Asymmetric connectivity of spawning aggregations of a - PeerJ

A peer-reviewed version of this preprint was published in PeerJ on 7August 2014.

View the peer-reviewed version (peerj.com/articles/511), which is thepreferred citable publication unless you specifically need to cite this preprint.

Munguia-Vega A, Jackson A, Marinone SG, Erisman B, Moreno-Baez M, Girón-Nava A, Pfister T, Aburto-Oropeza O, Torre J. 2014. Asymmetric connectivityof spawning aggregations of a commercially important marine fish using amultidisciplinary approach. PeerJ 2:e511 https://doi.org/10.7717/peerj.511

https://doi.org/10.7717/peerj.511

https://doi.org/10.7717/peerj.511

1

1

Asymmetric connectivity of spawning aggregations of a commercially important 1

marine fish using a multidisciplinary approach 2

3

Adrian Munguia-Vega1,2, Brad Erisman3, Alexis Jackson4, Silvio Guido Marinone5, 4

Marcia Moreno-Baez3, Alfredo Girón6, Tad Pfister7, Octavio Aburto-Oropeza3, Jorge 5

Torre1 6

7

1 Comunidad y Biodiversidad A.C., Guaymas, Sonora, México 8

2 Conservation Genetics Laboratory, School of Natural Resources and the Environment, 9

The University of Arizona, Tucson, AZ, USA 10

3 Marine Biology Research Division, Scripps Institution of Oceanography, University of 11

California, San Diego, La Jolla CA, USA 12

4 Department of Ecology and Evolutionary Biology, University of California Santa Cruz, 13

Santa Cruz, CA, USA 14

5 Departamento de Oceanografía Física, Centro de Investigación Científica y de 15

Educación Superior de Ensenada, Ensenada, Baja California, México 16

6Universidad Autónoma de Baja California, Ensenada, Baja California, México 17

7 School of Natural Resources ant the Environment, Center for Latin American Studies, 18

The University of Arizona, Tucson, AZ, USA 19

20

Corresponding Author: Adrian Munguia-Vega. Comunidad y Biodiversidad A.C., Isla del 21

Peruano 215, Col. Lomas de Miramar, Guaymas, Sonora CP 85448, Mexico. Phone: + 22

(52) 612-1232233. E-mail: [email protected] 23

PeerJ PrePrints | https://peerj.com/preprints/170v1/ | v1 received: 23 Dec 2013, published: 23 Dec 2013, doi: 10.7287/peerj.preprints.170v1

PrePrin

ts

2

2

Abstract 24

Understanding patterns of larval dispersal is key in determining whether no-take marine 25

reserves are self-sustaining, what will be protected inside reserves and where the benefits 26

of reserves will be observed. However, explicitly incorporating dispersal data into 27

designing reserves for fisheries and conservation is still uncommon in many places 28

around the world. We followed a multidisciplinary approach that merged detailed 29

descriptions of fishing zones and spawning time at 17 sites distributed in the Midriff 30

Island region of the Gulf of California (GC) with a biophysical oceanographic model that 31

simulated larval transport at Pelagic Larval Duration (PLD) 14, 21 and 28 days for the 32

most common and targeted predatory reef fish (leopard grouper Mycteroperca rosacea). 33

M. rosacea is endemic to the GC and considered ‘Vulnerable’ according to World 34

Conservation Union. We described metapopulation dynamics using graph theory and 35

employed empirical sequence data from a subset of 10 sites at two mitochondrial genes to 36

verify the model predictions. Our approach made sense of seemingly chaotic patterns of 37

genetic diversity and structure, and provided a mechanistic explanation of the location of 38

fishing zones. Most of the connectivity patterns observed were strictly asymmetric, 39

except for a small region in the Southeast. The best-supported gene flow model 40

confirmed a pulse of larvae from the Baja Peninsula, across the GC and northward up the 41

Sonoran coastline, in agreement with the cyclonic gyre present at the peak of spawning 42

(May). We found support that genetic diversity increased in sink sites that concentrated 43

larvae from many sources at the time of larval flexion (PLD 14 days), while diversity 44

decreased at important gateways identified at PLD 28 days with high betweenness 45

centrality that are key for multigenerational dispersal and population resilience. Heavily 46


PrePrin

ts

3

3

targeted fished areas seem to be sustained by high levels of local retention, contribution 47

of larvae from upstream sites and oceanographic patterns that concentrate larval density 48

from all over the region. The general asymmetry in marine connectivity observed 49

highlights that benefits from reserves are biased towards particular directions, that no-50

take areas need to be located upstream of targeted fishing zones, and that some fishing 51

localities might not directly benefit from avoiding fishing within reserves located 52

adjacent to their communities. We discuss the implications of marine connectivity for the 53

current network of marine protected areas and no-take zones, and identify ways of 54

improving it. 55

56

57

58

59

60

61

62

63

64

65

66

67

68

69


PrePrin

ts

4

4

Introduction 70

Knowledge of patterns of larval dispersal is essential to implement fully-protected 71

marine reserves (no-take zones), a tool frequently used to enhance the conservation of 72

biodiversity and the recovery of fisheries (Gaines et al. 2010). Reserves must either be 73

self-sufficient via local retention (larvae retained or returning to the reserve where they 74

were produced), or need to be linked by a network of reserves for persistence (larval 75

supply from reserves to other reserves/fished sites) (Hastings & Botsford 2006; White et 76

al. 2010). However, the efficacy of networks of reserves has been hindered by a lack of 77

knowledge regarding complex patterns of marine connectivity (Burgess et al. 2013; Sale 78

et al. 2005). A multidisciplinary approach could best address the intricacy of connectivity 79

by merging biophysical models of ocean currents that generate connectivity hypotheses 80

using detailed biological information on the spatial and temporal distribution of 81

propagules (larvae), followed by validation with empirical population genetics data 82

(Alberto et al. 2011; Crandall et al. 2012; Feutry et al. 2013; Foster et al. 2012; Soria et 83

al. 2012). A multi-prong approach could also help advance an increasing interest in 84

incorporating genetic information into marine spatial planning, for instance, by 85

identifying sites with high genetic diversity that hold evolutionary potential under future 86

environmental change (Beger et al. 2013). 87

Marine connectivity within a single species is influenced by multiple biological and 88

physical factors including: spawning time and location, pelagic larval duration (PLD), 89

their interaction with ocean current speed and direction, as well as the distribution of 90

suitable habitat for settlement (Cowen & Sponaugle 2009). Additionally, many 91

commercially exploited species of invertebrates and fishes display meta populations that 92


PrePrin

ts

5

5

are connected via larval dispersal (Cowen 2000). This is why relatively few attempts 93

have been done to establish multidisciplinary approaches to understand marine 94

connectivity, and the great challenge is to find a key species that can be a relevant case 95

study, and which can be use to gather this information relative easily and can be use as an 96

umbrella species to design marine reserves. 97

The leopard grouper Mycteroperca rosacea (Streets, 1877), is a large 98

predatory reef fish (Teleostei: Epinephelidae) endemic to the Gulf of California (GC) 99

bioregion. It ranges from Bahía Magdalena in the Pacific coast of the Baja California 100

Peninsula south to Bahía Banderas in Nayarit, Mexico, including all rocky-reefs within 101

the interior of GC (Hastings et al. 2010; Robertson & Cramer 2009; Thomson et al. 102

2000). Ecologically, it represents the most common and numerically abundant fish top 103

predator on reefs in the entire GC. Individuals can reach 1 m in length and at least 22 104

years of age (Diaz-Uribe et al. 2001). Histological and population data indicate 105

gonochorism, with no evidence of post-maturational sex change found in adults caught in 106

the wild (Erisman et al. 2007b). Adults form spawning aggregations of hundreds of 107

individuals during spring in the GC, with spawning occurring earlier in southern 108

locations (Erisman et al. 2007a; Sala et al. 2003). Spawning occurs in the evening within 109

groups of 6 to 40 individuals and is not correlated with the lunar cycle (Erisman et al. 110

2007a). 111

Small fisheries world-wide comprise most of the global catch, yet most lack formal 112

assessments and are thought to continue to decline (Costello et al. 2012). M. rosacea is 113

the most heavily targeted grouper by commercial, artisanal, and recreational fisheries in 114

the GC (Craig et al. 2012; Sala et al. 2003). Due to increased fishing pressure and 115


PrePrin

ts

6

6

observed declines in fisheries landings, sizes of harvested fish, and population 116

abundances in some areas of the GC over the past few decades (Sala et al. 2004), the 117

World Conservation Union (IUCN) currently lists M. rosacea as ‘Vulnerable’ (Craig & 118

Sadovy 2008). While commercial fishers are required to hold a finfish permit and record 119

their landings of leopard grouper to their local fisheries offices, no specific regulations 120

related to catch, size, or gear restrictions exist for this species. Currently, all M. rosacea 121

catches are aggregated in the finfish group with other 270 fish species according to the 122

National Fisheries Chart (CNP 2012). Marine Protected Areas (MPAs) represent the 123

primary, current conservation and management strategy that has been implemented for 124

this or any other reef fish in the GC, and are mainly concentrated in the western coast of 125

the GC and contain a few small no-take zones (Fig. 1) (Aburto-Oropeza et al. 2011; 126

Cudney-Bueno et al. 2009). 127

Our goal was to identify the potential larval connectivity of spawning aggregation 128

sites and fishing zones for M. rosacea where larvae recruit in the Midriff Island region of 129

the GC. We first determined the distinct spawning season for leopard grouper and 130

identified the spatial distribution of spawning aggregation sites and fishing zones across 131

the entire region. We modeled connectivity with a biophysical model and used graph 132

theory to describe metapopulation dynamics. We then contrasted distinct measures 133

derived from graph theory against empirical estimates of genetic diversity and 134

differentiation to corroborate model expectations' and identify sites that are likely self-135

sustaining and important sources and sinks for leopard grouper larvae, including locations 136

that may lie inside or outside the borders of existing MPAs. Results of this study provide 137

insights on validating biophysical models with empirical genetic data, on the benefits and 138


PrePrin

ts

7

7

limitations of the current network of MPAs to fishing communities in the Midriffs region 139

that harvest M. rosacea and help identify areas that may serve as ideal locations for 140

spawning or juveniles refuges of this economically important yet vulnerable species. 141

142

Methods 143

Spawning sites, season and period 144

We examined information acquired from underwater and fisheries surveys 145

conducted at locations throughout the Midriffs Islands region in the GC, Mexico (Fig. 1) 146

in order to identify representative sites to simulate the dispersal of leopard grouper eggs 147

and larvae from spawning aggregation sites. Underwater surveys were performed at 33 148

sites throughout the Midriffs during the spawning season of M. rosacea (April to June) in 149

2008, 2009, and 2010. Evidence of the formation of spawning aggregations were based 150

on standard protocols (Colin et al. 2003) and those adapted for leopard grouper (Erisman 151

et al. 2007b). Direct evidence of spawning aggregations included observations of 152

courtship or spawning behavior or the collection of females with hydrated or ovulated 153

oocytes. Indirect evidence involved observations of putative females with enlarged 154

abdomens indicative of imminent spawning, color patterns associated with courtship, the 155

collection of males with ripe testes, and abundances and densities of fish that were 156

markedly higher (e.g., 3-fold increases or greater) than observed during non-spawning 157

months. Additional indirect evidence of spawning aggregations was acquired through 158

interviews with commercial fishers at five fishing communities (Bahía de los Ángeles, 159

Bahía de Kino, Desemboque Seri, Puerto Libertad and Punta Chueca) during 2005 and 160

2006 (Moreno-Báez et al. 2012; Moreno-Baez et al. 2010). 161


PrePrin

ts

8

8

Spatial units (Fig. S1) were established to evaluate spatial connectivity by 162

combining physical and political boundaries, as well as local knowledge from fishers 163

(Moreno-Báez et al. 2012; Moreno-Baez et al. 2010). We incorporated coastline and 164

bathymetry developed by the National Geophysical Data Center 165

(http://www.ngdc.noaa.gov/mgg/shorelines/shorelines.html) and the marine protected 166

areas in Mexico (www.conanp.gob.mx), respectively. We used the spatial union function 167

to integrate the different boundaries and define the spatial units, under ArcGIS 10.1 168

(ESRI) with the Spatial Analyst Extension and Model Builder tools. The size of the 169

spatial units varied from 13 to 812 km2 (Fig. S1). 170

While the general spawning season for M. rosacea in the GC occurs from late 171

April to June (Erisman et al. 2007a), it was necessary to collect empirical data to narrow 172

the specific spawning season from the Midriff Islands region. We acquired gonad 173

samples of adult female leopard groupers (i.e. > 30 cm TL; Erisman et al. 2007a) from 174

commercial fishers on a monthly basis from December 2008 to June 2010. Fish were 175

captured by gill nets or handlines at various sites at or near San Pedro Martir and Tiburon 176

islands, and Bahia Kino (Fig. 1). We processed tissue taken from the central portion from 177

one gonad lobe for each sample using standard histological techniques (Humason 1972) 178

in order to determine sex and developmental stage. Classes of ovarian and testicular 179

development were adapted from previous studies (Erisman et al. 2007b), and stages of 180

gametogenesis followed previously established definitions (Wallace & Selman 1981). 181

We determined the duration of the spawning season using a combination of two 182

methods. First, we examined the histological preparations of all gonad samples to identify 183

the percentage of females capable of spawning or actively spawning during each month 184


PrePrin

ts

9

9

of sampling. Females categorized as ‘spawning capable’ included those with ovaries 185

dominated by oocytes in advanced stages of vitellogenesis (e.g., primary to tertiary yolk 186

stage), whereas those categorized as ‘actively spawning’ contained ovaries with oocytes 187

in the migratory nucleus, hydrated, or ovulated stage or post-ovulatory follicles were 188

present. Data were pooled by month to estimate the monthly proportion of spawning 189

females over a calendar year. Dates on which actively spawning females were collected 190

were used as indicators to determine exact dates of spawning. A second estimate of the 191

spawning season was obtained by calculating the mean monthly gonadosomatic index 192

(GSI = 100* gonad weight/ total body weight) of female M. rosacea over the study 193

period. Changes in monthly GSI were used to assess reproductive activity, associating 194

elevated levels with gonadal development and spawning. This information was used in 195

the release dates for larvae in the oceanographic model (see below). 196

197

Supporting fishing knowledge 198

The central component to documenting the fishing grounds for the M. rosacea was 199

captured through a series of interviews implemented across 17 fishing communities in the 200

northern GC (Moreno-Báez et al. 2010, Moreno-Báez et al., 2012). The methodology 201

entailed aggregating local knowledge of a representative set of individual fishers 202

(captains) through semi-structured interviews conducted between December 2005 and 203

July 2006 regarding what, where, when and how they fish. The interview included 204

questions regarding the spatial and temporal distribution of fishing activities but also, 205

their knowledge about spawning aggregations and juvenile sights. The maps were 206

digitized, georeferenced, and integrated into a geographic information systems (GIS) 207


PrePrin

ts

10

10

using ArcGIS 9.2 (ESRI, 1999 - 2008). 208

These interviews indicated that fishing activity frequently overlapped spatially with 209

spawning aggregation sites in three main regions: 1) the north end of Angel de la Guarda 210

Island; 2) on the south, western and northern edge of Tiburon Island, and on sites in 211

mainland Sonora north of Tiburon, around Las Cuevitas, Puerto Libertad and Puerto 212

Lobos (yellow areas Fig. 1). Other important fishing zones were identified around Puerto 213

Peñasco in northern Sonora. According to the interviews, the principal fishing season 214

starts in November and ends in June. 215

Oceanographic model 216

In a computer simulation exercise, four thousand particles were released at a 217

depth of 5 m at each of 17 spawning aggregation sites (see below). Particles were tracked 218

for 28 days, which is close to the maximum pelagic larval duration (PLD) for M. rosacea 219

(Aburto-Oropeza et al. 2007). We released particles during May 16th and May 25th of 220

2007, which based on our observations covers the peak spawning period around the 221

Midriff Islands region (see Results). Since seasonal oceanographic regimes are consistent 222

across years in the GC (Marinone 2003; Soria et al. 2013), the simulation year was 223

chosen arbitrarily (2007) while the dates covered spring and neap tides. The phase of the 224

springs-neaps cycle is simply shifted from year to year. 225

We used the velocity field from the GC implementation of the three-dimensional 226

baroclinic Hamburg Shelf Ocean Model (HAMSOM) (Backhaus 1985) to calculate the 227

particle trajectories. The model has been described in detail for the GC (Marinone 2003; 228

Marinone 2008). Its domain has a mesh size of 2.5′/3 × 2.5′/3 (~1.31 × 1.54 km) in the 229

horizontal and 12 layers in the vertical with nominal lower levels at 10, 20, 30, 60, 100, 230


PrePrin

ts

11

11

150, 200, 250, 350, 600, 1000 and 4000 m. The model equations are solved semi-231

implicitly with fully prognostic temperature and salinity fields, thus allowing time-232

dependent baroclinic motions. The model is started from rest with a 300 s time step and 233

becomes periodically stable after three years. Results for this study were obtained from 234

the fourth year of the model when it adequately reflects the main seasonal signals of 235

surface temperature, heat balance, tidal elevation and tidal currents and surface 236

circulation in the NGC (Lavin et al. 1997; Marinone 2003). The forcing includes at the 237

open boundary model tidal components (M2, S2, N2, K2, K1, O1, P1, Ssa, and the Sa), 238

climatological hydrography historical data and at the sea surface climatological heat and 239

fresh water fluxes. We used the seasonal climatology constructed from QUICKSCAT 240

data as forcing for wind. The Lagrangian trajectories are due to the Eulerian velocity field 241

plus a random-walk contribution related to turbulent eddy diffusion processes (Proehl et 242

al. 2005; Visser 1997). We obtained values of the diffusivities from the numerical model. 243

A pseudo-advective term was introduced, since the vertical diffusivity is not constant, to 244

prevent particles from walking away from areas of high to low diffusivities. The velocity 245

at each particle position and the vertical eddy coefficients are calculated by bilinear 246

interpolation of the instantaneous Eulerian velocity fields and the eddy coefficient from 247

the numerical model, which were saved every hour. The horizontal diffusivity is taken as 248

a constant (100 m2/s). We reasonably assumed that larvae are advected as passive 249

particles and do not migrate vertically downward to deep depths (Watson et al. 2010), 250

given that leopard grouper recruit to shallow Sargassum spp. beds of < 5 m deep 251

(Aburto-Oropeza et al. 2007). 252

253


PrePrin

ts

12

12

Modeled connectivity 254

Hourly latitude and longitude data for each modeled particle were imported into 255

MatLab (MATHWORKS). We estimated connectivity at different time intervals: 336 h 256

(14 days), 504 h (21 days) and 672 h (28 days) respectively after the released dates. 257

These PLDs were selected based on the average time of flexion in groupers that 258

corresponds to the onset of larval behavior (14 days) (Cowen 2002; Gracia-Lopez et al. 259

2005) and the maximum PLD (Aburto-Oropeza et al. 2007). A selection by location 260

function “inpolygon” was used to identify the intersection between particles and the 261

recruitment areas (spatial units). We then generated connectivity matrices using the 262

proportion of larvae that settled at each location relative to the total number of larvae 263

released at each site. We constructed matrices averaging for the two spawning dates May 264

16th and May 25th within each PLD (i.e. day 14, 21, 28). The probability of local retention 265

(i.e., diagonal in the connectivity matrix) was calculated as the proportion of particles 266

produced locally that remained within the spatial unit at the end of the PLD (Burgess et 267

al. 2013). The probabilities within each site were summarized with two statistics aimed at 268

describing source-sink dynamics. Export probability was defined as the proportion of 269

larvae produced within a site that successfully settled within any of the other 16 coastal 270

areas left at the end of the PLD. Import probability was defined as the proportion of all 271

larvae produced among the 17 sites that settled within each site. This later metric is 272

identical to self-recruitment as defined by Burgess et al. 2013. 273

Marine connectivity patterns were displayed using graph theory and a spatial 274

network approach (Treml et al. 2012) with the software NODEXL (Smith et al. 2010). 275

We calculated four statistics that describe the relationships among elements in complex 276


PrePrin

ts

13

13

networks (Newman 2003), including: 1) graph size (the total number of directed links 277

within a graph); 2) in-degree (number of links that enter a node); 3) out-degree (number 278

of links that leave a node); and 4) betweenness centrality, or the proportion of shortest 279

paths between all node pairs that pass through a particular node, which highlights ‘most 280

used’ dispersal pathways or stepping stones that act like gateways though which genes or 281

individuals have to pass to spread to other nodes, emphasizing key sites for 282

multigenerational connectivity (Andrello et al. 2013). Betweenness centrality can be 283

viewed as a measure of resilience by measuring how many paths will get longer when a 284

node is removed (Newman 2003). 285

286

Genetic connectivity 287

We collected tissue samples from the pectoral fins of M. rosacea from 10 sites 288

included in our modeling exercise around the Midriff Islands region (Fig. 1). Samples 289

were acquired in fish markets or directly from fishermen at harbors between 2009 and 290

2012 under IACUC protocol Berng1101. We interviewed both fish vendors and 291

fishermen to determine the approximate localities where fish were collected. 292

Immediately after collection, samples were stored in 95% ethanol and kept at -20ºC in the 293

laboratory. Genomic DNA was extracted using standard chloroform extraction protocols 294

(Sambrook et al. 1989). We amplified a 787 bp fragment of mitochondrial marker 295

cytochrome b using primers Gludgl and CB3H (Palumbi et al. 1991). Thermocycler 296

parameters were as follows: initial hold at 94°C/5 min, 35 cycles of 94°C/45 sec, 45°C/45 297

sec, 72°C/45 sec, with a final extension of 72°C/7 min. We developed species-specific 298

primers for M. rosacea (MYCROS Forward: TTCTCCCACTACCCTGATTC and 299


PrePrin

ts

14

14

MYCROS Reverse: TACGTAGGCTTGGATCATTG) to amplify a 726 bp fragment of 300

mitochondrial marker ATPase. Thermocycler parameters were as follows: initial hold at 301

94°C/5 min, 35 cycles of 94°C/30 sec, 54°C/30 sec, 72°C/30 sec, with a final extension 302

of 72°C/7 min. After purification of PCR products following ABI manufacturer’s 303

protocols (ABI, Perkin-Elmer), we sequenced clean PCR products on an ABI 3730xl 304

automated sequencer (Applied Biosystems, Foster City, CA). 305

We calculated molecular diversity indices including nucleotide diversity (π) and 306

haplotype diversity (h). We corrected haplotype diversity using CONTRIB (Petit et al. 307

1998) to account for differences in sample size between sites based on rarefaction to a 308

minimum sample size of n = 4. Theory predicts that genetic diversity levels observed 309

whiting sites is highly dependent upon the amount of migration from source populations 310

(Gaggiotti 1996), and that genetic diversity increases in sink sites that concentrate larvae 311

from multiple sources (Kool et al. 2011). To compare directly the ocean model with the 312

empirical genetic data, we performed simple linear regressions between corrected 313

haplotype diversity and in-degree, out-degree and betweenness centrality estimates for 314

networks at each PLD. To explore the possibility that patterns could be confounded if a 315

single outlier site if it would have been modeled wrong, we repeated the regressions 316

leaving out one site at a time. 317

Phylogenetic relationships among sequences were inferred from a haplotype 318

network based on pairwise differences between haplotypes generated using Arlequin 319

(Excoffier et al. 2005) and R software. To test for hierarchical population structure we 320

performed an Analysis of Molecular Variance (AMOVA) in Arlequin. AMOVA 321

significance was estimated using a permutation test of 10,000 replicates. The 10 sites 322


PrePrin

ts

15

15

were clustered into three regions: Baja Peninsula (La Ventana, La Poma and San 323

Francisquito), the Midriff Islands (San Pedro Martir, Salsipuedes, Datil, San Esteban) and 324

the Sonoran coast (Puerto Libertad, El Tecomate, Puerto Lobos, Puerto Libertad). Chi-325

squared analyses were concurrently performed using DnaSP version 5.10 (Librado & 326

Rozas 2009) to test for patterns of regional subdivision. Pairwise comparisons were 327

made between each location to assess patterns of genetic differentiation. 328

We evaluated three different migration models using Migrate-n 3.2.16 (Beerli & 329

Palczewski 2010). First, we tested an unrestricted full migration model between all 330

sampling localities. Next, we considered two models with three population sizes 331

comprised of a subsampling (n=30) from sampling localities in the Baja Peninsula, the 332

Midriff Islands, and the Sonoran coast. One model assessed unidirectional gene flow 333

from the Baja Peninsula, across the Midriff Islands, and northward up the Sonoran coast, 334

while the other model tested gene flow in the reverse direction. The latter two models 335

reflect seasonal differences in directionality of a cyclonic (May to September) and 336

anticlyclonic (October to April) gyres, respectively, present in the northern GC 337

(Marinone 2003; Marinone 2012). Using a Bezier approximation, we chose the most 338

appropriate model for our dataset by taking the natural log of the ratio of the marginal 339

likelihoods (Baye’s factors) for each model (Beerli & Palczewski 2010). Running 340

conditions for Migrate-n were as follows: 5,000,000 recorded steps, a burn-in of 341

2,500,000 steps, a static heating scheme using 20 temperatures, a tree swapping interval 342

of 1, and an upper prior boundary for migration set to 7,500. 343

344

345


PrePrin

ts

16

16

Results 346

Spawning sites, season and period 347

We chose seventeen sites representative of distinct spatial units for the release of 348

virtual larvae in the simulation model based on direct and indirect evidence of the 349

presence of spawning aggregations for leopard grouper (Table 1). Although some spatial 350

units had evidence of multiple spawning aggregations, we assumed their close proximity 351

along with the spatial resolution of the oceanographic model meant multiple aggregations 352

within the same unit would disperse larvae in similar directions. Individual sites were 353

distributed throughout the region and fulfilled a range of 3 to 7 criteria, with an average 354

of 4.76±1.48 (SD). A marked increase in the abundance of adult groupers during the 355

spawning season relative to the non-spawning season was the most common evidence 356

(recorded at all 17 sites). Other types of indirect evidence such as the observation of 357

gravid females with swollen abdomens, observations of fish exhibiting courtship 358

coloration, the collection of running-ripe males, or elevated catch rates by fishers during 359

the spawning season were recorded for the majority of sites. Direct evidence of spawning 360

via the collection of hydrated females was recorded for 65% of the sites, whereas 361

spawning was observed at only 35% of the sites. 362

A total of 162 samples of female M. rosacea were collected from commercial 363

fishers over the study period, with an average of 14 samples collected each month (range 364

= 8 to 27). Based on microscopic examinations of gonadal tissue samples, females in the 365

spawning capable phase were collected from March through June, and actively spawning 366

females were collected April to June. Similarly, the GSI of adult females showed 367

elevated levels from April to June, with a peak during May (Fig. 2). When the results of 368


PrePrin

ts

17

17

the gonadal phases and GSI were combined, they indicate that M. rosacea spawn from 369

April to June in the Midriffs region, with peak spawning activity occurring in May. 370

Actively spawning females were collected on three days in 2009 (14 May, 31 May, 25 371

June) and two days in 2010 (25 April, 7 May). 372

373

Modeled connectivity 374

From all simulated particles released (4,000 particles x 17 sites x two release 375

dates = 136,000), coastal areas that are suitable for larval recruitment captured 41.84% 376

(PLD 14 days), 35.24% (PLD 21 days) and 33.26% (PLD 28 days). Remaining particles 377

did not reach any coastal habitat by the end of the PLD. Simulations of ocean currents 378

produced the highest concentrations of larvae in coastal areas north of Tiburon Island 379

after PLD 28 days (around Las Cuevitas Fig. 3), followed by the south end of Tiburon 380

Island, Puerto Libertad, and the north end of Angel de la Guarda Island. Simulations at 381

PLD 14 and 21 days indicated similar trends (Fig. 3). 382

The trajectories of particles released from each site at the two dates (Fig. 4a and 383

4b) showed sites in the Baja California Peninsula generally followed a southward 384

direction (except those released from La Ventana that entered the Canal de Ballenas), 385

while sites on mainland Sonora and the north edge of Tiburon Island followed a 386

northward trajectory. Most locations around Angel de la Guarda Island and Tiburon 387

Island had particles dispersing north and south of the release site. In all cases, the 388

distance traveled by particles was directly proportional to the PLD. The graph size of the 389

connectivity networks increased from 38 edges at PLD 14 days, to 59 at PLD 21 days and 390

67 at PLD 28 days, indicating a longer PLD is associated with more connected and 391


PrePrin

ts

18

18

complex networks (Fig. 5). Although few differences were observed between networks 392

derived from the two release dates (see Fig. S2 for an example at PLD 28 days), 393

connectivity patterns showed similar trends. Based on the directionality of the links in the 394

networks, three main patterns were evident. First, northward links were prevalent along 395

the eastern coast of the GC, in the Canal de Ballenas and between the southern end of 396

Angel de la Guarda Island and San Lorenzo Island across the GC towards the northern 397

end of Tiburon Island and mainland Sonora (Fig. 5). Second, southward links were 398

present between the eastern coast of Angel de la Guarda Island towards southern 399

locations in San Lorenzo Island and Baja California and across the GC to Tiburon Island 400

and San Esteban Island. Third, bi-directional north-south links were evident only in a 401

small area located between San Pedro Martir Island, San Esteban Island and the southern 402

end of Tiburon Island. Overall, the strongest links (i.e., those showing the larger 403

probabilities) were observed between the western and northern coasts of Tiburon Island 404

towards northern localities situated in mainland Sonora and in the Canal de Ballenas (Fig 405

5). 406

The probability of local retention decreased with increasing PLD, except in Datil 407

(Fig. 6). According to the ocean model, local retention was most likely in Puerto Libertad 408

(range: 0.26 to 0.72 for PLD 28 and 14 days, respectively), the northern end of Tiburon 409

Island (El Tecomate, range: 0.21 to 0.69), followed by the northern end of Angel de la 410

Guarda Island (Punta Refugio, range: 0.28 to 0.46), the western coast of Tiburon Island 411

(La Tordilla, range: 0.07 to 0.35), and the southern coast of Tiburon Island (Datil Island, 412

range: 0.13 to 0.24). With one exception (Puerto Lobos at PLD 14), all other sites had 413

probabilities below 0.07. 414


PrePrin

ts

19

19

Our analyses about the performance of each site on the network using statistics 415

describing the probabilities of local retention, export/import and the number of 416

connections leaving and entering each site (degree) illustrated the spatial overlap of some 417

criteria, highlighting a few sites that were important despite variation in PLD (Fig. 6, 418

Table S1-S3). For instance, Tecomate on the north of Tiburon Island and Puerto Libertad 419

on mainland Sonora had the largest probabilities of export, import and local retention. 420

However, both sites exported larvae to only a few sites and showed relatively low 421

betweenness centrality. The north end of Angel de la Guarda Island (Puerto Refugio) also 422

had relatively large probability of import and local-retention, but low export probability 423

to few sites. Some sites acted exclusively as sources (e.g., La Ventana) with high export 424

probability to many sites despite low local retention and low import probabilities. Datil 425

Island was identified as a sink with intermediate levels of larvae imported from many 426

sites but relatively low export probabilities. Betweenness centrality was more sensitive 427

to PLD variation than the other measures, and identified Las Cuevitas on mainland 428

Sonora, the Southern end of Angel de la Guarda Island and La Ventana as key sites for 429

multigenerational larval dispersal through the entire network. 430

431

Genetic connectivity 432

We analyzed a 787 bp fragment of cytochrome b and a 726 bp fragment of 433

ATPase for 235 individuals. We identified a total of 77 haplotypes, with adjacent 434

haplotypes in the haplotype network separated by 1 to 4 bp (Fig. 7). There was limited 435

evidence of geographic separation of haplotypes, and the three most frequent haplotypes 436

were present in all locations, with the exception of the third most frequent which was 437


PrePrin

ts

20

20

absent in San Francisquito on the Baja Peninsula. Corrected estimates of haplotype 438

diversity were high, ranging from 0.844 (San Lorenzo Island) to 0.943 (San Esteban 439

Island) (Table 2). Linear regression models were unable to find significant correlations 440

between empirical observations (corrected haplotype diversity) and modeled connectivity 441

measured with in-degree, out-degree and betweenness centrality at any PLD (all R2 < 442

0.147, P values > 0.550, Table S1-S3). However, upon elimination of one site at a time, 443

we observed two significant trends when San Lorenzo Island was excluded from the 444

analyses (Fig. 8). First, genetic diversity was significantly higher at sites with larger in-445

degree values estimated from the model at PLD 14 days (R2 = 0.489, P = 0.035, Fig. 8a). 446

Second, genetic diversity was significantly lower at locations showing higher 447

betweenness centrality according to the model at PLD 28 days (R2 = 0.471, P = 0.041, 448

Fig. 8b). All other comparisons leaving out one site were not significant (R2 < 0.170, P > 449

0.175). 450

Statistically significant pairwise estimates of genetic structure were observed 451

(Table 3). Pairwise φST values suggest Puerto Libertad is genetically divergent from the 452

majority of other sampling localities. Pairwise FST values suggest greater overall genetic 453

differentiation between most sampling localities, except between San Esteban Island and 454

all the other sites. Global estimates of FST and φST suggest moderate levels of population 455

structure within the northern Gulf (φST = 0.04663, P = 0.0001; FST = 0.10836, P < 456

0.00001). Regional genetic subdivision was also observed when sampling sites were 457

clustered into the following groups – Baja Peninsula, Midriff Islands and the Sonoran 458

coast. Genetic subdivision of regional groups was supported by a chi-squared test (χ2 = 459

186.876, d.f. = 152, P = 0.0286). Rankings of proposed larval dispersal models between 460


PrePrin

ts

21

21

the aforementioned regions are listed in Table 4. The best-supported model was for 461

unidirectional larval dispersal from the Baja Peninsula to the mainland. There was minor 462

support for the full, unrestricted larval dispersal model and almost no support for 463

unidirectional larval dispersal from the mainland to the Baja Peninsula. 464

465

Discussion 466

Our study contribute to a growing body of literature (Alberto et al. 2011; Crandall 467

et al. 2012; Feutry et al. 2013; Foster et al. 2012; Galindo et al. 2010; Petitgas et al. 2012; 468

Selkoe et al. 2010; Soria et al. 2012) highlighting the inherent value of verifying outputs 469

of biophysical oceanographic models with empirical genetic data to inform larval 470

dispersal patterns and marine connectivity. Concordance of genetic and biophysical 471

modeling data for M. rosacea elucidate the role of oceanographic processes in driving 472

patterns of larval dispersal, while models helped to explain seemingly chaotic patterns of 473

genetic diversity and structure. The best supported gene flow model (based on genetic 474

subdivisions) confirmed biophysical model outputs demonstrating a pulse of larvae from 475

the Baja Peninsula, across the GC and northward up the Sonoran coastline, in agreement 476

with the cyclonic gyre present from May to September. Our modeled connectivity 477

networks mirrored these results. Furthermore, our study demonstrates that patterns of 478

oceanographic circulation in the GC may be a powerful predictor of source-sink 479

dynamics for M. rosacea. In our dataset we found support from empirical genetic data 480

that genetic diversity increases in sink sites that concentrate larvae from many sources. 481

However, the fact that patterns were significant only for PLD 14 days supports that larvae 482

reaching suitable habitat are already able to settle early at the average time of flexion in 483


PrePrin

ts

22

22

groupers. Our results could suggest that biophysical models in combination with graph 484

theory could be used as a proxy for predicting genetic diversity, but further studies are 485

needed to verify the validity of this relationship across markers and species. For instance, 486

San Esteban Island, the site with the largest number of incoming sources of larvae at PLD 487

14 days and with the largest diversity of haplotypes, showed the lowest levels of genetic 488

differentiation according to pairwise statistics, likely as a result of high levels of gene 489

flow towards this site. In contrast, Puerto Libertad on mainland Sonora showed the 490

largest genetic differences compared to other sites, which could be explained by a higher 491

proportion of kin than expected by chance due to high levels of local retention (Iacchei et 492

al. 2013), as supported by the model regardless of PLD. By coupling modeled and 493

empirical connectivity approaches, we are able to better understand the mechanisms 494

driving dispersal in the Gulf and to potentially inform spatially explicit management 495

efforts for M. rosacea as well as marine organisms with similar life histories. However, 496

validation of the passive dispersal model through subsequent studies such as those that 497

use parentage analyses and highly polymorphic microsatellite loci are recommended and 498

underway. 499

Our study had several limitations. Below we discuss some of the major ones. The 500

relationship between patterns of genetic connectivity and modeled larval connectivity 501

was significant only after removing San Lorenzo Island, where, according to observed 502

levels of genetic variation, our ocean model seemed to have overestimated connectivity 503

in terms of the number of sources supplying larvae to that site. San Lorenzo is a relatively 504

thin and small island, and a detailed examination of the connectivity matrices indicated 505

that the scenario where most of the larvae arrive from multiple sites to this one, has very 506


PrePrin

ts

23

23

small probabilities (< 0.01). These observations suggest that the large size of the spatial 507

unit in our study, relative to the small size of the island, coupled with a large bathymetric 508

profile around the island, likely biased real connectivity estimates, indicating room for 509

improving the spatial resolution of our analyses. While there is no direct evidence that 510

groupers larvae vertically migrate downward to escape advecting currents, there is a 511

growing body of evidence that suggests that local retention of larvae in groupers can be 512

quite high (Almany et al. 2013; Harrison et al. 2012). Thus our passive model could have 513

overestimated larval exchange rates and underestimated local retention (Cowen 2000). 514

However, the effects of vertical migration are comparable to those of reducing PLD 515

(Andrello et al. 2013). Investigations into larval behavior of groupers are warranted and 516

could greatly increase the precision and accuracy of the model. Our models did not 517

include an explicit description of the habitat for larval recruitment (Sargassum sp. beads), 518

neither considered larval mortality after settlement, thus we only assessed potential 519

connectivity, as opposed to realized connectivity. 520

Our multidisciplinary approach provided a mechanistic explanation of why some 521

areas in the Midriff Island region concentrate the fishing effort for leopard grouper in the 522

GC. Heavily targeted fished areas, including the north end of Angel de la Guarda Island, 523

the west, north and south edges of Tiburon Island and Las Cuevitas and Puerto Libertad 524

on mainland Sonora, showed the largest values of local retention of larvae, together with 525

a high probability of importing larvae from other spawning sites and for concentrating 526

larvae from all over the region. Notably, some of these are known to be sites that 527

historically have held huge spawning aggregations of leopard grouper that have been 528

harvested at high levels for decades, like the north end of Angel de la Guarda Island 529


PrePrin

ts

24

24

(Cannon 1966). Thus, the main fishing areas seem to depend both on local retention and 530

contributions of larvae from upstream sites, coupled to oceanographic patterns that focus 531

larval density towards these areas that sustain most of the fisheries. 532

A key result of our study is the observation that marine connectivity for M. 533

rosacea around the Midriff island region is predominantly asymmetric. Other studies 534

have previously shown the negative effects that asymmetric connectivity has on 535

population persistence (Bode et al. 2008; Vuilleumier et al. 2010). In the presence of 536

strong asymmetric currents, reserves can significantly outperform traditional quota based 537

management strategies in terms of fisheries yield, with considerably less risk (Gaines et 538

al. 2003). Asymmetry also constrains the notion that benefits of reserves in terms of 539

larval input are proportional to their distance to the reserve (Almany et al. 2009; Buston 540

et al. 2012). For example, one study using DNA parentage analyses found that reserves in 541

the Great Barrier Reef, which accounted for 28% of the local reef area, produced 542

approximately half of all juvenile recruitment of snappers and groupers to both reserve 543

and fished reefs within 30 km of the source spawning site inside the reserve (Harrison et 544

al. 2012), while a similar study in Papua New Guinea found that 50% of larvae in a coral 545

grouper settled within 14 km of the spawning aggregation sites (Almany et al. 2013). In 546

contrast, the benefits of reserves are completely biased towards one particular direction in 547

the GC, highlighting that the spatial location of no-take zones is even more important in 548

the Midriff Island region than in other systems. An exception to the general asymmetry in 549

connectivity was detected in small area between the south edge of Tiburon Island, San 550

Esteban Island, and San Pedro Martir Island. 551


PrePrin

ts

25

25

A network of no-take zones within the Midriff region might have a very well 552

defined zone of influence that does not include the eastern edge of Tiburon Island or any 553

locality towards the south in mainland Sonora. This observation has important practical 554

implications. For example, fishing localities on mainland Sonora South of Tiburon Island 555

are restricted from fishing at no-take areas within MPAs in the Midriffs, yet according to 556

this an other studies (Soria et al. 2013) they receive no benefit by not fishing there. 557

Conversely, fishing communities in mainland Sonora (Puerto Lobos, Puerto Libertad) 558

seem to receive great benefits from San Pedro Martir, San Esteban and Tiburon Islands, 559

even though they may not fish there. This brings up an important concept in highly 560

advective systems like the GC where there may be a spatial disconnect and strong 561

directionality between the location of no-take zones and the areas that benefit most from 562

them, and highlights that, in order of reserves to be effective, they need to be located 563

upstream of targeted fishing sites (Beger et al. 2010). Our analyses suggest that 564

establishment of smaller no-take zones at the north end of Angel de la Guarda Island 565

within the current MPA will likely boost local fisheries via local retention, while other 566

current no-take zones within the Canal de Ballenas and San Lorenzo MPA could export 567

larvae to fishing sites across the GC. The additional establishment of no-take zones 568

adjacent to current heavily fished areas in the western and northern edges of Tiburon 569

Island, and in the coast between Las Cuevitas-Puerto Lobos will likely increase not only 570

local fisheries (via local retention) but fisheries at downstream fished sites on mainland 571

Sonora as north as Puerto Peñasco (located ~300 km from Tiburon Island) via larval 572

dispersal. Notably, except for San Francisquito on the coast of Baja California, current 573

MPAs do not include those sink sites receiving larvae from multiple sources and that 574


PrePrin

ts

26

26

harbor the largest genetic diversity and evolutionary potential (San Esteban and Tiburon 575

Islands). In other hand, the fact that key gateway routes for multigenerational dispersal 576

showing high betweenness centrality at PLD 28 days had low genetic variation (e.g. La 577

Ventana on Baja California, Las Cuevitas on mainland Sonora), stress that important sites 578

for long-term metapopulation persistence and resilience are not necessarily aligned 579

spatially with other criteria for protection, such as preserving evolutionary potential via 580

genetic variation. 581

582

Acknowledgements 583

We would like to acknowledge Rene Loiaza, Ivan Martinez, Angeles Cruz, Mario Rojo, 584

Kimberly Tenggardjaja, Eva Salas, Gary Longo, Emmett Ziegler and Ana Liedke for 585

their assistance with acquiring samples in the field. This a scientific contribution of the 586

multidisciplinary PANGAS project (www.pangas.arizona.edu). 587

588

ADDITIONAL INFORMATION AND DECLARATIONS 589

590

Funding 591

This work was funded by the David and Lucile Packard Foundation, National Geographic 592

(Young Explorers Grant #8928-11), the University of California Office of the President 593

(UC-HBCU Initiative grant), Marine Conservation Institute (Mia J. Tegner memorial 594

research grant), Ecology Project International, CenTread, the Dr. Earl H. Myers and Ethel 595

M. Myers Oceanographic and Marine Biology Trust, FCT (SFRH/BPD/26901/2006) and 596

the Friends of Long Marine Lab. Additional funding for this research was provided by 597


PrePrin

ts

27

27

Dr. Exequiel Ezcurra through the Pew Fellowship Program on Marine Conservation, and 598

several grants from the Walton Family Foundation. 599

600

Grant Disclosures 601

The following grant information was disclosed by the authors: 602

The David and Lucile Packard Foundation (grant award #2008- 32210, #2013-39400) 603

University of California Office of the President (UC-HBCU Initiative grant) 604

Marine Conservation Institute (Mia J. Tegner memorial research grant) 605

Ecology Project International 606

CenTread 607

Dr. Earl H. Myers and Ethel M. Myers Oceanographic and Marine Biology Trust 608

FCT (SFRH/BPD/26901/2006) 609

Friends of Long Marine Lab 610

Pew Fellowship Program on Marine Conservation 611

Walton Family Foundation 612

613

Competing Interests 614

The authors declare that they have no competing interests. 615

616

Author Contributions 617

Adrian Munguia-Vega, Brad Erisman, Alexis Jackson, Jorge Torre and Tad Pfister 618

conceived and designed the experiments. 619


PrePrin

ts

28

28

Adrian Munguia-Vega, Alexis Jackson, Silvio Guido Marinone, Marcia Moreno-Baez 620

and Alfredo Giron perfomed the experiments and analyzed data. 621

Adrian Munguia-Vega, Brad Erisman, Alexis Jackson, Octavio Aburto-Oropeza and 622

Jorge Torre wrote the paper. 623

624

Field Study Permissions 625

Scientific collection permits were acquired from SAGARPA (No. 626

DGOPA.09151.260809.2885 and DAPA/2/020511/01197). 627

628

629

Supplemental Information 630

Table S1 Statistics describing the dynamics of larval dispersal in a network of 17 sites 631

after PLD 14 days. 632





Figure S1 Spatial units of analyses for studying connectivity in the Gulf of California. 637

Figure S2 Modeled networks of larval connectivity for PLD 28 days for larvae released 638

at two distinct dates. 639

640

641

642


PrePrin

ts

29

29

References 643 644

Aburto-‐Oropeza O, Erisman B, Galland GR, Mascarenas-‐Osorio I, Sala E, and Ezcurra 645 E. 2011. Large Recovery of Fish Biomass in a No-‐Take Marine Reserve. PloS 646 one 6:e23601. 647

Aburto-‐Oropeza O, Sala E, Paredes G, Mendoza A, and Ballesteros E. 2007. 648 Predictability of reef fish recruitment in a highly variable nursery habitat. 649 Ecology 88:2220-‐2228. 650

Alberto F, Raimondi PT, Reed DC, Watson JR, Siegel DA, Mitarai S, Coelho N, and 651 Serrao EA. 2011. Isolation by oceanographic distance explains genetic 652 structure for Macrocystis pyrifera in the Santa Barbara Channel. Molecular 653 ecology 20:2543-‐2554. 654

Almany GR, Connolly SR, Heath DD, Hogan JD, Jones GP, McCook LJ, Mills M, Pressey 655 RL, and Williamson DH. 2009. Connectivity, biodiversity conservation and 656 the design of marine reserve networks for coral reefs. Coral Reefs 28:339-‐657 351. 658

Almany GR, Hamilton RJ, Bode M, Matawai M, Potuku T, Saenz-‐Agudelo P, Planes S, 659 Berumen ML, Rhodes KL, Thorrold SR et al. . 2013. Dispersal of grouper 660 larvae drives local resource sharing in a coral reef fishery. Current biology : 661 CB 23:626-‐630. 662

Andrello M, Mouillot D, Beuvier J, Albouy C, Thuiller W, and Manel S. 2013. Low 663 Connectivity between Mediterranean Marine Protected Areas: A Biophysical 664 Modeling Approach for the Dusky Grouper Epinephelus marginatus. PloS one 665 8:e68564. 666

Backhaus JO. 1985. A three-‐dimensional model for the simulation of the shelf sea 667 dynamics. Dtsch Hydrogr 38:165-‐187. 668

Beerli P, and Palczewski M. 2010. Unified framework to evaluate panmixia and 669 migration direction among multiple sampling locations. Genetics 185:313-‐670 326. 671

Beger M, Linke S, Watts M, Game E, Treml E, Ball I, and Possingham HP. 2010. 672 Incorporating asymmetric connectivity into spatial decision making for 673 conservation. Conservation Letters 3:359-‐368. 674

Beger M, Selkoe KA, Treml EA, Barber PH, von der Heyden S, Crandall ED, Toonen RJ, 675 and Riginos C. 2013. Evolving coral reef conservation with genetic 676 information. Bulletin of Marine Science 90. 677

Bode M, Burrage K, and Possingham HP. 2008. Using complex network metrics to 678 predict the persistence of metapopulations with asymmetric connectivity 679 patterns. Ecological Modelling 214:201-‐209. 680

Burgess SC, Nickols KJ, Griesemer CD, Barnett LAK, Dedrick AG, Satterthwaite EV, 681 Yamane L, Morgan SG, White JW, and Botsford LW. 2013. Beyond 682 connectivity: how empirical methods can quantify population persistence to 683 improve marine protected area design. Ecological Applications In-‐press. 684

Buston PM, Jones GP, Planes S, and Thorrold SR. 2012. Probability of successful 685 larval dispersal declines fivefold over 1 km in a coral reef fish. Proceedings 686 Biological sciences / The Royal Society 279:1883-‐1888. 687

Cannon R. 1966. The Sea of Cortez: Lane Magazine and Book Company. 688


PrePrin

ts

30

30

CNP. 2012. Carta Nacional Pesquera. Secretaria de Agricultura, Ganaderia, Desarrollo 689 Rural, Pesca y Alimentacion 236 pp:available from: 690 http://www.inapesca.gob.mx. 691

Colin PL, Sadovy YJ, and Domeier ML. 2003. Manual for the study and conservation 692 of reef fish spawning aggregations. Society for the Conservation of Reef Fish 693 Aggregations, Special Publication 1:1-‐99. 694

Costello C, Ovando D, Hilborn R, Gaines SD, Deschenes O, and Lester SE. 2012. Status 695 and Solutions for the World's Unassessed Fisheries. Science 338:517-‐520. 696

Cowen RK. 2000. Connectivity of Marine Populations: Open or Closed? Science 697 287:857-‐859. 698

Cowen RK. 2002. Larval dispersal and retention and consequences for population 699 connectivity. In: Sale PF, ed. Ecology of Coral Reef Fishes: Recent Advances. San 700 Diego, CA: Academic Press, 149-‐170. 701

Cowen RK, and Sponaugle S. 2009. Larval Dispersal and Marine Population 702 Connectivity. Annual Review of Marine Science 1:443-‐466. 703

Craig MT, Sadovy de Mitcheson YJ, and Heemstra PC. 2012. Groupers of the World: a 704 field and market guide: Taylor & Francis Group. 705

Craig MT, and Sadovy Y. 2008. Mycteroperca rosacea. IUCN 2012 IUCN Red List of 706 Threatened Species Version 20122 <wwwiucnredlistorg> Downloaded on 16 707 June 2013. 708

Crandall ED, Treml EA, and Barber PH. 2012. Coalescent and biophysical models of 709 stepping-‐stone gene flow in neritid snails. Molecular ecology 21:5579-‐5598. 710

Cudney-‐Bueno R, Bourillón L, Sáenz-‐Arroyo A, Torre-‐Cosío J, Turk-‐Boyer P, and 711 Shaw WW. 2009. Governance and effects of marine reserves in the Gulf of 712 California, Mexico. Ocean & Coastal Management 52:207-‐218. 713

Diaz-‐Uribe JG, Elorduy-‐Garay JF, and Gonzalez-‐Valdovinos MT. 2001. Age and 714 Growth of the Leopard Grouper, Mycteroperca rosacea, in the Southern Gulf 715 of California, Mexico. Pacific Science 55:171-‐182. 716

Erisman BE, Buckhorn ML, and Hastings PA. 2007a. Spawning patterns in the 717 leopard grouper, Mycteroperca rosacea, in comparison with other 718 aggregating groupers. Marine Biology 151:1849-‐1861. 719

Erisman BE, Rosales-‐Casián JA, and Hastings PA. 2007b. Evidence of gonochorism in 720 a grouper, Mycteroperca rosacea, from the Gulf of California, Mexico. 721 Environmental Biology of Fishes 82:23-‐33. 722

Excoffier L, Laval G, and Schneider S. 2005. Arlequin (version 3.0): An integrated 723 software package for population genetics data analysis. Evolutionary 724 Bioinformatics 1:47-‐50. 725

Feutry P, Vergnes A, Broderick D, Lambourdiere J, Keith P, and Ovenden JR. 2013. 726 Stretched to the limit; can a short pelagic larval duration connect adult 727 populations of an Indo-‐Pacific diadromous fish (Kuhlia rupestris)? Molecular 728 ecology 22:1518-‐1530. 729

Foster NL, Paris CB, Kool JT, Baums IB, Stevens JR, Sanchez JA, Bastidas C, Agudelo C, 730 Bush P, Day O et al. . 2012. Connectivity of Caribbean coral populations: 731 complementary insights from empirical and modelled gene flow. Molecular 732 ecology 21:1143-‐1157. 733


PrePrin

ts

31

31

Gaggiotti OE. 1996. Population Genetic Models of Source-Sink Metapopulations. 734 Theoretical Population Biology 50:178-‐208. 735

Gaines SD, Gaylord B, and Largier JL. 2003. Avoiding current oversights in marine 736 reserve design. Ecological Applications 13:S32-‐S36. 737

Gaines SD, White C, Carr MH, and Palumbi SR. 2010. Designing marine reserve 738 networks for both conservation and fisheries management. Proceedings of 739 the National Academy of Sciences of the United States of America 107:18286-‐740 18293. 741

Galindo HM, Pfeiffer-‐Herbert AS, McManus MA, Chao Y, Chai F, and Palumbi SR. 742 2010. Seascape genetics along a steep cline: using genetic patterns to test 743 predictions of marine larval dispersal. Molecular ecology 19:3692-‐3707. 744

Gracia-‐Lopez V, Kiewek-‐Martinez M, Maldonado-‐Garcia M, Monsalvo-‐Spencer P, 745 Portillo-‐Clark G, Civera-‐Cerecedo R, Linares-‐Aranda M, Robles-‐Mungaray M, 746 and Mazon-‐Suastegui JM. 2005. Larvae and juvenile production of the 747 leopard grouper, Mycteroperca rosacea (Streets, 1877). Aquaculture 748 Research 36:110-‐112. 749

Harrison HB, Williamson DH, Evans RD, Almany GR, Thorrold SR, Russ GR, Feldheim 750 KA, van Herwerden L, Planes S, Srinivasan M et al. . 2012. Larval export from 751 marine reserves and the recruitment benefit for fish and fisheries. Current 752 biology : CB 22:1023-‐1028. 753

Hastings A, and Botsford LW. 2006. Persistence of spatial populations depends on 754 returning home. Proceedings of the National Academy of Sciences of the United 755 States of America 103:6067-‐6072. 756

Hastings PA, Findley LT, and Van der heiden AMV. 2010. Fishes of the Gulf of 757 California. In: Brusca RC, ed. The Gulf of California Biodiversity and 758 Conservation. Tucson, AZ: The University of Arizona Press, 96-‐118. 759

Humason GL. 1972. Animal tissue techniques, 3rd edn. San Francisco: WH Freeman 760 and Company. 761

Iacchei M, Ben-‐Horin T, Selkoe KA, Bird CE, Garcia-‐Rodriguez FJ, and Toonen RJ. 762 2013. Combined analyses of kinship and FST suggest potential drivers of 763 chaotic genetic patchiness in high gene-‐flow populations. Molecular ecology 764 22:3476-‐3494. 765

Kool JT, Paris CB, Barber PH, and Cowen RK. 2011. Connectivity and the 766 development of population genetic structure in Indo-‐West Pacific coral reef 767 communities. Global Ecology and Biogeography 20:695-‐706. 768

Lavin MF, Durazo R, Palacios E, Argote ML, and Carrillo L. 1997. Lagrangian 769 observations of the circulation in the northern Gulf of California. Journal of 770 Physical Oceanography 27:2298-‐2305. 771

Librado P, and Rozas J. 2009. DnaSP v5: a software for comprehensive analysis of 772 DNA polymorphism data. Bioinformatics (Oxford) 25:1451-‐1452. 773

Marinone SG. 2003. A three-‐dimensional model of the mean and seasonal circulation 774 of the Gulf of California. Journal of Geophysical Research 108:1-‐27. 775

Marinone SG. 2008. On the three-‐dimensional numerical modeling of the deep 776 circulation around Ángel de la Guarda Island in the Gulf of California. 777 Estuarine, Coastal and Shelf Science 80:430-‐434. 778


PrePrin

ts

32

32

Marinone SG. 2012. Seasonal surface connectivity in the Gulf of California. Estuarine, 779 Coastal and Shelf Science 100:133-‐141. 780

Moreno-‐Báez M, Cudney-‐Bueno R, Orr BJ, Shaw WW, Pfister T, Torre-‐Cosio J, Loaiza 781 R, and Rojo M. 2012. Integrating the spatial and temporal dimensions of 782 fishing activities for management in the northern Gulf of California, mexico. 783 Ocean & Coastal Management 55:111-‐127. 784

Moreno-‐Baez M, Orr BJ, Cudney-‐Bueno R, and Shaw WW. 2010. Using fishers local 785 knowledge to aid management at regional scales: spatial distribution of 786 small-‐scale fisheries in the Northern Gulf of California, Mexico. Bulletin of 787 Marine Science 86:339-‐353. 788

Newman MEJ. 2003. The Structure and Function of Complex Networks. SIAM Review 789 45:167-‐256. 790

Palumbi SR, Martin A, Romano S, McMillan WO, Stice L, and Grabowski G. 1991. The 791 Simple Fool's Guide to PCR, Version 2.0. Honolulu, HI: Privately published 792 document. 793

Petit RJ, El Mousadik A, and Pons O. 1998. Identifying populations for conservation 794 on the basis of genetic markers. Conservation Biology 12:844-‐855. 795

Petitgas P, Alheit J, Peck MA, Raab K, Irigoien X, Huret M, van der Kooij J, Pohlmann 796 T, Wagner C, Zarraonaindia I et al. . 2012. Anchovy population expansion in 797 the North Sea. Marine Ecology Progress Series 444:1-‐13. 798

Proehl JA, Lynch DR, McGillicuddy DJ, and Ledwell JR. 2005. Modeling turbulent 799 dispersion on the North Flank of Georges Bank using Lagrangian Particle 800 Methods. Continental Shelf Research 25:875-‐900. 801

Robertson DR, and Cramer KL. 2009. Shore fishes and biogeographic subdivisions of 802 the Tropical Eastern Pacific. Marine Ecology Progress Series 380:1-‐17. 803

Sala E, Aburto-‐Oropeza O, Paredes G, and Thompson G. 2003. Spawning 804 aggregations and reproductive behavior of reef fishes in the Gulf of 805 California. Bulletin of Marine Science 72:103-‐121. 806

Sala E, Aburto-‐Oropeza O, Reza M, Paredes G, and López-‐Lemus LG. 2004. Fishing 807 Down Coastal Food Webs in the Gulf of California. Fisheries 29:19-‐25. 808

Sale PF, Cowen RK, Danilowicz BS, Jones GP, Kritzer JP, Lindeman KC, Planes S, 809 Polunin NV, Russ GR, Sadovy YJ et al. . 2005. Critical science gaps impede use 810 of no-‐take fishery reserves. Trends in ecology & evolution 20:74-‐80. 811

Sambrook J, Fritsch EF, and Maniatis T. 1989. Molecular Cloning: A Laboratory 812 Manual Second Edition Vols. 1, 2, and 3. Molecular Cloning: A Laboratory 813 Manual, Second Edition, Vols 1, 2 and 3 Cold Spring Harbor, New York: Cold 814 Spring Harbor Laboratory Press. 815

Selkoe KA, Watson JR, White C, Horin TB, Iacchei M, Mitarai S, Siegel DA, Gaines SD, 816 and Toonen RJ. 2010. Taking the chaos out of genetic patchiness: seascape 817 genetics reveals ecological and oceanographic drivers of genetic patterns in 818 three temperate reef species. Molecular ecology 19:3708-‐3726. 819

Smith M, Milic-‐Frayling N, Shneiderman B, Mendes-‐Rodriguez E, Leskovec J, and 820 Dunne C. 2010. NodeXL: a free and open network overview, discovery and 821 exploration add-‐in for Excel 2007/2010 http://nodexlcodeplexcom. 822


PrePrin

ts

33

33

Soria G, Munguía-‐Vega A, Marinone SG, Moreno-‐Báez M, Martínez-‐Tovar I, and 823 Cudney-‐Bueno R. 2012. Linking bio-‐oceanography and population genetics to 824 assess larval connectivity. Marine Ecology Progress Series 463:159-‐175. 825

Soria G, Torre-‐Cosio J, Munguia-‐Vega A, Marinone SG, Lavín MF, Cinti A, and Moreno-‐826 Báez M. 2013. Dynamic connectivity patterns from an insular marine 827 protected area in the Gulf of California. Journal of Marine Systems 828 10.1016/j.jmarsys.2013.06.012. 829

Thomson DA, Findley LT, and Kerstitch AN. 2000. Reef fishes of the Sea of Cortez. 830 Austin, TX: University of Texas Press. 831

Treml EA, Roberts JJ, Chao Y, Halpin PN, Possingham HP, and Riginos C. 2012. 832 Reproductive output and duration of the pelagic larval stage determine 833 seascape-‐wide connectivity of marine populations. Integrative and 834 Comparative Biology 52:525-‐537. 835

Visser AW. 1997. Using random walk models to simulate the vertical distribution of 836 particles in a turbulent water column. Marine Ecology Progress Series 837 158:275-‐281. 838

Vuilleumier S, Bolker BM, and Leveque O. 2010. Effects of colonization asymmetries 839 on metapopulation persistence. Theoretical Population Biology 78:225-‐238. 840

Wallace RA, and Selman K. 1981. Cellular and dynamic aspects of oocyte growth in 841 teleosts. American Zoology 21:325-‐343. 842

Watson JR, Mitarai S, Siegel DA, Caselle JE, Dong C, and McWilliams JC. 2010. 843 Realized and potential larval connectivity in the Southern California Bight. 844 Marine Ecology Progress Series 401:31-‐48. 845

White JW, Botsford LW, Hastings A, and Largier JL. 2010. Population persistence in 846 marine reserve networks: incorporating spatial heterogeneities in larval 847 dispersal. Marine Ecology Progress Series 398:49-‐67. 848

849 850

851

852

853

854

855

856

857

858

859


PrePrin

ts

34

34

Tables 860

861

Table 1 Study sites in the Gulf of California and selection criteria. Each site was selected 862

based on seven criteria to define where spawning aggregations might act like source of 863

larvae. 864

865

866

867

868

Site

# Site Name

High

Abundance

of Fish

Elevated

Catch Rates

Hydrated

Females

Collected

Running-

Ripe Males

Collected

Courtship or

Spawning

Observed

Gravid

Females

Observed

Courtship

Coloration

Observed

# Criteria

observed

1 Punta Refugio ✔ ✔ ✔ ✔ ✔ ✔ ✔ 7

2 Punta Diablo ✔ ✔ ✔ ✔ 4

3 La Ventana ✔ ✔ ✔ ✔ ✔ ✔ ✔ 7

4 Chorros ✔ ✔ ✔ ✔ ✔ 5

5 San Pedro Martir Island ✔ ✔ ✔ ✔ ✔ ✔ 6

6 San Lorenzo Island ✔ ✔ ✔ ✔ 4

7 Las Cuevitas ✔ ✔ ✔ ✔ 4

8 Puerto Libertad ✔ ✔ ✔ ✔ 4

9 Datil Island ✔ ✔ ✔ ✔ ✔ ✔ ✔ 7

10 La Tordilla ✔ ✔ ✔ ✔ 4

11 El Tecomate ✔ ✔ ✔ ✔ 4

12 San Esteban Island ✔ ✔ ✔ 3

13 Punta Roja ✔ ✔ ✔ ✔ ✔ ✔ 6

14 Los Machos ✔ ✔ ✔ ✔ ✔ ✔ 6

15 Puerto Lobos ✔ ✔ ✔ ✔ 4

16 La Poma ✔ ✔ ✔ 3

17 San Francisquito ✔ ✔ ✔ ✔ 4

# Sites 17 13 11 12 6 12 11


PrePrin

ts

35

35

869

Table 2 Molecular diversity. Number of samples (n), number of haplotypes (nH), 870

haplotype diversity (h), corrected haplotype diversity (h†) and nucleotide diversity (π). 871

872

Location n nH h h† π

3. La Ventana* 52 23 0.870 ± 0.035 0.885 0.002 ± 0.001

5. San Pedro Martir Island* 17 9 0.860 ± 0.068 0.881 0.001 ± 0.001

6. San Lorenzo Island* 11 4 0.673 ± 0.123 0.844 0.001 ± 0.001

8. Puerto Libertad 55 21 0.874 ± 0.033 0.887 0.002 ± 0.001

9. Datil Island 20 13 0.947 ± 0.032 0.919 0.002 ± 0.001

11. El Tecomate 26 16 0.935 ± 0.034 0.914 0.002 ± 0.001

12. San Esteban Island 4 4 1.000 ± 0.177 0.943 0.001 ± 0.001

15. Puerto Lobos 11 7 0.909 ± 0.066 0.902 0.002 ± 0.001

16. La Poma 16 12 0.950 ± 0.040 0.920 0.002 ± 0.001

17. San Francisquito* 23 17 0.949 ± 0.034 0.920 0.002 ± 0.001

* Indicates sites within MPAs. 873

874

875

876

877

878

879

880

881

882

883

884


PrePrin

ts

36

36

885

Table 3 Pairwise F statistics between sites. Pairwise FST values are above diagonal and 886

pairwise φST values are below diagonal. Values in bold are statistically significant (P < 887

0.05). 888

889

890

Location 3 5 6 8 9 11 12 15 16 17

3. La Ventana - 0.1342 0.2089 0.1278 0.0939 0.0988 0.0837 0.1130 0.0935 0.0929

5. San Pedro Martir Island -0.0122 - 0.2262 0.1321 0.0955 0.1006 0.0868 0.1167 0.0951 0.0943

6. San Lorenzo Island 0.0381 0.0869 - 0.2064 0.1776 0.1800 0.2039 0.2091 0.1803 0.1747

8. Puerto Libertad 0.0912 0.1189 -0.0094 - 0.0920 0.0969 0.0812 0.1109 0.0915 0.0910

9. Datil Island -0.0001 -0.0018 -0.0090 0.0408 - 0.0587 0.0323 0.0706 0.0513 0.0520

11. El Tecomate 0.0101 -0.0094 0.0621 0.1234 0.0169 - 0.0401 0.0767 0.0576 0.0581

12. San Esteban Island -0.0421 -0.0391 0.0772 0.0386 -0.0605 0.0203 - 0.0542 0.0915 0.0317

15. Puerto Lobos 0.0567 0.0555 -0.0353 0.0122 -0.0035 0.0433 0.0193 - 0.0696 0.0697

16. La Poma -0.0124 -0.0196 0.0725 0.0891 -0.0169 0.0038 -0.0578 0.0554 - 0.0507

17. San Francisquito -0.0045 -0.0055 0.0193 0.0601 -0.0129 0.0092 -0.0455 0.0193 -0.0176 -

891

892

893

894

895

896

897

898

899

900

901


PrePrin

ts

37

37

902

Table 4 Probability of three larval dispersal models. Bayes factors and marginal log 903

likelihoods for proposed larval dispersal models estimated in Migrate-n version 3.2.16 904

using Bayesian approximation and thermal integration. 905

906

907

908

909

910

911

912

913

914

915

916

917

Model Bezier lmL Harmonic

lML

Choice

(Bezier)

Model

probability

Full Matrix -2673.20 -2514.33

2 0.00001

Baja to

Mainland

-2658.64 -2513.07 1 0.99999

Mainland to

Baja

-2974.33 -2529.21 3 0.00000


PrePrin

ts

38

38

Figures 918

919

920

921

Figure 1 Area of study around the Midriff Islands in the Gulf of California. Study sites, 922

in which modeled and genetic connectivity was measured (Release sites with genetic 923

samples) and those only for modeled connectivity (Release sites only), including marine 924

protected areas, no-take zones and fishing zones identified by interviews with fishers 925

(Moreno-Báez et al. 2012, 2010). 926

927

928

929


PrePrin

ts

39

39

930

931

932

933

934

935

936

Figure 2 Spawning season and period. Monthly proportion of actively spawning and 937

spawning capable females (left y axis) and female gonadosomatic index (GSI, right y 938

axis) for the Midriff Islands collected in 2009. 939

940

941

942

943

944

945

0

1

2

3

4

5

6

7

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

1 2 3 4 5 6 7 8 9 10 11 12

Mea

n Fe

mal

e G

SI (

%)

Pro

porti

on

Month

Actively Spawning

Spawning Capable

Female GSI


PrePrin

ts

40

40

946

947

948

949

950

Figure 3 Probability of larval density. Density is shown for larvae only in coastal areas 951

after PLD 14 days (A), 21 days (B) and 28 days (C) in each spatial unit of analysis. 952

953

954

955

956

957


PrePrin

ts

41

41

958

Figure 4 Larval dispersal from each spawning site. Maps showing the spatial distribution 959

of larvae released in May 16th (A) and May 25th (B) at each of 17 sites ("X" in each box) 960

for PLD 14, 21 and 28 days. Scale in kilometers. 961

962

963

964


PrePrin

ts

42

42

965

966

967 968

969

Figure 5 Modeled networks of larval connectivity. Spatial networks of larval dispersal 970

between sites for PLD 14 days (A), 21 days (B) and 28 days (C), showing dispersal 971

events (links) between sites (nodes). Line width is proportional to probability, according 972

to the scale to the right. The direction of the larval dispersal events is indicated by 973

different colors: northward (red), southward (blue) or both simultaneously (green). Open 974

nodes are sites within MPAs, solid nodes are fished sites. 975

976

977

978

979

980

981

982


PrePrin

ts

43

43

983

984

985

986

987

988

989

Figure 6 Performance of 17 sites for different aspects of marine connectivity. 990

Performance was measured with export probability, import probability, local-retention 991

probability, out-degree, in-degree and betweenness centrality, as estimated by an 992

oceanographic model at PLD 14 and 28 days. 993

994

995

996

997


PrePrin

ts

44

44

998

999

1000

1001

1002

1003

Figure 7 Minimum spanning network among haplotypes. The network shows the 1004

relationships among 77 haplotypes found in M. rosacea. Circles are sized proportionally 1005

to the number of individuals that possess each haplotype and colors indicate their 1006

geographic distribution in 10 sites shown to the left. All haplotypes are separated by one 1007

to four mutation steps as denoted by scaling provided. 1008

1009

1010

1011

1012

La VentanaDatil IslandEl TecomateSan Pedro Martir IslandSan Esteban IslandPuerto LibertadPuerto LobosPuerto LobosLa PomaSan Lorenzo IslandSan Francisquito


PrePrin

ts

45

45

1013

1014

1015

1016

1017

Figure 8 Correlations between genetic diversity and characteristics of modeled spatial 1018

networks derived from graph theory. Correlation between corrected haplotype diversity 1019

and properties of each site calculated from the modeled networks of larval dispersal, 1020

including: in-degree (A) and betweenness centrality (B). In both, the coefficient of 1021

correlation (R2) and the respective P values are shown for all sites (excluding San 1022

Lorenzo Island). Numbers refers to localities as in Table 1. 1023

1024

1025

1026

1027

1028


PrePrin

ts

Asymmetric connectivity of spawning aggregations of a - PeerJ

Documents

Transcript of Asymmetric connectivity of spawning aggregations of a - PeerJ