HANKIN; A MULTISTAGE RECRUITMENT PROCESS 



Table 7. — Total population numbers and, in parentheses, total biomass in grams for single species tguppy) populations at week 



of enumeration during Phase I, weeks 0-36. 



mass for each population was estimated from the 

 slope and intercept of a regression of the reciprocal 

 of population biomass at week t + 2 against the 

 reciprocal of biomass at week t as: maximum 

 biomass — Sj^ax = '1 ~ slope )/intercept. Esti- 

 mates of maximum biomass ranged from 27.8 to 

 37.6 g (Table 8) with a mean estimate of 31.8 g. 



The nearly uniform changes in total biomass 

 throughout Phase I, despite widely divergent 

 numerical growth patterns, implied that growth 

 was strongly density-dependent. Density-depen- 

 dence of growth was seen most dramatically in 

 the mean weights of large adult females. Mean 

 weights of adult females in size categories A^ and 



Ag were inversely related to mean population 

 numbers during weeks 0-36 (Figure 5). 



NUMERICAL DYNAMICS. — WTiile the guppy 

 populations appeared to approach a common maxi- 

 mum supportable biomass, no such commonality 

 was evident in total population numbers. Numeri- 

 cally, populations grew in erratic and independent 

 fashion with no clear indications of asymptotic 

 behavior. Numerical growth was steadily positive 

 \s\ only a single population (population 1). In all 

 other populations patterns of numerical increase 

 were unanticipated and often consisted of discrete 

 pulses of increase, followed by periods of static 



563 



