Exploration of the "larval pool": development and ground-truthing of a larval transport model off leeward Hawai'i.

Johanna L.K Wren,Donald R Kobayashi

doi:10.7717/peerj.1636

Abstract

Most adult reef fish show site fidelity thus dispersal is limited to the mobile larval stage of the fish, and effective management of such species requires an understanding of the patterns of larval dispersal. In this study, we assess larval reef fish distributions in the waters west of the Big Island of Hawai‘i using both in situ and model data. Catches from Cobb midwater trawls off west Hawai‘i show that reef fish larvae are most numerous in offshore waters deeper than 3,000 m and consist largely of pre-settlement Pomacanthids, Acanthurids and Chaetodontids. Utilizing a Lagrangian larval dispersal model, we were able to replicate the observed shore fish distributions from the trawl data and we identified the 100 m depth strata as the most likely depth of occupancy. Additionally, our model showed that for larval shore fish with a pelagic larval duration longer than 40 days there was no significant change in settlement success in our model. By creating a general additive model (GAM) incorporating lunar phase and angle we were able to explain 67.5% of the variance between modeled and in situ Acanthurid abundances. We took steps towards creating a predictive larval distribution model that will greatly aid in understanding the spatiotemporal nature of the larval pool in west Hawai‘i, and the dispersal of larvae throughout the Hawaiian archipelago.

Highlights

Population connectivity, defined as the exchange of individuals among geographically separate sub-populations (Cowen et al, 2007; Fogarty & Botsford, 2007; Pineda, Hare & Sponaugle, 2007; Cowen & Sponaugle, 2009) is widely recognized as important in effective marine conservation (e.g., Jones, Srinivasan & Almany, 2007; Toonen et al, 2011; Treml et al, 2012)
We used in situ Acanthurid abundances from trawl catches and compared them to modeled abundances from the forecast simulation parameterized after Zebrasoma flavescens
A key linkage between larval abundance and larval catch involves efficiency of the sampling gear (e.g., Clarke, 1983), the general additive model (GAM) was parameterized with variables that have been known to impact visual avoidance such as moon phase and moon angle

Summary

Introduction

Population connectivity, defined as the exchange of individuals among geographically separate sub-populations (Cowen et al, 2007; Fogarty & Botsford, 2007; Pineda, Hare & Sponaugle, 2007; Cowen & Sponaugle, 2009) is widely recognized as important in effective marine conservation (e.g., Jones, Srinivasan & Almany, 2007; Toonen et al, 2011; Treml et al, 2012). Management efforts focus on regulating the adult fish, but an effective management for healthy ecosystems and sustainable fisheries requires understanding of adult habitats and spawning areas, and larval dispersal and population connectivity (Gaines et al, 2010; Toonen et al, 2011). The importance of the larval life stage of marine fish and invertebrates in understanding population dynamics has long been recognized (Thorson, 1950; Knight-Jones, 1953; Scheltema, 1971) yet little is still known about dispersal of marine larvae, and factors affecting dispersal and near shore retention (Levin, 2006; Hellberg, 2009; Cowen & Sponaugle, 2009). An important driver of horizontal dispersal is the vertical strata occupied by these early life-history stages (Olivar & Sabate, 1997; Muhling & Beckley, 2007) since abiotic and biotic gradient vary greater on the vertical scale than the horizontal scale

Methods

Results

Conclusion