Stock and resource risk assessments of Kawakawa (Euthynnus affinis) and longtail tuna (Thunnus tonggol) in Southeast Asian waters using ASPIC. Alphabetical order by surname) Sallehudin bin Jamon (Malaysia). Stock and resource risk assessments of Kawakawa (Euthynnus affinis) and longtail tuna (Thunnus tonggol) in Southeast Asian waters using ASPIC. Then, the LOT in the Pacific Ocean is also in a safe position as the previous assessment in 2016.
Finally, the LOT in the Indian Ocean was in a safe situation (green area) compared to previous assessments in 2016, which was in a severely overfished situation (red area). The results should be viewed with caution due to uncertainties in the overall process. As can be seen in the comparisons between past and current assessments, the result from the stock and risk assessment shows that the current catch level of kawakawa in both areas should be considered reduced.
This practical workshop is one of the key activities in the SEAFDEC neritic tuna project. As a reference for the SEAFDEC neritic tuna project, we provide the summary of the stock status of four neritic tuna species (KAW, LOT, Spanish narrow-band mackerel and Indian Ocean king mackerel), including those in the IOTC, in Annex A. Member countries with high produce catches (>98% of total catch) by species and area (listed in order). 2) Uncertainties in the catch data.
METHODS
CPUE standardization
We use the original ASPIC application and batch software to perform the four inventory estimates. Initial values for MSY and K (point, minimum and maximum values), B1/K and q are selected using criteria suggested by experts. In serial jobs, four parameters are combined to find the most probable one in addition to the two excess production models (Schaefer or Fox).
All results are saved in Excel file including four estimated parameters, estimated metrics (total biomass, F, MSY, Fmsy, TBmsy, TB/TBmsy, F/Fmsy and [r] internal population growth rate) and fitness (r2 and Root Mean Square Error [RMSE]). To choose the best combination of parameters, we refer to the values of r2 (higher is better), RMSE (lower is better) and optimal value [r] (closer to the optimal value is better). We proceed with parameter lookup using the flag code (flag code =1 to evaluate parameters, while flag code = 0, not to evaluate parameters, but to provide values) available in ASPIC.
The first clue is to set flags for (B1/K, MSY, K and q) to (1111) respectively and estimate all four parameters. When we fix, we use several plausible values (scenarios), then run ASPIC by scenario and choose the most optimal (plausible) run in all results (scenarios). In this case we also use several values (scenarios) for B1/K and K, i.e. for example, if we have 4 scenarios for each parameter, we will have 16 scenarios.
Then run ASPIC by scenario and select the most optimal (plausible) run in all results. If it still doesn't converge, we check CPUE again to see if there is any implausible behavior (such as sudden jumps/drops, unrealistic trends, etc.) and then delete such data points if any. If it still does not produce any plausible results, we believe that the catch and CPUE data do not fit ASPIC at all, and we conclude that no results are obtained.
Using the most reliable estimated parameters, we make Kobe plots (stock status trajectories . 8 ) and four types of time series plots, i.e. catch vs. MSY, F vs. To estimate the uncertainties (banana-shaped confidence surface) of the last year in the Kobe plot, we use the bootstrap (1000 times) available in the original ASPIC application.
Risk assessment
We present the results of stock and risk assessments of KAW (P), KAW (I), LOT (P) and LOT (I) including CPUE standardization, relationships between catch vs.
KAW (Pacific Ocean side)
CPUE standardization by log-normal GLM was performed using Thai PS nominal CPUE in the Andaman Sea. Box 5 shows the results of standardized CPUE including 6 outputs, i.e. (a) trends of standardized CPUE with 95% CI, (b) ANOVA table, (c) frequency distribution of residuals, (d) the QQ plot, (e) time series relationship (catch vs. STD_CPUE) and (f) negative correlation (catch vs. STD_CPUE). ASPIC was conducted using the Indian Ocean catch and the standardized CPUE of Thai PS in the Andaman Sea.
BOX 6 shows five different types of ASPIC results, including a table for “estimated parameters and related metrics” and four types of time series graphs for “catch vs. 3) Kobe plots (stock status) and risk assessments (optimal catch level) ( Box 7) .Using ASPIC results, the Kobe plot was created (Box 7) and risk assessments were performed for TB and F. BOX 7 shows the results suggesting that the stock status of KAW(I) in 2018 is in the red zone of The Kobe plot with a probability of 76%, therefore it is in a serious overfished and overfished situation.
Frame 8 shows the results of standardized CPU for both including 4 outputs, i.e. (a) trends of standardized CPU with 95% CI, (b) ANOVA table, (c) frequency distribution of residuals and (d) time series relationships (catch) versus STD_CPUE). Box 9 shows the trend of combined STD_CPUE, the time series relationship (catch against combined STD_CPUE) and the negative correlation (catch against combined STD_CPUE). ASPIC was performed using the catch in the Pacific side and the combined standardized CPUE.
BOX 10 shows five different types of ASPIC results, including one table for "estimated parameters and related statistics" and four types of time series graphs for "catch vs. 3) Kobe plot (stock status) and risk assessments (optimum catch level) (box) 11). Using ASPIC results, the Kobe plot was created (BOX 11) and risk assessments were performed for TB and F. We combined two nominal CPU by taking their mean based on the scaled CPU (mean=1) for each CPU.
Box 12 shows the combined nominal CPUE trend, time series ratio (catch vs. combined nominal CPUE), and negative correlation (catch vs. combined CPUE). We then decided to use the combined nominal CPUE for ASPIC because it reflected the negative well. ASPIC was conducted using the catch on the Indian Ocean side and the combined nominal CPUE (PS and GILL) from Thailand.
Using the ASPIC results, the Kobe plot was generated (BOX 14) and risk estimates for TB and F were also performed.
Optimum catch levels
We compared the stock status available from the previous stock assessments by SEAFDEC based on ASPIC, and by IOTC based on Stock Reduction Analysis (SRA) for the Indian Ocean. Stock status for 2013 and 2018 is available based on ASPIC conducted by the SEAFEDC neritic tuna project (SEADFEC, 2017 and this document) (Figure 8). Within the green zone, the status of stocks deteriorated slightly between 2013 and 2018, as catch levels increased slightly between 2013 and 2018 (Figure 9).
In SE Asian waters, the stock status deteriorates significantly between 2014 and 2018 (from green to red zone), while for the entire Indian Ocean both points are in the green zone, although biomass has decreased from 2013 to 2018. A possible reason is that catches in SE Asian waters increased during the year, while catches in the entire Indian Ocean were stable (Figure 11). The increase in catches in SE Asian waters is not so great, but the state of the stock has changed significantly.
Stock status for 2013 and 2018 for Southeast Asian aquifers in ASPIC is available from two SEAFEC neritic tones project practical workshops (SEAFDEC 2017 and this document respectively). The status of the stock has deteriorated significantly from 2013 to 2018 due to significant increases in catch levels (Figure 13), but it is still in the green. In Southeast Asian waters, the stock status recovered greatly from 2014 to 2018 (from the red zone to the green zone) because the catch decreased continuously from 2011 to 2018 (Figure 15).
There are two possible reasons for the large discrepancy in stock status in 2018 between IOTC and SEAFDEC, i.e. a) there is less fishing pressure (F) in Southeast Asian waters and (b) different stock assessment approaches are used (SRA without CPUE and ASPIC with CPUE).
FINAL REMARKS
2020) Software and manuals for standardization of CPUE (ver2), ASPIC batch job (ver2), Kobe I (Kobe plot) (ver5) and Kobe II (Risk assessment) (ver1) http://www.seafdec.or.th /neritic-tunas/. 2004) ASPIC User Manual: A Stock-Production Model Incorporating Covariates (Neritic Tuna Home Page Version http://www.seafdec.or.th/neritic-tunas/. Syahida Kasim N, Mat Jaafar TNA, Mat Piah R, Mohd Arshaad W, Mohd Nor SA, Habib A, Abd. Ghaffar M, Sung YY, Danish-Daniel M and Tan MP. 2020) Recent population expansion of longtail tuna Thunnus tonggol (Bleeker, 1851) by mitochondrial DNA markers.
The progress of the SEAFDEC neritic tuna project is indicated here by comparing the stock status of four neritic tuna species, including those in the IOTC. Four species are kawakawa, longtail tuna, Spanish mackerel and Indo-Pacific king mackerel. In the Indian Ocean, the stock status between Sea Asian waters (SEAFDEC) and the entire Indian Ocean (IOTC) is different.
This is probably because the fishing pressure between two areas is different and/or the stock assessment methods (with and without CPUE) are different. Note) Colors are the same as in the Kobe chart and % is the probability in four quadrants of the Kobe chart. These two stock statuses are in the green zone, but the chances are less than 50%, making the stock status (safe) less certain.
According to Table 6, stock status in the Pacific side of the SE Asian waters are all in green zones and very healthy conditions except narrowband Spanish mackerel (red). This implies that the F in the Pacific side is probably less than in the Indian Ocean side. Summary and comparisons of four neritic species (Pacific side). Note) Colors are the same as in the Kobe plot and % is the probability in four quadrants of the Kobe plot.
