American Fisheries Society Symposium 41:439–460, 2005 © 2005 by the American Fisheries Society Effects of Commercial Otter Trawling on Benthic Communities in the Southeastern Bering Sea ELOISE J. BROWN,1 BRUCE FINNEY, AND SUE HILLS Institute of Marine Science, University of Alaska Fairbanks, 245 O’Neill Building, Post Office Box 757220, Fairbanks, Alaska 99775, USA MICHAELA DOMMISSE Monash University, Department of Geography and Environmental Science, Post Office Box 11A, Clayton 3168, Australia Abstract. The effects of commercial bottom trawling for yellowfin sole Limanda aspera on benthic communities were investigated in a sandy habitat exposed to high wave and tidal disturbance at 20–30 m depth in the southeastern Bering Sea. We compared an area that has been closed to commercial trawling for 10 years with an adjacent area that is now open to commercial trawling. In addition, we examined the immediate effects of experimental trawling on benthic community structure in the area closed to trawling. The fished area was characterized by reduced macrofauna density, biomass, and richness relative to the closed (unfished) area, but diversity was not different. Interannual variability of macrofauna assemblages was high in the system, yet assemblages in the two areas were distinguished using multivariate analyses and dominant taxa. After 10 years, sessile taxa (e.g., Maldanidae polychaetes) were prevalent in the closed area, and mobile scavengers (e.g., Lysianassidae amphipods) were more common in the fished area. Immediate responses of macrofauna to experimental trawling were subtle (i.e., reduced richness, absence of rare taxa, and patchy changes in assemblage biomass), but no differences were detected relative to controls for density, diversity, or total biomass. Fragile, structure-forming megafauna were rare. However, when trawled, they appeared mostly unaffected. Though we could not completely rule out other factors such as food supply from water column primary production, our results indicate that trawling altered macrofauna communities. Our findings also suggest that individual taxa respond differently to trawling but that commonly used summary measures such as total abundance do not capture these changes. Based on the functional attributes of individual taxa, hypotheses explaining different macrofaunal assemblages include bottom-up shifts caused by physical disturbance from trawling, altered benthic food webs from discards and processing waste, and top-down shifts caused by altered predator–prey interactions between Asterias amurensis and yellowfin sole. Stomach contents of yellowfin sole indicated that this flatfish did not target specific prey taxa; however, fish- processing waste was prominent in its diet. More ecological and bioenergetic data are required to determine how changes in benthic communities are linked specifically to the productivity of yellowfin sole. Introduction resources (e.g., Dayton et al. 1995; Jennings and Kai- ser 1998; Hall 1999; Norse and Watling 1999). Bottom Mobile fishing gear has recently become the focus of trawling has even been compared to forest clear-cut- international attention as scientists, fishers, and policy ting (Watling and Norse 1998). Yet habitat type, gear makers fear habitat degradation, loss of biological di- type, and gear configuration largely affect the degree to versity, and the subsequent consequences to fishery which biological communities are affected by fishing. Few experimental studies have quantified trawl effects using realistic fishing gear at the intensities imposed by 1 Corresponding author: [email protected]; commercial fleets (Collie et al. 2000). present address: University of Western Australia, School The southeastern Bering Sea supports the largest of Plant Biology (MO90), 35 Stirling Highway, Crawley flatfish fishery in the United States (>300,000 metric WA 6009, Australia. tons in 1997; NPFMC 1997; Fritz et al. 1998; NRC 439 439-460_BH_Brown.pmd 439 8/10/2005, 3:48 PM 440 BROWN ET AL. 2002). Targeted primarily by commercial catcher–pro- sediments in the fished zone were slightly better sorted, cessors, yellowfin sole Limanda aspera accounted for less variable, and contained less fine-grained sediment over half the harvest at the time of this study (NPFMC than those of the unfished zone (Brown et al., in press). 1997; Fritz et al. 1998). Despite the long history and We focused on three hypotheses: (1) Benthic com- economic importance of fishery resources in the Bering munity indices including density, biomass, diversity, Sea, few studies on otter–trawl impacts have been con- richness and multivariate patterns in macrofaunal as- ducted in this region (McConnaughey et al. 2000, 2005, semblages do not change in response to chronic expo- this volume). However, trawl impacts on benthos in sure to commercial trawling for flatfish; (2) Benthic other areas (e.g., North Sea, eastern Atlantic, South community indices and multivariate patterns do not Pacific) include reduced infauna density; reduced bio- change immediately after experimental trawling; and mass and diversity; selective elimination of large, sed- (3) Based on (1), (2), and yellowfin sole diet composi- entary, and fragile taxa by physical removal; or damage tion, trawling does not affect essential fish habitat for in heavily trawled areas (e.g., Bergman and Hup 1992; this species. Eleftheriou and Robertson 1992; Jones 1992; Collie et al. 1997; Engel and Kvitek 1998; Freese et al. 1999). In contrast, most research in shallow, high-energy regions Methods did not find detectable changes to benthic fauna caused by trawl gear (e.g., Gibbs et al. 1980; Brylinsky et al. Study Location 1994). Populations of certain prey and commercially Since the 1950s, a single commercial groundfish fish- important species may be enhanced by some level of ery (yellowfin sole) has been active in the federally trawl disturbance through increased food availability managed waters of the southeastern Bering Sea, within from damaged organisms and the replacement of large an area created in 1996 called the Nearshore Bristol benthic animals with smaller, opportunistic prey (e.g., Bay Closure (Figure 1a; Witherell and Pautzke 1997). Millner and Whiting 1996; Engel and Kvitek 1998). The catch is primarily yellowfin sole, and bycatch rates The 1996 Magnuson–Stevens Fishery Conserva- are particularly low (NPFMC 1991). Immediately ad- tion and Management Act (MSFC) requires the identi- jacent to the fishery area is the Walrus Islands State fication and description of essential fish habitat (EFH) Sanctuary, where the Alaska Department of Fish and for each commercially fished species within the United Game has prohibited all commercial fishing in state States and determination of any adverse impacts from waters (out to 5 km; 3 mi) since 1989. Within approxi- fishing (U.S. Code, title 16, section 1801 et seq.; Pub- mately 19 km (12 mi) of Round Island, fishing has lic Law 94-265 as amended October 11, 1996). Essen- been prohibited since 1992 by the North Pacific Fish- tial fish habitat is defined as those waters and substrate ery Management Council (P. Koehl, Alaska Depart- necessary for spawning, breeding, feeding, or growth ment of Fish and Game Wildlife Conservation Divi- to maturity and, hence, encompasses an enormous range sion, personal communication). Global positioning sys- of Alaska’s marine environment (DiCosimo 1999). tem (GPS)-based maps of fishing locations and gear Given the regional differences and spatial heterogene- type recorded by the National Marine Fisheries Service ity in benthic communities and substrate types, it is (NMFS) Observer Program indicate that the Round crucial that trawl impacts are evaluated specifically for Island 12-mi (~19 km) closure has not been fished for the southeastern Bering Sea. approximately 10 years (NPFMC 1997; NRC 2002). The objectives of this 2-year study were to evalu- Our study area spanned the Round Island closure and ate long-term effects of commercial flatfish trawling on adjoining commercial fishery areas (Figure 1b) and EFH in the southeastern Bering Sea and to investigate offered a rare opportunity for comparison of two areas initial responses of the benthic community to experi- that were otherwise extremely similar. This study is mental trawling. Macrofaunal and megafaunal commu- based on two research cruises to this site with R/V Big nity responses to trawling were described, and yellow- Valley in 1999 and 2000. fin sole prey were examined to determine potential im- pacts on EFH for this species. The incorporation of Chronic Study Design commercial fishing intensity and gear typical for the Bering Sea flatfish fleet ensured that the results of the Long-term effects of trawling on the benthic commu- experiment were relevant to local fisheries and manag- nity were evaluated based on a comparison between the ers. A concurrent study of physical habitat response to Round Island closure (unfished) and commercially trawling in this study area found that surface sediments fished habitats. In 1999, we selected a study area where were composed primarily of fine sand with low or- the fished and unfished areas were contiguous and had ganic content (~3 mg C/g). However, the top 3 cm of similar depths (20–30 m; mean = 26 m) and sediment 439-460_BH_Brown.pmd 440 8/10/2005, 3:48 PM EFFECTS OF COMMERCIAL OTTER TRAWLING ON BENTHIC COMMUNITIES 441 (a) Bering Sea Commercial Fishery Round Island 58°N 57°N 56°N Alaska 55°N 54°N Gulf of Alaska 53°N 175°W170°W 165°W 160°W (b) 59°N 25 m Round Island 15 m 12-mile closure Controls C5 C4 C3 Experimental block C2 C6 F3 F5 F4 F1 35 m N F2 25 m 58°N W E Commercial Fishery open April 1 - June 15 35 m S 45 m 0 10 20 30 40 50 Kilometers 160°W 159°W Figure 1. (a) Nearshore Bristol Bay trawl closure area showing location of Round Island (closed to fishing) and commercialy fished area, adapted from Witherell and Pautzke (1997). (b) Study area showing location of closed stations (C2–C5) within the unfished 12-mi (~19 km) Round Island closure and fished stations (F1–F5) within the commercial trawl fishery boundaries.