Size based indicators rjan stman Dept. Aquatic Resources, regrund - PowerPoint PPT Presentation

Size based indicators Örjan Östman Dept. Aquatic Resources, Öregrund SLU

What fish size indicate? • Diet size dependend – the ecological function of a stock/community depends on size distributions • High mortality or selective mortality results in few large fishes • Poor feeding conditions result in slow growth • Swedish Agency for Marine and Water Mangement has ” natural size distributions” of stocks as a primary goal

Can size based indicators be used for ecological assessment? • Size, i.e. body length, readily available in Swedish monitoring data • What indicators to use? • Which asepects? • Drivers of size distributions?

Potential indicators % Median • Some property of the size Mean distribution, e.g. mean, median quantile (Q90) Quantile • Easy, just measure length • Mean length of mature fish (mL mat ) Size or proportion fish expected to be mature (%Mat) #Ind • Must know maturity pattern, sex and site specific Mature • Can be few/low proportion • Mean length in relation to length at maturity (mL/Mat) or number of fish expected to be mature (#Mat) Size

Perch (key species) & Pike, pikeperch, whitefish (large fish group) • Perch ubiquitous along the Swedish Baltic Sea coast, diet shift around 15-20 cm, mature earlier in Gulf of Bothnia (females 15 cm) than in BS Proper (females 25 cm) • Pike, pikeperch and whitefish lumped into one generic group of large fishes. Mature (90%) at 35 cm

High redundancy among indicators across sites Variables factor map (PCA) Variables factor map (PCA) 1.0 1.0 p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat np_cpue_m90 np_cpue_m90 np_cpue_m90 np_cpue_m90 np_cpue_m90 np_cpue_m90 p_mL_M90 p_mL p_MedianL p_mL p_MedianL p_mL p_MedianL np_cpue_m90 np_cpue_m90 Dim 2 (20.70%) p_mL_M90 p_mL_M90 p_mL p_MedianL p_mL p_MedianL p_mL p_MedianL np_cpue_m90 p_mL_M90 p_mL_M90 p_mL_M90 p_mL p_MedianL p_mL p_MedianL p_mL_M90 p_mL_M90 p_mL_M90 p_mL p_MedianL 0.5 p_L90 p_L90 p_L90 p_cpue_m90 p_L90 p_L90 p_L90 p_cpue_m90 p_cpue_m90 p_L90 p_L90 p_mLmat p_mLmat p_mLmat Dim 4 (6.92%) p_cpue_m90 p_cpue_m90 p_cpue_m90 p_L90 p_mLmat p_mLmat p_mLmat 0.5 p_cpue_m90 p_cpue_m90 p_cpue_m90 p_mLmat p_mLmat p_mLmat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat p_P_Mat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat p_cpue_m90 p_cpue_m90 p_cpue_m90 p_cpue_m90 p_cpue_m90 p_cpue_m90 p_cpue_m90 p_cpue_m90 np_L90 p_cpue_m90 np_L90 np_L90 np_L90 np_L90 np_L90 np_L90 np_L90 np_L90 0.0 p_MedianL p_MedianL p_MedianL p_MedianL p_MedianL p_MedianL p_MedianL p_MedianL p_MedianL 0.0 p_mL p_mL p_mL p_mL_M90 p_mL_M90 p_mL_M90 p_mL p_mL p_mL p_mL_M90 p_mL_M90 p_mL_M90 p_mL p_mL p_mL p_mL_M90 p_mL_M90 p_mL_M90 p_L90 p_L90 p_L90 p_L90 p_L90 p_L90 np_P_Mat np_P_Mat np_P_Mat p_L90 p_L90 p_L90 np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_cpue_m90 np_cpue_m90 np_cpue_m90 np_mL np_MedianL np_mL np_MedianL np_mL np_MedianL np_cpue_m90 np_cpue_m90 np_cpue_m90 np_mL np_MedianL np_mL np_MedianL np_mL np_MedianL np_cpue_m90 np_cpue_m90 np_mL_M90 np_mL_M90 np_mL_M90 np_mL np_MedianL np_mL np_MedianL np_mL np_MedianL np_cpue_m90 np_mL_M90 np_mL_M90 np_mL_M90 np_mL_M90 np_mL_M90 np_mL_M90 p_mLmat p_mLmat p_mLmat p_mLmat p_mLmat p_mLmat p_mLmat p_mLmat np_L90 p_mLmat np_mL_M90 np_mL_M90 np_mL_M90 np_L90 np_L90 np_mL_M90 np_mL_M90 np_mL_M90 np_L90 np_L90 np_L90 np_L90 np_L90 np_mL_M90 np_mL_M90 np_mL_M90 np_L90 np_MedianL np_MedianL np_MedianL np_MedianL np_MedianL np_MedianL np_MedianL np_MedianL np_MedianL np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mLmat np_mL np_mL np_mL np_mLmat np_mL np_mL np_mL np_mL np_mL np_mL np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat np_P_Mat -1.0 -1.0 -2 -1 0 1 2 -2 -1 0 1 2 Dim 1 (48.33%) Dim 3 (11.06%)

I chose… mTemp GearNL GearTN 1 • Length at 90 percentile (L90) TOTP np_L90 np_mLmat np_mL np_MedianL np_cpue_m90 p_mLmat p_L90 p_mL p_MedianL 0 0 p_P_Mat GearNN np_mL_M90 • Mean length relative size-at- Lat GearKÖ -1 RDA2 -2 np_P_Mat maturation (mL/M90) p_cpue_m90 -4 • Catch per unit effort of mature -6 fish (cpue_m90) p_mL_M90 -10 -5 0 5 • Differ between gears and RDA1 latitude

Temporal drivers? 4 3 2 3 • At three long-term 1 2 L90 Perch L90 NonPerch 0 monitoring sites (Holmön and 1 -2.5 -1.5 -0.5 0.5 1.5 2.5 -1 0 Forsmark, Coastal nets, Kvädö -3 -2 -1 0 1 2 -2 -1 -3 Salinity and Vinö, Net links) -2 N load 4 4 3 • Negative associations with 3 CPUE M90 Perch 2 CPUE M90 NonPerch 2 1 nutrient loadings over time 1 0 0 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 -1 -1 -2 -2 N load P load • mL/M90 for both perch and 4 5 3 4 large fish positive with mL/M90 Perch 2 mL/M90 NonPerch 3 1 2 salinity? 0 1 -2.5 -1.5 -0.5 0.5 1.5 -1 0 -2 -3 -2 -1 0 1 2 3 -1 -3 -2 -4 P load P load -3

Cont • Several indicators negative Non- Non- Driver Perch Perch Perch Non- perch perch associations with temperature variable L 90 CPUE m90 mL m90 perch L 90 CPUE m90 mL m90 471.0 448.1 447.6 451.1 478.0 418.7 AICc 0.09 0.10 0.20 0.10 0.07 0.14 Adj. R 2 • Perch positive autocorrelation -10.0 -3.1 -16.2 +1.8 +1.2 +1.3 φ(t -1) Hydro- +1.5 -2.2 (+) -6.0 (-) -11.2(-) -11(-) -1.8 (-) Temp climate -2.4(+) +2.0 -0.6 (+) +1.8 +1.2 -4.2(+) variables Salinity +0.6 +1.9 +0.4 +1.9 +0.6 +1.8 BSI Anthropo +0.6 +1.3 +0.6 +0.4 -1.2(-) +1.2 Secchi -genic +2.0 +0.9 +2.0 +0.1 -7.2(-) +2.0 divers Nload +1.5 -0.6(-) -4.2 (-) -15.7(-) +1.6 -18 (-) Pload +1.4 +1.3 +0.7 +1.8 +0.8 +1.5 DIN +0.7 +0.6 +0.1 +2.0 +1.5 +1.9 DIP Landing +0.9 +0.2 +0.8 +1.9 +1.7 +0.9 s

No-take areas/refererence area Perch Non-perch Perch L 90 Perch mL m90 Non-perch L 90 Non-perch mL m90 CPUE m90 CPUE m90 Licknevarp NN (2013) NTA 32 cm 8.5 96% 49 cm 0.55 105% (26 cm) (1.5) (80%) (39 cm) (0.14) (66%) Kvädö Lännåker NN (2012-2015) 30 cm 1.5 84% 32 cm 0.34 53% NTA (31 cm) (3.75) (92%) (40 cm) (0.14) 62% Before Lännåker TN (2012-2015) 27 cm 8.6 82% 73 cm 3.0 105% NTA Before (30 cm) (20.2) (94%) (71 cm) (1.6) (115%) Storjungfrun NN (2013-2015) NA NA NA 43 cm 0.32 104% (30 cm) (0) Before Kalvhararna NN (2013-2015) 33 cm 0.84 170% 39 cm 0.19 95% Before (28 cm) (0.48) (166%) (37 cm) (0.03) (92%)

GES bounadries reference areas Long time-series Perch Non-perch Non-perch Perch L 90 Perch mL m90 Non-perch L 90 CPUE m90 CPUE m90 mL m90 Kvädö Net link 22 cm (3) 0.3 (2) 72% (1) 38 cm (4) 0.10 (2) 73% (7*) Kvädö Nordic Net 24 cm (3) 0.75 (2) 77% (2) 35 cm (1) 0.05 (1) 59% (2*) Vinö Net link 21 cm (1) 0.5 (1) 71% (1) 39 cm (1) 0.14 (3) 71% (1) Forsmark Coast net 24 cm (2) 2 (3) 75% (3) 31 cm (3) 0 (0) 57% (2) Forsmark Nordic Net 26 cm (1) 1.7 (0) 79% (2) 25 cm (1) 0 (0) 49% (3) Holmön Coast Net 23 cm (3) 30 (1) 125% (3) 38 cm (3) 0.04 (4) 80% (2) Holmön Nordic net 25 cm (2) 11 (1) 125% (1) 35 cm (3) 0.05 (3) 85% (3) (Years below boundary, n = 22-28 for Net link/ Coastal nets, n = 15 for Nordic nets) CS, NL: 22-23 cm NA 70% 38 cm 0.05? 70% NN: 24 cm NA 75% 35 cm 0.05? 70%

Nordic nets spatial comparison Gulf of Non-perch Non-perch Perch L 90 Perch CPUE m90 Perch mL m90 Non-perch L 90 CPUE m90 mL m90 Bothnia MPA (N = 4) 28.5 ± 1.2 7.7 ± 5.2 135 ± 17 42 ± 2.2 0.11 ± 0.05 96.3 ± 1.8 REF (N = 6) 25.7 ± 0.7 11.4 ± 2.2 123 ± 8.1 43 ± 2.9 0.13 ± 0.03 70.2 ± 4.4 URB (N = 3) 26.3 ± 1.2 12.2 ± 2.6 136 ± 3.9 33.7 ± 3.5 0.27 ± 0.10 69.7 ± 9.8 Baltic Sea Proper EUT (N = 4) 26 ± 1.2 1.0 ± 0.4 80.5 ± 1.2 36.5 ± 2.4 0.46 ± 0.07 66.6 ± 1.8 MPA (N = 5) 24.75 ± 1.9 1.45 ± 0.5 80.6 ± 3.5 47 ± 1.9 0.36 ± 0.03 109 ± 12 NOT (N =3) 29.7 ± 1.2 3.65 ± 2.4 86.9 ± 4.8 57.3± 16 0.35 ± 0.1 102 ± 27 OTH (N =7) 26.3 ± 0.7 1.4 ± 0.3 81.1 ± 1.6 41.8 ± 1.1 0.30 ± 0.06 84 ± 3.8 REF (N = 15) 24.9 ± 0.4 1.6 ± 0.3 78.6 ± 0.8 41.25 ± 2.0 0.11 ± 0.03 70.4 ± 5.9 URB (N = 4) 27.3 ± 0.6 1.7 ± 0.5 83.6 ± 1.3 57.5 ± 3.9 0.25 ± 0.02 105 ± 6.3

Summary pros • Size based indicators use readily (in Sweden) available information • Associated with nutrients and temperature in time-series • Respond positively to no-take areas • Identify non-site specific boundaries for L90 seem possible • Maybe non-site specific boundaries for mL/M90 as well • Good complement to abundance based indicators

Summary cons • Gear specific • Some human impacted sites show quite good status based on size distributions • Need size distributions • Would need more/better site-speific life-history data

Summary