SCIENTIFIC COMMITTEE NINTH REGULAR SESSION

6-14 August 2013 Pohnpei, Federated States of

Preliminary Results on Biology for Elagatis bipinnulata Associated with Drifting Aggregation Devices in the Western and Central Pacific WCPFC-SC9-2013/ EB-IP-04

WANG XUEFANG123, DAI XIAOJIE 1,2,3, XU LIUXIONG 1,2,3, WANG ZHENHUA1,2,3

1 College of Marine Sciences, Shanghai Ocean University, Shanghai 201306,China; 2 National Engineering Research Center for Oceanic , Ministry of Education, Shanghai 201306, China; 3 Key Laboratory of Sustainable Exploitation Oceanic Fisheries Resources, Ministry of Education, Shanghai 201306, China Preliminary Results on Fishery Biology for Rainbow Runner Elagatis

bipinnulata Associated with Drifting Fish Aggregation Devices in the

Western and Central Pacific Ocean

WANG XUEFANG1,2,3, DAI XIAOJIE1,2,3, XU LIUXIONG1,2,3, WANG ZHENHUA1,2,3 1 College of Marine Sciences, Shanghai Ocean University, Shanghai 201306,China; 2 National Engineering Research Center for Oceanic Fisheries, Ministry of Education, Shanghai 201306, China; 3 Key Laboratory of Sustainable Exploitation Oceanic Fisheries Resources, Ministry of Education, Shanghai 201306, China

Abstract: Many pelagic tend to aggregate under drifting floating objects. This has led to the development of drifting fish aggregation devices (FADs) that attract tropical for the tuna purse seine fishery. However, FADs can also attract other non-target small pelagic species such as rainbow runner Elagatis bipinnulata. Using biological data collected in the Western and Central Pacific Ocean by scientific observers on board Chinese tuna purse seine fishing vessels, we evaluated the biological characteristics for rainbow runner aggregated around drifting objects. This study indicates 1) Fork length of rainbow runner ranged from 30.0 to 90.6 cm, with dominant fork lengths of 60.0 to 80.0 cm, accounting for 76.3% of the total sampled fish and suggesting that large rainbow runner dominate around drifting objects; 2) Fork length at 50% maturity was 65.7 cm, and mature individuals were dominant under the FADs; and 3) Some commonly observed small fish species, such as macarellus, Kyphosus cinerascens, Caranx sexfasciatus, Katsuwonus pelamis and the juveniles of Thunnus obesus and Thunnus albacares, were found in the stomach of rainbow runner, which suggests that rainbow runners under FADs prey on other FAD-associated small pelagic species. This last finding suggests that feeding is perhaps one of the most likely motivations for adult rainbow runners to aggregate under the FAD. Key words: Western and Central Pacific Ocean; tuna purse seine fishery; fishery biology; drifting fish aggregation device (FAD); Elagatis bipinnulata

Introduction Natural drifting floating objects including logs, kelp, branches, debris, etc., often attract and thus cause tropical and other pelagic species to form aggregations (Fréon and Dagorn 2000). Since the presence of natural floating objects is not always predictable, fishermen often construct and deploy artificial fish aggregation devices (FADs) to stimulate these aggregations and increase fishing efficiency (Wang et al. 2012). Sets on FADs have been rapidly become popular worldwide because of the high catchability, and currently 40% of the global tropical tuna catch comes from this fishing method (Dagorn et al. 2012). However, the use of drifting FADs also raises the possibility of potential negative impacts, e.g. increased by-catch and perturbation of pelagic ecosystem balance (Dagorn et al. 2012). Rainbow runner (Elagatis bipinnulata), a species distributed widely in oceanic and coastal shallow waters, is one of the most common by-catch by the purse seine fishery. Using biological data collected in the Western and Central Pacific Ocean by the scientific observers on board Chinese tuna purse seine fishing vessels, we examined some biological characteristics for rainbow runner associated with drifting FADs. This report presents biological information for rainbow runner in the FAD sets and gives insight into some possible reasons they aggregate under FADs.

Materials and Methods Chinese tuna purse seine observer survey data The data were collected from three survey cruises by the Chinese tuna purse seine fishery observers during October 2010 to February 2013. The fishing area is from latitude 2ºN - 12ºS and longitude 145ºE- 170ºE, mainly within the Bismarck Sea. A total of 49 sampling stations were distributed within this area (Fig. 1). The tuna purse seiners, Jinhui No.6 and Jinhui No.7, set on both FADs and other floating objects such as logs, animal debris and parts of destroyed fishing gear (commonly attached to FADs), but FADs accounted for nearly 90% of all the floating objects that were set on.

Fig. 1 Sampling locations and survey area

Data collection and analyses The observers recorded the date, sampling station location and catch for each set. Approximately five rainbow runners were sampled randomly in each set; biological information for a total of 219 samples was measured: Fork length (FL) FL for all samples was measured to the nearest cm and grouped in 5 cm bins to examine size distribution. Sex ratio 187 individuals were gutted to determine their sex. Sex ratio is defined here as the ratio of females to males. A chi-squared test was used to test the deviation in sex ratio from the expected 1:1. Maturation stages The observers determined the maturity stage for each individual based on the external morphology and weight of the gonads. Six stages of maturity were defined as: (I) undeveloped stage; (II) early developing stage; (III) later developing stage; (IV) mature stage; (V) spawned stage; and (VI) spent stage (Chen 2004). A total of 134 samples were observed for maturity stage. A logistic curve was fitted to length-specific proportion of sexually mature individuals (as in King 1995); the equation is: 1 P=L (1) 1+ exp(-r(L- Lm ))

where PL is the proportion of sexually mature individuals in fork length L, r is the slope of the curve, and Lm is the length at 50% maturity, which is defined as the length at which 50 percent of individuals are sexually mature. Feeding status A total of 183 stomachs were collected for analysis. Stomach fullness was characterized using a categorical scale from 0-4. The stages were: 0 for empty stomachs, 1 for traces of food, 2 for 1/2 full, 3 for 3/4 full and 4 for full (Tanabe, 2001). Recognizable prey items such as some common species associated with floating objects were directly recorded, otherwise they had to be treated as unidentified fish.

Results Size distribution The FL of 219 individuals ranged from 30.0 to 90.6 cm. The average FL was 67.0 and the standard deviation 11.0 cm, but most of the individuals (76.3%) ranged from 60 to 80 cm (Fig.2).

Fig.2 Fork length distribution of Elagatis bipinnulata associated with drifting floating objects

Sex ratio All the sampled fish were identified as female or male. The observed ratio was 1.00 female to 0.80 males, which did not deviate statistically from the expected ratio

2 of 1:1 according to a chi-squared test (  = 3.556, p > 0.05).

Sexual maturity The observers had no difficulty sexing the sampled fish. The number of immature fish (maturity stagesⅡ～Ⅳ) was 52, accounting for 38.8% of the total. The number of mature fish (maturity stagesⅤ～Ⅵ) was 82, accounting for 61.2% of the total. Fig. 3 shows the variability in the proportion of mature fish across the observed size classes: no adult fish was found in the size class of 30-50 cm, but with increasing FL, the proportion of mature individuals increased rapidly within size classes between 50 and 80 cm, and nearly all the fish were sexually mature by the time they reached

80 cm. Size at 50% sexual maturity (L50) was estimated to be 65.7 cm using equation (1), and the maturation rate parameter r was 0.089 cm-1.

Fig.3 Proportion of sexually mature Elagatis bipinnulata associated with drifting floating objects by length

Stomach fullness Among the 183 sampled fish examined for their stomach content, 43.2% had empty stomachs when they were captured (stage 0) and 24.4% only had small amounts of prey (stage 1). In contrast, individuals more full stomachs (stage 4) accounted for 22.4% of the total (Fig.4).

Fig.4 Frequency distribution for Elagatis bipinnulata associated with drifting floating objects by feeding stage

Fig.5 shows the variation in the proportion of full stomach by size: the proportion of full stomach of six size classes from 30 to 80 cm are, respectively, 33.3%, 38.9%, 44.4%, 53.3%, 65.4% and 77.8%, clearly an ascendant trend. The Fisher’s test indicated that the proportion of full stomachs was correlated significantly with size class (p<0.001).

Fig.5 Proportion of empty and full stomachs of Elagatis bipinnulata associated with drifting floating objects

Prey composition The contents of 69 non-empty stomachs were examined and prey composition of identified species is displayed in Table. 1. Some common small species associated with drifting objects were found in the stomach of rainbow runners, such as scad (Decapterus macarellus), blue sea chub (Kyphosus cinerascens), bigeye trevally (Caranx sexfasciatus), skipjack tuna (Katsuwonus pelamis) and juveniles of bigeye (Thunnus obesus) and yellowfin (Thunnus albacares) tuna. A small species of , , frequently appeared. Other small and which are not typically attracted to floating objects were also found. In addition, unidentified species were found in 42% of the sampled stomachs.

Table1 Prey composition in the stomach of Elagatis bipinnulata associated with drifting floating objects Typically associated with drifting floating Prey items Frequency objects Decapterus macarellus + + + ● Kyphosus cinerascens + ● Katsuwonus pelamis + ● Caranx sexfasciatus + ● () + + () + + crustacean () + young tunas + ● unidentified fish + + + N/A *Symbol “+ + +” represents the highest frequency of occurrence; symbol “+ +” represents the higher frequency of occurrence; symbol “+” represents the lowest frequency of occurrence; symbol “●” represents the species belong to associated species.

Discussion Kumoru (2007) investigated catch information from the FAD-based tuna purse seine fishery out of Papua New Guinea during 1999, and noted that the size frequency of rainbow runner had two distinct modes at 36-38 cm and 68-70 cm. Our study presented a general similar pattern of length distribution and our two modes were 35-45 cm and 65-75 cm. There was a clear separation between our large and small modes (Fig. 2). The difference in the modes from the two studies may result from the choice of size intervals (2 cm vs 5cm). The size structure of rainbow runner provides additional biological information: the mean FL was 67.0±11.0 cm and lengths between 60 and 80 cm accounted for 76.3% of the sampled fish. This suggests that larger individuals dominate around drifting objects. In addition, the estimated L50 for rainbow runner was 65.7 cm, suggesting that floating object-associated schools of rainbow runner are dominated by mature individuals. Combining these two observations, we conclude that larger mature adults are the main component of aggregations of rainbow runner associated with floating objects. This contrast with the behavior of other species: skipjack tuna associated with floating objects are typically smaller that their conspecifics in free swimming schools (Fréon and Dagorn 2000). Klima and Wikham (1971) observed large pelagic predators, e.g., rainbow runner and ( spp .) with mixed schools of small , e.g., round scad (Decapterus punctatus), Spanish sardine (Sardinella anchovia), and scaled sardine (Harengula pensacolae), but the large fish were not feeding on the small ones. In our study, commonly associated small fish species such as mackerel scad, blue sea chub, bigeye trevally, skipjack tuna and juveniles of bigeye and yellowfin tuna, were indeed found in the stomach of rainbow runners, which suggests that rainbow runners associated with drifting floating objects do prey on other associated small pelagic species. Other observations also indicate that some occurs under floating objects (Fréon and Dagorn 2000). The “Concentration of food supply” hypothesis has been used to explain the presence of predatory fish, like the dolphin fish (Coryphaena hippurus), around floating objects (Kojima, 1956). Our observations suggest that feeding is perhaps one of the most probable motivations for adult rainbow runners to aggregate under FADs, in agreement with the “Concentration of food supply” hypothesis. This contrasts with the hypothesis of Kumoru (2007), who proposed FADs serve instead as a “meeting place” for individuals of the same species. Considering the large-scale deployment of FADs by purse seiners, base-line knowledge of the biology of species that are associated with floating objects is particularly important for fisheries management. The average school associated with an FAD shows a wide range of species and size distributions; thus the development of appropriate techniques to sample fish characteristics within and across species is critical. Fishery-dependent methods such as those used in this study, as with any sampling method, have certain biases, and a research goal should be the development of less biased sampling procedures for FAD-associated schools.

Reference Chen XJ. Fisheries resource and fishing ground research. Beijing: Ocean Press. 2004. Dagorn L, Holland KN, Restrepo V, Moreno G. Is it good or bad to fish with FADs? What are the real impacts of the use of drifting FADs on pelagic marine ecosystems? Fish and Fisheries, 2012, DOI: 10.1111/j.1467-2979.2012.00478.x Fréon P, Dagorn L. Review of fish associative behavior: toward a generalization of the meeting point hypothesis. Reviews in Fish Biology and Fisheries, 2000,10: 183-207 King M. Fisheries biology, assessment and management. Blackwell Science, Oxford. 1995. Klima EF, Wickham DA. Attraction of coastal pelagic with artificial structures. Trans. Amer. Fish. Soc., 1971, 1: 86-99. Kojima S. Fishings for dolphins in the western part of the Japan Sea -II. Why do the fish take shelter under floating materials? Bull. Japan. Soc. Sci. Fish. 1956, 21(10): 1049-1052. Kumoru L. Catch information from the FAD-based domestic tuna purse seine fishery in Papua New Guinea. Scientific committee third regular session of the Western and Central Pacific Fisheries Commission. WCPFC-SC3-FT SWG/WP-8. Honolulu, USA, 2007. Tanabe T. Feeding habits of skipjack tuna Katsuwonus pelamis and other tuna Thunnus spp. Juveniles in the tropical western Pacific, Fish. Sci. 67(4) (2001), pp. 563–570. Wang XF, Xu LX, Chen Y, Zhu GP, Zhu JF, Tian SQ. Impacts of fish aggregation devices on size structures of skipjack tuna Katsuwonus pelamis. Aquatic ecology, 2012, 46(3): 343–352.