Effect of escape vents on retention and size selectivity of crab pots for swimming crab Portunus trituberculatus in the East China Sea

Jin Zhng, Zhnwn Pi, Pingguo H, Jingo Shi

aCollege of Marine Sciences, Shanghai Ocean University, Shanghai, 201306, China

bNational Engineering Research Centre for Oceanic Fisheries, Shanghai Ocean University, Shanghai, 201306, China

cKey Laboratory of Sustainable Exploitation of Oceanic Fisheries Resources, Ministry of Education, Shanghai Ocean University, Shanghai, 201306, China

dSchool for Marine Science and Technology, University of Massachusetts Dartmouth, New Bedford, MA, 02747-2300, USA

eEast China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, Shanghai, 200090, China

Keywords:

ABSTRACT An increase in the crab pot fishery in the East China Sea has caused great pressure on swimming crab Portunus trituberculatus resources. Thus, it is essential to implement suitable measures to release sublegal-sized crabs to increase the number of recruits for legal-sized crabs. One of the measures considered is the installation of escape vents on crab pots. We tested crab pots with one and two escape vents and compared the catchability and size selectivity of these pots with control pots without an escape vent. Pots with one escape vent located on single side of the pot, top or bottom (SS), and another on both sides at top and bottom edges (BS) were tested in sea trial experiments following typical commercial fishing practice. The results show that both SS pots and BS pots caught significantly smaller number of sublegal-sized crabs (<116 mm carapace width, CW), but the difference in catch between the two types of crab pots is not significant. The 50% selective CWs (CW50) and the selective ranges(SRs) of both BS pots are slightly larger than those of SS pots. However, 95% confidence intervals of CW50s and SRs are overlapped, indicating that the discrepancy between them is small. Then, in a mixed effect model, the effect of locations of escape vents, taken as the fixed effect, on selectivity parameters and indexes was analyzed by hypothesis testing. The results show that null hypothesis of no effect of location of vents on size selectivity cannot be rejected, indicating that there is no significant difference in size selectivity between the two types of pots for P. trituberculatus.

1.Introduction

Although fishery for swimming crabPortunus trituberculatusin the East China Sea has a long history, the use of pots for crabP. trituberculatusstarted in the early 1980s. With continuous decline in commercial fish stocks in the early 1990s, shrimps and crabs thrived in the East China Sea (Song, Yu, Xue, & Yao, 2006). The crab fishery mainly targetingP. trituberculatushas become the most important fishery of Zhejiang and Fujian Province, and the pots have become the main fishing gear for crabs owing to simple structure, low cost, and easy operation (Song et al., 2006; Yu, Song, Yao, & Shen, 2003). Crabs caught by pots also have better quality as most of them are live when landing on deck. The growth of fishery, now accounting for about 400,000 tons of annual landings in the East China Sea and the Yellow Sea (Fisheries Administration Bureau, MARA, PRC, 2019), is apparently associated with an increase in fishing effort. This has caused tremendous pressure on crab resources, large fluctuations in landing and decline in catch per unit effort (CPUE) (Butler & Heinrich, 2007; He, Zhang, & Li, 2011;Panhwar et al., 2018; Zhang, Wang, & Guan, 2015). In Zhejiang Province fishery regulations, there is a minimum landing size (MLS) of 60 mm Carapace Length (CL) or 125 g weight since 2015 (Quality &Technical Supervision Bureau, 2015), recently updated to 49 mm CL and 70 g weight (Li, 2017) forP. trituberculatus. The MLS of carapace width(CW) should be about 116 mm according to the relationship between CW and CL forP. trituberculatus(Zhang et al., 2015).

Installation of escape vents or gaps on crustacean pots or traps,which would be more precise in their selective release of sublegal-sized crabs and retention of legal-sized crabs than enlargement of mesh(Nishiuchi, 2003; Winger & Walsh, 2007), has been proved to be an effective and convenient technical measure to release sublegal-sized crabs, conserve fisheries resources, and optimize catching efficiency in different pot or trap fisheries (Boutson, Mahasawasde, Mahasawasde,Tunkijjanukij, & Arimoto, 2009; Brown, 1982; Rotherham, Johnson,Macbeth, & Gray, 2013; Tallack, 2007). Wu (1996) and Zhang, Xu,Huang, and Zhou (2010) evaluated the effect of escape vents on releasing sublegal-sized crabs through tank observations and sea trials in the East China Sea. Based on the results of these studies, it was recommended that the use of one to three escape vents, not less than 33 mm height and 250 mm width in size and installed at the bottom edge of side panels of the pot, should be implemented as a fishery management measure in Zhejiang Province (Marine and Fisheries Bureau of Zhejiang Province, 2013). However, previous selectivity studies tested only one escape vent on single crab pot (Bai, Zhang, He, & Zhang, 2015; Wu,1996; Zhang, Lin, et al., 2010; Zhang, Xu, et al., 2010). No study has been conducted on the location of escape vents if two or three vents are installed in a pot; therefore, no specification on the location of vents is provided in the management measures. In commercial crab pot fisheries,the pots are set in a longline, and owing to the symmetrical shape of pot,the landing surface of pot cannot be controlled when they sink to the seabed. This means the escape vents, if mounted on one side only, would be located at either bottom or top side of the pot. Therefore, it is not sure whether the original intention of installing escape vents can be achieved. Unfortunately, the effect of location of vents on size selectivity forP. trituberculatusin crab pot fisheries in the East China Sea has not been evaluated.

In this study, comparative fishing experiments were carried out by using crab pots with different mounting configurations of escape vents.The main purpose of this study was to confirm that pots with escape vents would reduce the retention of sublegal-sized crabs, to evaluate the effect of location of escape vents on size selectivity, and to verify the applicability and effectiveness of these management measures for crab pot fisheries in the region.

2.Materials and methods

2.1.Fishing gear

A commercial crab pot was used as the control pot, made from a cylindrical metal frame with 60 cm diameter and 25 cm height, and covered by a 32 mm mesh size netting panel. Three entrance funnels were installed at equal intervals around the side of the pot (Fig. 1a).

The experimental pots had either one (Fig. 1b) or two escape vents(Fig. 1c). For pots with one escape vent, the escape vent was installed on one side of the pot (termed “SS”). For pots with two escape vents, the escape vents were installed on both sides of the pot at the opposite corner, one on the top and another on the bottom (termed “BS”).

Escape vents were constructed from rectangular polypropylene (PP)plates, 30 cm wide ×10 cm high ×0.3 cm thick (Fig. 1d). A rectangular gap (20 cm wide ×2.8 cm high) was cut from the plate, with a distance of 2.5 cm from the side to be tied to the frame of the pot.

2.2.Sea trials

Sea trials were conducted in the traditional fishing ground for crabs located on the East China Sea (31°54′N to 32°02′N and 123°12′E to 123°18′E) at 35–40 m depth and mud/sand seabed during the period of 24 Dec to 29 Dec 2012. The fishing experiments were carried out on board the commercial crab pot fishing vessel “Zheshengyu 090101”,with 46.0 m total length and 290 kW main engine power. The vessel has a capacity of operating over 5000 crab pots and normally conducts fishing in the studied area.

In a normal fishing operation, the pots were baited with 100–150 g fresh fish (such as chopped little yellow croakerLarimichthys polyactis) in a plastic mesh bait bag placed at the center of the pot and deployed using a longline with 7–8 m between the pots and about 0.6 m long branch lines.

One hundred test pots, including 50 SS pots and 50 BS pots were used for fishing with 50 control pots (termed “C”) following typical commercial fishing practice. These 150 pots were attached to a main line in a random order. A total of seven experimental hauls were used for fishing,with soak time ranging from 7 to 12 h.

2.3.Data collection

Upon retrieval, the total catch of each pot was sorted by species. All the swimming crabs caught were subsampled. The CW of swimming crab was measured to the nearest mm, and the number of crabs for each 1 mm CW group was recorded.

2.4.Catch comparison analysis

The retention of crabs in different types of experimental pots and control pots was modeled by using the generalized linear mixed model(GLMM). The proportion of catch retained at CW by using SS or BS pots[Φ(CW,SS),Φ(CW,BS)]can be expressed for each CW and for each haul as follows:

where NCW,SSand NCW,Care number of crabs at CW measured for the SS pot and control pot, respectively. A value of Φ =0.5 indicates no difference in the catch between the two types of pots at CW. The same was conducted for the BS pot. The catch proportions Φ(CW)for the crabs caught from the two types of experimental pots were analyzed using the GLMM with CW as the explanatory variable, Φ as the response variable,and individual haul as the random effect, following the technique described in Holst and Revill (2009). The GLMM analysis was conducted using the glmm PQL package in R, using a penalized quasi-likelihood method. The polynomial regression GLMM was used to fit curves for the expected proportion of the catch retained by the experiment pots,after logit transformation, as follows:

whereβ0,β1,β2,andβ3are coefficients, andδis the error term. The analyses were preceded by fitting the highest-order polynomials followed by subsequent reductions until all the terms showed a significance(p<0.05) based on Wald’s test, with removal of one term at each step to determine the best-fit model (Holst & Revill, 2009).

2.5.Selectivity analysis

The selectivity analyses were conducted by comparing the catch CW frequencies of crabs retained in the SS, BS and control pots. A logistic selection curve (Millar & Walsh, 1992; Wileman, Ferro, Fonteyne, &Millar, 1996) was fitted to the datasets in the following equation:

whereS(CW) is the probability of retention for each interval of CW,vi=(ai,bi)Tare the selectivity parameters for hauli. The 50% retention CW,CW50, and selection range,SR, of 25%–75% retention CW are given by

The SELECT model (Millar, 1992; Millar & Fryer, 1999) with equal fishing power was used to fit the data for each haul, and selectivity parametersviwere estimated using the maximum likelihood method.Mean selectivity curves of test pots were derived in the mixed effect model (Fryer, 1991; Millar & Fryer, 1999). The between-haul variations were considered as random effect, while the different locations of escape vents and soak time of each haul were considered as fixed effect (full model), i.e., conditional onvi,

Fig. 1.Configuration and mounting of escape vents of crab pots. a, control pot without an escape vent; b, SS (single-side) pot with one escape vent; c, BS (both-sides)pot with two escape vents located on the opposite position of the pot (top and bottom sides); d, detail design and dimension of escape vent.

whereXiis a known 2 ×6 matrix of explanatory variables (i.e. location of vents and soak time) for hauli,αis a 6 ×1 vector of unknown parameters to be estimated, matricesRimeasure the “known” binomial variation in estimatingviand matrixDmeasures the between-haul variation in selectivity parameters. More details on the model interpretation and parameter estimation could be found in the literature(Fryer, 1991; Millar & Fryer, 1999; Wileman et al., 1996).

In addition, whether the locations or soak time affected the size selectivity of escape vents was tested by hypothesis testing. The difference between model deviances was considered as approximate chisquare distribution, and the degree of freedom was given by the difference in number of parameters between nested models. Thus, two hypotheses were tested as follows:

H01.no effect of soak time on size selectivity, i.e., the fixed effect of soak time could be ignored, and then

H02.no effect of location of vents on size selectivity, i.e., the difference inCW50andSRfor swimming crabs between SS and BS pots is statistically insignificant.

3.Results

3.1.Distribution of catch

The total number of swimming crabP. trituberculatuscaught in the pots is 2413 with a mean catch rate of 3.7, 1.9, and 1.6 individuals per pot for the control, SS, and BS pots, respectively. Swimming crabs dominated in the catch and accounted for 93% of the total catch in numbers. The CW of crabs retained in control pots ranged from 57 mm to 195 mm with a mode of 10 cm (Table 1). Size frequency distributions of different pot groups in each trip are shown in Fig. 2. The results of the Kolmogorov-Smirnov sample tests show that the CW frequencies of control pots were significantly different from those of either SS (p=0.022) or BS pots (p<0.01), indicating that escape vents significantly increased the escape of sublegal crabs (Table 1). However, no significant difference was observed between the CW frequencies from SS and BS pots (p=0.771).

3.2.Size-related retention by different pots

The analysis using the GLMM shows that both the SS and BS pots reduced the retention of small crabs compared to the control pots(Fig. 3). The data were best modeled by using a third-order polynomial(cubic model) formula (Table 2). For SS pot, the reduction of crabs is less than 118 mm CW, while for BS pot, the reductions were those 120 mm CW and below. However, no statistical difference was observed between SS and BS pots.

3.3.Size selectivity analysis

The logistic selectivity curves of SS and BS pots for each set was fitted to CW frequency distribution, and a satisfactory fit was observed in SELECT model with equal fishing power. The estimated selectivity parameters and selectivity curves of SS and BS pots in each haul are shown in Table 1 and Fig. 4 (dashed lines).CW50s andSRs for SS and BS pots ineach haul are plotted in Fig. 5 (dots) and their 95% confidence intervals(CIs) are shown using solid-line circles. Considering between-haul variations, the mean selectivity curves for SS and BS pots were derived in the mixed effect model and shown in Fig. 4 as solid lines. The meanCW50s andSRs of SS and BS pots are shown in Table 3. Figs. 4 and 5 show that the meanCW50of BS pots is slightly larger than that of SS pots,while theSRs remained unchanged regardless of location of vents. This indicates that to some extent, BS pots would release more small crabs than SS pots.

Table 1 Summary of biological data of P. trituberculatus caught during replicate deployments of SS, BS, and control pots and estimates of selectivity parameters for each deployment. SS - single side, pot with one vent, BS - both sided, pot with two vents, C - control pot without an escape vent, SE - standard error, SD - standard deviation.

Fig. 2.Catch frequency with CW of swimming crab P.trituberculatus caught by different experimental (SS and BS) and control pots in seven hauls (a–g) and the combined carapace frequency of all hauls (h).

Fig. 3.Retention of swimming crab P.trituberculatusby SS (a) and BS (b) pots compared with that of control pots from GLMM. The top panels show catch proportion[SS/(SS +C), or BS/(BS +C)], the bottom panels show the distribution of residuals.

Table 2 Output of GLMM fit by maximum likelihood. SS - single side, pot with one vent,BS - both sided, pot with two vents, C - control pot without an escape vent, SE -standard error, DF - degree of freedom.

Fig. 4.Logistic selection curve of SS and BS pots for swimming crab, P.trituberculatus. Dashed lines show selectivity curves for each comparative haul, and solid lines show the mean curves from all the hauls.

Fig. 5.Plot of selectivity parameters (50% retention width, CW50, and selectivity range, SR) for SS (unfilled circles) and BS (filled diamonds) pots. Solidline circles show 95% confidence interval.

3.4.Effect of soak time and location of vents

The soak time and difference in location of escape vents between SS and BS pots, taken as the fixed effect, were set as qualitative controllable variables in mixed effect model analysis, and the hypotheses of effect of soak time and location of vents on size selectivity were tested. The results, as shown in Table 3, indicate that both hypothesis H01, i.e., soaktime have no effect, and hypothesis H02, i.e., the location of vents has no effect onSRs andCW50s, could not be rejected (p=0.738 for H01andp=0.410 for H02). The large overlap of 95% CIs for CW50and SR plots also indicates similar results, as shown in Fig. 5. In other words, the difference in size selectivity between SS and BS pots owing to the location of vents is insignificant. Therefore, the fisheries management measure in Zhejiang Province, which does not specify the location of escape vents for crab pots, is valid.

Table 3 Mean selectivity indexes of experimental pots under hypothesis testing models.SS - single side, pot with one vent, BS - both sided, pot with two vents, DF -degree of freedom.

4.Discussion

In this study, the effect of escape vents with 2.8 cm height on the size selectivity of crab pots for swimming crabP. trituberculatusin the East China Sea was evaluated. It was found that the installation of escape vents in cylindrical pots for swimming crabP. trituberculatusis an effective measure to reduce the number of sublegal-sized crabs less than 116 mm CW. Although two escape vents installed at the opposite side(BS pots) have some advantage in releasing more sublegal-sized crabs,no statistical difference was observed between the two types of experimental pots.

Since the 1940s, numerous studies have demonstrated that installation of escape vents, gaps, or rings is a simple and cost-effective measure for improving the size selectivity of crustacean pots, reducing overall mortality for crustaceans, especially sublegal-sized, retaining more large individuals and preventing ghost fishing of lost pots (Miller, 1990,1995). Increased catch rates of legal-sized swimming crabs in experimental pots, compared with control pots, are observed in this study, as shown in Figs. 2 and 3. Gear saturation is a key factor for fishing efficiency. Sublegal-sized swimming crabs comprise a relatively high percentage of saturation level in control pots. However, in the experimental pots, most sublegal-sized crabs escape through the vents and free up space inside the pots for others while larger crabs enter and retained.The advantage of escape vents has been proved in several occasions(Boutson et al., 2009; Guillory & Hein, 1998).

Selectivity and efficiency of pots with vents or gaps can be affected by various factors, including opening size (Arana, Orellana, & De Caso,2011; Gandy, Crowley, Leone, & Crawford, 2018; Jirapunpipat, Phomikong, Yokota, & Watanabe, 2008; Rotherham et al., 2013), number(Boutson et al., 2009; Nishiuchi, 2003), shape (Broadhurst, Millar, &Hughes, 2017; Miller, 1995), and location of vents, gaps, and rings(Arendt et al., 2018; Boutson et al., 2009; Havens, Bilkovic, Stanhope, &Angstadt, 2009; Stearns, Conrad, Winfrey, Shippentower-Games, &Finley, 2017). Winger and Walsh (2007) found that the number of snow crabsChionoecetes opilioattempting to escape from vents installed 5 cm from the bottom of crab pot is three times higher than the pots with vents installed 10 cm from the bottom. Havens et al. (2009) found that immature blue crabsCallinectes sapidusescaped at higher rates when escape rings were moved from the upper portion to the lower portion of trap. Similarly, higher efficiencies on releasingScylla olivaceaandPortunus pelagicuswere observed when the escape vents were mounted near the bottom sides of pots and traps, compared with those with their vents located on the top, as shown in Jirapunpipat et al. (2008) and Boutson et al. (2009). These studies showed that, the bottom edge was a more suitable location for escape vents in most instances, although the findings were from laboratory tests.

Few studies evaluated the effect of location of escape vents in pots on fishing efficiency and size selectivity in field experiments. Stearns et al.(2017) found that the escape rates of undersized Dungeness crabMetacarcinus magisterincreased when the position of escape rings was lowered towards the bottom, but no statistically significant difference was observed in the retention rates of sublegal crabs, similar to the findings of this study.

Fishing factors, including environmental conditions, soaking times,saturation, and orientation of pots, are more controllable in laboratory conditions than those during commercial fishing at sea. It is difficult to reproduce field environmental conditions, such as light condition, current, and food abundance reliably in the laboratory, which would lead to deviations in behaviors, including feeding, movement, and escape behavior (Miller & Addison, 1995). Furthermore, differences in the densities of individuals and size frequency distributions of population available to the pots, and differences in biological conditions such as levels of hungry, maternity stage, and sex between fields and laboratory would alter conspecific competitive behaviors for food and spaces between individuals. Interspecific competition and predation, which often exist in the field are usually not replicated in the laboratory. Therefore,whether the findings in laboratory would reliably represent the situation in fields should be treated with caution.

As mentioned above, the symmetrical feature of cylindrical crab pots, fished in a longline, inevitably led to their uncontrollable landing orientation with respect to the location of escape vents, if only one or two vents are mounted only on one side of the pot (single side). Grubert and Lee (2013) and Broadhurst, Butcher, and Cullis (2014) might have noticed this issue; they mounted two escape vents on the opposing side in each pot, similar to the arrangements in this study, to improve size selectivity for Giant Mud Crab,Scylla serratin their field experiments.However, they did not compare with pots with only one vent mounted on one side (single side).

The number of escape vents might be an important factor affecting the release efficiency of sublegal-sized crustaceans (Boutson et al., 2009;Eldridge, Burrell Jr, & Steele, 1979; Nishiuchi, 2003). Nishiuchi (2003)found that pots with two escape vents are more efficient in reducing the sublegal-sized crabs than those with one, but identical escape vent.Brown and Caputi (1986) suggested that multiple escape vents would increase the probability of individuals contacting a vent, thus more likely to escape through the vent. Inthis study, each BS pot had two escape vents while each SS pot had only one vent, i.e., the potential area for crabs to escape in each BS pot is twice that of a SS pot. Therefore,crabs inside a BS pot would have more chance to find and contact a vent than in a SS pot. This explains why the BS pots showed some advantage in releasing more small crab and had a largerCW50than SS pots. In other words, the location of vents might have no effect on the fishing efficiency.

Furthermore, according to the results of Wu (1996) and Winger and Walsh (2007), most crabs trapped in pots spent most of their time moving on the floor of the pot and made their escape attempts at that level. Winger and Walsh (2007) concluded that installing escape vents close to the floor of pots is a key factor for effectively releasing undersized crabs. As for Chinese crab pots, as shown in Fig. 1a, three funnel entrances extending to the bait box constructed a “floor” together,shortening the distance between crabs and escape vents, even if only one escape vent was located on the top side of pot. Besides, it has been observed that crabs can crawl comfortably on the walls of the pot(Winger & Walsh, 2007).

Trapped swimming crabs displayed intense conspecific antagonistic behavior, competing for food and space, as reported by Wu (1996), also observed in other crustacean species (Frusher & Hoenig, 2001; Ihde,Frusher, & Hoenig, 2006; Miller, 1979; Winger & Walsh, 2007). Smith and Sumpton (1989) found that trappedP. pelagicusattempted to defend the bait and guard the funnel entrances, causing less dominant individuals to flee from the pot. In this case, escape attempts of small-sized or weak crabs would be repeated for several times. This is probably the reason why the difference in the effect of location of vents on fishing efficiency becomes smaller in field experiments than those observed in the laboratory. For example, if it is assumed that the probabilities of a trapped crab contacting and escaping from the lower vent is 0.6, and that from the upper vent is 0.3, and if a crab makes five attempts, the difference in escape probability is decreased to 19% (1-0.45divided by 1–0.75) with the upper vent compared to the lower vent.

In conclusion, as an effective technical measure, installation of escape vents in Chinese cylindrical pots for swimming crabP. trituberculatuswould release more sublegal-sized crabs, wherever they are installed. Because of the lack of various catch compositions during all commercial operations in swimming crab pot fisheries, lack of field underwater observations and small length of soak time in this study,more research is warranted to further demonstrate the number, position,and size of swimming crab pots so that an optimal pot escape vent dimension and position of installation can be recommended for management to further conserve sublegal crab resources.

Declaration of Competing Interest

The authors declare that there is no conflicts of interest.

Acknowledgements

We would like to thank the crew of “Zheshengyu 090101” fishing vessel for their help and support during the sea trial. This study was jointly funded by National Natural Science Foundation of China(31001138) and Special Fund for Agro-scientific Research in the Public Interest (201203018).