Pollack report 13/06/2023
Simon Thomas
13/06/2023
FISP pollack
Pollack represents an economically and socially important fishery across the UK. Recreational and commercial fishers have highlighted a decline in Pollack (Pollachius pollachius) across the English channel. ICES stock assessments have similarly indicated there is insufficient evidence to identify exploitation and population trends, however have highlighted a decline of 72% in commercial landings since 1968. Data collection is required on life history parameters, abundance and spatial ecology to ensure sustainability in commercial and recreational fisheries.
Data summary
The pollack FISP project is a collaboration between charter skippers and scientists to address ongoing concerns of both the recreational and commercial fleet regarding the stock status for the species.13 skippers have already collected length data on over 6266 pollack from 323 trips, this will allow the assessment of stock abundance, composition and recruitment. 20.2 % of pollack were returned alive.
The total catch during this time period was 12.31 tonnes
The mean size of P. pollachius captured was 55.92 cm, equating to 3.28 years of age. The size distribution is shown in Figure 1.
There have been 10 otoliths collected from 5 fish as of 2023-06-14.
Figure 1. Size distribution of P. pollachius captured in ICES sea area viie from April 2022 to may 2023. Normal distributions are indicated by red lines.
The median size of P. pollachius captured was 58 cm (SD =13.74), equating to 3.3 years of age (SD =1.53). There was a significant difference in size of P. pollachius captured between wreck and reef marks (Table 1.)
Figure 2. Age distribution of P. pollachius captured in ICES sea area viie from April 2022 to may 2023. Normal distributions are indicated by red lines.
Table. 1. ANOVA for size difference between marks for P. pollachius.
| Df | Sum Sq | Mean Sq | F value | Pr(>F) |
|---|---|---|---|---|
| 1 | 18502.15 | 18502.146 | 99.585 | 0 |
| 6264 | 1163809.31 | 185.793 | NA | NA |
P. pollachius sizes followed a normal distribution (Figure 1.). The lack of evidence for knife edge selection suggest a degree of non-specificity of the gear used. The maximum sized fish captured during the study period was 107 cm and the the minimum size was 12 cm.
Figure 3. Q-Q plot for P. pollachius size distribution. Solid black line indicates theoretical normal distribution. N= 6266.
P. pollachius age distribution does not follow a normal distribution (Figure 4.) and is positively skewed. The mean age is 3.28 years (Stdev 1.53 years) with the median age of 3.3 years.The positive skew in age distribution is more obvious in fish captured from wrecks (Figure 2.). However as the age is calculated from a rearrangement of the von Bertalanffy equation,
\(Age=ln(1-Length/ Linf))/-K +T0\)
this result may be an artifact and needs to be validated by otolith analysis.
Table. 2. Catch Per Unit Effort (CPUE (Kg/Trip)) per month
| Month | Median | SDdev | Mean | Max | Min |
|---|---|---|---|---|---|
| 1 | 11.72 | 17.57 | 20.09 | 51 | 5.82 |
| 2 | 24.57 | 26.20 | 29.03 | 89 | 4.41 |
| 3 | 46.32 | 53.80 | 59.40 | 151 | 1.00 |
| 4 | 38.53 | 50.05 | 41.03 | 275 | 0.00 |
| 5 | 38.55 | 33.07 | 46.38 | 167 | 0.00 |
| 6 | 42.94 | 42.98 | 48.39 | 216 | 0.57 |
| 7 | 19.67 | 22.56 | 24.84 | 101 | 0.00 |
| 8 | 11.22 | 13.60 | 15.33 | 49 | 0.21 |
| 9 | 11.72 | 15.23 | 14.71 | 57 | 0.00 |
| 10 | 25.06 | 14.15 | 23.71 | 50 | 10.51 |
| 11 | 12.73 | 26.46 | 21.24 | 66 | 0.22 |
| 12 | 4.79 | 5.12 | 4.93 | 13 | 0.61 |
Catch Per Unit Effort
Raw Catch Per Unit Effort (median Kg/ Trip) showed an increasing trend during February, which was sustained until June when catches decreased during the summer months. The raw median CPUE per trip peaked in March with a nadir in August (Figure 5.). The adjusted CPUE (Kg/trip/time spent targeting P. pollachius) peaked in March, with a nadir in September (Figure 6.).Correcting for numbers of anglers produced confusing results, as the relationship between angler numbers and catches is non-linear, due to disturbance of the shoal by additional fishing gear.
Table. 3 Adjusted CPUE (Kg/Trip/Angler) per month
| Month | CPUE | Confint | Upper | Lower |
|---|---|---|---|---|
| 1 | 3.93 | 1.59 | 5.52 | 2.35 |
| 2 | 4.16 | 3.48 | 7.63 | 0.68 |
| 3 | 5.79 | 5.36 | 11.16 | 0.43 |
| 4 | 4.43 | 3.07 | 7.49 | 1.36 |
| 5 | 4.81 | 3.01 | 7.81 | 1.80 |
| 6 | 5.67 | 1.86 | 7.53 | 3.81 |
| 7 | 3.92 | 1.35 | 5.26 | 2.57 |
| 8 | 2.91 | 0.90 | 3.81 | 2.01 |
| 9 | 1.54 | 0.88 | 2.42 | 0.65 |
| 10 | 3.58 | 1.52 | 5.10 | 2.06 |
| 11 | 1.72 | 2.01 | 3.73 | -0.30 |
| 12 | 0.75 | 2.34 | 3.09 | -1.59 |
Figure 4. Q-Q plot for P. pollachius age distribution. Solid black line indicates theoretical normal distribution. N= 6266.
Figure 5. Raw median monthly CPUE (Kg/Trips). Ribbon indicates 95% confidence intervals
Figure 6. Median CPUE (Kg/ trip) adjusted for fishing time spent targeting P. pollachius . Ribbon indicates 95% confidence intervals
Von Bertalanffy growth parameters through ELEFAN
Von Bertalanffy growth parameters were obtained using ELEFAN for a subgroup of fish captured from May 2022 and May 2023. 95% confidence intervals were obtained by using the Jack knife procedure The results are shown is Table 4. The mean values for Linf (88.57 cm), K (0.168 year\(^{-1}\)) and PhiL (3.105) are similar to values from Southern Ireland in Fishbase. Natural mortality (M) was calculated as 0.25 year\(^{-1}\) (+/- 0.014).Total mortality (Z) was 0.616 year\(^{-1}\) giving a fishing mortality of 0.361 year\(^{-1}\) (+/- 0.014).
Table 4. Von Bertalanffy constants for P. pollachius from ICES sea area viie from May 2022 and May 2023.
| Five | Mean | Ninety_five | |
|---|---|---|---|
| Linf | 84.610 | 88.570 | 92.510 |
| K | 0.150 | 0.168 | 0.180 |
| t_anchor | 0.260 | 0.614 | 0.970 |
| PhiL | 3.070 | 3.105 | 3.140 |
| M | 0.241 | 0.255 | 0.269 |
| Z | 0.616 | 0.616 | 0.616 |
| F | 0.375 | 0.361 | 0.347 |
Figure 7. Size distribution of P. pollachius by mark from April 2022 to May 2023. Horizontal red lines show the mean size for wreck (Black) and reef (Red) caught fish.
The size distribution of through the study period is shown in figure 7. The largest fish are associated with the deeper wreck marks and the size of fish captured from wrecks decreased from April to July 2022 before increasing during January 2023. A boat from Penzance started providing data in January 2023 and much of the increase in sizes of wreck caught fish can be attributed to this new skipper fishing the far west of ICES sea area viie.
In both 2022 and 2023, the peak size and numbers of reef-caught fish peaked during May, which coincides with the arrival of large amounts of prey species on these reef systems, mainly Ammodytes and Hyperoplus sp. especially the sand launce Hyperoplus lanceolatus.
Hierarchical Generalised Additive Models (HGAM)
In order to determine differences in P. pollachius catches across ICES sea area viie, Hierarchical Generalised Additive Models (HGAM) were used to test five hypothesis.
There is a common pattern of catches across the area for the year.
There is a common pattern of catches across the area for the year, but with variations between ICES rectangles which share a common pattern.
There is a common pattern of catches across the area for the year, but with variations between ICES rectangles have different patterns.
There are no common patterns of catches across the area for the year.
To test these hypotheses the following models were used:
P0: NULL model: sqrt (CPUE)~s(ICES_Rectangle, bs=“re”)
P1: Square root (CPUE)~s(Day of Year, k=30)+s(ICES_Rectangle, bs=“re”)
P2: Square root (CPUE)~s(Day of Year, k=50,m=2)+s(Day of Year,ICES_Rectangle, bs=“fs” ,m=2)
P3: Square root (CPUE)~s(Day of Year, k=50)+s(Day of Year,ICES_Rectangle, k = 5, bs=“fs”,m=2)
P4: Square root (CPUE)~s(Day of Year, ICES_Rectangle, k=5, bs=“fs”, m=2)
nb. K= degree of “wiggliness, m=2 Marginal Thin Plate Regressive Spline basis for term is penalised on the squared second derivative.
Table 5. ANOVA of HGAM models for P. pollachius CPUE from ICES sea area viie.
## Analysis of Deviance Table
##
## Model 1: sqrt(Trip_CPUE) ~ s(ICES_Rec, bs = "re")
## Model 2: sqrt(Trip_CPUE) ~ s(DOY, k = 30) + s(ICES_Rec, bs = "re")
## Model 3: sqrt(Trip_CPUE) ~ s(DOY, k = 50, m = 2) + s(DOY, ICES_Rec, bs = "fs",
## m = 2)
## Model 4: sqrt(Trip_CPUE) ~ s(DOY, k = 50) + s(DOY, ICES_Rec, k = 5, bs = "fs",
## m = 2) + ICES_Rec
## Model 5: sqrt(Trip_CPUE) ~ s(DOY, ICES_Rec, k = 5, bs = "fs", m = 2)
## Resid. Df Resid. Dev Df Deviance F Pr(>F)
## 1 333.44 3330.2
## 2 326.46 2932.2 6.9717 398.06 7.0401 8.754e-08 ***
## 3 310.87 2571.5 15.5971 360.71 2.8516 0.0002687 ***
## 4 320.89 2722.6 -10.0238 -151.14 1.8592 0.0501444 .
## 5 322.73 2782.5 -1.8391 -59.85 4.0124 0.0219405 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Table 5. indicates that the HGAM models are significantly different from the Null model (P<0.05) and model P2 has the best fit to the data (Table 6.) and that the residuals for this data are (square root) normally distributed (Figure 8.), with some positive skew. A summary of the model is shown in Table 7.
This suggests that catches of P. pollachius share a common pattern across ICES sea area viie but there are different some variations in catches present between ICES rectangles.This is suggestive of a common stock which disperses differently in geographical areas (Hypothesis 3.). However, the data are still sparse in some areas and the errors large, so these findings should be treated with caution.
Table 6. Comparison of GAM models for P. pollachius CPUE from ICES sea area viie.
##
## Pollack CPUE models
## ======================================================
## Dependent variable:
## ---------------------------------------
## sqrt(Trip_CPUE)
## GAM
## (continuous)
## Null P1 P2 P3 P4
## (1) (2) (3) (4) (5)
## ------------------------------------------------------
## ICES_Rec28_E6
##
##
## ------------------------------------------------------
## AIC 1766.4 1734 1712.4 1718.4 1720.9
## Observations 342 342 342 342 342
## Adjusted R2 0.10 0.20 0.27 0.24 0.23
## Log Likelihood -883.21 -867.01 -856.19 -859.20 -860.43
## UBRE 10.19 9.28 8.75 8.88 8.94
## ======================================================
Figure 8. QQ plot of pearson residuals from GAM model P2 for P. pollachius CPUE from ICES sea area viie.
Figure 9. HGAM model P2 left: Day of Year; Right: Day of Year by ICES rectangle for P. pollachius CPUE from ICES sea area viie.
There is considerable uncertainty in the model P2 at both extremes of the year (Figure 9) which reflects the restricted data due to inclement weather conditions during this period. The deviance explained by model P2 is 0.3 with an adjusted R2 of 0.268
Spawning season 2022-2023
Up to 14/06/2023 198 P. pollachius containing roe were reported by the skippers. The smallest fish reported as mature was 32 cm. The first fish containing spawn during the study was captured on 2022-12-11 whilst the latest capture of a spawning fish was 2023-05-14.
The pollack containing identifiable roe after 2023-04-07 were all male with roe described as spent (Figure 10)
Figure 10. Temporal distribution of spawning for P. pollachius captured from ICES sea area viie.
Historical information
North Wales
There were 208 P. pollachius captured from ICES sea area viia during this study of which 85 were identified as female and 94 identified as males. The sex of fish was not ascertained for 29 fish.
The overal mean total length of fish sampled during the study was 51.88 cm and the median total length was 51.65 cm.
The total lengths (cm) by sex are shown in Table.8.
Table 8. Summary length data (cm) by sex of P. pollachius captured from ICES sea area viia during 1989
| Sex | Mean | Median | StDev |
|---|---|---|---|
| Female | 51.68 | 52.69 | 9.99 |
| Male | 52.05 | 51.65 | 10.07 |
Figure 11. Weight distribution (kg) of P. pollachius from ICES sea area viia during 1989.
Figure 12. Maturity ogive for P. pollachius from ICES sea area viia during 1989. blue dotted line indicates length at 50% maturity and black dotted line at 90% maturity
The length which 50% of pollack were mature at was 45.49 cm and 90% maturity occurred at 50.04 cm (Figure 12.)
Figure 13. Maturity ogive for P. pollachius by sex from ICES sea area viia during 1989. blue dotted line indicates length at 50% maturity and black dotted line at 90% maturity
Figure 13. Confidence intervals for growth from VB statistics from P. pollachius captured from ICES sea area viia during 1989. Grey shading indicate 95% confidence intervals. Points are otolith age data
The von Bertalanffy growth constants obtained from otolith-derived ages are show in Table 9.
Table 9. Von Bertalanffy constants for P. pollachius from ICES sea area viia during 1989
## Nonlinear regression model
## model: tl ~ vb(age, Linf, K, t0)
## data: wf14T
## Linf K t0
## 97.3006 0.1679 -0.8444
## residual sum-of-squares: 3107
##
## Number of iterations to convergence: 5
## Achieved convergence tolerance: 1.833e-06