Early post-settlement mortality in a coral reef fish and its effect on local population size

The relative importance of recruitment and post-recruitment factors in controlling population fluctuations and determining population sizes in open marine systems is still poorly understood. To address this issue, I conducted an experiment investigating whether very early post-recruitment mortality of a coral reef fish, Thalassoma bifasciatum, is dependent on recruitment density. To quantify the extent to which local populations are likely to be controlled primarily by recruitment or by post-settlement density-dependent mortality, I also measured natural mortality rates and recruitment variation at sites around the island of St. Croix, U.S. Virgin Islands. I manipulated densities of 1-3-d post-settlement T. bifasciatum on large natural patch reefs in Tague Bay, St. Croix, U.S. Virgin Islands. Manipulated recruits were followed for similar to 30 d. The loss rate over the initial 24-h period following placement of recruits on the reef was positively related to the density at the beginning of that period. After this period, initial recruit density had no effects on mortality. Natural monthly mortality rates approximating the monthly mortality monitored during the experiment were measured for naturally recruited fish at sites around the island. Mortality was positively correlated with recruitment density. The effect was strong enough that across sites the relationship between recruit density and juvenile density the following month was not proportional; that is, juvenile density tended to reach an asymptote with increasing recruitment. The relationship between total recruitment per season at a site and adult abundance at the end of the season was similarly asymptotic, indicating the persistence of a density-dependent effect at the scale of the island. Mortality rates around the island were extremely variable and density independent at densities below similar to 1 recruit/m(2). Above that level, mortality rates were higher and tended to increase with further increases in recruitment density. Around St. Croix, three out of eight sites consistently received recruitment above this "threshold," while five sites never experienced recruitment above that level during the study. Because spatial patterns of recruitment around St. Croix were consistent over time, some local populations may be controlled mainly by recruitment and density-independent mortality at all times, while others may be chronically affected by post-settlement density dependence. This study gives support to the idea that recruitment limitation and density dependence are not mutually exclusive. Rather, local populations must be characterized by the degree of recruitment, the resultant population density, and the subsequent degree of density-dependent regulation.