CORE Metadata, citation and similar papers at core.ac.uk

Provided by Research Repository

MURDOCH RESEARCH REPOSITORY

This is the author’s final version of the work, as accepted for publication following peer review but without the publisher’s layout or pagination. The definitive version is available at http://dx.doi.org/10.1016/j.fishres.2011.04.021

Beatty, S., de Graaf, M., Molony, B., Nguyen, V. and Pollock, K.H. (2011) Plasticity in population biology of Cherax cainii (Decapoda: Parastacidae) inhabiting lentic and lotic environments in south-western Australia: Implications for the sustainable management of the recreational fishery. Fisheries Research, 110 (2). pp. 312-324.

http://researchrepository.murdoch.edu.au/4693/

Copyright: © 2011 Elsevier B.V.

It is posted here for your personal use. No further distribution is permitted.

Accepted Manuscript

Title: Plasticity in population biology of Cherax cainii (Decapoda: Parastacidae) inhabiting lentic and lotic environments in south-western Australia: implications for the sustainable management of the recreational fishery

Authors: Stephen Beatty, Martin de Graaf, Brett Molony, Vinh Nguyen, Kenneth Pollock

PII: S0165-7836(11)00170-6 DOI: doi:10.1016/j.fishres.2011.04.021 Reference: FISH 3218

To appear in: Fisheries Research

Received date: 9-11-2010 Revised date: 29-4-2011 Accepted date: 30-4-2011

Please cite this article as: Beatty, S., de Graaf, M., Molony, B., Nguyen, V., Pollock, K., Plasticity in population biology of Cherax cainii (Decapoda: Parastacidae) inhabiting lentic and lotic environments in south-western Australia: implications for the sustainable management of the recreational fishery, Fisheries Research (2010), doi:10.1016/j.fishres.2011.04.021

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. *Highlights

 The study demonstrates considerable variability in the population biology of a large parastacid.  Growth rate and productivity of adult Cherax cainii was greater in the lotic system.  Stunting of adults occurred in the habitat limited lentic population and the stock was over exploited.  Sustainable increases in fishing exploitation may be possible in more pristine rivers.  Habitat and water quality decline and rainfall reductions are the major long-term threats to the fishery.

Accepted Manuscript

Page 1 of 44 *Manuscript including abstract

1 Plasticity in population biology of Cherax cainii (Decapoda: Parastacidae)

2 inhabiting lentic and lotic environments in south-western Australia:

3 implications for the sustainable management of the recreational fishery

4

5 Stephen Beattya,*, Martin de Graafbc, Brett Molonyb, Vinh Nguyenb, Kenneth Pollocka,b

6

7 a*Freshwater Fish Group and Fish Health Unit, Centre for Fish and Fisheries Research, Murdoch

8 University, 90 South St, Murdoch, Western Australia, 6150, Australia

9 bWestern Australian Fisheries and Marine Research Laboratories, Department of Fisheries,

10 Government of Western Australia, PO Box 20 North Beach, Western Australia, 6920, Australia

11 cCurrent address: IMARES Wageningen UR, PO Box 68, 1970 AB IJmuiden, The Netherlands

12 *Corresponding author. Tel.: +61 893602813; fax: +61 893607512. E-mail address:

13 [email protected] (S. Beatty)

14 15

Accepted Manuscript

1

Page 2 of 44 16 Abstract

17 The Smooth Marron Cherax cainii is endemic to south-western Australia and supports an iconic

18 recreational fishery that exploits stocks in both lentic and lotic systems. This study is the first to

19 determine and compare the population biology of a parastacid from both lentic and lotic systems

20 and aimed to gather the information necessary for more effective management of the fishery.

21 Modal progression demonstrated growth rates of juvenile C. cainii were greater in the lentic

22 (Wellington Dam) compared to the lotic (Warren River) system, however, mark-recapture

23 suggested the growth rate of the adult component of the lentic population was stunted whereas in

24 the lotic population 60-90 mm OCL individuals were common and had a faster growth rate. The

25 Wellington Dam stock appeared to be over-exploited and had very low productivity whereas the

26 Warren River stock had relatively low fishing exploitation and a high productivity resulting from

27 higher growth rates coupled with higher population densities (with the trappable population sizes

28 estimated using the open POPAN formulation in the MARK software program). Comparisons of

29 these biological parameters were made with populations elsewhere and there existed a considerable

30 plasticity that is probably due to differences in thermal regimes, degree of habitat complexity, food

31 resource availability and fisher accessibility. The findings demonstrate the need to determine the

32 level of intraspecific biological plasticity in freshwater crayfish in order to sustainably manage their

33 fisheries.

34

35 KEYWORDS: freshwater crayfish, biological plasticity, aquatic habitat, sustainable exploitation,

36 mark-recapture methods, estimation of demographic parameters.

37 Accepted Manuscript

2

Page 3 of 44 38 1. Introduction

39 The South West Coast Drainage Division of Australia is recognised as a global biodiversity

40 hotspot due to an exceptional concentration of endemic species undergoing a major loss of habitat

41 (Myers et al., 2000). This is reflected in the high degree of endemism of the freshwater fauna in the

42 region with 80% of native freshwater fish (Morgan et al., 1998; Allen et al., 2002) and 100% of

43 native freshwater crayfish (Austin and Knott, 1996; Crandall et al., 1999; Austin and Ryan, 2002)

44 being endemic to the region. The region’s aquatic ecosystems have been severely impacted by

45 habitat alterations, particularly secondary salinisation (Halse et al., 2003; Morgan et al., 2003;

46 Beatty et al., 2011), introduced aquatic fishes (Morgan et al., 2002, 2004; Tay et al., 2007) and an

47 eastern Australian freshwater crayfish (Beatty et al., 2005a; Beatty, 2006).

48 The ecosystems of south-western Australia are naturally low in productivity with allochthonous

49 material generally the dominant source of energy (Bunn and Davies, 1990). This low productivity

50 results in the region’s freshwater fishes being relatively small in size (generally <200 mm total

51 length (TL)) and of little value for recreational or commercial fisheries (Morgan et al., 1998; Allen

52 et al., 2002). However, the freshwater crayfish fauna of the region includes the third largest species

53 in the world, the Smooth Marron Cherax cainii Austin and Ryan, 2002. Cherax cainii supports a

54 relatively small (7,400 fishers, 20.8 tonnes of catch in 2007) yet iconic recreational fishery in both

55 lotic (that receive ~70% of the annual effort in terms of total number days fished) and artificial

56 lentic (~30% of annual effort) systems (Molony et al., 2002; de Graaf and Barharthah, 2008).

57 Freshwater crayfishes are known to exhibit considerable plasticity in biological and ecological

58 traits (e.g. Honan and Mitchell, 1995; Payne, 1996; Hayes et al., 2009). This plasticity has been 59 attributed to geneticAccepted (Austin, 1998; Vogt et al., 2008 ) orManuscript environmental (Parkyn et al., 2002; Beatty 60 et al., 2005b) variability as well as the presence of introduced species (Larson and Magoulick, 2008;

61 Hayes et al., 2009). In particular, the major environmental differences that exist between reservoirs

62 and rivers (Bunn and Arthington, 2002; Arthington et al., 2010; Olden and Naiman, 2010) are

63 known to cause variability in population biology parameters of aquatic species (e.g. Parker et al.,

64 1995). For example, warmer environments and/or systems with greater productivity may increase 3

Page 4 of 44 65 growth rates and lengths at first maturity of freshwater crayfishes (Parkyn et al., 2002; Beatty et al.,

66 2005b). Quantification of the biological plasticity of Southern Hemisphere parastacids is relatively

67 limited (Honan and Mitchell, 1995; Austin, 1998; Beatty et al., 2003a, 2005b; Jones et al. 2007),

68 which is somewhat surprising given that there are at least 140 species recognised in Australia alone

69 and a number of them support recreational and commercial fisheries and/or aquaculture industries

70 (Crandall et al., 1999).

71 Aside from two studies in the 1970s (see Morrissy, 1970, 1975), it is only recently that the

72 population and reproductive biology of C. cainii has been determined from wild aquatic systems

73 (Beatty et al., 2003a, 2005b). Those studies found that considerable differences existed between

74 populations for several biological parameters and suggested that this species may have a variable

75 life history (Morrissy, 1975; Beatty et al., 2003a, 2005b). Despite this apparent high variability,

76 comprehensive evaluation of key population biology parameters such as growth, mortality,

77 exploitation rates, abundance, density and productivity of populations have not yet been

78 determined; and are urgently required to ensure the ongoing sustainable management and

79 monitoring of the fishery (Beatty et al., 2005b).

80 The annual landings and catch per unit effort (CPUE, number of marron per fisher per day) of

81 the C. cainii fishery have gradually decreased since the 1970s and substantial restrictions on the

82 fishing effort have occurred. For example, a reduction of the fishing season occurred from 135 days

83 (~500,000 marron) in the 1970s and 1980s to just 24 days from 2007 to the present, which aimed to

84 significantly reduce the impact of fishing mortality (Molony et al., 2002; de Graaf and Barharthah,

85 2008). Although secondary salinisation is known to have greatly reduced the overall distribution 86 and catches of C. cainiiAccepted (Morrissy, 1978; Molony et Manuscriptal., 2002), the degree of winter rainfall and 87 river flow have recently been shown to influence catch rates by altering available aquatic habitat (de

88 Graaf et al., 2010). The Mediterranean climate of south-western Australia has already undergone a

89 10-15% annual rainfall reduction since 1975 and average forecasts estimate another 8% decline

90 resulting in a 25% reduction in surface flows by 2030 (CSIRO, 2009). Therefore, it is likely that

91 these considerable rainfall reductions will further reduce the range of C. cainii and the annual 4

Page 5 of 44 92 recruitment of juveniles to the fishery, which may have serious consequences for its long-term

93 viability.

94 This study aimed to determine the population biology of C. cainii from a lentic and lotic system.

95 Wellington Dam and the Warren River (Fig.1) were selected as they consistently receive a

96 considerable proportion of the annual recreational fishing effort (in terms of number of fishing

97 days); with the number of total fishing days in Wellington Dam representing ~25-60% of the annual

98 effort in all water supply dams (between 2000-2008) and the Warren River representing 5-15% of

99 the total annual effort in all rivers between that period (de Graaf and Barharthah, 2008). Therefore,

100 examining C. cainii in these systems will enable an assessment of the relative influence of fishing

101 exploitation on such heavily fished populations. By also comparing these data with the limited

102 number of other studies on this species, the study aimed to assess the degree of plasticity in key

103 biological parameters of this parastacid to enable a more effective, finer scale management

104 approach.

105

106 2. Materials and methods

107 2.1. Sampling protocols

108 Sampling of C. cainii in Wellington Dam and the Warren River occurred on eight occasions

109 between April 2005 and April 2006. An additional two opportunistic samples were carried out in

110 September and October 2006 in order to recapture previously marked individuals. On each of the

111 major sampling occasions, consistent effort of standardised trapping occurred in Wellington Dam

112 (fished area 250 x 150 m) and in the Warren River (fished area 800 x 20 m) using rectangular 113 crayfish box traps (60Accepted cm length x 45 cm width x 20 Manuscriptcm height, 10 mm square mesh) baited with 114 poultry pellets and fished up to a maximum depth of 10 and 35 m in the river and dam sites,

115 respectively. Traps were deployed between 16:00-18:00 hrs at 20 m intervals and were recovered

116 the following morning (8:00-12:00 hrs). As it was found that C. cainii only become fully trappable

117 at greater than 35mm Orbital Carapace Length (OCL) (see Results), manual scoop netting (mesh

118 width 2 mm) of juveniles also occurred during daylight hours from the banks of the study sites on 5

Page 6 of 44 119 each sampling occasion. Surface water temperatures were measured from each site at three

120 locations and a mean determined.

121 Each individual captured was sexed and measured to the nearest 1 mm OCL. Individual coded

122 marking using punches occurred for those animals captured during the sampling occasions in June,

123 August, October and December 2005. The marking occurred on the ventral side of the four uropods

124 (up to three punches per uropod, left to right) in the pattern of 1-2-4, 7-10-20, 40-70-100, 200-400-

125 700 (Abrahamsson, 1965). This marking technique was only carried out on larger (>35 mm OCL)

126 C. cainii (i.e. trappable with adequate uropod surface area) and allowed accurate individual

127 identification of individuals following multiple moults during the study period (Molony and Bird,

128 2005) (see Results). Subsequent recaptures were re-marked if they had moulted to ensure

129 continuation of identification.

130

131 2.2 Growth analysis

132 Modal progression in growth analysis of invertebrates and freshwater crayfishes in particular can

133 prove difficult due to the problem of the overlapping of older cohorts (e.g. Gutiérrez-Yurrita and

134 Latournerié-Cervera, 1999; Beatty et al., 2005c). However, this methodology has recently been

135 successfully used in fitting growth curves on younger age classes of freshwater crayfish in south-

136 western Australia, including C. cainii (see Beatty et al., 2005b, 2005c). The use of this technique

137 for the analysis of the populations in Wellington Dam and within the Warren River allowed data to

138 be gathered that are directly comparable with those of Beatty et al. (2005b) obtained for C. cainii in

139 the Hutt River. Conversely, individual mark-recapture of those younger (<35 mm OCL) cohorts via 140 tail punching was Accepted not possible due to the small size Manuscript of the telson and uropods. Therefore, both 141 modal progression analysis and mark-recapture methods of growth determination were undertaken

142 in order to fully describe and compare the growth of the populations in Wellington Dam and the

143 Warren River. The complementary use of both these growth estimate methods, with mark-

144 recapture also being used to estimate population sizes, enabled determination of key population

6

Page 7 of 44 145 biology parameters, including natural and fishing mortality, annual survivorship, abundance,

146 density and production.

147

148 2.3 Modal progression

149 The techniques for modal progression analysis followed those of de Lestang et al. (2003) and

150 Beatty et al. (2005a, 2005b, 2005c). September 1st was assigned as the birth (hatching) date as the

151 main spawning period of C. cainii in Wellington Dam and the Warren River occurred

152 approximately between August and October 2005 (de Graaf et al., 2010). Sexes were pooled for

153 analysis as initial examination of the sub-adult modes in the length-frequency distributions revealed

154 no discernable differences between sexes. Single, two and three normal distributions were fitted to

155 length-frequency distributions, in 2 mm OCL increments, on each sampling occasion in Wellington

156 Dam and the Warren River with the most appropriate normal distributions for each month

157 determined using the chi-square method (Schnute and Fournier, 1980) with the modification

158 described in de Lestang et al. (2003).

159 Once the modes were determined for each distribution in each month, the relative location of the

160 OCL distribution of a cohort within the total frequency distribution of each month and the

161 relationship of the OCL distribution in the adjacent months were used to assign the cohorts to 0+ or

162 1+ age classes (de Lestang et al., 2003). As the moult frequency of freshwater crayfish is greatest

163 in the first few months of life (Reynolds, 2002), a modified version of the von Bertalanffy growth

164 curve of Hanumara and Hoenig (1987) was fitted to the mean OCL distributions that assumed that

165 the maximum growth rate occurred in the previous 5 months (de Lestang et al., 2003): 166 Accepted Manuscript    Kt  t'0 CK 3  OCL1 exp   sin 2      12 2 12  if t < ts + 3 OCLt       Kt  t'0 CK t  ts if t > t + 3 OCL1 exp   sin 2  s     12 2 12  169

170 where OCLt is the estimate of orbital carapace length at age t months, OCL is the asymptotic

171 orbital carapace length, K is the curvature parameter, t0 is the theoretical age at which the estimated 7

Page 8 of 44 172 orbital carapace length is zero (t0 = t'0 - (6C/)sin(0.5)), C is the relative amplitude of the seasonal

173 oscillation (where 0 < C < 1) and ts is the phase of seasonal oscillation relative to t0. Growth curves

174 were fitted to the length-frequency data using Solver in Microsoft Excel™.

175

176 2.4 Mark-recapture

177 The mark-recapture study was undertaken between June and December 2005 (with subsequent

178 sampling for recaptures occurring in March, May and October 2006) and allowed both an analysis

179 of growth of these larger cohorts of trappable C. cainii (>~35 mm OCL) and also an estimation of

180 trappable population abundance and other demographic parameters (see next section).

181 Growth trajectories were determined for each individual male and female C. cainii recaptured in

-1 182 the Warren River and Wellington Dam. Mean absolute growth (Gabs mm OCLyr ) and relative

-1 183 growth (Grel % OCLyr ) rates were calculated for recaptured males and females (and pooled for

184 sexes) in both systems. The statistical significance of any differences in growth rates between sexes

185 and systems were tested first by employing Levene’s tests for equality of error variance on the un-

186 transformed data. Heteroscedastic data were ln-transformed prior to one-way ANOVAs being

187 performed. A probability level of ( = 0.05) was used to test all null hypotheses that the growth

188 rates were not different between sexes and sites.

189 To provide direct comparison between the growth as determined from modal progression

190 analysis of the sub-adult size classes (described above for each system) with those trappable

191 animals in the mark-recapture program, Gulland and Holt (1959) estimates of von Bertalanffy

192 parameters L and K were initially undertaken for each sex individually, and then for pooled sexes,

193 in each system of recaptured,Accepted trappable animals using theManuscript equation:

194 (ΔOCL/ Δt) = a + bOCLm

195 Where ΔOCL is the change in the OCL during the time at liberty Δt and OCLm is the mean OCL

196 during Δt. The parameter a is the y-axis intercept and b is the slope of the regression and –b is

197 equal to K and –a/b is equal to OCL. However, this analysis resulted in a positive slope of the

8

Page 9 of 44 198 regression line as no clear reduction in growth rate occurred with increasing size in the Warren

199 River. Therefore, forced Gulland-Holt plots were fitted whereby the length of infinity (OCLgh)

200 was implied by the size class of the largest C. cainii captured in both systems (i.e. 105 and 78 mm

201 OCL in Warren River and Wellington Dam, respectively) and the mean of the absolute growth rates

202 (Gabs) (i.e. y axis) and mean size (i.e. x axis) were used to determine Kgh using the equation:

203 Kgh = Y / (OCLgh - X )

204 The moult increments of C. cainii in the Warren River and in Wellington Dam were determined by

205 examination of the modal frequencies of absolute growth rates of individuals within 10 mm size

206 classes. The most appropriate normal distributions were fitted to the dominant mode using the same

207 technique as applied in the above modal progression analysis for growth rates of sub-adults. The

208 first (dominant) mode reflected a single growth increment with those subsequent (i.e. larger) modes

209 representing those individuals that had moulted more than once (Frisch, 2007). The average time at

210 liberty of those that moulted only once as determined from the modal analysis was determined for

211 each 10 mm category of C. cainii in both systems (Robertson and Butler, 2003).

212

213 2.5 Abundance and density estimates

214 The mark-recapture sampling that occurred in June, August, October and December 2005, and

215 March and May 2006 (as described above) was used in the estimation of the trappable population

216 abundance and density of C. cainii in the trapped section of the Warren River and the Wellington

217 Dam using the MARK software program (White and Burnham, 1999). Cherax cainii is known to

218 have a relatively high home range fidelity and both study areas were effectively totally blanket

219 trapped on each samplingAccepted occasion (traps set within Manuscript ~20 m from each other assuming a bait

220 attraction of ~300 m2 (Morrissy, 1975)). However, we could not reasonably assume closure of the

221 populations (i.e. no net changes due to births, deaths, immigration or emigration) due to the

222 relatively long length of the study period and the fact that animals could potentially migrate into and

223 out of the trapped area throughout the study period. Therefore, the open population POPAN

9

Page 10 of 44 224 function (Schwartz and Arnason, 1996) was employed in MARK in order to provide population

225 estimates of female and male C. cainii in the trapped areas of Warren River and Wellington Dam

226 during the study period. The MARK program allowed estimation of the trappable population on

227 each sampling occasion (Nj), the superpopulation size (N) (i.e. total number of trappable C. cainii

228 present in each population between the initial and final sampling occasion) (Schwartz and Arnason,

229 1996, 2010). Other parameters estimated are apparent survival rate j), probability of entry to the

230 population (j), and probability of capture (pj). Parameters in the models can be denoted as time

231 dependant (t) or constant (.) (Cooch and White, 2010). Following Schwartz and Arnason (1996),

232 for the fully time dependent model we imposed the constraints that the first two capture

233 probabilities were equal and the last two capture probabilities were equal. This ensured all

234 parameters could be estimated correctly. Model selection used the AIC criteria used in MARK

235 (White and Burnham, 1999). Population sizes for each sampling period were plotted for each sex in

236 each population. In order to provide an estimate of population density and productivity, the density

237 (D m-2) of the trappable population of male and female C. cainii in each sampling area in the

238 Warren River and in Wellington Dam was determined using the estimated population size in

239 December 2005; the approximate midpoint of the study and immediately following the major

240 breeding period and prior to the recreational fishing season. The approximate area of the trapped

241 section in each system is estimated from satellite imagery (Google EarthTM).

242

243 2.6 Productivity

244 The mean stock productivity (P g.m-2yr-1) of trappable (i.e. ~>35 mm OCL) C. cainii of both

245 sexes in the WarrenAccepted River and in Wellington Dam was Manuscriptestimated using the formula:

246 P = GwtxD

247 Where Gwt is the mean annual weight gain for each sex in each system determined by converting

248 Gabs to weight (using the previously determined length-weight relationship of C. cainii in Beatty et

249 al. (2003a)) and D is the mean density of trappable females and males in each system in December

250 2005. 10

Page 11 of 44 251

252 2.7 Mortality

253 To determine the instantaneous total mortality rate (Z, 1yr-1), a catch curve was employed that

254 plotted the natural logarithms of numbers of C. cainii surviving over age (Beverton and Holt, 1957;

255 Ricker, 1975) using the growth parameters estimated from the forced Gulland-Holt plots that better

256 represented the growth rates of the larger cohorts (approximating those that were trappable) in the

257 population. In order to provide an estimate of rates of fishing and natural mortality in the Warren

258 River and in Wellington Dam, estimates of Z were determined from the catch curves separately for

259 size classes that were unexploited (i.e., Zu, an estimate of natural mortality) by recreational fishing

260 (i.e., those <76 mm carapace length, or < 54.2 mm OCL (Beatty et al., 2003a)) and those open to

261 legal fishing (i.e. Ze, > 54.2 mm OCL, an estimate of total mortality). These estimates assumed

262 negligible illegal retention of undersize individuals. As length-frequency data precluded accurately

263 identifying frequency at age for older animals, the data were used to create an age-frequency

264 distribution via the generation of a length-converted catch curve (Pauly, 1983; King, 1995):

265  Ni  ln  α - Zti Δt  266

267 where Ni is the number of individuals in a size class, Δt the time taken to grow through the size

268 class i, ti is the relative age of the size class i (t0 = 0 as only relative ages are required),  is a

269 constant, and Z is either the total rate of instantaneous mortality of those that were not subject to

-1 -1 270 recreational fishing (Zu, 1yr ) or those that were exploited (Ze, 1yr ). The annual percentage

271 survivorship (Se) of fully exploited C. cainii was then determined using the equation:

272 Accepted Se = 100 x exp( -ZManuscripte)

273 Subsequently, the instantaneous rate of fishing mortality (F, 1yr-1) was determined using modified

274 versions (i.e., using the estimate of Ze and Zu) of the King (1995) equation:

275 F = Ze – Zu

276 An estimate of the exploitation rate E (the proportion of Ze attributed to fishing mortality) was then

277 determined using modified versions of the Quinn and Deriso (1999) equation: 11

Page 12 of 44 278 E = F/Ze

279 3. Results

280 3.1 Growth

281 A total of 5308 and 5117 C. cainii were captured in the Warren River and in Wellington Dam,

282 respectively over a period of 18 months. Of these, 3847 and 3229 were captured in the standardised

283 box-trapping programs that were used in the multiple mark-recapture analysis of growth, mortality

284 and production. The remainder were sub-adults captured during the trialling of methods to

285 determine juvenile abundance and were included in the modal progression analysis of growth. The

286 sex ratio (female:male) of those captured in the standardised trapping program was 1:1.52 and

287 1:1.95 in the Warren River and Wellington Dam, respectively.

288 Examination of the length-frequency distributions in the Warren River and Wellington Dam

289 populations and the fitting of normal distributions, revealed a clearly discernable single 0+ cohort

290 between April and December 2005; that were 1+ in February 2006 (Fig. 2). A new 0+ cohort, the

291 result of spawning in spring 2005 (as mentioned, September 1st was the estimated hatching date),

292 was first discernable in February 2006 and then again in April 2006. Considerable overlap in the

293 larger size cohorts prevented fitting normal distributions to those sizes >~40mm OCL. The

294 seasonal von Bertalanffy growth curves fitted to the monthly modes of the 0+ and 1+ size classes

295 (up to the age of 19 months) in the Warren River and in Wellington Dam clearly indicated a faster

296 growth rates of those sub-adult animals in Wellington Dam compared to the Warren River (Fig. 3,

-1 -1 297 Table 1). This was highlighted by a Kmp of 0.32.yr cf 0.23.yr in Wellington Dam and the Warren

298 River, respectively (Table 1). Based on extrapolation of the seasonal von Bertalanffy curve of sub- 299 adult growth, the timeAccepted taken to reach the current legal Manuscriptsize of 80 mm CL (~57 mm OCL) would be 300 ~75 and ~42 months in the Warren River and in Wellington Dam, respectively (Table 1). This

301 compares with only ~25 months for the Hutt River.

302 During the multiple mark-recapture sampling program of trappable (>~35 mm OCL) adult C.

303 cainii, a total of 1172 and 1039 were marked in the Warren River and Wellington Dam with

304 average size of 55.8 (±0.35) and 49.7 (±0.19) mm OCL in these systems, respectively (Table 2). 12

Page 13 of 44 305 Recapture rates were similar in both systems (39.8 and 36.1% for Warren River and Wellington

306 Dam, respectively) with a greater percentage of males being recaptured in both systems; probably

307 due to a degree of trap shyness of berried females coupled with higher aggression by males (Table

308 2). Mean time at liberty of recaptured C. cainii was 239 (±6) and 193 (±6) days in the Warren

309 River and Wellington Dam, respectively (Table 2). The multiple mark-recapture program in these

310 systems revealed an opposite trend in growth rates to those revealed in the modal progression

311 analysis of sub-adult C. cainii, with the Warren River C. cainii growing significantly faster than

312 those in Wellington Dam (Fig. 4). The average size of C. cainii re-captured in the trapping program

313 in Warren River (61.9 ±0.54 mm OCL) was significantly greater (ANOVA, p < 0.01) than those in

314 Wellington Dam (51.9 ± 0.29 mm OCL) (Table 2). The greater growth rate of the trappable Warren

315 River population was evidenced by the significantly greater overall absolute growth of females,

316 males and pooled sexes when compared to those in Wellington Dam (Fig. 4). Similarly, relative

317 growth rate of females, males and pooled sexes was also significantly greater in the Warren River

318 population (Fig. 4).

319 Figure 5 shows the mean absolute annual growth rate of 10 mm OCL size categories and reveals

320 that the growth rate slowed with increasing size in Wellington Dam, however, this trend was not

321 observed in the Warren River. A similar trend in declining relative growth rate with size occurred

322 in Wellington Dam (Fig. 6), and a sharp decline in the relative annual growth occurred in the

323 Warren River between 30-40 mm OCL and the larger size categories, with a slight negative trend

324 occurring in the subsequent larger sizes.

325 Using a specified OCLgh of 105 and 78 mm OCL in the Warren River and in Wellington Dam,

-1 -1 326 the forced GullandAccepted-Holt formula estimated the Kgh inManuscript these systems to be 0.14.yr and 0.11.yr ,

327 respectively (Table 1). There were very weak, non-significant negative relationships between the

328 relative growth rates and the mean OCL of C. cainii in the Warren River, i.e., Grel = 19.0-

2 2 329 0.102OCL (r = 0.008, p = 0.22) and Grel = 14.6 - 0.081OCL (r = 0.009, p = 0.12) for females and

330 males, respectively. There were very weak, significant negative relationships between the relative

2 331 growth rates and the mean OCL of C. cainii in Wellington Dam, i.e., Grel = 38.3-0.643OCL (r = 13

Page 14 of 44 2 332 0.07, p = 0.01) and Grel = 14.8 - 0.182OCL (r = 0.142, p = 0.02) for females and males,

333 respectively.

334 The frequency of absolute growth increments are shown in Fig. 7 and the subsequent modes of

335 these increments are plotted in Fig. 8. A greater moult increment occurred in the 40-50 mm and 50-

336 60 mm OCL size classes in the Warren River (3.34 and 4.02 mm OCL, respectively) compared to

337 Wellington Dam (3.18 and 3.50, respectively). A lack of C. cainii >60 mm OCL in Wellington

338 Dam prevented determining their moult increments, however, those categories in the Warren River

339 showed an increase in moult increment to 5.89 mm OCL in the 80-90 mm size category (Fig. 8).

340 There was no significant difference between the populations in the mean time at liberty of C. cainii

341 that moulted once within the 40-50 mm (ANOVA, p = 0.267) or the 50-60 mm (ANOVA, p =

342 0.944) OCL size categories with the mean time in the Warren River being 0.74 and 0.80 yrs cf 0.68

343 and 0.80 yrs in the Wellington Dam for those size categories, respectively (Fig. 8). Therefore, a

344 slight increase in the average time at liberty occurred in both systems between the 40-50 and 50-60

345 mm OCL size categories.

346

347 3.2. Abundance, density and productivity

348 The multiple mark-recapture program and analysis estimated the total number of trappable

349 female and males present in the study area in the Warren River during the entire study period (i.e.

350 N, the super population) were 1357.0 (±159.9) females and 1490.0 (±107.1) males, respectively. In

351 the Wellington Dam, the super population of females and males throughout the study period were

352 821.7 (±70.8) females and 1052.4 (±44.2) males, respectively. The abundances of trappable female 353 and males in the Accepted Warren River study area were highlyManuscript variable over the first four sampling 354 occasions between showing only a gradual decline between December 2005 and May 2006. The

355 abundances of trappable female and males in the Wellington Dam study were more stable and

356 showed similar trends in abundances (Fig. 9). The estimated trappable population sizes of female

357 and male C. cainii in the Warren River in December 2005 (following the major breeding period and

358 prior to the recreational fishing season) were 478.8 (±58.4) and 486.5 (±43.8), respectively (Table 14

Page 15 of 44 359 2, Fig. 9). This equated to an approximate density of 0.030 (±0.004) females.m-2 and 0.030

360 (±0.003) males.m-2 (pooled sexes = 0.060 (±0.005) C. cainii.m-2) (Table 2, Fig. 9). The estimated

361 number of trappable female and male C. cainii in the Wellington Dam study area in December 2005

362 were 397.1 (±84.3) and 498.7 (35.2), respectively (Table 2, Fig. 9). This equated to an approximate

363 density of 0.011 (±0.002) females.m-2 and 0.013 (±0.001) males.m-2 (pooled sexes = 0.024 (±0.002)

364 C. cainii.m-2) (Table 2, Fig. 9). The subsequent estimates of C. cainii productivity were much

365 greater for the Warren River (pooled sexes = 5.07 (±0.473) g.m-2.yr-1 or 3.03 (±0.370) and 2.23

366 (±0.201) g.m-2.yr-1 for females and males, respectively) than Wellington Dam (pooled sexes = 0.63

367 (±0.082) g.m-2.yr-1 or 0.28 (±0.059) and 0.36 (±0.025) g.m-2.yr-1 for females and males,

368 respectively) (Table 2).

369

370 3.3. Mortality

371 The instantaneous total rate of mortality of fully recruited C. cainii in the Warren River was

372 slightly higher than that recorded in Wellington Dam (0.59 and 0.47 yr-1 respectively, Fig. 10, Table

373 1). The unexploited mortality (Zu) in the Warren River was considerably higher than that recorded

374 in Wellington Dam (0.42 and 0.18 yr-1, respectively) and this resulted in a greater fishing mortality

375 (F) being calculated for Wellington Dam (0.29 yr-1) than in the Warren River (0.17 yr-1). Therefore,

376 the exploitation rate (E) was much greater in Wellington Dam (0.62) than in the Warren River

377 (0.29). However, these rates were much lower than those calculated for the Hutt River population

378 with F and E being 1.38 yr-1 and 0.77, respectively (Table 1). The annual survivorship of fully

379 recruited and exploited (legal size) C. cainii in the Wellington Dam was 62% compared with 55% 380 in Warren River. Accepted Manuscript 381

382 4. Discussion

383 This study is the first to comprehensively quantify and compare the population biology of a

384 parastacid in a lentic and lotic system. There were marked differences in the growth, mortality,

385 exploitation and survivorship rates between these two populations of C. cainii and also the more 15

Page 16 of 44 386 northern Hutt River population examined in Beatty et al. (2005b) (Table 1). The growth rate of

387 sub-adults in the Warren River was less than those in Wellington Dam; however, facilitated by

388 differences in moult increments, adults in the river population grew much faster than those in the

389 dam. Based on the mark-recapture analysis, the size of the trappable population and subsequent

390 density estimates of females and male C. cainii in the Warren River were more variable over the

391 duration of the study compared to the Wellington Dam population; although a decline in trappable

392 abundances occurred in both sexes in both systems between December 2005 and March 2006. A

393 major proportion of those declines can be attributed to fishing mortality as the recreational season

394 occurred from January 20th – February 5th 2006 (see below). Although trappable density of C.

395 cainii varied over time, a much higher density and mean annual weight gain were recorded in C.

396 cainii in the Warren River (i.e. 83.9 ±4.6 g.yr-1) cf Wellington Dam (i.e. 26.4 ±2.1 g.yr-1) which

397 resulted in a higher productivity estimate for the Warren River population (Table 2). The

398 underlying low productivity in the Wellington Dam population is also reflected in the low

399 proportion of legal sized animals that were recorded.

-1 400 A higher total mortality of exploited (Ze) C. cainii was recorded in the Hutt River (1.79 yr ) than

401 in both the Warren River (0.59 yr-1) and Wellington Dam (0.47 yr-1) and this resulted in far lower

402 annual survivorship in the Hutt River compared with the Warren River and Wellington Dam (Beatty

403 et al., 2005b). Furthermore, fishing mortality and exploitation rate was also far greater in the Hutt

404 River population than in the other systems; with Wellington Dam also being considerably greater in

405 these parameters than the Warren River. The exploitation rates in the Hutt River and Wellington

406 Dam suggest that these populations may be over-exploited. Fishing in water supply dams accounts 407 for ~30% of the overallAccepted annual C. cainii recreational fishingManuscript effort (in terms of fishing days) with 408 the remaining ~70% occurring in rivers (de Graaf and Baharthah, 2009). The relatively high fishing

409 mortality and exploitation rate in the Wellington Dam may reflect the fact that this system accounts

410 for ~35% of the total annual effort within all dams (de Graaf and Baharthah, 2009).

411 The relatively low level of fishing mortality in the Warren River compared with Wellington Dam

412 may have contributed to the presence of a considerably greater number of legal-sized C. cainii in 16

Page 17 of 44 413 the Warren River during this study, however, it should be noted that our trapping included areas

414 away from fisher access therefore other factors influence the observed size distributions. By

415 comparison, Beatty et al. (2005b) recorded very high fishing mortality and exploitation rates in the

416 Hutt River C. cainii population (Table 1). At the time of the latter study, breeding females were not

417 actually protected from fishing prior to reaching maturity (L50 = 95 mm CL cf minimum legal size

418 of 76 mm CL); however, despite this high fishing mortality and lack of brood-stock protection, a

419 considerable proportion of legal sized C. cainii were captured during that study and the Hutt River

420 continues to produce considerable recreational CPUE despite an increase in legal size to 90 mm CL

421 and reduced daily bag limit of five since 2007 (de Graaf and Barharthah, 2008).

422 The growth of C. cainii in Wellington Dam was previously described by Morrissy (1975) and the

423 size of individuals at approximately 14 months of age (November) was found to be 34 mm OCL,

424 which is considerably greater than that recorded in the current study (~21 mm OCL, Fig. 3) and

425 suggests that growth rate has declined over that period. That length at age was similar to that

426 recorded by Beatty et al. (2005b) in the fast growing Hutt River population of 33 mm OCL.

427 Furthermore, Morrissy (1970) estimated a growth coefficient K of 0.2 for the Warren River

428 population; similar to the 0.23 estimated by modal progression analysis in the current study. This

429 suggests that there are both spatial and temporal variations in growth rates of populations in the C.

430 cainii fishery (likely driven by interannual variability in environmental conditions; particularly

431 rainfall and discharge (de Graaf et al., 2010)). This has implications for monitoring programs that

432 need to account for this variability, i.e., by ongoing monitoring of indicator stocks from a number of

433 habitats, which has recently been implemented (de Graaf and Barharthah, 2009). The differences in 434 the population biologyAccepted parameters between these stocks Manuscript are likely to be driven by several factors 435 that are discussed below.

436 Growth rates of freshwater crayfish are influenced by temperature and are generally positively

437 related up to a thermal optima (e.g., Jones, 1981; Jussila and Evans, 1997; Whitmore and Huryn,

438 1999). For example, Paranephrops planifrons grew faster in warmer pasture streams compared

439 with cooler forested streams due to a greater moult increment in adults and a higher moult 17

Page 18 of 44 440 frequency of smaller individuals in the pasture streams (Parkyn et al., 2002). The Hutt River

441 population had a greater growth rate than the majority of the more southern river and dam

442 populations; however, it was less than the rate recorded in the Harvey Dam (Morrissy, 1975; Beatty

443 et al., 2005b). Water temperatures generally decline with increases in latitude, for example, mean

444 surface water temperatures in Wellington Dam and Warren River were 15.35°C (±0.89) and 13.8°C

445 (±0) in October 2005 and 17.7°C (±0.4) and 15.2°C (±0.11) in May 2006, respectively. Based

446 solely on temperature regime, the higher growth rate in the Hutt River compared to those examined

447 in higher latitudes in this study and by Morrissy (1975) could be attributed to higher water

448 temperatures in the former system that often approximate the known optimum temperature for

449 growth of C. cainii (i.e., 24°C) (Morrissy, 1990; Beatty et al., 2005b). Furthermore, a greater

450 growth rate could be expected in the lower latitude Wellington Dam population cf the more

451 southern Warren River population. However, although younger cohorts in the Wellington Dam

452 population grew faster than in the Warren River, this was reversed in adults. Therefore, growth

453 rates of these populations are not linearly related to temperature regimes, and more complex factors

454 are involved.

455 The importance of intact riparian vegetation and adequate water quality in sustaining C. cainii

456 populations results in the species being known as a keystone species and an ideal indicator of river

457 health in this region (Nickoll and Horwitz, 2000). One of the most obvious ecological differences

458 between reservoirs and rivers of this region is the influence of the riparian vegetation zone; such as

459 the degree to which it contributes to both productivity and benthic habitat (Bunn and Davies, 1990;

460 Beatty et al., 2003b). Food availability is a key factor influencing growth rates of freshwater 461 crayfish (e.g. Sokal,Accepted 1988; Whitmore and Huryn, 1999). Manuscript Regardless of temperature, rapid growth 462 rates are only possible if adequate energy sources are available to support them (Reynolds, 2002).

463 The high degree of allochthonous inputs into rivers with intact riparian zones provides an important

464 source of organic matter entering the detrital ecosystem that would subsequently become available

465 to C. cainii either by direct consumption of detritus and biofilm, or consumption of higher taxa such

466 as invertebrates and even fish (Parkyn et al., 2001; Hollows et al., 2002; Beatty, 2006). Although 18

Page 19 of 44 467 the majority of recreationally fished dams such as Wellington and Waroona may have large areas of

468 shallow, warmer habitat; they have low habitat diversity (dominated by mud or sand substrata) with

469 very limited riparian vegetation inputs (Beatty, 2003b) and food webs that are reliant on potentially

470 limited autochthonous production. However, one anomaly to this is Harvey Dam as C. cainii are

471 believed to have an exceptionally high growth rate in this system; reflected by a greater minimum

472 legal size of 90 mm OCL and its ‘trophy fishery’ status (de Graaf and Barharthah, 2008). This may

473 be due to unusually high productivity in this lentic system due to nutrient inputs from adjacent

474 farmland. Despite this exception, the differences in growth rates and productivities between the

475 river and reservoir populations revealed in the current study are likely related to the degree of

476 allochthonous input.

477 Benthic habitat complexity also structures freshwater crayfish populations and influences

478 density, growth, and mortality from predation (France et al., 1991). Complex habitat facilitates

479 increased survivorship of juvenile freshwater crayfishes by reducing the degree of inter- and

480 intraspecific predation (France et al., 1991; Lodge and Hill, 1994; de Graaf et al., 2010). Both the

481 Warren River and Wellington Dam house the same species of introduced fishes that are known to

482 predate on C. cainii, i.e., rainbow trout Oncorhynchus mykiss, brown trout Salmo trutta and redfin

483 perch Perca fluviatilis (Morgan et al., 2004; Tay et al., 2007; Beatty and Morgan, 2010). The

484 relatively intact riparian vegetation and high bank to surface area ratio contributes to the complex

485 instream habitat in the Warren River whereas a completely opposite scenario exists in the dams of

486 the region (Beatty et al., 2003b). Such habitat complexity may reduce the predation rates by

487 predatory fishes and increase the survivorship of juvenile C. cainii thereby contributing to the 488 consistently higher Accepteddensity of adults recorded in the WarrManuscripten River cf Wellington Dam during the 489 mark-recapture program. This may also be the case in other rivers of the region that house those

490 introduced predators but continue to have largely intact riparian zones and a high degree of instream

491 habitat complexity. Protection and rehabilitation of riparian zones should be an important part of

492 helping to ensure the long-term sustainability of the C. cainii fishery.

19

Page 20 of 44 493 The influence of complex habitat availability on the productivity of C. cainii was highlighted in

494 the artificial habitat study by Molony and Bird (2005). That study showed that C. cainii (including

495 newly recruited individuals) rapidly populated artificial habitats in Big Brook Dam (in the Warren

496 River catchment). This colonisation effectively increased productivity probably via reducing teleost

497 predation (Molony and Bird, 2005). Furthermore, increasing the complexity of habitat in that

498 system was also thought to enhance the biofilm community and thus total species richness of the

499 system (Molony and Bird, 2005). Those results, coupled with the findings of the current study,

500 have direct management implications for enhancing the productivity of freshwater crayfish fisheries

501 in habitat limited reservoir environments (see below).

502 Predation rates influence activity, growth and abundance of freshwater crayfishes (e.g. Foster

503 and Slater, 1995; Reynolds, 2002). The presence of the above mentioned large introduced teleost

504 predators in other habitat limited reservoir environments of south-western Australia (Beatty et al.,

505 2003b, Morgan et al., 2004) may result in both the selection of faster growing juveniles that are able

506 to attain a size that precludes inter and intra-specific predation, and also a reduction in the density of

507 the adult population by reducing juvenile recruitment. Moreover, harvest and natural selection can

508 work in opposite directions in terms of influencing somatic growth and reproductive investment

509 (Edeline et al., 2007). The high exploitation rates recorded in Wellington Dam may be selectively

510 removing faster growing individuals thus selecting for slower growing adults; contributing to the

511 stunting of adult growth. Therefore, along with competition for limited food resources, predation

512 and fishing exploitation may be influencing the growth, abundance, productivity and annual rate of

513 recruitment to the fishery in Wellington Dam. 514 Although the HuttAccepted River and the Harvey Dam populationsManuscript of C. cainii have unusually high 515 growth rates and lengths at first maturity, the population biology parameters determined for the

516 river and reservoir populations in the current study may be broadly similar to these habitat types

517 throughout the majority of the fishery and these aquatic system types should therefore be managed

518 as separate units. However, along with conducting a broader study examining additional

519 populations in order to quantify the biotic and a-biotic drivers of the biological variability, it would 20

Page 21 of 44 520 also be useful for the ongoing monitoring of the fishery to also include sentinel stocks (i.e., closed

521 to recreational fishing such as all potable water supply dams in south-western Australia) to help

522 quantify the effects of (legal) fishing mortality on the productivity of these systems.

523 The findings of the current study suggest that a sustainable increase in the annual catch may be

524 possible in more pristine river habitats. The relatively low fishing mortality and exploitation in the

525 Warren River suggests that recreational fishing has a relatively minor contribution to total mortality

526 of the fully recruited stock due to high population abundances and a lack of fisher access to the

527 majority of the population stemming from stringent fishery regulations that restrict fishing effort

528 (see de Graaf and Barhartha, 2008). Morrissy (1975) concluded that there is an upper limit to

529 fishing mortality for the species where increased effort is independent of legal sized catch. Other

530 freshwater crayfish studies have also highlighted the resilience of populations to overexploitation

531 (e.g. Orconectes virilis in Momot, 1991). Morrissy (1975) also postulated that the majority of the

532 recreational catch consisted of the faster growing 2+ and 3+ individuals as a correlation between

533 growth rate and catchability was found. Following the former hypothesis, Morrissy (1975)

534 concluded that increasing recruitment, not limiting fishing effort, was the most practical method of

535 increasing legal-sized stocks. However, caution must be exercised in allowing increased fishery

536 effort as yield reductions due to increased exploitation and removal of undersized crayfishes has

537 previously been demonstrated for other stocks of freshwater crayfish (e.g., Skurdal et al., 1993).

538 A large proportion of freshwater crayfishes (up to one half of known species) are under threat

539 globally (Taylor, 2002) and habitat alteration has previously been identified as a key threat to the

540 group (Jones et al., 2007). The reduction in geographical range of C. cainii has undoubtedly 541 occurred due to declineAccepted in habitat quality; in particular Manuscript, the salinisation of waterways that has 542 severely impacted the endemic freshwater fishes of the region (Morgan et al., 2003; Beatty et al.,

543 2011). This would have resulted in concentrating the current (reduced) fishing effort to areas that

544 continue to be of adequate water quality. Overfishing may still occur on a localised scale as C.

545 cainii has a relatively high home range fidelity and low dispersal ability (Morrissy, 1975).

546 Therefore spreading additional fishing effort over a greater area rather than merely increasing daily 21

Page 22 of 44 547 bag limits may be a more appropriate way of sustainably increasing the annual recreational catch

548 from rivers.

549 In terms of reservoir stocks, increasing the habitat complexity (e.g. addition of rock reefs) has

550 the potential to augment the productivity via increasing juvenile survivorship (reducing inter- and

551 intraspecific competition thereby increasing densities), and providing increased substrate for food

552 resources (e.g. biofilm accumulation) (Beatty et al., 2003b; Molony and Bird, 2005). This increase

553 in productivity may translate into a measurable increase in the recruitment to the fishery in dams

554 due to the increased densities and survivorship of the juveniles (see Molony and Bird, 2005).

555 The long-term sustainable management of the C. cainii fishery will undoubtedly be impacted by

556 climatic change. South-western Australia has undergone a 10-15% reduction in rainfall since 1970

557 that has resulted in a 50% reduction in inflows into water supply dams (IOCI, 2002; CSIRO, 2009).

558 The region is forecasted to continue to undergo a large reduction in annual rainfall due to climate

559 change; with 2030 rainfall levels predicted to decline by 8% and mean annual runoff by 25 % from

560 post 1975 levels (CSIRO, 2009). As river flow is positively related to C. cainii catches (de Graff et

561 al., 2010), the forecasted decline in rainfall in this region should be incorporated in the long term

562 management of the C. cainii fishery that may need to adapt to reduced recruitment.

563 The current study has demonstrated considerable inter-specific variability in key population

564 biology parameters of C. cainii between lentic and lotic systems and highlights the likely

565 mechanisms driving the variability. Although a limited number of exceptions may exist, these

566 findings support the separate management of lentic and lotic stocks. However, a much broader

567 study of additional populations is required to fully quantify the variability across its range. This 568 research should alsoAccepted include examination of those Manuscript biotic (such as levels of predation and 569 productivity) and a-biotic (such as temperature regime) parameters discussed above in order to

570 specifically identify and quantify those mechanisms responsible for the variability.

571

572

22

Page 23 of 44 573 Acknowledgements

574 Funding for the study was provided by the Australian Government through the Fisheries Research

575 and Development Commission and the Department of Fisheries, Government of Western Australia.

576 The field assistance of Cameron Hugh was greatly appreciated as was the revision of earlier drafts

577 of the manuscript by David Morgan, Howard Gill and Rick Fletcher.

578

Accepted Manuscript

23

Page 24 of 44 579 References

580 Abrahamsson, S.A.A., 1965. A method of marking crayfish Astacus astacus Linné in population

581 studies. Oikos 16, 228-231.

582 Allen, G.R., Midgley, S.H., Allen, M., 2002. Field guide to the freshwater fishes of Australia.

583 Western Australian Museum, Perth, Western Australia.

584 Arthington, A.H., Naiman, R.J., McClain, M.E, Nilsson, C., 2010. Preserving the biodiversity and

585 ecological services of rivers, new challenges and research opportunities. Freshwater Biol. 55,

586 1–16.

587 Austin, C.M., 1998. Intra-specific variation in clutch and brood size and rate of development in the 588 yabby, Cherax destructor (Decapoda: Parastacidae). Aquaculture 167, 147–159. 589 Austin, C.M., Knott, B., 1996. Systematics of the freshwater crayfish genus Cherax Erichson

590 (Decapoda: Parastacidae) in south-western Australia: electrophoretic, morphological and

591 habitat variation. Aust. J. Zool. 44, 223–258.

592 Austin, C.M., Ryan, S.G., 2002. Allozyme evidence for a new species of freshwater crayfish of the

593 genus Cherax Erichson (Decapoda: Parastacidae) from the south-west of Western Australia.

594 Invertebr. Syst. 16, 357–367.

595 Beatty, S.J., 2006. The diet and trophic positions of translocated, sympatric populations of Cherax

596 destructor and Cherax cainii in the Hutt River, Western Australia, evidence of resource

597 overlap. Mar. Freshwater Res. 57, 825-835.

598 Beatty, S., Molony, B., Rhodes, M., Morgan, D., 2003b. A methodology to mitigate the negative

599 impacts of dam refurbishment on fish and crayfish values in a south-western Australian 600 reservoir. EcologicalAccepted Management and Restoration Manuscript 4, 147-149. 601 Beatty, S.J., Morgan, D.L., 2010. Teleosts, agnathans and macroinvertebrates as bioindicators of

602 ecosystem health in a south-western Australian river. Journal of the Royal Society of Western

603 Australia 93, 65-79.

604 Beatty, S.J., Morgan, D.L., Gill, H.S., 2003a. Reproductive biology of the large freshwater crayfish

605 Cherax cainii in south-western Australia. Mar. Freshwater Res. 54, 597-608. 24

Page 25 of 44 606 Beatty, S.J., Morgan, D.L., Gill, H.S., 2005b. Biology of a translocated population of the large

607 freshwater crayfish, Cherax cainii Austin & Ryan, 2002 in a Western Australian river.

608 Crustaceana 77, 1329-1351.

609 Beatty, S.J., Morgan, D.L., Gill, H.S., 2005a. Role of life history strategy in the colonisation of

610 Western Australian aquatic systems by the introduced crayfish Cherax destructor Clark, 1936.

611 Hydrobiologia 549, 219-237.

612 Beatty, S.J., Morgan, D.L., Gill, H.S., 2005c. Life history and reproductive biology of the gilgie

613 Cherax quinquecarinatus, a freshwater crayfish endemic to south-western Australia. J.

614 Crustacean Biol. 252, 251-262.

615 Beatty, S.J., Morgan, D.L., McAleer, F.J., Ramsay, A., 2010. Groundwater contribution to baseflow

616 maintains habitat connectivity for Tandanus bostocki (Teleosteii: Plotosidae) in a south-western

617 Australian river. Ecol. Freshw. Fish 19, 595-608.

618 Beatty, S.J., Morgan, D.L., Rashnavidi, M., Lymbery, A.J., 2011. Salinity tolerances of endemic

619 freshwater fishes of south-western Australia: implications for conservation in a biodiversity

620 hotspot. Mar. Freshwater Res. 62, 91-100.

621 Beverton, R.J.H., Holt, S.J., 1957. On the dynamics of exploited fish populations. Ministry of

622 Agriculture, Fisheries and Food, Fisheries Investigations London, United Kingdom, 219 pp.

623 Bunn S.E., Arthington A.H., 2002. Basic principles and ecological consequences of altered flow

624 regimes for aquatic biodiversity. Environ. Manage. 30, 492–507.

625 Bunn, S.E., Davies, P.M., 1990. Why is the stream fauna of south-western Australia so

626 impoverished? Hydrobiologia 194, 169–176. 627 Cooch, E., White, G.Accepted 2010. Chapter 3: First steps in usingManuscript Program MARK. In: Cooch, E., White, 628 G. (Eds.), Program MARK: A gentle introduction, ninth edition, pp. 3.1-3.18. Available at:

629 http://www.phidot.org/software/mark/docs/book/

630 Crandall, K.A., Fetzner, J.W., Lawler, S.H., Kinnersley, M., Austin, C.M., 1999. Phylogenetic

631 relationships among the Australian and New Zealand genera of freshwater crayfishes

632 (Decapoda : Parastacidae). Aust. J. Zool. 47, 199–214. 25

Page 26 of 44 633 CSIRO, 2009. Surface water yields in south-west Western Australia. A report to the Australian

634 Government from the CSIRO. South-West Western Australia Sustainable Yields Project.

635 CSIRO Water for a Healthy Country Flagship, Australia, 171 pp.

636 de Graaf, M., Baharthah, T., 2008. Southern inland bioregion. In: Fletcher W.J, Santoro K. (Eds.),

637 State of the fisheries report 2007/08. Department of Fisheries, Government of Western

638 Australia, Perth, Western Australia.

639 de Graaf, M., Baharthah, T., 2009. Southern inland bioregion. In: Fletcher W.J, Santoro K. (Eds.),

640 State of the fisheries report 2008/09. Department of Fisheries, Government of Western

641 Australia, Perth, Western Australia.

642 de Graaf, M., Beatty, S.J., Molony, B.M., 2010. Evaluation of the recreational marron fishery

643 against environmental change and human interaction. Final report to Fisheries Research and

644 Development Corporation on Project No. 2003/027. Baxter, D., Larsen, R. (Eds.), Fisheries

645 Research Report. No. 211. Department of Fisheries, Western Australia. 188 pp.

646 de Lestang, S., Hall, N., Potter, I.C., 2003. Changes in density, age composition, and growth rate of

647 Portunus pelagicus in a large embayment in which fishing pressures and environmental

648 conditions have been altered. J. Crustacean Biol. 23, 908–19.

649 Edeline E., Carlson, S.M., Stige, L.C., Winfield, I.J., Fletcher, J.M., James, B., Haugen, T.O.,

650 Vollstad, L.A., Stenseth, N.C., 2007. Trait changes in a harvested population are driven by a

651 dynamic tug-of-war between natural and harvest selection. P. Natl. Acad. Sci. USA 104,

652 15799-15804.

653 Foster, J., Slater, F.M., 1995. A global review of crayfish predation with observations on the 654 possible loss Accepted of Austropotamobius pallipes in Manuscript the Welsh Wye due to crayfish plague. 655 Freshwater Crayfish 8, 589-613.

656 France, R.L., Holmes, J., Lynch, A., 1991. Use of size-frequency data to estimate the age

657 compositioin of crayfish populations. Can. J. Fish. Aquat. Sci. 48, 2324-2332.

658 Frisch, A.J., 2007. Growth & reproduction of the painted spiny lobster Panulirus versicolor on the

659 Great Barrier Reef Australia. Fish. Res. 85, 61-67. 26

Page 27 of 44 660 Gulland, J.A., Holt, S.J., 1959. Estimation of growth parameters for data at unequal time intervals.

661 J. Conseil 25, 47-49.

662 Gutiérrez-Yurrita, P.J., Latournerié-Cervera, J.R., 1999. Ecological features of Procambarus digueti

663 and Procambarus bouvieri Cambaridae, two endemic crayfish species of Mexico. Freshwater

664 Crayfish 12, 605–619.

665 Halse, S.A., Ruprecht, J.K., Pinder, A.M., 2003. Salinisation and prospects for biodiversity in rivers

666 and wetlands of south-west Western Australia. Aust. J. Bot. 51, 673–688.

667 Hanumara, R.C., Hoenig, N.A., 1987. An empirical comparison of a fit of linear and non-linear

668 models for seasonal growth in fish. Fish. Res. 5, 351–381.

669 Hayes, N.M., Butkas, K.J., Olden, J.D., Vander Zanden, M.J., 2009. Behavioural and growth

670 differences between experienced and naive populations of a native crayfish in the presence of

671 invasive rusty crayfish. Freshwater Biol. 54, 1876-1887.

672 Hollows, J.W., Townsend, C.R., Collier, K.J., 2002. Diet of the crayfish Paranephrops zealandicus

673 in bush and pasture streams, insights from stable isotopes and stomach analysis. New. Zeal. J.

674 Mar. Fresh. 36, 129–142.

675 Honan, J.A., Mitchell, B.D., 1995. Reproduction of Euastacus bispinosus Clark (Decapoda:

676 Parastacidae), and trends in the reproductive characteristics of freshwater crayfish. Mar.

677 Freshwater Res. 46, 485–499.

678 IOCI, 2002. Climate variability and change in south Western Australia. Indian Ocean climate

679 initiative, Perth, Australia, 34 pp.

680 Jones, J.B., 1981. Growth of two species of freshwater crayfish Paranephrops spp. in New Zealand. 681 New. Zeal. J. Mar.Accepted Fresh. 15, 15–20. Manuscript 682 Jones, J.G., Andriahajaina, F.B., Hockley, N.J., Crandall, K.A., Ravoahangimalala, O.R., 2007. The

683 ecology and conservation status of Madagascar’s freshwater crayfish (Parastacidae;

684 Astacoides). Freshwater Biol. 52, 1820-1833.

685 Jussila, J., Evans, L.H., 1997. On the factors affecting marron, Cherax tenuimanus, growth in

686 intensive culture. Freshwater Crayfish 11, 428–440. 27

Page 28 of 44 687 King, M., 1995. Fisheries biology, assessment and management, Fishing News Books; Blackwell

688 Science, Oxford, 341 pp.

689 Larson, E.R., Magoulick, D.D., 2008. Comparative life history of native Orconectes eupunctus and

690 introduced Orconectes neglectus crayfishes in the Spring River drainage of Arkansas and

691 Missouri. Am. Midl. Nat. 160, 323-341.

692 Lodge, D.M., Hill, A.M., 1994. Factors governing species composition, population size, and

693 productivity of cool-water crayfishes. Nordic Journal of Freshwater Research 69, 111-136.

694 Molony, B.W., Bird, C., 2005. Are marron, Cherax tenuimanus (Crustacea: Decapoda), populations

695 in irrigation reservoirs habitat limited? A trial using artificial habitats. Lakes & Reservoirs:

696 Research & Management 10, 39–50.

697 Molony, B.W., Bird, C., Nguyen, V., Baharthah, T., 2002. Assessment of the recreational marron

698 fishery for the 2002 season. Western Australian fisheries assessment document. Department of

699 Fisheries, Western Australia.

700 Momot, W.T., 1991. Potential for exploitation of freshwater crayfish in coolwater systems:

701 management guidelines and issues. Fisheries 16, 14-21.

702 Morgan, D.L., Gill, H.S., Maddern, M.G., Beatty, S.J., 2004. Distribution and impacts of introduced

703 freshwater fishes in Western Australia. New Zeal. J. Mar. Fresh. 38, 511-523.

704 Morgan, D.L., Gill, H.S., Potter, I.C., 1998. Distribution, identification and biology of freshwater

705 fishes in south-western Australia. Records of the Western Australian Museum Supplement No.

706 56, 97 pp. 707 Morgan, D.L., Hambleton,Accepted S.J., Gill, H.S., Beatty, S.J.,Manuscript 2002. Distribution, biology and likely 708 impacts of the introduced redfin perch (Perca fluviatilis) (Percidae) in Western Australia. Mar.

709 Freshwater Res. 53, 1211-1221.

710 Morgan, D.L., Thorburn, D.C., Gill, H.S., 2003. Salinization of south-western Western Australian

711 rivers and the implications for the inland fish fauna – the Blackwood River, a case study. Pac.

712 Conserv. Biol. 9, 161-171.

28

Page 29 of 44 713 Morrissy, N.M., 1970. Spawning of marron, Cherax tenuimanus Smith in Western Australia.

714 Fisheries Research Bulletin of Western Australia 112, 1–55.

715 Morrissy, N.M., 1975. Spawning variation and its relationship to growth rate and density in the

716 marron, Cherax tenuimanus Smith. Fisheries Research Bulletin of Western Australia 16, 1–32.

717 Morrissy, N.M., 1978. The past and present distribution of marron, Cherax tenuimanus, in Western

718 Australia. Fisheries Research Bulletin of Western Australia 22, 1–38.

719 Morrissy, N.M., 1990. Optimum and favourable temperatures for growth of Cherax tenuimanus

720 Smith 1912 Decapoda, Parastacidae. Aust. J. Mar. Fresh. Res. 41, 735–746.

721 Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Biodiversity

722 hotspots for conservation priorities. Nature 403, 853–858.

723 Nickoll, R., Horwitz, P., 2000. Evaluating flagship species for ecosystem restoration, a case study

724 involving freshwater crayfish and a river in southwestern Australia. In: Craig, J.I., Mitchell, N.,

725 Saunders, D.A. (Eds.), Nature conservation 5, nature conservation in production environments,

726 managing the matrix. Surrey Beatty and Sons, pp. 557–564.

727 Olden, J.D., Naiman, R.J., 2010. Incorporating thermal regimes into environmental flows

728 assessments, modifying dam operations to restore freshwater ecosystem integrity. Freshwater

729 Biol. 55, 86–107.

730 Parker, R.M., Zimmerman, M.P., Ward, D.L., 1995. Variability in biological characteristics of

731 northern squawfish in the lower Columbia and Snake Rivers. T. Am. Fish. Soc. 124, 335-346.

732 Parkyn, S.M., Collier, K.J., Hicks, B.J., 2001. New Zealand crayfish, functional omnivores but

733 trophic predators? Freshwater Biol. 46, 641–652. 734 Parkyn, S.M., Collier,Accepted K.J., Hicks, B.J., 2002. Growth Manuscript and population dynamics of crayfish 735 Paranephrops planifrons in streams within native forest and pastoral leases. New Zeal. J. Mar.

736 Fresh. 36, 847–861.

737 Pauly, D., 1983. Length-converted catch curves, a powerful tool for fisheries research in the tropics

738 part 1. Fishbyte 1, 9–13.

739 Payne, J.F., 1996. Adaptive success with the cambarid life cycle. Freshwater Crayfish 11, 1–12. 29

Page 30 of 44 740 Quinn, T.J.II., Deriso, R.B., 1999. Quantitative fish dynamics. Oxford University Press, New York.

741 Reynolds, J.D., 2002. Growth and reproduction. In: Holdich, D.M. (Ed.), The biology of freshwater

742 crayfish, Blackwell Science, Oxford, U.K., pp. 152–191.

743 Ricker, W.E., 1975. Computation and interpretation of biological statistics of fish populations.

744 Fisheries Research Board of Canada Bulletin 191.

745 Robertson, D.N., Butler, M.J., 2003. Growth and size at maturity in the spotted spiny lobster,

746 Panulirus guttatus. J. Crust Biol 23, 265-272.

747 Schnute, J., Fournier, D.A., 1980. A new approach to length-frequency analysis, growth structure.

748 Can. J. Fish. Aquat. Sci. 37, 1337–1351.

749 Schwarz, C.J., Arnason, A.N., 1996. A general methodology for the analysis of open-model capture

750 recapture experiments. Biometrics 52, 860-873.

751 Schwarz, C.J., Arnason, A.N., 2010. Chapter 13: Jolly-Seber models in MARK. In: Cooch, E.,

752 White, G. (Eds.), Program MARK: A gentle introduction, ninth edition, pp. 13.1-13.51.

753 Available at: http://www.phidot.org/software/mark/docs/book/

754 Skurdal, J., Qvenild, T., Taugbøl, T., Garnås, E., 1993. Long term study of exploitation, yield and

755 stock structure of noble crayfish Astacus astacus in Lake Steinsfjorden. Freshwater Crayfish 9,

756 118-133.

757 Sokal, A., 1988. The Australian yabby. In: Holdich, D.M. (Ed.), Biology of freshwater crayfish.

758 Blackwell Science, Oxford, U.K., pp. 401–426.

759 Tay, M.Y., Lymbery, A.J., Beatty, S.J., Morgan, D.L., 2007. Predation by Rainbow Trout

760 (Oncorhynchus mykiss) on a Western Australian icon: Marron (Cherax cainii). New Zeal. J. 761 Mar. Fresh. 41:Accepted 197–204. Manuscript 762 Taylor, C.A., 2002. Taxonomy and Conservation of Native Crayfish Stocks. In: The biology of

763 freshwater crayfish. Holdich D.M. (Ed.), Blackwell Science, Oxford, U.K., pp. 236-257.

764 Vogt, G., Huber, M., Thiemann, M., van den Boogarart, G., Schultz, O.J., Schubart, C.D., 2008.

765 Production of different phenotypes from the same genotype in the same environment by

766 developmental variation. J. Exp. Biol. 211, 510-523. 30

Page 31 of 44 767 White, G.C., Burnham, K.P., 1999. Program MARK: survival rate estimation from both live and

768 dead encounters. Bird Study 46(Supplement), 120-139.

769 Whitmore, N., Huryn, A.D., 1999. Life-history and production of Paranephrops zealandicus in a

770 forest stream, with comments about the sustainable harvest of a freshwater crayfish.

771 Freshwater Biol. 42, 467–478.

772

773 FIGURE CAPTIONS

774 Fig. 1. Location of the C. cainii populations examined in the current study in south-western Western 775 Australia. 776 Fig. 2 Orbital carapace length-frequency histograms for C. cainii between April 2005 and April 777 2006 in the Wellington Dam and Warren River. Normal distributions have been fitted to the 778 one or two size cohorts present in each month that were subsequently used in creating the 779 seasonal von Bertalanffy growth curve, n = sample size. 780 Fig. 3. Modified seasonal von Bertalanffy growth curves of C. cainii in the Warren River and in 781 Wellington Dam. Curves were fitted to the monthly mean orbital carapace lengths at age of 782 the 0+ or 1+ cohorts. 783 Fig. 4. Mean (±1 S.E.) sizes of female and male C. cainii from the Warren River and in Wellington 784 Dam that were tagged and subsequently recaptured, mean annual absolute growth rate, and 785 mean annual percentage growth (relative to overall OCL). Those categories with different 786 superscripts were significantly different (p < 0.05) and * denotes overall differences between 787 rivers (p < 0.05). 788 Fig. 5. Mean (±1 S.E.) absolute annual growth rates of 5 mm OCL size classes of female and male 789 C. cainii in the Warren River and Wellington Dam. The bottom graph shows absolute growth 790 rates of pooled sexes at both sites. 791 Fig. 6. Mean (±1 S.E.) relative annual growth rates (percentage increase of total OCL) of 5 mm 792 OCL size classes of female and male C. cainii in the Warren River and Wellington Dam. The 793 bottom graph shows relative annual growth rates of pooled sexes at both sites. 794 Fig. 7. Frequency of the absolute growth increments of various 10 mm OCL size categories of C. 795 cainii that were marked and recaptured in the Warren River and Wellington Dam between 796 June 2005 and November 2006. Moult increments implied by the sequential modes clearly 797 indicated an increase in the moult increment with size class in the Warren River; and those 798 individuals that did not grow were excluded. 799 Fig. 8. Moult increments for size categories (10 mm OCL) of C. cainii in the Warren River and 800 Wellington Dam, as implied by sequential modes of the frequency of absolute growth 801 increments, and the mean time at liberty of C. cainii that moulted once in each system. 802 Fig. 9. Mean (±1S.E.)Accepted trappable population numbers Manuscript(top graph) and densities (bottom graph) of 803 female, male and total C. cainii in the Warren River and Wellington Dam as estimated from 804 the POPAN formulation in the MARK software program. 805 Fig. 10. Length-converted catch curve of C. cainii in the Warren River and Wellington Dam. N.B.: 806 Slope of the regression line represents the instantaneous mortality rate of the exploited 807 population (Ze). Data points represented by triangles were excluded from regression as they 808 were size classes not subjected to legal fishing and those with open circles were excluded as 809 they represent mean ages that were not fully recruited (ascending data points) or those with 810 small sample sizes (<10 individuals).

31

Page 32 of 44 Table

Table 1 Parameters for the seasonal von Bertalanffy growth curves and estimates of mortality of C. cainii in the Hutt River (Beatty et al., 2005a), Wellington Dam and Warren River where: OCLmp is the

asymptotic orbital carapace length from modal progression, Kmp is the curvature parameter from modal progression, t0 is the theoretical age at which the estimated orbital carapace length is zero, C determines the relative amplitude of the seasonal oscillation (where 0 < C < 1), ts determines the 2 phase of seasonal oscillation relative to t0, R is the coefficient of determination, Ze is the instantaneous rate of mortality of C. cainii >legal size, OCLgh is the implied length of infinity for the forced Gulland-Holt plots, Kgh is the growth coefficient estimated from the Gulland-Holt plots, Zu is the instantaneous rate of mortality of unexploited C. cainii (legal size) C. cainii. N.B. All mortality estimates determined by the length converted catch curves used the growth parameters determined from the forced Gulland-Holt plots.

Parameter Hutt River Wellington Dam Warren River

Number of C. cainii 1275 5117 5308 measured

OCL mp(mm) 101.9 88.9 74.6

Kmp 0.42 0.32 0.23

t0 (month) 1.54 1.36 -2.69 C 0.37 0.69 1.0

ts 3.85 6.12 2.52 R2 0.99 0.99 0.99 Months to legal size (80 mm 27 42 75 CL) using modal progression

OCLgh (mm) 78 105

Kgh 0.11 0.14 Months to legal size (80 mm 147 66 CL) using Gulland-Holt -1 Zu (1year ) 0.41 0.18 0.42 -1 Ze (1year ) 1.79 0.47 0.59 F (1year-1) 1.38 0.29 0.17 0.77 0.62 0.29 E Accepted Manuscript -1 66 83 66 Su (% year )

-1 17 62 55 Se (% year )

Page 33 of 44 Table 2 Mark-recapture data from Wellington Dam and the Warren River using box traps between June 2005 and November 2006. N.B. All estimates are for fully recruited (trappable) animals (>~40 mm OCL). Productivity estimates for trappable C. cainii were calculated using density estimates from open population mark-recapture models and mean annual growth rates derived from mark-recapture methods, see text for details.

Mark-recapture Wellington Wellington Wellington Warren River Warren Warren information Dam females Dam males Dam females River males River TOTAL TOTAL Total number captured 1092 2137 3229 1524 2323 3847 Number marked 359 680 1039 505 667 1172 Mean size marked mm 47.9 (0.26) 50.6 (0.24) 49.7 (0.19) 56 (0.51) 55.7 (0.47) 55.8 (0.35) (±1 S.E.) Percentage recaptured 25.1 (90) 41.9 (285) 36.1 (375) 36.0 (182) 42.6 (284) 39.8 (466) (number) Mean time at liberty 178 (11) 197 (7) 193 (6) 217 (9) 253 (8) 239 (6) (days ±1 S.E.) Mean size mark- 50.1 (0.47) 52.1 (0.34) 51.9 (0.29) 60.9 (0.84) 62.4 (0.70) 61.9 (0.54) recaptured (mm ±1 S.E.) Percentage recaptured 46 (41) 47 (135) 47 (176) 80 (145) 69 (197) 73 (342) with ≥1 moult (number) Mean absolute growth 2.94 (0.53) 2.71 (0.22) 2.77 (0.21) 7.68 (0.56) 5.81 (0.37) 6.54 (0.32) rate (mm/yr ±1S.E.) Mean relative growth 6.1 (1.14) 5.3 (0.44) 5.48 (0.43) 12.8 (0.93) 9.5 (0.61) 10.80 (0.52) rate (% OCL /yr ±1S.E.) Mean annual weight 26.12 (5.00) 26.7 (2.29) 26.4 (2.11) 101.3 (8.37) 73.4 (5.15) 83.9 (4.58) increase (g ±1 S.E.) Trappable population 397.1 (84.28) 498.67 (35.20) 895.78 (91.34) 478.81 (58.44) 486.47 (43.76) 965.29 (73.00) estimate in December 2005 (± 1 S.E.)

Density estimate in 0.011 (0.002) 0.013 (0.001) 0.024 (0.002) 0.030 (0.004) 0.030 (0.003) 0.060 (0.005) December 2005 (m-2) Estimated productivity 0.28 (0.059) 0.36 (0.025) 0.63 (0.082) 3.03 (0.370) 2.23 (0.201) 5.07 (0.473) (g.m-2.year-1)

Accepted Manuscript

Page 34 of 44 Figure

1

Hutt River

Perth

Harvey Dam Wellington Dam

Margaret River Esperance

Big Brook Dam 0 100 200 k m Warren River 2 3 4 Fig. 1

Accepted Manuscript

1

Page 35 of 44 Wellington Dam Warren River 100 100 April April n = 442 75 n = 586 75

50 50

25 25

0 0 100 100 June June n = 420 75 n = 575 75

50 50

25 25

0 0 100 100 August August n = 775 75 n = 699 75

50 50

25 25

0 0 150 150 October October 100 100 n = 1025 n = 1021

50 50

0 Adundance 0 100 100 December December n = 1094 75 n = 589 75

50 50

25 25

0 0 150 150 February February n = 965 n = 1045 100 100

50 50

0 0 100 100 April April n = 511 75 n = 678 75

50 50 25 Accepted Manuscript25 0 0

0 10 20 30 40 50 60 70 80 90100 0 10 20 30 40 50 60 70 80 90100 Orbital carapace length (mm) Orbital carapace length (mm)

5 6 Fig. 2 7

2

Page 36 of 44 8 9 10

Warren River modal length 50 Warren River curve Wellington Dam curve Wellington Dam modal length 40

30

20

10

Orbital carapace length (mm) 0 S O N D J F M A M J J A S O N D J F M A Month

0 2 4 6 8 10 12 14 16 18

11 Age (months) 12 13 14 Fig. 3 15 16

Accepted Manuscript

3

Page 37 of 44 Mean size of recaptures Females 70 Males c c * Overall 60 b a * 50 40 30 20 10 0

Mean orbital carapace length (mm) length carapace orbital Mean Wellington Warren

Females Mean absolute growth 10 Males Overall c 8 d * 6

4 a a * 2

Mean absolute growth (mm/yr) growth absolute Mean 0 Wellington Warren

Females 16 Mean relative growth Males c Overall 14 12 * d 10 8 a 6 a * 4 Accepted2 Manuscript Mean relative growth (%OCL/yr) growth relative Mean 0 Wellington Warren

17 18 Fig. 4 19 20

4

Page 38 of 44 Warren River Females 18 Males 16 14 12 10 8 6 4 2 0 30 40 50 60 70 80 90 100

Wellington Dam 18 16 14 12 10 8 6 4

Absolute growth(mm/yr) 2 0 30 40 50 60 70 80 90 100

Combined sites 18 16 Warren River 14 Wellington Dam 12 10 8 Accepted6 Manuscript 4 2 0 30 40 50 60 70 80 90 100

21 Orbital carapace length (mm) 22 23 Fig. 5 24 5

Page 39 of 44 Warren River Females 60 Males

50

40

30

20

10

0 30 40 50 60 70 80 90 100

Wellington Dam 60

50

40

30

20

Relative growthOCL/yr) Relative (% 10

0 30 40 50 60 70 80 90

Combined sexes 50 Warren River 40 Wellington Dam

30 Accepted20 Manuscript 10

0 30 40 50 60 70 80 90 100

25 Orbital carapace length (mm) 26 27 Fig. 6 28

6

Page 40 of 44 Warren River Wellington Dam 40-50 mm 40 40

30 30

20 20

10 10

0 0

2 4 6 8 10 12 14 16 18 20 22 2 4 6 8 10 12 14 16 18 20 22

50-60 mm Frequency 40 40

30 30

20 20

10 10

0 0

2 4 6 8 10 12 14 16 18 20 22 2 4 6 8 10 12 14 16 18 20 22

40 60-70 mm Absolute growth (mm OCL)

30

20

Frequency 10

0

2 4 6 8 10 12 14 16 18 20 22

40 70-80 mm

30

20

10

0

2 4 6 8 10 12 14 16 18 20 22

40 80-90 mm 30 Accepted Manuscript 20

10

0

2 4 6 8 10 12 14 16 18 20 22 29 Absolute growth (mm OCL) 30 Fig. 7 31

7

Page 41 of 44 32 33 34

Warren River moult increment Wellington Dam moult increment Warren River mean moult interval Wellington Dam mean moult interval

7 0.82

6 0.80

0.78 5

0.76 4 0.74 3 0.72

Moult incrementMoult mode 2

0.70 (yrs)liberty at Mean time

1 0.68

0 0.66

40 50 60 70 80 90 Orbital carapace length category (mm) 35 36 37 Fig. 8 38 39 Accepted Manuscript

8

Page 42 of 44 40

1800 Warren River females Warren River males 1600 Wellington Dam females Wellington Dam males 1400

1200

1000

800

600

400

Trappable population abudance population Trappable

200

0

June August October December March May Month

0.12 Warren River females

)

-2 Warren River males Wellington Dam females 0.10 Wellington Dam males

0.08

0.06

0.04

0.02 Accepted Manuscript

Trappable population density (crayfish.m density population Trappable 0.00

June August October December March May

41 Month 42 Fig. 9

9

Page 43 of 44 43 Wellington Dam

Pre legal size 8 y = -0.183x + 8.51 r2 = 0.84 7 Post legal size y = -0.473x + 11.52 2 6 r = 0.94

5

4

ln (F/dt) ln 3

2

1

0 0 5 10 15 20 25 Relative age (t years)

Warren River Pre legal size 8 y = -0.421x + 9.01 r2 = 0.91 7 Post legal size y = -0.589x + 9.97 r2 = 0.98 6

5

ln (F/dt) ln 4

3

2 1 Accepted Manuscript 0 2 4 6 8 10 12 14 16 44 Relative age (t years) 45 Fig. 10 46 47

10

Page 44 of 44