Seascape and life-history traits do not predict self-recruitment in a coral reef fish.

Marcela Herrera,Gerrit B Nanninga,Simon R Thorrold,Serge Planes,Pablo Saenz-Agudelo,Glenn R Almany,Michael L Berumen,Geoffrey P Jones

doi:10.1098/rsbl.2016.0309

Abstract

The persistence and resilience of many coral reef species are dependent on rates of connectivity among sub-populations. However, despite increasing research efforts, the spatial scale of larval dispersal remains unpredictable for most marine metapopulations. Here, we assess patterns of larval dispersal in the angelfish Centropyge bicolor in Kimbe Bay, Papua New Guinea, using parentage and sibling reconstruction analyses based on 23 microsatellite DNA loci. We found that, contrary to previous findings in this system, self-recruitment (SR) was virtually absent at both the reef (0.4–0.5% at 0.15 km2) and the lagoon scale (0.6–0.8% at approx. 700 km2). While approximately 25% of the collected juveniles were identified as potential siblings, the majority of sibling pairs were sampled from separate reefs. Integrating our findings with earlier research from the same system suggests that geographical setting and life-history traits alone are not suitable predictors of SR and that high levels of localized recruitment are not universal in coral reef fishes.

Highlights

Connectivity in marine metapopulations is predominantly driven by the exchange of pelagic larvae among relatively sedentary adult populations
Sibship analysis performed in COLONY confidently identified three pairs as full siblings; a further 42 pairs and one triplet had similar likelihoods of being full or half siblings
–0.010 5 distance class size species of coral reef fishes [8,10,11], we found a near lack of SR and no apparent spatial structure in recruitment

Summary

Introduction

Connectivity in marine metapopulations is predominantly driven by the exchange of pelagic larvae among relatively sedentary adult populations. The relative importance of dispersal versus local retention of larvae has received considerable attention Numerous studies have produced estimates of self-recruitment (SR; the proportion of all sampled recruits at a given location that had been locally produced) from different systems in a variety of coral reef fishes (figure 1). These studies have shown that levels of SR can be highly variable temporally both within a species [4,5] and among closely related species with similar life-history traits. SR rates may be remarkably consistent over time within species [8,9] and even among species with very different life-history characteristics [10]

Methods

Results

Conclusion