LIFE-HISTORY EVOLUTION IN GUPPIES: 2. REPEATABILITY OF HELD OBSERVATIONS AND THE EFFECTS OF SEASON ON LIFE HISTORIES.

Natural populations of guppies that co-occur with the pike cichlid Crenicichla alta and associated predators mature at smaller body sizes, produce more and smaller offspring per litter reproduce more frequently, and have higher reproductive allotments (weight of developing embryos/total body weight) than guppies that co-occur with just the killifish Rivulus harti (Reznick and Endler, 1982). I here consider three forms of repeatability in these life-history patterns: i) among replicate samples collected on the same day from the same locality, ii) between Crenicichla and Rivulus communities among a new series of localities, and iii) among a smaller series of Crenicichla and Rivulus localities sampled in two wet and two dry seasons. In the analysis of replicate collections from two localities, seven of eight statistical comparisons revealed no significant difference. The usual methodology for estimating these variables therefore accurately represents guppy life-history patterns at a given locality. Differences among guppies from Rivulus and Crenicichla localities, covering a wider geographical area than considered by Reznick and Endler (1982), were virtually identical to the previous comparison. Wet-season samples were associated with significant decreases in reproductive allotment and fecundity and significant increases in the size of mature males and the minimum size of reproducing females. Differences between guppies from Rivulus and Crenicichla localities persisted across all samples and were consistent with all other observations, although they tended to be smaller during the wet season. Discriminant analyses on female reproductive traits showed that fecundity and offspring size made strong, independent contributions to discriminating between guppies from the two types of localities. The contribution from reproductive allotment was considerably smaller. There was more overlap between predator treatments during the wet season. Only 8.5% of the individuals were misclassified during the dry season, but 19.5% were misclassified during the wet season.