DISPERSAL, FISHING, AND THE CONSERVATION OF MARINE SPECIES

A DISSERTATION

SUBMITTED TO THE DEPARTMENT OF BIOLOGY

AND THE COMMITTEE ON GRADUATE STUDIES

OF STANFORD UNIVERSITY

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

Malin La Farge Pinsky

June 2011

http://creativecommons.org/licenses/by-nc/3.0/us/

This dissertation is online at: http://purl.stanford.edu/fk096nf3828

© 2011 by Malin La Farge Pinsky. All Rights Reserved.

Re-distributed by Stanford University under license with the author.

This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.

ii

I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.

Stephen Palumbi, Primary Adviser

I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.

Rodolfo Dirzo

I certify that I have read this dissertation and that, in my opinion, it is fully adequatein scope and quality as a dissertation for the degree of Doctor of Philosophy.

Elizabeth Hadly

Approved for the Stanford University Committee on Graduate Studies.

Patricia J. Gumport, Vice Provost Graduate Education

This signature page was generated electronically upon submission of this dissertation in electronic format. An original signed hard copy of the signature page is on file inUniversity Archives.

iii

iv

v

Abstract

A central goal of ecology is to understand the forces driving the distribution

and abundance of organisms. However, understanding the population dynamics of

high-dispersal species, their conservation, and the connections between population

dynamics and evolution remains difficult. It is in this context that marine organisms

provide a particularly intriguing and challenging study system. Their population

dynamics are often highly stochastic, most species have a great ability to disperse, and

as the last group of wild species exploited commercially, their ecology and evolution

can be strongly influenced by human behavior. By using population genetics,

modeling, and meta-analysis, this thesis investigates the spatial ecology of reef fish

and the causes and evolutionary consequences of global fisheries collapse.

One of the first challenges in understanding spatial population dynamics is

obtaining accurate measurements of dispersal abilities. This has been especially

difficult for marine species with pelagic larvae. In Chapter 1, I apply a new approach

to measuring single-generation dispersal kernels in Clark’s anemonefish (Amphiprion

clarkii) in the central Philippines (Pinsky et al. 2010 Evolution 64(9): 2688-2700).

After developing two methods for measuring the strength of local genetic drift, my

results suggest that larval dispersal kernels in A. clarkii had a spread near 11 km (4-27

km). This study shows that ecologically relevant larval dispersal can be estimated with

widely available genetic methods when effective density is measured carefully

through cohort sampling and ecological censuses.

In Chapter 2, I use dispersal kernels to develop a model for population

openness. Openness refers to the degree to which populations are replenished by

immigrants or by local production, a factor that has strong implications for population

dynamics, species interactions, and response to exploitation. It is also a population

trait that has been increasingly measured empirically, though we have until now

lacked theory for predicting population openness. I show that considering habitat

isolation elegantly explains the existence of surprisingly closed populations in high

vi

dispersal species, and that relatively closed populations are expected when patch

spacing is more than twice the standard deviation of a species’ dispersal kernel. In

addition, empirical scales of habitat patchiness on coral reefs are sufficient to create

both largely open and largely closed populations. We predict that habitat patchiness

has strong control over population replenishment pathways for a wide range of marine

and terrestrial species with a highly dispersive life stage.

While the first tow chapters have strong implications for the design of regional

marine protected areas, I turn to global conservation questions in Chapters 3 and 4. I

first ask which marine fishes are most vulnerable to human impacts (Pinsky et al. 2011

Proceedings of the National Academy of Sciences doi/10.1073/pnas.1015313108).

Surveys of terrestrial species have suggested that large-bodied species and top

predators are the most at risk, but there has been no global test of this hypothesis in the

sea. Contrary to expectations, two datasets compiled from around the world suggest

that up to twice as many fisheries for small, low trophic level species have collapsed

as compared to those for large predators.

I then show that collapsed and overfished species have lower genetic diversity

than their close relatives (Pinsky & Palumbi, in prep). While the ecological and

ecosystem impacts of harvesting wild populations have long been recognized, it has

been controversial how widespread evolutionary impacts are. Using a meta-analytical

approach across 37 taxonomically paired comparisons, I find on average 19% fewer

alleles per locus in overfished species, but little difference in heterozygosity. I confirm

with simulations that these results are consistent with a recent population bottleneck.

These results suggest that the genetic impacts of overharvest are widespread, even

among abundant species. A loss of allelic richness has implications for the long-term

evolutionary potential of species.

vii

Acknowledgements

I could not have completed these last five years without the support, encouragement,

and enthusiasm of those around me. While this list is incomplete, I am indebted to:

My Advisor

Steve Palumbi

My Committee

Elizabeth Hadly

Fiorenza Micheli

Steve Gaines

Rodolfo Dirzo

Chris Lowe

Collaborators

Olaf Jensen, Dan Ricard, Boris Worm, Ray Hilborn, Trevor Branch, Katie Arkema,

Greg Guannel, Mary Ruckelshaus, Anne Guerry, Marcel van Tuinen, Doug Kennett,

Seth Newsome, Serge Andréfouët, Humberto Montes, Jr. and the Visayas State

University Marine Lab, Amado Blanco and the Project Seahorse Foundation, Rose-

Liza Eisma-Osorio and the Coastal Conservation and Education Foundation

Colleagues in the Palumbi Lab

Melissa Pespeni, Jason Ladner, Carolyn Tepolt, Alison Haupt, Ryan Kelly, Dan

Barshis, Tom Oliver, Mollie Manier, Heather Galindo, Liz Alter, Emily Jacobs-

Palmer, Kelly Barr, Kristen Ruegg, Vanessa Michelou, Arjun Sivasundar, Pierre De

Wit, Marina Oster, Hannah Jaris, Veronica Searles, and Mark Walker

My Cohort

Julie Stewart, Nishad Jayasundara, Kevin Miklasz, Aaron Carlisle, Posy Busby,

Camila Donati, Beth Pringle, and Jason Ladner

Hopkins Marine Station Scientists and Staff

Kristy Kroeker, Cheryl Logan, Judit Pungor, Giulio de Leo, Salvador Jorgensen,

Chelsea Wood, Steve Litvin, Doug McCauley, Mark Denny, Ashley Greenley, Ashley

viii

Booth, Tom Hata, Megan Jensen, Ishbel Kerkez, Dane Klinger, Judy Thompson, Joe

Wible, Doreen Zelles, Chris Patton, Freya Sommer, Carol Reeb, John Lee, Jim

Watanabe, Bob Doudna, Peter Ferrante, Barbara Compton, and Vicki Pearse

Stanford University Scientists and Staff

Jessica Blois, Judsen Bruzgul, Sarah McMenamin, Paula Spaeth, Lily Li, Dmitri

Petrov, Valeria Kiszka, Matt Pinheiro, and Jennifer Mason

Funding

National Science Foundation Graduation Fellowship, National Defense Science and

Engineering Graduation Fellowship, International Society for Reef Studies and the

Ocean Conservancy, Earl & Ethel Myers Oceanographic Trust, Jane Miller Scholars,

Friends of Hopkins, Woods Institute for the Environment, Center for Ocean Solutions,

National Center for Ecological Analysis and Synthesis, and Stanford Department of

Biology

and My Family

Kristin Hunter-Thomson

Rob, Margaret, and Maia Pinsky

ix

Table of Contents

Abstract........................................................................................................................... v

Acknowledgements ......................................................................................................vii

Table of Contents .......................................................................................................... ix

List of Tables...............................................................................................................xiii

List of Figures.............................................................................................................. xiv

Introduction .................................................................................................................... 1

Statement on Multiple Authorship ..................................................................... 5

References .......................................................................................................... 6

Chapter 1: Using isolation by distance and effective density to estimate dispersal

scales in anemonefish ..................................................................................... 11

1.1. Abstract...................................................................................................... 11

1.2. Introduction ............................................................................................... 12

1.3. Materials and Methods .............................................................................. 14

1.3.1. Study system............................................................................... 14

1.3.2. Ecological surveys...................................................................... 15

1.3.3. Genetic samples.......................................................................... 16

1.3.4. Genetic analysis.......................................................................... 17

1.3.5. Isolation by distance ................................................................... 17

1.3.6. Effective density from temporal method.................................... 18

1.3.7. Effective density from census density........................................ 20

1.3.8. Uncertainty ................................................................................. 21

1.4. Results ....................................................................................................... 22

1.4.1. Adult density .............................................................................. 22

1.4.2. Genetic analysis.......................................................................... 23

1.4.3. Temporal estimates of effective density..................................... 24

1.4.4. Effective density from census density........................................ 24

1.4.5. Dispersal distance....................................................................... 25

x

1.5. Discussion.................................................................................................. 25

1.5.1. The central role of effective density and census density ............ 26

1.5.2. Effective density from genetic diversity?................................... 27

1.5.3. Assumptions of dispersal calculations........................................ 28

1.5.4. Comparison to other measures of larval dispersal...................... 29

1.5.5. Reef patchiness and comparison to self-recruitment.................. 30

1.6. Conclusions and future directions ............................................................. 31

1.7. Acknowledgments ..................................................................................... 32

1.8. References ................................................................................................. 33

1.9. Tables ........................................................................................................ 40

1.10. Figures ..................................................................................................... 43

Chapter 2: Open and closed seascapes: where does habitat patchiness create

populations with low immigration? .............................................................. 49

2.1. Abstract...................................................................................................... 49

2.2. Introduction ............................................................................................... 49

2.3. Materials and Methods .............................................................................. 53

2.3.1. Model.......................................................................................... 53

2.3.2. Simplifications for applying the model ...................................... 55

2.3.3. Application to simple seascapes................................................. 56

2.3.4. Application to remotely sensed seascapes.................................. 56

2.4. Results ....................................................................................................... 56

2.4.1. Population openness in simple seascapes................................... 56

2.4.2. Comparison to empirical studies ................................................ 57

2.4.3. Closed populations within empirical seascapes ......................... 58

2.5. Discussion.................................................................................................. 60

2.5.1. A new language for marine connectivity: immigration and

emigration................................................................................. 61

2.5.2. Immigration in naturally patchy landscapes............................... 62

2.5.3. Model limitations........................................................................ 63

xi

2.5.4. Future directions......................................................................... 64

2.6. Acknowledgments ..................................................................................... 65

2.7. References ................................................................................................. 65

2.8. Figures ....................................................................................................... 70

Chapter 3: Unexpected patterns of fisheries collapse in the world’s oceans ........ 79

3.1. Abstract...................................................................................................... 79

3.2. Introduction ............................................................................................... 79

3.3. Materials and Methods .............................................................................. 81

3.3.1. Data sources................................................................................ 81

3.3.2. Fishery collapses – Assessment data.......................................... 82

3.3.3. Fishery collapses – Landings data.............................................. 82

3.3.4. Fisheries characteristics.............................................................. 84

3.3.5. Life history traits ........................................................................ 84

3.3.6. Statistical models........................................................................ 85

3.3.7. Phylogenetically independent contrasts ..................................... 86

3.4. Results ....................................................................................................... 87

3.5. Discussion.................................................................................................. 89

3.6. Acknowledgements ................................................................................... 92

3.7. References ................................................................................................. 93

3.8. Tables ........................................................................................................ 98

3.9. Figures ..................................................................................................... 108

3.10. Supplementary Material ........................................................................ 117

Chapter 4: Genetic impacts of overfishing are widespread.................................. 127

4.1. Abstract and Introduction ........................................................................ 127

4.2. Results and Discussion ............................................................................ 128

4.3. Materials and Methods ............................................................................ 132

4.3.1. Literature selection ................................................................... 132

4.3.2. Analysis .................................................................................... 133

xii

4.3.3. Simulations ............................................................................... 134

4.4. Acknowledgements ................................................................................. 135

4.5. References ............................................................................................... 135

4.6. Figures ..................................................................................................... 139

4.7. Supplementary Material .......................................................................... 143

xiii

List of Tables

Table 1.1. Microsatellite loci used in Amphiprion clarkii............................................ 40

Table 1.2. Densities of A. clarkii compiled from the literature.................................... 41

Table 1.3. Sample sizes of A. clarkii at each site ......................................................... 42

Table 3.1. Species with stocks that had collapsed according to stock assessments and

in the landings database.................................................................................... 98

Table 3.2. Parameters for models predicting the proportion of stocks collapsed within

each species .................................................................................................... 104

Table 3.S1. Additional sources consulted for data on egg diameter and fecundity ... 117

Table 4.S1. References for the overfished and control species ................................... 143

xiv

List of Figures Figure 1.1. Cebu and Leyte Islands in the central Philippines ..................................... 43

Figure 1.2. Census densities of A. clarkii in Cebu and Leyte, Philippines .................. 44

Figure 1.3. Genetic distance between A. clarkii populations in Cebu and Leyte......... 45

Figure 1.4. Graph illustrating the calculation of dispersal spread................................ 46

Figure 1.5. Effects of reef patchiness on the self-recruitment fraction ........................ 47

Figure 2.1. Diagram illustrating habitat patchiness and dispersal................................ 70

Figure 2.2. Maps of coral reef seascapes that were analyzed for openness ................. 71

Figure 2.3. Percent immigration on a one-dimensional coastline or a two-dimensional

grid.................................................................................................................... 72

Figure 2.4. Percent immigration with advection .......................................................... 73

Figure 2.5. Variation in openness across patchy seascapes.......................................... 74

Figure 2.6. Population openness within each of 17 coral reef seascapes ..................... 75

Figure 2.7. Mean immigration plotted against mean nearest neighbor distance .......... 76

Figure 2.8. The exact level of immigration depends on grid size ................................ 77

Figure 3.1. Life history patterns of fished species...................................................... 108

Figure 3.2. Collapses in the assessment database in relation to life history traits...... 109

Figure 3.3. Collapses in the landings database in relation to life history traits .......... 110

Figure 3.4. Life history trends in the proportion of overfished stocks (assessment data)

........................................................................................................................ 111

Figure 3.5. Life history patterns with an alternative definition of collapse (landings

data) ................................................................................................................ 112

Figure 3.6. Life history trends in the magnitude of decline ....................................... 113

Figure 3.7. Life history patterns after correcting for relative fishery mortality

(assessment data) ............................................................................................ 114

Figure 3.8. Life history patterns after correcting for phylogeny (assessment data) ... 115

Figure 3.9. Life history patterns after correcting for phylogeny (landings data) ....... 116

Figure 4.1. Illustration of paired comparisons............................................................. 139

xv

Figure 4.2. Overfished species have consistently lower allelic richness than closely

related control species ..................................................................................... 140

Figure 4.3. Overfished species have similar heterozygosity when compared to closely

related control species ..................................................................................... 141

Figure 4.4. Expected loss of allelic richness and heterozygosity from simulated

bottlenecks ...................................................................................................... 142

xvi

1

Introduction

A central goal of ecology has long been to understand the forces driving the

distribution and abundance of organisms. Asking questions about why populations

fluctuate, or why animals are found here but not there, help us to understand the

structure of the natural world around us. In addition, the answers to these questions

provide guidance as we try to conserve the beauty and value that we derive from

nature, both on land and in the sea.

Early research on fluctuations in natural populations emphasized external

influences on the abundance of animals, such as the effects of disease or weather on

mortality (Andrewartha & Birch 1954). In the 1950s and 1960s, the emphasis

switched to intrinsic forces that regulate populations, including density-dependence, or

the tendency for populations to grow when rare and to decline when overly abundant

(Krebs 1995). Further developments have revealed the influence of non-linear

dynamics within populations (Higgins et al. 1997) and uncovered the importance of

large-scale and predictable climate forcing (Bjørnstad & Grenfell 2001). Areas of

active research include investigations on the role that dispersal between populations

plays in determining population persistence (Levins 1969; Hanski 1998; Hastings &

Botsford 2006) and the interplay between evolution and ecology (Stockwell et al.

2003). In addition, the fingerprint of human influences are being seen across an ever-

greater number of systems (Vitousek et al. 1997; Palumbi 2001). In the current

synthesis, the challenge is to understand how this range of forces drives population

dynamics through time and space, and how their relative influence changes with

species and setting (Bjørnstad & Grenfell 2001).

It is in this context that marine species provide a particularly intriguing and

challenging study system. Their population dynamics tend to be highly stochastic,

most species have a great ability to disperse, and as the last group of wild species

exploited commercially, their dynamics and evolution can be strongly influenced by

human behavior. These factors have largely foiled traditional approaches to population

dynamics, and many aspects of marine population dynamics remain poorly known

2

despite the substantial effort we invest in counting and enumerating those that are

commercially important. In addition, harvest was the only major human influence on

marine populations for centuries, but this is now rapidly changing as mineral

extraction, shipping, renewable energy, and recreation begin to crowd the seascape

(Crowder et al. 2006). Where, when, and how much development will be consistent

with vibrant, healthy, and productive marine ecosystems in the future are all pressing

questions.

For many years, marine ecologists viewed populations as largely open, with

dynamics determined by the vagaries of ocean currents that deliver marine larvae each

year (Roughgarden et al. 1988; Caley et al. 1996; Siegel et al. 2008). Most marine

fish, and nearly three-quarters of marine invertebrates, have pelagic larvae that

disperse for weeks or months (Scheltema 1986). These factors lead marine species to

generally have greater dispersal abilities than species on land (Kinlan & Gaines 2003).

Taken to an extreme, the view of open populations posits that local interactions

between individuals of the same or different species, or local levels of exploitation, are

largely irrelevant, and that chance arrival or the direction of ocean currents are all that

matters for population dynamics and community composition (Sale 1977; Roberts

1997). Even if not taken so far, this view of well-mixed marine populations permeates

fisheries management, and populations are often managed as large stocks that span

hundreds of kilometers.

In more recent years, however, evidence has accumulated that some marine

populations are actually quite closed at scales less than a kilometer, and instead

replenished primarily by local reproduction (Jones et al. 1999; Swearer et al. 1999;

Almany et al. 2007). Discovery of genetic breaks within marine populations also hint

that dispersal may not be as extensive as once thought (Barber et al. 2000; Taylor &

Hellberg 2003). This has helped to drive a growing realization that dispersal distances

in marine species may not be as extensive as previously thought (Levin 2006).

Dramatic strides have also been made in modeling the oceanographic currents that

presumably carry larvae from birth until settling (James et al. 2002; Siegel et al.

3

2008), but without empirical measurements of larval dispersal distances, it remains

uncertain how accurate such models are.

Basic quantities such as how far marine larvae actually travel have remained

difficult to measure because their small size, large number, and high mortality rates

preclude standard tracking methods. Therefore, dispersal distances for marine species

remain largely unknown, even though this basic quantity sets the spatial scale of

marine population dynamics and by extension, the appropriate spatial scales of

conservation (Botsford et al. 2001). We currently sit at a point where substantial

theory for marine metapopulations exists (e.g., Armsworth 2002; Hastings & Botsford

2006; Figueira 2009), but we have yet to ground this theory in empirical research. My

first chapter begins to address this problem by developing a new method for

measuring larval dispersal kernels with population genetics (Pinsky et al. 2010).

One of the practical needs driving this interest in spatial marine populations

dynamics is the dramatic trend in marine conservation towards spatial management.

Marine protected areas (MPAs), marine reserves, and now marine spatial management

(Crowder & Norse 2008) are all premised on the idea of setting aside particular areas

of the ocean for particular humans uses or for the lack thereof. How will these choices

affect marine populations? These questions are inherently spatial and have helped

spark efforts to map the ocean’s habitats. These data are often used for choosing

appropriate sites for MPAs (Leslie et al. 2003), and yet on land, such habitat

patchiness has also long been recognized as an important influence on the dispersal of

organisms between populations (Schumaker 1996). In my second chapter, I expand

from the larval dispersal distances in my first chapter to show that, in coral reefs, the

spacing between habitat patches has a strong influence on the openness of marine

populations as well (Pinsky et al. in review). More generally, there is a need to focus

more attention on the effects of habitat patchiness on the spatial dynamics of marine

populations and their conservation, something that is just now becoming possible.

As mentioned above, there are also strong fingerprints of human influence on

marine populations dynamics, particularly those species that are directly harvested for

food or other uses. In some locations, fishing has denuded reefs of large predators

4

(Stevenson et al. 2006) or removed habitat important for many species (Thrush et al.

1998). How the uncoordinated actions of many fisheries around the world add up to a

coherent pattern of impacts across marine species, however, has been controversial

(Pauly et al. 1998; Worm et al. 2009; Branch et al. 2010). These questions inherently

address broad trends in the nature of human preferences, economic forces, the practice

of natural resource management, and, at the core, the sensitivity of species to human

impacts. While large species are most vulnerable to human impacts on land (Cardillo

et al. 2005), I show in my third chapter that fishing acts in a fundamentally different

way across species and has driven both large and small species to collapse with similar

probabilities (Pinsky et al. 2011). Part of the lesson from this chapter is that we

severely impact even those populations that grow the fastest and are presumably the

least sensitive. There are also hints that environmental variability plays a role in some

of these collapses, and the interaction between exploitation and environmental

variability remains an interesting area of research.

In addition to these cross-species impacts, fishing may also have evolutionary

impacts within species (Sharpe & Hendry 2009). This linkage between evolutionary

impacts and population dynamics challenges traditional, non-evolutionary approaches

to understanding population ecology and reflects a growing trend towards recognizing

that evolutionary and ecological processes can act on similar timescales (Stockwell et

al. 2003). In the short term, the collapse and lack of recovery in Atlantic cod (Gadus

morhua) may be due in part to evolutionary changes towards smaller size and early

maturation that were induced by fishing (Olsen et al. 2004). Over the longer term,

researchers have also suggested that fishing can drive a loss of genetic diversity within

species, with implications for the ability of species to adapt to environmental change

(Hauser et al. 2002). My fourth chapter shows that this loss of genetic diversity affects

not only a few isolated species, but is widespread across those that are overfished.

This finding suggests a strong connection between the short-term demographic

consequences of fishing, and of any harvest of wild species, and long-term

evolutionary impacts.

5

Throughout, my dissertation has integrated a diverse set of approaches to

advance marine conservation biology, with a focus on spatial ecology and meta-

analysis to understand large-scale trends. Population genetics in particular has proved

a useful tool, both for tracking larvae through the ocean to understand dispersal and to

track evolutionary consequences of population declines. The growing availability of

ecological information across wide temporal, spatial, and taxanomic scales has also

opened an entirely new set of questions for inquiry, and harnessing this information

was crucial for my third and fourth chapters.

Moving forward, considerable work continues to be needed on the spatial

ecology and spatial dynamics of marine populations. The high levels of dispersal and

stochasticity continue to challenge the simple models developed so far, and the link

between theory and empirical patterns continues to be weak. Linking these will require

innovative fieldwork and practical modeling. Continuing to apply genetic tools to

these ecological questions will likely be an important approach, as will greater

integration with oceanographic information. The analysis of large spatial and temporal

datasets are also likely provide new advances, and there have been some exciting

developments in this area already (Gouhier et al. 2010). The large datasets from

fisheries that are now available are likely to provide important sources of data against

which to test hypotheses. Finally, the influences of climate change on marine

populations are both substantial and largely unknown. Dispersal provides many

marine species with a method for reaching more suitable conditions when climate

changes, but scaling up from the local impacts of climate change to the large scale

spatial dynamics of marine species, including range shifts, remains challenging and in

important area of future work.

Statement on Multiple Authorship

Throughout my tenure as a Ph.D. student, I have benefited tremendously from

collaborations both within Stanford and outside. However, I am the first author and

primary contributor to each of my dissertation chapters, including the design, data

collection, analysis, and writing. The paragraphs below explain the roles of my co-

6

authors, while acknowledgments after each chapter thank the many other individuals

who helped make these projects a reality.

The fieldwork in the central Philippines that forms the core of Chapter 1

benefited greatly from the assistance of Humberto Montes, Jr. in navigating the local

culture, providing access to a field station, and arranging the logistics of fieldwork.

Steve Palumbi helped to design the project, design the analysis, and write the

manuscript.

In Chapter 2, Serge Andréfouët and Sam Purkis provided the remote sensing

data and provided input on the analysis methods. Steve Palumbi helped to design the

analysis and write the manuscript.

For Chapter 3, Olaf Jensen and Steve Palumbi helped to design the analysis

and write the paper, while Dan Ricard provided stock assessment data.

Steve Palumbi helped to design and write Chapter 4.

References

Almany G.R., Berumen M.L., Thorrold S.R., Planes S. & Jones G.P. (2007). Local

replenishment of coral reef fish populations in a marine reserve. Science, 316,

742-744.

Andrewartha H.G. & Birch L.C. (1954). The Distribution and Abundance of Animals.

The University of Chicago Press, Chicago, IL.

Armsworth P.R. (2002). Recruitment limitation, population regulation, and larval

connectivity in reef fish metapopulations. Ecology, 83, 1092-1104.

Barber P.H., Palumbi S.R., Erdmann M.V. & Moosa M.K. (2000). A marine Wallace's

line? Nature, 406, 692-693.

Bjørnstad O.N. & Grenfell B.T. (2001). Noisy clockwork: time series analysis of

population fluctuations in animals. Science, 293, 638-43.

doi:10.1126/science.1062226

7

Botsford L.W., Hastings A. & Gaines S.D. (2001). Dependence of sustainability on

the configuration of marine reserves and larval dispersal distance. Ecol. Lett.,

4, 144-150.

Branch T.A., Watson R., Fulton E.a., Jennings S., McGilliard C.R., Pablico G.T.,

Ricard D. & Tracey S.R. (2010). The trophic fingerprint of marine fisheries.

Nature, 468, 431-435. doi:10.1038/nature09528

Caley M.J., Carr M.H., Hixon M.A., Hughes T.P., Jones G.P. & Menge B.A. (1996).

Recruitment and the local dynamics of open marine populations. Annu. Rev.

Ecol. Syst., 27, 477-500.

Cardillo M., Mace G.M., Jones K.E., Bielby J., Bininda-Emonds O.R.P., Sechrest W.,

Orme C.D.L. & Purvis A. (2005). Multiple causes of high extinction risk in

large mammal species. Science, 309, 1239-1241.

Crowder L. & Norse E. (2008). Essential ecological insights for marine ecosystem-

based management and marine spatial planning. Mar. Policy, 32, 772-778.

doi:10.1016/j.marpol.2008.03.012

Crowder L.B., Osherenko G., Young O.R., Airamé S., Norse E.A., Baron N., Day

J.C., Douvere F., Ehler C.N., Halpern B.S., Langdon S.J., McLeod K.L.,

Ogden J.C., Peach R.E., Rosenberg A.A. & Wilson J.A. (2006). Resolving

mismatches in U.S. ocean governance. Science, 313, 617-618.

Figueira W.F. (2009). Connectivity or demography: Defining sources and sinks in

coral reef fish metapopulations. Ecol. Model., 220, 1126-1137.

Gouhier T.C., Guichard F. & Menge B.A. (2010). Ecological processes can

synchronize marine population dynamics over continental scales. P Natl Acad

Sci USA.

Hanski I. (1998). Metapopulation dynamics. Nature, 396, 41-50.

Hastings A. & Botsford L.W. (2006). Persistence of spatial populations depends on

returning home. Proc. Natl. Acad. Sci. USA, 103, 6067-6072.

Hauser L., Adcock G.J., Smith P.J., Bernal Ramírez J.H. & Carvalho G.R. (2002).

Loss of microsatellite diversity and low effective population size in an

8

overexploited population of New Zealand snapper (Pagrus auratus). P Natl

Acad Sci USA, 99, 11742-11747.

Higgins K., Hastings A., Sarvela J.N. & Botsford L.W. (1997). Stochastic dynamics

and deterministic skeletons: population behavior of dungeness crab. Science,

276, 1431-1435. doi:10.1126/science.276.5317.1431

James M.K., Armsworth P.R., Mason L.B. & Bode L. (2002). The structure of reef

fish metapopulations: modelling larval dispersal and retention patterns. P Roy

Soc Lond B Bio, 269, 2079-2086.

Jones G.P., Milicich M.J., Emslie M.J. & Lunow C. (1999). Self-recruitment in a coral

reef fish population. Nature, 402, 802-804.

Kinlan B.P. & Gaines S.D. (2003). Propagule dispersal in marine and terrestrial

environments: a community perspective. Ecology, 84, 2007-2020.

Krebs C.J. (1995). Two paradigms of population regulation. Wildl. Res., 22, 1.

doi:10.1071/WR9950001

Leslie H., Ruckelshaus M., Ball I.R., Andelman S. & Possingham H.P. (2003). Using

siting algorithms in the design of marine reserve networks. Ecol. Appl., 13,

S185-S198.

Levin L.A. (2006). Recent progress in understanding larval dispersal: new directions

and digressions. Integr. Comp. Biol., 46, 282-297.

Levins R. (1969). Some demographic and genetic consequences of environmental

heterogeneity for biological control. Bulletin of the Entomological Society of

America, 15, 237-240.

Olsen E.M., Heino M., Lilly G.R., Morgan M.J., Brattey J., Ernande B. & Dieckmann

U. (2004). Maturation trends indicative of rapid evolution preceded the

collapse of northern cod. Nature, 428, 932-935.

Palumbi S.R. (2001). Humans as the world’s greatest evolutionary force. Science, 293,

1786-1790.

Pauly D., Christensen V., Dalsgaard J., Froese R. & Torres F., Jr. (1998). Fishing

down marine food webs. Science, 279, 860-863.

9

Pinsky M., Jensen O.P., Ricard D. & Palumbi S.R. (2011). Unexpected patterns of

fisheries collapse in the world’s oceans. Proceedings of the National

Academies of Science. doi:10.1073/pnas.1015313108

Pinsky M., Montes H.R., Jr. & Palumbi S.R. (2010). Using isolation by distance and

effective density to estimate dispersal scales in anemonefish. Evolution, 64,

2688-2700. doi:10.1111/j.1558-5646.2010.01003.x

Pinsky M.L., Palumbi S.R., Andréfouët S. & Purkis S.J. (in review). Open and closed

seascapes: where does habitat patchiness create populations with low

immigration? P Roy Soc B-Biol Sci.

Roberts C.M. (1997). Connectivity and management of Caribbean coral reefs. Science,

278, 1454-1457.

Roughgarden J., Gaines S.D. & Possingham H. (1988). Recruitment dynamics in

complex life cycles. Science, 241, 1460-1466.

Sale P.F. (1977). Maintenance of high diversity in coral reef fish communities. The

American Naturalist, 111, 337-359.

Scheltema R.S. (1986). On dispersal and planktonic larvae of benthic invertebrates: an

eclectic overview and summary of problems. Bull. Mar. Sci., 39, 290-322.

Schumaker N.H. (1996). Using landscape indices to predict habitat connectivity.

Ecology, 77, 1210-1225.

Sharpe D.M.T. & Hendry A.P. (2009). Life history change in commercially exploited

fish stocks: an analysis of trends across studies. Evolutionary Applications, 2,

260-275. doi:10.1111/j.1752-4571.2009.00080.x

Siegel D.A., Mitarai S., Costello C.J., Gaines S.D., Kendall B.E., Warner R.R. &

Winters K.B. (2008). The stochastic nature of larval connectivity among

nearshore marine populations. P Natl Acad Sci USA, 105, 8974-8979.

Stevenson C., Katz L.S., Micheli F., Block B.A., Heiman K.W., Perle C., Weng K.,

Dunbar R. & Witting J. (2006). High apex predator biomass on remote Pacific

islands. Coral Reefs, 26, 47-51.

Stockwell C.A., Hendry A.P. & Kinnison M.T. (2003). Contemporary evolution meets

conservation biology. Trends Ecol. Evol., 18, 94-101.

10

Swearer S.E., Caselle J.E., Lea D.W. & Warner R.R. (1999). Larval retention and

recruitment in an island population of a coral-reef fish. Nature, 402, 799-802.

Taylor M.S. & Hellberg M.E. (2003). Genetic evidence for local retention of pelagic

larvae in a Caribbean reef fish. Science, 299, 107-109.

Thrush S.F., Hewitt J.E., Cummings V.J., Dayton P.K., Cryer M., Turner S.J., Funnell

G.A., Budd R.G., Milburn C.J. & Wilkinson M.R. (1998). Disturbance of the

marine benthic habitat by commercial fishing: Impacts at the scale of the

fishery. Ecol. Appl., 8, 866-879.

Vitousek P.M., Mooney H.A., Lubchenco J. & Melillo J.M. (1997). Human

domination of Earth's ecosystems. Science, 277, 494-499.

Worm B., Hilborn R., Baum J.K., Branch T.A., Collie J.S., Costello C., Fogarty M.J.,

Fulton E.A., Hutchings J.A., Jennings S., Jensen O.P., Lotze H.K., Mace P.M.,

McClanahan T.R., Minto C., Palumbi S.R., Parma A.M., Ricard D., Rosenberg

A.A., Watson R. & Zeller D. (2009). Rebuilding global fisheries. Science, 325,

578.

11

Chapter 1

Using isolation by distance and effective density

to estimate dispersal scales in anemonefish

1.1 Abstract

Robust estimates of dispersal are critical for understanding population

dynamics and local adaptation, as well as for successful spatial management. Genetic

isolation by distance patterns hold clues to dispersal, but understanding these patterns

quantitatively has been complicated by uncertainty in effective density. In this study,

we genotyped populations of a coral reef fish (Amphiprion clarkii) at 13 microsatellite

loci to uncover fine-scale isolation by distance patterns in two replicate transects.

Temporal changes in allele frequencies between generations suggested that effective

densities in these populations are 4-21 adults/km. A separate estimate from census

densities suggested that effective densities may be as high as 82-178 adults/km.

Applying these effective densities with isolation by distance theory suggested that

larval dispersal kernels in A. clarkii had a spread near 11 km (4-27 km). These kernels

predicted low fractions of self-recruitment in continuous habitats, but the same kernels

were consistent with previously reported, high self-recruitment fractions (30-60%)

when realistic levels of habitat patchiness were considered. Our results suggested that

ecologically relevant larval dispersal can be estimated with widely available genetic

methods when effective density is measured carefully through cohort sampling and

ecological censuses, and that self-recruitment studies should be interpreted in light of

habitat patchiness.

12

1.2 Introduction

Dispersal drives population dynamics, range limits, and local adaptation, and

can thereby enhance ecosystem resilience (Roughgarden et al. 1988; Gaston 1996;

Nyström & Folke 2001; Lenormand 2002). Dispersal also sets the spatial scale of

ecological and evolutionary processes, and therefore determines the relative

importance of local and regional forces within ecosystems. In the ocean, marine

currents and long pelagic larval stages for most organisms create a high potential for

long-distance dispersal, despite relatively sedentary adults (Scheltema 1986;

Roughgarden et al. 1988; Mora & Sale 2002). High levels of genetic similarity across

wide ocean distances support this view of open populations (Palumbi 1992; Mora &

Sale 2002), where individual populations primarily receive recruits from other

populations rather than from themselves (Jones et al. 1999). However, recent tagging

evidence for short-distance larval dispersal (Swearer et al. 1999; Jones et al. 2005;

Almany et al. 2007; Planes et al. 2009; Saenz-Agudelo et al. 2009) and sharp genetic

breaks in species thought to have high dispersal (Barber et al. 2000; Taylor &

Hellberg 2003) suggest that marine dispersal may instead be surprisingly local. As

fisheries decline and coastal habitats degrade, identifying typical scales of marine

larval dispersal is critical for ecosystem-based management (Sale et al. 2005).

One difficulty in research to date is that most genetic and self-recruitment

analyses only measure a small portion of the dispersing individuals. Many genetic

methods are strongly influenced by rare, long-distance dispersal events (Slatkin 1987;

Waples & Gaggiotti 2006), and these methods are therefore most useful where

migration rates are low. However, rare events over evolutionary timescales may be

irrelevant to current ecological processes and management decisions. On the other

hand, tagging or parentage studies require recapture of individuals or their offspring,

and are therefore conducted over areas of limited extent. One danger of these studies is

that typical dispersal distances may be underestimated (Koenig et al. 1996).

Genetic approaches based on isolation by distance theory may present a middle

ground that estimates ecologically relevant dispersal parameters (Slatkin 1993;

13

Rousset 1997; Palumbi 2003). When sampled over small spatial scales, these genetic

patterns are driven by effective population density and typical dispersal over the past

few generations, and are less affected by evolutionary or rare events (Rousset 1997;

Hardy & Vekemans 1999; Leblois et al. 2004). Highly polymorphic genetic markers

such as microsatellites makes sampling over small spatial scales both possible and

informative (Selkoe & Toonen 2006).

Isolation by distance patterns represent a balance between genetic drift and

dispersal, and strong isolation patterns can therefore result from either strongly limited

dispersal or low effective density. To date, isolation by distance patterns are often

interpreted as evidence that dispersal is limited by distance, but that distance remains

unknown. To understand dispersal distances quantitatively, we need information on

effective population density. While some studies have taken guesses at what effective

densities may be (Kinlan & Gaines 2003; Palumbi 2003; Buonaccorsi et al. 2004),

estimation of effective density in marine species is difficult. Census sizes for many

marine organisms are in the millions, but estimates of effective size in some species

are up to six orders of magnitude smaller (Hedgecock 1994; Hauser et al. 2002;

Árnason 2004; Hoarau et al. 2005). Practical approaches to empirically estimate

effective density are needed for a more accurate understanding of dispersal.

Anemonefish (genera Amphiprion and Premnas) provide a productive system

in which to develop these methods because previous research provides initial

expectations for their dispersal scales. They are also intensely exploited for the

aquarium trade (Shuman et al. 2005) and therefore strong candidates for conservation

within marine protected areas if dispersal scales are known. Larvae hatch after

approximately 7 days from benthic eggs laid adjacent to the parents’ anemone, then

spend 7-11 days in the pelagic ocean (Thresher et al. 1989) before settling onto a host

anemone for the rest of their lives (Fautin & Allen 1992). Genetic studies have

revealed low genetic distance between populations 1000 km apart on the Great Barrier

Reef, and this has been interpreted as 5 migrants per generation dispersing this

distance (Doherty et al. 1995). On the other hand, 25-60% self-recruitment fractions

have also been measured in anemonefish with artificial otolith tags and genetic

14

parentage analysis (Jones et al. 2005; Almany et al. 2007; Planes et al. 2009; Saenz-

Agudelo et al. 2009). These high fractions suggest highly localized dispersal, though

the studies were conducted on relatively isolated islands. These disparate pieces of

evidence for dispersal scales in anemonefish have appeared difficult to reconcile.

In this study, we searched for isolation by distance patterns in a common coral

reef fish (A. clarkii) and take multiple approaches to estimate effective density. We

then used these estimates to derive more robust estimates of larval dispersal scales

than have been available previously. Finally, we determined whether our dispersal

estimates were consistent with long-distance or local dispersal.

1.3 Materials and Methods

1.3.1 Study system

Clark’s anemonefish (A. clarkii) is a species distributed throughout the Indo-

Pacific. Along the islands of Cebu and Leyte in the central Philippines, populations are

relatively continuous at the scale of kilometers, except where coral reefs are disrupted

by sandy sediment near major river outflows. Currents in this region reverse with the

seasonal monsoons, flowing primarily northwards along each coast during the

northeast monsoon in January and primarily southwards in August with the southwest

monsoon (USAID 2007). Based on an oceanographic model of the region, currents are

likely weakest in Ormoc Bay (< 15 cm/s) on the west coast of Leyte and strongest in

the shallow water between Bohol and Leyte (up to 100 cm/s) (USAID 2007).

Our study sites were 25 km apart along the east coast of Cebu (n = 10) and the

west coast of Leyte (n = 8) (Fig. 1.1). The two coastlines were chosen as replicates to

examine common processes affecting dispersal. We intentionally designed our

sampling over narrower spatial scales (223-252 km of coastline) than most marine

dispersal studies. Genetic differentiation of samples that are close together spatially

are more likely to represent recent rather than past migration rates because time to

equilibrium is shorter (Slatkin 1993; Hardy & Vekemans 1999).

15

1.3.2 Ecological surveys

Census density of A. clarkii along coral reef coastlines was measured with

underwater visual transects while on SCUBA during August-October 2008. Visual

transects were swum parallel to the fringing reef. Two divers recorded the number and

size of each anemonefish on each anemone in two 5 m swaths that were randomly

located from 3 to 12 m deep. The two largest fish per anemone were considered the

breeding adults if they were at least 8 cm long (Ochi 1989). One diver towed a GPS

unit that recorded position every 15 seconds in order to precisely measure the length

of each transect path.

Transects were located exactly 25 km apart on each island, with locations

chosen by GPS prior to visiting the site (n = 10 on Cebu and n = 6 on Leyte).

Transects were not relocated if habitat was poor to ensure that we estimated an

unbiased, mean coastal density (D). Each transect was on average 71 ± 8 minutes and

655 ± 51 m long, for a total area surveyed of 111,000 m2. We chose to conduct fewer

but longer transects so as to average over small-scale spatial variability. The surveys

covered 1/40th of the length of the 475 km study area.

Area of each transect was calculated in ArcGIS 9.2 (ESRI, Redlands, CA)

from the GPS tracks. We multiplied census densities (fish per m2) by reef width to

calculate linear fish density (fish per km). We measured reef width at our sampling

sites from satellite photos in Google Earth.

In addition, two to six additional “reef surveys” were conducted during sample

collection dives in the vicinity of each study site specifically on high quality coral reef

habitat in Cebu (n = 27) and Leyte (n = 12). The underwater protocol for these reef

surveys was the same as for the census transects described above. These reef surveys

covered an additional 160,000 m2 of reef, spanning in total about 1/15th of the length

of our study area. When analyzing these reef surveys, however, we had to account for

the fact that they represented density on coral reef habitats rather than coast-wide

density. Therefore, we used ArcGIS and the Reefs at Risk coral reef map (Burke et al.

2002) to calculate the area of Cebu and Leyte’s coastline covered by reef habitat. We

then multiplied reef area (m2) by the density of fish on reef habitats (fish per m2) and

16

divided by coastline length to get linear fish density (fish per km). This is the same

approach as applied by Puebla et al. (2009).

Finally, we compared our density estimates to those in the literature to ensure

that we were not greatly under- or over-estimating typical A. clarkii densities.

1.3.3 Genetic samples

We collected non-lethal finclips underwater from A. clarkii in August-October

2008 after capturing fish with dip and drive nets at our sampling sites. Sampling was

conducted at nearby locations if few or no fish were present at the precise sampling

site. The first twenty fish of any size were sampled, and samples were stored in 70%

ethanol. Size of each specimen was recorded to the nearest cm and location was

marked by GPS.

We extracted DNA from all samples with Nucleospin (Machery-Nagel,

Bethlehem, PA) or DNEasy 96 (Qiagen, Valencia, CA) column extraction kits. We

amplified and genotyped 13 microsatellite loci (Table 1.1). Two loci were found

through cross-species amplification of loci screened by Beldade et al. (2009), though

not published by them. These loci were B6 (F: 5’-3’ TGTCTTCTCCCCAAGTCAG,

R: 5’-3’ ACGAGGCTCAACATACCTG) and C1 (F: 5’-3’

GCGACCTTGTTATCACTGTC, R: 5’-3’ TTGGTTGGACTTTCTTTGTC).

Final concentrations in 10 µl PCR reactions were 1 µl genomic DNA, 1x

Fermentas PCR buffer, 3mM MgCl2, 500 nM fluorescently labeled primer, 500 nM

unlabeled primer, 40 µM each dNTP, and 0.1 µl (0.5 U) Fermentas Taq. Thermal

cycling consisted of a 94°C denaturing step for 2 minutes, followed by 30 cycles of

94°C for 45 seconds, annealing temperature for 45 seconds (Table 1.1), and 72°C for

45 seconds, followed by a final extension at 72°C for 1 minute. Some loci were

multiplexed using the Type-it Microsatellite PCR kit (Qiagen) and the manufacturer’s

PCR protocol with a 60°C or 57°C annealing temperature. PCR products were

genotyped on an Applied Biosystems 3730 (MRDDRC Molecular Genetics Core

Facility at Children's Hospital Boston) and analyzed in GeneMapper 4.0 (Applied

Biosystems, Foster City, CA). All genotypes were checked by eye.

17

Genotyping error rate was assessed with duplicate, independent PCRs and

genotypes for 11 to 66 samples per locus.

1.3.4 Genetic analysis

We assessed genetic linkage and departure from Hardy-Weinberg Equilibrium

(HWE) in Genepop with 5000 iterations (Rousset 2008). Linkage and HWE were

assessed independently for each locus within each population, then p-values were

combined across populations with Fisher’s method (Sokal & Rohlf 1995). We report

Weir & Cockerham’s FIS estimate (Weir & Cockerham 1984). We calculated FST and

expected heterozygosity (He) in Arlequin 3.11 using the number of different alleles

between multilocus genotypes (Excoffier et al. 2005). We use α = 0.05 as our Type I

error rate throughout and apply Bonferroni corrections where appropriate (Rice 1989).

We assessed the presence of an isolation by distance pattern with a Mantel test

for each island and then calculated a combined p-value across both islands using

Fisher’s method (Sokal & Rohlf 1995). This combination is appropriate because data

from each island independently test the hypothesis of isolation by distance. We used

the smatr package in R 2.8.1 (Warton & Ormerod 2007) to calculate the slope of the

relationship with reduced major axis regression. This method is appropriate when

distance between populations is measured with error (Hellberg 1994; Sokal & Rohlf

1995). We also jackknifed over populations to ensure that one outlier population was

not having a large influence on our slope estimate.

1.3.5 Isolation by distance

Population genetics theory predicts that the balance between drift and

migration in a continuous population will result in a positive correlation of genetic and

geographic distance between samples (Rousset 1997). This relationship is called

isolation by distance. If the organism is distributed in a linear habitat and samples are

taken in discrete locations,

18

(1)

where σ is the spread of the dispersal kernel, De is effective density, and m is

the slope of the relationship between FST/(1-FST) and geographic distance (Rousset

1997). Technically, spread is the standard deviation of parental position relative to

offspring position (Rousset 1997), otherwise known as the standard deviation of the

dispersal kernel. Dispersal spread (σ) can be estimated from Eq. 1 if the slope (m) and

effective density (De) are known. Effective density can be thought of as effective

population size (Ne) divided by the area occupied by this population. The same set of

factors that reduce Ne below census population size (N) (Frankham 1995) also reduce

De below census density (D) (Watts et al. 2007).

Isolation by distance theory is built on a Wright-Fisher model of reproduction,

assumes no selection, and assumes that the population is at drift-migration

equilibrium. The one-dimensional formula used here is appropriate when the length of

the habitat is greater than the width (Rousset 1997). This assumption seems

appropriate on our study reefs, which are hundreds of kilometers long and only

hundreds of meters wide.

1.3.6 Effective density from temporal method

Methods to estimate effective density from genetic data are not readily

available for continuous populations, though continuous populations are common in

the natural world. We take two, independent approaches to estimating effective

density in this paper. Our first approach uses temporal genetic change, while our

second approach is derived from census density (D).

The change in allele frequencies between cohorts contains information about

the effective size of the population, but this information can be confounded by

migration from surrounding populations. In general, allele frequencies will become

more similar to the source population if migration is strong, while frequencies will

19

change independently of the source if drift is strong. The pseudo-maximum likelihood

method of Wang and Whitlock (2003) uses this information to estimate effective size

independently from immigration rates. Their method considers both temporal changes

in gene frequencies in a focal population and the gene frequencies in a source

population from which migrants arrive. The method assumes that there is no selection

and negligible mutation, that the source population can be identified, that gene

frequencies in the source population are stable, and that sampling does not impact the

availability of reproductive individuals (Wang & Whitlock 2003).

For the estimate of effective size as implemented in MNe 2.0 (Wang &

Whitlock 2003), we defined two cohorts of A. clarkii: the breeding adults (largest pair

on each anemone if ≥8cm) and the juveniles (≤6cm). A. clarkii grows to 6 cm in 2-3

years, reaches reproductive size at 5-6 years, and is known to live as long as 11 years

(Moyer 1986; Ochi 1986). We therefore assume these cohorts are parental and

offspring samples about one generation apart, though individual pairs of fish may be a

bit closer or further apart in age.

To define the source population, we first combined all non-focal samples

because differentiation between populations was low (“MNe-All”). As an alternative

definition of the source, we used the two populations flanking the focal population

(“MNe-Flanking”). In each case, we repeated the calculation separately with each

sampling site as the focal site. We report the median effective size across sites and

bootstrap percentile confidence intervals from 10,000 resamples with replacement

(Davison & Hinkley 1997).

These two approaches with MNe gave us estimates of local effective

population size. These estimates of effective size excluded the fish in other

populations centered 25 km and more in each direction. Therefore, we assumed that

the spatial extent of each local population extended halfway to each flanking

population (12.5 km in each direction). We converted local effective size to effective

density by dividing the effective size by the spatial extent of each local population (25

km).

20

1.3.7 Effective density from census density

A number of factors reduce effective size below census size, including

fluctuations in population size through time, unequal sex ratios, variance in family

size, and variance in reproductive success (Frankham 1995). Marine species may be

most affected by variance in reproductive success because of family-correlated

survival through larval dispersal (sweepstakes recruitment) and age-related increases

in fecundity and offspring survival (Hauser et al. 2002; Hedrick 2005). Variance in

reproductive success is strongly affected by the mating system, number of mates,

fecundity, and longevity of a species, as well as the environmental variability a

population experiences (Clutton-Brock 1988). Because anemonefish have similar

mating systems and number of mates (permanent pair bonds, (Fautin & Allen 1992)),

similarly high fecundity (thousands to tens of thousands of eggs per year, (Richardson

et al. 1997)), similar lifespans (around a decade, (Fautin & Allen 1992)), and

experience relatively low environmental variability in tropical climates, we expect that

variance in reproductive success will be similar across anemonefish species.

To estimate variance in reproductive success for Amphiprion, we analyzed

genetic parentage studies by Jones et al. (2005) in A. polymnus and Planes et al. (2009)

in A. percula. These studies found that 15 of 33 or 77 of 270 potential breeding pairs

(respectively) produced locally recruiting larvae in 23 or 108 larvae sampled (Jones et

al. 2005; Planes, pers. comm.). Since observations of non-locally recruiting offspring

were not made, we needed to consider the limited sampling of these studies. We did

this by simulating breeding pairs (33 or 270) whose reproductive success was modeled

by a negative binomial distribution with a mean of two offspring per breeding pair (a

stable population) and a variance that we fit to the data. While a Poisson distribution is

often used when all parents have the same probability of reproducing, the negative

binomial can allow probability of reproducing to vary among parents (Bolker 2008).

From each simulation, we selected a subsample of the offspring (23 or 108) and

compared the number of parents represented in this sample to the number of parents

observed by Jones et al. or Planes et al. (15 or 77). We conducted 10,000 simulations

21

for each of two hundred reproductive variance values between 2 and 100 on a log10

scale and selected the variance most likely to reproduce the observed values.

Equation 2 in Hedrick (2005) allows us to calculate the ratio of effective (Ne)

to census (N) population size as

(2)

where Vk is the variance in reproductive success. We then calculate De as

(3)

1.3.8 Uncertainty

Each step of our calculations contained a certain degree of uncertainty. Some

uncertainty resulted from uncertainty in parameter estimation, while other resulted

from uncertainty in which method to use to estimate effective density. We used both

temporal genetic and census methods to explore methodological uncertainty, and we

propagated parameter uncertainty through all of our calculations with bootstrap

resampling. To do the latter, we sampled the parameter values in Eqs. 1 and 3 from

probability distributions that reflected the uncertainty for each parameter (m, De, D, or

Ne/N) and repeated this 10,000 times. We used normal distributions for D and Ne/N,

but used a lognormal distribution for m because confidence intervals from reduced

major axis regression are asymmetric. For the MNe estimates of De, we sampled from

the bootstrapped distribution of medians (see above). In addition, we replicated our

calculations independently on two islands to examine the reliability of our results.

22

1.4 Results

1.4.1 Adult density

Our surveys revealed coast-wide A. clarkii densities on Cebu (0.53 ± 0.16

fish/100 m2, n = 10) that were similar to densities on Leyte (0.55 ± 0.45 fish/100 m2, n

= 6), though Leyte densities were more variable and included numerous zeros (Fig

1.2). Leyte’s reefs are often found as small patches, and some surveys landed on sandy

habitat. The densities that we observed do not appear unusual for A. clarkii (Table

1.2). A literature search revealed mean (1.1 ± 0.47 fish/100 m2, n = 7) and median

densities (0.36 fish/100 m2, n = 7) that were similar to our observations.

Adults made up 13-66% of the fish in each survey, and adult densities were

higher on Cebu (0.21 ± 0.071 adults/100 m2, n = 10) than on Leyte (0.096 ± 0.064

adults/100 m2, n = 6). From satellite photos, we estimated that reefs in our study

region were approximately 150 m wide. Therefore, mean density (D) of A. clarkii in

Cebu was 317 ± 107 adults/km and 144 ± 96 adults/km in Leyte.

As an estimate of census density with a larger sample size, we also analyzed all

of our reef surveys on Cebu (0.25 ± 0.061 adults/100 m2, n = 27) and Leyte (0.16 ±

0.056 adults/100 m2, n = 12). We also estimated reef area to be 47 km2 along the 252

km east coast of Cebu and 32 km2 along the 223 km west coast of Leyte. Our

calculations therefore suggested a mean linear density from our reef surveys of 457 ±

114 adults/km (Cebu) and 231 ± 80 adults/km (Leyte). The coast-wide densities

calculated from reef surveys were higher than those from our census transects, but not

significantly so (p > 0.38). We would expect these densities to be higher because the

reef surveys were biased towards high quality coral reef habitats. In addition, the error

bounds on the reef surveys were of similar width to those from the census transects

despite twice the sample size, suggesting that additional survey effort would not

greatly reduce uncertainty in census density. Because the reef surveys are likely biased

high, and because the extra computational steps required to analyze the reef surveys

likely introduces additional error and bias that is difficult to quantify (particularly in

23

the calculation of reef area), we use our census transects for all further calculations of

effective density.

1.4.2 Genetic analysis

Among 369 A. clarkii samples (Table 1.3) genotyped at 13 microsatellite loci,

the number of alleles per locus ranged from 3 to 18 (mean: 9.5) (Table 1.1). None of

the loci showed significant departure from HWE after combining p-values across

populations (p > 0.052 for all loci), though 13 of the 234 locus-by-population

comparisons (5.6%) were significant. Only one of 78 locus pairs (1.3%) showed

significant linkage (APR_Cf8 and NNG_028, p = 0.047), but this is likely due to the

number of comparisons we made rather than actual linkage. The genotyping error rate

was zero for eleven loci, and 3.2% or less for the remaining two loci (Table 1.1).

Expected heterozygosity ranged from 0.102 to 0.890 with a mean of 0.615 ± 0.067.

FSTs between any two sites were low (< 0.028) and 17 of 153 pair-wise

comparisons (11%) were significant (0.05 > p > 0.002). While more than the 5% of

comparisons expected to be significant by chance, none of these comparisons

remained significant after Bonferroni corrections. One interpretation of these data

would be to conclude that gene flow across the study area is common and that, if any

restrictions to gene flow exist, they are weak and not detectable with the current

sampling design.

However, as predicted by a drift-migration balance in continuous populations

with distance-limited dispersal, a positive relationship between genetic and geographic

distance was observed on both Cebu and Leyte with a combined p-value of 0.009 (Fig

1.3). The slope (m) in Leyte (1.89 x 10-4, 95% CI: 1.60 x 10-4 - 2.23 x 10-4) was higher

than that in Cebu (0.847 x 10-4, 95% CI: 0.630 x 10-4 - 1.14 x 10-4). To ensure that one

outlier population was not having a large influence on our slope estimates, we

jackknifed over populations. Jackknifed mean slopes were slightly steeper that linear

model slopes for both Leyte (2.04 x 10-4 ± 0.819 x 10-4) and Cebu (0.908 x 10-4 ±

0.256 x 10-4), but were well within the 95% CI for the original estimates. Similarly,

24

removing the two loci with non-zero error rates (Cf29 and 65) led to somewhat steeper

slope estimates (2.2 x 10-4 in Leyte and 0.92 x 10-4 in Cebu).

1.4.3 Temporal estimates of effective density

As our first approach to calculating effective density, we estimated population

size for single sampling sites with our MNe-All method (the source defined as all non-

focal populations). MNe failed while profiling 95% confidence intervals for some

sampling sites, potentially due to low sample sizes, but still reported the maximum

likelihood estimates of effective size (Ne) in all cases. Median effective size across

sampling sites was 92 in Cebu (95% CI: 77 – 196) and 101 in Leyte (95% CI: 54 –

286). Based on 25 km of reef between sampling sites, this would be equivalent to a De

of 3.7 (Cebu, 95% CI: 3.1 – 7.8) or 4.0 (Leyte, 95% CI: 2.2 – 11) adults/km.

Because MNe is sensitive to misidentification of the source population of

immigrants (Wang & Whitlock 2003), we also reran the analysis while defining the

source as the two populations flanking each focal sampling site. The Ne estimates for

MNe-Flanking were generally larger, with a median size of 526 in Cebu (95% CI: 330

– 2360) and 327 in Leyte (95% CI: 150 – 5020). These higher sizes suggested higher

De of 21 (Cebu, 95% CI: 13 – 94) or 13 (Leyte, 95% CI: 6 – 200) adults/km.

1.4.4 Effective density from census density

Our second approach to estimating effective density was to consider previously

published information on reproductive success in Amphiprion and our observed census

densities on Cebu and Leyte. Our simulations of reproductive variance revealed that a

variance of 4.3 was most likely to produce the Jones et al. (2005) observations in A.

polymnus, while a variance of 6.0 was most likely to produce the Planes et al. (2009)

observations in A. percula. Using equation 2 suggested Ne/N ratios in Amphiprion of

0.63 or 0.50, with a mean of 0.57 ± 0.065.

Therefore, considering uncertainty in both D and Ne/N, our demographic

estimates of effective density are 178 ± 64 (Cebu) and 82 ± 55 (Leyte) adults/km.

25

1.4.5 Dispersal distance

By using Eq. 1 and our isolation by distance slopes on Cebu and Leyte, we

could define the relationship between effective density and dispersal spread (the

diagonal lines in Fig 1.4), but we could not calculate dispersal spread (σ) without

knowing De. Our three estimates of De on each island provided us with six estimates

of dispersal spread (Fig. 1.4). From MNe-All, we estimated a dispersal spread of 27

km (Cebu, 95% CI: 20 - 33) or 18 km (Leyte, 95% CI: 10 - 25). Our estimates from

MNe-Flanking were lower because the effective density estimates were higher: 12 km

(Cebu, 95% CI: 6 – 16) and 10 km (Leyte, 95% CI: 2.5 - 15). Finally, our

demographic estimates of dispersal were the lowest at 4.1 km (Cebu, 95% CI: 2.9 –

7.2) and 3.9 km (Leyte, 95% CI: 2.6 – 12.4). Our estimates of dispersal were generally

higher in Cebu because the observed isolation by distance slope was shallower than on

Leyte, though this difference was partially compensated by the higher effective

density on Cebu.

Comparing the 95% confidence intervals around each estimate to the

differences between estimates, it became clear that the greatest source of uncertainty

was not in any one parameter’s estimate, but rather in which method to use to

calculate effective density. Our final range of dispersal spread estimates spanned a

factor of seven (4 to 27 km with median 11 km), reflecting remaining uncertainty in

the effective density of A. clarkii.

1.5 Discussion

Understanding dispersal scales in many organisms has been notoriously

difficult, and our research demonstrates that increased attention to effective density

can aid in the estimation of dispersal spread. Across two replicate coastlines of 220-

250 km in the Philippines, we found isolation by distance patterns in populations of A.

clarkii. These data suggest that dispersal distances are less than 220-250 km, but such

patterns cannot easily be compared to high self-recruitment rates in anemonefish. In

addition, we used multiple approaches to measure dispersal scale by estimating

26

effective density. We observed temporal shifts in allele frequencies between adult and

juvenile cohorts that suggested effective densities of 4-21 adults/km. Census densities

of adults in our study area and low variance in reproductive success in Amphiprion

implied that effective density was perhaps as high as 82-178 adults/km. Using these

estimates of effective density with isolation by distance theory suggested that A.

clarkii dispersal spread is in the range of 4-27 km (median 11 km).

1.5.1 The central role of effective density and census density

Effective density is a central concept in the population genetics of continuous

populations because it is needed to convert isolation by distance signals into dispersal

estimates. However, little attention has been paid to the estimation of this quantity. By

collecting genetic data from multiple cohorts of A. clarkii along with ecological census

data, we were able to develop two independent estimates of effective density.

The first approach is based on shifts in genetic composition from generation to

generation using an explicit model that permits some local retention from a local

population as well as input from surrounding populations (Wang & Whitlock 2003).

The second is based on census density, because effective population sizes are typically

lower than current census sizes (Frankham 1995). For marine species such as cod,

snapper, plaice and oysters, evidence suggests that effective density is lower than

census density by five to six orders of magnitude (Hedgecock 1994; Hauser et al.

2002; Hoarau et al. 2002; Árnason 2004). If this were true for anemonefish, census

density would provide an upper bound that was much too high, and as a result, our

dispersal estimate would be much too low.

However, anemonefish occupy individual, easily observed, breeding habitats,

and adult density is the density of breeding pairs. This fact brings effective and census

densities into closer alignment. Data from parentage studies confirm this assumption

for anemonefish (A. polymnus and A. percula), where 29-45% of parents produced

local offspring. This observation suggested that about half of the census density might

provide a reasonable effective density estimate. Even using overall census density as

an upper bound would provide an informative guideline for these species.

27

Without empirical estimates of effective density, the range of possible effective

densities in marine species is extremely large, and simple assumptions about effective

density could be dramatically incorrect. The methods we proposed with A. clarkii

narrowed this uncertainty considerably and allowed us to estimate dispersal distance

within an order of magnitude. Even with the remaining uncertainty, this is a

substantial improvement over previous knowledge.

1.5.2 Effective density from genetic diversity?

Beyond the temporal genetic and census data that we used in this paper, it may

also be possible to estimate effective density from genetic diversity. For example,

Puebla et al. (2009) proposed a method that used the program MIGRATE to estimate

effective population size, and then divided population size by reef length to estimate

density. However, this method assumes the island model of migration and requires

discrete and isolated populations within which isolation by distance processes do not

affect genetic diversity. These criteria are often difficult to meet in widely dispersing

marine species.

An alternative approach for estimating effective density may come from the

population genetic theory for continuous populations. For example, Wright showed

that in a continuous population Ne is affected not only by the number of individuals in

the population, but also by the size of what he called the genetic neighborhood

(Wright 1969). The neighborhood refers to the number of adults from which an

individual’s parents can be treated as if drawn at random. Wright provides equations

for the effective size, length, and neighborhood size of one-dimensional, continuous

populations (Wright 1969, pp. 298 and 302). These equations can be combined to

show that:

(4)

28

where a varies from about 1.5 to 3.5 depending on the shape of the dispersal

kernel, and k is the length of habitat occupied by the population. Effective population

size (Ne) can be estimated from genetic diversity (e.g., Ne = θ/(4µ), where θ is a

measure of genetic diversity and µ is mutation rate). Eq. 4 may be most useful for

estimating bounds on effective density (De), as it relies on specifying a maximum or

minimum dispersal spread (σ). However, the scale of analysis is crucial, and it is

difficult to know what value to assume for k. If we use the diversity in our data set

measured over 475 km of coastline and assume that σ < 475 km, then De > 1.6

adults/km. If we instead assume that the diversity we observe applies across the

Philippines (k ≈ 24,000 km), then De > 0.23 adults/km. Estimating local θ in the

context of an isolation by distance model based on empirical data would be a

valueable topic for future research and theory.

1.5.3 Assumptions of dispersal calculations

A major assumption of our calculations is that the isolation by distance pattern

has reached a stationary phase close to drift-migration equilibrium under the current

demographic parameters. This stationary phase is reached within a few generations for

populations separated by 10σ, but may take tens or hundreds of generations for

populations separated by 100-1000σ (Hardy & Vekemans 1999; Vekemans & Hardy

2004). If A. clarkii spatial genetic patterns are not yet stationary, they are likely

becoming stronger with time because A. clarkii is exploited for the aquarium trade in

the Philippines and its density has likely declined as a result (Shuman et al. 2005).

Given the relatively small spatial scale of our study and the 5-10 yr generation time in

A. clarkii (see Methods), it appears that our estimate of dispersal should reflect an

ensemble average over the last few decades or perhaps century of dispersal. This

period is likely an ecological timescale relevant to ecology, conservation, and

management.

Isolation by distance estimates of dispersal also assume that effective density is

constant across space (Leblois et al. 2004). In reality, densities vary at both large and

small spatial scales. Simulations by Leblois revealed that isolation by distance patterns

29

can be biased upwards if small sampling areas are immediately surrounded by a lower

density (Leblois et al. 2004), as might occur if coastlines are scouted for high

population densities and only sampled in those locations. We did not select our study

areas based on high density, and therefore do not expect this to be a problem. We are

not aware of analyses that examine the effects of spatial variation in effective density

within a study region.

In addition, we assume that demography, not selection, drives patterns of

genetic differentiation. Hardy-Weinberg Equilibrium at all of our loci and consistent

patterns of isolation by distance across multiple loci support this assumption (data not

shown). While the large variation in allelic richness across loci (3 to 18) suggests that

mutation rate may vary among the loci we examined, mutation rate does not strongly

affect isolation by distance patterns unless rates are much higher or lower than typical

microsatellites (Leblois et al. 2003).

1.5.4 Comparison to other measures of larval dispersal

Previous evidence for larval dispersal in Amphiprion appeared contradictory

because separate studies reported both relatively high effective migrant exchange (5

migrants/generation) between populations 1000 km apart on the Great Barrier Reef

(Doherty et al. 1995) and high fractions of self-recruitment (30-60%) to small, local

reefs (Jones et al. 2005; Almany et al. 2007). In comparison, the dispersal kernel that

we estimated predicts an effectively zero probability of any larvae traveling 1000 km.

However, many larvae moving shorter distances over multiple generations (stepping-

stone dispersal) can produce relatively low genetic divergence over large distances. In

fact, extrapolating the isolation by distance pattern that we observed to 1000 km

predicts that FST should equal 0.03 at this distance. This low predicted FST matches

well to the FST of 0.05 observed by Doherty et al. (1995), though our interpretation

based on an isolation by distance framework differs from their island model

calculation.

When compared to observations of high self-recruitment, our estimated

dispersal kernel for A. clarkii does not at first appear compatible. Typical dispersal of

30

4-27 km appears unlikely to provide 30-60% self-recruitment. To investigate this

further, we simulated larval dispersal across a continuous habitat as a Gaussian

random number with the median dispersal spread calculated in our study (11 km) (Fig.

1.5a). We then measured self-recruitment to a 500 m section of reef, which is similar

in size to those studied for Amphiprion self-recruitment (Jones et al. 2005; Almany et

al. 2007). Under this model, we found that only 2% of arriving larvae on a continuous

reef were born by parents on the same reef (2% self-recruitment) (Fig. 1.5a). The self-

recruitment rate was similarly low for Laplacian dispersal kernels.

1.5.5 Reef patchiness and comparison to self-recruitment

While our results were not compatible on continuous habitats, another

possibility is that reef patchiness may strongly influence self-recruitment. Most self-

recruitment studies have been conducted on small habitat patches with the nearest

population more than 10 km away (Jones et al. 2005; Almany et al. 2007). We tested

the idea that reef patchiness is important by simulating larval dispersal as above, but

used 500 m patch reefs spaced every 10 or 15 km in place of a continuous reef. In this

patchy environment, self-recruitment rose to 36% or 56% (10 or 15 km spacing,

respectively) (Fig. 1.5b). Levels of self-recruitment similar to this have been measured

for a number of reef fish in habitats that are patchy at this spatial scale (Jones et al.

1999; Jones et al. 2005; Almany et al. 2007).

This high self-recruitment fraction in patchy habitats results from a low influx

of non-local larvae, not from large numbers of larvae remaining on local reefs. We

note that self-recruitment as measured by Jones et al. (2005) and similar studies

(percent of recruiting larvae that are from local parents) is different from local

retention (percent of dispersing larvae that recruit locally) (Botsford et al. 2009). Our

finding suggests that habitat patchiness may play an important role in creating high

self-recruitment but low local retention (Fig. 1.5c). This hypothesis should be testable

by conducting self-recruitment studies in both continuous and patchy habitats.

A number of other explanations for this discrepancy are possible. Some

authors have suggested that marine fish larvae may have a bimodal strategy in which

31

some larvae are actively retained while others passively disperse (Armsworth et al.

2001). This possibility could be represented by a strongly leptokurtic dispersal kernel

with a very strong mode at zero distance and long tails away from the parents.

Another possibility is that dispersal spread (σ) varies dramatically among

species, from A. clarkii (this study) to A. polymnus (Jones et al. 2005) and A. percula

(Almany et al. 2007). Alternatively, Amphiprion dispersal spread might vary between

regions, from the Philippines (this study) to Papua New Guinea (Jones et al. 2005;

Almany et al. 2007), perhaps as a result of larval behavior or oceanographic currents.

Using our continuous habitat model above, dispersal spread would have to be very low

(600 m) for self-recruitment to reach 30%, and even lower (300 m) to reach 60% self-

recruitment. These dispersal spreads are one to two orders of magnitude below our

estimates of A. clarkii dispersal spread in the Philippines, and such strong variation

among species or regions appears unlikely.

We suggest that high fractions of self-recruitment in patchy habitats as well as

regular dispersal to surrounding reefs are both consistent with a single larval dispersal

strategy and do not require dispersal kernels to change shape dramatically either

among Amphiprion species or among study regions.

1.6 Conclusions and future directions

In the future, we predict that greater attention to effective density will provide

more robust estimates of dispersal and greater ability to interpret isolation by distance

patterns. Our study suggests that two additions to typical genetic sampling can aid in

the estimation of effective density. First, samples from two (or more) distinct cohorts

can be used to examine temporal changes in allele frequencies with MNe and derive

point estimates of effective density. Second, ecological surveys can put an upper

bound on effective density. These additions entail more field effort, but the gain is an

enhanced ability to understand dispersal.

Going forward, there is a clear need for further development of effective

density methods relevant to continuous populations. Our study showed that remaining

32

uncertainty in dispersal distances derives largely from differences among methods for

calculating effective density rather than from uncertainty in parameters. New theory or

simulations that indicate the most appropriate methods or suggest new methods would

be quite useful at this point. As discussed above, genetic diversity may provide

insights into effective density if certain challenges can be resolved. In addition, the

intercept of the isolation by distance pattern may contain important but rarely used

information on effective density if one can make assumptions about the shape of the

dispersal kernel (see Rousset 1997).

Understanding ecological scales of dispersal in a wide range of organisms has

been complicated by methods that focus on exceptional rather than typical dispersers,

but isolation by distance approaches can address this problem when effective density

is estimated. In A. clarkii, our median dispersal spread estimates of 11 km appear

consistent with high self-recruitment rates if habitat patchiness is considered. Our

estimates of dispersal spread suggest that marine reserves for anemonefish would need

to be ten or more kilometers wide to be self-sustaining (Lockwood et al. 2002), or

integrated in dense marine reserve networks (Kaplan et al. 2006; Gaines et al. 2010).

Further efforts to integrate multiple sources of information on dispersal, such as

studies that combine both isolation by distance and parentage methods, will continue

to improve our understanding of dispersal.

As populations of many species continue to decline, accurate measurement of

dispersal distances will aid in effective management and conservation. Isolation by

distance genetic studies that account for the effective density of populations can

provide this important information.

1.7 Acknowledgments

MLP gratefully acknowledges support from a NSF Graduate Research

Fellowship, a NDSEG Fellowship, an International Society for Reef Studies/Ocean

Conservancy grant, a Stanford Biology SCORE grant, and a Myers Oceanographic

Trust grant. The authors also want to thank the Project Seahorse Foundation, the

Coastal Conservation and Education Foundation, G. Sucano, and A. Vailoces for

33

critical assistance in the field. G. Bernardi generously shared primer sequences for us

to screen. R. Waples, S. Planes, and an anonymous reviewer provided helpful

comments on earlier versions of the manuscript.

1.8 References

Almany G.R., Berumen M.L., Thorrold S.R., Planes S. & Jones G.P. (2007). Local

replenishment of coral reef fish populations in a marine reserve. Science, 316,

742-744.

Armsworth P.R., James M.K. & Bode L. (2001). When to press one or turn back:

dispersal strategies for reef fish larvae. Am. Nat., 157, 434-450.

Árnason E. (2004). Mitochondrial cytochrome b DNA variation in the high-fecundity

Atlantic cod: trans-Atlantic clines and shallow gene genealogy. Genetics, 166,

1871-1885.

Barber P.H., Palumbi S.R., Erdmann M.V. & Moosa M.K. (2000). A marine Wallace's

line? Nature, 406, 692-693.

Beldade R., Holbrook S.J., Schmitt R.J., Planes S. & Bernardi G. (2009). Isolation and

characterization of 8 polymorphic microsatellite markers from the orange-fin

anemonefish, Amphiprion chrysopterus. Conservation Genetics Resources, 1,

333-335. doi:10.1007/s12686-009-9077-9

Bolker B.M. (2008). Ecological models and data in R. Princeton University Press,

Princeton.

Botsford L.W., White J.W., Coffroth M.-A., Paris C.B., Planes S., Shearer T.L.,

Thorrold S.R. & Jones G.P. (2009). Connectivity and resilience of coral reef

metapopulations in marine protected areas: matching empirical efforts to

predictive needs. Coral Reefs, 28, 327-337.

Buonaccorsi V.P., Westerman M., Stannard J., Kimbrell C., Lynn E. & Vetter R.D.

(2004). Molecular genetic structure suggests limited larval dispersal in grass

rockfish, Sebastes rastrelliger. Mar. Biol., 145, 779-788.

Burke L., Selig E.R. & Spalding M. (2002). Reefs at Risk in Southeast Asia. World

Resources Institute, Washington, DC.

34

Buston P.M., Bogdanowicz S.M., Wong A. & Harrison R.G. (2007). Are clownfish

groups composed of close relatives? An analysis of microsatellite DNA

variation in Amphiprion percula. Mol. Ecol., 16, 3671-3678.

Clutton-Brock T.H. (1988). Reproductive success: studies of individual variation in

contrasting breeding systems. The University of Chicago Press, Chicago.

Davison A.C. & Hinkley D.V. (1997). Bootstrap methods and their application.

Cambridge University Press, New York.

Doherty P.J., Planes S. & Mather P. (1995). Gene flow and larval duration in seven

species of fish from the Great Barrier Reef. Ecology, 76, 2373-2391.

Elliott J.K. & Mariscal R.N. (2001). Coexistence of nine anemonefish species:

differential host and habitat utilization, size and recruitment. Mar. Biol., 138,

23-36.

Excoffier L., Laval G. & Schneider S. (2005). Arlequin (version 3.0): an integrated

software package for population genetics data analysis. Evolutionary

Bioinformatics, 1, 47-50.

Fautin D.G. & Allen G.R. (1992). Field guide to anemonefishes and their host sea

anemones. Western Australian Museum, Perth, Australia.

Frankham R. (1995). Effective population size/adult population size ratios in wildlife:

a review. Genet. Res., 66, 95-107.

Frisch A.J. & Hobbs J.-P.A. (2009). Rapid assessment of anemone and anemonefish

populations at the Keppel Islands [electronic resource]: a report to the Great

Barrier Reef Marine Park Authority. In: Townsville, QLD.

Gaines S.D., White C., Carr M.H. & Palumbi S.R. (2010). Designing marine reserve

networks for both conservation and fisheries management. P Natl Acad Sci

USA. doi:10.1073/pnas.0906473107

Gaston K.J. (1996). Species-range-size distributions: patterns, mechanisms and

implications. Trends Ecol. Evol., 11, 197-201.

Hardy O.J. & Vekemans X. (1999). Isolation by distance in a continuous population:

reconciliation between spatial autocorrelation analysis and population genetics

models. Heredity, 83, 145-154.

35

Hattori A. (1994). Inter--group movement and mate acquisition tactics of the

protandrous anemonefish, Amphiphrion clarkii, on a coral reef, Okinawa. Jap.

J. Ichthyol., 41, 159-165.

Hauser L., Adcock G.J., Smith P.J., Bernal Ramírez J.H. & Carvalho G.R. (2002).

Loss of microsatellite diversity and low effective population size in an

overexploited population of New Zealand snapper (Pagrus auratus). P Natl

Acad Sci USA, 99, 11742-11747.

Hedgecock D. (1994). Does variance in reproductive success limit effective

population sizes of marine organisms? In: Genetics and evolution of aquatic

organisms (ed. Beaumont AR). Chapman & Hall New York, pp. 122-134.

Hedrick P.W. (2005). Large variance in reproductive success and the Ne/N ratio.

Evolution, 59, 1596-1599.

Hellberg M.E. (1994). Relationships between inferred levels of gene flow and

geographic distance in a philopatric coral, Balanophyllia elegans. Evolution,

48, 1829-1854.

Hoarau G., Boon E., Jongma D.N., Ferber S., Palsson J., Van der Veer H.W.,

Rijnsdorp A.D., Stam W.T. & Olsen J.L. (2005). Low effective population size

and evidence for inbreeding in an overexploited flatfish, plaice (Pleuronectes

platessa L.). P Roy Soc B-Biol Sci, 272, 497-503.

Hoarau G., Rijnsdorp A.D., Van der Veer H.W., Stam W.T. & Olsen J.L. (2002).

Population structure of plaice (Pleuronectes platessa L.) in northern Europe:

microsatellites revealed large-scale spatial and temporal homogeneity. Mol.

Ecol., 11, 1165-1176.

Jones G.P., Milicich M.J., Emslie M.J. & Lunow C. (1999). Self-recruitment in a coral

reef fish population. Nature, 402, 802-804.

Jones G.P., Planes S. & Thorrold S.R. (2005). Coral reef fish larvae settle close to

home. Curr. Biol., 15, 1314-1318.

Kaplan D.M., Botsford L.W. & Jorgensen S. (2006). Dispersal per recruit: an efficient

method for assessing sustainability in marine reserve networks. Ecol. Appl., 16,

2248-2263.

36

Kinlan B.P. & Gaines S.D. (2003). Propagule dispersal in marine and terrestrial

environments: a community perspective. Ecology, 84, 2007-2020.

Koenig W.D., Van Vuren D. & Hooge P.N. (1996). Detectability, philopatry, and the

distribution of dispersal distances in vertebrates. Trends Ecol. Evol., 11, 514-

517.

Leblois R., Estoup A. & Rousset F. (2003). Influence of Mutational and Sampling

Factors on the Estimation of Demographic Parameters in a ‘‘Continuous’’

Population Under Isolation by Distance. Mol. Biol. Evol., 20, 491-502.

Leblois R., Rousset F. & Estoup A. (2004). Influence of spatial and temporal

heterogeneities on the estimation of demographic parameters in a continuous

population using individual microsatellite data. Genetics, 166, 1081-1092.

Lenormand T. (2002). Gene flow and the limits to natural selection. Trends Ecol.

Evol., 17, 183-189.

Liu S.-Y.V., Yu H.-T. & Dai C.-F. (2007). Eight microsatellite loci in Clark's

anemonefish, Amphiprion clarkii. Mol. Ecol. Notes, 7, 1169-1171.

Lockwood D.R., Hastings A. & Botsford L.W. (2002). The effects of dispersal

patterns on marine reserves: does the tail wag the dog? Theor. Popul. Biol., 61,

297-309.

Mora C. & Sale P.F. (2002). Are populations of coral reef fish open or closed? Trends

Ecol. Evol., 17, 422-428.

Moyer J.T. (1980). Influence of temperate waters on the behavior of the tropical

anemonefish Amphiprion clarkii at Miyake-jima, Japan. Bull. Mar. Sci., 30,

261-272.

Moyer J.T. (1986). Longevity of the anemonefish Amphiprion clarkii at Miyake-Jima,

Japan with notes on four other species. Copeia, 1986, 135-139.

Nyström M. & Folke C. (2001). Spatial resilience of coral reefs. Ecosystems, 4, 406-

417.

Ochi H. (1985). Temporal patterns of breeding and larval settlement in a temperate

population of the tropical anemonefish, Amphiprion clarkii. Jap. J. Ichthyol.,

32, 248-256.

37

Ochi H. (1986). Growth of the anemonefish Amphiprion clarkii in temperate waters,

with special reference to the influence of settling time on the growth of 0-year

olds. Mar. Biol., 92, 223-229.

Ochi H. (1989). Mating behavior and sex change of the anemonefish, Amphiprion

clarkii, in the temperate waters of southern Japan. Environ. Biol. Fishes, 26,

257-275.

Palumbi S.R. (1992). Marine speciation on a small planet. Trends Ecol. Evol., 7, 114-

118.

Palumbi S.R. (2003). Population genetics, demographic connectivity, and the design

of marine reserves. Ecol. Appl., 13, S146-S158.

Planes S., Jones G.P. & Thorrold S.R. (2009). Larval dispersal connects fish

populations in a network of marine protected areas. P Natl Acad Sci USA, 106,

5693-5697.

Puebla O., Bermingham E. & Guichard F. (2009). Estimating dispersal from genetic

isolation by distance in a coral reef fish (Hypoplectrus puella). Ecology, 90,

3087-3098.

Quenouille B., Bouchenak-Khelladi Y., Hervet C. & Planes S. (2004). Eleven

microsatellite loci for saddleback clownfish Amphiprion polymnus. Mol. Ecol.

Notes, 4, 291-293.

Rice W.R. (1989). Analyzing tables of statistical tests. Evolution, 43, 223-225.

Richardson D.L., Harrison P.L. & Harriott V.J. (1997). Timing of spawning and

fecundity of a tropical and subtropical anemonefish (Pomacentridae:

Amphiprion) on a high latitude reef on the east coast of Australia. Mar. Ecol.

Prog. Ser., 156, 175-181.

Roughgarden J., Gaines S.D. & Possingham H. (1988). Recruitment dynamics in

complex life cycles. Science, 241, 1460-1466.

Rousset F. (1997). Genetic differentiation and estimation of gene flow from F-

statistics under isolation by distance. Genetics, 145, 1219-1228.

Rousset F. (2008). GENEPOP'007: a complete re-implementation of the GENEPOP

software for Windows and Linux. Molecular Ecology Resources, 8, 103-106.

38

Saenz-Agudelo P., Jones G.P., Thorrold S.R. & Planes S. (2009). Estimating

connectivity in marine populations: an empirical evaluation of assignment tests

and parentage analysis under different gene flow scenarios. Mol. Ecol., 18,

1765-1776.

Sale P.F., Cowen R.K., Danilowicz B.S., Jones G.P., Kritzer J.P., Lindeman K.C.,

Planes S., Polunin N.V.C., Russ G.R., Sadovy Y.J. & Steneck R.S. (2005).

Critical science gaps impede use of no-take fishery reserves. Trends Ecol.

Evol., 20, 74-80.

Scheltema R.S. (1986). On dispersal and planktonic larvae of benthic invertebrates: an

eclectic overview and summary of problems. Bull. Mar. Sci., 39, 290-322.

Selkoe K.A. & Toonen R.J. (2006). Microsatellites for ecologists: a practical guide to

using and evaluating microsatellite markers. Ecol. Lett., 9, 615-629.

Shuman C.S., Hodgson G. & Ambrose R.F. (2005). Population impacts of collecting

sea anemones and anemonefish for the marine aquarium trade in the

Philippines. Coral Reefs, 24, 564-573.

Slatkin M. (1987). Gene Flow and the Geographic Structure of Natural-Populations.

Science, 236, 787-792.

Slatkin M. (1993). Isolation by distance in equilibrium and non-equilibrium

populations. Evolution, 47, 264-279.

Sokal R.R. & Rohlf F.J. (1995). Biometry. 3rd edn. W. H. Freeman and Company,

New York.

Swearer S.E., Caselle J.E., Lea D.W. & Warner R.R. (1999). Larval retention and

recruitment in an island population of a coral-reef fish. Nature, 402, 799-802.

Taylor M.S. & Hellberg M.E. (2003). Genetic evidence for local retention of pelagic

larvae in a Caribbean reef fish. Science, 299, 107-109.

Thresher R.E., Colin P.L. & Bell L.J. (1989). Planktonic duration, distribution and

population structure of western and central Pacific damselfishes

(Pomacentridae). Copeia, 420-434.

USAID (2007). Establishing Larval Exchange and Reef Connectivity Using Larval

Dispersal Models. In. USAID Cebu City, Cebu.

39

Vekemans X. & Hardy O.J. (2004). New insights from fine-scale spatial genetic

structure analyses in plant populations. Mol. Ecol., 13, 921-935.

Wang J. & Whitlock M.C. (2003). Estimating effective population size and migration

rates from genetic samples over space and time. Genetics, 163, 429-446.

Waples R.S. & Gaggiotti O. (2006). What is a population? An empirical evaluation of

some genetic methods for identifying the number of gene pools and their

degree of connectivity. Mol. Ecol., 15, 1419-1439.

Warton D. & Ormerod J. (2007). smatr: (Standardized) Major Axis estimation and

testing routines. In: R package version 2.1.

Watts P.C., Rousset F., Saccheri I.J., Leblois R., Kemp S.J. & Thompson D.J. (2007).

Compatible genetic and ecological estimates of dispersal rates in insect

(Coenagrion mercuriale: Odonata: Zygoptera) populations: analysis of

'neighborhood size' using a more precise estimator. Mol. Ecol., 16, 737-751.

Watts P.C., Veltsos P., Soffa B.J., Gill A.B. & Kemp S.J. (2004). Polymorphic

microsatellite loci in the black-and-gold chromis, Neoglyphidodon nigroris

(Teleostei: Pomacentridae) Mol. Ecol. Notes, 4, 93-95.

Weir B.S. & Cockerham C.C. (1984). Estimating F-statistics for the analysis of

population structure. Evolution, 38, 1358-1370.

Wright S. (1969). Evolution and the genetics of populations. Volume 2: the theory of

gene frequencies. University of Chicago, Chicago.

40

1.9 Tables

Table 1.1. Microsatellite loci used in Amphiprion clarkii.

Locus Annealing Temp (°C)

# of alleles

He FIS % error (# tested)

Reference

1359 53 7 0.602 0.040 0 (23) (Liu et al.

2007)

1578 58 5 0.439 0.117 0 (63) (Liu et al.

2007)

915 53 4 0.102 -0.012 0 (29) (Liu et al.

2007)

B6 53 14 0.878 -0.009 0 (32) this paper

C1 53 8 0.756 0.002 0 (30) this paper

D1 53 18 0.890 0.015 0 (27) (Beldade et al.

2009)

Cf29 53 18 0.823 -0.038 3.2 (31) (Buston et al.

2007)

Cf8 53 3 0.621 0.034 0 (27) (Buston et al.

2007)

45 58 12 0.695 0.038 0 (66) (Quenouille et

al. 2004)

65 53 12 0.700 -0.017 1.9 (52) (Quenouille et

al. 2004)

LIST12_004 58 3 0.203 0.026 0 (31) (Watts et al.

2004)

LIST12_012 53 4 0.558 0.060 0 (11) (Watts et al.

2004)

LIST12_028 53 15 0.730 0.027 0 (29) (Watts et al.

2004)

41

Table 1.2. Densities of A. clarkii compiled from the literature.

Country Site Density (fish/100 m2)

Reference

Japan Murote Beach, Shikoku Island

3.2 (Ochi 1985)

Japan Miyake-jima 0.25 (Moyer 1980)

Japan Sesoko Island,

Okinawa 0.36 (Hattori 1994)

Philippines Olango inside MPA* 2 (Shuman et al. 2005) Philippines Olango outside MPA 0.15 (Shuman et al. 2005) Papua New

Guinea Madang 1.8 (Elliott & Mariscal 2001)

Australia Keppel Islands 0.0036 (Frisch & Hobbs 2009)

* MPA: Marine Protected Area

42

Table 1.3. Sample sizes of A. clarkii at each site. Sites are listed south to north

on each island. Adults are the two largest fish on each anemone if they are at least 8

cm. Juveniles are defined as fish up to 6 cm.

Site Name Island # adults # juveniles Total samples Santander Cebu 10 5 20 Boljoon Cebu 8 7 20 Argao Cebu 9 8 20 Carcar Cebu 15 4 20

Minglanilla Cebu 12 6 20 Danao Cebu 12 7 20 Sogod Cebu 4 15 21

Tabogon Cebu 13 3 21 Daanbantayan Cebu 10 9 20 Malapascua Cebu 14 4 20

Pintuyan Leyte 11 4 17 Padre Burgos Leyte 11 6 19

Maasin Leyte 11 5 18 Inopacan Leyte 23 6 34 Baybay Leyte 12 6 19 Albuera Leyte 10 7 20

Ormoc City Leyte 11 9 20 Palompon Leyte 14 3 20

Total 369

43

1.10 Figures

Figure 1.1. Cebu and Leyte Islands in the central Philippines. Black dots

indicate study sites. Rectangle in inset map shows location of primary map within the

Philippines.

44

Figure 1.2. Census densities of A. clarkii in Cebu and Leyte, Philippines.

Lines are the adult density (solid with dots) and total density (dashed with dots) at

standard survey sites. Also shown are the total densities on sites chosen for having

high quality coral reefs (x).

45

Figure 1.3. Genetic distance between A. clarkii populations in Cebu and Leyte,

shown with a reduced major axis regression against geographic distance. Cebu: p =

0.11, r2 = 0.04, m = 0.847 x 10-4, 95% CI: 0.630 x 10-4 - 1.14 x 10-4. Leyte: p = 0.01, r2

= 0.31, m = 1.89 x 10-4, 95% CI: 1.60 x 10-4 - 2.23 x 10-4).

46

Figure 1.4. Graph illustrating the calculation of dispersal spread from the slope

of the isolation by distance relationship and estimates of effective density. By knowing

the slope, we can draw the solid lines (Cebu in black, Leyte in grey). By also knowing

the effective density, we can calculate the corresponding dispersal spread (various

dashed and dotted lines). Three estimates of effective density are shown for each

island: MNe-All (dashed), MNe-Flanking (dotted), and a demographic estimate from

census density (dash-dotted).

Effective density (adults/km)

Dis

pers

al s

prea

d (k

m)

0.001 0.01 0.1 1 10 100 1000 10000

0.1

110

100

1000

47

Figure 1.5. Effects of reef patchiness on the self-recruitment fraction. In

continuous habitats (a), many of the recruiting larvae on a small patch of reef (center

of diagram) come from surrounding reefs and the self-recruitment fraction (fraction of

recruiting larvae that are from local parents) is low. In a patchy reef seascape (b), there

are few surrounding reefs from which larvae can arrive and the number of non-local

recruiting larvae will be low. Therefore, the self-recruitment fraction is high.

However, because self-recruitment is a measure of the larvae arriving at a local reef,

the few larvae that self-recruit may in reality be only a small fraction of all larvae that

disperse from a reef, leading to both high self-recruitment and low retention (c).

48

49

Chapter 2

Open and closed seascapes: where does habitat

patchiness create populations with low

immigration?

2.1 Abstract

Which populations are replenished primarily by immigrants and which by local

production remains a fundamental question in ecology with implications for

population dynamics, species interactions, and response to exploitation. However, we

lack methods for predicting population openness. Here, we develop a model for

openness and show that considering habitat isolation elegantly explains the existence

of surprisingly closed populations in high dispersal species, including marine

organisms. Relatively closed populations are expected when patch spacing is more

than twice the standard deviation of a species’ dispersal kernel. In addition, natural

scales of habitat patchiness on coral reefs are sufficient to create both largely open and

largely closed populations. Contrary to previous interpretations, largely closed marine

populations do not require unusually short dispersal distances. We predict that habitat

patchiness has strong control over population replenishment pathways for a wide

range of marine and terrestrial species with a highly dispersive life stage.

2.2 Introduction

Ecologists frequently classify populations as open or closed, depending on

whether they are replenished primarily by immigrants or by local production (Thomas

& Kunin 1999; Hixon et al. 2002). The degree to which a population is open to

regional inputs has important implications for a wide range of ecological and

50

evolutionary processes, including population dynamics, recovery from disturbance,

community assembly, local adaptation, and response to exploitation (Palmer et al.

1996; Roberts 1997; Hixon et al. 2002; Lenormand 2002; Leibold et al. 2004). In each

of these examples, local patterns are more tightly coupled to local processes in closed

populations, but can only be understood in the context of regional processes in open

populations. For example, local recruitment does not depend on local production of

offspring in open populations, making them highly resilient to local disturbance and

exploitation (Roughgarden et al. 1988; Palmer et al. 1996; Roberts 1997).

An important question is therefore which populations are open, which closed,

and which somewhere in between. To date, only qualitative guidelines have existed

for predicting when regional processes should be more or less important. For example,

we expect largely open populations for species with a highly dispersive life stage,

including many plants, wind-dispersed insects, stream-dwelling species, and marine

organisms (Palmer et al. 1996; Thomas & Kunin 1999; Hixon et al. 2002; Mora &

Sale 2002). For many of these species, the defining characteristic is a bipartite life

history with a seed, larval or juvenile stage that disperses much further than relatively

sedentary adults. However, open populations are not expected for all species. For

example, local processes are expected to dominate for species with discrete

populations and little to no dispersal. In addition, the scale of investigation is can be

important (Hixon et al. 2002; Kinlan et al. 2005). At the widest scale of species’

ranges, populations are by definition closed.

Population openness is defined as one minus the probability that an arriving

recruit was born within the population in question, and it can therefore be measured

empirically (Hixon et al. 2002). A number of high-profile studies have revealed that

both direct and indirect measurements of openness do not always match expectations,

particularly in marine species thought to have high dispersal abilities (Jones et al.

1999; Swearer et al. 1999; Taylor & Hellberg 2003; Jones et al. 2005; Almany et al.

2007). These observations have prompted a range of explanations for how largely

closed populations can arise despite long-dispersal dispersal, including natural

selection against immigrants, mortality of dispersing individuals, and physical

51

retention mechanisms (Paris & Cowen 2004; Shanks 2009; Marshall et al. 2010;

Shima et al. 2010). For marine species, one widespread interpretation is that dispersal

distances are much shorter than previously suspected (Warner & Cowen 2002; Levin

2006), and perhaps only hundreds of meters (Shanks 2009). The implicit assumption

in many of these interpretations is that surprisingly closed populations must be

evidence of very short dispersal distances.

What, however, constitutes a “surprisingly closed” population, and what is the

relationship between dispersal abilities and population openness? To date, this

question has been left to qualitative judgment and imprecise terms like high, low,

local, and regional. Surprisingly, we lack the quantitative theory to predict what level

of population openness we should expect in a given situation. Valuable guidance in

this context would set expectations by developing a baseline model against which

observations could be compared. In this way, true deviations from expectations could

more easily be detected. A model would also facilitate increased communication

among empiricists measuring population openness and theoreticians considering

spatial population dynamics, as has been called for repeatedly (Kinlan et al. 2005;

Botsford et al. 2009).

In this paper, we develop a simple model for population openness and use

marine species as an illustrative example because, as mentioned above, recent

empirical papers have highlighted what appear to be surprisingly closed populations in

these species. Our model considers dispersal ability as well as habitat patchiness. This

latter addition facilitates the application of our model to realistic landscapes and

seascapes, and, as we will show, is a critical consideration for openness. Most habitats

exist as patches in a less suitable matrix, including meadows, tree-fall gaps, forests,

mountain tops, riffle-pool arrays, and estuaries (Saunders et al. 1991; Andréfouët et al.

2006). While sometimes overlooked, habitats are also highly patchy in the sea,

including kelp forests, coral and rocky reefs, sheltered bays, and deep-sea vents

(Kritzer & Sale 2006). It remains unclear, however, whether natural scales of habitat

patchiness are likely to have an impact on population-level processes. In general,

patchiness is unimportant if long distance dispersal easily crosses habitat gaps (Wiens

52

1989). Given the perception of long-distance dispersal for some species, it is perhaps

for this reason that habitat patchiness is often overlooked in empirical studies of

marine populations. However, there have yet to be any quantitative comparisons

between dispersal distances and empirical scales of marine habitat patchiness.

Renewed attention to habitat patchiness is also timely given the strong focus

on patchiness in metapopulation and landscape theory, including the effects of patch

number, patch spacing, patch quality, and matrix quality on population persistence and

density (e.g., Bascompte et al. 2002; Hastings & Botsford 2006; Moilanen & Hanski

2006; Figueira 2009; Kaplan et al. 2009; Shima et al. 2010). Similarly, protected area

and marine reserve theory have examined how population survival depends on reserve

size and spacing (Botsford et al. 2001; Lockwood et al. 2002; Gaines et al. 2010;

White et al. 2010). In this body of theory, however, the focus has largely been on

understanding a patch’s contribution to regional processes and persistence. In contrast,

the focus of our paper is on understanding where and when regional processes have

important impacts on local dynamics, and how this dynamic is mediated by landscape

structure. The relative roles of local and regional dynamics are measured by

population openness, and so far, a quantitative treatment of population openness has

been overlooked. From a local manager’s perspective, population openness critically

determines whether local over-harvest or conservation will have future consequences

for local persistence, or whether activities elsewhere will be a more important

consideration.

The goal of this paper, therefore, is to determine the conditions under which

habitat patchiness can create closed populations and ask whether these conditions are

likely to be common in marine ecosystems. To do so, we first develop a simple model

to connect habitat patchiness and dispersal ability to the degree of population

openness. We then examine whether previous empirical studies are likely to fit the

conditions for relatively closed populations. Finally, we ask whether such conditions

are likely to be common in the natural world by using a variety of coral reef seascapes

mapped by remote sensing. We argue that studies measuring self-recruitment or

population openness must be interpreted in the context of habitat patchiness. Some of

53

the apparent conflict over open and closed marine populations may result from

previously underappreciated impacts of habitat patchiness on the source of larvae.

2.3 Materials and Methods

2.3.1 Model

To connect habitat patchiness to population openness, we start from a simple

connectivity metric inspired by metapopulation theory (Moilanen & Hanski 2006). We

calculate the number of immigrants (Si) into patch i:

where D(j,i) is the probability of a larva from patch j settling on patch i, and

N(j) is the number of offspring produced by patch j. We next use the same logic to

define the absolute number of individuals that return to the patch in which they were

born:

Ri = D(i,i)N(i)

where D(i,i) is the probability of returning to patch i.

Following previous convention, we include any mortality that occurs during

dispersal in the dispersal kernel (D), and so D sums to much less than one for most

marine species with high larval mortality (Botsford et al. 2009). The kernel

summarizes the impacts of many ‘biological barriers’ that occur during dispersal,

including predation and starvation that are related to time in the plankton or distance

(Marshall et al. 2010), but does not include any post-settlement mortality that might

favor (Hamilton et al. 2008) or select against (Marshall et al. 2010) immigrants. We

54

define the kernel at settlement because most empirical measurements of openness are

made at or quite near the time of settlement rather than substantially later. Empirical

deviations from our model can suggest where such processes may be important,

however.

The two equations above allow us to model population openness. We define

immigration for a population as the fraction of settling individuals that are immigrants

(Fig. 2.1a-c):

(1)

where Ri and Si are the numbers of self-recruiting individuals and the number

of immigrants (respectively), as defined above. This equation directly measures

population openness as it has been defined previously (Hixon et al. 2002).

To connect our equation to empirical research, we note that immigration is

simply one minus self-recruitment, where self-recruitment is the probability that an

arriving recruit was born within the local population (Botsford et al. 2009). Self-

recruitment is commonly measured and reported by empirical studies of marine larval

dispersal (Jones et al. 1999; Swearer et al. 1999; Jones et al. 2005; Almany et al.

2007).

We also note that self-recruitment has commonly been confused with retention

(e.g., Kinlan et al. 2005), despite their substantial differences. Retention measures the

proportion of larvae produced in a local population that stays in that population:

(2)

55

The key difference between self-recruitment and retention is that the former

indicates the source of locally settling individuals (Fig. 2.1a-c), while the latter

specifies the destination of locally produced individuals (Fig. 2.1d). The distinction

lies in the denominator: while self-recruitment is calculated as a fraction of all

recruiting individuals (Si + Ri), retention is calculated as a fraction of all locally

produced individuals (N). For further discussion, see Botsford et al. (2009).

To avoid future confusion, we define emigration as the fraction of locally

produced larvae that leave the local patch (Fig. 2.1d):

Ei = 1-Ti = 1 - D(i,i) (4)

Emigration is simply one minus retention. We use immigration and emigration

throughout our paper because we find these terms more intuitive than self-recruitment

and retention.

2.3.2 Simplifications for applying the model

To implement this model, we calculate immigration and emigration using

simple approximations for D(j,i) and N(j). We use a Gaussian dispersal kernel, a form

that arises both from a random walk dispersal process (Skellam 1951) and from

averaging across many quasi-random larval trajectories in a coastal ocean (Siegel et al.

2008). Dispersal ability is determined by the standard deviation of this kernel, which is

called the dispersal spread (Siegel et al. 2003). We center the dispersal kernel on zero

as our base scenario, but we investigate the effects of directional dispersal (e.g.,

advection by currents) by offsetting the kernel from zero by a mean displacement. We

scale advection to increase proportionally to the dispersal spread so that the ratio of

the two (the Peclet number) remains a constant (White et al. 2010). We also assume

that production of offspring depends only on patch area. Both the dispersal kernel and

production approximations could be replaced if oceanographic models of connectivity

(e.g., Cowen et al. 2006; Siegel et al. 2008) or metrics of local production (e.g.,

56

Watson et al. 2010) were available. Finally, we conduct our calculations on a regular

grid, as recommended by Thomas & Kunin (1999) for assessing spatial population

structure.

2.3.3 Application to simple seascapes

To explore our model, we apply it first to highly simplified “dashed line”

coastlines. The dashed line is similar to many marine reserve models (e.g., Botsford et

al. 2001) and consists of an infinite, 1D array of 500 m habitat patches separated by

uninhabitable spaces. We also investigate a 2D grid of habitat patches. A half-

kilometer patch width is arbitrary, but allows comparison to previous field studies.

Spacing between adjacent patches is measured from patch center to patch center.

2.3.4 Application to remotely sensed seascapes

To determine whether closed populations are likely to be common, we

analyzed 17 coral reef seascapes. These seascapes were previously classified from 30

m spatial resolution Landsat satellite images (Andréfouët et al. 2006; Wabnitz et al.

2010). The individual seascapes (Fig. 2.2) were chosen to represent the global

diversity of coral reefs. For analysis, the reefs were converted to a 500 x 500 m grid.

Grid cells were specified as suitable habitat if greater than 50% of the cell was covered

by coral reef. We calculated I for each grid cell with Eq. 1 and a symmetrical,

bivariate Gaussian dispersal kernel. Because few patches were isolated by land (see

Fig. 2.2), we did not prevent larvae from dispersing across land.

2.4 Results

2.4.1 Population openness in simple seascapes

We first tested our model on a dashed line of habitat patches (Fig. 2.1b). In this

context, I can vary from nearly 0% to nearly 100%, depending on the relative values

of dispersal distance and habitat spacing (Fig. 2.3). As expected, high spacing between

57

habitat patches and short dispersal creates closed populations with low immigration

(lower right of Fig. 2.3), while the opposite creates open populations (upper left of

Fig. 2.3). Changing the scale of analysis (different patch size) has negligible impact on

this graph because all patches in this simple model are the same size, and larger

patches have both more local recruits and more immigrants. In two dimensions (Fig.

2.3b), patches tend to be somewhat more open for the same patch spacing and

dispersal distance because there are a greater number of surrounding patches to

contribute immigrants. Patches are also more open in advective environments because

fewer larvae return to the patch from which they were born (Fig. 2.4).

As a general guideline, our model predicts relatively closed populations if

spacing is more than about twice the mean dispersal (diagonal of Fig. 2.3). The precise

transition point can be somewhat lower if the kernel is leptokurtic rather than

Gaussian, and so this guideline tends to be conservative. The transition point is higher

in advective environments (Fig. 2.4).

An example illustrates the important difference between self-recruitment and

retention. For a species with 10 km mean dispersal in a habitat with 25 km gaps

between patches, populations have both low retention (>98% of larvae emigrate) and

high self-recruitment (only 21% of settling larvae are immigrants). This situation

corresponds to the lower right of Fig. 2.3.

When habitat spacing is about twice the mean dispersal distance, small

differences in dispersal or spacing will have relatively large effects on I (along the

diagonal of Fig. 2.3). If marine environments are patchy at scales similar to an

organism’s dispersal ability, seascapes should contain both relatively open and

relatively closed populations (high and low I, respectively).

2.4.2 Comparison to empirical studies

It appears that many recent marine larval tagging studies have been conducted

in the zone where both open and closed populations are likely, and particular care

must therefore be used when interpreting their results. For example, three studies

observed immigration of 40-70% (self-recruitment of 30-60%) to ~500 m habitat

58

patches in four species of coral reef fish (genera Amphiprion and Chaetodon) (Jones et

al. 2005; Almany et al. 2007; Planes et al. 2009). Previous interpretations of these

studies inferred a dispersal distance of 100-500 m (Shanks 2009). However, this

interpretation did not consider that, in all three studies, the nearest habitat patch was 5-

20 km away (Jones et al. 2005; Almany et al. 2007; Planes et al. 2009). Using the

patch spacing and immigration reported by these studies, we delineated their

parameter space with the dashed polygon in Fig. 2.3 (vertical sides: 5-20km; top and

bottom sides: 40-70% immigration).

While none of these studies measured mean dispersal distance, our model

allowed us to infer what it might be. In particular, the model suggests that mean

dispersal of 3.5-26 km (1D) or 2.5-14 km (2D) would be most compatible with the

studies’ observations (the range along the y-axis for the polygons in Fig. 2.3). The 1D

approximation is most appropriate for Jones et al. 2005 (near a coastline), while 2D is

more appropriate for the other two studies (in the middle of a bay). Our models shows

that these populations should have been dramatically more closed if average dispersal

was <2 km. By considering habitat patchiness, we show that dispersal distances may

be one to two orders of magnitude greater than previous interpretations suggested.

Recent field studies suggest that our model result is reasonable, despite its

substantial simplifications. Population genetics suggest a dispersal spread of

approximately 10 km in another Amphiprion species (Pinsky et al. 2010). Similarly,

other studies suggest that dispersal distances in reef fish are usually tens of kilometers

or greater (Puebla et al. 2009).

2.4.3 Closed populations within empirical seascapes

Upon examining empirical seascapes, we found that they had both open and

relatively closed populations. We first show examples from the Bahamas with 2 km

dispersal spread and from Papua New Guinea with 5 km dispersal (Fig. 2.5). These

distances were chosen to illustrate the range of I predicted within each seascape. In

both, the patches embedded in continuous sections of reef had high I (up to 98%) and

would be classified as open at this scale. In addition, a small number of patches were

59

more isolated and had I as low as 44%. These latter patches received up to 56% of

their recruits from local parents (56% self-recruitment) and would appear quite closed.

We next applied our models across 17 coral reef seascapes (Fig. 2.2) to ask

whether all seascapes are likely to contain closed populations. We found surprising

similarities among seascapes (Fig. 2.6). While seascapes on average were open across

all dispersal distances (Fig. 2.6a), we also found that all seascapes contained at least

some isolated patches that were largely closed and had low I (Fig. 2.6b). All seascapes

exhibited a similar relationship between openness and dispersal distance, with a

threshold near 10 km (Fig. 2.6b). Closed patches were most likely for species with

mean dispersal < 5 km. Some fish and many marine invertebrates likely fall in this

category (Shanks 2009).

Substantial variation between seascapes is also apparent. On continuous reefs

such as Northwest Belep (New Caledonia 2), patches were generally open even at

short dispersal distances. In highly patchy seascapes such as Kimbe Bay (Papua New

Guinea 3), some patches were moderately closed even for dispersal distances up to 10

km. The somewhat unique curve in the Java Sea (Indonesia 1) was created because

two adjacent patches were the most isolated and therefore had a predicted I near 50%

for dispersal distances of 2-5 km (Fig. 2.6b). However, all seascapes were open for

species with average dispersal >20 km.

Average nearest neighbor distance was a reasonable predictor of average

openness (I) within seascapes (Fig. 2.7) (p < 0.003, r2 > 46%). Seascapes with wider

spacing had lower I, and the populations in these patchier seascapes were more closed.

We conducted the above analyses at only a single grid scale to illustrate a

general pattern. Following the logic of our models, analyses with a larger grid size

(e.g. 1 x 1 km) would generally show populations that are more closed (Fig. 2.8).

Similarly, a smaller grid size would show populations that are more open. Our chosen

grid scale made our results relevant to previous studies, but higher spatial resolution

remote sensing products (e.g., 30 m) are available to resolve small reefs that may be

important stepping stones for low dispersal species (Andréfouët et al. 2006).

60

2.5 Discussion

By examining a simple model, we have determined conditions under which

habitat patchiness can have strong impacts on immigration and population openness.

We found that relatively closed populations with few immigrants are more common

where patch spacing is more than twice the dispersal spread. When investigating

realistic levels of habitat patchiness, we found that a wide range of coral reef

seascapes are likely to contain a mixture of isolated, relatively closed populations with

low immigration and open populations with substantially more immigration. We

found that patch spacing alone can explain why a number of recent studies have found

surprisingly high fractions of self-recruitment in marine fishes otherwise characterized

by long-distance dispersal.

Our criteria for closed populations provide quantitative guidance on an issue

that has been discussed qualitatively for many years (Hixon et al. 2002; Mora & Sale

2002; Warner & Cowen 2002). Our models of patchy habitats indicated that largely

closed populations could arise for many species even at commonly encountered levels

of patch isolation. Where a patch is isolated by more than twice the average dispersal

distance, relatively few individuals will immigrate from other patches and we should

expect the population to be relatively closed. This criterion should be relevant not only

to marine species with larval dispersal, but also to insects, stream-dwelling aquatic

organisms, plants with seeds, and other organisms with one life stage that disperses

further than others.

We also note that our two-times rule for closed populations should not be

confused with previous criteria for self-persistent populations. Theory suggests that

isolated populations are more likely to survive on patches at least twice as wide as

mean dispersal distance (Botsford et al. 2001; Lockwood et al. 2002; White et al.

2010). Under such conditions, the fraction of larvae emigrating from the patch is low

enough that the population can be sustained on local production alone. As a result,

self-persistence depends on the relation between patch size and dispersal, while in

contrast, population openness depends on patch spacing and dispersal.

61

Appreciating the difference between persistence and openness is especially

important for interpreting recent studies investigating the source of larvae recruiting to

populations (e.g., Jones et al. 1999; Swearer et al. 1999; Jones et al. 2005; Almany et

al. 2007). For example, it would be incorrect to assume that dispersal distances must

be less than 500 m to explain largely closed populations of reef fish on 500 m habitat

patches (Jones et al. 2005; Almany et al. 2007; Planes et al. 2009), as some authors

have done (e.g., Shanks 2009). If average dispersal was <500 m, these populations

should have had virtually 0% immigration, rather than the 40-70% immigration

observed. Instead, our model for population openness provides a quantitative method

for interpreting these studies, and suggests that dispersal is one to two orders of

magnitude greater than previously suspected. Therefore, these populations appear

unlikely to be self-persistent. These differences between criteria for persistence and

for population openness, and between immigration and emigration, are critical for

understanding marine larval dispersal and for appropriately interpreting the empirical

evidence for larval connectivity that is now available (Botsford et al. 2009).

2.5.1 A new language for marine connectivity: immigration and

emigration

More generally, past misinterpretations of larval dispersal may result in part

from confusion over the terms often used to describe marine connectivity. Self-

recruitment and retention sound like similar ideas, and yet refer to substantially

different concepts (Botsford et al. 2009). As our example from the dashed line habitats

showed, self-recruitment can be high even if retention is very low.

For this reason, we advocate a new, more intuitive vocabulary to describe

population connectivity: the fraction of immigrants and the fraction of emigrants. The

fraction of immigrants is simply the proportion of all settling recruits that are

immigrants, which describes the source of recruits to the next generation. This is the

same as population openness: open populations have high immigration fractions, while

fully closed populations have no immigrants. In contrast, the fraction of emigrants is

the proportion of locally produced larvae that emigrate. This describes the destination

62

of local larvae. The Methods section of this paper included mathematical formulations

for both terms. The fractions of immigrants and emigrants are simply one minus self-

recruitment or retention (respectively), and so can easily be inserted in the literature.

We believe that this change is important because there is much less chance that

immigration and emigration will be misinterpreted or confused.

2.5.2 Immigration in naturally patchy landscapes

By analyzing a broad range of coral reef seascapes, we showed that natural

variation in patch spacing is sufficient to have strong control over immigration rates,

determining both whether and where populations are more open or closed. It has long

been clear that extremely isolated islands harbor closed populations with little

immigration (Robertson 2001), but it has not been clear that habitats are also

sufficiently patchy at scales of 10s of km to create largely closed populations. Our

results indicated that across 17 coral reef seascapes, populations with little

immigration were always present, particularly for species with dispersal spread less

than 10 km. Our maps of predicted immigration fractions provide initial hypotheses

for where openness should be higher or lower, and these hypotheses can be tested with

field observations. These simple models, however, do not consider oceanographic

currents that are also likely to be important.

More generally, our analysis showed that the location of a population is

important for determining how that population is replenished. A study undertaken

within a continuous string of habitat patches is likely to encounter open populations

with high immigration in which larval supply or recruitment is independent of local

production. In contrast, studies on more isolated habitat patches may encounter

relatively closed populations with little immigration where local production has a

strong impact on larval supply. This realization adds habitat configuration as an

important aspect to consider in debates about the sources of regulation and density

dependence in populations (Hixon et al. 2002). Models assuming open population

dynamics (e.g., Bascompte et al. 2002) will be more appropriate in relatively

continuous habitats, while models assuming partially closed dynamics (e.g., Bolker &

63

Pacala 1999) will be more appropriate in more isolated patches. The conclusion is

appropriate for a wide range of marine and terrestrial species with a dispersive life

stage.

We chose to focus on coral reefs in this paper because there are readily

available global data on their distribution, but we expect that our results will be

broadly relevant across many habitats, both marine and terrestrial. While the

patchiness of terrestrial habitats have long been recognized (Saunders et al. 1991), we

also emphasize that many marine habitats are patchy as well, including estuaries,

rocky reefs, deep sea vents, seamounts, rocky intertidal habitats, kelp forests,

mangroves, seagrasses, tide pools, and sheltered bays (Kritzer & Sale 2006). Many

marine species specialize on a single or small number of these habitat types, and hence

many species have patchy, fragmented distributions. We therefore predict that these

species will have both open and more closed populations. New efforts to map marine

habitats, including with remote sensing, will be an important step towards

understanding the role of patchiness in local demography (Andréfouët et al. 2006;

Purkis et al. 2007).

As may have become clear, immigration fractions and population openness are

matters of scale (Wiens 1989; Levin 1992). As other authors have noted, picking a

wider spatial scale for investigation will reveal more strongly closed populations,

while at a finer scale (e.g., a single coral head), populations will be almost entirely

open (Wiens 1989; Hixon et al. 2002). Temporal scale is also an important

consideration, and our focus in this paper has been on single-generation, ecological

time scales. Over evolutionary time scales of many generations, however, rare long

distance dispersal events and multigenerational dispersal across intermediate stepping

stones may keep populations evolutionarily connected even if they are ecologically

quite closed (Waples 1998).

2.5.3 Model limitations

Our model represents the interactions of larval transport, behavior, and

survival as a dispersal kernel, and it captures the locations of larval production and

64

settlement with a habitat map. These are clearly simplifications of marine dispersal,

but our model’s complexity is comparable to many useful metapopulation models

(Moilanen & Hanski 2006). Models that also consider ocean currents would add

accuracy to our predictions, but won’t change our general conclusion that both

relatively open and closed populations are likely to co-exist within seascapes.

Oceanographic features such as jets and gyres can create consistent places that favor

transport or retention of larvae (Cowen et al. 2006), while heterogeneities in water

quality can alter larval survival and affect population openness (Shima et al. 2010).

Larval transport is also temporally stochastic (Siegel et al. 2008), and this temporal

variability may be sufficient to switch populations between open and closed.

2.5.4 Future directions

Moving forward, it is interesting to note that conditions for persistence

typically depend on fractions of emigration (Botsford et al. 2001; Lockwood et al.

2002; Byers & Pringle 2006; Hastings & Botsford 2006), while empirical studies

measure immigration or dispersal distance. Dispersal kernels, as used in this study,

provide one method for translating between dispersal distance, immigration and

emigration. Parameterizing such kernels requires at least knowledge of kernel width,

but we lack such estimates for the vast majority of marine species because measuring

larval dispersal remains difficult.

Existing methods for estimating dispersal distances include isolation-by-

distance genetic methods (Puebla et al. 2009; Pinsky et al. 2010), invasion rate

estimates (Shanks 2009), and mark-recapture studies or their recent variations with

natural and artificial tags (Jones et al. 1999; Swearer et al. 1999; Jones et al. 2005;

Planes et al. 2009). All of these methods come with caveats, however. Isolation-by-

distance methods require knowledge of genetic effective density, which remains

difficult to estimate empirically. Invasion rates are dominated by rare long-distance

dispersal events, and therefore overestimate average dispersal distance (Higgins &

Richardson 1999). Mark-recapture experiments tend to underestimate average

dispersal because short-distance dispersers are easier to find (Koenig et al. 1996).

65

In conclusion, we argue that natural scales of habitat patchiness are likely to

have strong impacts on immigration and population openness for marine species.

Seascape geography likely has a larger and more easily detectable role in determining

population openness than has been appreciated to date, particularly for species with

mean dispersal less than a few tens of kilometers.

2.6 Acknowledgements

We thank S. Gaines for helpful conversations during this project, and three

anonymous referees for constructive suggestions on earlier drafts. M. L. P. was

supported by a NSF Graduate Research Fellowship.

2.7 References

Almany G.R., Berumen M.L., Thorrold S.R., Planes S. & Jones G.P. (2007). Local

replenishment of coral reef fish populations in a marine reserve. Science, 316,

742-744.

Andréfouët S., Muller-Karger F.E., Robinson J.A., Kranenburg C.J., Torres-Pulliza

D., Spraggins S.A. & Murch B. (2006). Global assessment of modern coral

reef extent and diversity for regional science and management applications: a

view from space. Proceedings of the 10th International Coral Reef

Symposium, Okinawa, Japan. June 28-July 2, 2004, 1732-1745.

Bascompte J., Possingham H.P. & Roughgarden J.D. (2002). Patchy populations in

stochastic environments: critical number of patches for persistence. The

American Naturalist, 159, 128-37.

Bolker B.M. & Pacala S.W. (1999). Spatial Moment Equations for Plant Competition:

and the Spatial Strategies Understanding Advantages of Short Dispersal. The

American Naturalist, 153, 575-602.

66

Botsford L.W., Hastings A. & Gaines S.D. (2001). Dependence of sustainability on

the configuration of marine reserves and larval dispersal distance. Ecol. Lett.,

4, 144-150.

Botsford L.W., White J.W., Coffroth M.-A., Paris C.B., Planes S., Shearer T.L.,

Thorrold S.R. & Jones G.P. (2009). Connectivity and resilience of coral reef

metapopulations in marine protected areas: matching empirical efforts to

predictive needs. Coral Reefs, 28, 327-337.

Byers J.E. & Pringle J.M. (2006). Going against the flow: retention, range limits and

invasions in advective environments. Mar. Ecol. Prog. Ser., 313, 27-41.

Cowen R.K., Paris C.B. & Srinivasan A. (2006). Scaling of connectivity in marine

populations. Science, 311, 522-527.

Figueira W.F. (2009). Connectivity or demography: Defining sources and sinks in

coral reef fish metapopulations. Ecol. Model., 220, 1126-1137.

Gaines S.D., White C., Carr M.H. & Palumbi S.R. (2010). Designing marine reserve

networks for both conservation and fisheries management. P Natl Acad Sci

USA. doi:10.1073/pnas.0906473107

Hamilton S.L., Regetz J. & Warner R.R. (2008). Postsettlement survival linked to

larval life in a marine fish. P Natl Acad Sci USA, 105, 1561-1566.

Hastings A. & Botsford L.W. (2006). Persistence of spatial populations depends on

returning home. Proc. Natl. Acad. Sci. USA, 103, 6067-6072.

Higgins S.I. & Richardson D.M. (1999). Predicting plant migration rates in a changing

world: the role of long‐distance dispersal. The American Naturalist, 153, 464-

475.

Hixon M.A., Pacala S.W. & Sandin S.A. (2002). Population regulation: historical

context and contemporary challenges of open vs. closed systems. Ecology, 83,

1490-1508.

Jones G.P., Milicich M.J., Emslie M.J. & Lunow C. (1999). Self-recruitment in a coral

reef fish population. Nature, 402, 802-804.

Jones G.P., Planes S. & Thorrold S.R. (2005). Coral reef fish larvae settle close to

home. Curr. Biol., 15, 1314-1318.

67

Kaplan D.M., Botsford L.W., O'Farrell M.R., Gaines S.D. & Jorgensen S. (2009).

Model-based assessment of persistence in proposed marine protected area

designs. Ecol. Appl., 19, 433-448.

Kinlan B.P., Gaines S.D. & Lester S.E. (2005). Propagule dispersal and the scales of

marine community process. Divers. Distrib., 11, 139-148.

Koenig W.D., Van Vuren D. & Hooge P.N. (1996). Detectability, philopatry, and the

distribution of dispersal distances in vertebrates. Trends Ecol. Evol., 11, 514-

517.

Kritzer J.P. & Sale P.F. (eds.) (2006). Marine metapopulations. Elsevier Academic

Press, New York.

Leibold M.A., Holyoak M., Mouquet N., Amarasekare P., Chase J.M., Hoopes M.F.,

Holt R.D., Shurin J.B., Law R., Tilman D., Loreau M. & Gonzalez A. (2004).

The metacommunity concept: a framework for multi-scale community

ecology. Ecol. Lett., 7, 601-613.

Lenormand T. (2002). Gene flow and the limits to natural selection. Trends Ecol.

Evol., 17, 183-189.

Levin L.A. (2006). Recent progress in understanding larval dispersal: new directions

and digressions. Integr. Comp. Biol., 46, 282-297.

Levin S.A. (1992). The problem of pattern and scale in ecology. Ecology, 73, 1943-

1967.

Lockwood D.R., Hastings A. & Botsford L.W. (2002). The effects of dispersal

patterns on marine reserves: does the tail wag the dog? Theor. Popul. Biol., 61,

297-309.

Marshall D.J., Monro K., Bode M., Keough M.J. & Swearer S. (2010). Phenotype-

environment mismatches reduce connectivity in the sea. Ecol. Lett., 13, 128-

40.

Moilanen A. & Hanski I. (2006). Connectivity and metapopulation dynamics in highly

fragmented landscapes. In: Connectivity Conservation. Cambridge University

Press Cambridge, pp. 44-71.

68

Mora C. & Sale P.F. (2002). Are populations of coral reef fish open or closed? Trends

Ecol. Evol., 17, 422-428.

Palmer M.A., Allan J.D. & Butman C.A. (1996). Dispersal as a regional process

affecting the local dynamics of marine and stream benthic invertebrates.

Trends Ecol. Evol., 11, 322-326.

Paris C.B. & Cowen R.K. (2004). Direct evidence of a biophysical retention

mechanism for coral reef fish larvae. Limnol. Oceanogr., 49, 1964-1979.

Pinsky M., Montes H.R., Jr. & Palumbi S.R. (2010). Using isolation by distance and

effective density to estimate dispersal scales in anemonefish. Evolution, 64,

2688-2700. doi:10.1111/j.1558-5646.2010.01003.x

Planes S., Jones G.P. & Thorrold S.R. (2009). Larval dispersal connects fish

populations in a network of marine protected areas. P Natl Acad Sci USA, 106,

5693-5697.

Puebla O., Bermingham E. & Guichard F. (2009). Estimating dispersal from genetic

isolation by distance in a coral reef fish (Hypoplectrus puella). Ecology, 90,

3087-3098.

Purkis S.J., Kohler K.E., Riegl B.M. & Rohmann S.O. (2007). The statistics of natural

shapes in modern coral reef landscapes. The Journal of Geology, 115, 493-508.

Roberts C.M. (1997). Connectivity and management of Caribbean coral reefs. Science,

278, 1454-1457.

Robertson D.R. (2001). Population maintenance among tropical reef fishes: Inferences

from small-island endemics. P Natl Acad Sci USA, 98, 5667-5670.

Roughgarden J., Gaines S.D. & Possingham H. (1988). Recruitment dynamics in

complex life cycles. Science, 241, 1460-1466.

Saunders D.A., Hobbs R.J. & Margules C.R. (1991). Biological consequences of

ecosystem fragmentation: a review. Conserv. Biol., 5, 18-32.

Shanks A.L. (2009). Pelagic larval duration and dispersal distance revisited.

Biological Bulletin, 216, 373-385.

69

Shima J.S., Noonburg E.G. & Phillips N.E. (2010). Life history and matrix

heterogeneity interact to shape metapopulation connectivity in spatially

structured environments. Ecology, 91, 1215-24.

Siegel D.A., Kinlan B.P. & Gaines S.D. (2003). Lagrangian descriptions of marine

larval dispersion. Mar. Ecol. Prog. Ser., 260, 83-96.

Siegel D.A., Mitarai S., Costello C.J., Gaines S.D., Kendall B.E., Warner R.R. &

Winters K.B. (2008). The stochastic nature of larval connectivity among

nearshore marine populations. P Natl Acad Sci USA, 105, 8974-8979.

Skellam J.G. (1951). Random dispersal in theoretical populations. Biometrika, 38,

196-218.

Swearer S.E., Caselle J.E., Lea D.W. & Warner R.R. (1999). Larval retention and

recruitment in an island population of a coral-reef fish. Nature, 402, 799-802.

Taylor M.S. & Hellberg M.E. (2003). Genetic evidence for local retention of pelagic

larvae in a Caribbean reef fish. Science, 299, 107-109.

Thomas C.D. & Kunin W.E. (1999). The spatial structure of populations. J. Anim.

Ecol., 68, 647-657.

Wabnitz C.C., Andréfouët S. & Muller-Karger F.E. (2010). Measuring progress

toward global marine conservation targets. Front. Ecol. Environ., 8, 124-129.

Waples R.S. (1998). Separating the wheat from the chaff: Patterns of genetic

differentiation in high gene flow species. J. Hered., 89, 438-450.

Warner R.R. & Cowen R.K. (2002). Local retention of production in marine

populations: evidence, mechanisms, and consequences. Bull. Mar. Sci., 70,

S245-S249.

Watson J.R., Mitarai S., Siegel D.A., Caselle J.E., Dong C. & McWilliams J.C.

(2010). Realized and potential larval connectivity in the Southern California

Bight. Mar. Ecol. Prog. Ser., 401, 31-48.

White J.W., Botsford L.W., Hastings A. & Largier J.L. (2010). Population persistence

in marine reserve networks: incorporating spatial heterogeneities in larval

dispersal. Mar. Ecol. Prog. Ser., 398, 49-67.

Wiens J.A. (1989). Spatial scaling in ecology. Funct. Ecol., 3, 385-397.

70

2.8 Figures

Figure 2.1. Habitats can be continuous (a) or patchy (b, c, and d). (a) In

continuous habitats, large numbers of immigrants into the focal population (in black)

create a population that is largely open. (b) In patchy environments, fewer immigrants

reach the focal population, and the population is therefore more closed. (c) If the

spacing between patches is small relative to dispersal abilities, the openness of the

population will not be affected by habitat patchiness (compare to a). (d) Low

immigration and high emigration can both occur at the same time in patchy habitats.

71

Figure 2.2. Maps of coral reef seascapes that were analyzed for openness.

Scale bars are 10 km long and north is up. Each seascape has been simplified to a 500

x 500 m grid, and only grid cells with > 50% reef coverage were retained. Seascapes

with sparse reefs < 500 x 500 m may look quite different after this processing (e.g.,

Glover’s Reef in Belize). The gray shading is land.

Bahamas 1 − Nassau Bahamas 2 − Ragged Is. Indonesia 1 − Java Sea Indonesia 2 − SW Sulawesi Indonesia 3 − N Sulawesi

New Caledonia 1 − NE Noumea New Caledonia 1 − NW Belep Panama − N San Blas Papua New Guinea 1 − Numata Is. Papua New Guinea 2 − Panasia Is.

Papua New Guinea 3 − Kimbe Bay Tuamotu − Fakarava Vanuatu − Lakatoro Western Solomon − Choiseul Belize 1 − Yucatan − Harvest Cay

Belize 2 − Yucatan − Columbus Cay Belize 3 − Yucatan − Glovers Reef

72

Figure 2.3. Percent immigration (I) into patches predicted from a Gaussian

dispersal kernel in a patchy habitat that is a) a one-dimensional coastline (similar to

Fig. 2.1b) or b) a two-dimensional grid. Patches are 500 m wide and spacing measures

the distance between patch centers. When spacing between patches is more than about

twice the dispersal spread, I is low and % self-recruitment is high (lower right).

Dashed polygon outlines the habitat spacing and values of I measured by three marine

population openness studies (Jones et al. 1999; Almany et al. 2007; Planes et al.

2009).

73

Figure 2.4. Percent immigration (I) into a 1D array of patches with a Gaussian

dispersal kernel displaced from zero to model advection. The Peclet number (kernel

displacement/dispersal spread) is a) 0.25, b) 0.75, or c) 1.5. Otherwise, compare to

Fig. 2.3a.

0 10 20 30 40

510

1520

2530

Dis

pers

al s

prea

d (k

m)

a)

0 10 20 30 40

510

1520

2530

Habitat spacing (km)

Dis

pers

al s

prea

d (k

m)

c)

0 10 20 30 40

510

1520

2530

Habitat spacing (km)

b)

● 0% Immigration

20%

40%

60%

80%

100% Immigration

74

Figure 2.5. Variation in openness (% immigration, I) across patchy seascapes.

(a) Ragged Island, Bahamas (2 km dispersal spread). (b) Kimbe Bay, Papua New

Guinea (5 km dispersal spread). Grid cells were 500 m squares and the dispersal

kernel was bivariate Gaussian. Each seascape has both many well-connected reefs

with high immigration and a smaller number of relatively closed populations with as

little as 44% immigration (56% self-recruitment).

75

Figure 2.6. Population openness (% immigration, I) within each of 17 coral

reef seascapes, as related to dispersal spread. Mean (a), minimum (b), and maximum

(c) are calculated across all grid cells within a seascape. Minimum and maximum

immigration indicates the most or least closed habitat patch (respectively) in each

seascape. Each line represents one seascape. Immigration was calculated to 500 x 500

m grid cells. See Fig. 2.2 for maps and locations of each seascape.

76

Figure 2.7. Mean immigration (I) across 500 x 500 m grid cells plotted against

mean nearest neighbor distance within each seascape. Each point is one of the 17

seascapes analyzed in this paper.

77

Figure 2.8. The exact level of immigration (I) depends in part on the grid size

used for the analysis. Larger grids tend to have lower I. (a and d) Average I across all

patches in a seascape, (b and e) minimum I within a seascape, and (c and f) maximum

I within a seascape (similar to Fig. 2.4). Top row shows NW Belep in New Caledonia.

Lower row shows Kimbe Bay, Papua New Guinea.

1 2 5 10 20 50 100

040

80

Dispersal spread (km)

Imm

igra

tion

(%) a)

1 2 5 10 20 50 100

040

80

Dispersal spread (km)

b)

1 2 5 10 20 50 100

040

80

Dispersal spread (km)

c)

1 2 5 10 20 50 100

040

80

Dispersal spread (km)

Imm

igra

tion

(%) d)

1 2 5 10 20 50 100

040

80Dispersal spread (km)

e)

1 2 5 10 20 50 100

040

80

Dispersal spread (km)

f)

Grid size (m)2505007501000

78

79

Chapter 3

Unexpected patterns of fisheries collapse in the

world’s oceans

3.1 Abstract

Understanding which species are most vulnerable to human impacts is a

prerequisite for designing effective conservation strategies. Surveys of terrestrial

species have suggested that large-bodied species and top predators are the most at risk,

and it is commonly assumed that such patterns also apply in the ocean. However, there

has been no global test of this hypothesis in the sea. We analyzed two fisheries

datasets (stock assessments and landings) to determine the life history traits of species

that have suffered dramatic population collapses. Contrary to expectations, our data

suggest that up to twice as many fisheries for small, low trophic level species have

collapsed as compared to those for large predators. These patterns contrast with those

on land, suggesting fundamental differences in the ways that industrial fisheries and

land conversion affect natural communities. Even temporary collapses of small, low

trophic level fishes can have ecosystem-wide impacts by reducing food supply to

larger fish, seabirds, and marine mammals.

3.2 Introduction

Overfishing is one of the most serious conservation concerns in marine

ecosystems (Worm et al. 2009), but understanding which species are most at risk

remains a challenge. On land, life history traits are strong predictors of extinction risk

(Fisher & Owens 2004), and vulnerable species often have large body size and high

trophic level (Fisher & Owens 2004; Cardillo et al. 2005). In marine ecosystems, the

80

well-publicized declines of large predatory fishes (Baum et al. 2003; Myers & Worm

2003) suggest that similar trends may also be common in the sea. However, research

to date has found or proposed a wide range of life history characteristics that cause

high vulnerability, including large body size (Jennings et al. 1998; Dulvy et al. 2003;

Reynolds et al. 2005; Olden et al. 2007), late maturity (Jennings et al. 1998; Reynolds

et al. 2005), long lifespan (Jennings et al. 1998; Denney et al. 2002; Dulvy et al.

2003; Reynolds et al. 2005; Winemiller 2005), low fecundity and high parental

investment in offspring (King & McFarlane 2003; Winemiller 2005), or high trophic

level (Pauly et al. 1998; Fisher & Owens 2004). Understanding which traits, or

combinations of traits, are most useful for predicting vulnerability has been difficult

because analyses have been limited to regional comparisons or narrow species groups,

and because reliable global data have not been available to more broadly test which

types of fishes are most likely to suffer fisheries collapse.

In addition, there are reasons to believe that regional or terrestrial life history

trends might not apply globally in the ocean. For example, fishery biologists often

recommend higher harvest rates for fast-growing, highly productive species, and lower

harvest rates for species with lower productivity (Hilborn & Walters 1992). Where

implemented, these adjustments might reduce the resilience of fast-growing species

and put all harvested species at similar risks of decline. In addition, economic forces

or management regime may be more important than life history in determining

whether fishing effort is successfully controlled (Costello et al. 2008; Sethi et al.

2010). Small pelagic species, while often possessing a rapid growth rate, are also

highly catchable and therefore susceptible to overfishing (Beverton 1990). Finally, the

conflict with human development that is particularly acute for large, terrestrial

mammals (Cardillo et al. 2005) may be smaller in open ocean ecosystems far from

coastlines.

In this paper, we used two independent fisheries databases to determine which

stocks have collapsed to low population abundance. Our first database contained 223

scientific stock assessments for 120 species. For these assessments, a stock was

defined as collapsed if its minimum annual biomass (BMIN) fell to < 20% of the

81

biomass necessary to support maximum sustainable yield (BMSY) (Worm et al. 2009).

In addition, we examined global landings reported by the Food and Agriculture

Organization (FAO) for 1950-2006. We treated each FAO statistical area as a stock,

for a total of 891 stocks across 458 species. For landings data, a stock was defined as

collapsed if landings remained below 10% of the average of the five highest landings

recorded for more than two years. We found this definition to have the lowest

misclassification rate (18%) when we evaluated it against stocks for which we had

both biomass and catch timeseries (see Methods). In most misclassifications (14 of

24), we failed to detect collapses that had occurred, suggesting that our landings

definition is relatively conservative. Finally, we assessed the prevalence of collapse

across a broad range of life history traits, including lifespan, age of maturity, body

size, trophic level (TL), growth rate, fecundity, and parental investment in offspring

(egg size) (Froese & Pauly 2010).

3.3 Materials and Methods

3.3.1 Data sources

We downloaded stock assessments on June 9, 2010 from the RAM Legacy

database (Worm et al. 2009). The database was compiled in 2009 and 2010 from

countries around the world, and assessments were the most current available at that

time. In particular, we extracted time series of catch, model-estimated biomass, and

fishing mortality rates from 1950-2008. The final year included in the timeseries was

on average 2006 ± 1.5 (s.d.) (range: 2000-2008) and duration was on average 39.6 ±

12.4 (s.d.) yrs (range: 10 to 59 yrs). We also extracted reference points for BMSY and

FMSY (the biomass and instantaneous fishing mortality rate, respectively, that result in

maximum sustainable yield). We only used assessments for fishes (not invertebrates).

Our landings database contained statistics reported to the Food and Agriculture

Organization of the United Nations (downloaded from

http://www.fao.org/fishery/statistics/software/fishstat/en, December 2009). Only

species-level records for fish with cumulative landings > 1000 tons were retained,

82

while invertebrates and records for species groups (e.g., “Cods” or “Flatfishes”) were

removed. We removed records with low cumulative catches to avoid minor and

experimental fisheries, and results were similar when we only retained records with

total landings > 10,000 tons. Data were reported in one of 19 major statistical areas,

and one species in one area was considered a stock.

3.3.2 Fishery collapses – Assessment data

For the assessments, we analyzed biomass relative to BMSY. Stocks that fell

below 20% or 50% of BMSY were defined as collapsed or overfished, respectively. We

also recorded the maximum depletion as BMIN/ BMSY, where BMIN is the minimum

biomass in the timeseries. The length of collapse was the maximum number of

consecutive years that a stock was below 20% of BMSY.

In cases where neither BMSY nor a proxy used in place of BMSY was reported in

the stock assessment, we followed (Worm et al. 2009) and estimated BMSY from

Schaefer surplus-production models fit to the assessment time series of annual total

biomass and total catch or landings. Surplus-production models are commonly used in

fisheries science and allow calculation of both carrying capacity and maximum

sustainable yield (Worm et al. 2009). Models were fit in AD Model Builder

(http://admd-project.org) assuming normally distributed errors. We only used time

series greater than 20 years. We used surplus-production models to find reference

points for 92 of 223 stocks.

To examine sensitivity to our choice of model form, we also fit a Fox surplus-

production model. Compared to the Schaefer model, the Fox model assumes that BMSY

is a smaller fraction of unfished biomass (37% instead of 50%). Only five stocks

(2.2%) were reclassified using this approach, and in all cases this was from collapsed

to not collapsed. This small change did not affect our results.

3.3.3 Fishery collapses – Landings data

The choice of appropriate definitions for fisheries collapse in landings data has

been contentious because some apparent collapses may result from stochasticity or

83

changes in reporting, management, or fishing practices rather than population status

(Wilberg & Miller 2007; de Mutsert et al. 2008). By focusing only on true species

(rather than species groups), we avoid false collapses that would otherwise appear

when reporting improves and species groups begin to be reported as individual

species. However, if management measures severely restrict landings to allow

rebuilding of an overfished stock, a stock’s landings could appear collapsed even if

biomass was overfished but not fully collapsed. Changes in fleet capacity or fishing

efficiency could also reduce landings independently from changes to population

abundance.

To avoid false detection of collapses in landings to the extent possible, we

evaluated a range of potential collapse definitions. For each definition, a stock was

defined as collapsed when annual landings fell below 10% of a reference level for a

specified window of time. The reference level was the maximum annual landings

averaged over one or five years. While one year has been used before (Worm et al.

2006), the five-year average was used here to avoid false collapses triggered by a

single, spuriously high year of landings (Wilberg & Miller 2007). As our time

window, we used either one (Worm et al. 2006), two, or four consecutive years

(Mullon et al. 2005). We only looked for collapses in the years following the

maximum annual landings.

We tested the landings-based definitions against the stock assessments for

which we had both biomass and catch or landings data (n = 131). Taking the

assessment-based collapse definitions to be accurate, the 5-yr reference/2-yr window

collapse definition for landings data had a somewhat lower error rate than the others

and misclassified 24 stocks (18%). Of these, 14 collapsed stocks were not detected by

the landings definition, while 10 un-collapsed stocks were falsely detected. Falsely

detected collapses tended to be for species with longer lifespans than those that we

failed to detect (p = 0.006). In comparison, the lax 1-yr maximum/1-yr window

collapse definition misclassified 28 stocks (21%). Other combinations of threshold and

time window produced intermediate numbers of misclassified stocks. Therefore, we

84

used the 5-yr maximum/2-yr window definition of collapse for landings data in our

paper.

3.3.4 Fisheries characteristics

We defined fishery initiation as the year in which landings reached 10% of the

maximum annual landings within a stock. Relative fishing mortality for a stock in the

assessment database was defined as the maximum instantaneous fishing mortality rate

(maximum F) divided by the fishing mortality predicted to produce maximum

sustainable yield (FMSY). Where FMSY was not available, we estimated FMSY from a

Schaefer surplus-production model, as above (Worm et al. 2009). We used a 5-yr

running mean to average out noise, producing maximum F5-yr/FMSY. Both metrics were

averaged across all stocks within a species.

3.3.5 Life history traits

We extracted information on each species’ maximum total length (cm),

maximum weight (kg), lifespan (yr), age of maturity (yr), trophic level, growth rate,

fecundity (eggs/individual), and parental investment in offspring from Ref. (Froese &

Pauly 2010). Growth rate was measured as the exponent (K) in the von Bertalanffy

growth function (von Bertalanffy 1938). Offspring investment was measured as egg

diameter (mm). We supplemented fecundity and egg diameter data with literature

searches because sample size was low for these traits (Table 3.S1). We used the

average if multiples values were available. For the one stock assessment conducted on

a species group (“Redfish species” on the Newfoundland-Labrador Shelf), we

averaged the life history characteristics for the two species targeted by this fishery

(Sebastes mentella and S. fasciatus).

For comparison of collapse rates among species with opposing traits, we

compared the upper and lower quartiles for each trait. For example, we define top

predators as species in the upper quartile of all trophic levels, and low trophic level

species as those in the lowest quartile. Thresholds were chosen independently for

85

species in the assessment and landings databases, and so are slightly different between

the two.

We used log-transformed values for lifespan, age of maturity, weight, length,

growth rate, fecundity, and egg diameter. Because length and weight are highly

correlated (Pearson correlation: p < 0.0001, ρ = 0.89 on log-transformed variables),

we only report results for weight. We did not consider reproductive lifespan (lifespan

minus age of maturity) because it is highly correlated to lifespan (Pearson correlation:

p < 0.0001, ρ = 0.99 on log-transformed variables). Other life history traits remain

somewhat correlated (e.g., lifespan and growth rate or length and trophic level), but

our conclusions do not depend on multiple regressions that would be affected by this

lack of independence.

3.3.6 Statistical models

We fit generalized linear models (GLMs) with binomial errors and a logit link

(Dobson 2002) to predict the probability of a stock collapsing within each species

(either assessment or landings data). In other words, the proportion of stocks collapsed

within each species i was assumed to follow a binomial distribution with mean pi. The

linear predictor of pi was

log(pi/(1-pi)) = ß1,j + ß2,jxi,j

where ß1,j and ß2,j are the fitted coefficients for trait j, and xi,j is the value of

trait j for species i. The traits (x) were life history or fisheries characteristics. The

binomial error model accounted for the fact that variance changes with the number of

stocks within each species. Fitting the models to species-level (rather than stock-level)

data avoided pseudoreplication of species. We used the same format to fit models to

the proportion of overfished stocks. For maximum depletion in the assessment data,

we used standard linear models.

86

Next, we evaluated models with all possible combinations of life history

variables:

log(pi/(1-pi)) = ßxi

where ß is a vector of parameters and xi a vector of life history traits for

species i. The models did not include interactions between variables because of the

large number of potential combinations. We only tested these models on species for

which we had complete life history data (N = 55 for assessment data, N = 67 for

landings) .We evaluated all models within a model-choice framework (Burnham &

Anderson 2002) and retained the minimal adequate models with an Akaike’s

Information Criterion (AIC) within 2 of the lowest AIC. We also tested the specific

hypothesis that species adapted to stable environments, as indicated by low fecundity

and high parental investment in offspring (Winemiller 2005), would be more

vulnerable to fisheries collapse. We did this by fitting a model with fecundity, egg

diameter, and their interaction. In all cases, we evaluated a model’s significance with a

chi-squared test comparing the reduction in deviance between a null model (only the

mean) and the focal model.

To test whether fishery characteristics affected our results, we used the

assessment data to build a GLM for the proportion of stocks collapsed as a function of

relative fishing mortality. We then used the model residuals in linear regressions that

included each of the life history characteristics.

3.3.7 Phylogenetically independent contrasts

To correct for shared evolutionary history among species, we fit linear

regressions through the origin on phylogenetically independent contrasts generated

with the Analyses of Phylogenetics and Evolution (ape) package (Paradis et al. 2004)

in R 2.12.1. We used a simple phylogeny based upon the taxonomic classification of

each species and equal branch lengths. In all cases, the response variable was the

87

proportion of stocks collapsed within each species. This approach cannot account for

differences in variance driven by the number of stocks in each species.

3.4 Results

We first examined life history traits of species targeted by global fisheries.

Though fisheries caught the entire range of trophic levels and growth rates seen among

marine fish, fisheries tended to catch larger, higher trophic level, and slower-growing

species (p < 10-7) (Fig. 3.1). In addition, the very smallest species (< 8 cm) did not

appear in global, industrial fisheries. There were no substantial size or trophic level

differences between the species that appear in the global landings database and those

that appear in the scientific stock assessments, though the latter are on average

somewhat slower growing (p = 10-10).

Overall, we found that 17.0% or 25.1% of the stocks in each species had

collapsed, on average, in the assessment (n = 52) or landings data (n = 223),

respectively. In addition, 23.3% or 34.9% of species had experienced at least one stock

collapse (assessments n = 28, or landings n = 160, respectively, Table 3.1).

One hypothesis was that large, high trophic level species would show a higher

incidence of collapse than small, low trophic level species. Instead, the assessment

data revealed fisheries collapses across the range of life history traits (Fig. 3.2a-e).

Among top predators (TL > 4.2), 12% of stocks had collapsed, while twice the

percentage (25%) had collapsed among low trophic level fishes (TL < 3.3). Among

large species (> 16 kg), 16% of stocks had collapsed, while 29% had collapsed in

small species (< 2.5 kg).

While the above comparisons suggested more collapses in small, low trophic

level fishes, Generalized Linear Models fit to the assessment data showed that the

proportion of stock collapses was not significantly related to trophic level (p = 0.15),

weight (p = 0.26), longevity (p = 0.10), age of maturity (p = 0.92), fecundity (p =

0.77), or investment in offspring (p = 0.99) (Table 3.2). The incidence of collapse,

however, was somewhat higher for fast growing species (p = 0.019).

88

The landings data generally supported the conclusion that large, top predators

are not more vulnerable than small, low trophic level fishes (Fig. 3.3a-e). Among top

predators (TL > 4.2), 26% of stocks had collapsed, while a similar percentage (21%)

had collapsed in low trophic level species (TL < 3.4). Similarly, 36% of stocks had

collapsed in large species (> 30 kg), while 31% had collapsed in small species (< 2.4

kg). Models suggested that the proportion of stocks collapsed did not change with

trophic level (p = 0.22), fecundity (p = 0.23), or investment in offspring (p = 0.52).

Trends were weakly towards more collapses among long-lived species (p = 0.012),

those with later maturity (p = 0.036), heavier species (p = 0.014), and those with

slower growth (p = 0.048), opposite to the trends in the assessment data. Trends were

not significant after correction for multiple comparisons.

In addition, we examined whether combinations of life history traits might

predict vulnerability, since multiple traits are often necessary for defining a life history

strategy (Winemiller 2005). However, combining life history traits in multiple

regression models could only predict 14% or 8% of the deviance in the data (p = 0.046

or p = 0.15, assessments and landings, respectively, Table 3.2), suggesting that

collapse incidence was not strongly related to life history in any combination. For the

models based on assessment data, higher growth rate was the most frequently included

trait, but these models suggested that species with higher growth rates experienced a

higher incidence of fisheries collapse. Previous authors have proposed that fishes with

low fecundity and high investment in offspring (large egg diameter) may be

particularly vulnerable to fishing (Winemiller 2005), but we did not find evidence that

this life history combination could explain the incidence of fishery collapses (0.8% or

1.1% of deviance explained, p = 0.79 or 0.59 within the assessment and landings data,

respectively).

We also explored the relationship between life history and alternative measures

of vulnerability, including incidence of overfished stocks (BMIN < 50% BMSY),

maximum depletion (BMIN/BMSY), and a less strict definition of landings collapse

(landings below 10% of the single highest landings recorded for a single year). All

trends were similar to those reported above (Fig. 3.4, 3.5, and 3.6). Removing small

89

pelagic species (families Engraulidae and Clupeidae) known to fluctuate strongly with

climate (Chavez et al. 2003) also did not change our results.

As we would expect, incidence of collapse was higher for species that

experienced greater relative fishing mortality (p = 1 x 10-5, 22% of deviance explained

among assessment data) and a longer history of developed fisheries (p = 5 x 10-5, 5.8%

of deviance explained among landings data) (Fig. 3.2f-g and 3.3f). However,

correcting for fisheries characteristics did not change the relationships between

collapse and life history (Fig. 3.7).

We also corrected for evolutionary relationships because phylogenetic history

can reduce the independence of species-level data (Felsenstein 1985). However, using

phylogenetically independent contrasts only revealed weaker relationships between

life history and incidence of collapse. These relationships reduced the discrepancy in

model results between assessment and landings data (Fig. 3.8 and Fig. 3.9).

Finally, we examined whether collapses last longer for certain life histories. In

the assessment data, collapses are longer (18.5 yrs) for long-lived species (lifespan >

44 yrs) than for short-lived species (5.1 yrs, lifespan < 14 yrs, p = 0.028). Other life

history comparisons were not significant (p > 0.21).

3.5 Discussion

Small, short-lived species have what is sometimes called a “fast” life history

strategy that is presumed to make them less vulnerable to fisheries (Jennings et al.

1998; Reynolds et al. 2005). In contrast, our review of global fisheries revealed that

these fast species collapse just as often as species with slower life histories. We found

collapsed stocks in short-lived species such as summer flounder (Paralichthys

dentatus) and Spanish mackerel (Scomberomorus maculatus) and among small, fast-

growing species like capelin (Mallotus villosus) and herring (Clupea harengus and C.

pallasii). Species low in the food chain had also collapsed, including winter flounder

(Pseudopleuronectes americanus) and chub mackerel (Scomber japonicus). While

these collapses are well known to local fishermen and managers, the general

prevalence of collapse among these types of species has not been recognized.

90

Our data suggest that species with fast life histories have at least as high a

probability (per stock) of declining to low abundance as larger, slower species, which

is dramatically different than the pattern among terrestrial species (Fisher & Owens

2004; Cardillo et al. 2005). Why might fast species be more vulnerable in the ocean

than we would expect? One explanation may be that fisheries management often

recommends higher exploitation rates for species with faster life histories and greater

productivity. For example, our assessment database revealed that the fishing mortality

predicted to supply maximum sustainable yield (FMSY) for long-lived rockfishes

(Sebastes spp.) is 0.07 (average of 20 stocks) while FMSY for short-lived skipjack tuna

is 0.42 (average of three stocks). Managing with these or similar reference points can

thus equalize the impact of fishing across species, a process with no widespread

equivalent on land. When fishing rate is correctly determined relative to species’

biology, life history should not be an important determinant of fish collapse.

This rationale may also explain differences between the assessment and

landings data. Many of the species in the landings database are not managed with

scientific stock assessments, and, presumably, fishing mortality is less closely matched

to stock productivity. Under such conditions, we might expect life history to be more

important for determining fisheries collapse and slow species to collapse more often.

In fact, collapses were slightly more common among long-lived species in the

landings database (Fig. 3.3a). Among the more closely managed stocks in the

assessment database, this trend was absent.

In addition, a fast life history may actually increase vulnerability to collapse.

Populations of short-lived species can grow or decline quickly in response to climatic

shifts (Chavez et al. 2003), and a rapid decline in productivity often requires similarly

rapid reductions in fishing effort to prevent collapse (Bakun & Broad 2003). If

fisheries management lags behind these biophysical changes, a population can be

driven to collapse (Bakun & Broad 2003; Fryxell et al. 2010). The high harvest rates

on many short-lived species also mean that errors in setting harvest rates can have

particularly severe consequences. In addition, we note that environmental variability

along can drive variation in fish abundance (Baumgartner et al. 1992; Chavez et al.

91

2003), and when of sufficiently large magnitude, this variation will be detected as a

collapse by our methods. Short-lived species may be particularly sensitive to such

environmental variability because of their fast growth rates and short generation times

(Winemiller 2005). Long-lived species respond more slowly to changes in climatic

conditions because they store more biomass in older age groups and are less dependent

on recent recruitment success (Winemiller 2005).

Our findings contrast with previous studies suggesting that population declines

in marine species are correlated with large size, late age of maturity, slow growth rate,

and high trophic level (Jennings et al. 1998; Jennings et al. 1999a; Jennings et al.

1999b; Dulvy et al. 2000; Dulvy & Reynolds 2002; Byrnes et al. 2007; Olden et al.

2007). However, previous analyses focused primarily on the North Atlantic (Jennings

et al. 1998; Jennings et al. 1999a; Dulvy et al. 2000) or on bycatch and artisanally

fished species (Jennings et al. 1999b; Dulvy et al. 2000; Dulvy & Reynolds 2002;

Byrnes et al. 2007). The particular decline of large, high trophic level species appears

to be a unique pattern of the North Atlantic that does not apply globally (Essington et

al. 2006; Fisher et al. 2010). Bycatch and artisanally fished species are less likely to

appear in the stock assessments or global landings that we analyzed. The adjustment

of fishing pressure as a function of species’ productivity, as explained above, is also

unlikely to occur for these weakly or un-managed species. The small proportion of

marine fishes that have been assessed under the World Conservation Union’s

(IUCN’s) Red List of Threatened Species (Baillie et al. 2004) have tended to be larger

species, perhaps helping to explain why listed, threatened species tend to be larger

than unlisted species (Olden et al. 2007).

While our paper has focused on testing whether lifespan, age of maturity, size,

trophic level, growth rates, fecundity, and offspring investment are useful predictors of

fisheries collapse, life histories are multifaceted strategies that include traits we could

not examine. However, many life history traits among fishes are strongly correlated,

including body size, natural mortality, size at maturity, and population growth rate at

low abundance (Roff 1984; Charnov 1993; Frisk et al. 2001; Denney et al. 2002). In

addition, our use of multivariate models allowed us to test whether combinations of

92

traits might be useful predictors of collapse. We note that life history evolution often

reflects a complex adaptive response to the scales of environmental variation (Roff

2002; Winemiller 2005), and life histories often diverge from a simple slow vs. fast

dichotomy (Winemiller 2005).

The high incidence of collapse that we uncover among small, short-lived, low

trophic level species has important implications for ecosystem structure and function,

especially in “wasp-waisted” ecosystems where a few species play a large role in

transferring food energy to higher trophic level fishes, birds, and marine mammals

(Cury et al. 2000). For example, sandeels (Ammodytes marinus) are targeted by the

largest single-species fishery in the North Sea, and declining sandeel abundance

causes severely reduced breeding success in seabirds (Frederiksen et al. 2004). Other

studies suggest similar sensitivity to a few small fish species across a range of seabirds

and pinnipeds (Duffy 1983; Guénette et al. 2006; Crawford 2007). Even though

short-lived species may recover more quickly from collapse than other fishes

(Hutchings 2000), collapses in small, low trophic level species can last from years to

decades (Beverton 1990). These durations are long enough to have substantial impacts

on the food web (Duffy 1983; Frederiksen et al. 2004; Crawford 2007).

In summary, analysis of stock assessment and global landings databases

revealed that patterns of vulnerability in the ocean are dramatically different from

those on land, and that both small and large fishes are vulnerable to collapse. A major

driver of differences between marine and terrestrial vulnerabilities may be the

importance of harvest versus habitat loss in these different ecosystems. A halt to

overfishing is needed across the full spectrum of life histories, not just for top

predators, to reduce the incidence of fishery collapses and to avoid the ecological,

economic, and social disruption that they cause.

3.6 Acknowledgements

This research was part of an NCEAS Distributed Graduate Seminar. The

authors thank R. Froese, T. Branch, R. Hilborn, B. Worm, and S. Tracey for insightful

feedback and access to data. NSF and NDSEG graduate fellowships (MLP), a David

93

H. Smith Postdoctoral Fellowship (OPJ), NSF CAMEO grant #1041678 (OPJ), and

the Census of Marine Life/Future of Marine Animal Populations (DR) provided

generous support. Financial support for the assessment database was provided by

NSERC grants to J. A. Hutchings and a Canadian Foundation for Innovation Grant to

H. Lotze.

3.7 References

Baillie J.E.M., Hilton-Taylor C. & Stuart S. (eds.) (2004). 2004 IUCN Red List of

threatened species: a global species assessment. IUCN, Gland, Switzerland

and Cambridge, UK.

Bakun A. & Broad K. (2003). Environmental 'loopholes' and fish population

dynamics: comparative pattern recognition with focus on El Niño effects in the

Pacific. Fish. Oceanogr., 12, 458-473.

Baum J.K., Myers R.A., Kehler D.G., Worm B., Harley S.J. & Doherty P.A. (2003).

Collapse and conservation of shark populations in the Northwest Atlantic.

Science, 299, 389-392.

Baumgartner T.R., Soutar A. & Ferreira-Bartrina V. (1992). Reconstruction of the

history of Pacific sardine and northern anchovy populations over the past two

millenia from sediments of the Santa Barbara basin, California. CalCOFI

Reports, 33, 24-40.

Beverton R.J.H. (1990). Small marine pelagic fish and the threat of fishing: Are they

endangered? J. Fish Biol., 37, 5-16.

Burnham K.P. & Anderson D.R. (2002). Model selection and multimodel inference: a

practical information-theoretic approach. 2nd edn. Springer, New York.

Byrnes J.E., Reynolds P.L. & Stachowicz J.J. (2007). Invasions and extinctions

reshape coastal marine food webs. PLoS One, 2, e295.

Cardillo M., Mace G.M., Jones K.E., Bielby J., Bininda-Emonds O.R.P., Sechrest W.,

Orme C.D.L. & Purvis A. (2005). Multiple causes of high extinction risk in

large mammal species. Science, 309, 1239-1241.

94

Charnov E.L. (1993). Life history invariants: some explorations of symmetry in

evolutionary ecology. Oxford University Press, Oxford.

Chavez F.P., Ryan J., Lluch-Cota S.E. & Ñiquen M.C. (2003). From anchovies to

sardines and back: multidecadal change in the Pacific Ocean. Science, 299,

217-21.

Costello C., Gaines S.D. & Lynham J. (2008). Can catch shares prevent fisheries

collapse? Science, 321, 1678-1681.

Crawford R.J.M. (2007). Food, fishing and seabirds in the Benguela upwelling

system. Journal of Ornithology, 148, 253-260.

Cury P., Bakun A., Crawford R.J.M., Jarre A., Quiñones R.A., Shannon L.J. &

Verheye H.M. (2000). Small pelagics in upwelling systems: patterns of

interaction and structural changes in “wasp-waist” ecosystems. ICES J. Mar.

Sci., 57, 603-618.

de Mutsert K., Cowan J.H., Essington T.E. & Hilborn R. (2008). Reanalyses of Gulf

of Mexico fisheries data: landings can be misleading in assessments of

fisheries and fisheries ecosystems. P Natl Acad Sci USA, 105, 2740-2744.

Denney N.H., Jennings S. & Reynolds J.D. (2002). Life-history correlates of

maximum population growth rates in marine fishes. P Roy Soc Lond B Bio,

269, 2229-2237.

Dobson A.J. (2002). An Introduction to Generalized Linear Models. 2nd edn.

Chapman & Hall, Boca Raton.

Duffy D.C. (1983). Environmental uncertainty and commercial fishing: effects on

Peruvian guano birds. Biol. Conserv., 26, 227-238.

Dulvy N.K., Metcalfe J.D., Glanville J., Pawson M.G. & Reynolds J.D. (2000).

Fishery stability, local extinctions, and shifts in community structure in skates.

Conserv. Biol., 14, 283-293.

Dulvy N.K. & Reynolds J.D. (2002). Predicting extinction vulnerability in skates.

Conserv. Biol., 16, 440-450.

Dulvy N.K., Sadovy Y. & Reynolds J.D. (2003). Extinction vulnerability in marine

populations. Fish Fish., 4, 25-64.

95

Essington T.E., Beaudreau A.H. & Wiedenmann J. (2006). Fishing through marine

food webs. P Natl Acad Sci USA, 103, 3171-3175.

Felsenstein J. (1985). Phylogenies and comparative method. Am. Nat., 125, 1-15.

Fisher D.O. & Owens I.P.F. (2004). The comparative method in conservation biology.

Trends Ecol. Evol., 19, 391-398.

Fisher J.A.D., Frank K.T. & Leggett W.C. (2010). Global variation in marine fish

body size and its role in biodiversity–ecosystem functioning. Mar. Ecol. Prog.

Ser., 405, 1-13.

Frederiksen M., Wanless S., Harris M.P., Rothery P. & Wilson L.J. (2004). The role

of industrial fisheries and oceanographic change in the decline of North Sea

black-legged kittiwakes. J. Appl. Ecol., 41, 1129-1139.

Frisk M.G., Miller T.J. & Fogarty M.J. (2001). Estimation and analysis of biological

parameters in elasmobranch fishes: a comparative life history study. Can. J.

Fish. Aquat. Sci., 58, 969-981.

Froese R. & Pauly D. (eds.) (2010). FishBase. World Wide Web electronic

publication. www.fishbase.org. Visited 4/01/2011.

Fryxell J.M., Packer C., McCann K., Solberg E.J. & Saether B.E. (2010). Resource

Management Cycles and the Sustainability of Harvested Wildlife Populations.

Science, 328, 903-906.

Guénette S., Heymans S.J.J., Christensen V. & Trites A.W. (2006). Ecosystem models

show combined effects of fishing, predation, competition, and ocean

productivity on Steller sea lions (Eumetopias jubatus) in Alaska. Can. J. Fish.

Aquat. Sci., 63, 2495-2517.

Hilborn R. & Walters C.J. (1992). Quantitative fisheries stock assessment: choice,

dynamics, and uncertainty. Kluwer Academic Publishers, Boston.

Hutchings J.A. (2000). Collapse and recovery of marine fishes. Nature, 406, 882-885.

Jennings S., Greenstreet S.P.R. & Reynolds J.D. (1999a). Structural change in an

exploited fish community: a consequence of differential fishing effects on

species with contrasting life histories. J. Anim. Ecol., 68, 617-627.

96

Jennings S., Reynolds J.D. & Mills S.C. (1998). Life history correlates of responses to

fisheries exploitation. P Roy Soc Lond B Bio, 265, 333-339.

Jennings S., Reynolds J.D. & Polunin N.V.C. (1999b). Predicting the vulnerability of

tropical reef fishes to exploitation with phylogenies and life histories. Conserv.

Biol., 13, 1466-1475.

King J.R. & McFarlane G.A. (2003). Marine fish life history strategies: applications to

fishery management. Fish. Manage. Ecol., 10, 249-264.

Mullon C., Fréon P. & Cury P. (2005). The dynamics of collapse in world fisheries.

Fish Fish., 6, 111-120.

Myers R.A. & Worm B. (2003). Rapid worldwide depletion of predatory fish

communities. Nature, 423, 280-283.

Olden J.D., Hogan Z.S. & Zanden M.J.V. (2007). Small fish, big fish, red fish, blue

fish: size-biased extinction risk of the world's freshwater and marine fishes.

Global Ecol. Biogeogr., 16, 694-701.

Paradis E., Claude J. & Strimmer K. (2004). APE: Analyses of phylogenetics and

evolution in R language. Bioinformatics, 20, 289-290.

Pauly D., Christensen V., Dalsgaard J., Froese R. & Torres F., Jr. (1998). Fishing

down marine food webs. Science, 279, 860-863.

Reynolds J.D., Dulvy N.K., Goodwin N.B. & Hutchings J.A. (2005). Biology of

extinction risk in marine fishes. P Roy Soc B-Biol Sci, 272, 2337-2344.

Roff D.A. (1984). The evolution of life history parameters in teleosts. Can. J. Fish.

Aquat. Sci., 41, 989-1000.

Roff D.A. (2002). Life history evolution. Sinauer Associates, Sunderland, MA.

Sethi S.A., Branch T.A. & Watson R. (2010). Global fishery development patterns are

driven by profit but not trophic level. P Natl Acad Sci USA, 107, 12163-12167.

von Bertalanffy L. (1938). A Quantitative Theory of Organic Growth (Inquiries on

Growth Laws. II.). Hum Biol, 10, 181-213.

Wilberg M.J. & Miller T.J. (2007). Comment on "Impacts of biodiversity loss on

ocean ecosystem services". Science, 316, 1285c.

97

Winemiller K.O. (2005). Life history strategies, population regulation, and

implications for fisheries management. Can. J. Fish. Aquat. Sci., 62, 872-885.

Worm B., Barbier E.B., Beaumont N., Duffy J.E., Folke C., Halpern B.S., Jackson

J.B.C., Lotze H.K., Micheli F., Palumbi S.R., Sala E., Selkoe K.A., Stachowicz

J.J. & Watson R. (2006). Impacts of biodiversity loss on ocean ecosystem

services. Science, 314, 787-790.

Worm B., Hilborn R., Baum J.K., Branch T.A., Collie J.S., Costello C., Fogarty M.J.,

Fulton E.A., Hutchings J.A., Jennings S., Jensen O.P., Lotze H.K., Mace P.M.,

McClanahan T.R., Minto C., Palumbi S.R., Parma A.M., Ricard D., Rosenberg

A.A., Watson R. & Zeller D. (2009). Rebuilding global fisheries. Science, 325,

578.

98

3.8 Tables

Table 3.1. Species with stocks that had collapsed a) according to stock

assessments (<20% of BMSY) and b) in the landings database (< 10% of the average of

the five highest years of landings, for at least two consecutive years). The species have

been sorted by family.

a) Species Family Number of

stocks Proportion of

stocks collapsed Trachurus murphyi Carangidae 1 1 Clupea harengus harengus Clupeidae 10 0.4 Clupea pallasii pallasii Clupeidae 5 1 Sardinops sagax Clupeidae 2 1 Scorpaenichthys marmoratus Cottidae 2 0.5 Gadus macrocephalus Gadidae 4 0.5 Gadus morhua Gadidae 17 0.59 Melanogrammus aeglefinus Gadidae 8 0.38 Pollachius virens Gadidae 4 0.25 Rexea solandri Gempylidae 2 0.5 Lutjanus campechanus Lutjanidae 1 1 Merluccius hubbsi Merlucciidae 2 0.5 Mallotus villosus Osmeridae 2 0.5 Paralichthys dentatus Paralichthyidae 1 1 Hippoglossoides elassodon Pleuronectidae 2 0.5 Hippoglossoides platessoides Pleuronectidae 2 0.5 Lepidopsetta polyxystra Pleuronectidae 1 1 Limanda aspera Pleuronectidae 1 1 Limanda ferruginea Pleuronectidae 4 1 Pseudopleuronectes americanus Pleuronectidae 3 0.67 Reinhardtius hippoglossoides Pleuronectidae 3 0.33 Cynoscion regalis Sciaenidae 1 1 Scomber japonicus Scombridae 1 1 Scomberomorus maculatus Scombridae 1 1 Sebastes alutus Sebastidae 3 0.33 Sebastes levis Sebastidae 1 1 Sebastes paucispinis Sebastidae 1 1 Epinephelus niveatus Serranidae 1 1

99

b) Species Family Number of

stocks Proportion of

stocks collapsed Albula vulpes Albulidae 1 1 Anoplopoma fimbria Anoplopomatidae 3 0.33 Ariomma indica Ariommatidae 1 1 Caranx crysos Carangidae 2 0.5 Caranx hippos Carangidae 3 0.33 Decapterus russelli Carangidae 3 0.33 Gnathanodon speciosus Carangidae 1 1 Megalaspis cordyla Carangidae 4 0.25 Seriola lalandi Carangidae 3 0.33 Trachinotus carolinus Carangidae 1 1 Trachurus declivis Carangidae 2 0.5 Trachurus japonicus Carangidae 2 0.5 Trachurus lathami Carangidae 1 1 Trachurus symmetricus Carangidae 2 1 Nemadactylus bergi Cheilodactylidae 1 1 Chirocentrus dorab Chirocentridae 4 0.5 Alosa pseudoharengus Clupeidae 2 0.5 Alosa sapidissima Clupeidae 4 0.75 Brevoortia aurea Clupeidae 1 1 Brevoortia pectinata Clupeidae 1 1 Brevoortia tyrannus Clupeidae 2 0.5 Clupea pallasii pallasii Clupeidae 3 0.33 Clupeonella cultriventris Clupeidae 1 1 Dussumieria elopsoides Clupeidae 1 1 Etrumeus teres Clupeidae 3 0.33 Opisthonema oglinum Clupeidae 3 0.67 Sardinella zunasi Clupeidae 1 1 Sardinops sagax Clupeidae 7 0.43 Spratelloides gracilis Clupeidae 2 0.5 Sprattus fuegensis Clupeidae 1 1 Tenualosa ilisha Clupeidae 2 0.5 Coryphaena hippurus Coryphaenidae 9 0.11 Cyclopterus lumpus Cyclopteridae 1 1 Eleginops maclovinus Eleginopidae 2 1 Elops saurus Elopidae 3 0.33 Cetengraulis edentulus Engraulidae 1 1

100

Engraulis mordax Engraulidae 2 0.5 Engraulis ringens Engraulidae 1 1 Gadus morhua Gadidae 1 1 Gadus ogac Gadidae 1 1 Melanogrammus aeglefinus Gadidae 1 1 Microgadus tomcod Gadidae 1 1 Thyrsites atun Gempylidae 4 0.5 Conodon nobilis Haemulidae 1 1 Orthopristis chrysoptera Haemulidae 1 1 Ophiodon elongatus Hexagrammidae 2 0.5 Hypoptychus dybowskii Hypoptychidae 1 1 Istiophorus platypterus Istiophoridae 7 0.86 Makaira indica Istiophoridae 5 0.8 Makaira nigricans Istiophoridae 11 0.55 Tetrapturus albidus Istiophoridae 4 0.75 Tetrapturus audax Istiophoridae 7 0.71 Lamna nasus Lamnidae 2 0.5 Lates calcarifer Latidae 4 0.25 Lutjanus argentiventris Lutjanidae 2 0.5 Lutjanus purpureus Lutjanidae 2 0.5 Coryphaenoides rupestris Macrouridae 1 1 Macrourus berglax Macrouridae 1 1 Lopholatilus chamaeleonticeps Malacanthidae 2 1 Megalops atlanticus Megalopidae 2 0.5 Megalops cyprinoides Megalopidae 1 1 Mene maculata Menidae 2 0.5 Merluccius bilinearis Merlucciidae 1 1 Merluccius productus Merlucciidae 2 0.5 Stephanolepis cirrhifer Monacanthidae 1 1 Salilota australis Moridae 2 0.5 Morone saxatilis Moronidae 2 1 Joturus pichardi Mugilidae 2 0.5 Valamugil seheli Mugilidae 1 1 Dissostichus eleginoides Nototheniidae 5 0.2 Patagonotothen brevicauda Nototheniidae 1 1 Patagonotothen ramsayi Nototheniidae 1 1 Genypterus chilensis Ophidiidae 1 1 Hypomesus pretiosus Osmeridae 1 1 Mallotus villosus Osmeridae 2 1 Thaleichthys pacificus Osmeridae 1 1

101

Paralichthys dentatus Paralichthyidae 2 0.5 Pseudopentaceros richardsoni Pentacerotidae 1 1 Urophycis chuss Phycidae 1 1 Prolatilus jugularis Pinguipedidae 1 1 Atheresthes stomias Pleuronectidae 2 0.5 Glyptocephalus cynoglossus Pleuronectidae 1 1 Hippoglossoides elassodon Pleuronectidae 2 1 Hippoglossoides platessoides Pleuronectidae 1 1 Hippoglossus hippoglossus Pleuronectidae 1 1 Hippoglossus stenolepis Pleuronectidae 2 0.5 Lepidopsetta bilineata Pleuronectidae 2 0.5 Limanda aspera Pleuronectidae 2 0.5 Limanda ferruginea Pleuronectidae 1 1 Pleuronectes quadrituberculatus Pleuronectidae 2 0.5 Reinhardtius hippoglossoides Pleuronectidae 3 0.33 Eleutheronema tetradactylum Polynemidae 3 0.67 Polyprion americanus Polyprionidae 3 0.33 Polyprion oxygeneios Polyprionidae 2 0.5 Stereolepis gigas Polyprionidae 1 1 Pomatomus saltatrix Pomatomidae 10 0.1 Rhinobatos percellens Rhinobatidae 1 1 Rhinobatos planiceps Rhinobatidae 1 1 Oncorhynchus kisutch Salmonidae 3 0.67 Oncorhynchus tshawytscha Salmonidae 3 0.33 Salmo salar Salmonidae 1 1 Salvelinus alpinus alpinus Salmonidae 1 1 Argyrosomus hololepidotus Sciaenidae 5 0.2 Argyrosomus regius Sciaenidae 3 0.33 Atractoscion aequidens Sciaenidae 1 1 Atractoscion nobilis Sciaenidae 1 1 Atrobucca nibe Sciaenidae 1 1 Cynoscion regalis Sciaenidae 2 1 Genyonemus lineatus Sciaenidae 1 1 Menticirrhus littoralis Sciaenidae 1 1 Menticirrhus saxatilis Sciaenidae 1 1 Micropogonias undulatus Sciaenidae 2 0.5 Paralonchurus peruanus Sciaenidae 1 1 Pogonias cromis Sciaenidae 3 0.67 Sciaenops ocellatus Sciaenidae 1 1 Totoaba macdonaldi Sciaenidae 1 1

102

Umbrina canosai Sciaenidae 1 1 Acanthocybium solandri Scombridae 6 0.17 Euthynnus affinis Scombridae 4 0.25 Euthynnus lineatus Scombridae 2 0.5 Katsuwonus pelamis Scombridae 13 0.31 Sarda chiliensis chiliensis Scombridae 2 1 Sarda sarda Scombridae 6 0.17 Scomber japonicus Scombridae 11 0.09 Scomber scombrus Scombridae 2 0.5 Scomberomorus cavalla Scombridae 3 0.67 Scomberomorus guttatus Scombridae 4 0.25 Thunnus alalunga Scombridae 14 0.29 Thunnus albacares Scombridae 13 0.23 Thunnus maccoyii Scombridae 6 0.83 Thunnus obesus Scombridae 12 0.42 Thunnus orientalis Scombridae 3 0.67 Thunnus thynnus Scombridae 5 0.6 Thunnus tonggol Scombridae 4 0.25 Lepidorhombus whiffiagonis Scophthalmidae 1 1 Scophthalmus aquosus Scophthalmidae 1 1 Sebastes alutus Sebastidae 3 1 Sebastes entomelas Sebastidae 2 1 Sebastes goodei Sebastidae 1 1 Sebastes mentella Sebastidae 1 1 Sebastes paucispinis Sebastidae 1 1 Sebastes pinniger Sebastidae 1 1 Epinephelus analogus Serranidae 1 1 Epinephelus morio Serranidae 2 1 Mycteroperca xenarcha Serranidae 1 1 Argyrops spinifer Sparidae 1 1 Pagellus bellottii Sparidae 2 0.5 Pagrus pagrus Sparidae 2 0.5 Stenotomus chrysops Sparidae 1 1 Pampus argenteus Stromateidae 4 0.25 Peprilus paru Stromateidae 2 0.5 Peprilus triacanthus Stromateidae 2 0.5 Saurida undosquamis Synodontidae 2 0.5 Sphoeroides maculatus Tetraodontidae 1 1 Galeorhinus galeus Triakidae 3 0.33 Chelidonichthys capensis Triglidae 1 1

103

Chelidonichthys kumu Triglidae 3 0.33 Xiphias gladius Xiphiidae 14 0.14 Zeus faber Zeidae 5 0.4 Zoarces americanus Zoarcidae 1 1

104

T

able 3.2. Parameters for m

odels predicting the proportion of stocks collapsed within each species. a) Single-variable

regressions with G

eneralized Linear M

odels (GL

Ms) for stock assessm

ent species, b) single-variable GL

Ms for species w

ith landings

data, c) minim

al adequate models from

multiple regressions w

ith stock assessment species, d) m

inimal adequate m

odels from m

ultiple

regressions for species with landings data. C

onfidence Intervals (CIs) for single-variable G

LM

s refer to the coefficient. The L

og

column indicates w

hether the explanatory variable has been log-transformed. M

inimal adequate m

odels were chosen as those w

ith the

lowest A

IC (∆

AIC

< 2) from am

ong all possible GL

Ms w

ithout interactions.

a) Life H

istory Trait

n Intercept

Coefficient

Low

er 95%

CI

Upper

95% C

I N

ull D

eviance R

esidual D

eviance p

Log

Longevity (yr)

97 0.24

-0.41 -0.92

0.068 139.51

136.69 0.10

* A

ge of maturity (yr)

96 -1.1

-0.03 -0.63

0.56 130.7

130.69 0.92

* L

ength (cm)

120 -0.17

-0.23 -0.75

0.28 157

156.25 0.39

* W

eight (kg) 93

-0.12 -0.1

-0.28 0.074

137.68 136.41

0.26 *

Trophic level

120 0.32

-0.4 -0.96

0.15 157

154.97 0.15

G

rowth rate (K

) 120

-0.0047 0.71

0.13 1.3

157 151.28

0.019 *

Fecundity (eggs) 93

-1.4 0.025

-0.14 0.2

134.07 133.98

0.77 *

Egg diam

eter (mm

) 97

-1.2 -0.0047

-0.82 0.72

134.46 134.46

0.99 *

Fishery Initiation (yr) 46

60 -0.031

-0.083 0.015

67.54 65.83

0.20

Relative Fishing

Mortality

99 -2.4

1.6 0.94

2.4 128.17

99.7 1.2x10

-5 *

105

b) L

ife

His

tory

Tra

it

n In

terc

ept

Coe

ffic

ient

L

ower

95

% C

I U

pper

95

% C

I N

ull

Dev

ianc

e R

esid

ual

Dev

ianc

e p

Log

Lon

gevi

ty (

yr)

206

-1.9

0.

33

0.07

3 0.

59

318.

88

312.

54

0.01

2 *

Age

of

mat

urity

(yr

) 21

6 -1

.4

0.3

0.02

1 0.

57

302.

91

298.

46

0.03

6 *

Len

gth

(cm

) 45

7 -2

.0

0.19

0.

0047

0.

37

616.

05

612.

01

0.04

5 *

Wei

ght (

kg)

267

-2.0

0.

11

0.02

2 0.

19

390.

13

383.

98

0.01

4 *

Tro

phic

leve

l 45

7 -1

.7

0.16

-0

.093

0.

43

616.

05

614.

5 0.

22

G

row

th r

ate

(K)

447

-1.4

-0

.22

-0.4

3 -0

.003

4 60

3.48

59

9.51

0.

048

* Fe

cund

ity (

eggs

) 17

2 -1

.7

0.05

-0

.029

0.

14

243.

22

241.

73

0.23

*

Egg

dia

met

er (

mm

) 15

5 -0

.79

-0.1

2 -0

.52

0.25

24

9.64

24

9.23

0.

52

* Fi

sher

y In

itiat

ion

(yr)

20

8 69

-0

.035

-0

.053

-0

.019

31

2.24

29

4.07

5.

4x10

-5

106

c) Intercept

log L

ongevity log

Maturity

log W

eight T

rophic level

log G

rowth

Rate

log F

ecundity log E

gg D

iameter

Residual

Deviance

(Null =

96.9)

%

Deviance

AIC

p

0.71

0.99

89.6 7.50%

114.93

0.05 0.33

0.67

1.3

87.8

9.40%

115.11 0.063

0.83

1.2

0.79

88 9.20%

115.32

0.057 2.1

-0.46

0.73

88.1

9.10%

115.41 0.039

2.1

-0.45 0.91

0.76

86.6 11%

115.92

0.1 0.49

0.6

1.5

0.67 86.7

11%

115.95 0.081

3.9 -0.58

-0.8

88.7 8.50%

116.03

0.031 0.96

-0.8

1.1 0.21

1.2 84.8

12%

116.08 0.032

1.5

0.57

-0.38 1.1

86.9 10%

116.17

0.086 2.1

-0.78

91.1

6%

116.43 0.037

0.056

0.69

-0.76 1.5

0.25 1.1

83.3 14%

116.57

0.046 4.5

-0.74

-0.87

0.7 87.4

9.80%

116.72 0.093

1.4

-0.65 0.81

0.11

87.5 9.70%

116.76

0.077 0.92

-0.034

0.93

89.5

7.60%

116.84 0.064

-0.34

0.75

1.5

0.059

87.6 9.60%

116.88

0.093 0.59

0.059

1

89.6

7.50%

116.91 0.082

0.64

1

0.007

89.6 7.50%

116.92

0.062

107

d)

Inte

rcep

t lo

g L

onge

vity

lo

g M

atur

ity

log

Wei

ght

Tro

phic

le

vel

log

Gro

wth

R

ate

log

Fec

undi

ty

log

Egg

D

iam

eter

Res

idua

l D

evia

nce

(Nul

l = 1

07.7

)

%

Dev

ianc

e A

IC

p

-0.4

1

0.45

0.

23

-0.8

1

99.4

7.

70%

14

5.8

0.2

-0.0

11

0.19

-0

.83

-0.4

1

10

0.2

7%

146.

58

0.15

-1

.3

0.

39

10

4.4

3.10

%

146.

8 0.

14

0.53

-0

.32

0.72

0.

24

-0.9

3

98.7

8.

40%

14

7.07

0.

15

-1.4

-0

.36

104.

9 2.

60%

14

7.28

0.

6 0.

0042

-0.3

9 -0

.44

102.

9 4.

50%

14

7.28

0.

31

-0.4

8

0.5

0.23

-0

.79

-0.1

1 99

.2

7.90

%

147.

62

0.27

-0

.35

0.

35

0.22

-0

.82

-0.1

4

99

.2

7.90

%

147.

63

0.2

-0.6

9

0.47

0.

22

-0.7

9

0.02

8

99.3

7.

80%

14

7.67

0.

09 9 0.

26

0.2

-0.7

7

103.

4 4%

14

7.78

0.

1

108

3.9 Figures

Figure 3.1. Life history patterns of fished species. Histograms for all marine

fish (black), species in the landings database (grey), and species in the assessment

database (white) for a) length (N = 16548/457/120 for all/landings/assessments), b)

trophic level (N = 16548/457/120), and c) growth rate (N = 14118/447/120).

Length (cm)

Prop

ortio

n0.

00.

20.

4

100 101 102 103

a)

Trophic Level0.

000.

150.

30

2 3 4 5

b)

Growth Rate (K)

0.00

0.15

0.30

10!2 10!1 100 101

c)

109

Figure 3.2. Collapses in the assessment database in relation to life history

traits. Traits include a) lifespan (N = 97), b) age of maturity (N = 96), c) weight (N =

93), d) trophic level (N = 120), e) growth rate (N = 120), f) fecundity (N = 93), g)

investment in offspring (egg diameter, N = 97), h) year of fishery initiation (N = 46),

and i) relative fishing mortality (N = 99). Each dot represents the proportion of stocks

collapsed within a species. All x-axes are log-transformed except those for trophic

level and fishery initiation. Dashed line is the best fit from a Generalized Linear

Model.

Longevity (yr)

Prop

ortio

n co

llaps

ed

3 10 30 100

0.0

0.5

1.0

a)

Age of maturity (yr)1 3 10 30

0.0

0.5

1.0

b)

Weight (g)102 103 104 105

0.0

0.5

1.0

c)

Trophic Level

Prop

ortio

n co

llaps

ed

2 3 4

0.0

0.5

1.0

d)

Growth rate (K)0.04 0.10 0.40

0.0

0.5

1.0

e)

Fecundity (# eggs)101 103 105 107

0.0

0.5

1.0

f)

Egg diameter (mm)

Prop

ortio

n co

llaps

ed

1 3 10 30

0.0

0.5

1.0

g)

Fishery initiation (yr)1960 1980 2000

0.0

0.5

1.0

h)

Fishing mortality0.1 1.0 10.0

0.0

0.5

1.0

i)

110

Figure 3.3. Collapses in the landings database in relation to life history traits.

Traits include a) lifespan (N = 206), b) age of maturity (N = 216), c) weight (N = 267),

d) trophic level (N = 457), e) growth rate (N = 447), f) fecundity (N = 172), g)

investment in offspring (egg diameter, N = 155), and h) year of fishery initiation (N =

208). Also see notes for Figure 2.

Longevity (yr)

Prop

ortio

n co

llaps

ed

3 10 30 100

0.0

0.5

1.0

a)

Age of maturity (yr)1 3 10 30

0.0

0.5

1.0

b)

Weight (g)102 103 104 105

0.0

0.5

1.0

c)

Trophic Level

Prop

ortio

n co

llaps

ed

2 3 4

0.0

0.5

1.0

d)

Growth rate (K)0.04 0.40

0.0

0.5

1.0

e)

Fecundity (# eggs)101 103 105 107

0.0

0.5

1.0

f)

Egg diameter (mm)

Prop

ortio

n co

llaps

ed

1 3 10 30

0.0

0.5

1.0

g)

Fishery initiation (yr)1960 1980 2000

0.0

0.5

1.0

h)

111

Figure 3.4. Life history trends in the proportion of overfished stocks

(BMIN/BMSY < 50%) from the assessment data. Compare to Figure 2. Life history

characteristics include a) lifespan (p = 0.064, N = 97), b) age of maturity (p = 0.0095,

N = 96), c) maximum weight (p = 0.27, N = 93), d) trophic level (p = 0.0042, N =

120), e) growth rate (p = 0.010, N = 120), f) fecundity (p = 0.47, N = 93), g)

investment in offspring (egg diameter, p = 0.22, N = 97), h) year of fishery initiation

(p = 0.084, N = 46), and i) relative fishing mortality (p = 4 x 10-7, N = 99). All x-axes

are log-transformed except those for trophic level and fishery initiation. Each dot

represents one species. Dashed line is the best fit from a Generalized Linear Model.

Longevity (yr)

Prop

ortio

n ov

erfis

hed

3 10 30 100

0.0

0.5

1.0

a)

Age of maturity (yr)1 3 10 30

0.0

0.5

1.0

b)

Weight (g)102 103 104 105

0.0

0.5

1.0

c)

Trophic Level

Prop

ortio

n ov

erfis

hed

2 3 4

0.0

0.5

1.0

d)

Growth rate (K)0.04 0.10 0.40

0.0

0.5

1.0

e)

Fecundity (# eggs)101 103 105 107

0.0

0.5

1.0

f)

Egg diameter (mm)

Prop

ortio

n ov

erfis

hed

1 3 10 30

0.0

0.5

1.0

g)

Fishery initiation (yr)1960 1980 2000

0.0

0.5

1.0

h)

Fishing mortality0.1 1.0 10.0

0.0

0.5

1.0

i)

112

Figure 3.5. Life history patterns with an alternative definition of collapse for

landings data. For this figure, a stock is considered collapsed if it falls below 10% of

the maximum annual landings for a single year (our lax definition). Life history

characteristics include a) lifespan (p = 0.78, N = 206), b) age of maturity (p = 0.087, N

= 216), c) maximum weight (p = 0.0020, N = 267), d) trophic level (p = 0.23, N =

457), e) growth rate (p = 0.025, N = 447), f) fecundity (p = 0.13, N = 172, g)

investment in offspring (egg diameter, p = 0.29, N = 155), and h) year of fishery

initiation (p = 3 x 10-6, N = 208).

Longevity (yr)

Prop

ortio

n co

llaps

ed

3 10 30 100

0.0

0.5

1.0

a)

Age of maturity (yr)1 3 10 30

0.0

0.5

1.0

b)

Weight (g)102 103 104 105

0.0

0.5

1.0

c)

Trophic Level

Prop

ortio

n co

llaps

ed

2 3 4

0.0

0.5

1.0

d)

Growth rate (K)0.04 0.40

0.0

0.5

1.0

e)

Fecundity (# eggs)101 103 105 107

0.0

0.5

1.0

f)

Egg diameter (mm)

Prop

ortio

n co

llaps

ed

1 3 10 30

0.0

0.5

1.0

g)

Fishery initiation (yr)1960 1980 2000

0.0

0.5

1.0

h)

113

Figure 3.6. Life history trends in the magnitude of decline (BMIN/BMSY). Small

values represent a species that reached low abundance. Life history characteristics

include a) lifespan (p = 0.22, N = 97), b) age of maturity (p = 0.035, N = 96), c)

maximum weight (p = 0.24, N = 93), d) trophic level (p = 0.0016, N = 120), e) growth

rate (p = 0.034, N = 120), f) fecundity (p = 0.44, N = 93), g) investment in offspring

(egg diameter, p = 0.04, N = 97), h) year of fishery initiation (p = 0.41, N = 46), and i)

relative fishing mortality (p = 5 x 10-10, N = 99). Compare to Figure 2. Each dot

represents one species. Y-axes are log-transformed.

Longevity (yr)

Max

imum

dep

letio

n

3 10 30 100

10!2

10!1

100

a)

Age of maturity (yr)1 3 10 30

10!2

10!1

100

b)

Weight (g)102 103 104 105

10!2

10!1

100

c)

Trophic Level

Max

imum

dep

letio

n

2 3 4

10!2

10!1

100

d)

Growth rate (K)0.04 0.10 0.40

10!2

10!1

100

e)

Fecundity (# eggs)101 103 105 107

10!2

10!1

100

f)

Egg diameter (mm)

Max

imum

dep

letio

n

1 3 10 30

10!2

10!1

100

g)

Fishery initiation (yr)1960 1980 2000

10!2

10!1

100

h)

Fishing mortality0.1 1.0 10.0

10!2

10!1

100

i)

114

Figure 3.7. Correcting for relative fishery mortality has little impact on the

sign of the relationship between collapse probability and life history traits (assessment

data). Life history characteristics include a) lifespan (p = 0.025, N = 83), b) age of

maturity (p = 0.82, N = 82), c) maximum weight (p = 0.21, N = 75), d) trophic level

(p = 0.45, N = 99), e) growth rate (p = 0.039, N = 99), f) fecundity (p = 0.89, N = 77),

and g) investment in offspring (egg diameter, p = 0.43, N = 78). Y-axes represent the

residuals from a GLM that predicted the proportion of stocks collapsed from relative

fishing mortality. Positive values on the y-axes represent species that are more

collapsed than expected from fishery characteristics.

Longevity (yr)

Res

idua

ls

3 10 30 100

−10

12

a)

Age of maturity (yr)1 3 10 30

−10

12

b)

Weight (g)102 103 104 105

−10

12

c)

Trophic Level

Res

idua

ls

2 3 4

−10

12

d)

Growth rate (K)0.04 0.10 0.40

−10

12

e)

Fecundity (# eggs)101 103 105 107

−10

12

f)

Egg diameter (mm)

Res

idua

ls

1 3 10 30

−10

12

g)

115

Figure 3.8. Correcting for phylogeny with the assessment data also suggests

that incidence of collapse does not increase with life history traits, but does increase

when overfishing occurs. The y-axes are phylogenetically independent contrasts on the

proportion of stocks collapsed. The x-axes are contrasts on life history traits, including

a) lifespan (p = 0.37, N = 96), b) age of maturity (p = 0.64, N = 95), c) maximum

weight (p = 0.60, N = 92), d) trophic level (p = 0.90, N = 119), e) growth rate (p =

0.88, N = 119), f) fecundity (p = 0.55, N = 92), g) investment in offspring (egg

diameter, p = 0.60, N = 96), h) year of fishery initiation (p = 0.59, N = 45), and i)

relative fishing mortality (p = 1.2 x 10-5, N = 98). Dashed line is the best fit from a

linear regression through the origin.

−1.0 0.0 1.0

−1.0

0.0

1.0

Longevity contrast

Col

laps

ed c

ontra

st a)

−1.5 −0.5 0.5

−1.0

0.0

1.0

Maturity contrast

b)

−3 −1 1 2 3

−1.0

0.0

1.0

Weight contrast

c)

−1.0 0.0 0.5 1.0

−1.0

0.0

1.0

Trophic contrast

Col

laps

ed c

ontra

st d)

−1.0 0.0 1.0

−1.0

0.0

1.0

Growth rate contrast

e)

−4 0 2 4 6

−1.0

0.0

1.0

Fecundity contrast

f)

−1.5 −0.5 0.5

−1.0

0.0

1.0

Egg diam. contrast

Col

laps

ed c

ontra

st g)

−30 −10 10 30

−1.0

0.0

0.5

Initiation contrast

h)

−2 −1 0 1 2 3

−1.0

0.0

0.5

Mortality contrast

i)

116

Figure 3.9. Correcting for phylogeny with the landings data also suggests that

incidence of collapse does not increase with life history traits. The y-axes are

phylogenetically independent contrasts on the proportion of stocks collapsed. The x-

axes are contrasts on life history traits, including a) lifespan (p = 0.098, N = 205), b)

age of maturity (p = 0.36, N = 215), c) maximum weight (p = 0.099, N = 266), d)

trophic level (p = 0.12, N = 456), e) growth rate (p = 0.041, N = 446), f) fecundity (p =

1.0, N = 171), g) investment in offspring (egg diameter, p = 0.50, N = 154), and h)

year of fishery initiation (p = 0.13, N = 207). Dashed line is the best fit from a linear

regression through the origin.

−1.5 −0.5 0.5 1.5

−1.0

0.0

1.0

Longevity contrast

Col

laps

ed c

ontra

st a)

−2 −1 0 1 2

−1.0

0.0

1.0

Maturity contrast

b)

−4 −2 0 2 4

−1.0

0.0

0.5

1.0

Weight contrast

c)

−1.5 −0.5 0.5 1.5

−1.0

0.0

1.0

Trophic contrast

Col

laps

ed c

ontra

st d)

−2 −1 0 1

−1.0

0.0

1.0

Growth rate contrast

e)

−5 0 5

−1.0

0.0

1.0

Fecundity contrast

f)

−1.5 −0.5 0.5

−1.0

0.0

0.5

1.0

Egg diam. contrast

Col

laps

ed c

ontra

st g)

−40 −20 0 20 40

−1.0

0.0

0.5

Initiation contrast

h)

117

3.10 Supplementary Material

Table 3.S1. Additional sources consulted for data on a) egg diameter and b)

fecundity.

a) Species Data References Atheresthes stomias 1 Balistes capriscus 2 Brevoortia patronus 3 Dissostichus mawsoni 4 Engraulis anchoita 5 Eopsetta jordani 6, 7 Epinephelus morio 8 Genypterus blacodes 9 Genypterus capensis 10 Glyptocephalus zachirus 6 Hexagrammos decagrammus 11 Hippoglossus stenolepis 7, 12 Limanda aspera 12 Lophius americanus 13, 14 Lopholatilus chamaeleonticeps 15 Lutjanus analis 16 Mallotus villosus 17, 18 Merluccius australis 19 Merluccius capensis 10, 20 Merluccius paradoxus 21 Merluccius hubbsi 22 Micromesistius australis 23 Microstomus pacificus 12 Mycteroperca microlepis 24 Ocyurus chrysurus 25 Pagrus pagrus 26 Parophrys vetulus 12 Pleurogrammus monopterygius 27 Pleuronectes quadrituberculatus 28, 29 Raja rhina 7 Reinhardtius hippoglossoides 12, 20 Rhomboplites aurorubens 30 Scomberomorus cavalla 31, 32 Scomberomorus maculatus 33 Sebastes aleutianus 7

118

Sebastes alutus 7 Sebastes entomelas 7 Sebastes flavidus 7 Sebastes melanops 34 Sebastes paucispinis 7 Sebastes pinniger 7 Sebastes polyspinis 35 Sebastes ruberrimus 7 Sebastolobus alascanus 7 Sebastolobus altivelis 7 Seriola dumerili 36 Seriolella brama 37 Seriolella punctata 33 Squalus acanthias 7 Theragra chalcogramma 7, 32 Thunnus alalunga 7 Thunnus albacares 33 Thunnus obesus 33 Thunnus thynnus 32 Trachurus capensis 10 Urophycis tenuis 14

b) Species Data References Atheresthes stomias 38 Balistes capriscus 2 Centropristis striata 39 Dissostichus eleginoides 4 Dissostichus mawsoni 4 Engraulis anchoita 40 Eopsetta jordani 6, 7 Genypterus blacodes 9 Glyptocephalus cynoglossus 41 Glyptocephalus zachirus 6 Hexagrammos decagrammus 42 Hippoglossus stenolepis 7 Katsuwonus pelamis 43 Lepidopsetta bilineata 7 Merluccius capensis 21 Merluccius paradoxus 21 Micromesistius australis 44, 45 Micromesistius australis 46 Pagrus pagrus 47 Paralichthys dentatus 32

119

Parophrys vetulus 7 Platichthys stellatus 7 Pleurogrammus monopterygius 48 Pleuronectes quadrituberculatus 28 Pollachius virens 49 Pseudopleuronectes americanus 50 Raja rhina 7 Rhomboplites aurorubens 30 Sebastes aleutianus 7 Sebastes alutus 51 Sebastes alutus 7 Sebastes entomelas 52 Sebastes entomelas 53 Sebastes entomelas 51 Sebastes entomelas 7 Sebastes goodei 51 Sebastes levis 51 Sebastes paucispinis 53 Sebastes paucispinis 51 Sebastes paucispinis 7 Sebastes pinniger 53 Sebastes pinniger 51 Sebastes pinniger 7 Sebastes ruberrimus 7 Seriolella brama 37 Tautoga onitis 54 Thunnus alalunga 7 Thunnus obesus 55 Thunnus thynnus 32 Urophycis tenuis 56

1. Blood, DM, AC Matarese, and MS Busby. 2007. Spawning, egg development,

and early life history dynamics of arrowtooth flounder (Atheresthes stomias) in

the Gulf of Alaska. U.S. Dept. of Commerce. National Marine Fisheries

Service Professional Paper NMFS 7.

2. Simmons, CM. 2008. Gray triggerfish, Balistes capriscus, reproductive

behavior, early life history, and competitive interactions between red snapper,

Lutjanus campechanus, in the

northern Gulf of Mexico. Auburn University, Ph.D. dissertation.

120

3. Hettler, WF. 1984. Description of eggs, larvae, and early juveniles of Gulf

menhaden, Brevoortia patronus, and comparisons with Atlantic menhaden, B.

tyrannus, and yellowfin menhaden, B. smithi. Fishery Bulletin 82(1): 85-95.

4. Kock, K-H and A Kellerman. 1991. Reproduction in Antarctic notothenioid

fish. Antarctic Science 3(2): 125-150.

5. de Ciechomski, JD. 1973. The size of the eggs of the Argentine anchovy,

Engraulis anchoita (Hubbs & Marini) in relation to the season of the year and

to the area of spawning. J Fish Biol 5: 393-398.

6. Pearcy, WG, MJ Hosie, and SL Richardson. 1977. Distribution and duration of

pelagic life of larvae of Dover sole, Microstomus pacificus; Rex sole,

Glyptocephalus zachirus; and Petrale sole, Eopsetta jordani, in waters off

Oregon. Fishery Bulletin 75(1): 173-183.

7. King, JR and GA McFarlane. 2003. Marine fish life history strategies:

applications to fishery management. Fisheries Management and Ecology

10(4): 249-264.

8. Colin, PL, CC Koenig, and WA Laroche. 1996 in Biology, fisheries and

culture of tropical groupers and snappers, eds Arreguin-Sanchez, F, JL

Munro, MC Balgos, and D Pauly (ICLARM, Makati City, Philippines), pp

399-414.

9. Graham, DH. 1939. Breeding habits of the fishes of Otago Harbour and

adjacent seas. Transactions and Proceedings of the Royal Society of New

Zealand 69: 361-372.

10. Brownell, CL. 1979. Stages in the early development of 40 marine fish species

with pelagic eggs from the Cape of Good Hope. Ichthyological Bulletin of the

J. L. B. Smith Institute of Ichthyology 40: 1-84.

11. DeMartini, EE. 1986. Reproductive colorations, parental behavior and egg

masses of kelp greenling, Hexagrammos decagrammus, and whitespotted

greenling, H. stelleri. Northwest Science 60(1): 32-35.

12. Ahlstrom, EH, K Amaoka, DA Hensley, HG Moser, and BY Sumida. 1984 in

Ontogeny and Systematics of Fishes, ed Moser, HG (American Society of

Ichthyologists and Herpetologists), pp 640-670.

121

13. Everly, AW. 2002. Stages of development of the goosefish, Lophius

americanus, and comments on the phylogenetic significance of the

development of the luring apparatus in Lophiiformes. Environmental Biology

of Fishes 64: 393-417.

14. Markle, DF and L-A Frost. 1985. Comparative morphology, seasonality, and a

key to planktonic fish eggs from the Nova Scotian shelf. Canadian Journal of

Zoology 63: 246-257.

15. Grimes, CB, CF Idelberger, KW Able, and SC Turner. 1988. The reproductive

biology of tilefish, Lopholatilus chamaeleonticeps Goode and Bean, from the

United States Mid-Atlantic Bight, and the effects of fishing on the breeding

system. Fishery Bulletin 86(4): 745-762.

16. Watanabe, WO, EP Ellis, SC Ellis, J Chaves, and C Manfredi. 1998. Artificial

propagation of mutton snapper Lutjanus analis, a new candidate marine fish

species for aquaculture. Journal of the World Aquaculture Society 29(2): 176-

187.

17. Gjosaeter, H. 1998. The population biology and exploitation history of capelin

(Mallotus villosus) in the Barents Sea. Sarsia 83: 453-496.

18. Stergiou, KI. 1989. Capelin Mallotus villosus (Pisces: Osmeridae), glaciations,

and speciation: a nomothetic approach to fisheries ecology and reproductive

biology. Marine Ecology Progress Series 56: 211-224.

19. Bustos, CA, F Balbontin, and MF Landaeta. 2007. Spawning of the southern

hake Merluccius australis (Pisces: Merlucciidae) in Chilean fjords. Fisheries

Research 83: 23-32.

20. Pepin, P and RA Myers. 1991. Significance of egg and larval size to

recruitment variability of temperate marine fish. Canadian Journal of

Fisheries and Aquatic Sciences 48: 1820-1828.

21. Osborne, RF, YC Melo, MD Hofmeyer, DW Japp. 1999. Serial spawning and

batch fecundity of Merluccius capensis and M. paradoxus. South African

Journal of Marine Science 21: 211-216.

122

22. Brown, DR, E Leonarduzzi, and LM Andiarena. 2004. Age, growth and

mortality of hake larvae (Merluccius hubbsi) in the north Patagonian shelf.

Scientia Marina 68(2): 273-283.

23. Laptikhovsky, V and P Brickle. 2009. Aspects of embryonic development in

two southwest Atlantic gadiform fish: tadpole codling, Salilota australis

(Mordiae), and southern blue whiting, Micromesistius australis (Gadidae).

Acta Icthyologica et Piscatoria 39(2): 127-130.

24. Roberts, DE, Jr. and RA Schlieder. 1983. Induced sex inversion, maturation,

spawning and embryogeny of the protogynous grouper, Mycteroperca

microlepis. Journal of the World Mariculture Society 14: 639-649.

25. Riley, CM, GJ Holt, and CR Arnold. 1995. Growth and morphology of larval

and juvenile captive bred yellowtail snapper, Ocyurus chrysurus. Fishery

Bulletin 93:179-185.

26. Mihelakakis, A, T Yoshimatsu, and C Tsolkas. 2001. Spawning in captivity

and early life history of cultured red porgy, Pagrus pagrus. Aquaculture 199:

333-352.

27. McDermott, SF and SA Lowe. 1997. The reproductive cycle and sexual

maturity of Atka mackerel, Pleurogrammus monopterygius, in Alaska Waters.

Fishery Bulletin 95: 321-333.

28. Zhang, CI, TK Wilderbuer, and GE Walters. 1998. Biological characteristics

and fishery assessment of Alaska plaice, Pleuronectes quadrituberculatus, in

the Eastern Bering Sea. Marine Fisheries Review 60(4): 16-27.

29. Matarese, AC, AW Kendall, Jr., DM Blood, and BM Vinter. 1989. Laboratory

guide to early life history stages of Northeast Pacific fishes. U.S. Department

of Commerice, NOAA Technical Report NMFS 80. 652 p.

30. Grimes, CB and GR Huntsman. 1980. Reproductive biology of the vermillion

snapper, Rhomboplites aurorubens, from North Carolina and South Carolina.

Fishery Bulletin 78(1): 137-146.

31. Finucane, JH, LA Collins, HA Brusher, and CH Saloman. 1986. Reproductive

biology of king mackerel, Scomberomorus cavalla, from the southeastern

United States. Fishery Bulletin 84(4): 841-50.

123

32. Elgar, MA. 1990. Evolutionary compromise between a few large and many

small eggs: comparative evidence in teleost fish. Oikos 59(2): 283-287.

33. Pauly, D and RSV Pullin. 1988. Hatching time in spherical, pelagic, marine

fish eggs in response to temperature and egg size. Environmental Biology of

Fishes 22(4): 261-271.

34. Boehlert, GW and MM Yoklavich. 1984. Reproduction, embyronic energetics,

and the maternal-fetal relationship in the viviparous genus Sebastes (Pisces:

Scorpaenidae). Biol Bull 167: 354-370.

35. Chilton, E. 2007. Maturity of female northern rockfish Sebastes polyspinis in

the central Gulf of Alaska. Alaska Fishery Research Bulletin 12(2): 264-269.

36. Mylonas, CC, N Papandroulakis, A Smboukis, M Papadaki, and P Divanach.

2004. Induction of spawning of cultured greater amberjack (Seriola dumerili)

using GnRHa implants. Aquaculture 237: 141-154.

37. Knuckey, IA and KP Sivakumaran. 2001. Reproductive characteristics and

per-recruit analyses of blue warehou (Seriolella brama): implications for the

South East Fishery of Australia. Marine & Freshwater Research 52: 575-587.

38. Zimmermann, M. 1997. Maturity and fecundity of arrowtooth flounder,

Atheresthes stomias, from the Gulf of Alaska. Fishery Bulletin 95: 598-611.

39. Waltz, W, WA Roumillat, PK Ashe. 1979. Distribution, age structure, and sex

composition of the black sea bass, Centropristis striata, sampled along the

southeastern coast of the United States. South Carolina Marine Resources

Center Technical Report 32, pp. 18.

40. Pajaro, M, GJ Macchi, and RP Sanchez. 1997. Fertility and reproductive

frequency of Patagonia and Buenos Aires stocks of the Argentine anchovy

(Engraulis anchoita). Revista de Investigacion y Desarollo Pesquero 11: 19-

39.

41. Burnett, J, MR Ross, SH Clark. 1992. Several biological aspects of the witch

flounder (Glyptocephalus cynoglossus (L.)) in the Gulf of Maine-Georges

Bank region. J Northw Atl Fish Sci 12: 15-25

124

42. Crow, KD, DA Powers, and G Bernardi. 1997. Evidence for multiple maternal

contributors in nests of kelp greenling (Hexagrammos decagrammus,

Hexagrammidae). Copeia 1997(1): 9-15.

43. Batts, BS. 1972. Sexual maturity, fecundity, and sex ratios of the skipjack tuna,

Katsuwonus pelamis (Linnaeus), in North Carolina waters. Transactions of the

American Fisheries Society 101: 626-637.

44. Macchi, GJ and M Pajaro. 1999. Features of the reproductive biology of the

southern blue whiting (Micromesistius australis). INIDEP Documento

Cientifico 5: 67-79.

45. Macchi, GJ, M Pajaro, OC Wohler, MJ Acevedo, RL Centurion, DG Urteaga.

2005. Batch fecundity and spawning frequency of southern blue whiting

(Micromesistius australis) in the southwest Atlantic Ocean. New Zealand

Journal of Marine and Freshwater Research 39: 993-1000.

46. Pajaro, M and GJ Macchi. 2001. Spawning pattern, length at maturity, and

fecundity of the southern blue whiting (Micromesistius australis) in the south-

west Atlantic Ocean. New Zealand Journal of Marine and Freshwater

Research 35: 375-385.

47. Manooch, CS III. 1976. Reproductive cycle, fecundity, and sex ratios of the

red porgy, Pagrus pagrus (Pisces: Sparidae) in North Carolina. Fishery

Bulletin 74: 775-781.

48. McDermott, SF, KP Maslenikov, DR Gunderson. 2007. Annual fecundity,

batch fecundity, and oocyte atresia of Atka mackerel (Pleurogrammus

monopterygius) in Alaskan waters. Fishery Bulletin 105: 19-29.

49. Cargnelli, LM, SJ Griesbach, DB Packer, PL Berrien, DL Johnson, WW

Morse. 1999. Pollock, Pollachius virens, life history and habitat

characteristics. U.S. Dept. of Commerce, NOAA Technical Memorandum,

NMFS-NE-131: 1-30.

50. Buckley, LJ, AS Smigielski, TA Halavik, EM Caldarone, BR Burns, GC

Laurence. 1991. Winter flounder Pseudopleuronectes americanus reproductive

success. II. Effects of spawning time and female size on size, composition and

viability of eggs and larvae. Marine Ecology Progress Series 74: 125-135.

125

51. Haldorson, L and M Love. 1991. Maturity and fecundity in the rockfishes,

Sebastes spp., a review. Marine Fisheries Review 53(2): 25-31.

52. Boehlert, GW, WH Barss, and PB Lamberson. 1982. Fecundity of the widow

rockfish, Sebastes entomelas, off the coast of Oregon. Fishery Bulletin 80:881-

884.

53. Phillips, JB. 1964. Life history studies on ten species of rockfish (Genus

Sebastodes). California Department of Fish & Game Fish Bulletin 126:1-70.

54. White, GG, TA Munroe, and HM Austin. 2003. Reproductive seasonality,

fecundity, and spawning frequency of tautog (Tautoga onitis) in the lower

Chesapeake Bay and coastal waters of Virginia. Fishery Bulletin 101(2): 424-

442.

55. Sun, C-L, S-L Chu, S-Z Yeh. 2006. Reproductive biology of bigeye tuna in the

western and central Pacific Ocean. Western and Central Pacific Fisheries

Commission WCPFC-SC2-2006/BI WP-1. pp. 22

56. Beacham, TD and SJ Nepszy. 1980. Some aspects of the biology of white

hake, Urophycis tenuis, in the southern Gulf of St. Lawrence. Journal of

Northwest Atlantic Fishery Science 1: 49-54.

126

127

Chapter 4

Genetic impacts of overfishing are widespread

4.1 Abstract and Introduction

The harvesting of wild populations has a range of important impacts, including

changes in life history traits, depression of abundance, and alterations of community

composition (Dayton et al. 1998; Jennings & Kaiser 1998; Hutchings & Baum 2005;

Myers et al. 2007). Evolutionary impacts of harvest, however, have been less

frequently considered as part of routine management (Jørgensen et al. 2007). While

harvest can cause loss of genetic variation and the evolution of life history traits

(Allendorf et al. 2008), it is controversial how widespread such impacts are in the

wild. The greatest uncertainty surrounds species with high abundance, including most

marine fishes, because genetic drift during harvest is not expect to be strong enough to

reduce genetic diversity. However, a small number of case studies have reported such

reductions (Hauser et al. 2002). If these impacts are indeed rare, genetic diversity

across a range of overfished species should be similar to that in more stable congeners

and confamilials. In contrast, we found that allelic richness in overfished species was

on average 18% lower (p = 0.008) across 37 comparisons involving 88 species.

Heterozygosity was not significantly lower (-1.2%, p = 0.66), and we confirmed with

simulations that these results are consistent with a recent population bottleneck.

Genetic impacts of overharvest appear to be widespread even among abundant

species. A loss of allelic richness has implications for the long-term evolutionary

potential of species, and more generally, these evolutionary impacts deserve greater

attention as a routine aspect of natural resource management.

128

4.2 Results and Discussion

Using a meta-analytical framework for this study, we compiled data on 11,259

microsatellite loci from 185 studies of 88 species (Table 4.S1). These data allowed us

to ask whether overfished species have lower diversity than their close relatives (Fig.

4.1). Such an approach allows a wider range of species to be studied than those few

with historical samples (Hauser et al. 2002). If being overfished has no genetic impact,

then diversity should be equivalent in the two groups. Conversely, if diversity is

consistently lower in overfished species, this would suggest that population declines in

marine fish have evolutionary impacts. We used other species known to be fished as

controls so as to avoid known life history differences, including differences in pre-

fishing abundance, between fished and unfished species (Sethi et al. 2010; Pinsky et

al. 2011).

In 26 of our 37 comparisons, there were fewer alleles per locus (A50) among

overfished species (Fig. 4.2), including for species that ranged from small menhaden

(Brevoortia sp.) and herring (Clupea sp.) to large bluefin tuna (Thunnus thynnus) and

long-lived rockfish (Sebastes crameri and levis). On average, overfished species had

18% fewer alleles at microsatellite loci than closely related control species, and this

was significantly different from zero in both a random effects model (p = 0.008) and a

nonparametric Wilcoxon signed rank text (p = 0.015). As would be expected from a

recent bottleneck, expected heterozygosity (He) was slightly but not significantly

negative (Fig. 4.3, -1.2% lower, p = 0.66).

Across taxon groups, the response to fishing appears to differ (QT = 3053, p <

0.0001 and QT = 1503, p < 0.0001 for A50 and He respectively). Population size,

magnitude of decline, average duration overfished, and time since the species became

overfished did not explain significant amounts of this heterogeneity (p > 0.1, mixed

effects models).

The differences in allelic richness could suggest a loss of diversity in

overfished species, or they could represent consistent differences between overfished

and control species that are unrelated to overfishing. Because our evidence for loss is

129

indirect (a cross-taxa rather than cross-time comparison), it is important to consider

alternative explanations. In microsatellites, amplifications of loci developed in non-

focal species (cross-species loci), selection of high diversity loci, differences in

sampling effort, or differences in repeat size are all possible biases (Rico et al. 1996;

Ellegren 2004; Väli et al. 2008). Cross-species loci, however, were more prevalent

among our control species (42% of loci) than among our overfished species (27%, p <

0.0001, chi-squared test). Similarly, loci from primer notes – where low-diversity loci

are more likely to be reported – were more common among our control species (28%

vs. 8% of loci, p < 0.0001). Both trends reduce diversity among the control species

and make our comparison more conservative. In addition, sampling effort was higher

among the overfished species (n = 122 vs. n = 55, p < 0.0001, t-test), suggesting a

greater ability to detect alleles and a more conservative test. We did not detect a trend

in our data towards lower allelic richness among loci with longer repeat size (p = 0.68,

linear regression), though repeat sizes were on average somewhat longer among our

overfished species (2.7 bp) than our control species (2.3 bp, p < 0.001, t-test).

In addition, consistent differences in natural (pre-fishing) population

abundance between overfished and control species could create the differences in

allelic richness that we observed. However, abundant species are typically fished first

(Sethi et al. 2010), and species with longer histories of fishing are more likely to

collapse (Pinsky et al. 2011). It therefore seems likely that our overfished species had

a higher pre-fishing abundance than the control species. In addition, overfished species

do not exhibit unique life histories when compared against other fished species

(Pinsky et al. 2011). Finally, average landings and body size have both been suggested

as proxies for abundance in fishes (McCusker & Bentzen 2010). Our overfished

species had higher but not significantly different landings (9000 vs. 2820 metric tons

per year, p = 0.09) and lower but not significantly different body size (73 vs. 85 cm, p

= 0.26). Both trends suggest, if anything, a higher pre-fishing abundance among

overfished species (McCusker & Bentzen 2010).

Finally, our data from overfished species likely include some populations that

are not overfished. Similarly, our control species likely include some that are in reality

130

overfished, but that haven’t yet been identified as such. Our comparison is therefore

even more conservative and the true difference may be greater than we measured.

While our data are correlational rather than direct evidence, we believe the most

parsimonious explanation is that overfishing led to a reduction in diversity.

We next simulated and analytically calculated population bottlenecks for a

range of initial population sizes to determine whether a 19% loss of allelic richness is

reasonable for marine fishes (Fig. 4.4). Our analysis suggested that these results in fact

should be common if populations with an initial effective size less than 5,000 declined

90% more than five generations ago (Fig. 4.4b,d). A 90% decline from pre-fishing

biomass is reasonable to expect for heavily overfished marine species (Myers &

Worm 2003; Worm et al. 2009). Weaker bottlenecks (50% decline) could also

produce these patterns in populations with smaller effective sizes and bottlenecks

further in the past (Fig. 4.4a,c).

In summary, across a significant majority of overfished species, we found that

genetic diversity was lower than in closely related fishes that had not been overfished.

These results suggest that the genetic impacts of overfishing are widespread. While

previous studies have indicated that overfishing can reduce genetic diversity under

some circumstances (Hauser et al. 2002), this is the first study to suggest that such

impacts are common.

A reduction in genetic diversity among overfished species could be caused by

selection or by genetic drift. Fishing can exert strong selective pressures on target

species (Jørgensen et al. 2007), potentially reducing effective population size (Ne) by

increasing the variance in reproductive success among individuals (Hare et al. 2010).

However, recombination should quickly narrow this effect to the region of the genome

near selected loci, leaving only “outlier loci” affected by genetic draft. Obvious outlier

loci in our overfished species were not apparent, but the available data did not allow

formal statistical tests for outliers and we cannot entirely discount this process.

In contrast, a reduction in effective population size can drive genetic drift to

erode diversity throughout the genome. Census and effective population sizes are

often correlated by a scaling factor (Ne/N) that depends in part on life history

131

characteristics (Frankham 1995; Turner et al. 2006; Portnoy et al. 2008), and a decline

in census size can therefore increase the strength of genetic drift. The Ne/N ratio may

be quite small for marine species (Hauser et al. 2002). Drift is expected to act more

quickly and severely on allelic richness than on heterozygosity (Maruyama & Fuerst

1985), and, as shown by our simulations, the loss of diversity we observed was

consistent with effective population size decreases of 50-90% from initial effective

sizes of 102-104. These initial sizes appear reasonable for a wide range of marine

fishes, though the upper end is large enough so as to be difficult to measure precisely

with existing population genetic methods (Hare et al. 2010). Therefore, we propose

that widespread genetic impacts should be expected when harvest causes 90% declines

in abundance.

The microsatellite loci we examined are putatively neutral, but our results

suggest that genetic drift in marine fishes is also strong enough to reduce diversity at

neutral or weakly selected loci that would otherwise be adaptive in future

environments. The ability of a species to evolve and survive in future conditions

depends on the number of alleles and the traits conferred by them, particularly in

rapidly changing environments (Allendorf et al. 2008). Once lost, creation of new

alleles ultimately depends on the slow process of mutation, and recovery of lost allelic

richness will therefore be slow.

The demographic and ecological impacts of harvest have long been considered

and documented, while the evolutionary impacts, particularly in abundant species,

have been less widely appreciated. Our results suggest that a wide range of species

have lost alleles as a result of overfishing. Milder loss is also likely to accompany

even typical levels of fishing, which frequently reduce stocks by half or more. It is

difficult to put a precise cost on such loss, but it is clear that the ability to adapt in

coming decades will be crucial as species are faced with rapidly changing climatic

conditions.

132

4.3 Materials and Methods

4.3.1 Literature selection

For our research, we collected population genetic studies on overfished focal

fishes and control species closely related to these focal species. We focused on

microsatellites because of their widespread use, large number of loci available per

species, presumed neutrality, and high variability. We used the RAM Legacy stock

assessment database (Worm et al. 2009) to identify 71 species of fish that have had at

least one overfished population (biomass less than 50% of that predicted to sustain

maximum sustainable yield, BMSY) (Pinsky et al. 2011). We then conducted searches

in ISI Web of Science and Google Scholar with species name and “microsatellite” as

keywords to identify relevant studies. We also searched species synonyms as listed in

Fishbase (Froese & Pauly 2010).

For comparison against these overfished species, we compiled studies from

congeners not known to be overfished, or from confamilials when more than one

congener was not available. We restricted our control groups to fished species that

appeared in global landings reported to the UN Food and Agricultural Organization

(FAO) (downloaded from http://www.fao.org/fishery/statistics/software/fishstat/en,

December 2009). We limited our set of controls because fished species, including

those that are overfished, generally have a higher abundance than unfished species

(Sethi et al. 2010). Overfished species and those that are fished but not overfished,

however, have quite similar life histories (Pinsky et al. 2011). In addition, we only

compared marine species to other marine species, and only anadromous to

anadromous (DeWoody & Avise 2000). We also excluded endangered or critically

endangered species from the control species, as defined by the IUCN Redlist (Baillie

et al. 2004). Studies were collected until May 1, 2011.

From each study, we recorded allelic richness (A), expected heterozygosity

(He), sample size for each locus, whether the locus was originally developed in a

different species (cross-species amplification), and whether the study was a primer

133

note. Because A is dependent on sample size, we used Ewen’s sampling formula to

correct all measures to a standard sample size of 50 (A50), the median in our dataset

(Ewens 1972). To best estimate neutral diversity for all of our species, we only

included data from wild populations sampled after the species became overfished and

with a sample size > 10. We excluded loci linked to Expressed Sequence Tags (ESTs),

as these are more likely to be under selection. Because most studies focus on high

diversity microsatellites, we also excluded monomorphic loci and those expressly

chosen for low diversity (e.g., some used for forensic analysis).

4.3.2 Analysis

The effect of population decline was measured as the natural log of the

response ratio (Hedges et al. 1999),

where XE and XC are the mean diversity in the overfished species and in the

control species. Ln-transformed response ratios are commonly used in metaanalysis

for their ease of interpretation and approximate normality. However, we discuss our

results in terms of un-transformed response ratios for ease of interpretation.

To calculate mean diversity within a species or group of species, we weighted

each locus (i) by its sample size (wi),

where Xi is diversity at a locus in a population from a species. By counting

each locus separately, this approach has the effect of weighting more heavily those

populations, studies, and species with a higher sample size and more loci. Because a

134

small number of studies had very high sample size, we used the minimum of 200 or a

study’s sample size as wi. Not using this cap did not alter our conclusions.

In meta-analysis, individual effect sizes are typically weighted by the inverse

of the variance to account for each effect size’s precision (Hedges et al. 1999). We

calculated variance of L for each comparison as

where S and n are the standard deviation and the sample size for either the

overfished (E) or the control (C) species. The total number of loci, summed across

populations and across studies, was used as sample size. This approach treats each

locus within a population within a study as an independent replicate of the diversity

within a species.

We fit a random-effects model to calculate a mean response across all taxa

(Hedges et al. 1999). This model allows for variation between taxa to account for the

wide range of species examined. We used the test statistic QT to determine if there was

significant heterogeneity between groups in the effect size (Hedges et al. 1999). To

examine whether we could explain this variation, we fit mixed effects models with a

range of predictor variables. These variables included average annual landings from

1950-2006 as a proxy for population size (McCusker & Bentzen 2010), magnitude of

decline as the minimum fraction of BMSY reached by a species, average duration in

years for which a species was overfished, and the average number of years after the

species became overfished that the samples were collected. These additional data were

extracted from the stock assessments and the FAO landings. We conducted all

analyses in R with the metafor package (Viechtbauer 2010).

4.3.3 Simulations

To examine expected loss of allelic richness following a bottleneck, we

implemented a forward Wright-Fisher genetic model by sampling 2Nb alleles each

135

generation, where Nb is the effective size after the bottleneck. We calculated allelic

richness from a random sample of 100 alleles (50 individuals) in each generation. The

model was initialized with equilibrium allele frequency distributions from a locus with

a stepwise mutation model (SMM) in a population with 2Ne alleles (pre-bottleneck)

provided by the coalescent simulator Simcoal2 (Laval & Excoffier 2004). We

conducted 1000 independent simulations at each of Ne = 10x, where x = [2, 2.5, 3, 3.5,

4] and where Nb/Ne = 50 or 90%. We sampled populations 1-30 generations after the

bottleneck.

Because heterozygosity declines predictably as

where Ho is pre-bottleneck heterozygosity and Ht is heterozygosity in

generation t (Frankham et al. 2002), we evaluated this equation to generate

expectations for heterozygosity after a bottleneck. Neither this equation nor our allelic

richness simulations consider mutation, which is reasonable given the short time frame

that we examined (30 generations).

4.4 Acknowledgments

The authors thank Dan Ricard and Olaf Jensen for help with the RAM Legacy

database. M.L.P. was supported by a NSF Graduate Research Fellowship.

4.5 References

Allendorf F.W., England P.R., Luikart G., Ritchie P.A. & Ryman N. (2008). Genetic

effects of harvest on wild animal populations. Trends Ecol. Evol., 23, 327-337.

136

Baillie J.E.M., Hilton-Taylor C. & Stuart S. (eds.) (2004). 2004 IUCN Red List of

threatened species: a global species assessment. IUCN, Gland, Switzerland

and Cambridge, UK.

Dayton P.K., Tegner M.J., Edwards P.B. & Riser K.L. (1998). Sliding baselines,

ghosts, and reduced expectations in kelp forest communities. Ecol. Appl., 8,

309-322.

DeWoody J.A. & Avise J.C. (2000). Microsatellite variation in marine, freshwater,

and anadromous fishes compared with other animals. J. Fish Biol., 56, 461-

473.

Ellegren H. (2004). Microsatellites: simple sequences with complex evolution. Nature

Reviews Genetics, 5, 435-45. doi:10.1038/nrg1348

Ewens W.J. (1972). The sampling theory of selectively neutral alleles. Theor. Popul.

Biol., 3, 87-112.

Frankham R. (1995). Effective population size/adult population size ratios in wildlife:

a review. Genet. Res., 66, 95-107.

Frankham R., Ballou J.D. & Briscoe D.A. (2002). Introduction to conservation

genetics. Cambridge University Press, New York.

Froese R. & Pauly D. (eds.) (2010). FishBase. World Wide Web electronic

publication. www.fishbase.org. Visited 4/01/2011.

Hare J.A., Alexander M.A., Fogarty M.J., Williams E.H. & Scott J.D. (2010).

Forecasting the dynamics of a coastal fishery species using a coupled climate-

population model. Ecol. Appl., 20, 452-64.

Hauser L., Adcock G.J., Smith P.J., Bernal Ramírez J.H. & Carvalho G.R. (2002).

Loss of microsatellite diversity and low effective population size in an

overexploited population of New Zealand snapper (Pagrus auratus). P Natl

Acad Sci USA, 99, 11742-11747.

Hedges L.V., Gurevitch J. & Curtis P.S. (1999). The Meta-Analysis of Response

Ratios in Experimental Ecology. Ecology, 80, 1150.

Hutchings J.A. & Baum J.K. (2005). Measuring marine fish biodiversity: temporal

changes in abundance, life history and demography. Philosophical

137

Transactions of the Royal Society of London B: Biological Sciences, 360, 315-

338. doi:10.1098/rstb.2004.1586

Jennings S. & Kaiser M.J. (1998). The effects of fishing on marine ecosystems and

communities. Adv. Mar. Biol., 34, 201-352. doi:10.1006/jfbi.2000.1373

Jørgensen C., Enberg K., Dunlop E.S., Arlinghaus R., Boukal D.S., Brander K.,

Ernande B., Gårdmark A., Johnston F., Matsumura S., Pardoe H., Raab K.,

Silva A., Vainikka A., Dieckmann U., Heino M. & Rijnsdorp A.D. (2007).

Managing evolving fish stocks. Science, 318, 1247-1248.

Laval G. & Excoffier L. (2004). SIMCOAL 2.0: a program to simulate genomic

diversity over large recombining regions in a subdivided population with a

complex history. Bioinformatics, 20, 2485-2487.

Maruyama T. & Fuerst P.A. (1985). Population bottlenecks and nonequilibrium

models in population genetics. II. Number of alleles in a small population that

was formed by a recent bottleneck. Genetics, 111, 675-689.

McCusker M.R. & Bentzen P. (2010). Positive relationships between genetic diversity

and abundance in fishes. Mol. Ecol., 19, 4852-4862.

Myers R.A., Baum J.K., Shepherd T.D., Powers S.P. & Peterson C.H. (2007).

Cascading effects of the loss of apex predatory sharks from a coastal ocean.

Science, 315, 1846-1850.

Myers R.A. & Worm B. (2003). Rapid worldwide depletion of predatory fish

communities. Nature, 423, 280-283.

Pinsky M., Jensen O.P., Ricard D. & Palumbi S.R. (2011). Unexpected patterns of

fisheries collapse in the world’s oceans. Proceedings of the National

Academies of Science. doi:10.1073/pnas.1015313108

Portnoy D.S., McDowell J.R., McCandless C.T., Musick J.a. & Graves J.E. (2008).

Effective size closely approximates the census size in the heavily exploited

western Atlantic population of the sandbar shark, Carcharhinus plumbeus.

Conserv. Genet., 10, 1697-1705. doi:10.1007/s10592-008-9771-2

138

Rico C., Rico I. & Hewitt G.M. (1996). 470 million years of conservation of

microsatellite loci among fish species. Proceedings. Biological sciences / The

Royal Society, 263, 549-57.

Sethi S.A., Branch T.A. & Watson R. (2010). Global fishery development patterns are

driven by profit but not trophic level. P Natl Acad Sci USA, 107, 12163-12167.

Turner T.F., Osborne M.J., Moyer G.R., Benavides M.A. & Alò D. (2006). Life

history and environmental variation interact to determine effective population

to census size ratio. Proceedings of the Royal Society B: Biological Sciences,

273, 3065-73.

Väli U., Einarsson A., Waits L. & Ellegren H. (2008). To what extent do microsatellite

markers reflect genome-wide genetic diversity in natural populations? Mol.

Ecol., 17, 3808-17. doi:10.1111/j.1365-294X.2008.03876.x

Viechtbauer W. (2010). Conducting meta-analyses in R with the metafor package.

Journal of Statistcal Software, 36, 1-48.

Worm B., Hilborn R., Baum J.K., Branch T.A., Collie J.S., Costello C., Fogarty M.J.,

Fulton E.A., Hutchings J.A., Jennings S., Jensen O.P., Lotze H.K., Mace P.M.,

McClanahan T.R., Minto C., Palumbi S.R., Parma A.M., Ricard D., Rosenberg

A.A., Watson R. & Zeller D. (2009). Rebuilding global fisheries. Science, 325,

578.

139

4.6 Figures

Figure 4.1. Paired comparisons between an overfished species and its close

relatives. For example, average allelic richness (A50) and expected heterozygosity (He)

are lower in the overfished Pagrus auratus than in other members of the Sparidae

family.

A50 = 10.2He = 0.736

A50 = 16.6He = 0.846

140

Figure 4.2. Overfished species have consistently lower allelic richness (A50)

than closely related control species. Each box represents the ratio of overfished/control

diversity with 95% confidence intervals. The average ratio (82%) with 95% CI is

indicated by the diamond.

RE Model

0.22 0.37 0.61 1.00 1.65 2.72 4.48

Aoverfished Acontrol

Lutjanus analisScomber japonicusMallotus villosusSebastes pinnigerSebastes mentellaTrachurus murphyiSebastes mystinusSebastes fasciatusMelanogrammus aeglefinusMycteroperca microlepisPleuronectes platessaSebastes paucispinisCynoscion regalisSebastes ruberrimusPseudopleuronectes americanusSebastes alutusGadus macrocephalusGadus morhuaLutjanus campechanusMerlangius merlangusTheragra chalcogrammaThunnus thynnusPagrus pagrusSardinops sagaxMerluccius hubbsiMicromesistius poutassouEngraulis encrasicolusLepidorhombus whiffiagonisPagrus auratusClupea pallasiiSprattus sprattusSebastes crameriBrevoortia patronusClupea harengusBrevoortia tyrannusSebastes levisPlatichthys stellatus

141

Figure 4.3. Overfished species have similar heterozygosity (He) as compared

to closely related control species. Each box represents the ratio of overfished/control

diversity with 95% confidence intervals. The average ratio (99%) with 95% CI is

indicated by the diamond.

RE Model

0.37 0.61 1.00 1.65

Hoverfished Hcontrol

Lutjanus analisScomber japonicusSebastes mentellaSebastes alutusSebastes mystinusSebastes fasciatusSebastes pinnigerMycteroperca microlepisSebastes ruberrimusGadus morhuaGadus macrocephalusMicromesistius poutassouTheragra chalcogrammaMelanogrammus aeglefinusSebastes paucispinisSebastes crameriMerlangius merlangusMallotus villosusTrachurus murphyiThunnus thynnusSardinops sagaxEngraulis encrasicolusPleuronectes platessaMerluccius hubbsiClupea pallasiiLutjanus campechanusPagrus auratusSprattus sprattusPlatichthys stellatusClupea harengusPagrus pagrusBrevoortia tyrannusCynoscion regalisPseudopleuronectes americanusBrevoortia patronusLepidorhombus whiffiagonis

142

Figure 4.4. Expected loss of allelic richness (a,b) and heterozygosity (c,d)

from simulated bottlenecks of 50% (a,c) and 90% (b,d). Contour lines indicate

proportion of diversity remaining. Initial effective population size (Ne) is indicated

along the y-axis, and generations post-bottleneck are indicated along the x-axis.

Contours bend back toward the right for low Ne because few alleles are present pre-

bottleneck, and therefore only a small proportion can be lost.

Ne

0.85

0.9

0.95

1

1

102

103

104

a)

0.6

5

0.7

0.75

0.8

0.85

0.9

0

.95

1

b)

Generations

Ne

0.8

0.9

1

0 5 10 20 30

102

103

104

c)

Generations

Ne

0.3 0.4 0.5 0.6

0.7 0.8

0.9

1

0 5 10 20 30

d)

143

4.7 Supplemental Information

Table 4.S1. References for the overfished species and the control species

against which they were compared.

Overfished species

Common name Control taxon Control species References

Brevoortia patronus

Gulf menhaden

Clupeidae Sardina pilchardus 1-6

Brevoortia tyrannus

Atlantic menhaden

Clupeidae Sardina pilchardus 1, 3-6

Clupea harengus Atlantic herring

Clupeidae Sardina pilchardus 4, 5, 7-11

Clupea pallasii Pacific herring

Clupeidae Sardina pilchardus 4, 5, 12-22

Cynoscion regalis

Gray weakfish

Sciaenidae

Argyrosomus regius, Cynoscion nebulosus, Larimichthys polyactis, Micropogonias furnieri, Miichthys miiuy, Sciaenops ocellatus

23-39

Engraulis encrasicolus

European anchovy

Engraulis Engraulis japonicus 40-43

Gadus macrocephalus

Pacific cod

Gadidae Trisopterus minutus 44-47

Gadus morhua Atlantic cod

Gadidae Trisopterus minutus 47-60

Lepidorhombus whiffiagonis Megrim Scophthalmidae

Lepidorhombus boscii, Psetta maxima, Scophthalmus rhombus

61-69

Lutjanus analis Mutton snapper

Lutjanus Lutjanus argentimaculatus, L. synagris

70-75

Lutjanus campechanus

Northern red snapper

Lutjanus Lutjanus argentimaculatus, L. synagris

70-72, 75-78

Mallotus villosus Capelin Osmeridae Osmerus mordax, Thaleichthys pacificus

79-85

Melanogrammus aeglefinus Haddock Gadidae Trisopterus minutus

47, 86-88

144

Merlangius merlangus Whiting Gadidae Trisopterus minutus

47, 89, 90

Merluccius hubbsi

Argentine hake

Merluccius

Merluccius albidus, M. australis, M. bilinearis, M. gayi, M. merluccius

91-95

Micromesistius poutassou

Blue whiting

Gadidae Trisopterus minutus 47, 96-98

Mycteroperca microlepis Gag Mycteroperca

Mycteroperca bonaci, M. phenax

99-103

Pagrus auratus Squirefish Sparidae

Acanthopagrus schlegelii, Dentex dentex, Lithognathus mormyrus, Pagellus erythrinus, Sparus aurata

67, 104-117

Pagrus pagrus Common seabream

Sparidae

Acanthopagrus schlegelii, Dentex dentex, Lithognathus mormyrus, Pagellus erythrinus, Sparus aurata

67, 104, 105, 108, 110-118

Platichthys stellatus

Starry flounder

Pleuronectidae Platichthys flesus, Pseudopleuronectes herzensteini

61, 119-122

Pleuronectes platessa

European plaice

Pleuronectidae Platichthys flesus, Pseudopleuronectes herzensteini

61, 119, 120, 122-129

Pseudopleuronectes americanus

Winter flounder

Pleuronectidae Platichthys flesus, Pseudopleuronectes herzensteini

61, 119, 120, 122, 130-132

Sardinops sagax South American pilchard

Clupeidae Sardina pilchardus 4, 5, 133

145

Scomber japonicus

Chub mackerel

Scombridae

Auxis rochei, Katsuwonus pelamis, Scomber australasicus, S. brasiliensis, Scomberomorus cavalla, S. commerson, S niphonius, T. alalunga, T. albacares, T. atlanticus, T. obesus, T. orientalis

134-159

Sebastes alutus Pacific ocean perch

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-168

Sebastes crameri Darkblotched rockfish

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 169

Sebastes fasciatus

Acadian redfish

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 170-174

Sebastes levis Cowcod Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167

Sebastes mentella

Deepwater redfish

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 170-177

Sebastes mystinus

Blue rockfish

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 178-180

146

Sebastes paucispinis Bocaccio Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 181

Sebastes pinniger

Canary rockfish

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 182, 183

Sebastes ruberrimus

Yelloweye rockfish

Sebastes

Sebastes aleutianus, S. carnatus, S. entomelas, S. flavidus, S. goodei, S. melanops, S. melanostomus

160-167, 184

Sprattus sprattus European sprat

Clupeidae Sardina pilchardus 4, 5, 185, 186

Theragra chalcogramma

Alaska pollock

Gadidae Trisopterus minutus 47, 187-190

Thunnus thynnus Northern bluefin tuna

Thunnus

Thunnus alalunga, T. albacares, T. atlanticus, T. obesus, T. orientalis

134, 135, 139, 141-144, 147, 149, 151-153, 191-193

Trachurus murphyi Inca scad Trachurus

Trachurus japonicus, T. trachurus

194-197

1. Anderson J.D. & Karel W.J. (2007). Genetic evidence for asymmetric

hybridization between menhadens (Brevoortia spp.) from peninsular Florida. J.

Fish Biol., 71, 235-249. doi:10.1111/j.1095-8649.2007.01597.x

2. Anderson J.D. & McDonald D.L. (2007). Morphological and genetic

investigations of two western Gulf of Mexico menhadens (Brevoortia spp.). J.

Fish Biol., 70, 139-147. doi:10.1111/j.1095-8649.2007.01326.x

3. Anderson J.D. (2007). Systematics of the North American menhadens:

molecular evolutionary reconstructions in the genus Brevoortia (Clupeiformes:

Clupeidae). Fish. Bull., 205, 368-378.

147

4. Gonzalez E.G. & Zardoya R. (2007). The relative role of life-history traits and

historical factors in shaping genetic population structure of sardines (Sardina

pilchardus). BMC Evol. Biol., 7, 1-12. doi:10.1186/1471-2148-7-197

5. Gonzalez E.G. & Zardoya R. (2007). Isolation and characterization of

polymorphic microsatellites for the Sardine Sardina pilchardus (Clupeiformes:

Clupeidae). Mol. Ecol. Notes, 7, 519- 521. doi:10.1111/j.1471-

8286.2006.01640.x

6. Lynch A.J., Mcdowell J.R. & Graves J.E. (2010). A molecular genetic

investigation of the population structure of Atlantic Menhaden (Brevoortia

tyrannus). Fish. Bull., 108, 87-97.

7. Bekkevold D., André C., Dahlgren T.G., Clausen L.A.W., Torstensen E.,

Mosegaard H., et al. (2005). Environmental correlates of population

differentiation in Atlantic Herring. Evolution, 59, 2656-2668.

8. Bekkevold D., Clausen L.A.W., Mariani S., André C., Christensen T.B. &

Mosegaard H. (2007). Divergent origins of sympatric herring population

components determined using genetic mixture analysis. Mar. Ecol. Prog. Ser.,

337, 187-196.

9. Jørgensen H.B.H., Hansen M.M. & Loeschcke V. (2005). Spring-spawning

Herring (Clupea harengus L.) in the southwestern Baltic Sea: do they form

genetically distinct spawning waves? ICES J. Mar. Sci., 62, 1065-1075.

doi:10.1016/j.icesjms.2005.04.007

10. Jørgensen H.B.H., Hansen M.M., Bekkevold D., Ruzzante D.E. & Loeschcke

V. (2005). Marine landscapes and population genetic structure of Herring

(Clupea harengus L.) in the Baltic Sea. Mol. Ecol., 14, 3219-3234.

doi:10.1111/j.1365-294X.2005.02658.x

11. Mariani S., Hutchinson W.F., Hatfield E.M.C., Ruzzante D.E., Simmonds E.J.,

Dahlgren T.G., et al. (2005). North Sea Herring population structure revealed

by microsatellite analysis. Mar. Ecol. Prog. Ser., 303, 245-257.

12. Beacham T.D., Schweigert J.F., MacConnachie C., Le K.D., Labaree K. &

Miller K.M. (2001). Population structure of Herring (Clupea pallasi) in British

148

Columbia: an analysis using microsatellite loci. Canadian Science Advisory

Secretariat Research Document 2001/128. Department of Fisheries and Oceans

Canada, Nanaimo, BC.

13. Beacham T.D., Schweigert J.F., MacConnachie C., Le K.D., Labaree K. &

Miller K.M. (2002). Population structure of Herring (Clupea pallasi) in British

Columbia determined by microsatellites, with comparisons to southeast Alaska

and California. Research Canadian Science Advisory Secretariat Document

2002/109. Department of Fisheries and Oceans Canada, Nanaimo, BC.

14. Beacham T.D., Schweigert J.F., MacConnachie C., Le K.D. & Flostrand L.

(2008). Use of microsatellites to determine population structure and migration

of Pacific Herring in British Columbia and adjacent regions. Trans. Am. Fish.

Soc., 137, 1795-1811. doi:10.1577/T08-033.1

15. Kitada S., Kitakado T. & Kishino H. (2007). Empirical Bayes inference of

pairwise FST and its distribution in the genome. Genetics, 177, 861-873.

doi:10.1534/genetics.107.077263

16. Sugaya T., Sato M., Yokoyama E., Nemoto Y., Fujita T., Okouchi H., et al.

(2008). Population genetic structure and variability of Pacific Herring Clupea

pallasii in the stocking area along the Pacific coast of northern Japan. Fish.

Sci., 74, 579-588. doi:10.1111/j.1444-2906.2008.01561.x

17. Miller K.M., Laberee K., Schulze A.D. & Kaukinen K.H. (2001).

Development of microsatellite loci in Pacific Herring (Clupea pallasi). Mol.

Ecol. Notes, 1, 131-132.

18. Mitchell D.M. (2006). Biocomplexity and metapopulation dynamics of Pacific

Herring (Clupea pallasii) in Puget Sound, Washington. Master of Science.

University of Washington, Seattle.

19. O'Connell M., Dillon M.C. & Wright J.M. (1998). Development of primers for

polymorphic microsatellite loci in the Pacific Herring (Clupea harengus

pallasi). Mol. Ecol., 7, 358-360.

149

20. O'Connell M., Dillon M.C., Wright J.M., Bentzen P., Merkouris S.E. & Seeb

J.E. (1998). Genetic structuring among Alaskan Pacific Herring populations

identified using microsatellite variation. J. Fish Biol., 53, 150-163.

21. Olsen J.B., Lewis C.J., Kretschmer E.J., Wilson S.L. & Seeb J.E. (2002).

Characterization of 14 tetranucleotide microsatellite loci derived from Pacific

Herring. Mol. Ecol., 2, 101-103.

22. Small M.P., Loxterman J.L., Frye A.E., Von Bargen J.F., Bowman C. &

Young S.F. (2005). Temporal and spatial genetic structure among some Pacific

Herring populations in Puget Sound and the southern Strait of Georgia. Trans.

Am. Fish. Soc., 134, 1329-1341. doi:10.1577/T05-050.1

23. Anderson J.D. & Karel W.J. (2010). Population genetics and dynamics of

Spotted Seatrout in the estuarine waters of Texas. Fisheries and Aquaculture

Journal, 2010, FAJ2.

24. Chapman R.W., Ball A.O. & Mash L.R. (2002). Spatial homogeneity and

temporal heterogeneity of Red Drum (Sciaenops ocellatus) microsatellites:

Effective population sizes and management implications. Mar. Biotechnol., 4,

589-603. doi:10.1007/s10126-002-0038-5

25. Chen S.-L., Xing S.-C., Xu G.-B., Liao X. & Yang J.-F. (2009). Isolation and

characterization of 10 polymorphic microsatellite loci from Small Yellow

Croaker (Pseudosciaena polyactis). Conserv. Genet., 10, 1469-1471.

doi:10.1007/s10592-008-9762-3

26. Cordes J.F. & Graves J.E. (2003). Investigation of congeneric hybridization in

and stock structure of Weakfish (Cynoscion regalis) inferred from analyses of

nuclear and mitochondrial DNA loci. Fish. Bull., 101, 443-450.

27. D'Anatro A., Pereira A.N. & Lessa E.P. (2011). Genetic structure of the White

Croaker, Micropogonias furnieri Desmarest 1823 (Perciformes: Sciaenidae)

along Uruguayan coasts: contrasting marine, estuarine, and lacustrine

populations. Environ. Biol. Fishes. doi:10.1007/s10641-011-9799-x

28. Gold J.R. & Turner T.F. (2002). Population structure of Red Drum (Sciaenops

ocellatus) in the northern Gulf of Mexico, as inferred from variation in

150

nuclear-encoded microsatellites. Mar. Biol., 140, 249-265.

doi:10.1007/s002270100692

29. Gold J.R., Stewart L.B. & Ward R. (2003). Population structure of Spotted

Seatrout (Cynoscion nebulosus) along the Texas Gulf coast, as revealed by

genetic analysis. In: Biology of the Spotted Seatrout (ed. Bortone SA). CRC

Press Boca Raton, FL, pp. 17-29.

30. Karlsson S., Renshaw M.A., Rexroad III C.E. & Gold J.R. (2008). PCR

primers for 100 microsatellites in Red Drum (Sciaenops ocellatus). Molecular

Ecology Resources, 8, 393-398. doi:10.1111/j.1471-8286.2007.01969.x

31. Karlsson S., Renshaw M.A., Rexroad III C.E. & Gold J.R. (2008).

Microsatellite primers for Red Drum (Sciaenops ocellatus). Fish. Bull., 106,

476-482.

32. Li J., Feng F. & Yue G.H. (2006). Twelve novel polymorphic microsatellites

in a marine fish species, Yellow Croaker Larimichthys polyactis. Mol. Ecol.

Notes, 6, 188- 190. doi:10.1111/j.1471-8286.2005.01188.x

33. Molecular Ecology Resources Primer Development Consortium, Andree K.,

Axtner J., Bagley M.J., Barlow E.J., Beebee T.J.C., et al. (2010). Permanent

genetic resources added to Molecular Ecology Resources Database 1 April

2010 – 31 May 2010. Molecular Ecology Resources, 10, 1098-1105.

34. Renshaw M.A., Gawriluk T.R. & Gold J.R. (2009). Characterization of Red

Drum microsatellite markers in Spotted Seatrout. N. Am. J. Aquacult., 71, 374-

379. doi:10.1577/A08-070.1

35. Wang R.X., Xu T.J., Sun Y.N. & He G.Y. (2010). Polymorphic microsatellite

loci from two enriched genomic libraries for the genetic analysis of the Miiuy

Croaker, Miichthys miiuy (Sciaenidae). Gen. Mol. Res., 9, 931-934.

doi:10.4238/vol9-2gmr806

36. Molecular Ecology Resources Primer Development Consortium, Anderson

C.M., Aparicio G.J., Atangana A.R., Beaulieu J., Bruford M.W., et al. (2010).

Permanent genetic resources added to Molecular Ecology Resources Database

151

1 December 2009–31 January 2010. Molecular Ecology Resources, 10, 576-

579.

37. Ward R., Bowers K., Hensley R., Mobley B. & Belousi E. (2007). Genetic

variability in Spotted Seatrout (Cynoscion nebulosus), determined with

microsatellite DNA markers. Fish. Bull., 105, 197-206.

38. Wiley B.A. & Chapman R.W. (2003). Population structure of Spotted Seatrout,

Cynoscion nebulosus, along the Atlantic coast of the U.S. In: Biology of the

Spotted Seatrout (ed. Bortone SA). CRC Press Boca Raton, FL.

39. Molecular Ecology Resources Primer Development Consortium, Almany G.R.,

Arruda M.P.D., Arthofer W., Atallah Z.K., Beissinger S.R., et al. (2009).

Permanent genetic resources added to Molecular Ecology Resources Database

1 May 2009–31 July 2009. Molecular Ecology Resources, 9, 1460-1466.

40. Landi M., Garoia F., Piccinetti C. & Tinti F. (2005). Isolation of polymorphic

microsatellite loci from the European Anchovy, Engraulis encrasicolus. Mol.

Ecol. Notes, 5, 266-268. doi:10.1111/j.1471-8286.2005.00892.x

41. Pakaki V., Magoulas A. & Kasapidis P. (2009). New polymorphic

microsatellite loci for population studies in the European Anchovy, Engraulis

encrasicolus (L.). Molecular Ecology Resources, 9, 1406-1409.

doi:10.1111/j.1755-0998.2009.02678.x

42. Yu H.-T., Lee Y.-J., Huang S.-W. & Chiu T.-S. (2002). Genetic analysis of the

populations of Japanese Anchovy (Engraulidae: Engraulis japonicus) using

microsatellite DNA. Mar. Biotechnol., 4, 471-479. doi:10.1007/s10126-002-

0035-8

43. Zarraonaindia I., Pardo M.A., Iriondo M., Manzano C. & Estonba A. (2009).

Microsatellite variability in European Anchovy (Engraulis encrasicolus) calls

for further investigation of its genetic structure and biogeography. ICES J.

Mar. Sci., 66, 2176-2182.

44. Canino M.F., Spies I.B. & Hauser L. (2005). Development and

characterization of novel di- and tetranucleotide microsatellite markers in

152

Pacific Cod (Gadus macrocephalus). Mol. Ecol. Notes, 5, 908-910.

doi:10.1111/j.1471-8286.2005.01109.x

45. Cunningham K.M. (2007). Population genetics of Pacific Cod (Gadus

macrocephalus): evidence for large and small scale population structure.

Master of Science. University of Washington, Seattle.

46. Kim M.-J., An H.S. & Choi K.-H. (2010). Genetic characteristics of Pacific

Cod populations in Korea based on microsatellite markers. Fish. Sci., 76, 595-

603. doi:10.1007/s12562-010-0249-z

47. Zhang D.-X., Smith J.A., Rico C. & Hewitt G.M. (2001). Isolation and

characterization of 10 microsatellite loci in Poor Cod Trisopterus minutus (L).

Mol. Ecol. Notes, 1, 50-52.

48. Beacham T.D., Brattey J., Miller K.M., Le K.D. & Withler R.E. (2002).

Multiple stock structure of Atlantic Cod (Gadus morhua) off Newfoundland

and Labrador determined from genetic variation. ICES J. Mar. Sci., 59, 650-

665. doi:10.1006/jmsc.2002.1253

49. Bentzen P., Taggart C.T., Ruzzante D.E. & Cook D. (1996). Microsatellite

polymorphism and the population structure of Atlantic Cod (Gadus morhua) in

the northwest Atlantic. Can. J. Fish. Aquat. Sci., 53, 2706-2721.

50. Bradbury I.R., Snelgrove P.V.R., Bentzen P., de Young B., Gregory R.S. &

Morris C.J. (2009). Structural and functional connectivity of marine fishes

within a semi-enclosed Newfoundland fjord. J. Fish Biol., 75, 1393-1409.

doi:10.1111/j.1095-8649.2009.02391.x

51. Brooker A.L., Cook D., Bentzen P., Wright J.M. & Doyle R.W. (1994).

Organization of microsatellites differs between mammals and cold-water

teleost fishes. Can. J. Fish. Aquat. Sci., 51, 1959-1966.

52. Hutchinson W.F., Carvalho G.R. & Rogers S.I. (2001). Marked genetic

structuring in localised spawning populations of Cod Gadus morhua in the

North Sea and adjoining waters, as revealed by microsatellites. Mar. Ecol.

Prog. Ser., 223, 251-260.

153

53. Hutchinson W.F., van Oosterhout C., Rogers S.I. & Carvalho G.R. (2003).

Temporal analysis of archived samples indicates marked genetic changes in

declining North Sea Cod (Gadus morhua). Proceedings of the Royal Society B:

Biological Sciences, 270, 2125-2132. doi:10.1098/rspb.2003.2493

54. Miller K.M., Le K.D. & Beacham T.D. (2000). Development of tri- and

tetranucleotide repeat microsatellite loci in Atlantic Cod. Molcular Ecology, 9,

238-239.

55. Ruzzante D.E., Taggart C.T., Cook D. & Goddard S.V. (1996). Genetic

differentiation between inshore and offshore Atlantic Cod (Gadus morhua) off

Newfoundland: microsatellite DNA variation and antifreeze level. Can. J.

Fish. Aquat. Sci., 53, 634-645.

56. Ruzzante D.E., Taggart C.T. & Cook D. (1996). Spatial and temporal variation

in the genetic composition of a larval Cod (Gadus morhua) aggregation: cohort

contribution and genetic stability. Can. J. Fish. Aquat. Sci., 53, 2695-2705.

57. Ruzzante D.E., Taggart C.T., Cook D. & Goddard S.V. (1997). Genetic

differentiation between inshore and offshore Atlantic Cod (Gadus morhua) off

Newfoundland: a test and evidence of temporal stability. Can. J. Fish. Aquat.

Sci., 54, 2700-2708.

58. Ruzzante D.E., Taggart C.T. & Cook D. (1998). A nuclear DNA basis for

shelf- and bank-scale population structure in northwest Atlantic Cod (Gadus

morhua): Labrador to Georges Bank. Mol. Ecol., 7, 1663-1680.

59. Ruzzante D.E., Taggart C.T., Lang S. & Cook D. (2000). Mixed-stock analysis

of Atlantic Cod near the Gulf of St. Lawrence based on microsatellite DNA.

Ecol. Appl., 10, 1090-1109.

60. Ruzzante D.E., Wroblewski J.S., Taggart C.T., Smedbol R.K., Cook D. &

Goddard S.V. (2000). Bay-scale population structure in coastal Atlantic Cod in

Labrador and Newfoundland, Canada. J. Fish Biol., 56, 431-447.

doi:10.1006/jfbi.1999.1168

61. Bouza C., Presa P., Castro J., Sánchez L. & Martínez P. (2002). Allozyme and

microsatellite diversity in natural and domestic populations of Turbot

154

(Scophthalmus maximus) in comparison with other Pleuronectiformes. Can. J.

Fish. Aquat. Sci., 59, 1460-1473. doi:10.1139/F02-114

62. Coughlan J.P., Imsland A.K., Galvin P.T., Fitzgerald R.D., Naevdal G. &

Cross T.F. (1998). Microsatellite DNA variation in wild populations and

farmed strains of Turbot from Ireland and Norway: a preliminary study. J. Fish

Biol., 52, 916-922.

63. Coughlan J.P., McCarthy E., McGregor D., O'Donoghue P., Galvin P.T.,

Fitzgerald R.D., et al. (1996). Four polymorphic microsatellites in Turbot

Scophthalmus maximus. Anim. Genet., 27, 433-442.

64. Danancher D. & Garcia-Vazquez E. (2009). Population differentiation in

Megrim (Lepidorhombus whiffiagonis) and Four Spotted Megrim

(Lepidorhombus boscii) across Atlantic and Mediterranean waters and

implications for wild stock management. Mar. Biol., 156, 1869-1880.

doi:10.1007/s00227-009-1219-9

65. Florin A.-B. & Höglund J. (2007). Absence of population structure of Turbot

(Psetta maxima) in the Baltic Sea. Mol. Ecol., 16, 115-126.

doi:10.1111/j.1365-294X.2006.03120.x

66. Iyengar A., Piyapattanakorn S., Heipel D.A., Stone D.M., Howell B.R., Child

A.R., et al. (2000). A suite of highly polymorphic microsatellite markers in

Turbot (Scophthalmus maximus L.) with potential for use across several

flatfish species. Mol. Ecol., 9, 365-378.

67. Liu Y.G., Liu L.X., Li Z.X., Gao A.Y., Lin H., Li B.F., et al. (2007).

Development of polymorphic microsatellite from RAPD bands of Black Sea

Bream Acanthopagrus schlegeli. Mol. Ecol. Notes, 7, 1178-1180.

doi:10.1111/j.1471-8286.2007.01823.x

68. Nielsen E.E., Nielsen P.H., Meldrup D. & Hansen M.M. (2004). Genetic

population structure of Turbot (Scophthalmus maximus L.) supports the

presence of multiple hybrid zones for marine fishes in the transition zone

between the Baltic Sea and the North Sea. Mol. Ecol., 13, 585 - 595.

doi:10.1046/j.1365-294X.2004.02097.x

155

69. Pardo G.B., Casas L., Fortes G.G., Bouza C., Martínez P., Clark M.S., et al.

(2005). New microsatellite markers in Turbot (Scophthalmus maximus) derived

from an enriched genomic library and sequence databases. Mol. Ecol. Notes, 5,

62-64. doi:10.1111/j.1471-8286.2004.00834.x

70. Karlsson S., Saillant E. & Gold J.R. (2009). Population structure and genetic

variation of Lane Snapper (Lutjanus synagris) in the northern Gulf of Mexico.

Mar. Biol., 156, 1841-1855. doi:10.1007/s00227-009-1217-y

71. Ovenden J.R. & Street R. (2003). Genetic population structure of Mangrove

Jack, Lutjanus argentimaculatus (Forsskål). Marine And Freshwater Research,

54, 127-137.

72. Renshaw M.A., Karlsson S. & Gold J.R. (2007). Isolation and characterization

of microsatellites in Lane Snapper (Lutjanus synagris), Mutton Snapper

(Lutjanus analis), and Yellowtail Snapper (Ocyurus chrysurus). Mol. Ecol.

Notes, 7, 1084-1087. doi:10.1111/j.1471-8286.2007.01785.x

73. Shulzitski K. (2005). A genetic assessment of population connectivity in

Mutton Snapper, Lutjanus analis. Master of Science. University of North

Carolina, Wilmington.

74. Shulzitski K., McCartney M.A. & Burton M.L. (2009). Population

connectivity among Dry Tortugas, Florida, and Caribbean populations of

Mutton Snapper (Lutjanus analis), inferred from multiple microsatellite loci.

Fish. Bull., 107, 501-509.

75. Zhang J., Cai Z. & Huang H. (2006). Isolation and characterization of

microsatellite loci from Mangrove Red Snapper Lutjanus argentimaculatus.

Mol. Ecol. Notes, 6, 408-411. doi:10.1111/j.1471-8286.2006.01251.x

76. Gold J.R., Pak E. & Richardson L.R. (2001). Microsatellite variation among

Red Snapper (Lutjanus campechanus) from the Gulf of Mexico. Mar.

Biotechnol., 3, 293-304. doi:10.1007/s10126-001-0004-7

77. Saillant E. & Gold J.R. (2006). Population structure and variance effective size

of Red Snapper (Lutjanus campechanus) in the northern Gulf of Mexico. Fish.

Bull., 104, 136-148.

156

78. Saillant E., Bradfield S.C. & Gold J.R. (2006). Genetic impacts of shrimp

trawling on Red Snapper (Lutjanus campechanus) in the northern Gulf of

Mexico. ICES J. Mar. Sci., 63, 705-713. doi:10.1016/j.icesjms.2005.12.005

79. Bradbury I.R., Coulson M.W., Campana S.E. & Bentzen P. (2006).

Morphological and genetic differentiation in anadromous Smelt Osmerus

mordax (Mitchill): disentangling the effects of geography and morphology on

gene flow. J. Fish Biol., 69, 95-114. doi:10.1111/j.1095-8649.2006.01263.x

80. Bradbury I.R., Campana S.E. & Bentzen P. (2008). Low genetic connectivity

in an estuarine fish with pelagic larvae. Can. J. Fish. Aquat. Sci., 65, 147-158.

doi:10.1139/F07-154

81. Coulson M.W., Paterson I.G., Green A., Kepkay R. & Bentzen P. (2006).

Characterization of di- and tetranucleotide microsatellite markers in Rainbow

Smelt (Osmerus mordax). Mol. Ecol. Notes, 6, 942- 944. doi:10.1111/j.1471-

8286.2006.01409.x

82. Gordos K., Kenchington E., Hamilton L.C., Nakashima B.S. & Taggart C.T.

(2005). Atlantic Capelin (Mallotus villosus) tetranucleotide microsatellites.

Mol. Ecol. Notes, 5, 220-222. doi:10.1111/j.1471-8286.2005.00882.x

83. McLean J.E. & Taylor E.B. (2001). Resolution of population structure in a

species with high gene flow: microsatellite variation in the Eulachon

(Osmeridae: Thaleichthys pacificus). Mar. Biol., 139, 411-420.

doi:10.1007/s002270100483

84. Præbel K., Westgaard J.I., Fevolden S.-E. & Christiansen J.S. (2008).

Circumpolar genetic population structure of Capelin Mallotus villosus. Mar.

Ecol. Prog. Ser., 360, 189-199. doi:10.3354/meps07363

85. Røed K.H., Fevolden S.-E., Christiansen J.S. & Midthjell L. (2003). Isolation

and characterization of DNA microsatellites for Capelin Mallotus villosus

population genetic studies. Mar. Ecol. Prog. Ser., 262, 299-303.

86. Lage C.R. & Kornfield I. (1999). Isolation and characterization of

microsatellite loci in Atlantic Haddock (Melanogrammus aeglefinus). Mol.

Ecol., 8, 1355-1357.

157

87. Lage C.R., Purcell M., Fogarty M. & Kornfield I. (2001). Microsatellite

evaluation of Haddock (Melanogrammus aeglefinus) stocks in the northwest

Atlantic Ocean. Can. J. Fish. Aquat. Sci., 58, 982-990. doi:10.1139/cjfas-58-5-

982

88. O'Reilly P.T., Mcpherson A.A., Kenchington E., Taggart C.T., Jones M.W. &

Bentzen P. (2002). Isolation and characterization of tetranucleotide

microsatellites from Atlantic Haddock (Melanogrammus aeglefinus). Mar.

Biotechnol., 4, 418-422. doi:10.1007/s10126-002-0010-4

89. Charrier G., Coombs S.H., McQuinn I.H. & Laroche J. (2007). Genetic

structure of Whiting Merlangius merlangus in the northeast Atlantic and

adjacent waters. Mar. Ecol. Prog. Ser., 330, 201-211.

90. Rico C., Ibrahim K.M., Rico I. & Hewitt G.M. (1997). Stock composition in

North Atlantic populations of Whiting using microsatellite markers. J. Fish

Biol., 51, 462-475.

91. Castillo A.G.F., Martínez J.L. & García-Vázquez E. (2003). Identification and

Atlantic hake species by a simple PCR-based methodology employing

microsatellite loci. J. Food Prot., 66, 2130-2134.

92. Castillo A.G.F., Alvarez P. & Garcia-Vazquez E. (2005). Population structure

of Merluccius merluccius along the Iberian Peninsula coast. ICES J. Mar. Sci.,

62, 1699-1704. doi:10.1016/j.icesjms.2005.06.001

93. Lundy C.J., Moran P., Rico C., Milner R.S. & Hewitt G.M. (1999).

Macrogeographical population differentiation in oceanic environments: a case

study of European Hake (Merluccius merluccius), a commercially important

fish. Mol. Ecol., 8, 1889 -1898.

94. Lundy C.J., Rico C. & Hewitt G.M. (2000). Temporal and spatial genetic

variation in spawning grounds of European Hake (Merluccius merluccius) in

the Bay of Biscay. Mol. Ecol., 9, 2067 - 2079.

95. Machado-Schiaffino G. & Garcia-Vazquez E. (2009). Isolation and

characterization of microsatellite loci in Merluccius australis and cross-species

158

amplification. Molecular Ecology Resources, 9, 585-587. doi:10.1111/j.1755-

0998.2008.02442.x

96. Moran P., Ryan A.W., Rico C. & Hewitt G.M. (1999). Four microsatellite loci

in the gadoid fish, Blue Whiting Micromesistius poutassou (Riso 1826). Anim.

Genet., 30, 463-464.

97. Ryan A.W., Mattiangeli V. & Mork J. (2005). Genetic differentiation of Blue

Whiting (Micromesistius poutassou Risso) populations at the extremes of the

species range and at the Hebrides-Porcupine Bank spawning grounds. ICES J.

Mar. Sci., 62, 948-955. doi:10.1016/j.icesjms.2005.03.006

98. Was A., Gosling E., McCrann K. & Mork J. (2008). Evidence for population

structuring of Blue Whiting (Micromesistius poutassou) in the Northeast

Atlantic. ICES J. Mar. Sci., 65, 216-225.

99. Chapman R.W., Sedberry G.R., Koenig C.C. & Eleby B.M. (1999). Stock

identification of Gag, Mycteroperca microlepis, along the southeast coast of

the United States. Mar. Biotechnol., 1, 137-146.

100. Cushman E.L., Jue N.K., Strand A.E. & Sotka E.E. (2009). Evaluating the

demographic significance of genetic homogeneity using a coalescent-based

simulation: a case study with Gag (Mycteroperca microlepis). Can. J. Fish.

Aquat. Sci., 66, 1821-1830. doi:10.1139/F09-140

101. Jue N.K. (2010). The role of larval dispersal in the population genetics and

ecology of Gag, Mycteroperca microlepis, in the Gulf of Mexico. Doctor of

Philosophy. Florida State University.

102. Zatcoff M.S., Ball A.O. & Chapman R.W. (2002). Characterization of

polymorphic microsatellite loci from Black Grouper, Mycteroperca bonaci

(Teleostei: Serranidae). Mol. Ecol. Notes, 2, 217-219.

103. Zatcoff M.S., Ball A.O. & Sedberry G.R. (2004). Population genetic analysis

of Red Grouper, Epinephelus morio, and Scamp, Mycteroperca phenax, from

the southeastern U.S. Atlantic and Gulf of Mexico. Mar. Biol., 144, 769-777.

doi:10.1007/s00227-003-1236-z

159

104. Alarcón J.A., Magoulas A., Georgakopoulos T., Zouros E. & Alvarez M.C.

(2004). Genetic comparison of wild and cultivated European populations of the

Gilthead Sea Bream (Sparus aurata). Aquaculture, 230, 65-80.

105. An H.S., Hong S.W., Lee J.U., Park J.Y. & Kim K.-K. (2010). Genetic

diversity of wild and farmed Black Sea Bream populations in Jeju. Animal

Cells and Systems, 14, 37-44. doi:10.1080/19768351003765020

106. Bernal-Ramírez J.H., Adcock G.J., Hauser L., Carvalho G.R. & Smith P.J.

(2003). Temporal stability of genetic population structure in the New Zealand

Snapper, Pagrus auratus, and relationship to coastal currents. Mar. Biol., 142,

567-574. doi:10.1007/s00227-002-0972-9

107. Brown R.C., Tsalavouta M., Terzoglou V., Magoulas A. & Mcandrew B.J.

(2005). Additional microsatellites for Sparus aurata and cross-species

amplification within the Sparidae family. Mol. Ecol. Notes, 5, 605-607.

doi:10.1111/j.1471-8286.2005.01007.x

108. de Innocentiis S., Lesti A., Livi S., Rossi A.R., Crosetti D. & Sola L. (2004).

Microsatellite markers reveal population structure in Gilthead Sea Bream

Sparus auratus from the Atlantic Ocean and Mediterranean Sea. Fish. Sci., 70,

852-859.

109. Hauser L., Adcock G.J., Smith P.J., Ramírez J.H.B. & Carvalho G.R. (2002).

Loss of microsatellite diversity and low effective population size in an

overexploited population of New Zealand Snapper (Pagrus auratus). P Natl

Acad Sci USA, 99, 11742-11747.

110. Jeong D.-S., Umino T., Kuroda K., Hayashi M., Nakagawa H., Kang J.-C., et

al. (2003). Genetic divergence and population structure of Black Sea Bream

Acanthopagrus schlegeli inferred from microsatellite analysis. Fish. Sci., 69,

896-902.

111. Karaiskou N., Triantafyllidis A., Katsares V., Abatzopoulos T.J. &

Triantaphyllidis C. (2009). Microsatellite variability of wild and farmed

populations of Sparus aurata. J. Fish Biol., 74, 1816-1825.

doi:10.1111/j.1095-8649.2009.02186.x

160

112. Molecular Ecology Resources Primer Development Consortium, An J., Bechet

A., Berggren Å., Brown S.K., Bruford M.W., et al. (2010). Permanent genetic

resources added to Molecular Ecology Resources Database 1 October 2009 –

30 November 2009. Molecular Ecology Resources, 10, 404-408.

doi:10.1111/j.1755-0998.2009.02827.x

113. Palma J., Alarcón J.A., Alvarez C., Zouros E., Magoulas A. & Andrade J.P.

(2001). Developmental stability and genetic heterozygosity in wild and

cultured stocks of Gilthead Sea Bream (Sparus aurata). J. Mar. Biol. Assoc.

U.K., 81, 283-288.

114. Piñera J.A., Bernardo D., Blanco G., Vázquez E. & Sánchez J.A. (2007).

Usefulness of microsatellite markers developed from Pagellus bogaraveo to

genetically study five different species of Sparidae. Mar. Ecol., 28, 184-187.

115. Ramšak A., Garoia F., Guarniero I., Mannini P. & Tinti F. (2003). Novel

polymorphic microsatellite markers for the Common Pandora (Pagellus

erythrinus). Mol. Ecol. Notes, 3, 553 - 555. doi:10.1046/j.1471-

8286.2003.00509.x

116. Sala-Bozano M., Ketmaier V. & Mariani S. (2009). Contrasting signals from

multiple markers illuminate population connectivity in a marine fish. Mol.

Ecol., 18, 4811-4826. doi:10.1111/j.1365-294X.2009.04404.x

117. Sala-Bozano M., Tsalavouta M. & Mariani S. (2009). Isolation and

characterisation of new polymorphic microsatellite markers for the Striped Sea

Bream (Lithognathus mormyrus). Conserv. Genet., 10, 1507-1509.

doi:10.1007/s10592-008-9770-3

118. Ball A.O., Beal M.G., Chapman R.W. & Sedberry G.R. (2007). Population

structure of Red Porgy, Pagrus pagrus, in the Atlantic Ocean. Mar. Biol., 150,

1321-1332. doi:10.1007/s00227-006-0425-y

119. Hemmer-Hansen J., Nielsen E.E., Grønkjær P. & Loeschcke V. (2007).

Evolutionary mechanisms shaping the genetic population structure of marine

fishes; lessons from the European Flounder (Platichthys flesus L.). Mol. Ecol.,

16, 3104- 3118. doi:10.1111/j.1365-294X.2007.03367.x

161

120. Kim S.G., Morishima K. & Arai K. (2007). Isolation and characterization of

polymorphic microsatellite DNA markers in the Brown Sole, Pleuronectes

herzensteini. Mol. Ecol. Notes, 7, 79- 81. doi:10.1111/j.1471-

8286.2006.01535.x

121. Miao G., Shao C., Ma H., Liao X. & Chen S. (2009). Isolation and

characterization of polymorphic microsatellite loci from a dinucleotide-

enriched genomic library of Starry Flounder (Platichthys stellatus) and cross-

species amplification. Conserv. Genet., 10, 631-633. doi:10.1007/s10592-008-

9594-1

122. Tysklind N., Neuparth T., Ashcroft G.R., Taylor M.I., Lyonds B.P., McCarthy

I.D., et al. (2009). Isolation and characterization of 28 new microsatellite

markers for European Flounder (Platichthys flesus L.). Mol. Ecol., 9, 1065-

1068. doi:10.1111/j.1755-0998.2009.02552.x

123. Hoarau G., Cook D., Stam W.T. & Olsen J.L. (2002). New microsatellites

primers for Plaice, Pleuronectes platessa L. (Teleostei: Pleuronectidae). Mol.

Ecol. Notes, 2, 60-61.

124. Hoarau G., Rijnsdorp A.D., Van Der Veer H.W., Stam W.T. & Olsen J.L.

(2002). Population structure of Plaice (Pleuronectes platessa L.) in northern

Europe: microsatellites revealed large-scale spatial and temporal homogeneity.

Mol. Ecol., 11, 1165 -1176.

125. Was A., Gosling E. & Hoarau G. (2010). Microsatellite analysis of Plaice

(Pleuronectes platessa L.) in the NE Atlantic: weak genetic structuring in a

milieu of high gene flow. Mar. Biol., 157, 447-462. doi:10.1007/s00227-009-

1331-x

126. Watts P.C., Nash R.D.M., George S.G. & Kemp S.J. (1999). Isolation and

characterization of microsatellite loci in the European Plaice, Pleuronectes

platessa L. (Teleostei: Pleuronectidae). Mol. Ecol., 8, 2151-2152.

127. Watts P.C., Nash R.D.M. & Kemp S.J. (2001). Polymorphic microsatellite loci

in the European Plaice, Pleuronectes platessa, and their utility in Flounder,

Lemon Sole and Dover Sole. J. Mar. Biol. Assoc. U.K., 81, 367-368.

162

128. Watts P.C., Nash R.D.M. & Kemp S.J. (2004). Genetic structure of juvenile

Plaice Pleuronectes platessa on nursery grounds within the Irish Sea. J. Sea

Res., 51, 191 - 197. doi:10.1016/j.seares.2003.09.003

129. Watts P.C., Kay S.M., Wolfenden D., Fox C.J., Geffen A.J., Kemp S.J., et al.

(2010). Temporal patterns of spatial genetic structure and effective population

size in European Plaice (Pleuronectes platessa) along the west coast of

Scotland and in the Irish Sea. ICES J. Mar. Sci., 67, 607-616.

130. Buckley L., Collie J., Kaplan L.A.E. & Crivello J.F. (2008). Winter Flounder

larval genetic population structure in Narragansett Bay, RI: recruitment to

juvenile young-of-the-year. Estuaries and Coasts, 31, 745-754.

doi:10.1007/s12237-008-9065-4

131. Crivello J.F., Danila D.J., Lorda E., Keser M. & Roseman E.F. (2004). The

genetic stock structure of larval and juvenile Winter Flounder larvae in

Connecticut waters of eastern Long Island Sound and estimations of larval

entrainment. J. Fish Biol., 65, 62-76. doi:10.1111/j.1095-8649.2004.00424.x

132. McClelland G., Melendy J., Osborne J., Reid D. & Douglas S. (2005). Use of

parasite and genetic markers in delineating populations of Winter Flounder

from the central and south-west Scotian Shelf and north-east Gulf of Maine. J.

Fish Biol., 66, 1082-1100. doi:10.1111/j.1095-8649.2005.00659.x

133. Pereyra R.T., Saillant E., Pruett C.L., Rexroad III C.E., Rocha-Olivares A. &

Gold J.R. (2004). Characterization of polymorphic microsatellites in the

Pacific Sardine Sardinops sagax sagax (Clupeidae). Mol. Ecol. Notes, 4, 739-

741. doi:10.1111/j.1471-8286.2004.00801.x

134. Appleyard S.A., Grewe P.M., Innes B.H. & Ward R.D. (2001). Population

structure of Yellowfin Tuna (Thunnus albacares) in the western Pacific Ocean,

inferred from microsatellite loci. Mar. Biol., 139, 383-393.

doi:10.1007/s002270100578

135. Appleyard S.A., Ward R.D. & Grewe P.M. (2002). Genetic stock structure of

Bigeye Tuna in the Indian Ocean using mitochondrial DNA and

microsatellites. J. Fish Biol., 60, 767-770. doi:10.1006/jfbi.2002.1866

163

136. Broughton R.E., Stewart L.B. & Gold J.R. (2002). Microsatellite variation

suggests substantial gene flow between King Mackerel (Scomberomorus

cavalla) in the western Atlantic Ocean and Gulf of Mexico. Fisheries

Research, 54, 305-316.

137. Catanese G., Infante C., Crespo A., Zuasti E., Ponce M., Funes V., et al.

(2007). Development and characterization of eight microsatellite markers in

Bullet Tuna (Auxis rochei). Mol. Ecol. Notes, 7, 842-844. doi:10.1111/j.1471-

8286.2007.01723.x

138. Cha H.K., An H.S., Choi J.H., Kang S., Park J.Y. & Kim K.K. (2010).

Isolation and characterization of polymorphic microsatellite markers for

genetic analysis of Chub Mackerel (Scomber japonicus). Conservation

Genetics Resources, 2, 7-9. doi:10.1007/s12686-009-9123-7

139. Clark T.B., Ma L., Saillant E. & Gold J.R. (2004). Microsatellite DNA

markers for population-genetic studies of Atlantic Bluefin Tuna (Thunnus

thynnus thynnus) and other species of genus Thunnus. Mol. Ecol. Notes, 4, 70-

73. doi:10.1046/j.1471-8286.2003.00572.x

140. Dammannagoda S.T. (2007). Genetic stock structure and inferred migratory

patterns of Skipjack Tuna (Katsuwonus pelamis) and Yellowfin Tuna

(Thunnus albacares) in Sri Lankan waters. Doctor of Philosophy. Queensland

University of Technology, Brisbane.

141. Dammannagoda S.T., Hurwood D.A. & Mather P.B. (2008). Evidence for fine

geographical scale heterogeneity in gene frequencies in Yellowfin Tuna

(Thunnus albacares) from the north Indian Ocean around Sri Lanka. Fisheries

Research, 90, 147-157. doi:10.1016/j.fishres.2007.10.006

142. Díaz-Jaimes P. & Uribe-Alcocer M. (2006). Spatial differentiation in the

eastern Pacific Yellowfin Tuna revealed by microsatellite variation. Fish. Sci.,

72, 590-596.

143. Durand J.-D., Collet A., Chow S.L., Guinand B. & Borsa P. (2005). Nuclear

and mitochondrial DNA markers indicate unidirectional gene flow of Indo-

Pacific to Atlantic Bigeye Tuna (Thunnus obesus) populations, and their

164

admixture off southern Africa. Mar. Biol., 147, 313-322. doi:10.1007/s00227-

005-1564-2

144. Farnham T.T. (2003). Genetic variation in Atlantic Yellowfin Tuna (Thunnus

albacares) to assess stock structure and reproductive variance. Master of

Science. Texas A&M University.

145. Gold J.R., Pak E. & DeVries D.A. (2002). Population structure of King

Mackerel (Scomberomorus cavalla) around peninsular Florida, as revealed by

microsatellite DNA. Fish. Bull., 100, 491-509.

146. Gold J.R., Jobity A.M.C., Saillant E. & Renshaw M.A. (2010). Population

structure of Carite (Scomberomorus brasiliensis) in waters offshore of Trinidad

and northern Venezuela. Fisheries Research, 103, 30-39.

doi:10.1016/j.fishres.2010.01.009

147. Gonzalez E.G., Beerli P. & Zardoya R. (2008). Genetic structuring and

migration patterns of Atlantic Bigeye Tuna, Thunnus obesus (Lowe, 1839).

BMC Evol. Biol., 8, 1-14. doi:10.1186/1471-2148-8-252

148. Menezes M.R., Noguchi D., Nakajima M. & Taniguchi N.K. (2008).

Microsatellite development and survey of genetic variation in Skipjack Tuna

Katsuwonus pelamis. J. Fish Biol., 73, 463-473. doi:10.1111/j.1095-

8649.2008.01912.x

149. Morishima K., Yamamoto H., Sawada Y., Miyashita S. & Kato K. (2009).

Developing 23 new polymorphic microsatellite markers and simulating

parentage assignment in the Pacific Bluefin Tuna, Thunnus orientalis.

Molecular Ecology Resources, 9, 790-792.

150. Renshaw M.A., Douglas K.C., Rexroad III C.E., Jobity A.M.C. & Gold J.R.

(2009). Isolation and characterization of microsatellite markers in the Serra

Spanish Mackerel, Scomberomorus brasiliensis. Molecular Ecology

Resources, 9, 835-838. doi:10.1111/j.1755-0998.2008.02284.x

151. Saxton B.L. (2009). Historical demography and genetic population structure of

the Blackfin Tuna (Thunnus atlanticus) from the northwest Atlantic Ocean and

the Gulf of Mexico. Master of Science. Texas A&M University.

165

152. Takagi M., Okamura T., Chow S.L. & Taniguchi N.K. (2001). Preliminary

study of Albacore (Thunnus alalunga) stock differentiation inferred from

microsatellite DNA analysis. Fish. Bull., 99, 697-701.

153. Takagi M., Okamura T., Chow S.L. & Taniguchi N.K. (1999). PCR primers

for microsatellite loci in tuna species of the genus Thunnus and its application

for population genetic study. Fish. Sci., 65, 571-576.

154. Tang C.Y., Tzeng C.-H., Chen C.S. & Chiu T.S. (2009). Microsatellite DNA

markers for population-genetic studies of Blue Mackerel (Scomber

australasicus) and cross-specific amplification in S. japonicus. Molecular

Ecology Resources, 9, 824-827. doi:10.1111/j.1755-0998.2008.02276.x

155. Tzeng C.-H., Chen C.-S., Tang P.-C. & Chiu T.-S. (2009). Microsatellite and

mitochondrial haplotype differentiation in Blue Mackerel (Scomber

australasicus) from the western North Pacific. ICES J. Mar. Sci., 66, 816-825.

156. van Herwerden L., McIlwain J., Al-Oufi H., Al-Amry W. & Reyes A. (2006).

Development and application of microsatellite markers for Scomberomorus

commerson (Perciformes; Teleostei) to a population genetic study of Arabian

Peninsula stocks. Fisheries Research, 79, 258-266.

doi:10.1016/j.fishres.2006.04.004

157. Xing S.-C., Xu G.-B., Liao X.-L., Yang G.-P. & Chen S.-L. (2009). Twelve

polymorphic microsatellite loci from a dinucleotide-enriched genomic library

of Japanese Spanish Mackerel (Scomberomorus niphonius). Conserv. Genet.,

10, 1167-1169. doi:10.1007/s10592-008-9735-6

158. Yagishita N. & Kobayashi T. (2008). Isolation and characterization of nine

microsatellite loci from the Chub Mackerel, Scomber japonicus (Perciformes,

Scombridae). Molecular Ecology Resources, 8, 302-304. doi:10.1111/j.1471-

8286.2007.01935.x

159. Yokoyama E., Sakamoto T., Sugaya T. & Kitada S. (2006). Six polymorphic

microsatellite loci in the Japanese Spanish Mackerel, Scomberomorus

niphonius. Mol. Ecol. Notes, 6, 323-324. doi:10.1111/j.1471-

8286.2005.01217.x

166

160. Gharrett A.J. (2004). Distribution of two species of Rougheye Rockfish

(Sebastes aleutianus) in the northeastern Gulf of Alaska. North Pacific

Research Board, Anchorage, AK.

161. Miller K.M., Schulze A.D. & Withler R.E. (2000). Characterization of

microsatellite loci in Sebastes alutus and their conservation in congeneric

rockfish species. Mol. Ecol., 9, 237-244.

162. Miller J.A., Banks M.A., Gomez-Uchida D. & Shanks A.L. (2005). A

comparison of population structure in Black Rockfish (Sebastes melanops) as

determined with otolith microchemistry and microsatellite DNA. Can. J. Fish.

Aquat. Sci., 62, 2189-2198. doi:10.1139/F05-133

163. Narum S.R., Buonaccorsi V.P., Kimbrell C.A. & Vetter R.D. (2004). Genetic

divergence between Gopher Rockfish (Sebastes carnatus) and Black and

Yellow Rockfish (Sebastes chrysomelas). Copeia, 2004, 926-931.

164. Pearse D.E., Wooninck L., Dean C.A. & Garza J.C. (2007). Identification of

northeastern Pacific rockfish using multilocus nuclear DNA genotypes. Trans.

Am. Fish. Soc., 136, 272-280. doi:10.1577/T06-051.1

165. Seeb L.W. (2007). Genetic markers distinguish populations of Black Rockfish

in the Gulf of Alaska. In: Conservation of North Pacific rockfishes: Ecological

genetics and stock structure. Proceedings of the workshop March 2-3, 2004,

Seattle, Washington (eds. Berntson EA, Levin PS & Moran PC). U.S. Dept.

Commer., NOAA Tech. Memo. NMFS-NWFSC-80 Seattle, WA, pp. 57-60.

166. Sivasundar A. & Palumbi S.R. (2010). Life history, ecology and the

biogeography of strong genetic breaks among 15 species of Pacific rockfish,

Sebastes. Mar. Biol., 157, 1433-1452. doi:10.1007/s00227-010-1419-3

167. Westerman M.E., Buonaccorsi V.P., Stannard J.A., Galver L., Taylor C., Lynn

E.A., et al. (2005). Cloning and characterization of novel microsatellite DNA

markers for the Grass Rockfish, Sebastes rastrelliger, and cross-species

amplification in 10 related Sebastes spp. Mol. Ecol. Notes, 5, 74-76.

doi:10.1111/j.1471-8286.2004.00837.x

167

168. Withler R.E., Beacham T.D., Schulze A.D., Richards L.J. & Miller K.M.

(2001). Co-existing populations of Pacific Ocean Perch, Sebastes alutus, in

Queen Charlotte Sound, British Columbia. Mar. Biol., 139, 1-12.

doi:10.1007/s002270100560

169. Gomez-Uchida D. & Banks M.A. (2005). Microsatellite analyses of spatial

genetic structure in Darkblotched Rockfish (Sebastes crameri): is pooling

samples safe? Can. J. Fish. Aquat. Sci., 62, 1874-1886. doi:10.1139/F05-084

170. Pampoulie C. & Daníelsdóttir A.K. (2008). Resolving species identification

problems in the genus Sebastes using nuclear genetic markers. Fisheries

Research, 93, 54-63. doi:10.1016/j.fishres.2008.02.007

171. Roques S., Pallotta D., Sévigny J.-M. & Bernatchez L. (1999). Isolation and

characterization of polymorphic microsatellite markers in the North Atlantic

redfish (Teleostei: Scorpaenidae, genus Sebastes). Mol. Ecol., 8, 655-702.

172. Roques S., Duchesne P. & Bernatchez L. (1999). Potential of microsatellites

for individual assignment: the North Atlantic redfish (genus Sebastes) species

complex as a case study. Mol. Ecol., 8, 1703 -1717.

173. Roques S., Sévigny J.-M. & Bernatchez L. (2001). Evidence for broadscale

introgressive hybridization between two redfish (genus Sebastes) in the North-

west Atlantic: a rare marine example. Mol. Ecol., 10, 149-165.

174. Schmidt C. (2005). Molecular genetic studies on species and population

structure of North Atlantic redfish (Genus Sebastes; Cuvier 1829). Doctor of

Philosophy. University of Hamburg.

175. Roques S., Sévigny J.-M. & Bernatchez L. (2002). Genetic structure of Deep-

water Redfish, Sebastes mentella, populations across the North Atlantic. Mar.

Biol., 140, 297-307. doi:10.1007/s002270100705

176. Stefánsson M.Ö., Reinert J., Sigurðsson o., Kristinsson K. & Nedreaas K.

(2009). Depth as a potential driver of genetic structure of Sebastes mentella

across the North Atlantic Ocean. ICES J. Mar. Sci., 66, 680-690.

177. Stefánsson M.Ö., Sigurdsson T., Pampoulie C., Daníelsdóttir A.K.,

Thorgilsson B., Ragnardóttir A., et al. (2009). Pleistocene genetic legacy

168

suggests incipient species of Sebastes mentella in the Irminger Sea. Heredity,

102, 514-524. doi:10.1038/hdy.2009.10

178. Burford M.O. & Larson R.J. (2007). Genetic heterogeneity in a single year-

class from a panmictic population of adult Blue Rockfish (Sebastes mystinus).

Mar. Biol., 151, 451-465. doi:10.1007/s00227-006-0475-1

179. Burford M.O. & Bernardi G. (2008). Incipient speciation within a subgenus of

rockfish (Sebastosomus) provides evidence of recent radiations within an

ancient species flock. Mar. Biol., 154, 701-717. doi:10.1007/s00227-008-0963-

6

180. Burford M.O. (2009). Demographic history, geographical distribution and

reproductive isolation of distinct lineages of Blue Rockfish (Sebastes

mystinus), a marine fish with a high dispersal potential. J. Evol. Biol., 22,

1471-1486. doi:10.1111/j.1420-9101.2009.01760.x

181. Matala A.P., Gray A.K. & Gharrett A.J. (2004). Microsatellite variation

indicates population genetic structure of Bocaccio. N. Am. J. Fish. Manage.,

24, 1189-1202.

182. Gomez-Uchida D., Hoffman A., Ardren W.R. & Banks M.A. (2003).

Microsatellite markers for the heavily exploited Canary (Sebastes pinniger)

and other rockfish species. Mol. Ecol. Notes, 3, 387- 389. doi:10.1046/j.1471-

8286.2003.00458.x

183. Hyde J.R., Kimbrell C.A., Budrick J.E., Lynn E.A. & Vetter R.D. (2008).

Cryptic speciation in the Vermilion Rockfish (Sebastes miniatus) and the role

of bathymetry in the speciation process. Mol. Ecol., 17, 1122-1136.

doi:10.1111/j.1365-294X.2007.03653.x

184. Yamanaka K.L., Withler R.E. & Miller K.M. (2000). Structure of Yelloweye

Rockfish (Sebastes ruberrimus) populations in British Columbia. Canadian

Stock Assessment Secretariat Research Document 2000/172. Fisheries and

Oceans Canada, Nanaimo, BC.

185. Dailianis T., Limborg M.T., Hanel R., Bekkevold D., Lagnel J., Magoulas A.,

et al. (2008). Characterization of nine polymorphic microsatellite markers in

169

Sprat (Sprattus sprattus L.). Molecular Ecology Resources, 8, 861-863.

doi:10.1111/j.1755-0998.2008.02091.x

186. Limborg M.T., Pedersen J.S., Hemmer-Hansen J., Tomkiewicz J. & Bekkevold

D. (2009). Genetic population structure of European Sprat Sprattus sprattus:

differentiation across a steep environmental gradient in a small pelagic fish.

Mar. Ecol. Prog. Ser., 379, 213-224. doi:10.3354/meps07889

187. O'Reilly P.T., Canino M.F., Bailey K.M. & Bentzen P. (2000). Isolation of

twenty low stutter di- and tetranucleotide microsatellites for population

analyses of Walleye Pollock and other gadoids. J. Fish Biol., 56, 1074-1086.

doi:10.1006/jfbi.2000.1230

188. O'Reilly P.T., Canino M.F., Bailey K.M. & Bentzen P. (2004). Inverse

relationship between FST and microsatellite polymorphism in the marine fish,

Walleye Pollock (Theragra chalcogramma): implications for resolving weak

population structure. Mol. Ecol., 13, 1799-1814. doi:10.1111/j.1365-

294X.2004.02214.x

189. Shubina E.A., Mel'nikova M.N., Glubokov A.I. & Mednikov B.M. (2004).

Analysis of the genetic structure of northwestern Bering Sea Walleye Pollock,

Theragra chalcogramma. Environ. Biol. Fishes, 69, 177-185.

190. Shubina E.A., Ponomareva E.V. & Glubokov A.I. (2009). Population genetic

structure of Walleye Pollock Theragra chalcogramma (Gadidae, Pisces) from

the Bering Sea and Sea of Okhotsk. Mol. Biol., 43, 855-866.

doi:10.1134/S0026893309050161

191. Carlsson J., McDowell J.R., Díaz-Jaimes P., Carlsson J.E.L., Boles S.B., Gold

J.R., et al. (2004). Microsatellite and mitochondrial DNA analyses of Atlantic

Bluefin Tuna (Thunnus thynnus thynnus) population structure in the

Mediterranean Sea. Mol. Ecol., 13, 3345-3356. doi:10.1111/j.1365-

294X.2004.02336.x

192. McDowell J.R., Díaz-Jaimes P. & Graves J.E. (2002). Isolation and

characterization of seven tetranucleotide microsatellite loci from Atlantic

170

Northern Bluefin Tuna Thunnus thynnus thynnus. Mol. Ecol. Notes, 2, 214-

216. doi:10.1046/j.1471-8286

193. Riccioni G., Landi M., Ferrara G., Milano I., Cariani A., Zane L., et al. (2010).

Spatio-temporal population structuring and genetic diversity retention in

depleted Atlantic Bluefin Tuna of the Mediterranean Sea. P Natl Acad Sci

USA, 107, 2102-7. doi:10.1073/pnas.0908281107

194. Canales-Aguirre C., Ferrada S. & Galleguillos R. (2010). Isolation and

characterization of microsatellite loci for the Jack Mackerel (Trachurus

murphyi Nichols, 1920). Conserv. Genet., 11, 1235-1237. doi:10.1007/s10592-

009-9932-y

195. Cárdenas L., Silva A.X., Magoulas A., Cabezas J., Poulin E. & Ojeda F.P.

(2009). Genetic population structure in the Chilean Jack Mackerel, Trachurus

murphyi (Nichols) across the South-eastern Pacific Ocean. Fisheries Research,

100, 109-115. doi:10.1016/j.fishres.2009.06.013

196. Chang D.S., An H.S., Oh T.Y., Lee J.B., Park J.Y. & Kim K.K. (2009). Eleven

new microsatellite markers in Jack Mackerel (Trachurus japonicus) derived

from an enriched genomic library. Genes & Genomics, 31, 397-402.

197. Kasapidis P. & Magoulas A. (2008). Development and application of

microsatellite markers to address the population structure of the Horse

Mackerel Trachurus trachurus. Fisheries Research, 89, 132-135.

doi:10.1016/j.fishres.2007.09.015