THE MATING PATTERNS OF FEMALES WHEN TERRITORIAL DIFFERENCES AMONG MALES ARE REDUCED: A TEST IN THE POLYGYNOUS BEAUGREGORY DAMSELFISH

Abstract

Abstract Using the Caribbean beaugregory damselfish (Stegastes leucostictus) we tested whether patterns of male reproduction could be modified by reducing differences among males, in this case by giving males identical artificial breeding sites. Previous studies have shown that very few males using the natural small rubble habitat reproduced and variations in male reproductive success were based mainly on the differences in the quality of their natural breeding sites. By providing males with identical artificial breeding sites, we tested whether females would be less likely to confine their reproduction to only a few of the available males. We examined male reproductive patterns by first simulating females choosing males (1) 'randomly,' (i.e. the number of males that receive eggs is based on each female randomly choosing a mate from a group of males), or (2) 'exclusively' (i.e. only one female mates with one male on a given day), or (3) 'highly selectively' (i.e. this is the typical polygynous pattern in which many females select the same few males). We tested these simulations against the actual daily amounts of eggs deposited and the number of different males that received eggs in each of 4 summers. Although the regression trend line from the daily patterns did resemble the Random Mating Pattern, egg clutches were more widely dispersed than random but less dispersed than the Exclusive Mating Pattern. That is, with more females mating on a particular day, more males received eggs. We speculated that this might have resulted from females aggressively excluding other females from mating with the same male on a given day. In contrast, males using the variable quality natural sites rarely mated and those that did receive eggs, received them in larger amounts than males using the artificial sites. Thus, for natural sites, more females seemed to be mating with the same few males. Perhaps for these natural sites any intra-female aggression may have been ineffective when so few superior spawning sites were available. For males using the artificial sites, the total amount of eggs received over a two-month period was nonrandom and resembled the 'highly selected pattern.' This resulted from some males receiving egg clutches on more days. Other traits, besides breeding site structure, may have been important in causing different females to select the same males on subsequent days.