draft - university of toronto t-space · 104 < 20 km from the embayments (fig. 2a; wakefield et...

Draft

Does a spatio-temporal closure to fishing Chrysophrysauratus (Sparidae) spawning aggregations also protect

individuals during migration?

Journal: Canadian Journal of Fisheries and Aquatic Sciences

Manuscript ID cjfas-2017-0449.R2

Manuscript Type: Article

Date Submitted by theAuthor: 15-Jun-2018

Complete List of Authors: Crisafulli, Brett; Government of Western Australia Departmentof Primary Industries and Regional Development, FisheriesDivisionFairclough, David; Government of Western AustraliaDepartment of Primary Industries and Regional Development,Fisheries DivisionKeay, Ian; Government of Western Australia Department ofPrimary Industries and Regional Development, FisheriesDivisionLewis, Paul; Government of Western Australia Department ofPrimary Industries and Regional Development, FisheriesDivisionHow, Jason; Government of Western Australia Department ofPrimary Industries and Regional Development, FisheriesDivisionRyan, Karina; Government of Western Australia Department ofPrimary Industries and Regional Development, FisheriesDivisionTaylor, Steve; Government of Western Australia Department ofPrimary Industries and Regional Development, FisheriesDivisionWakefield, Corey; Government of Western Australia Departmentof Primary Industries and Regional Development, FisheriesDivision

Keyword:FISHERY MANAGEMENT < General, MARINE PROTECTED AREAS< Environment/Habitat, REPRODUCTION < General,RECREATIONAL FISHERIES < General, MIGRATION < General

Is the invited manuscriptfor consideration in a

Special Issue? :Oceans Tracking Network

https://mc06.manuscriptcentral.com/cjfas-pubs

Canadian Journal of Fisheries and Aquatic Sciences

Draft

1

Does a spatio-temporal closure to fishing Chrysophrys auratus (Sparidae) 1

spawning aggregations also protect individuals during migration? 2

3

4

5

6

1Crisafulli, B.M., 1Fairclough, D.V., 1Keay, I.S., 1Lewis, P., 1How, J.R., 7

1Ryan, K.L., 1Taylor, S.M. and 1Wakefield, C.B. 8

9

10

11

12

13

14

Corresponding author: 15

Brett M. Crisafulli 16

E: [email protected] 17

Ph: +61 8 9203 0111 18

1Department of Primary Industries and Regional Development, Fisheries Division 19

Government of Western Australia 20

39 Northside Drive, Hillarys 21

Western Australia 6025 22

Page 1 of 58



Draft

2

Abstract 23

Understanding migration dynamics of fishes that aggregate-spawn is critical if spatio-24

temporal closures to fishing are expected to protect them. Concern over fishing of 25

Chrysophrys auratus spawning aggregations in embayments near a west Australian city led to 26

an annual four-month spatial fishing closure. However, the extent to which it protects fish 27

migrating to and from aggregations is unclear. Acoustic telemetry demonstrated a bimodal 28

pattern of entry to and departure from the main embayment via only one of several pathways. 29

Among years, 33-56% of fish occurred in the pathway prior to the closure, but most left 30

before it ceased. Fish were detected within the closure in multiple but not always consecutive 31

years. Variation in migration timing and aggregation philopatry may alter capture risk, but 32

pre- and post-spawning migratory fish are fished in the main pathway and adjacent reefs, 33

which would presumably impact spawning aggregation biomass. Assessment of this would 34

assist in understanding whether expansion of the closure’s spatial and temporal limits is 35

necessary to ensure spawning biomass or if current management is sufficient. 36

37

38

39

40

Keywords: recreational fisheries, migration, marine protected areas, reproduction, fishery 41

management 42

43

Page 2 of 58



Draft

3

Introduction 44

A diverse array of fishes aggregate to spawn, including epinephelids, carangids, 45

labrids and sparids (Sadovy de Mitcheson and Colin 2012). Fish spawning aggregations 46

(FSA) are usually predictable, occurring repeatedly at specific sites at the same time of year, 47

during particular environmental cycles (e.g. lunar phases, water temperatures, diel cycles, 48

tidal regimes etc.) and with fish using the same migratory pathways to join them (Zeller 49

1998; Wakefield 2010; Domeier 2012; Nanami et al. 2014). This predictability and associated 50

increase in catchability has led, in many cases, to overexploitation of spawning aggregations 51

via commercial, subsistence and/or recreational fishing (Sadovy de Mitcheson and Erisman 52

2012; Grüss et al. 2014; Sadovy de Mitcheson 2016). 53

Spatial and temporal closures to fishing are often implemented to protect FSAs. 54

However, of 17 reviewed closures, less than half (43%) led to purported species benefits, 55

such as increases in abundance or body size, and only 17% indicated specific fisheries 56

benefits, e.g. increased yield or catch per unit effort (Grüss et al. 2014). Expected benefits of 57

fishing closures to FSAs or the broader stocks are influenced by the (1) often transient nature 58

of aggregating fish vs the fixed spatio-temporal constraints of closures to fishing, potentially 59

leaving adults exposed to exploitation when not aggregating, (2) variation in whether all or 60

only part of the adult population aggregates to spawn and (3) limited understanding of the 61

complex patterns of behaviour and movement of species that aggregate-spawn, such as 62

predictability of migration pathways, spawning site philopatry and habitat use (Zeller 1998; 63

Erisman et al. 2017). This information is critical if implemented closed areas are expected to 64

contribute to resource management (Heupel et al. 2006). 65

The sparid Chrysophrys auratus (Forster 1801) supports large commercial and 66

recreational fisheries in Australia and New Zealand, including the targeting of its spawning 67

aggregations (e.g. Parsons et al. 2009; Hamer et al. 2011; Sadovy de Mitcheson and Colin 68

Page 3 of 58



Draft

4

2012). On the west coast of Australia, this functional gonochorist only aggregates to spawn in 69

two regions almost 1 000 km apart at ~26 and 32°S (Wakefield 2010; Jackson and Moran 70

2012; Fig. 1). Peak spawning occurs when water temperatures are within a narrow range 71

(~19-21°C) and thus at different times of the year in the latitudinally separate regions. At 72

32°S on the lower west coast, this usually occurs during austral spring (around November), 73

although spawning can extend from August to January (Wakefield 2010; Wakefield et al. 74

2015). The restricted environmental parameters for peak spawning by C. auratus would 75

contribute to its high interannual recruitment variation, which may include spawning 76

omission in some years and latitudinal variation in recruit abundance (Francis 1993; Fowler 77

and Jennings 2003; Sim-Smith et al. 2012; Wakefield et al. 2015; 2017). 78

On the lower west coast of Australia, C. auratus can reach over 1 300 mm total length 79

and 40 y, and matures at ~6 y and ~600 mm, making it inherently vulnerable to fishing 80

(Randall et al. 1997; Norriss and Crisafulli 2010; Wakefield et al. 2015). These large fish 81

aggregate to spawn in three adjacent marine embayments at ~32°S, that are close to the major 82

population centre of this region. Here, they are exposed to substantial anthropogenic 83

activities, including industry and ports, naval and trade vessels, recreational boating and 84

fishing (Wakefield 2010; Wakefield et al. 2013a). The aggregations are easily accessible and 85

historically they were fished both commercially and recreationally until an annual six-week 86

closure (15 Sept-31 Oct) to fishing for C. auratus in two of the embayments (Cockburn 87

Sound and Warnbro Sound) was introduced in 2000 (Fig. 1). This was extended to four 88

months (1 Oct-31 Jan) by 2005 to better protect fish during their spawning period 89

(Wakefield, 2010). As demersal species such as C. auratus are a major target of commercial 90

and recreational fisheries along the broader west coast and they had experienced overfishing, 91

significant changes to management were made between late 2007 and early 2010 to recover 92

stocks (Wise et al. 2007). Commercial fishing for these species became limited to permitted 93

Page 4 of 58



Draft

5

vessels in late 2007, which were prohibited from fishing in the Metropolitan Area of this 94

coast (31-33°S, Fig. 1). Recreational fishing from private- and charter-boats is allowed in the 95

Metropolitan Area under input and output controls introduced at the end of 2009. These 96

included a two-month (15 Oct - 15 Dec) fishing closure for demersal species in the West 97

Coast Bioregion (WCB; Fig. 1), boat-fishing licences and small daily bag limits (two 98

demersal fish per person). The minimum length for retention of C. auratus was also increased 99

to 500 mm total length for both sectors in the Metropolitan and South-west Areas of the 100

WCB (see Fig. 1). 101

Adult C. auratus tagged in the embayments in the spawning periods from 2003 to 102

2016 during this and previous studies disperse after spawning, with ~79% of fish recaptured 103

< 20 km from the embayments (Fig. 2a; Wakefield et al. 2011). About 91% of recaptures 104

have occurred within 4 y of tagging and 72% during the same or subsequent spawning 105

seasons, i.e. Aug-Jan (Fig. 2b; Wakefield et al. 2011; 2015). A small number of C. auratus 106

have been recaptured ~100-700 km from the embayments (Fig. 2a). However, otolith 107

microchemistry of C. auratus in this region indicated that adults typically do not move 108

distances greater than the scale of the large management areas along this coast, e.g. the 109

Metropolitan Area, which encompasses the embayments (Fairclough et al. 2013). Such 110

patterns are broadly consistent with movement variation described for this species in small- 111

to large-scale studies across Australia and New Zealand (e.g. Moran et al. 2003; Parsons et al. 112

2014; Harasti et al. 2015; Fowler et al. 2017). The recapture of fish in the lower west coast 113

embayments indicates that, if they did leave following the spawning period in which they 114

were tagged, they also returned. However, this does not demonstrate whether they return to 115

spawn in multiple or consecutive years, given the nature of mark-recapture methods. While 116

spawning also occurs in surrounding marine waters and along the extensive coast beyond the 117

Metropolitan Area (Lenanton 1974; Lenanton et al. 2009; Wakefield et al. 2011), there may 118

Page 5 of 58



Draft

6

be selective advantages in returning, as oceanography in the embayment facilitates the 119

retention of eggs and larvae, ensuring recruits remain in relatively more productive waters 120

during early life (Lenanton 1974; Wakefield 2010; Wakefield et al. 2011). 121

Biological studies have contributed critical information to the spatial management of 122

C. auratus across this extensive coastline. Knowledge of the reproductive behaviour and peak 123

spawning times in the embayments specifically, informed the introduction of the spatio-124

temporal fishing closure to protect aggregations (Wakefield 2010; Wakefield et al. 2015). 125

However, despite this knowledge and the comprehensive management of this species, 126

locations and timing of migrations in relation to the closure on the lower west coast have not 127

been assessed and thus neither has the exposure of migrating C. auratus to fishing. 128

Acoustic telemetry is frequently used to identify fine-scale movement patterns, that 129

mark-recapture data does not. It has been used to investigate the potential effectiveness of 130

spatial closures for protecting species or stocks (e.g. Meyer et al. 2010; Parsons et al. 2010; 131

Abecasis et al. 2009; 2015), but rarely in fisheries management (Sadovy de Mitcheson et al. 132

2008; Donaldson et al. 2014; Crossin et al. 2017). Over three years, we examined the spatial 133

and temporal patterns of occurrence of adult C. auratus in relation to the three embayments 134

on the lower west coast of Australia where it aggregates to spawn. Data were used to infer 135

migration patterns to and from the main embayment of Cockburn Sound. Detection data from 136

an acoustic receiver array, which included receivers provided by the Canada Foundation for 137

Innovation-funded international Ocean Tracking Network, were used to (1) determine the 138

entrance/exit pathways used by individuals to migrate, (2) identify the timing of immigration 139

and emigration and (3) evaluate the hypothesis that individuals exhibit spawning site 140

philopatry by returning to Cockburn Sound in consecutive years. These lines of evidence 141

were then used to investigate the potential for exploitation of migrating C. auratus given 142

Page 6 of 58



Draft

7

current spatial management arrangements, broader stock management measures and available 143

data on fishing in the vicinity of the embayments. 144

145

Materials and methods 146

Sampling regime 147

Chrysophrys auratus were caught by line fishing during the spawning periods in 2009 148

(n = 30) and 2010 (n = 10) on the lower west coast of Australia, either between Garden and 149

Carnac Island (Site A), or at locations within Cockburn Sound (Sites B and C; Supplementary 150

Table 1; Fig. 1). Each fish was placed in a circular 250 L plastic tank with aerated seawater 151

and anaesthetised using clove oil (0.27 ml L-1 seawater) until stage II anaesthesia, i.e. ~60 s 152

(Munday and Wilson 1997). Fork length was then measured to the nearest 1 mm and sex 153

determined by cannulation or during surgery if possible. Each fish was laid inverted in a v-154

shaped foam support covered in wet plastic and its gills irrigated with fresh seawater. A 155

Vemco© acoustic transmitter (V16-4H, V13-1L, V13-1H) was inserted into the body cavity 156

via a ~3 cm incision in the abdominal wall, ~2 cm anterior of the cloaca and 1 cm lateral of 157

the ventral mid-line. The incision was closed using absorbable sutures (Ethicon Vicryl CP-1 158

with 36 mm, ½ circle reverse cutting needle) and the fish injected with Bivatop antibiotic 159

(200mg mL-1 tetracycline) at 0.1 ml kg-1 of body weight, estimated using the fork length-total 160

weight relationship of Wakefield et al. (2015) (Supplementary Table 1). Two 13 cm dart tags 161

(PDS: Hallprint ©) were inserted into the dorsal musculature of each fish for potential 162

recapture identification. Fish recovered from anaesthesia in a 500 L tank of aerated sea water 163

until they oriented and swam independently, which took 2-11.5 min (mode = 3 min). Fish 164

were then immediately released. The time between capture and release ranged from ~7.5-165

17 min (mode = 9 min; based on data recorded for 21 fish). In Western Australia, the Animal 166

Welfare Act 2002 does not require the Department of Primary Industries and Regional 167

Page 7 of 58



Draft

8

Development to obtain a permit to use animals (fish) for scientific purposes unless the species 168

are outside the provisions of the governing legislation (i.e. Fish Resources Management Act 169

1994 and Fish Resources Management Regulations 1995). However, all sampling was 170

undertaken in strict adherence to the Department’s Policy for the handling, use and care of 171

marine fauna

for research purposes. In addition, surgery was conducted following methods of 172

previous ethically approved work (Fairclough et al. 2011) and with input from the 173

Department’s Animal Health staff and is consistent with the guidelines of the Committee for 174

the Update of the Guide for the Care and Use of Laboratory Animals (National Research 175

Council 2011). 176

No significant difference (Kolmogorov-Smirnov two-sample test, Da = 0.035; df = 1; 177

p > 0.05) was identified between the length frequency distributions of the C. auratus tagged 178

for telemetry and a larger sample of C. auratus (n = 549) captured from the same spawning 179

aggregations during 2009 and 2010 (Fig. 3). Thus, the sample of C. auratus used for 180

telemetry was considered representative of the broader population that aggregated to spawn 181

in Cockburn Sound during the study period. 182

183

Receiver array 184

Vemco© receivers were deployed across five potential pathways to Cockburn Sound 185

to produce double-gates that could detect immigration and emigration of C. auratus (Fig. 1). 186

Cockburn Sound is bounded at its northern end by a shallow bank, while at the southern end, 187

a narrow, shallow passage between Garden Island and the mainland delineates the 188

embayment from the open ocean. The five pathways each comprised two lines of receivers: 189

(1) Rottnest Island to Fremantle and Garden Island to Woodman Point, 190

(2) Rottnest Island to Straggler Rocks and Garden Island to Woodman Point 191

(3) Straggler Rocks to Carnac Island and Garden Island to Woodman Point 192

Page 8 of 58



Draft

9

(4) Carnac Island to Garden Island and Garden Island to Woodman Point, and 193

(5) the southern end of Garden Island and the mainland. 194

The array comprised 47 VR2W receivers (the ‘aggregation array’), 53 VR2Ws in the 195

Canada Foundation for Innovation-funded international Ocean Tracking Network’s Perth line 196

(OTN; http://oceantrackingnetwork.org; Cooke et al., 2011) and 19 VR2Ws and 20 VR4Gs 197

that form the Metropolitan component of the Department's Shark Monitoring Network (SMN; 198

McAuley et al. 2016). For analyses, the gate represented by Site A comprised the western line 199

of seven receivers between Garden and Carnac islands and the nine most western receivers 200

between Garden Island and Woodman Point (Fig. 1). As part of the ‘aggregation array’, a 201

VR2W was deployed at site B and C in Cockburn Sound and at three locations in Warnbro 202

Sound, where C. auratus also aggregate to spawn. A temporary array of 19 receivers was 203

deployed in Cockburn Sound from Dec 2009–Feb 2010 and Aug 2011–Mar 2012 to increase 204

the potential for detection of C. auratus. The SMN also comprised cross-shelf lines of 205

VR2Ws deployed in Apr/May 2012 off the south-west (48 VR2Ws) and western south coasts 206

of Australia (two lines, 44 and 33 VR2Ws; Fig. 1). Receivers in the SMN and OTN arrays 207

were deployed at 800 m intervals (McAuley et al. 2016). 208

Each VR2W in the aggregation array was attached to a stainless-steel pole mounted 209

vertically in a concrete-filled car tyre. VR2Ws were deployed ~500 m apart in water depths 210

of 4–17 m, at locations not obstructed by reef structure. Distances between receivers were 211

determined from (1) knowledge of an expected range of 525 m for V16-4Hs (158 dB re 1 µPa 212

@ 1 m) in wind conditions of 28–34 knots (https://vemco.com/range-calculator/) and (2) 213

range testing of a V16–4H (fixed 120 s delay) in the study area, where detections decreased 214

by 50 and 95% of the theoretical maximum number per day by 531 and 717 m, respectively 215

(How and deLestang 2012). Receivers were retrieved from the water annually to remove 216

fouling, download data via Vemco VUE (version 1.6.5) and install new batteries. 217

Page 9 of 58



http://oceantrackingnetwork.org/

https://vemco.com/range-calculator/

Draft

10

218

Analysis of detection data 219

Detection data from all receivers in all the arrays were kept within a central data 220

management system. This contained transmitter ID, date and time, spatial coordinates of 221

receiver (decimal degrees), Receiver Station ID and depth of Receiver Station associated with 222

transmitters detected from this study and tagged sharks from other studies (McAuley et al. 223

2016; Braccini et al. 2017). The False Detection Analyser in Vemco VUE v2.30 was used to 224

identify false transmitter detections; i.e. Type A (invalid transmitter codes, which can be 225

produced as a result of environmental sounds or transmitter code collisions) and Type B 226

(false detections of valid transmitters due to detections occurring on a single receiver at 227

shorter time intervals than the programmed ping rate) (Heupel et al. 2006; Simpfendorfer et 228

al. 2015). Approximately 0.01% of a total 3.5 million detections were Type A false 229

detections, with the vast majority occurring at site B. These would have been influenced by 230

the six V16 transmitters which which produced almost continuous detections at that site and 231

may have been deceased fish. These were removed from analyses (see later in Methods). 232

While the cause of false detections could not be ascribed to transmitters used in this study, 233

due to transmitters also having been deployed for other studies simultaneously, they 234

represented a very low percentage of total detections and would have had little effect on the 235

outcomes of the analyses. No Type B false transmitter detections were identified. A further 236

195 detections from one fish were removed due to a receiver station code transcription error. 237

Data were analysed using R (R Core Team, 2017). Six C. auratus with V16 238

transmitters contributed ~2 million detections to the total 3.5 million during the study and 239

were all recorded at site B (Fig. 1). As these detections were essentially continuous, it was 240

not possible to distinguish whether these fish had become resident or were deceased and still 241

within proximity of the receiver. Daily variations in detection rates may have been caused by 242

Page 10 of 58



Draft

11

environmental or anthropogenic noise interference, rather than movement towards and away 243

from the Site B receiver or a combination of all these factors. These data were omitted from 244

analyses. One fish (#60358) was predated on by a shark at the time of release (pers. obs.) and 245

thus its detections were removed from analyses. 246

Detections from the remaining 33 C. auratus, with 23 V16 and 10 V13 transmitters 247

were used to determine the locations of entry to or exit from Cockburn Sound, while only the 248

V16 transmitters were used to investigate migration timing and philopatry after 2009/10. 249

These transmitters were programmed to last for ~ 1000 d, with short and long delay periods 250

to coincide with spawning months and non-spawning months, respectively. The V13 251

transmitters were surplus from other studies and were beyond their recommended shelf life. 252

Thus, they were potentially less reliable for analyses of timing of migrations. The daily 253

presence/absence of each C. auratus within the entire array (aggregation array, OTN and 254

SMN) was determined from the day it was released until either its last detection within the 255

array (if this occurred before the end of the transmitter’s expected life) or the estimated end 256

of the transmitter’s life (see Supplementary Table 1). The sequence of detections of each fish 257

in double gate arrays at each entrance/exit and/or at receivers inside Cockburn Sound were 258

used to evaluate directionality of movement. 259

As detections of C. auratus among the five potential gates occurred almost 260

exclusively at the gate between Garden and Carnac islands, i.e. Site A (see Results), these 261

data were used to evaluate whether there was evidence of immigration and emigration via a 262

bimodal pattern of occurrence at Site A and whether it was consistent among years. This was 263

predicted to occur based on the expectation that C. auratus move into and out of Cockburn 264

Sound prior to and after the spawning period, respectively. These analyses were also used to 265

relate this movement to the timing of the spatio-temporal closure, commencing on 1 Oct each 266

year. The number of unique C. auratus detected from 1 Aug to 28 Feb in 2010/11 and 267

Page 11 of 58



Draft

12

2011/12 in each discrete week (7 days) prior to and after 1 Oct in each year was expressed as 268

a proportion of the total possible number of transmitters that could be detected in each year 269

(i.e. 20 fish, which did not include the three fish which became resident at Site A or inside 270

Cockburn Sound). The data were assumed to be normally distributed around a mean week of 271

immigration and emigration. Single, bimodal and tri-modal finite mixture distribution models 272

were fitted to the data for each year according to the function 𝐹𝐹(𝑥𝑥) = ∑ 𝑤𝑤𝑖𝑖𝑛𝑛𝑖𝑖=1 𝑃𝑃𝑖𝑖(𝑥𝑥), where 273

𝑃𝑃𝑖𝑖(𝑥𝑥) is the probability density function for each distribution (with mean �̅�𝑥 and standard 274

deviation 𝑠𝑠) and 𝑤𝑤𝑖𝑖 is the weight of each distribution, such that ∑𝑤𝑤𝑖𝑖 = 1. Model parameter 275

estimates were derived using a maximum likelihood approach. Akaike Information Criterion 276

(AIC) values were calculated to determine which of the models provided the best fit (lowest 277

AIC), where AIC = -2LL + 2K and K is the number of parameters in the regression equation 278

(including the estimate of the residual variance). Models were compared after adjusting for 279

sample size, as the ratio of n/K was less than 40 (n is the sum of the number of transmitters 280

observed per week from 1 Aug to 28 Feb), as follows: AICc = AIC + 2𝐾𝐾(𝐾𝐾+1)𝑛𝑛−𝐾𝐾−1

(Burnham and 281

Anderson 2002). Note that the assumption of the presence/absence data being gamma 282

distributed was also evaluated using single, bi- and tri-modal mixture models. However, the 283

shapes of the fitted distributions and their log likelihoods were similar to those of the normal 284

distributions in each case in each year. The bimodal model again produced the lowest AICc 285

values. Thus, the normal distribution fits were considered appropriate. Daily sea surface 286

water temperature (SST) data (multi-scale ultra-high resolution SST analysis fv04.1, 1 km 287

resolution) were obtained for an area bounded by 32.1-32.2 ºS and 115.6-115.7 ºE from 1 288

Aug to 28 Feb in 2010/11 and 2011/12 from https://coastwatch.pfeg.noaa.gov/erddap/griddap 289

(downloaded 6 April 2018). Moon fraction (percentage full) data were obtained for the same 290

period from http://aa.usno.navy.mil/data/docs/MoonFraction.php (downloaded 6 April 2018). 291

Mean weekly SST and moon fraction were then calculated, as per detection data. 292

Page 12 of 58



https://coastwatch.pfeg.noaa.gov/erddap/griddap

http://aa.usno.navy.mil/data/docs/MoonFraction.php

Draft

13

293

Fishery and biological data 294

Fishery and biological data were obtained from time periods as close as possible to 295

that of the telemetry study. These data were used to provide general patterns of fishing for 296

C. auratus and of its reproductive biology. Available retained catch data for C. auratus from 297

the commercial and recreational sectors were used to provide an indication of the typical 298

spatial distribution of catches per year by the two sectors across the four management areas 299

of the west coast. These include the Kalbarri (from 26°30′–28°S), Mid-west (28–31°S), 300

Metropolitan (from 31–33°S) and South-west areas (from 33°S on the west coast to 115°30′E 301

on the south coast; Fig. 1). Commercial fishing for C. auratus in the Metropolitan Area has 302

been prohibited since 2008, other than by a small line fishery which occasionally catches 303

C. auratus (Fairclough et al. 2014a). Catch data from compulsory logbooks of commercial 304

line fishers between 2008 and 2016 were used to calculate the average annual commercial 305

catch from each area. The recreational sector comprises both charter- and private-boat fishers. 306

Catch data from charter-boat fishing between 2008 and 2016 were obtained from compulsory 307

Tour Operator Returns and used to calculate their average annual catch for each area. As 308

catches from private-boat based recreational fishing are not surveyed annually, estimated 309

annual catches from each area were derived from a recent one year survey of recreational 310

fishers during 2013/14, which integrated phone-diary surveys and boat-ramp validation 311

surveys (Ryan et al. 2015). These estimates were combined with the average charter catches 312

by area to provide an estimate of the typical average annual catch of the recreational sector in 313

each area. The average annual catches of C. auratus in each area by the commercial and 314

recreational sectors were then used to calculate the percentage of the total catch taken by each 315

sector in each area. 316

Page 13 of 58



Draft

14

Charter- and private-boat data were further examined to infer any changes in fishing 317

activity (targeting) associated with the presence of migrating or aggregating C. auratus, prior 318

to, during and after the spawning period. Reported catches by charter-boats in each 10×10 nm 319

block in the Metropolitan Area (encompassing the area of the study) in each month in 320

2015/16 were plotted to identify the spatio-temporal distribution of their catch. These data are 321

not available for private vessels. Thus, to provide information on private vessel activity in the 322

area of the study, the estimated number of powered-vessel launches was determined for each 323

hour of the day in each month from March 2011 to February 2012, calculated from footage 324

obtained from a remote camera installed at Woodman Point, typically the Metropolitan 325

Area’s most-used boat ramp (Fig. 1; Ryan et al. 2013; 2015). Although not all vessels 326

launched are used in fishing activities, the number of launches is positively correlated with 327

recreational fishing effort and thus provides an index of fishing effort (Steffe et al. 2017). 328

Data on the frequency of occurrence of C. auratus in the Metropolitan Area, including 329

Cockburn Sound, that contained either stage III (i.e. developed or pre-spawning) or stage IV 330

(i.e. ripe or spawning) ovaries between 2002 and 2006 were obtained with permission from 331

Wakefield et al. (2015). The percentage frequency of these two stages in each month of the 332

year was used to compare the timing of the spawning period to the timing of detections 333

within the array and the annual closure to fishing in the embayments. Although this work was 334

conducted prior to the telemetry study, other studies of C. auratus in Cockburn Sound that 335

have focused on aspects of its reproductive biology (e.g. egg production) provide evidence of 336

consistent interannual timing of the spawning period (cf. Wakefield 2010; Dias et al. 2016; 337

Partridge et al. 2017). 338

Page 14 of 58



Draft

15

Results 339

Over one million valid detections of C. auratus with transmitters were recorded 340

throughout the study by the aggregation, OTN and SMN arrays combined. The total number 341

of detections of individual transmitters ranged from 113 to 265,742, while V16 transmitters 342

were detected on up to 675 d (�̅�𝑥 = 148.5 d; Supplementary Table 1) over the three year study. 343

The majority of C. auratus detections (~83%) occurred between September and January, with 344

> 99% of those recorded from 29 of the 33 valid transmitters at Site A (the line of seven 345

receivers between Garden and Carnac islands and the nine most western receivers between 346

Garden Island and Woodman Point) and sites B and C inside Cockburn Sound (Figs 1, 4, 5). 347

Relatively few C. auratus were detected in the other pathways to Cockburn Sound, i.e. 348

eastern receivers between Garden Island and Woodman Point (< 1% of detections, 10 fish), 349

Rottnest Island to Fremantle (< 1%, 13 fish), Rottnest Island to Straggler Rocks and Carnac 350

Island (< 1%, 8 fish) and the southern end of Cockburn Sound (< 1%, 4 fish) (Fig. 4). 351

352

Timing of detections in relation to the fishing closure and environmental parameters 353

Detections of V13 and V16 transmitters provided strong circumstantial evidence of 354

sequential movement to and/or from Cockburn Sound. After fish were tagged in 2009, 25 of 355

the 29 V16 and V13 transmitters detected at sites A (Garden Island to Carnac Island), B or C 356

(inside Cockburn Sound) between October and December 2009 were subsequently detected 357

at Site A before the end of the closure to fishing on 31 Jan 2010 (Fig. 5). Twenty three of 358

those 25 ceased to be detected at Site A by 31 Jan. One remained at Site A for a further 1.5 y 359

and then was no longer detected (#60352), while another (#1367) was detected again at Site 360

A in early February 2010. Fourteen fish were subsequently detected outside the embayment 361

on SMN or OTN receivers between Garden and Rottnest islands, or briefly east of Site A, 362

towards Woodman Point. Four of the 29 fish continued to be detected at Site B in the 363

Page 15 of 58



Draft

16

embayment where they were tagged. However, three of those (#60361, 60639, 60371) ceased 364

to be detected by 31 January, while one (#51857) was detected until early February 2010 at 365

Site B and then once in April 2010 outside the embayment, immediately north of Carnac 366

Island (Fig. 5). 367

Of a possible 23 fish with valid V16s, six of the 17 fish detected during the 2010/11 368

closure and 9 of the 15 fish detected during the 2011/12 closure were first detected prior to 369

the closure commencing on Oct 1 (Fig. 5). This occurred 32-36 d before the closure in those 370

years, respectively, either beyond the Site A gate on SMN or OTN receivers, at Site A or at 371

Site B. In contrast, the last detections of almost all fish at either Site A or B in the 2010/11 372

and 2011/12 spawning seasons were recorded before the end of the closure to fishing on 31 373

Jan. During the 2010/11 spawning period, six of 19 fish with V16s were recorded 374

consecutively at Site A, followed by Site B or C within Cockburn Sound and then five were 375

later recorded at Site A (Fig. 5). Seven fish were only detected at Site A, two of which 376

demonstrated evidence of residency and one was only detected late in November, thus later in 377

the spawning period. A hiatus in detections occurred at Site A with four of the seven fish, 378

during which time they were not detected elsewhere. However, three were last detected on 379

the eastern (or inside) line of receivers leading in to Cockburn Sound. While three fish were 380

not detected that year, the remaining three were detected at only Site A early in the spawning 381

season or only at Site B intermittently during the spawning season. Of the four fish tagged at 382

Site B in the 2010/11 spawning season with valid V16s, each was detected for a period at Site 383

B and three were subsequently detected at Site A before the end of the closure. 384

Of 23 valid V16s, detections at Site A, followed by Site B and then Site A were 385

recorded for two fish in 2011/12. Another 10 fish were detected at Site A, but not inside 386

Cockburn Sound. Four of those exhibited a hiatus in detections, preceded by detections on the 387

eastern (or inside) line of receivers at Site A. Three fish were detected intermittently only 388

Page 16 of 58



Draft

17

inside Cockburn Sound, while one was only recorded outside Cockburn Sound, detections of 389

another two indicated residency at Site A or B and five fish were not detected. 390

Across all years of the study, the relatively small number of detections (n = 5576) of 391

22 transmitters on receivers other than at Site A, B or C between September and January 392

occurred over short periods and for the majority of these fish (n = 17) indicated sequential 393

movement to and from Site A. For example, two individuals (#60366 and 60370) were 394

detected on OTN and SMN receivers near Fremantle in Oct 2010 over one and five 395

consecutive days, respectively (808 detections not visible on Fig. 4). They were preceded by 396

detections at Site A 2.2 and 1.25 d before, respectively, and followed by detections at site A 397

0.4 d later and Site B 0.9 d after that. Fish #60370 was then detected at Site B for 44 d and at 398

Site A 0.25 d later. Two C. auratus (#60364, 60368) were detected briefly (12 detections 399

over 33 min; 6 detections over 17 min) on the two western-most receivers of the southern 400

entrance of Cockburn Sound in Aug and Oct 2010 and subsequently on the western receivers 401

at Site A 15 h and 3.8 h later, respectively. Another two individuals were detected in 402

Warnbro Sound for short periods (#60360 for 2 d in Nov 2010 and 5 d in Oct 2011; #60368 403

for 13 d in Oct 2010) and subsequently at Site A within ~1 d. 404

The number of fish with V16 transmitters detected at Site A in each week between 1 405

Aug and 28 Feb in both 2010/11 and 2011/12 had a bimodal frequency pattern. AICc values 406

indicated that bimodal mixture distributions produced the best fit to these data in each year, in 407

comparison to both single mode and tri-modal models (Table 1). In both the 2010/11 and 408

2011/12 spawning seasons, the number of fish detected at Site A began to increase prior to 409

the commencement of the closure on 1 October, while mean water temperatures were < 19 ºC 410

(Fig. 6). The first of the means of the bimodal distributions occurred at ~ 1.0 and 0.6 weeks 411

after the commencement of the closure in 2010/11 and 2011/12, respectively, when water 412

temperatures were ~19 ºC (Table 2, Fig. 6). Fewer fish were detected during subsequent 413

Page 17 of 58



Draft

18

weeks, with the lowest frequencies recorded while water temperatures were < 21 ºC and at 414

the time of the new moon in November in 2010 and between the approximate times of the 415

new moons in October and November in 2011 (Fig. 6). This occurred when the percentage of 416

female C. auratus in Cockburn Sound with ovaries in pre-spawning/spawning condition is 417

typically increasing to or is at its maximum (Fig. 6; cf. Figs 4,8,9). The second mode 418

occurred at 9.6 and 8.0 weeks after the closure commenced, when water temperatures had 419

reached 21 ºC in both years (Table 2; Fig. 6). Frequencies of occurrence then declined as 420

water temperatures increased above 21 ºC, which occurred before the end of the fishing 421

closure on 31 Jan in both years and as the percentage of females with ovaries in pre-422

spawning/spawning condition declined (Fig. 6, cf. Figs 4,7,9). The standard deviation of the 423

mean number of fish detected per week indicated relatively consistent variation around the 424

two means in 2010/11, but the second mean in 2011/12 exhibited less variation than the first 425

and less than both means in 2010/11 (Table 2; Fig. 6). 426

427

Spawning aggregation philopatry 428

The majority (n = 21) of the 23 C. auratus with V16 transmitters were detected either 429

at Site A, B or C in at least one spawning season subsequent to being tagged. Seven fish were 430

detected in only one spawning season following tagging with two of those only detected 431

again two years after being tagged. Ten and four fish were respectively detected in two and 432

three consecutive seasons after being tagged. However, two of the latter four fish appeared to 433

remain at site A or site B for extended periods including the non-spawning period. Although 434

unexpected, two of the ten fish with V13 transmitters were also detected in one or two 435

subsequent spawning seasons (Fig. 5). After detection at sites A and B in two consecutive 436

spawning seasons after being tagged, one fish (#60360) was detected once in March 2012 on 437

Page 18 of 58



Draft

19

the SMN array #1, ~240 km south of the embayments, while another (#60362) was captured 438

and retained by a recreational angler ~135 km north of the embayments. 439

440

Fishing Effort 441

The majority of the average annual retained catch of C. auratus in the WCB is taken 442

by the commercial fishery in the Kalbarri (26°30′-28°S) and Mid-west areas (28-31°S), i.e. 443

~74% (Fig. 7). The recreational sector lands ~25% of the average annual catch in the WCB, 444

of which, ~14% is taken in the Metropolitan Area (31-33°S; Fig. 7). Average annual retained 445

charter fishing catches comprise 32% of the recreational sector catch in the WCB and the 446

largest component of that (16%) is from the Metropolitan Area (not shown in Fig. 7). 447

In the Metropolitan Area, the largest monthly catches (102-200 fish per 10×10 nm 448

block) from charter boat recreational fishing typically occurred west, north or north-west of 449

Rottnest and/or Garden islands (Fig. 8). In July to October, leading up to the peak of the 450

spawning season, the closure to fishing for C. auratus in the embayments (1 Oct-31 Jan) and 451

the closure to fishing for all demersal species, including C. auratus, across the WCB (15 Oct-452

15 Dec), catches were reported only in blocks north of Garden Island (~36.15°S; Fig. 8). The 453

largest of these catches per block were concentrated mostly between Rottnest and Garden 454

Islands and particularly the block encompassing Site A. Between December and March, i.e. 455

towards the end of, and following the spawning season, the largest total catches per block 456

were reported from the same area between Rottnest and Garden islands, but catches were also 457

widely dispersed across the Metropolitan Area, including southwards from Garden Island. 458

From April to June, charter catches were typically low and dispersed northwards from 459

Garden Island. 460

At Woodman Point, the majority of vessel launches in each month occurred in the 461

morning from ~06:00-12:00, followed by relatively fewer launches for the rest of the day and 462

Page 19 of 58



Draft

20

night (Fig. 9). However, in August and September, just prior to the October commencement 463

of the fishing closure, a prominent second mode of launches was detected in the late 464

afternoon/dusk. During these two months, ~30 and 44% of adult female C. auratus in the 465

Metropolitan Area contained ovaries which were in pre-spawning or spawning condition 466

(Fig. 9). This is in contrast to the 68% of adult females with ovaries in pre-spawning or 467

spawning condition in October and > 92% in November and December, when spawning is 468

typically at its maximum in that area. These months also exhibited a unimodal frequency of 469

boat launches in the morning, when fishing for C. auratus is prohibited. By January, adult 470

female C. auratus with ovaries in pre-spawning or spawning condition have declined below 471

50%, and after January when fishing is allowed, little spawning activity occurs (Fig. 9). 472

473

Discussion 474

This study used acoustic telemetry to determine the migration patterns of C. auratus 475

to and from an embayment on the lower west coast of Australia where it aggregates to spawn 476

and is protected by a spatio-temporal fishing closure. Detections of C. auratus indicated 477

predictable and repeated movement behaviours consistent with the relationship between 478

reproduction and environmental parameters in this species. Individuals (1) migrated via only 479

one of several pathways into the main embayment of Cockburn Sound where spawning 480

occurs, (2) often occurred in that pathway before the fishing closure began and spent varying 481

periods of time there, but most left before the closure ceased, (3) were detected in the 482

pathway and/or aggregation locations in the embayment in more than one year, confirming 483

spawning site philopatry, but (4) may not always return in consecutive years. The spatio-484

temporal closure offers limited protection to migrating and pre-spawning fish, as individuals 485

begin arriving at the main pathway before the closure commences and are fished there and 486

over adjacent reef systems at that time. This is likely to impact spawning aggregation 487

Page 20 of 58



Draft

21

biomass. Assessment and monitoring of mortality or spawning biomass would help to 488

identify whether current management is sufficient to ensure the sustainability of the 489

aggregations and their contribution to the broader stock. 490

491

Biological relevance of Chrysophrys auratus movements 492

Recurrent detections of C. auratus in only one of the several pathways to the main 493

embayment in which they aggregate-spawn implies that this pathway provides navigational 494

cues for migration. The pathway lies just to the east of two north-south oriented reef systems 495

(the Garden Island ridge on the western side of the island and Five Fathom Bank further to 496

the west), which both extend southwards from the south-eastern end of Rottnest Island, past 497

Garden Island and, in the case of Five Fathom Bank, beyond 33°S (Fig. 1). Migrating fish 498

may use these to navigate to and from the pathway to Cockburn Sound, which is supported 499

by the recapture of tagged C. auratus along Five Fathom Bank and is consistent with the 500

locations of charter catches of C. auratus. These and other reef lines in deeper water (~80 and 501

110 m) along the west coast may provide important pathways that some C. auratus use to 502

disperse substantial distances from Cockburn Sound as recorded during this and previous 503

studies (Wakefield et al. 2011). Reef lines and shelf margins or ridges comprise obvious 504

bathymetric characteristics which are commonly-used by demersal fishes, such as 505

epinephelids, to migrate to aggregation sites. This usage may be a learned behaviour, 506

particularly in the case of young maturing fish (Domeier and Colin 1997; Starr et al. 2007; 507

Colin 2012; Nemeth 2012; Sadovy de Mitcheson 2016). In southern Australia, C. auratus 508

also demonstrates systematic habitat/location use patterns, dispersing across areas of > 100 509

km2 from a location, but revisiting that location at the same time in different years (Fowler et 510

al. 2017). The pathway in to Cockburn Sound used by C. auratus may also provide resources 511

that others do not, such as habitat that offers protection from predation or food. As some 512

Page 21 of 58



Draft

22

C. auratus may migrate substantial distances, continued feeding may be necessary to provide 513

energy for gonadal development and spawning, as in the carnivorous Epinephelus guttatus 514

and herbivorous Chlorurus sordidus (Sancho et al. 2000; Nemeth 2012; Kawabata et al. 515

2015). 516

Acoustic telemetry has demonstrated that the timing of entry to and departure from 517

Cockburn Sound is consistent with the strong link between the reproductive periodicity of 518

spawning aggregations of C. auratus in the embayment and annual, lunar, diel and 519

oceanographic cycles (cf. Scott et al. 1993; Wakefield 2010; Saunders et al. 2012; Wakefield 520

et al. 2015). A bimodal frequency of occurrence of C. auratus was identified in the main 521

pathway to Cockburn Sound in both 2010/11 and 2011/12, with the first and second modes in 522

each year corresponding with the time when water temperatures were ~19 and 21ºC, 523

respectively. Spawning peaks in these waters within that narrow temperature range, which 524

typically occurs in November (late austral spring), while gonadal atresia occurs when water 525

temperatures exceed 21°C, usually in December/January (summer) (Wakefield 2010; 526

Wakefield et al. 2015). Thus, the two modes presumably reflect periods of accumulation in 527

the pathway, as individuals migrate to and from the aggregations and this migration is linked 528

with water temperature. The trough between the two modes would represent the time during 529

which most individuals are within Cockburn Sound, consistent with the detections of tagged 530

fish at sites in the embayment. This period coincided with one or more new and full moons in 531

2010/11 and 2011/12, during which greater spawning activity typically occurs, particularly 532

on the new moon that coincides with water temperatures of 19-21ºC (Wakefield 2010). The 533

difference in the number of new moons experienced in the time during which water 534

temperatures rise from 19 to 21ºC may thus influence the extent of spawning activity in any 535

one year. 536

Page 22 of 58



Draft

23

As there is a strong relationship between water temperatures and gonadal maturation, 537

migration and/or formation of spawning aggregations of C. auratus, predicted ocean 538

temperature increases (Pearce et al. 2016) may influence the timing or location of those 539

behaviours. This would alter the efficacy of the associated spatio-temporal fishing closure 540

(Sheaves 2006; Jackson et al. 2015; Wakefield et al. 2015). Furthermore, marine ‘heatwaves’ 541

have increased in duration and frequency in the last century (Lenanton et al. 2017; Oliver et 542

al. 2018), which may have rapid and unpredictable effects on C. auratus spawning success in 543

the embayments. Individuals may no longer enter the embayment if water temperatures 544

exceed the preferred range for successful spawning. Such phenological variations in relation 545

to water temperature are common in marine organisms in shallow and deep waters across 546

tropical and temperate regions, including among cephalopods (e.g. Loligo forbesi) and 547

teleosts (Dascyllus albisella, Platichthys flesus, Clupea pallasi, Gadus morhua, Hyporthodus 548

octofasciatus) (Lawson and Rose 2000; Sims et al. 2001; Asoh and Yoshikawa 2002; Sims et 549

al. 2004; Wakefield et al. 2013b). 550

551

Exploitation of migratory and aggregatory fish 552

The predictability of spawning aggregations is the key reason they are susceptible to 553

over-exploitation and consequent negative impacts on abundance and/or biology (e.g. 554

Coleman et al. 1996; Sadovy and Cheung 2003; Erisman et al. 2012; Sadovy de Mitcheson 555

and Erisman 2012). Fishing has the potential to extirpate aggregations and negate any 556

selective advantage conferred by such behaviour and has often led to the need for significant 557

management intervention. For example, while relatively few of the ~150 sparids aggregate-558

spawn (Nemeth 2012), some that do in southern Africa (Petris rupestris, Polysteganus 559

undulosus and Chrysoblephus gibbiceps) are now endangered or critically endangered due to 560

exploitation. This has led to complete closures to fishing for those species (Mann et al. 2014a, 561

Page 23 of 58



Draft

24

2014b; Sadovy de Mitcheson 2016). Commercial and recreational fishing of aggregations of 562

demographically separate but genetically-related C. auratus stocks on the upper west coast of 563

Australia also resulted in stock depletions (Jackson and Moran 2012; Gardner et al. 2017). 564

However, restrictive management to limit catches via a range of measures, including 565

temporal closures and harvest tags, has allowed recovery of some of those stocks over 15+ 566

years (Jackson and Moran 2012; Jackson et al. 2016). Recent overfishing of the coastal 567

Arripis georgianus in the same region would be influenced by the exploitation of spawning 568

biomass both during its pre-spawning migration along 100s to 1 000s of km of the southern 569

Australian coastline and during the spawning period (Fairclough et al. 2000; Smith et al. 570

2013). Similarly, impacts of recreational fishing of migrating Albula vulpes in Kiribati 571

resulted in the need for a fishing closure during its spawning period (Sadovy de Mitcheson 572

and Erisman 2012). 573

Exploitation of C. auratus on the lower west coast of Australia by the recreational 574

sector occurs along the reef systems between Rottnest and Garden islands, adjacent to the 575

fishing closure, and in the main migratory pathway between Garden and Carnac islands. An 576

increase in evening launches of private vessels at Woodman Point, the most proximal boat 577

ramp to the main pathway, in August and September just prior to the fishing closure, is 578

consistent with anecdotal information that recreational fishers target C. auratus at that time. 579

The combination of telemetry and reproductive data (Wakefield 2010; Wakefield et al. 2015) 580

confirms that fishers would be catching both pre-spawning migratory fish and fish that have 581

become periodically resident in the area. However, not all fish may be exposed to fishing 582

each year. Telemetry data also demonstrated that while the majority of fish were detected in 583

subsequent years after tagging, exhibiting philopatry, other fish may have returned but were 584

not detected, or may return but not in consecutive years. As some fish can travel substantial 585

distances (100s of km) from the embayments, they may never return, instead living and 586

Page 24 of 58



Draft

25

reproducing elsewhere. Such long distance movements have been documented in several 587

studies of C. auratus (e.g. Parsons et al. 2014; Harasti et al. 2015), but the reasons for them 588

are not known. Whether such individuals would return and what proportion of fish undertake 589

this type of movement remains unclear. Thus, philopatry rates may be underestimated. This is 590

similarly not well understood for aggregations of C. auratus in Port Phillip Bay in southern 591

Australia (Hamer et al. 2011). Fish that travel substantial distances along the west coast of 592

Australia are exposed to different levels of fishing mortality, associated with the varying 593

management regimes, including different size limits and commercial fishing. Despite 594

restrictive management of the spawning aggregations and broader stocks of C. auratus along 595

the west coast, migratory fish are not protected until they have crossed the boundary of the 596

spatio-temporal closure, after it has commenced. Thus, if exploitation of migratory pre-597

spawning C. auratus in the pathway and over adjacent reef systems is high enough, it may 598

influence migratory patterns and/or reduce spawning biomass. 599

600

Social importance and monitoring of Chrysophrys auratus spawning aggregations 601

While recreational fishers enjoy targeting the seasonally amassing C. auratus in 602

waters adjacent to or within the embayments, many presumably understand that targeting 603

aggregations would have negative consequences. This recognition can increase their 604

willingness to contribute to ensuring suitable management is in place (e.g. Hamilton et al. 605

2012; Heyman and Granados-Dieseldorff 2012). For example, public outcry occurred when 606

over 250 C. auratus died during an environmentally driven fish-kill in Cockburn Sound 607

around the time of peak spawning in 2015 (Department of Fisheries 2016). With recreational 608

sector support and crowd-sourced funding, hatchery-reared juvenile C. auratus, originally 609

bred for scientific purposes (Partridge et al. 2017), were released into the embayments on the 610

premise that stocks would benefit. Government subsequently committed additional funding 611

Page 25 of 58



Draft

26

for further releases of hatchery-reared juveniles in 2017 and 2018. The otoliths of released 612

fish are marked with alizarin red via immersion, allowing them to be identified at a later stage 613

if collected via C. auratus stock assessment sampling regimes (see Fairclough et al. 2014b). 614

However, the probability of this is low and there were no plans to evaluate the release 615

program with respect to its purported benefits to stocks or recreational fishing. The releases 616

are promoted as a mechanism to ensure ongoing occurrence of C. auratus in Cockburn 617

Sound, but given juveniles leave this embayment by ~18 months of age, disperse widely 618

along the coast and may not return, as occurs elsewhere in Australia, such benefits may never 619

be realised (Lenanton 1974; Fowler et al. 2005; Hamer et al. 2011; Wakefield et al. 2011). 620

Furthermore, such enhancement was not required to recover stocks of C. auratus on the upper 621

west coast of Australia, where traditional management methods, such as closures to fishing, 622

were adopted, accepted and successful (Jackson and Moran 2012, Molony et al. 2003). Thus, 623

education on the biology of C. auratus and the realistic benefits of stock enhancement vs 624

traditional input/output management controls is needed. 625

There are currently no data to demonstrate any decline in abundance of aggregating 626

C. auratus on the lower west coast since the fishing closure was introduced, which was 627

supported strongly by the recreational sector. Furthermore, the extent and effect of 628

recreational fishing of migratory C. auratus have not been measured. Surveys of private boat-629

based recreational fishers’ catch and effort are conducted at larger spatial scales and are more 630

relevant to the broader stock on the west coast (Ryan et al. 2017). These surveys have not 631

detected a significant change in the total retained catches of C. auratus by private recreational 632

fishers between 2010/11 and 2015/16, but estimated catches have been at levels around the 633

predicted maximum that would allow stocks of C. auratus on the west coast to recover (Ryan 634

et al. 2017). However, the stocks are yet to recover and continued management is needed 635

(Fairclough and Holtz 2017). In addition, the number of private boat-based recreational 636

Page 26 of 58



Draft

27

fishing licences in the metropolitan region, which is adjacent to the embayments, has 637

increased by ~20% from 2010/11 to 2015/16, i.e. from ~57 000 to 69 000. Also, the numbers 638

of C. auratus released by recreational fishers is very high, i.e. ~78% of fish landed in the 639

Metropolitan Area in 2015/16, where a majority of the retained recreational catch is taken 640

(Ryan et al. 2017). If post-release mortality is high in C. auratus, caused by factors such as 641

barotrauma, gut-hooking and on-board handling (e.g. Stewart 2008; Grixti et al. 2010), or 642

there are sub-lethal effects, such as skipped spawning, recovery of stocks or increases in 643

biomass of aggregations may be restricted or prevented. 644

The social importance of the C. auratus spawning aggregations and their likely 645

importance to its broader stocks along the west coast of Australia warrants further 646

investigation and potential direct monitoring. This could benefit both management and the 647

social desire for the aggregations to be sustained. A variety of methods may be applicable, 648

but may vary in their cost-effectiveness. Assessment of age-based fishing mortality rates 649

requires the collection of large numbers of fish from the aggregations, which would have 650

undesirable impacts on biomass and reproductive behaviour. Mark-recapture models could be 651

used to estimate fishing mortality, whereby aggregating fish are tagged and released. 652

Estimation of aggregation biomass may be a more direct approach. This possibly could be 653

achieved via daily egg production methods (DEPM), which have been used for C. auratus 654

elsewhere (e.g. Jackson et al. 2012). However, DEPMs require significant laboratory time, as 655

C. auratus eggs must be identified and counted from plankton surveys (Wakefield, 2010). 656

This is challenging due to their visual similarities to eggs of other species, but the use of 657

recently tested molecular identification techniques (real-time PCR and in-situ hybridisation) 658

may help (Dias et al. 2016; Oxley et al. 2016). Acoustic surveys of aggregations using multi- 659

or single-beam technology have been tested on other demersal species in this region and are 660

commonly used elsewhere to estimate biomass (Rose 2003; Gastauer et al. 2017). However, 661

Page 27 of 58



Draft

28

the relatively shallow waters of Cockburn Sound (< ~20 m) would require modifications to 662

the technique, given they are normally used in deeper water. In addition, a dynamic survey 663

design would be required that allows for the potential interannual variation in the timing of 664

peak spawning due to environmental change and thus the potential to bias annual relative 665

spawning biomass estimates. Determining the cost-effectiveness of the various methods and 666

comparing the information they could provide to management in terms of responsiveness to 667

changes in status (e.g. biomass) of the C. auratus aggregations would inform a discussion on 668

which method may be preferable for long-term monitoring. 669

670

This study has explored the relevance of a spatio-temporal closure to fishing for 671

C. auratus migrating to and from spawning aggregation sites in embayments on the lower 672

west coast of Australia. This species exhibits predictable migratory patterns to and from the 673

main embayment and, as a result, such fish are exploited by the recreational sector outside the 674

embayment. Individual variation in the timing of migration and interannual philopatry to the 675

aggregations may reduce the risk of capture. In addition, the current temporal and spatial 676

extent of the closure protects the aggregations specifically, which is widely accepted by 677

fishers. In conjunction with the restrictive management of stocks at the broader west coast 678

scale, overexploitation of the spawning aggregations of C. auratus in lower west coast 679

embayments is likely to have been avoided, but ongoing monitoring is required to ensure this 680

does not occur in the future as the recreational sector grows. In addition, the effect that 681

targeting of migratory fish may have on reproductive success of spawning aggregations may 682

limit the benefits offered by current management, by interruption of migrations and removal 683

of spawning biomass. Thus expansion of the spatial limits of the closure, to include the reef 684

systems adjacent to the embayments where C. auratus are fished, or extension of the closure 685

Page 28 of 58



Draft

29

to include August and September, when fish are beginning to arrive at the main pathway to 686

the embayment, may be required. 687

The importance of C. auratus spawning aggregations to the broader population along 688

the west coast of Australia is assumed, with progeny expected to disperse widely prior to 689

maturing. However, it is unclear to what extent this is the case. Indeed, aggregations in 690

southern Australia have been demonstrated to produce >70% of the recruits to ~800 km of 691

coastline (Hamer et al. 2011). Thus, identifying such source-sink relationships is important 692

and could be explored using contemporary genetic or possibly otolith microchemistry 693

techniques (Hamer et al. 2011; Ovenden et al. 2015; diBattista et al. 2017). Furthermore, such 694

an understanding would help to identify whether it is critical to monitor the status (e.g. 695

spawning biomass) of the spawning aggregations directly from a sustainability perspective, or 696

whether monitoring and appropriate management of fishing mortality of the broader stocks of 697

the west coast is sufficient to ensure the ongoing reproductive function of the aggregations 698

and also satisfy social expectations. 699

700

Acknowledgements 701

This project was supported by the Canada Foundation for Innovation’s international 702

Ocean Tracking Network 703

(http://oceantrackingnetwork.org; https://members.oceantrack.org/project?ccode=CCK) and 704

Western Australia’s Shark Monitoring Network, which each provided acoustic receivers and 705

access to detection data. Gratitude is expressed to staff of the Department of Primary 706

Industries and Regional Development, for assistance in the field, particularly S. Mountford 707

and other SMN staff, to R. McAuley and T. Pinnel for database support; S. Blight for boat-708

ramp camera support; A. Hesp and A. Denham for statistical advice and internal reviewers R. 709

McAuley, G. Jackson and B. Molony. We are extremely grateful to D. Wimpress and 710

Page 29 of 58



http://oceantrackingnetwork.org/

https://members.oceantrack.org/project?ccode=CCK

Draft

30

R. Thipthorp for assistance in tagging, to other recreational fishers who have reported 711

recaptured snapper and to the Department of Transport and Fremantle Ports for assistance in 712

deployment of acoustic arrays. We also thank two anonymous reviewers for their 713

constructive feedback on the manuscript. 714

715

Page 30 of 58



Draft

31

References 716

Abecasis, D., Bentes, L., and Erzini,K. 2009. Home range, residency and movements of 717

Diplodus sargus and Dilpodus vulgaris in a coastal lagoon: Connectivity between nursery 718

and adult habitats. Estuarine, Coastal and Shelf Science 85(4): 525-719

529. doi.org/10.1016/j.ecss.2009.09.001 720

Abecasis, D., Horta e Costa, B., Afonso, P., Gonçalves, E.J., and Erzini, K. 2015. Early 721

reserve effects linked to small home ranges of a commercial fish, Diplodus sargus, 722

Sparidae. Mar. Ecol. Prog. Ser. 518: 255-266. doi.org/10.3354/meps11054 723

Asoh, K., and Yoshikawa, T. 2002. The role of temperature and embryo development time in 724

the diel timing of spawning in a coral-reef damselfish with high-frequency spawning 725

synchrony. Environ. Biol. Fishes 64(4): 379–392. doi.org/10.1023/A:10161775 726

Braccini, M., Rensing, K., Langlois, T., and McAuley, R. 2017. Acoustic monitoring 727

reveals the broad-scale movements of commercially important sharks. Mar. Ecol. Prog. 728

Ser. 577: 121-129. doi.org/10.3354/meps12251 729

Burnham, K.P., Anderson, D.R., 2002. Model Selection and Multimodel Inference: A 730

Practical Information-Theoretic Approach. Springer-Verlag, New York. 731

Coleman, F.C., Koenig, C.C., and Collins, L.A. 1996. Reproductive styles of shallow-water 732

groupers (Pisces: Serranidae) in the eastern Gulf of Mexico and the consequences of 733

fishing spawning aggregations. Environ. Biol. Fishes 47(2): 129–141. 734

doi:10.1007/BF00005035. 735

Colin, P.L. 2012. Timing and location of aggregation and spawning in reef fishes. In Reef 736

fish spawning aggregations: biology, research and management. Fish and Fisheries 737

Series, vol 35. Edited by Y.J. Sadovy and P.L. Colin. Springer, Dordrecht pp. 117–158. 738

Cooke, S.J., Iverson, S.J., Stokesbury, M.J.W., Hinch, S.G., Fisk, A.T., VanderZwaag, D.L., 739

Apostle, R., and Whoriskey, F. 2011. Ocean Tracking Network Canada: A Network 740

Page 31 of 58



https://doi.org/10.1016/j.ecss.2009.09.001

https://doi.org/10.3354/meps11054

Draft

32

Approach to Addressing Critical Issues in Fisheries and Resource Management with 741

Implications for Ocean Governance. Fisheries 36: 583–592. 742

Crossin, G.T., Heupel, M.R., Holbrook, C.M., Hussey, N.E., Lowerre-Barbieri, S.K., 743

Nguyen, V.M., Raby, G.D., and Cooke, S.J. 2017. Acoustic telemetry and fisheries 744

management. Ecol. Appl. 27(4): 1031–1049. doi:10.1002/eap.1533. 745

Department of Fisheries. 2016. Fish kill incident: Cockburn Sound, Western Australia 746

November-December 2015. Department of Fisheries Western Australia, Perth. Available 747

from http://www.fish.wa.gov.au/Documents/corporate_publications/fish-kill-incident-748

cockburn-sound-2015.pdf 749

Dias, P.J., Wakefield, C.B., Fairclough, D.V., Jackson, G., Travers, M.J., and Snow, M., 750

2016. Real-time PCR validation of visually identified snapper Chrysophrys auratus 751

(Sparidae) eggs. J. Fish Biol. 88: 811–819. 752

DiBattista, J.D., Travers, M.J., Moore, G.I., Evans, R.D., Newman, S.J., Feng, M., Moyle, 753

S.D., Gorton, R.J., Saunders, T., and Berry, O. 2017. Seascape genomics reveals fine-scale 754

patterns of dispersal for a reef fish along the ecologically divergent coast of Northwestern 755

Australia. Mol. Ecol. 26(22): 6206-6223. doi.org/10.1111/mec.14352 756

Domeier, M.L. 2012. Revisiting spawning aggregations: Definitions and challenges. In Reef 757

Fish Spawning Aggregations: Biology, Research and Management. Fish and Fisheries 758

Series, vol 35. Edited by Y.J. Sadovy and P.L. Colin. Springer, Dordrecht. pp. 1–20. 759

doi:10.1007/978-94-007-1980-4_1. 760

Domeier, M.L., and Colin, P.L. 1997. Tropical Reef Fish Spawning Aggregrations: Defined 761

and Reviewed. Bull. Mar. Sci. 60(3): 698–726. 762

Donaldson, M.R., Hinch, S.G., Suski, C.D., Fisk, A.T., Heupel, M.R., and Cooke, S.J. 2014. 763

Making connections in aquatic ecosystems with acoustic telemetry monitoring. Front. 764

Ecol. Environ. 12(10): 565–573. doi:10.1890/130283. 765

Page 32 of 58



http://www.fish.wa.gov.au/Documents/corporate_publications/fish-kill-incident-cockburn-sound-2015.pdf

http://www.fish.wa.gov.au/Documents/corporate_publications/fish-kill-incident-cockburn-sound-2015.pdf

https://doi.org/10.1111/mec.14352

Draft

33

Erisman, B., Aburto-Oropeza, O., Gonzalez-Abraham, C., Mascareñas-Osorio, I., Moreno-766

Báez, M., and Hastings, P.A. 2012. Spatio-temporal dynamics of a fish spawning 767

aggregation and its fishery in the Gulf of California. Sci. Rep. 2(1): 284. 768

doi:10.1038/srep00284. 769

Erisman, B., Heyman, W., Kobara, S., Ezer, T., Pittman, S., Aburto-Oropeza, O., and 770

Nemeth, R.S. 2017. Fish spawning aggregations: where well-placed management actions 771

can yield big benefits for fisheries and conservation. Fish Fish. 18(1): 128–144. 772

doi:10.1111/faf.12132. 773

Fairclough, D. V, and Holtz, M. 2017. West Coast Demersal Scalefish Resource Status 774

Report 2016. In: Status Reports of the Fisheries and Aquatic Resources of Western 775

Australia 2015/16: The State of the Fisheries Edited by W.J. Fletcher, M.D. Mumme and 776

F.J. Webster. Department of Fisheries Western Australia, Perth. pp. 66-71. Available 777

from http://www.fish.wa.gov.au/Documents/sofar/status_reports_of_the_fisheries_and_aq778

uatic_resources_2015-16.pdf 779

Fairclough, D. V., Edmonds, J.S., Jackson, G., Lenanton, R.C.J., Kemp, J., Molony, B.W., 780

Keay, I.S., Crisafulli, B.M., and Wakefield, C.B. 2013. A comparison of the stock 781

structures of two exploited demersal teleosts, employing complementary methods of 782

otolith element analysis. J. Exp. Mar. Bio. Ecol. 439: 181–195. 783

doi:10.1016/j.jembe.2012.10.023. 784

Fairclough, D. V., Dimmlich, W.F., and Potter, I.C. 2000. Length and age compositions and 785

growth rates of the Australian herring Arripis georgiana in different regions. Mar. 786

Freshwater Res. 51(6): 631–640. doi:10.1071/MF99118. 787

Fairclough, D. V, Lek, I.C., Bivoltsis, and Babcock, A.K. 2011. The fish communities and 788

main fish populations of the Jurien Bay Marine Park. Murdoch University, Perth. 789

Available from 790

Page 33 of 58



http://www.fish.wa.gov.au/Documents/sofar/status_reports_of_the_fisheries_and_aquatic_resources_2015-16.pdf

http://www.fish.wa.gov.au/Documents/sofar/status_reports_of_the_fisheries_and_aquatic_resources_2015-16.pdf

Draft

34

http://researchrepository.murdoch.edu.au/id/eprint/6573/1/fish_populations_of_Jurien_Ba791

y.pdf [accessed 27 September 2017]. 792

Fairclough, D.V., Brown, J.I., Carlish, B.J., Crisafulli, B.M., and Keay, I.S. 2014b. Breathing 793

life into fisheries stock assessments with citizen science, Sci. Rep. 4: 7249. 794

Fairclough, D. V, Molony, B.W., Crisafulli, B.M., Keay, I.S., Hesp, S.A., and Marriott, R.J. 795

2014a. Status of demersal finfish stocks on the west coast of Australia. Fisheries Research 796

Report No. 253. Department of Fisheries Western Australia, Perth. Available from 797

http://www.fish.wa.gov.au/Documents/research_reports/frr253.pdf [accessed 28 August 798

2017]. 799

Fowler, A. J., and Jennings, P. R. 2003. Dynamics in 0+ recruitment and early life history for 800

snapper (Pagrus auratus, Sparidae) in South Australia. Mar. Freshw. Res. 54: 941–956. 801

Fowler, A.J., Gillanders, B.M., Hall, K.C. 2005. Relationship between elemental 802

concentration and age from otoliths of adult snapper (Pagrus auratus, Sparidae): 803

implications for movement and stock structure. Mar. Freshw. Res. 56(5): 661–676. 804

doi.org/10.1071/MF04157 805

Fowler, A.J., Huveneers, C., Lloyd, M.T. 2017. Insights into movement behaviour of snapper 806

(Chrysophrys auratus, Sparidae) from a large acoustic array. Mar. Freshw. Res. 68(8): 807

1438-1453. doi.org/10.1071/MF16121 808

Francis, M.P. 1993. Does water temperature determine year class strength in New Zealand 809

snapper (Pagrus auratus)? Fish. Oceanog. 2(2): 65–72. doi: 10.1111/j.1365-810

2419.1993.tb00121.x 811

Gastauer, S., Scoulding, B., Parsons, M., 2017. Estimates of variability of goldband snapper 812

target strength and biomass in three fishing regions within the Northern Demersal 813

Scalefish Fishery (Western Australia). Fish. Res. 193: 250–262. 814

10.1016/j.fishres.2017.05.001 815

Page 34 of 58



Draft

35

Gardner, M.J., Chaplin, J.A., Potter, I., Fairclough, D. V., and Jackson, G. 2017. The genetic 816

structure of a marine teleost, Chrysophrys auratus, in a large, heterogeneous marine 817

embayment. Environ. Biol. Fishes. 100(11): 1411-1425. doi:10.1007/s10641-017-0652-8. 818

Grixti, D., Conron, S.D., Morison, A. (2010). Post-release survival of recreationally caught 819

snapper, Pagrus auratus, in Port Phillip Bay, south-eastern Australia. Fish. Manag. Ecol. 820

17(1): 1-9. doi.org/10.1111/j.1365-2400.2009.00704.x 821

Grüss, A., Robinson, J., Heppell, S.S., Heppell, S.A., and Semmens, B.X. 2014. Conservation 822

and fisheries effects of spawning aggregation marine protected areas: What we know, 823

where we should go, and what we need to get there. ICES J. Mar. Sci. 71(7): 1515–1534. 824

doi:10.1093/icesjms/fsu038. 825

Hamer, P.A., Acevedo, S., Jenkins, G.P., and Newman, A. 2011. Connectivity of a large 826

embayment and coastal fishery: Spawning aggregations in one bay source local and broad-827

scale fishery replenishment. J. Fish Biol. 78(4): 1090–1109. doi:10.1111/j.1095-828

8649.2011.02921.x. 829

Hamilton, R., De Mitcheson, Y.S., and Aguilar-Perera, A. 2012. The role of local ecological 830

knowledge in the conservation and management of reef fish spawning aggregations. In 831

Reef Fish Spawning Aggregations: Biology, Research and Management. Fish and 832

Fisheries Series, vol 35. Edited by Y.J. Sadovy and P.L. Colin. Springer, Dordrecht. pp. 833

331–369. doi:10.1007/978-94-007-1980-4_10. 834

Harasti, D., Lee, K.A., Gallen, C., Hughes, J.M., Stewart, J. 2015. Movements, home range 835

and site fidelity of snapper (Chrysophrys auratus) within a temperate marine protected 836

area. PLoS ONE 10(11): e0142454. doi.org/10.1371/journal.pone.0142454 837

Heupel, M.R., Semmens, J.M., and Hobday, A. J. 2006. Automated acoustic tracking of 838

aquatic animals: Scales, design and deployment of listening station arrays. Mar. 839

Freshwater Res. 57: 113. doi:10.1071/MF05091. 840

Page 35 of 58



Draft

36

Heyman, W.D., and Granados-Dieseldorff, P. 2012. The voice of the fishermen of the Gulf of 841

Honduras: Improving regional fisheries management through fisher participation. Fish. 842

Res. 125–126: 129–148. doi:10.1016/j.fishres.2012.02.016. 843

How, J.R., and de Lestang, S. 2012. Acoustic tracking: issues affecting design, analysis and 844

interpretation of data from movement studies. Mar. Freshwater Res. 63(4): 312-324. 845

doi:10.1071/MF11194. 846

Jackson, G., and Moran, M. 2012. Recovery of inner Shark Bay snapper (Pagrus auratus) 847

stocks: relevant research and adaptive recreational fisheries management in a World 848

Heritage Property. Mar. Freshw. Res. 63(11): 1180–1190. 849

Jackson, G., Cheung, Y.W., and Wakefield, C.W. 2012. An evaluation of the daily egg 850

production method to estimate spawning biomass of snapper (Pagrus auratus) in inner 851

Shark Bay, Western Ausralia, following more than a decade of surveys 1997-2007. Fish. 852

Res. 117-118: 22-34. doi:10.1016/j.fishres.2010.12.009 853

Jackson, G., Fairclough, D., and Wakefield, C. 2015. Snapper (Chrysophrys auratus). In N. 854

Caputi, M. Feng, A. Pearce, J. Benthuysen, A. Denham, Y. Hetzel, R. Matear, G. Jackson, 855

B. Molony, L. Joll and A. Chandrapavan. 2015. Management implications of climate 856

change effect on fisheries in Western Australia Part 2: Case studies. FRDC Project 857

2010/535. Fisheries Research Report No. 261, Department of Fisheries Western Australia, 858

Perth. pp. 123-133 859

Jackson, G., Ryan, K.L., Green, T.J., Pollock, K.H., and Lyle, J.M. 2016. Assessing the 860

effectiveness of harvest tags in the management of a small-scale, iconic marine 861

recreational fishery in Western Australia. ICES J. Mar. Sci. 73(10): 2666–2676. 862

Kawabata, Y., Nanami, A., Yamamoto, K., Sato, T., Kuwahara, K., Koga, M., Kawaguchi, 863

K., Yamaguchi, T., Ohta, I. Kawabe, R., Nishihara, G.N., Yagi, M. and Soyano, K. 2015. 864

Page 36 of 58



Draft

37

Duration of migration and reproduction in males is dependent on energy reserve in a fish 865

forming spawning aggregations. Mar. Ecol. Prog. Ser. 534:149-161 866

Lam, T.J. 1983. Environmental influences on gonadal activity in fish. In Reproduction 867

Behavior and Fertility Control. Edited by W.S. Hoar, D.J. Randall and E.M. Donaldson. 868

Fish Physiology Vol. 9, Part B. Academic Press, New York. pp 65–116. 869

Lawson, G.L., and Rose, G.A. 2000. Small-scale spatial and temporal patterns in spawning of 870

Atlantic cod (Gadus morhua) in coastal Newfoundland waters. Can. J. Fish. Aquat. Sci. 871

57(5): 1011–1024. doi.org/10.1139/f00-026 872

Lenanton, R.C.J. 1974. The abundance and size composition of trawled juvenile snapper 873

Chrysophrys unicolor (Quoy and Gaimard) from Cockburn Sound, Western Australia. 874

Aust. J. Mar. Freshwat. Res. 25(2): 281-285. 875

Lenanton, R.C.J., Dowling, C.E., Smith, K.A., Fairclough, D.V. and Jackson, G. 2017. 876

Potential influence of a marine heatwave on range extensions of tropical fishes in the 877

eastern Indian Ocean – invaluable contributions from amateur observers. Regional Studies 878

in Marine Science 13: 19–31. 879

Lenanton, R., StJohn, J., Keay, I. Wakefield, C., Jackson, G., Wise, B. and Gaughan, D. 880

2009. Spatial scales of exploitation among populations of demersal scalefish: implications 881

for management. Part 2. Stock structure and biology of two indicator species, West 882

Australian dhufish (Glaucosoma hebraicum) and pink snapper (Pagrus auratus), in the 883

West Coast Bioregion. Final FRDC report, Project 2003/052. Department of Fisheries, 884

Perth. http://www.fish.wa.gov.au/Documents/research_reports/frr174.pdf 885

Mann, B.Q., Buxton, C.D., Pollard, D., Carpenter, K.E. and Winker, H. 2014a. Petrus 886

rupestris. The IUCN Red List of Threatened Species 2014: e.T170247A1300995. 887

Available from http://dx.doi.org/10.2305/IUCN.UK.2014-3.RLTS.T170247A1300995.en 888

Page 37 of 58



http://www.fish.wa.gov.au/Documents/research_reports/frr174.pdf

Draft

38

Mann, B.Q., Buxton, C.D., Pollard, D., Carpenter, K.E. and Sadovy, Y. 2014b. Polysteganus 889

undulosus. The IUCN Red List of Threatened Species 2014: e.T170206A1293194. 890

Available from http://dx.doi.org/10.2305/IUCN.UK.2014-3.RLTS.T170206A1293194.en 891

McAuley, R., Bruce, B., Keay, I., Mountford, S., and Pinnell, T. 2016. Evaluation of passive 892

acoustic telemetry approaches for monitoring and mitigating shark hazards off the coast of 893

Western Australia. Fisheries Research Report No. 273. Department of Fisheries Western 894

Australia, Perth. Available from 895

http://www.fish.wa.gov.au/Documents/research_reports/frr273.pdf [accessed 19 June 896

2017]. 897

Meyer, C.G., Papastamatiou, Y.P., and Clark, T.B. 2010. Differential movement patterns and 898

site fidelity among trophic groups of reef fishes in a Hawaiian marine protected area. Mar. 899

Biol. 157(7): 1499–1511. doi:10.1007/s00227-010-1424-6. 900

Molony, B.W., Lenanton, R., Jackson, G., and Norriss, J. 2003. Stock enhancement as a 901

fisheries management tool. Rev. Fish Biol. Fish. 13(4): 409–432. 902

Moran, M., Burton, C., Jenke, J., 2003. Long-term movement patterns of continental shelf 903

and inner gulf snapper (Pagrus auratus, Sparidae) from tagging in the Shark Bay region of 904

Western Australia. Mar. Freshwat. Res. 54(8): 913–922. doi.org/10.1071/MF03012 905

Munday, P.L., and Wilson, S.K. 1997. Comparative efficacy of clove oil and other chemicals 906

in anaesthetization of Pomacentrus amboinensis, a coral reef fish. J. Fish Biol. 51(5): 931–907

938. doi:10.1006/jfbi.1997.0498. 908

Nanami, A., Kawabata, Y., Sato, T., Yamaguchi, T., Kawabe, R., and Soyano, K. 2014. 909

Spawning migration and returning behavior of white-streaked grouper Epinephelus ongus 910

determined by acoustic telemetry. Mar. Biol. 161 (3): 669−680. 911

Page 38 of 58



Draft

39

National Research Council (US). 2011. Committee for the Update of the Guide for the Care 912

and Use of Laboratory Animals. 8th ed. National Academies Press (US), Washington, 913

D.C. 914

Nemeth, R.S. 2012. Ecosystem aspects of species that aggregate to spawn. In Reef fish 915

spawning aggregations: biology, research and management. Fish and Fisheries Series, vol 916

35. Edited by Y.J. Sadovy and P.L. Colin. Springer, Dordrecht. pp. 21–55. 917

Norriss, J., and Crisafulli, B.M. 2010. Longevity in Australian snapper Pagrus auratus 918

(Sparidae). Journal of the Royal Society of Western Australia 93(3): 129–132. 919

Oliver, E.C.J., Donat, M.G., Burrows, M.T., Moore, P.J., Smale, D.A., Alexander, L.V., 920

Benthuysen, J.A., Feng, M., Gupta, A.S., Hobday, A.J., Holbrook, N.J., Perkens-921

Kilpatrick, S.E., Scannell, H.A., Straub, S.C. and Wernberg, T. 2018. Longer and more 922

frequent marine heatwaves over the past century. Nat. Commun. 9: 1324, 923

doi:10.1038/s41467-018-03732-9 924

Ovenden, J.R., Berry, O., Welch, D.J., Buckworth, R.C., and Dichmont, C.M. 2015. Ocean’s 925

eleven: a critical evaluation of the role of population, evolutionary and molecular genetics 926

in the management of wild fisheries. Fish Fish. 16(1): 125–159. doi:10.1111/faf.12052 927

Oxley, A.P.A., Catalano, S.R., Wos-Oxley, M.L., Westlake, E.L., Grammer, G.L., and Steer, 928

M.A. 2017. Using in situ hybridization to expand the daily egg production method to new 929

fish species. Mol. Ecol. Resour. 17(6): 1108-1121. doi: 10.1111/1755-0998.12707 930

Parsons, D.M., Morrison, M. A., MacDiarmid, A.B., Stirling, B., Cleaver, P., Smith, I.W.G., 931

and Butcher, M. 2009. Risks of shifting baselines highlighted by anecdotal accounts of 932

New Zealand’s snapper (Pagrus auratus) fishery. New Zeal. J. Mar. Freshwater Res. 933

43(4): 965–983. doi:10.1080/00288330909510054. 934

Page 39 of 58



Draft

40

Parsons, D.M., Morrison, M. A., and Slater, M.J. 2010. Responses to marine reserves: 935

Decreased dispersion of the sparid Pagrus auratus (snapper). Biol. Conserv. 143(9): 936

2039–2048. doi:10.1016/j.biocon.2010.05.009. 937

Parsons, D.M., Sim-Smith, C.J., Cryer, M., Francis, M.P., Hartill, B., Jones, E.G., Le Port, 938

A., Lowe, M., McKenzie, J., Morrison, M., Paul, L.J., Radford, C., Ross, P.M., Spong, 939

K.T., Trnski, T., Usmar, N., Walsh, C., Zeldis, J. 2014. Snapper (Chrysophrys auratus): a 940

review of life history and key vulnerabilities in New Zealand. New Zeal. J. Mar. 941

Freshwater Res. 48(2): 256-283. 942

Partridge, G.J., Ginbey, B.M., Woolley, L.D., Fairclough, D. V., Crisafulli, B., Chaplin, J., 943

Prokop, N., Dias, J., Bertram, A., and Jenkins, G.I. 2017. Development of techniques for 944

the collection and culture of wild-caught fertilised snapper (Chrysophrys auratus) eggs for 945

stock enhancement purposes. Fish. Res. 186: 524–530. doi:10.1016/j.fishres.2016.08.025. 946

Pearce, A., Hutchins, B., Hoschke, A., and Fearns, P. 2016. Record high damselfish 947

recruitment at Rottnest Island, Western Australia, and the potential for climate-induced 948

range extension. Reg. Stud. Mar. Sci. 8: 77–88. doi:10.1016/j.rsma.2016.09.009. 949

R Core Team. 2017. R: A language and Environment for Statistical Computing. R 950

Foundation for Statistical Computing. Vienna. ISBN 3-900051-07-9. http://www.R-951

project.org. 952

Randall, J.E., Allen, G.R. Steene, R. 1997. Fishes of the Great Barrier Reef and Coral Sea. 953

University of Hawaii Press, Honolulu. 954

Rose, G.A. 2003. Monitoring coastal northern cod: towards an optimal survey of Smith 955

Sound, Newfoundland. ICES J. Mar. Sci. 60(3): 453-462. doi.org/10.1016/S1054-956

3139(03)00044-4 957

Ryan, K., Wise, B., Hall, N., Pollock, K., Sulin, E., and Gaughan, D. 2013. An integrated 958

system to survey boat-based recreational fishing in Western Australia 2011/12. Fisheries 959

Page 40 of 58



Draft

41

Research Report No. 249. Department of Fisheries Western Australia, Perth. Available 960

from: http://www.fish.wa.gov.au/Documents/research_reports/frr249.pdf 961

Ryan, K. L., Hall, N. G., Lai, E. K., Smallwood, C. B., Taylor, S. M., and Wise, B. S. 2015. 962

State-wide survey of boat-based recreational fishing in Western Australia 2013/14. 963

Fisheries Research Report No. 268. Department of Fisheries Western Australia, Perth. 964

Available from: http://www.fish.wa.gov.au/Documents/research_reports/frr268.pdf 965

Ryan, K. L., Hall, N. G., Lai, E. K., Smallwood, C. B., Taylor, S. M., and Wise, B. S. 2017. 966

State-wide survey of boat-based recreational fishing in Western Australia 2015/16. 967

Fisheries Research Report No. 287. Department of Primary Industries and Regional 968

Development, Western Australia, Perth. Available 969

from: http://www.fish.wa.gov.au/Documents/research_reports/frr287.pdf 970

Sadovy de Mitcheson, Y. 2016. Mainstreaming Fish Spawning Aggregations into Fishery 971

Management Calls for a Precautionary Approach. Bioscience 66(4): 295–306. 972

doi:10.1093/biosci/biw013. 973

Sadovy de Mitcheson, Y., and Colin, P.L. (Editors). 2012. Reef Fish Spawning 974

Aggregations: Biology, Research and Management. Fish and Fisheries Series, vol 35. 975

Springer, Dordrecht. doi:10.1007/978-94-007-1980-4. 976

Sadovy de Mitcheson, Y., Cornish, A., Domeier, M., Colin, P.L., Russell, M., and Lindeman, 977

K.C. 2008. A global baseline for spawning aggregations of reef fishes. Conserv. Biol. 978

22(5): 1233–1244. doi:10.1111/j.1523-1739.2008.01020.x. 979

Sadovy de Mitcheson, Y., and Erisman, B. 2012. Fishery and biological implications of 980

fishing spawning aggregations, and the social and economic importance of aggregating 981

fishes. In Reef Fish Spawning Aggregations: Biology, Research and Management. Fish 982

and Fisheries Series, vol 35. Edited by Y.J. Sadovy and P.L. Colin. Springer, Dordrecht. 983

pp. 225–284. doi:10.1007/978-94-007-1980-4_8. 984

Page 41 of 58





Draft

42

Sadovy, Y., and Cheung, W.L. 2003. Near extinction of a highly fecund fish: the one that 985

nearly got away. Fish Fish. 4(1): 86–99. doi:10.1046/j.1467-2979.2003.00104.x. 986

Sancho, G., Solow, A.R., and Lobel, P.S. 2000. Environmental influences on the diel timing 987

of spawning in coral reef fishes. Mar. Ecol. Prog. Ser. 206: 193-202. 988

doi:10.2307/24863764. 989

Saunders, R.J., Fowler, A.J., and Gillanders, B.M. 2012. The spawning dynamics of snapper 990

(Chrysophrys auratus) in northern Spencer Gulf, South Australia. New Zeal. J. Mar. 991

Fresh. 46(4): 491-510. doi.org/10.1080/00288330.2012.709522 992

Scott S. G., Zeldis J. R., and Pankhurst N. W. 1993. Evidence of daily spawning in natural 993

populations of the New Zealand snapper Pagrus auratus (Sparidae). Env. Biol. Fish. 994

36(2): 149–156. doi.org/10.1007/BF00002794 995

Sheaves, M. 2006. Is the timing of spawning in sparid fishes a response to sea temperature 996

regimes? Coral Reefs 25(4): 655–669. doi.org/10.1007/s00338-006-0150-5 997

Simpfendorfer, C.A., Huveneers, C., Steckenreuter, A., Tattersall, K., Hoenner, X., 998

Harcourt, R. and Heupel, M.R. 2015. Ghosts in the data: false detections in VEMCO 999

pulse position modulation acoustic telemetry monitoring equipment. Anim. Biotelem. 1000

3: 55. doi.org/10.1186/s40317-015-0094-z 1001

Sims, D.W., Genner, M.J., Southward, A.J., and Hawkins, S.J. 2001. Timing of squid 1002

migration reflects North Atlantic climate variability. Proc. R. Soc. London B Biol. Sci. 1003

268(1485): 2607–2611. 1004

Sims, D.W., Wearmouth, V.J., Genner, M.J., Southward, A.J., and Hawkins, S.J. 2004. Low-1005

temperature-driven early spawning migration of a temperate marine fish. J. Anim. Ecol. 1006

73(2): 333–341. doi:10.1111/j.0021-8790.2004.00810.x. 1007

Page 42 of 58



https://doi.org/10.1186/s40317-015-0094-z

Draft

43

Sim-Smith, C.J., Jeffs, A.G., and Radford, C.A. 2012. Localised spawning omission in 1008

snapper, Chrysophrys auratus (Sparidae). Mar. Freshw. Res. 63(2): 150-159. 1009

doi.org/10.1071/MF11203 1010

Smith, K., Brown, J., Lewis, P., Dowling, C., Howard, A., Lenanton, R., and Molony, B. 1011

2013. Status of nearshore finfish stocks in south-western Western Australia. Part 1: 1012

Australian herring. Fisheries Research Report No. 246. Department of Fisheries Western 1013

Australia, Perth. Available from 1014

http://www.fish.wa.gov.au/Documents/research_reports/frr246.pdf [accessed 29 1015

September 2017]. 1016

Starr, R., Sala, E., Ballesteros, E., and Zabala, M. 2007. Spatial dynamics of the Nassau 1017

grouper Epinephelus striatus in a Caribbean atoll. Mar. Ecol. Prog. Ser. 343: 239–249. 1018

doi:10.3354/meps06897. 1019

Steffe, A. S., Taylor, S. M., Blight, S. J., Ryan, K. L., Desfosses, C. J., Tate, A. C., 1020

Smallwood, C. B., Lai, E. K., Trinnie, F. I. and Wise, B. S. 2017. Framework for 1021

Integration of Data from Remotely Operated Cameras into Recreational Fishery 1022

Assessments in Western Australia. Fisheries Research Report No. 286. Department of 1023

Primary Industries and Regional Development, Western Australia. Available 1024

from http://www.fish.wa.gov.au/Documents/research_reports/frr286.pdf [accessed 31 1025

March 2018]. 1026

Stewart, J., 2008. Capture depth related mortality of discarded snapper (Pagrus auratus) and 1027

implications for management. Fish. Res. 90(1-3): 289–295. 1028

doi.org/10.1016/j.fishres.2007.11.003 1029

Wakefield, C.B. 2010. Annual, lunar and diel reproductive periodicity of a spawning 1030

aggregation of snapper Pagrus auratus (Sparidae) in a marine embayment on the lower 1031

Page 43 of 58




Draft

44

west coast of Australia. J. Fish Biol. 77(6): 1359–1378. doi:10.1111/j.1095-1032

8649.2010.02756.x. 1033

Wakefield, C.B., Fairclough, D. V., Lenanton, R.C.J., and Potter, I.C. 2011. Spawning and 1034

nursery habitat partitioning and movement patterns of Pagrus auratus (Sparidae) on the 1035

lower west coast of Australia. Fish. Res. 109(2–3): 243–251. 1036

doi:10.1016/j.fishres.2011.02.008. 1037

Wakefield, C.B., Lewis, P.D., Coutts, T.B., Fairclough, D.V., and Langlois, T.J. 2013a. Fish 1038

assesmblages associated with natural and anthropogenically-modified habitats in a marine 1039

embayment: comparison of baited videos and opera-house traps. PLOS One 8(3): e59959. 1040

doi:10.1371/journal.pone.0059959 1041

Wakefield, C.B., Newman, S.J., Marriott, R.J., Boddington, D.K., and Fairclough, D. V. 1042

2013b. Contrasting life history characteristics of the eightbar grouper Hyporthodus 1043

octofasciatus (Pisces: Epinephelidae) over a large latitudinal range reveals spawning 1044

omission at higher latitudes. ICES J. Mar. Sci. 70(3): 485–1045

497. doi.org/10.1093/icesjms/fst020 1046

Wakefield, C.B., Potter, I.C., Hall, N.G., Lenanton, R.C.J., and Hesp, S.A. 2015. Marked 1047

variations in reproductive characteristics of snapper (Chrysophrys auratus, Sparidae) and 1048

their relationship with temperature over a wide latitudinal range. ICES J. Mar. Sci. 72(8): 1049

2341–2349. doi:10.1093/icesjms/fsv108. 1050

Wakefield, C.B., Potter, I.C., Hall, N.G., Lenanton, R.C.J., and Hesp, and Hesp, S.A. 2017. 1051

Timing of growth zone formations in otoliths of the snapper, Chrysophrys auratus, in 1052

subtropical and temperate waters differ and growth follows a parabolic relationship with 1053

latitude. ICES J. Mar. Sci. 74(1): 180-192. doi.org/10.1093/icesjms/fsw137 1054

Wise, B.S., St. John, J., and Lenanton, R.C. 2007. Spatial scales of exploitation among 1055

populations of demersal scalefish: implications for management. Part 1: stock status of the 1056

Page 44 of 58



https://doi.org/10.1093/icesjms/fst020

https://doi.org/10.1093/icesjms/fsw137

Draft

45

key indicator species for the demersal scalefish fishery in the West Coast Bioregion. Final 1057

FRDC report - project 2003/052. Fisheries Research Report No. 163. Department of 1058

Fisheries Western Australia, Perth. Available from 1059

http://www.fish.wa.gov.au/Documents/research_reports/frr163.pdf [accessed 28 August 1060

2017]. 1061

Zeller, D.C. 1998. Spawning aggregations: Patterns of movement of the coral trout 1062

Plectropomus leopardus (Serranidae) as determined by ultrasonic telemetry. Mar. Ecol. 1063

Prog. Ser. 162: 253–263. doi:10.3354/meps162253. 1064

Page 45 of 58



Draft

46

Figure captions 1065

Fig. 1. Study location in the inshore Metropolitan Area (cross-hatched on inset) of the West 1066

Coast Bioregion (WCB) of Western Australia. Inset shows main fishery areas of Kalbarri 1067

(K), Mid-West (MW) and South-West (SW) and Shark Monitoring Network arrays outside 1068

the Metropolitan Area (1,2,3). Main map shows (i) acoustic receiver arrays comprising the 1069

aggregation array (○) around Cockburn (CS) and Warnbro (WS) sounds and Owen 1070

Anchorage (OA), the Ocean Tracking Network (▲) and the Shark Monitoring Network 1071

(VR2W ■ and VR4W □) in the Metropolitan Area, (ii) the area closed to fishing for 1072

Chrysophrys auratus from 1 Oct-31 Jan (diagonal hatching), (iii) the five pathways to/from 1073

Cockburn Sound (arrows). OTN Receivers (▲) west of Rottnest Island extend to ~115°13´E 1074

(not all shown). A, tagging location and double gate array between Garden Island (GI) and 1075

Carnac Island (CI); B and C are tagging locations within Cockburn Sound; F, Fremantle; RI, 1076

Rottnest Island; S, Straggler Rocks; WP, Woodman Point. Map produced in ArcGIS v10.3 1077

(ESRI 2014). 1078

1079

Fig. 2. (a) Recapture frequency of Chrysophrys auratus at different distances from tagging 1080

locations within the embayments and (b) frequency of recaptures in each month inside 1081

(black) and outside (grey) the area annually closed to fishing. Mark-recapture data derived 1082

with permission from Wakefield et al. (2011) for 2003-2009 is combined with ongoing data 1083

from the Department of Primary Industries and Regional Development for 2009-2016. 1084

1085

Fig. 3. Length-frequency distributions of Chrysophrys auratus tagged with dart tags (grey) 1086

and acoustic tags (white) in Cockburn Sound during the spawning seasons in 2009 and 2010. 1087

1088

Page 46 of 58



Draft

47

Fig. 4. Proportion of total detections per month (combined across years) of Chrysophrys 1089

auratus at each receiver in the aggregation array, Ocean Tracking Network and Shark 1090

Monitoring Network calculated as 𝑛𝑛 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑖𝑖𝑑𝑑𝑛𝑛𝑑𝑑 𝑝𝑝𝑑𝑑𝑝𝑝 𝑚𝑚𝑑𝑑𝑛𝑛𝑑𝑑ℎ 𝑝𝑝𝑑𝑑𝑝𝑝 𝑝𝑝𝑑𝑑𝑑𝑑𝑖𝑖𝑑𝑑𝑟𝑟𝑑𝑑𝑝𝑝𝑁𝑁 𝑑𝑑𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑖𝑖𝑑𝑑𝑛𝑛𝑑𝑑

; nfish = number of fish 1091

detected per month; %III/IV = percentage of female C. auratus in each month with ovaries in 1092

pre-spawning (III) and spawning (IV) condition (from Wakefield et al. 2015). Red shading 1093

indicates the spatial area closed to fishing for Chrysophrys auratus from 1 Oct-31 Jan. CI = 1094

Carnac Island, CS = Cockburn Sound, F = Fremantle, GI = Garden Island, OA = Owen 1095

Anchorage, RI = Rottnest Island, S = Straggler Rocks, WP = Woodman Point. Maps 1096

produced in ArcGIS v10.3 (ESRI 2014). 1097

1098

Fig. 5. Daily detections of 34 Chrysophrys auratus at site A (black), outside the embayments 1099

(blue) and sites B and C inside Cockburn Sound (red), between the day they were tagged (Oct 1100

2009-Dec 2009 or Sept 2010) and the day they were last detected. Expected end of each 1101

transmitter's life shown as vertical dashes. Grey shading indicates timing of closure to fishing 1102

for C. auratus from 1 Oct-31 Jan. †fish last detected on Shark Monitoring Network array 1 in 1103

Fig. 1; *fish recaptured at latitude 31° S. 1104

1105

Fig. 6. Frequency of occurrence of acoustic transmitters in tagged Chrysophrys auratus 1106

recorded at Site A between Garden and Carnac Islands (see Fig. 1) in each week of the 1107

spawning season in (a) 2010/11 and (b) 2011/12 relative to 1 Oct (Week 0 begins on 1 Oct 1108

when the fishing closure commences), with bimodal mixture models fitted to the frequency 1109

distributions. Mean weekly sea surface temperatures adjacent to Site A and mean weekly 1110

moon fraction (% full) shown above for the same period. Solid arrows identify where SSTs 1111

reached ~19 and 21ºC and dashed arrow identifies the time of the November new moon 1112

(lowest moon fraction) in each year. 1113

Page 47 of 58



Draft

48

1114

Fig. 7. Average percentage of the annual commercial and recreational sector catch of 1115

Chrysophrys auratus in each management area of the West Coast Bioregion; based on the 1116

average annual catch of the main commercial fishery (WCDSIMF) from 2008 to 2016 and 1117

the recreational sector catch, which combines catches of private vessels from the 2013/14 1118

survey of Ryan et al. (2015) and the average annual charter vessel catch from 2008 to 2016. 1119

1120

Fig. 8. Monthly numbers of Chrysophrys auratus retained by charter fishers in 2015/16 in 1121

each 5nm × 5nm block of the Metropolitan Area of the West Coast Bioregion (31 - 33°S). 1122

%III/IV = percentage of female C. auratus in each month with ovaries in pre-spawning (III) 1123

and spawning (IV) condition (created using reproductive data from Wakefield et al. 2015 1124

with permission). Red shading indicates the spatial area closed to fishing for Chrysophrys 1125

auratus from 1 Oct-31 Jan. Nil catch in November is due to an annual closure to recreational 1126

fishing for demersal fishes in the West Coast Bioregion from 15 Oct–15 Dec (inclusive). 1127

Depth contours are 20, 50, 100, 200 m. 1128

1129

Fig. 9. Estimated number of private recreational vessels launched from Woodman Point boat 1130

ramp during each hour of the day (24 h clock) in each month in 2011/12 (from Ryan et al. 1131

2013 with permission). Grey shading indicates closure to fishing for Chrysophrys auratus 1132

from 1 Oct-31 Jan. Bars on right show relative reproductive activity represented as 1133

percentage of female C. auratus in each month that contained ovaries in pre-spawning (III) 1134

and spawning (IV) condition (created using reproductive data from Wakefield et al. 2015 1135

with permission). 1136

1137

Page 48 of 58



Draft

Table 1. Results of adjusted Akaike Information Criterion (AICc) comparisons between

single, bi- and tri-modal mixture distributions fitted to frequencies of occurrence of

Chrysophrys auratus at Site A (Garden Island to Carnac Island pathway) during the 2010/11

and 2011/12 spawning seasons. n = sum of occurrence of C. auratus with transmitters per

week between 1 Aug and 28 Feb. K = number of parameters estimated by the models.

Spawning Season n Model K AICc

2010/11 68 Single mode 2 415

Bimodal 5 407

Tri-modal 8 412

2011/12 87 Single mode 2 516

Bimodal 5 503

Tri-modal 8 507

Table 2. Means (𝑥𝑥𝑖𝑖) and standard deviations (𝑠𝑠𝑖𝑖) of bimodal mixture distributions fitted to

frequencies of occurrence of n Chrysophrys auratus at Site A (Garden Island to Carnac Island

pathway) during the 2010/11 and 2011/12 spawning seasons.

Spawning Season n 𝒙𝒙𝟏𝟏, 𝒔𝒔𝟏𝟏 𝒙𝒙𝟐𝟐, 𝒔𝒔𝟐𝟐

2010/11 15 0.96, 2.34 9.57, 2.56

2011/12 12 0.57, 1.56 8.05, 3.22

Page 49 of 58



Draft

Page 50 of 58



Draft

Page 51 of 58



Draft

Page 52 of 58



Draft

Page 53 of 58



Draft

Page 54 of 58



Draft

Page 55 of 58



Draft

Page 56 of 58



Draft

Page 57 of 58



Draft

Page 58 of 58

