Assessment of the Impact on 20 Pelagic Fish Species by the Taiwanese Small-Scale Longline Fishery in the Western North Pacific Using Ecological Risk Assessment

Simple Summary In this study, a semi-quantitative ecological risk assessment (ERA) was used to evaluate the ecological risks of fishing impact on 20 pelagic fish species by the small-scale longline fisheries in the western North Pacific Ocean. More than 2.38 million individual landing records at Nangfangao and Hsinkang fishing ports, eastern Taiwan from 2001–2021 were used in this study. The productivity was estimated based on the mean ranking (high, median, and low) of seven life history parameters and the susceptibility was calculated by the multiplication of the catchability, selectivity and post-capture mortality. The ERA results indicated sharks have higher ecological risk than those of tunas and billfishes, except yellowfin tuna (Thunnus albacares). The shortfin mako shark (Isurus oxyrinchus) and dusky shark (Carcharhinus obscurus) have the highest risk. Yellowfin tuna, other shark species, and sailfish (Istiophorus platypterus) have medium risk. While the striped marlin (Kajikia audax), and albacore tuna (T. alalunga) have the lowest risk. Although ERA cannot replace the conventional stock assessment methods that can produce solid management information on catch and effort, yet it can provide useful information for precautionary management measures. Abstract Ecological risk assessment (ERA) has been applied on assessing the relative risk of bycatch species in recent years. ERA index is calculated by productivity of species and susceptibility of fisheries on fish species. In this study, a semi-quantitative method was used to evaluate the risks of exploitation for 20 pelagic fish species by the small-scale longline fisheries in the western North Pacific Ocean. The productivity was estimated based on the ranking (high, median, and low) of seven life history parameters. The susceptibility was calculated by the multiplication of the catchability, selectivity and post-capture mortality. The ERA results indicated the risks of sharks are higher than those of tunas and billfishes, except yellowfin tuna (Thunnus albacares). The shortfin mako shark (Isurus oxyrinchus) and dusky shark (Carcharhinus obscurus) have the highest risk. Other shark species, yellowfin tuna, and sailfish (Istiophorus platypterus) have medium risk. While the striped marlin (Kajikia audax), and albacore tuna (T. alalunga) have the lowest risk. Stock assessment and rigorous management measures such as catch quota and size limit are recommended for the species in high or medium ecological risk and a consistent monitoring management scheme is suggested for those in low ecological risk.


Introduction
The Taiwanese small-scale tuna longline fishing vessels (<100 gross tonnages, GRT) operating in the coastal and offshore waters of Taiwan, western North Pacific are mainly based on Nanfngao fishing port in the northeastern Taiwan and Hsinkang fishing port in the southeastern Taiwan. According to the sales records, annual landings of Nanfangao

Source of Data
The sales data of two major fishing ports of Taiwanese small-scale longline fishery-the Nanfanao and Hsinkang fishing ports-were used in the analysis as the landings of these two fishing ports comprised of 80% of the catch of Taiwanese longline fishery in the study area of the western North Pacific. Each of the individual tuna, billfishes, and sharks landed at the Nanfangao and Hsinkang fishing ports, eastern Taiwan, was weighed before auction. Therefore, we were able to collect the species-specific whole weight data of each individual from the sales records. These data, including 20 pelagic fish species from 2001-2021, were used to estimate the species-specific catch in number, the catch composition, and the mean and median weight. The length was estimated from the individual weight based on existing length-weight relationships. The age of each individual was then estimated by substituting the length into the growth equation of each species from the literature [Supplemental Table S1]. As most blue sharks that landed at the Nanfangao fishing port were not sold in a regular fish market channel, only weight without catch in number data were available. A mean whole weight of 32 kg reported by the on-board scientific observers of Taiwanese longline vessels in the North Pacific was used to estimate the catch in number for blue shark. Various length measurements have been applied to different fish species. For example, total length (TL) was commonly used in pelagic sharks, fork length (FL) in tunas, and eye-fork length (EFL) or lower jaw fork length (LJFL) in billfishes. For consistency, EFL was used for all billfish species in this study.
In addition to the 20 species examined in this study, the dolphinfish, Coryphaena hippurus, is the most abundant pelagic fish species caught by the Taiwanese small-scale longline fishery in the western North Pacific. Its catch in number was much more than other pelagic fish species and the information was not available, and the dolphinfish was usually caught by shallow set longline fishery, which was different from those targeting tuna, billfishes, and sharks. Therefore, the dolphinfish was excluded in this study.

ERA
The ERA method considers the productivity of fish and their susceptibility to fisheries. The estimation of the productivity score of 20 pelagic fish species followed Hobday et al. [32] based on seven life history parameters. The rank score (R = 1-3) was assigned to each of the life history parameters and the mean value of these scores represented the productivity index (P) for each species. The higher productivity score meant higher productivity (less risk) of a fish species. The seven life history parameters and their productivity scores are categorized based on the criteria in Table 2.
Susceptibility (S) is the impact of a fish species from fisheries and is estimated by the multiplication of the probabilities of the following three parameters: (1) catchability, the species composition (percentage of catch in weight) of 20 species based on the landing data from 2001-2021 at Nanfangao and Hsinkang fishing ports, eastern Taiwan; (2) selectivity, the ratio of age range of catch and longevity, can be expressed as: (A max_c -A min_c )/A max , where A max_c is the maximum age at catch, A min_c is the minimum age at catch, A max is the longevity [33]; (3) post-capture mortality, including the mortality of retention and after discard or live release. The post-mortality was referred to Cortés et al. [30] that estimated from the observed data of US far sea fishery. Susceptibility was estimated from the multiplication of the aforementioned three parameters. When the productivity and susceptibility were estimated, as these two were in different scale, a modified standardized Euclidean distance (D) from [29] was calculated: where P is productivity ranging from 1 to 2.5, and 1.5 is the range, S is susceptibility ranging from 0 to 0.2, and 0.2 is the range. Larger D value indicates higher risk.

Sales Data Analysis
In total, more than 2.38 million individual landing records at Nangfangao and Hsinkang fishing ports from 2001-2021 were used in this study. The sales records of Nanfangao (northeastern Taiwan) and Hsinkang (southeastern Taiwan) fishing ports showed that shark landings ranged from 4738 tons in 2007 to 1404 tons in 2012 and maintained stability at around 1702 tons after then ( Figure 2). Of this, the blue shark (27%), shortfin mako (20%), and bigeye thresher (16%) were the top three species in shark landings. The blue marlin (44%), swordfish (25%), and white marlin (14%) were the major billfish species landed at Nanfanao fishing port, while the white marlin (36%), sailfish (34%), and blue marlin (19%) dominated the billfish landings at Hsinkang fishing port. As for tuna landings, the yellowfin tuna was the top one for both fishing ports, followed by the Pacific bluefin tuna, the bigeye tuna in Nanfanao, and the bigeye and albacore in Hsinkang fishing port.

Length
Asymptotic length (L∞): The largest L∞ of sharks was 422 cm TL for the bigeye thresher, and the smallest L∞ was 210 cm TL for sandbar shark; the maximum L∞ of tunas was 366.7 cm FL for the Pacific bluefin tuna, and the smallest L∞ was 103.5 cm FL for albacore tuna; the maximum L∞ of billfishes was 421.8 cm EFL for the blue marlin, and the smallest L∞ was 277.4 cm EFL for striped marlin (Supplemental Table S1).
Maximum observed length (Lmax): The largest Lmax of sharks was 422 cm TL for the The sales records of Nanfangao fishing port indicated sharks had the highest catch in number followed by tunas and billfishes in 2001-2010; tunas had the highest catch in number followed by sharks and billfishes since 2011. The catch in number at Hsinkang fishing port was dominated by billfishes, followed by tunas and sharks. Shark catch in number of the two fishing ports ranged from 128,397 in 2006 to 27,008 in 2012. Of which, the blue shark (41%), shortfin mako (16%), and scalloped hammerhead (11%) were the top three species in shark landings. The catch in number of tunas ranged from 19,454 in 2005 to 47,091 in 2016. Of which, the yellowfin tuna comprised 76%, followed by bigeye tuna (13%) and albacore (8%). The billfish catch in number ranged from 19,836 in 2021 to 69,240 in 2010. Of which, the sailfish (45%), blue marlin (20%), and swordfish (16%) were the major billfish species landed.

Length
Asymptotic length (L ∞ ): The largest L ∞ of sharks was 422 cm TL for the bigeye thresher, and the smallest L ∞ was 210 cm TL for sandbar shark; the maximum L ∞ of tunas was 366.7 cm FL for the Pacific bluefin tuna, and the smallest L ∞ was 103.5 cm FL for albacore tuna; the maximum L ∞ of billfishes was 421.8 cm EFL for the blue marlin, and the smallest L ∞ was 277.4 cm EFL for striped marlin (Supplemental Table S1).
Maximum observed length (L max ): The largest L max of sharks was 422 cm TL for the bigeye thresher, the smallest L max was 210 cm TL for sandbar shark; the largest L max of tunas was 252 cm TL for the Pacific bluefin tuna, and the smallest L max was 101 cm FL for albacore tuna; the largest L max of billfishes was 324.5 cm EFL for the blue marlin, and the smallest L max was 191 cm EFL for striped marlin (Supplemental Table S1).
Size at birth (L b ): The largest L b of sharks was 174 cm TL for the bigeye thresher, and the smallest L b was 45 cm TL for blue shark. No L b information for tunas and billfishes was available as they were oviparous species (Supplemental Table S2).
Mean size at maturity (L m ): The largest L m of sharks was 336.6 cm TL for the bigeye thresher and the smallest L m was 172.5 cm TL for sandbar shark. The largest L m of tunas was 190 cm FL for the Pacific bluefin tuna and the smallest L m was 83 cm FL for albacore. The largest L m of billfish was 194.5 cm EFL for the white marlin and the smallest L m was 143.54 cm EFL for sailfish (Supplemental Tables S2 and S3).

1.
Maximum age (A max ): The largest A max of sharks was 50 years for the dusky shark, the smallest A max was 11.6 years for the scalloped hammerhead shark; the largest A max of tunas was 26 years for the Pacific bluefin tuna, and the smallest A max was 7.7 years for yellowfin tuna; the largest A max of billfishes was 14 years for the blue marlin, and the smallest A max was 6 years for striped marlin (Supplemental Table S1).

2.
Age at maturity (A m ): The largest A m of sharks was 16.4 years for the dusky shark, the smallest A m was 4.7 years for the scalloped hammerhead shark; the largest A m of tunas was 8 years for the Pacific bluefin tuna, and the smallest A m was 2.4 years for yellowfin tuna; the largest A m of billfishes was 7.4 years for the blue marlin, and the smallest A max was 4.8 years for striped marlin (Supplemental Tables S2 and S3).

1.
Fecundity/litter size: The largest litter size for sharks was the scalloped hammerhead of 30, and smallest was two for bigeye and pelagic thresher. The largest batch fecundity of tunas was the Pacific bluefin tuna of 5.8-25.2 million eggs, and the smallest was albacore of 0.94 million eggs; the largest of billfishes was the blue marlin of 6.94 million and the smallest was sailfish of 1.30 million (Supplemental Table S3).

ERA
The shortfin mako and dusky shark had the lowest productivity index value (p = 1.00) which corresponded to the lowest productivity, followed by the bigeye, pelagic thresher and smooth hammerhead (p = 1.14), and the striped marlin and yellowfin tuna had the highest productivity index value of 2.43 (highest productivity) ( Table 3). The yellowfin tuna had the highest susceptibility value (S = 0.1864), followed by the blue shark (S = 0.1344), sailfish (S = 0.1053), shortfin mako shark (S = 0.0613), and blue marlin (S = 0.0563). The oceanic whitetip shark had the lowest susceptibility value of 0.0009 due to its smallest catchability (0.0013) ( Table 4). Table 3. Productivity parameters by rank in the ecological risk assessment of the 20 pelagic species in the western North Pacific. The productivity indicator (P) reflects the risk of productivity.  The shortfin mako, dusky shark, and yellowfin tuna were the top three species with the highest ecological risk with D = 1.0459, 1.0028, and 0.9330, respectively. While the striped marlin, albacore, and bigeye tuna had the lowest ecological risk with D = 0.0564, 0.1612, and 0.2228, respectively (Table 5). Three groups can be categorized for the 20 species in this study as (1) the high ecological risk group of SMA and DUS, (2) the median ecological risk group including YFT, other shark species, and sailfish, and (3) the low ecological risk group of other tuna species and billfishes (Figure 3).

Discussion
This study assessed the impact on 20 pelagic fish species by the Taiwanese scale longline fishery in the western North Pacific using ERA. The results can p useful information for implementing ecosystem-based management and prioritizi management and conservation measures of pelagic species including tuna, billfishe sharks in this region.

Catch Data
In addition to Taiwanese longline fishing vessels, some of Japanese longline f vessels also operated in the western North Pacific Ocean. Future work, if possible, s refer to the study of Murua et al. [35], Cortés et al. [36], Murua et al. [28], and Corté [30] to include Japanese catch and observer data to improve the ERA of pelagic fis the western North Pacific. For example, only adult BFT was caught by the Taiw longline fishing vessels and examined in the present study, but both young and adu were caught by Japanese fishing vessels. Better estimate of susceptibility of BFT obtained by examining the pooled data in the future.

Discussion
This study assessed the impact on 20 pelagic fish species by the Taiwanese smallscale longline fishery in the western North Pacific using ERA. The results can provide useful information for implementing ecosystem-based management and prioritizing the management and conservation measures of pelagic species including tuna, billfishes, and sharks in this region.

Catch Data
In addition to Taiwanese longline fishing vessels, some of Japanese longline fishing vessels also operated in the western North Pacific Ocean. Future work, if possible, should refer to the study of Murua et al. [35], Cortés et al. [36], Murua et al. [28], and Cortés et al. [30] to include Japanese catch and observer data to improve the ERA of pelagic fishes in the western North Pacific. For example, only adult BFT was caught by the Taiwanese longline fishing vessels and examined in the present study, but both young and adult BFT were caught by Japanese fishing vessels. Better estimate of susceptibility of BFT can be obtained by examining the pooled data in the future.

Life History Parameters
Age estimation is the fundamental information of fish biology study and is one of the important life history parameters in stock assessment and population dynamics. The age of pelagic sharks used in this study was estimated based on the band pair counts of vertebral centra from the literature. As for tunas and billfishes used in this study, lengthfrequency analysis was used in growth parameter estimations for the yellowfin [37] and bigeye tuna [12]; otolith was used for the albacore [14], while the spine was used in the age estimation for the swordfish [38], blue marlin [39], sailfish [15], white marlin [40], and striped marlin [41]. The scale was used in the age estimate for the Pacific bluefin tuna [42]. The growth parameters derived from different hard parts of fish may be different which may lead to the bias of age-structure [43]. As the productivity was estimated based on the growth information of each species in this study, the bias and uncertainty of age estimation may affect the subsequent ERA results.
The A mat and A max and ERA of the scalloped hammerhead were based on a biannual deposition estimation [44] in this study. Although similar assumption was made by Anislado-Tolentino and Robinson-Mendoza [45] in Mexican waters, annual deposition was assumed by other authors in various regions of the Pacific Ocean [46][47][48]. Piercy et al. [49] reported A max of 30.5 years, which was much larger than 11.6 years used in the present study. If annual deposition was used in the ERA as a different scenario, the risk ranking of the scalloped hammerhead was promoted to be 4 from 9 (Table 5). If the body weight variation trend and the IUCN Red List category were taken into account, this species was at the highest risk among 11 pelagic shark species [1].
The uncertainty existed in life history parameters used in this study, which may also be related to the small sample size. For example, the sample size was 188 for the age, growth, and reproductive biology study of the oceanic whitetip [50]. However, only two pregnant females were collected in that study. Thus, the small sample size may lead to the bias of little size and size at maturity estimations. Although the life history parameters used in this study were from the best available information, some of the references were published many years ago and may not represent current biological condition of fish. Fortunately, as the productivity was estimated by the mean rank of seven life history parameters, the productivity indicator did not change even if the uncertainty of life history parameters were taken into account. Future work should focus on collecting more specimens and using consistent ageing characteristics to get more accurate and updated estimates of life history parameters and improve the ERA results.

ERA
Two methods were commonly used to estimate the productivity of fish. The first method was based on the rank of life history parameters of each species, and the mean of rank was used as the productivity index (p) [32,51,52]. The p value can be estimated by another method using an empirical equation based on the reproductive strategy, size at maturity, and the maximum length [24,26,35]. The latter had higher uncertainty because p was estimated based on few life history parameters. Thus, the first method was used in estimation of p of each species in this study. The shortfin mako and dusky shark had the highest value of, p which was corresponding to the dusky shark in Australian waters [33].
The higher value of p indicated higher reproductive potential, which may experience lower risk. Overall, pelagic sharks are slow growing and late maturing, with extended longevity and low reproductive potential species. The shortfin mako and dusky shark mature at older ages than other pelagic shark species (Supplemental Table S2), and thus experienced higher risk than other shark species.
The intrinsic rate of population growth (r) has been used as productivity index for pelagic sharks [1,29,30,36]. Although the intrinsic rate of population growth was a better productivity index for pelagic sharks, this index might not be suitable for tuna and billfishes because demographic parameters of these oviparous species were difficult to estimate. Due to the remarkable difference in life history such as the fecundity (litter size) and longevity between sharks and other species, a rank-based index was believed to be a better choice as it can be applied to all species. Susceptibility was commonly estimated by the multiplication of the availability, encounterability, selectivity, and post capture mortality in previous studies [29,30,32,36,51,52]. Due to the difficulty of obtaining the information of availability and encounterability, these two indices were replaced by the species-specific catchability estimated based on its mean percentage of catch in catch from 2001 to 2021 at Nanfangao and Hsinkang fishing ports. The long-term and large-size historical landing data (n > 2.38 million) can better describe the species-specific vulnerability to longline fishery. The low catchability may represent low abundance of fish species which is likely resulted from overfishing. Unfortunately, the susceptibility used in this study may not reflect the vulnerability of fishing for certain species after they were ban retention such as OCS and FAL as no discard data of these two species were available for susceptibility estimation. In addition, different sizes in fishing vessels and target species may affect the catchability. Future study should focus on collecting the information of number of hooks, main line length, branch line length, and species-specific capture depth from observer's records and tagging results to estimate availability and encounterability to validate the catchability used in this study.
The selectivity estimated in this study was based on the ratio of age range of catch and the longevity. The age range was estimated by converting individual shark landing data (body weight) to length and then to age, which were used to estimate the minimum and maximum ages of each species. Uncertainty may occur in the two-step converting process. Those weights greater than the maximum weight in the literature were excluded in our analyses, yet these individuals may be pregnant females but could not be confirmed due to lack of sex information of sales records. Selectivity may be affected by the hunger condition of fish, bait type, hook type and size, and material of branch line (steel wire or monofilament) [53]. Since the selectivity was estimated by pooling the long-term data together, those effects aforementioned could be ignored. We believed that our estimates based on the best available data with large sample size were representative. As individual body weight data were not available for most blue sharks, the selectivity of blue shark being set as 1 was assumed due to its wide range of size at catch. We believe this assumption was reasonable.
ERA has been mainly applied to the bycatch species in the past. In this study, due to the remarkable difference in life history traits among sharks and other teleost species, the ranking method was used in assessing the vulnerability of 20 pelagic fish species. The ERA indicated that sharks have higher risk than tunas and billfishes, except the yellowfin tuna, which was comparable with the findings in the waters off eastern Taiwan [31], but some differences were found between the two studies. The SMA was identified as the highest ecological risk shark species in the present study but Lin et al. [31] suggested that FAL had the higher risk than BSH and SMA. As for other species, the present study concluded that YFT had the highest risk, followed by BLM and BUM, but Lin et al. [31] reported that BFT had the highest risk, followed by BET and BLM. The discrepancy may be due to different definitions and scores of productivity and susceptibility in these two studies. Previous ERA was based on reproductive strategy, the maximum size, size at maturity as productivity, and size at catch, and post capture mortality of animals as susceptibility from observer's records [24,35,45]. However, high uncertainty existed in these studies due to the limited data available. In the present study, we used catchability, selectivity, and post capture mortality to estimate the susceptibility. If the availability and encounterability can be derived from observer's data in the future, more robust results of ERA can be obtained.

Comparison with Single-Species Approach
Liu and Chen [3] used an age-structured demographic analysis in scalloped hammerhead stock assessment and concluded that the species could not withstand the long-term exploitation on juveniles. Liu et al. [4] and Tsai et al. [5] reported the stock status of pelagic thresher shark was overexploited using yield per recruit model (YPR) and stage-based demographic analysis. Several studies with various approaches indicated that the shortfin mako shark, bigeye thresher shark, and smooth hammerhead in the western North Pacific were overexploited [6][7][8][9][10][11][12]. As for tunas and billfishes, Liu [13] suggested that the fishing mortality (F) of the bigeye tuna in Taiwan waters was greater than the biological reference point (F 0.1 and F SSB30 ) and was in full exploitation. Chung [14] evaluated the North Pacific bluefin tuna using a non-equilibrium production model and suggested a total allowable catch of 18,000 tons. Chen [15], Chiang [16] and Wang [17] conducted the stock assessment of the albacore, sailfish and swordfish in the North Pacific and concluded that the F of these species were at optimum levels.
Although the two approaches were based on different theories, the existing results of single species stock assessment of several species mentioned above were comparable with those derived from this study using the ERA method, suggesting that sharks have higher ecological risk than tuna and billfishes.

Conclusions
In the present study, the semi-quantitative ERA has prioritized the risk of 20 pelagic fish species in the western North Pacific. Sharks have higher ecological risk than tunas (except yellowfin tuna) and billfishes. The shortfin mako and dusky shark have the highest ecological risk; yellowfin tuna, other shark species, and sailfish have medium risk, while the striped marlin, and albacore tuna had the lowest risk. Although ERA cannot replace the conventional stock assessment methods that can produce solid management information on catch and effort, it can provide useful information for precautionary management measures. Rigorous management measures such as catch quota and size limit are recommended for the species at high or medium ecological risk and a consistent monitoring management measure is suggested for other species at low ecological risk. To improve the results, the ERA should be updated regularly based on best available information of the productivity and susceptibility of these species.
Supplementary Materials: The following supporting information can be downloaded at https:// www.mdpi.com/article/10.3390/ani12162124/s1, Table S1: Age and growth parameters of the 20 pelagic species in the Northwest Pacific Ocean. Table S2: Reproductive parameters of 11 pelagic shark species in the Northwest Pacific Ocean. Table S3: Reproductive parameters of 5 billfish and 4 tuna species in the Northwest Pacific Ocean. References

Institutional Review Board Statement:
This study was based on published data and no animal experiment was conducted. So, no animal ethics should be concerned.

Data Availability Statement:
The data used in this study can be found in Supplemental Tables S1-S3.

Conflicts of Interest:
The authors declare no conflict of interest.