Jeffrey S. Shima

RECRUITMENT OF A CORAL REEF FISH: ROLES OF SETTLEMENT, HABITAT, AND POSTSETTLEMENT LOSSES

Multiple processes typically influence patterns of abundance. Despite this widely accepted view, many studies continue to approach ecological questions from a single- factor, or, at most, a two-factor perspective. Here, I evaluate the consequences of consid- ering, separately and jointly, the effects of three factors (larval settlement, reef resources, and postsettlement losses) on spatial patterns of abundance of a marine reef fish, the six bar wrasse (Thalassoma hardwicke). Using correlational methods commonly employed in single-factor studies, I show that local patterns of abundance of juvenile wrasse could be attributed entirely to either (1) patterns of abundance of settlement habitat, or (2) patterns of larval settlement. This result occurred because habitat and presumed larval delivery covaried in space. I manipulated abundance of settlement habitat in a field experiment to uncouple this covariation and found subsequent settlement to be simultaneously influenced by both factors. However, joint effects of habitat and settlement failed to account for patterns of abundance of juvenile wrasse without also considering a third factor-postsettlement losses-which were density-dependent and substantially modified patterns of settlement. These results illustrate (1) how multifactorial explanations may be falsely refuted when incomplete sets of multiple factors are considered, and (2) how single-factor explanations may misrepresent underlying multifactorial causation of ecological patterns. Uncovering the interactive role of multiple factors in determining ecological patterns of interest requires a shift from single-factor approaches to more pluralistic perspectives.

Ontogenetic changes in habitat selection during settlement in a coral reef fish: ecological determinants and sensory mechanisms

The behavior of marine larvae during and after settlement can help shape the distribution and abundance of benthic juveniles and therefore the intensity of ecological interactions on reefs. Several laboratory choice-chamber experiments were conducted to explore sensory capabilities and behavioral responses to ecological stimuli to better understand habitat selection by “pre-metamorphic” (larval) and “post-metamorphic” (juvenile) stages of a coral reef fish (Thalassoma hardwicke). T. hardwicke larvae were attracted to benthic macroalgae (Turbinaria ornata and Sargassum mangarevasae), while slightly older post-metamorphosed juveniles chose to occupy live coral colonies (Pocillopora damicornis). Habitat choices of larvae were primarily based upon visual cues and were not influenced by the presence of older conspecifics. In contrast, juveniles selected live coral colonies and preferred those occupied by older conspecifics; choices made by juveniles were based upon both visual and olfactory cues from conspecifics. Overall, the laboratory experiments suggest that early life-history stages of T. hardwicke use a range of sensory modalities that vary through ontogeny, to effectively detect and possibly discriminate among different microhabitats for settlement and later occupation. Habitat selection, based upon cues provided by environmental features and/or by conspecifics, might have important consequences for subsequent competitive interactions.

Guard crabs alleviate deleterious effects of vermetid snails on a branching coral

Stony corals provide important structural habitat for microbes, invertebrates, and fishes, which in some cases has led to the evolution of beneficial interactions that may protect corals from environmental factors such as thermal stress, nutrient limitation, competitors, or predators. For example, guard crabs (Trapezia spp.) protect corals (Pocillopora sp.) from attacks by crown-of-thorn seastar and sedimentation. Here, a field experiment demonstrates that guard crabs (Trapezia serenei) also ameliorate the strong negative effects of the giant vermetid (Dendropoma maximum) on growth of Pocillopora. This experiment highlights the importance of this crab-coral mutualism: guard crabs facilitate the growth of corals in stressful environments (e.g., where vermetids are abundant), thereby preserving the ecological goods and services (e.g., food and shelter) that these corals may provide to other reef-associated species.

The vermetid gastropod Dendropoma maximum reduces coral growth and survival.

Coral reefs are one of the most diverse systems on the planet; yet, only a small fraction of coral reef species have attracted scientific study. Here, we document strong deleterious effects of an often overlooked species—the vermetid gastropod, Dendropoma maximum—on growth and survival of reef-building corals. Our surveys of vermetids on Moorea (French Polynesia) revealed a negative correlation between the density of vermetids and the per cent cover of live coral. Furthermore, the incidence of flattened coral growth forms was associated with the presence of vermetids. We transplanted and followed the fates of focal colonies of four species of corals on natural reefs where we also manipulated presence/absence of vermetids. Vermetids reduced skeletal growth of focal corals by up to 81 per cent and survival by up to 52 per cent. Susceptibility to vermetids varied among coral species, suggesting that vermetids could shift coral community composition. Our work highlights the potential importance of a poorly studied gastropod to coral dynamics.

Life history and matrix heterogeneity interact to shape metapopulation connectivity in spatially structured environments

Metapopulation models have historically treated a landscape as a collection of habitat patches separated by a matrix of uniformly unsuitable habitat. This perspective is still apparent in many studies of marine metapopulations, in which recruitment variation is generally assumed to be primarily the result of variability in ocean currents and interactions with disperser behavior, with little consideration of spatial structure that can affect disperser viability. We use a simple model of dispersal of marine larvae to demonstrate how heterogeneity in dispersal habitat (i.e., the matrix) can generate substantial spatial variation in recruitment. Furthermore, we show how this heterogeneity can interact with larval life-history variation to create alternative patterns of source-sink dynamics. Finally, we place our results in the context of spatially structured matrix population models, and we propose the damping ratio of the connectivity matrix as a general and novel measure of landscape connectivity that may provide conceptual unification to the fields of metapopulation biology and landscape ecology.

Algal and invertebrate bioindicators detect sewage effluent along the coast of Titahi Bay, Wellington, New Zealand

Abstract Sewage effluent contains both dissolved and particulate matter that can affect coastal ecosystems and cascade through food webs. We used δ15N and δ13C ratios to explore performance of a kelp (Carpophyllum maschalocarpum) and two invertebrates (a grazing isopod, Amphoroidea media, and a filter-feeding crab, Petrolisthes elongatus) as bioindicators along a gradient of sewage exposure. These species vary in trophic status and, consequently, are likely to assimilate sewage constituents differently into food webs. δ15N values in C. maschalocarpum and A. media were highest near the sewage outfall and decreased with increasing distance from the outfall. δ13C values in these two species did not appear to be reliable predictors of sewage concentrations in seawater. In contrast, δ13C ratios for P. elongatus were consistently depleted at sites nearest the sewage outfall, indicating the consumption of 13C-depleted sewage particulates. These results suggest there is value in multiple-isotope and multi-species bioindicator approaches both for detecting sewage dispersal patterns and understanding the incorporation of sewage-derived nutrients into food webs.

A Framework for Assessing Impacts of Marine Protected Areas in Moorea (French Polynesia)

ABSTRACT Marine Protected Areas (MPAs) have been promoted as effective management tools to protect biodiversity at local and global scales, but there remains considerable scientific uncertainty about effects of MPAs on species abundances and biodiversity. Commonly used assessment designs typically fail to provide irrefutable evidence of positive effects. In contrast, Before-After-Control-Impact (BACI) designs potentially remedy many of these problems by explicitly dealing with both spatial and temporal variation. Here, we document the historical context of implementation and the scientific assessment of MPAs recently established at eight sites around the island of Moorea, French Polyne-sia. In 2004, we designed and implemented a monitoring plan that uses a BACI-Paired Series (BACIPS) design to quantify the effect of the MPAs. Twice per year, we monitor fish, corals, and other benthic invertebrates at 13 sites (eight within MPAs and five outside MPAs) around Moorea, in three distinct reef habitats (fringing, barrier reef, and outer slope). We present statistical analyses of data collected during five surveys (July 2004 to July 2006), before the initiation of enforcement. We also assessed the potential of our program to detect future responses to the established MPA network. Our estimates of biomass for five categories of fishes (Acanthuridae, Chaetodontidae, Serranidae, Scaridae, and fisheries target species) within MPA sites generally track estimates in paired Control sites through time. Estimated statistical power to detect MPA effects (a 192% biomass increase within the MPA) was high at the MPA network scale but varied among taxonomic categories and reef habitats: power was high on the reef outer slope and lower in the lagoon, and generally high for acanthurids and chaetodontids. It did not vary significantly between sites. We discuss limitations of our approach (shared by all MPA assessments to date) and describe solutions and unique opportunities to redress these limitations in French Polynesia.

Inferring dispersal and migrations from incomplete geochemical baselines: analysis of population structure using Bayesian infinite mixture models

Summary Geochemical and stable isotope tags are often used to attribute individual animals in a sample of mixed origins to distinct sources, be it spawning, overwintering or foraging habitats. In order for individuals to be uniquely classified to one source, modelling approaches generally assume that all potential sources have been characterized in terms of their geochemical signature. This assumption is rarely met in applications of geochemistry in environments where species distributions and spawning grounds are poorly known; statistical methods that can accommodate this problem are therefore essential. We develop nonparametric Bayesian mixture models for geochemical signatures that estimate the most likely number of sources represented in a mixed sample, both in the absence and presence of baseline data. We then use a marginal clustering framework to evaluate the probability that a fish comes from a particular source. Using both simulations and a previously analysed data set, we illustrate the method and highlight the potential merits and difficulties. These examples reveal how our interpretations of geochemistry data sets can change when potentially un-sampled sources are taken into account.

Geographic and sex-specific variation in growth of yellow-eyed mullet, Aldrichetta forsteri, from estuaries around New Zealand

Abstract Survival and reproductive rates in fish are often a function of body size. Consequently, spatial‐ and sex‐specific variation in somatic growth rates can have important consequences for population growth and resilience. We used otolith‐based approaches to estimate geographic‐ and sex‐specific growth rates of yellow‐eyed mullet (Aldrichetta forsteri) collected from 14 estuaries and harbours around New Zealand. Aldrichetta forsteri is an abundant and dominant component of New Zealands estuarine fish fauna. We extracted otoliths from 511 fish, validated daily and annual increments, and prepared transverse thin sections of otoliths to determine age. “Size‐at‐age” relationships were estimated using both linear‐ and non‐linear (von Bertalanffy) growth models, and model performance was evaluated using Akaikes Information Criterion. Because growth rates of sampled fish were best approximated by linear functions, we used ANCOVA to test the null hypothesis that growth rates of A. forsteri were homogeneous between sexes and among geographic locations around New Zealand. Our analyses suggest heterogeneous growth rates between sexes and among locations. Interestingly, relative growth rates between sexes appeared to vary across separate latitudinal gradients for North Island and South Island. Within each island (but not across islands), female A. forsteri generally grew faster than males at the lowest latitudes; relative growth rates of females declined progressively below males with increasing latitude.

Consequences of variable larval dispersal pathways and resulting phenotypic mixtures to the dynamics of marine metapopulations

Larval dispersal can connect distant subpopulations, with important implications for marine population dynamics and persistence, biodiversity conservation and fisheries management. However, different dispersal pathways may affect the final phenotypes, and thus the performance and fitness of individuals that settle into subpopulations. Using otolith microchemical signatures that are indicative of ‘dispersive’ larvae (oceanic signatures) and ‘non-dispersive’ larvae (coastal signatures), we explore the population-level consequences of dispersal-induced variability in phenotypic mixtures for the common triplefin (a small reef fish). We evaluate lipid concentration and otolith microstructure and find that ‘non-dispersive’ larvae (i) have greater and less variable lipid reserves at settlement (and this variability attenuates at a slower rate), (ii) grow faster after settlement, and (iii) experience similar carry-over benefits of lipid reserves on post-settlement growth relative to ‘dispersive’ larvae. We then explore the consequences of phenotypic mixtures in a metapopulation model with two identical subpopulations replenished by variable contributions of ‘dispersive’ and ‘non-dispersive’ larvae and find that the resulting phenotypic mixtures can have profound effects on the size of the metapopulation. We show that, depending upon the patterns of connectivity, phenotypic mixtures can lead to larger metapopulations, suggesting dispersal-induced demographic heterogeneity may facilitate metapopulation persistence.