1University of Queensland, School of Veterinary Science, Gatton, QLD 4343, Australia

Integrating acoustic telemetry into mark-recapture models to improve the precision of apparent survival and abundance estimates

Christine L. Dudgeon1*, Kenneth H. Pollock2, J. Matias Braccini3, Jayson M. Semmens4, Adam Barnett5,6

1University of Queensland, School of Veterinary Science, Gatton, QLD 4343, Australia

2Department of Applied Ecology, North Carolina State University, Raleigh, NC 27695-7617, USA

3Western Australian Fisheries and Marine Research Laboratories, PO Box 20, North Beach, WA 6920, Australia

4Fisheries and Aquaculture Centre, Institute for Marine and Antarctic Studies, University of Tasmania, Hobart, Tasmania, 7000, Australia

5School of Life and Environmental Sciences, Deakin University, Melbourne, Victoria, Australia,

6 Centre for Tropical Water & Aquatic Ecosystem Research (TropWATER), Estuary and Tidal Wetland Ecosystems Research Group, School of Marine and Tropical Biology, James Cook University, Townsville, Queensland 4811, Australia

*Corresponding author.

E-mail:

Phone: + 61 423366398

Fax: + 61 (7) 5460 1922

Electronic Supplementary Material - 1

Cormack-Jolly-Seber (CJS) AIC model lists for the 4 different time series as corresponds with Table 2: (i) all 3 Years; (ii) Years 2 and 3 combined; (iii) Year 2 only and (iv)Year 3 only. Model lists are presented for the (1) Longline only data; (2) Acoustic telemetry only data and (3) the combined analysis with both data types included as two groups. Model parameters are apparent survival (φ), recapture probability (p), modeled as constant (.) or varying over time (t) and with group effect (g). The output values are given for the Akaike Information Criterion adjusted for small sample size (AICc); change in AICc from the most parsimonious model ( AICc); Akaike weights (AICc weight); Model Likelihood values; number of parameters (# Par) and model Deviance.

1. Longlineonly data

(i) Time series: 3 Years

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ(.)p(.)275.655 0.00 1.000 1.000 2 130.731

φ(.)p(t) 301.601 25.95 0.000 0.000 27 100.860

φ(t)p(.)316.504 40.85 0.000 0.000 27 115.763

φ(t)p(t) 351.661 76.01 0.000 0.000 49 92.017

(ii) Time series: Years 2&3

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ (.) p(.)148.877 0.00 0.999 1.000 2 65.409

φ (.) p(t)162.552 13.68 0.001 0.000 17 44.728

φ (t) p(.)176.694 27.82 0.000 0.000 17 58.870

φ (t) p(t)189.120 40.22 0.000 0.000 28 40.920______

(iii) Time series: Year 2 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ (.) p(.)48.461 0.00 0.955 1.000 2 21.474

φ (.) p(t)54.572 6.11 0.045 0.047 9 11.499

φ (t) p(.)63.899 15.44 0.001 0.000 9 20.826

φ (t) p(t) 70.750 22.29 0.000 0.000 15 11.499

(iv)Time series: Year 3 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ (.) p(.)36.453 0.00 0.979 1.000 2 14.901

φ (.) p(t)44.902 8.45 0.015 0.015 8 8.513

φ (t) p(t) 47.751 11.30 0.004 0.004 9 8.513

φ (t) p(.)47.876 11.42 0.004 0.003 8 11.487______

2. Acoustic data only

(ii) Time series: Years 2&3

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ (.) p(t)297.601 0.00 1.000 1.000 16 197.066

φ (.) p(.)313.515 15.91 0.000 0.000 2 247.825

φ (t) p(t) 318.830 21.23 0.000 0.000 24 192.127

φ (t) p(.)319.642 22.04 0.000 0.000 8 240.407______

(iii) Time series: Year 2 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ (.) p(t)104.362 0.00 0.954 1.000 8 36.894

φ (.) p(.)110.554 6.19 0.043 0.045 2 58.425

φ (t) p(.)116.524 12.16 0.002 0.002 8 49.056

φ (t) p(t)120.509 16.15 0.000 0.000 13 36.241______

(iv) Time series: Year 3 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ (.) p(.)100.216 0.00 0.997 1.000 2 56.608

φ (.) p(t)112.675 12.46 0.002 0.002 8 53.354

φ (t) p(.)114.171 13.96 0.001 0.001 8 54.849

φ (t) p(t)129.466 29.25 0.000 0.000 13 52.402______

3. Acoustic and Longline data combined

(ii) Time series: Years 2&3

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ(.)p(g) 427.353 0.00 0.685 1.000 3 302.207

φ(g)p(g) 429.280 1.93 0.261 0.382 4 302.059

φ(.)p(g+t) 432.946 5.59 0.042 0.061 33 236.036

φ(g)p(g+t) 435.436 8.08 0.012 0.01834 235.739

φ(t)p(g) 453.859 26.51 0.000 0.000 18 295.468

φ(g)p(t) 462.328 34.97 0.000 0.000 18 303.937

φ(t)p(g+t) 466.541 39.19 0.000 0.000 46 230.936

φ(g+t)p(g) 488.942 61.59 0.000 0.000 34 289.245

φ(g)p(.) 490.583 63.23 0.000 0.000 3 365.436

φ(g+t)p(g+t) 499.994 72.64 0.000 0.000 57 226.884

φ(g+t)p(t) 506.146 78.79 0.000 0.000 46 270.540

φ(.)p(t) 536.850 109.50 0.000 0.000 17 380.828

φ(g+t)p(.) 538.411 111.06 0.000 0.000 33 341.501

φ(t)p(t) 553.288 125.94 0.000 0.000 31 361.864

φ(.)p(.) 571.983 144.63 0.000 0.000 2 448.891

φ(t)p(.) 584.313 156.96 0.000 0.000 17 428.291______

(iii) Time series: Year 2 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ(g)p(g) 140.786 0.00 0.450 1.0004 78.223

φ(.)p(g)141.836 1.05 0.266 0.592 3 81.416

φ(.)p(t+g)142.470 1.68 0.193 0.431 17 48.141

φ(g)p(g+t)144.827 4.04 0.060 0.133 18 47.709

φ(t)p(g)146.297 5.51 0.029 0.064 10 70.032

φ(g)p(t)152.994 12.21 0.001 0.002 10 76.729

φ(t)p(t+g)157.247 16.46 0.000 0.000 23 45.293

φ(t)p(g+t)157.247 16.46 0.000 0.000 23 45.293

φ(g+t)p(g)162.256 21.47 0.000 0.000 18 65.138

φ(.)p(t)165.816 25.03 0.000 0.000 9 91.940

φ(g+t)p(g+t)169.525 28.74 0.000 0.000 27 44.516

φ(g+t)p(t)176.160 35.37 0.000 0.000 22 67.299

φ(g)p(.)176.537 35.75 0.000 0.000 3 116.118

φ(t)p(t)177.712 36.93 0.000 0.000 15 88.797

φ(g+t)p(.)195.311 54.52 0.000 0.000 17 100.982

φ(t)p(.)195.825 55.04 0.000 0.000 9 121.950

φ(.)p(.)203.224 62.44 0.000 0.000 2 144.910______

(iv) Time series: Year 3 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par Deviance______

φ(.)p(g)133.783 0.00 0.729 1.000 3 73.645

φ(g)p(g)135.762 1.98 0.271 0.372 4 73.444

φ(.)p(t+g)154.800 21.02 0.000 0.000 15 64.850

φ(g)p(t+g)156.763 22.98 0.000 0.000 16 63.914

φ(g)p(.)158.792 25.01 0.000 0.000 3 98.654

φ(g)p(t)168.802 35.02 0.000 0.000 9 94.810

φ(t)p(t+g)169.341 35.56 0.000 0.000 20 64.116

φ(t+g)p(t+g)181.980 48.20 0.000 0.000 24 62.962

φ(t)p(g)188.162 54.38 0.000 0.000 26 61.640

φ(.)p(.)196.532 62.75 0.000 0.000 2 138.527

φ(.)p(t)205.781 72.00 0.000 0.000 8 134.232

φ(t)p(.)219.060 85.28 0.000 0.000 25 96.345

φ(g+t)p(g)234.944 101.16 0.000 0.000 38 52.141

φ(g+t)p(.)236.755 102.97 0.000 0.000 37 59.547

φ(t)p(t)265.578 131.80 0.000 0.000 39 76.977

φ(g+t)p(t)350.343 216.56 0.000 0.000 54 39.968______

Electronic Supplementary Material - 2

Jolly-Seber (JS) AIC model lists for the 4 different time series as corresponds with the parameter estimates shown in Table 3: (i) all 3 Years; (ii) Years 2 and 3 combined; (iii) Year 2 only and (iv) Year 3 only. Model lists are presented for the (1) Longline only data. Model parameters are apparent survival (φ), recapture probability (p), permanent entry (β) modeled as constant (.) or varying over time (t). β(0) denotes the JS model with losses only where all permanent immigration is restricted to prior sampling. The output values are given for the Akaike Information Criterion adjusted for small sample size (AICc); change in AICc from the most parsimonious model (AICc); Akaike weights (AICc weight); Model Likelihood values; and number of parameters (# Par).

(i) Time series: 3 Years

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par_____

φ(.)p(t)β(0) 433.197 0.00 0.490 1.000 29

φ(t)p(t)β(0) 434.177 0.98 0.300 0.613 32

φ(.)p(t)β(t) 434.901 1.70 0.209 0.427 33

φ(.)p(.)β(0) 970.450 537.25 0.000 0.000 2______

(ii) Time series: Years 2&3

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par_____

φ(.)p(t)β(0) 244.290 0.00 0.923 1.000 19

φ(.)p(t)β(t) 249.267 4.98 0.077 0.083 22

φ(t)p(t)β(t) 265.920 21.63 0.000 0.000 30

φ(.)p(.)β(0) 301.562 57.27 0.000 0.000 3______

(iii) Time series: Year 2 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par_____

φ(.)p(t)β(0) 98.395 0.00 0.914 1.000 10

φ(t)p(t)β(t) 103.249 4.85 0.081 0.088 12

φ(.)p(t)β(t) 108.604 10.21 0.006 0.006 14

φ(.)p(.)β(0) 140.292 41.90 0.000 0.000 2______

(iv) Time series: Year 3 only

______

AICc Model

Model AICc ΔAICc Weight Likelihood # Par_____

φ(.)p(t)β(0) 82.206 0.00 0.662 1.000 9

φ(.)p(t)β(t) 83.638 1.43 0.327 0.489 11

φ(t)p(t)β(t) 89.984 7.78 0.014 0.021 13

φ(.)p(.)β(0) 96.873 14.67 0.000 0.001 3______