1 INTRODUCTION

The Asian paddle crab, Charybdis japonica, is widely distributed along the rocky shores of China, Korea, and Japan (Vazquez Archdale and Kuwahara, 2005 ; Yu et al., 2005) and is an important commercial crab species in China. Accordion-shaped traps (Fig. 1a) are largely used by Chinese fishermen to catch these crabs in reef waters and artificial reef areas. The accordion-shaped trap is a unique pot used only in China; its target species are bottom-dwelling fish and crabs. This trap consists of a series of 30 small connected sections (each 25 cm×15 cm×18 cm). Each section has a funnel entrance, except for the last three sections at each end, which are used to collect the crabs. The funnel entrances to the accordionshaped traps alternate between opposite sides to ensure that crabs can enter from both sides (Fig. 1b). The accordion-shaped trap is the most commonly used gear in Chinese artificial reef areas, where trawls and many other types of equipment are unsuccessful. The accordion-shaped trap has many advantages over other designs. Its low cost, low labor requirements and high portability allow fishermen to operate a large number of traps on a small deck. These traps are collapsible and are connected together during the fishing operation by tying the cod-ends of adjacent traps. The connected traps form long trap walls underwater to capture bottom-dwelling species. Because of the high catch rate, fishermen do not bait these traps. After 2 days of soaking, the trap at one end is pulled onto the boat and the others are hauled in manually. However, the mesh size (stretched) of the accordion-shaped trap is only 1.50 cm and the catch often contains many undersized crabs and nontarget species that must be discarded because of their low economic value. This approach suggests that mitigation measures to conserve the crab resource are urgently needed, but no practical successful measures have been taken so far.

Many approaches have been used to reduce the catch of undersized target species, by-catch, and discards during pot fishing, e.g., larger mesh size (Guillory and Prejean, 1997 ; Zhou and Shirley, 1997 ; Vazquez Archdale et al., 2006 ; Winger and Walsh, 2011), modified mesh shape (Guillory and Hein, 1998), modified pot structure and entrance design (Vazquez Archdale and Kuwahara, 2005 ; Vazquez Archdale et al., 2006, 2007 ; Barry et al., 2010), and installation of escape openings (Brown, 1982 ; Guillory and Merrell, 1993 ; Boutson et al., 2009 ; Iskandar, 2011 ; Winger and Walsh, 2011 ; Grubert and Lee, 2013). Escape openings are widely used in decapod fisheries to minimize the catch of undersized animals, particularly rigid-shelled lobsters and crabs (Weber and Briggs, 1983 ; Everson et al., 1992). Gear selectivity can be improved by changing the shape, location, size, and construction material of escape vents (Boutson et al., 2009). For example, a circular escape vent is suitable for lobsters, round fish (Treble et al., 1998) and blue crabs Callinecte s sapidus (Eldridge et al., 1979). A circular escape vent is more effective than a square vent because it allows lobsters to escape while retaining commercial-sized crabs (Stasko, 1975). A rectangular escape vent is optimal for catching the blue swimming crab, Portunus pelagicus, with low risk of injury, and provides a more efficient escape for undersized crabs than square, circular, and elliptical vents (Boutson et al., 2009). Fitting a rectangular vent is also cheaper than fitting vents of other shapes and can be easily performed by artisanal fishermen (Jirapunpipat et al., 2008). The position of the escape vent can also affect escape frequency from the trap (Jeong et al., 2000 ; Boutson et al., 2009). The bottom of the side panel of a collapsible pot is the best position for a vent to allow crabs to escape, because it takes advantage of their behavioral characteristics (Boutson et al., 2009). Crabs mainly move by crawling sideways around the corners of the bottom pot panel. As the accordionshaped trap has no side panels, the bottom of the front panel is prioritized. Given the possible availability and acceptance of a modified trap by pot fishermen, the simplest modification of an accordion-shaped trap with minimum cost would be to install a rectangular escape vent at the bottom of the front panel, opposite the entrance.

A laboratory experiment was conducted previously to identify the most appropriate size for an escape vent (Zhang and Zhang, 2013). Size selectivity of the escape vent was assessed precisely by placing crabs of known size in one section of an accordion-shaped trap with an escape vent. Crabs usually escape from vents by side-crawling, so the selectivity for carapace length (CL) depends on the vent length, and the carapace height (CH) is related to vent height (Boutson et al., 2009). In that experiment, the carapace length for 50% selection (L₅₀) of C. japonica was found to be 92.30% of vent length. According to the minimum market-acceptable CL of 4.00 cm, the vent dimensions were set at 4.3 cm×3.0 cm (L×W). In this study, 12 modified accordion-shaped traps with vents were constructed and compared with conventional accordion-shaped traps in fishing experiments in an artificial reef. The objectives of this study were to test the actual effects of the escape vents on crab catch and size, and to provide a basis for appropriate utilization of the Asian paddle crab resource.

2 MATERIAL AND METHOD 2.1 Design of the vented accordion-shaped trap

The conventional accordion-shaped trap is the most common commercial trap used by local fishermen. Vented accordion-shaped traps were modified from conventional traps. Iron-wire escape vents of 4.3 cm×3.0 cm were installed in the center of the bottom, opposite to the entrances of each section (Fig. 2); each accordion-shaped trap contained a total of 24 escape vents.

2.2 Experimental method

A fishing experiment was conducted to compare the conventional and vented accordion-shaped traps in the artificial reef area of Zhuwang, Laizhou Bay, Shandong Province, China (37°16.400′-37°16.600′N and 119°51.750′-119°52.350′E) from June 23 to August 22, 2012. Three conventional and six vented traps were tied together in sequence as a single group, and pairs of groups were deployed at positions separated by 50-150 m at depths of 7-12 m (Fig. 3). One 7.5-kg stone anchor was set at each end of a trap group. No bait was used in any of the traps during the experiment. To ensure that the traps extended fully, they were shot from a boat sailing at slow speed. The traps were hauled up in the morning after 2 days of soaking, provided there were no exceptional weather conditions. If there were exceptional conditions, the traps were left longer. The catch was identified immediately after each fishing operation. The number of individuals caught during two trap days divided by the number traps used was defined as the catch per unit effort (CPUE) for each trap type. After the catch was removed, the traps were re-tied together and deployed at two other positions. A total of 30 fishing trials were conducted following the local fishermen's normal fishing procedures. Traps were repaired if breakages occurred. Species, size, and body weights of individuals caught by the two trap types were recorded separately.

2.3 Analytical methods

The SELECT model was used to evaluate selectivity of the vented accordion-shaped traps (Xu and Millar, 1993 ; Treble et al., 1998 ; Jeong et al., 2000):

where p is relative fishing intensity; a and b are selective parameters; cl is the characteristic CL of C. japonica ; and S_CL is the selectivity rate of the trap for cl -size C. japonica. The a, b, and p values were estimated using the maximum-likelihood method in Microsoft Excel solver, and a master curve was simulated according to the characteristic CL and selective rate.

A paired t-test was used to detect differences in CPUE and catch size between the conventional and vented traps. A P-value<0.05 was considered significant.

3 RESULT 3.1 Catch species composition of the two trap types

The species compositions of the catch of the two trap types are shown in Figs. 4 and 5. A total of 2 422 individuals and 19 species were caught during the 30 fishing trials, including 852 individuals and 16 species caught by the conventional traps and 1 570 individuals and 18 species by the vented traps. The dominant species were C. japonica and Rapana venosa. The total numbers of C. japonica and R. venosa were 1 272 and 653, respectively, accounting for 52.52% and 26.96%, respectively, of the total catch. The catch species of the two trap types were similar. However, the presence of the escape vent increased the catch rates of C. japonica to 55.48%, compared with 47.07% in the conventional traps. However, the catch rates of fish, such as Sebastes schlegelii, Kareius bicoloratus, and Platycephalus sp. 1, were lower in the vented traps, while the catch rates of R. venosa were similar (vented 26.88%; conventional 27.01%). Overall, the vented traps had a smaller non-target catch and an optimized catch composition compared with the conventional traps.

3.2 C. japonica catch comparison

Asian paddle crabs of CL<4.00 cm were considered undersized and CL≥4.00 cm were considered marketable. A total of 136 undersized crabs and 265 marketable crabs were caught in 174 conventional trap hauls, and 112 undersized crabs and 759 marketable crabs were caught in 311 vented trap hauls.

Statistical analysis revealed significant differences in selectivity performance between the two gear types; a significantly lower CPUE for undersized crabs was observed in the vented traps than in the conventional traps (paired t-test: P<0.001)(Fig. 6a). The ratio of undersized crabs was significantly lower (paired t-test: P<0.001) in vented traps (12.53±0.69)% than in conventional traps (35.05±2.57)%(Fig. 6b). Consequently, mean crab CL was significantly higher (paired t-test: P<0.001) in vented traps (4.90±0.06 cm) than in conventional traps (4.42±0.16 cm)(Fig. 6c). Importantly, the CPUE of marketable crabs was significantly higher in the vented traps (2.61±1.29) than in the conventional traps (1.60±1.24)(Fig. 6d). The distributions of CL of C. japonica from the two trap types are shown in Fig. 7. The crabs were grouped into 0.50 cm CL classes, with a range of 2.00-7.90 cm. The numbers in each CL class caught by the 174 conventional and 311 vented hauls were converted to the numbers caught per 100 hauls. The vented type traps captured fewer undersized crabs and had higher marketable crab capture efficiency. The C. japonica catch composition details are shown in the Appendix.

Although the vented traps caught a greater number of larger crabs than the conventional traps did, the female: male ratio was much lower in the vented traps (1:1.90 vs. 1:1.03). This phenomenon may reflect their different body sizes; female crabs are generally smaller than male crabs, enabling more female crabs to escape through the escape vent. The relationships among carapace width (CW; excluding spines), CH (measured from the base of the second sternal segment to the highest part of the gastric region), body weight (W), and CL of C. japonica are shown in Fig. 8. The calculated parameters were as follows:

3.3 Size selectivity of the vented accordion-shaped trap for C. japonica

The SELECT estimated split model was used to assess the selectivity of the escape vent by considering the L₅₀ value. A master curve was simulated based on the selectivity rate and the characteristic CL, which was the mean CL of every CL class (Fig. 9). The parameter estimates were: a = -6.928 23, b = 1.679 34, p = 0.674 67, and L₅₀ =- a / b ≈4.13 cm. The estimated selectivity curve from the field experiment had a higher L₅₀^than the theoretical L₅₀ of 4.00 cm (Zhang and Zhang, 2013), indicating a lower probability of retaining crabs with CL>4.00. This result may have been caused by tears in the netting of the experimental traps, which were operated in an artificial reef area full of sharp rocks, permitting more crabs to escape.

3.4 Catch comparisons of other species

Comparisons of total catch, CPUE, and mean sizes of other species captured by the two trap types are shown in Table 1. No differences were observed in the R. venosa catch, which was one of the two dominant species caught. The vented trap caught fewer smallersized species, such as S. schlegelii, Platycephalus sp. 1, K. bicoloratus, Pseudopleuronectes yokohamae, and Paralichthys olivaceus, but caught larger S. schlegelii, Platycephalus sp. 1, and Octopus variabilis, indicating higher selectivity of the vented traps for these fish, although most fish caught were juveniles. Some species accidentally caught by the two trap types are also shown but their CPUE was not calculated.

4 DISCUSSION

Pots are important fishing gear for crabs and lobsters and are commonly used in topographically complex waters, such as tropical waters, coral reefs, and outcrops. The Laizhou Bay reef area is shallow (7-12 m) and the small accordion-shaped trap is much more suitable for fishing there than other pot types. The flexibility of the accordion-shaped traps allows them to be deployed in any type of bed, even on piles of stones. Although other types of gear can be very selective and do not capture small crabs, with limited by-catch (Vazquez Archdale et al., 2006), local fishermen prefer to use the accordion-shaped trap in artificial reef areas, because they are more suitable for uneven sea bottoms. However, the 2a mesh size of the accordion-shaped trap is only 1.50 cm and the funnel entrance was designed as a non-return device. Thus, many undersized crabs, fish, and non-target species are caught, potentially causing damage and stress to these resources.

Adding an escape opening successfully excludes small animals and improves pot selectivity (Eldridge et al., 1979 ; Brown, 1982 ; Guillory and Merrell, 1993 ; Guillory and Hein, 1998 ; Winger and Walsh, 2011). The present study showed that fitting traps with escape vents markedly reduced their retention of undersized crabs by increasing the L₅₀ to 4.13 cm; fewer non-target species and commercially less valuable species were caught. Moreover, catching fewer undersized crabs and less by-catch shortens the sorting time for fishermen and reduces the mortality of the discarded by-catch. Discards are subjected to exposure, displacement, and loss of appendages during sorting (Furevik and Løkkeborg, 1994 ; Tallack, 2007) and it is difficult to guarantee their survival (Treble et al., 1998 ; DiNardo et al., 2002). The 4.30 cm×3.00 cm escape vent in the bottom of the panel was appropriately sized for C. japonica, as it reduced the catch rate of undersized crabs from 35.05% to 12.53% and improved mean crab CL from 4.42 to 4.90 cm. Blue crab, C. sapidus, studies have shown that escape ports in the top of pots are superior to those in the bottom (Eldridge et al., 1979), as large crabs prefer to remain on the bottom of blue crab pots, which forces small crabs into other locations, particularly the top of the pot. Installing an escape vent in the top of an accordion-shaped trap should be evaluated in a future study.

In this study, the vented trap increased the catch of marketable C. japonica compared with that of the conventional trap. Similar results were reported for Cancer pagurus, Homarus gammarus (Brown, 1982), and Homarus americanu s (Fogarty and Borden, 1980). This increased catch may be related to the change in appearance of the accordion-shaped traps with escape vents. Crabs, including C. japonica, prefer holes or crevices. The escape vent resembles a hole or crevice and may entice crabs to enter. Conversely, the undersized crabs caught in the accordion-shaped trap may be more likely to escape through the escape vent because of food scarcity or competition for space with larger crabs, leaving the marketable crabs. However, further research is needed to determine the exact reason for the observed selectivity; many parameters affect catch efficiency and selectivity during pot fishing, making it difficult to isolate the effect of a single parameter.

Artificial reefs are attractive to many creatures, particularly bottom-dwelling crabs and fish. Fish species that inhabit rocky environments constitute a large proportion of the catch in some artificial reef areas. However, in this study, the dominant species caught were C. japonica and R. venosa rather than fish species associated with rocky areas. Installing escape vents in the traps improved C. japonica selectivity. The traps did catch some fish that inhabited rocky areas but most were undersized. The escape vent may have released some of these young fish, as the CPUE of S. schlegelii and Platycephalus sp. 1 was lower in the vented traps than in the non-vented traps. As a consequence, the mean body lengths of the captured S. schlegelii and Platycephalus sp. 1 were 9.37 and 9.67 cm, respectively, in the non-vented traps and 12.67 and 11.31 cm, respectively, in the vented traps. This will help protect the fish resources in the artificial reef area. In short, installing escape vents in accordion-shaped traps is a low-cost, convenient solution with high ecological benefit.

Reducing the catch of undersized crabs and increasing the catch of marketable crabs by modifying the accordion-shaped trap with an escape vent may help to establish a more sustainable and efficient crab fishery in an artificial reef area. Nevertheless, the minimum market-acceptable C. japonica CL, which has been set at 4.00 cm, is not based on scientific research or a legal regulation but on the experience of fishermen. No minimum landing size is currently enforced to protect C. japonica in China. Therefore, further studies are required to improve the selectivity of the accordion-shaped trap and to protect and use the crab resource.

5 ACKNOWLEDGMENT

We thank Shandong Oriental Ocean Group Co., Ltd. for their continuing support of our research.