 decline in guppy abundance (Figure 5). 



By week 74, it became apparent that the sword- 

 tail would require a lengthy period for recovery, 

 even if guppy abundance were further reduced. To 

 accelerate the study of exploitation, the popula- 

 tions v/ere reconstructed during weeks 74-85, us- 

 ing fish from exploited populations that had been 

 placed in a reserve tank. After reconstruction, the 

 populations approximated fairly closely their 

 number and weight at the time exploitation was 

 started (compare week 85 with week 30 in Figure 

 5). Exploitation rates after week 85 were adjusted 

 to keep both the guppy and swordtail at productive 

 levels while trying as wide a range of pairs of 

 exploitation rates as possible. 



RECRUITMENT RELATIONS 



Juvenile fish were counted both when removed 

 from and returned to the adult tanks; it was thus 

 possible to obtain a measure of recruitment. 

 Numbers were converted to weights by use of 

 factors (mean w^eights per fish) based on 

 weighings of the juveniles: guppy, 0.0656 g based 

 on 1,417 fish in 126 weighings spread over 199 wk; 

 swordtail 0.0678 g, 337, 61, 196, respectively. At the 

 beginning of the experiments, when few fish were 

 in the nursery tanks, it was possible to distinguish 

 individual groups of recruits by size, count them, 

 and thereby estimate the "reproductive lag" from 

 birth to recruitment. The lag was found to be 

 approximately one brood interval (guppy, 3 wk; 

 swordtail, 4 wk) for each species. In constructing 

 the stock-recruitment relations, the recruitment 

 for each brood interval was compared with the 

 mean stock in the adult tanks during the preced- 

 ing brood interval. During periods of exploitation, 

 the catch was subtracted from the stock at the 

 beginning of the interval. 



878 



