Ecology of Fishes on Coral Reefs

C:/ITOOLS/WMS/CUP-NEW/5644691/WORKINGFOLDER/AROM/9781107089181PRE.3D

iii

[1–14] 4.12.2014 6:17PM

Ecology of Fishes on Coral Reefs EDITED BY

Camilo Mora Department of Geography, University of Hawai‘i at Manoa, USA

C:/ITOOLS/WMS/CUP-NEW/5644691/WORKINGFOLDER/AROM/9781107089181PRE.3D

iv

[1–14] 4.12.2014 6:17PM

University Printing House, Cambridge CB2 8BS, United Kingdom Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107089181 © Cambridge University Press 2015 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2015 Printed in the United Kingdom by [XX] A catalog record for this publication is available from the British Library Library of Congress Cataloging in Publication data Ecology of fishes on coral reefs / edited by Camilo Mora, Department of Geography, University of Hawaii at Manoa, USA. pages cm Includes bibliographical references. ISBN 978-1-107-08918-1 1. Coral reef fishes. I. Mora, Camilo. QL620.45.E26 2015 597.1770 89–dc23 2014043414 ISBN 978-1-107-08918-1 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

C:/ITOOLS/WMS/CUP-NEW/5644691/WORKINGFOLDER/AROM/9781107089181PRE.3D

v

[1–14] 4.12.2014 6:18PM

CONTENTS Preface Foreword Peter F. Sale List of contributors

PART I 1 2 3 4 5

BASIC ECOLOGY

Sensory biology and navigation behavior of reef fish larvae Jelle Atema, Gabriele Gerlach, and Claire B. Paris Mission impossible: unlocking the secrets of coral reef fish dispersal Geoffrey P. Jones Recruitment of coral reef fishes: linkages across stages Su Sponaugle Competition in reef fishes Graham E. Forrester Predation: piscivory and the ecology of coral reef fishes Mark A. Hixon

PART II PATTERNS AND PROCESSES The evolution of fishes on coral reefs: fossils, phylogenies, and functions David R. Bellwood, Christopher H.R. Goatley, Peter F. Cowman, and Orpha Bellwood 7 Phylogeography of coral reef fishes Jeff A. Eble, Brian W. Bowen, and Giacomo Bernardi 8 How many coral reef fish species are there? Cryptic diversity and the new molecular taxonomy Benjamin C. Victor 9 Large-scale patterns and processes in reef fish richness Camilo Mora 10 Patterns and processes in geographic range size in coral reef fishes Benjamin I. Ruttenberg and Sarah E. Lester

ix xi xiii

1 3

16

28

34

41

53

6

55

64

76

88

97

C:/ITOOLS/WMS/CUP-NEW/5644691/WORKINGFOLDER/AROM/9781107089181PRE.3D

vi

vi

[1–14] 4.12.2014 6:18PM

Contents

11 Patterns and processes in reef fish body size Michel Kulbicki, Valeriano Parravicini, and David Mouillot 12 Multi-scale patterns and processes in reef fish abundance M. Aaron MacNeil and Sean R. Connolly

PART III HUMAN FINGERPRINTS 13 Effects of climate change on coral reef fishes Morgan S. Pratchett, Shaun K. Wilson, and Philip L. Munday 14 Effects of fishing on the fishes and habitat of coral reefs Edward E. DeMartini and Jennifer E. Smith 15 Effects of sedimentation, eutrophication, and chemical pollution on coral reef fishes Amelia S. Wenger, Katharina E. Fabricius, Geoffrey P. Jones, and Jon E. Brodie 16 Impacts of invasive species on coral reef fishes Isabelle M. Côté and John F. Bruno 17 Cashing in on coral reefs: the implications of exporting reef fishes Yvonne Sadovy de Mitcheson and Xueying Yin

PART IV CONSERVATION 18 Resilience in reef fish communities Tim McClanahan 19 Phase shifts and coral reef fishes Nicholas A.J. Graham 20 Extinction risk in reef fishes Loren McClenachan 21 A perspective on the management of coral reef fisheries Alan M. Friedlander 22 Linkages between social systems and coral reefs Joshua E. Cinner and John N. Kittinger

PART V DEBATES AND PARADIGM SHIFTS 23 Is dispersal of larval reef fishes passive? Jeffrey M. Leis 24 Density dependence and independence and the population dynamics of coral reef fishes Nick Tolimieri

104

116

125 127

135

145

154

166

181 183

191

199

208

215

221 223

227

C:/ITOOLS/WMS/CUP-NEW/5641473/WORKINGFOLDER/AROM/9781107089181C34.3D

274

[274–282] 4.12.2014 11:38AM

© Simon Pittman, NOAA

34 Seascape ecology of fishes on coral reefs Simon J. Pittman and Andrew D. Olds Ecological study of coral reef fishes is shifting from a reef patch-centric approach to a broader scale seascape perspective using concepts and techniques from landscape ecology. This change has occurred in response to increasing evidence that many “reef” fishes move among multiple reef patches and also connect different patch types, such as seagrasses and mangroves, through multi-habitat movements. As such, the spatial distribution of patches across the seascape is ecologically relevant. Seascape ecology, the marine counterpart of terrestrial landscape ecology, is concerned with the causes and ecological consequences of spatial patterning in the marine environment. It draws on concepts and analytical techniques from terrestrial landscape ecology supported by advances in remote sensing and spatial statistics. In the past decade, seascape ecologists have quantified effects on reef fish populations and critical ecological processes from variation in reef geometry, such as the spatial arrangement of patches and the three-dimensional morphology of seafloor terrains. This chapter provides a rationale for adopting a seascape ecology perspective for the ecological study of fishes on coral reefs. It describes key concepts central to the implementation of seascape studies, such as how seafloor spatial structure is represented, the ecological significance of mosaics and terrains and the importance of connectivity and corridors in ecology and marine ecosystem-based management. Lastly, we offer a selection of priority research themes to help guide future research.

Ecology of Fishes on Coral Reefs, ed. C. Mora. Published by Cambridge University Press. © Cambridge University Press 2015.

C:/ITOOLS/WMS/CUP-NEW/5641473/WORKINGFOLDER/AROM/9781107089181C34.3D

275

[274–282] 4.12.2014 11:39AM

34 Seascape ecology of fishes on coral reefs

S

tudies of reef fish movements (home range, ontogenetic shifts, and spawning migrations) have shown that many fish move across multiple patches of coral reef and also spend considerable time using non-reef patch types (e.g. seagrasses, mangroves, and macroalgal beds) [635,1111,1824]. Not all species are multi-habitat users, but even reef “residents” may move among patches of coral reef to forage, shelter, spawn, or relocate their home range [65,1841] (Figure 34.1). As such, coral reefs exist as structurally complex patches that are ecologically connected, directly and indirectly, to a broader mosaic of different patch types [993,1870,1924]. A mosaic of patch types provides structurally different resources that allow fishes to optimize trade-offs between maximizing growth and minimizing predation [603,1338]. Where the spatial configuration of tropical seascapes provides optimal habitat requirements and connectivity for fishes, geographical hotspots of productivity and diversity will occur [1972]. Geographically, coral reefs exist within a wide range of seascape types, where the structure of the surroundings, i.e. the seascape context, can profoundly influence species distributions and ecological processes on coral reefs. For instance, the composition, shape and spatial arrangement of patches including proximity to seagrasses, mangroves, land, deep water, and other coral reefs can be ecologically relevant. Recognition of the importance of connectivity across tropical seascapes has resulted in the emergence of new terminology that describes the utilization of patch mosaics, such as mangrove–seagrass–reef continuum [24,635] and seascape nursery [1825]. Nevertheless, many ecological studies of reef fishes still perceive coral reefs as isolated patches, where measurements on environmental structure are taken only within patches without considering the heterogeneous geographical context. The notion that the habitat patch is the natural spatial unit for measurement in ecology comes from the assumption that patch boundaries contain or delimit populations and communities [788]. For reef fishes, this is unlikely to be true for many species. Studies that have measured both within-patch and surrounding seascape structure demonstrate that inclusion of seascape variables increases the explanatory power of ecological models [238,992,1331,1970]. Consequently, for many species we propose that the focus of study must be expanded from single patches and single patch types (e.g. coral reef or mangrove) to include broader “seascape types” (sensu Pittman et al. [1971]). We argue that inclusion of seascape variables is required to gain a more complete picture of the key drivers in the ecology of fishes on coral reefs. Furthermore, a seascape ecology approach also offers great utility for identifying, characterizing, and prioritizing fish habitat in spatial management [1774,1874,1876] through consideration of ecologically meaningful structure at operationally relevant spatial scales. In the face of rapid human-induced changes to the patterning of marine ecosystems, it is critical that we seek to better understand relationships between spatial seascape patterns and the

ecological processes that underpin ecosystem services and resilience [593,1857]. Direct human activity can modify patchmosaic structure and global climate change drives geographical shifts in species, community structure, and seascape configuration [2004]. This chapter will introduce the seascape ecology perspective and explain why this approach is appropriate for the ecological study of coral reef fish. We advocate a broadening of perspective to include seascape mosaics and terrains, which will provide fresh insights on factors that drive fish species distributions and influence the functioning of coral reefs. Seascape ecology studies from the tropical Atlantic and IndoPacific regions are reviewed and priority questions for future research are presented. We focus primarily on studies that have quantified some component of the seascape geometry to explain ecological patterns and processes relevant to fishes on coral reefs.

TAKING LANDSCAPE ECOLOGY INTO THE SEA Landscape ecology provides a framework for ecological investigation that is spatially explicit, multidisciplinary, and multiscale [2532]. More specifically, landscape ecologists, and by extension seascape ecologists, are interested in the spatially explicit geometry of environmental patterns and their ecological consequences. A central tenet in landscape ecology is that patch context matters. The premise is that what happens in a patch is a function of both patch characteristics and the surrounding seascape and, therefore, cannot be adequately explained without quantification of patch context. A marine ecologist looking at the world through a landscape ecology perspective will ask different questions focused at different scales than other ecologists, such as: What are the ecological consequences of different shaped patches and different seascape composition or configuration? At what scale(s) is structure most influential? Concepts and analytical techniques from modern landscape ecology have proliferated in recent years and now permeate mainstream terrestrial ecology, yet with relatively little crossover into marine ecology. The study of spatial patchiness in the environment, however, is not new to marine ecologists. Pioneering marine applications of spatial theory were instrumental in testing hypotheses from island biogeography and patch dynamics [1898]. Island biogeography experiments were conducted within patches of mangrove [2355], coral reef [3,1723,2603], and seagrasses [1672,2116], but were limited to questions about individual patch geometry rather than the seascape as a whole [301,540]. The questions now being asked are more diverse, and for fishes, are guided by the scale of their movements [1969]. Technological advances in remote sensing [951], together with

275

C:/ITOOLS/WMS/CUP-NEW/5641473/WORKINGFOLDER/AROM/9781107089181C34.3D

276

276

[274–282] 4.12.2014 11:39AM

Simon J. Pittman and Andrew D. Olds

development of digital tools, such as geographical information systems (GIS) and spatial pattern metrics, have enabled quantification of seascape terrain at functionally relevant scales for fishes [1975,2637]. Using seafloor maps and spatial analytical tools, seascape ecology studies are now demonstrating that both the two-dimensional characteristics of patch mosaics (i.e. seascape composition and configuration) and three-dimensional structural characteristics of reef terrain (i.e. topographic complexity) strongly influence coral reef fish communities [301,1975]. A table of definitions for landscape ecology terminology with coral reef examples is provided at the end of the chapter.

WHAT IS A SEASCAPE? In seascape ecology, a seascape is a spatially defined heterogeneous area of the marine environment that can be quantified at

Figure 34.1 Conceptual illustration of multi-habitat use by coral reef fish and related seascape metrics. (A) Juveniles settle in non-reef nursery habitats before ontogenetic migration back to coral reefs. (B) Migration among reef patches to spawn or during dispersal. (C) Regular movement between reefs and adjacent habitats for foraging or sheltering. (D) Seascape view of reef patches that contrast with a surrounding homogenous matrix. (E) Seascape view of a spatially complex mosaic of habitats that cannot be characterized by simple binary metrics (Symbols courtesy of the IAN Network, ian.umces.edu/ symbols/).

any spatial and temporal scale(s) relevant to the ecological patterns and processes of interest. As such, the use of the terms seascape scale or seascape level when communicating scale selection is ambiguous without providing a numeric value and should be avoided [45]. Seascapes, just like landscapes, can be

C:/ITOOLS/WMS/CUP-NEW/5641473/WORKINGFOLDER/AROM/9781107089181C34.3D

277

[274–282] 4.12.2014 11:39AM

34 Seascape ecology of fishes on coral reefs Hypothetical species response A. Patch-mosaics

patch

edge effects

size

High quality

Patch B Patch A

Patch A

Discrete patch boundaries

Patch B matrix

Low quality

inter-patch distances High quality

B. Continuous spatial gradients Patch B

gradient

Low quality gradient

High quality gradient

Patch A Continuous variability No patch boundaries

Low quality

C. Continuous spatial gradients with discrete patch boundaries edge effect gradients

High quality

Patch B Patch A Discrete patch boundaries Internal heterogeneity preserved

Patch A

Patch B matrix

patch size

Low quality

inter-patch distances

Figure 34.2 Representations of coral reef seascapes in southwest Puerto Rico. (A) Benthic habitat map showing seascape complexity as a patch-mosaic model (with homogeneous patches and discrete patch boundaries). (B) Three-dimensional model of bathymetry. (C) Hybrid model (with discrete habitat patches and continuously varying structural complexity). The cross-sections show responses for a hypothetical species, which preferentially utilizes patches of complex coral reef (high quality) surrounded by less complex (and less

preferred) sandy sediments (lower quality). The size and spatial configuration of reefs is important and is represented adequately by the patch-mosaic model (Model A). However, species may also respond to continuously changing topographic complexity at finer scales (including within-patch structural heterogeneity and topographic complexity in the surrounding sand matrix) as represented in Model B and C.

conceptualized and modeled in different ways requiring informed selection by investigators. Three different conceptual models have been applied to map and interpret seascape patterning for tropical coastal environments: patch-matrix, patch-mosaic, and gradient models (Figure 34.2). Patch-matrix and patch-mosaic models represent seascapes as discrete homogeneous patches across a two-dimensional planar surface. These are typically applied in production of thematic benthic habitat maps. In

contrast, the spatial gradient or continuum model represents continuously varying data without discrete polygons. For example, seafloor depth variation, or bathymetry, is usually represented as a digital terrain model (DTM), providing spatially continuous information on attributes such as topographic complexity. The patch-matrix model has its origin in island biogeography and depicts the environment as a binary seascape composed of

277

C:/ITOOLS/WMS/CUP-NEW/5641473/WORKINGFOLDER/AROM/9781107089181C34.3D

278

278

[274–282] 4.12.2014 11:39AM

Simon J. Pittman and Andrew D. Olds focal habitat patches or “islands” surrounded by an inhospitable “matrix” [834]. In tropical waters, patch characteristics, such as size and isolation, have been examined relative to fish recruitment, turnover, and diversity on artificial reefs [2288,2603], coral reefs [113,1689], and reef islands [2258]. The binary patch-matrix model, however, has proven too simplistic for many species–environment relationships, particularly for highly mobile multi-habitat species [217]. As an alternative, the patch-mosaic model represents seascapes as collections of patch types where the composition and spatial arrangement of the whole mosaic influences ecological functioning [726]. Its underlying premise is that ecological processes vary among seascapes based on their composition and spatial configuration [2673]. In contrast, the spatial gradient model represents seascapes as continuously varying surfaces without discrete boundaries, although discontinuities and ecotones may still be represented [1663]. This model recognizes that species respond individually to gradual changes along spatial gradients in resource availability and environmental conditions [816]. We hypothesize here that hybrid habitat maps, which combine continuous spatial gradients and discrete patch mosaics, will explain more variability in response variables than either of the distinct types alone. Such maps, however, are rare and determining their ecological utility remains an area for future research.

IMPORTANCE OF CONSIDERING SCALE Species and life-stage responses to patchiness and gradients in environmental structure are likely to be scale dependent, therefore, scale selection is an important task in any ecological study [2672]. Since most species will respond to structure at multiple spatial scales and rarely are explicit focal scales known a priori, one solution is to examine species–environment relationships across a range of spatial scales. A multi-scale approach is advocated by landscape ecologists [1756,2304] and has become more common in ecological studies. For marine organisms, Pittman and McAlpine [1969] offer a multi-scale framework for scaling ecological studies that integrates hierarchy theory with movement ecology. Here the focal scale is guided by the spatial and temporal scales relevant to an ecological process of interest. The focal scale is nested within a spatial hierarchy that incorporates patterns and processes at both finer and broader scales. Seascape studies have now quantified species-environment relationships for various life stages at a range of spatial scales including the scale of tidal and diel movements (i.e. hundreds of meters) [1111,1971], ontogenetic habitat shifts (i.e. kilometers) and larval dispersal (i.e. tens of kilometers) (see reviews by [237,301,993]). In general, multiscale studies indicate that seascape structure within 100 m of the reef is most influential in explaining species distributions and assemblage richness.

When thematic maps (e.g. benthic habitat maps) are used to represent structure in the marine environment, issues related to spatial and thematic resolution, map accuracy, and error propagation become increasingly important [2637]. Studies on scale effects, particularly with regard to the impact on spatial pattern metrics are common in landscape ecology. Kendall and Miller [1331] found that changing the spatial resolution of benthic habitat maps resulted in disproportionate changes in the area, perimeter, and other spatial pattern metrics that influenced the strength of correlations for fish species–seascape relationships in the Caribbean [1332] (Figure 34.3). It is important to determine sensitivity to map characteristics (spatial and thematic resolution) and account for inherent data errors in landscape ecology studies, requiring some knowledge of cartography and remote sensing. Overall, the research questions and the suitability of data, concepts, tools, and techniques to address those questions will need careful evaluation for suitability at each scale of interest.

MOSAICS MATTER In landscape terminology, seascape composition is measured as the variety and abundance of patch types (e.g. habitat richness and percentage cover) and seascape configuration as the spatial arrangement of patches (e.g. nearest neighbor distance, contagion, edge density). Evidence is emerging to support the hypothesis that some of these seascape variables are key ecological drivers for coral reef fish [301,2637]. The landscape ecology perspective has been applied most often in the Caribbean [992,1330,1971,2748], but also in the Western Indian Ocean [237,1007] and Pacific Ocean [1877], where the proximity of marine vegetation (i.e. seagrasses, mangroves) to coral reefs was an important predictor for several reef fish metrics. In Japan, landscape ecology has been applied to understand habitat use for resident anemonefish on coral reefs where patch area multiplied by patch perimeter was a useful predictor of fish distributions [1060,1061]. Although most studies examine pattern–pattern relationships, some have linked pattern with ecological processes to reveal that seascape effects can influence the structure of reef food webs [1074], predator–prey dynamics [711], herbivory [1876], movement behavior [1111], and marine reserve performance [1206]. Interestingly, in some areas the impact of herbivory on coral reef ecosystems can be clearly observed from space with highresolution satellite imagery (Figure 34.4; [713,1560]) allowing spatial measurements to be made. Marine examples and guidance on the application of spatial pattern metrics to quantify seascape structure is provided by Wedding et al. [2637].

INFLUENCE OF TERRAINS Coral reef ecosystems have been mapped at a range of spatial scales using data from airborne laser (LiDAR), ship-based acoustic

C:/ITOOLS/WMS/CUP-NEW/5641473/WORKINGFOLDER/AROM/9781107089181C34.3D

279

[274–282] 4.12.2014 11:39AM

34 Seascape ecology of fishes on coral reefs

Low thematic Low spatial

Low thematic High spatial

High thematic Low spatial

High thematic High spatial

Map attributes Number of patch types

4

4 m2

17 m2

17 m2

100 m2

Minimum mapping unit

4048

Number of polygons

13

1069

311

2497

Total edge density

225

465

423

790

Mean shape index

1.7

1.5

1.9

1.4

Mean nearest neighbour distance (m)

131

44

152

55

100

sonar (e.g. multibeam echosounders), and air- and space-borne optical sensors (e.g. high-resolution multispectral imagery) [951]. Remote sensing data provides continuous data on seafloor structure that can be categorized into maps of seafloor habitat types, or as bathymetry using a DTM. A suite of spatial pattern metrics, which are commonly used in geomorphology and industrial engineering to quantify surface morphology, have proven valuable for predicting reef fish distributions and diversity in the Caribbean Sea [e.g. 1972,1973,1974] and Indo-Pacific Oceans [e.g. 1365,2047,2637]. Topographic complexity, water depth and cross-shelf location have proven to be the most useful predictors of fish diversity and abundance across broad (hundreds of meters to kilometers) geographical areas [e.g. 1974,2637]. The importance of topographic complexity to reef fishes means that digital terrains also provide an opportunity to forecast impacts to species distributions from declining topographic complexity caused by coral mortality, erosion, and structural collapse [1975]. Comparison with finer-scale in situ measurements of topographic complexity (i.e. chain-tape rugosity), however, indicate that finer resolution LiDAR terrain (