528 



Fishery Bulletin 105(4) 



goal of a MPA is to enhance local fisheries by spillover 

 in addition to conserving spawning biomass, then it 

 must be designed and situated so that net movement 

 of fishes from the MPA to fished areas can take place. 

 In most tropical island nations, enhancement of local 

 fisheries by adult spillover may be critical for contin- 

 ued support of the preserve system by the local fishing 

 community (Russ and Alcala, 1999; Galal et al., 2002). 

 In general, arguments for larval spillover carry little 

 weight with fishermen because dispersal may not occur 

 on spatiotemporal or "visual" scales that are meaning- 

 ful to them (Russ and Alcala, 1996; Russ, 2002; Abesa- 

 mis et al., 2006). This situation is doubtless the case 

 in Guam, where little is known about the movement of 

 fishes on coral reefs. 



One characteristic of heavily exploited reefs is the 

 very low abundance (in some cases virtual absence) of 

 large carnivorous fishes, particularly groupers (Ser- 

 ranidae) and snappers (Lutjanidae) (Russ, 1991, 2002; 

 Medley et al., 1993). In Guam, grouper biomass is no- 

 ticeably lower than at the less heavily populated islands 

 in Micronesia (e.g., Palau). Much research has therefore 

 been directed at determining the effects of implement- 

 ing MPAs on populations of large predatory fishes (e.g., 

 Russ and Alcala, 1996, 2004). In Micronesia, however, 

 herbivorous fishes, particularly surgeonfishes and uni- 

 cornfishes (Acanthuridae) and parrotfishes (Scaridae) 

 are equally important to local fisheries, and in many 

 areas they have dominated the catch (Amesbury et al., 

 1986). Thus, it is important to understand the effects 

 of MPAs on herbivorous and carnivorous fishes. 



To address whether MPAs in Guam can increase fish 

 biomass and provide spillover to adjacent reef areas, 

 biomass of five important reef fish species inside three 

 MPAs in Guam and on adjacent exploited reefs was 

 determined by using underwater visual census. Net 

 movements of both herbivorous and carnivorous reef 

 fish across MPA boundaries were determined by using 

 mark-recapture experiments. 



Materials and methods 



Study sites 



In May of 1997, the Guam Department of Aquatic and 

 Wildlife Resources (DAWR) established a network of five 

 MPAs, termed "marine preserves," around the island, 

 which accounted for 11.8% of Guam's shoreline and 

 15.3% of Guam's reef area. Fishing within these MPAs 

 is restricted to shore-based cast netting and hook-and- 

 line fishing for select reef species, except for the Piti 

 Bomb Holes Marine Preserve, where no reef fishing (i.e., 

 fishing from the shore to the reef margin) is allowed. 

 Trolling seaward of the reef margin for pelagic fish is 

 allowed in all preserves. Despite being implemented 

 in 1997, the marine preserves did not become enforced 

 until October 1999. Even then, only warnings were 

 issued until January 2001, at which time the regulations 

 became fully enforced, and all violators were subject to 



any or all penalties applicable under the law. This study 

 was conducted in 2003-04 at three MPA sites (Achang 

 Reef Flat Marine Preserve, Piti Bomb Holes Marine 

 Preserve, and Tumon Bay Marine Preserve — hereafter 

 referred to as Achang, Piti, and Tumon, respectively) on 

 the western and southern coasts of Guam, Micronesia 

 (Fig. 1). All three sites consist of shallow reef flats that 

 lead seaward to a reef crest and then drop off to a reef 

 slope. The two remaining marine preserves, Sasa Bay 

 and Pati Point, were not surveyed. Sasa Bay is a man- 

 grove swamp with no coral reefs within its boundaries, 

 and Pati Point is located within a restricted military 

 area (Anderson Air Force Base). 



The Achang preserve is the largest of the three pre- 

 serves (4.8 km-). It includes a wide variety of habitats: 

 mangroves, seagrass, sand, coral, and three channels 

 that cut through the fringing reef from the outer reef 

 slope to the reef flat. The largest of these, Manell Chan- 

 nel, separates Achang reef flat from Cocos Lagoon, to 

 the west of Achang Reef Flat. To the west, the reef flat 

 narrows and is interrupted by a rocky headland. Refer- 

 ence fished sites for the Achang Preserve were located 

 in Cocos Lagoon. Seasonal traditional fishing is permit- 

 ted in the Achang Preserve for juvenile streamlined 

 spinefoot iSiganus argenteus), juvenile fusiliers [Pter- 

 caesio tile), and big eye scad (Selar crumenophthalmus) 

 under special permit. 



The Piti preserve covers 3.6 km- and consists of a 

 broad reef flat (1.4 km^) in Piti Bay bordered by Tepun- 

 gan Channel to the west. The fringing reef is continu- 

 ous from Piti Bay eastward to the fished sites at Asan 

 Bay. The reef flat includes unique features known as 

 "bomb holes," which provide sheltered areas of deeper 

 water. The deepest of these sinkholes were 9-10 m deep 

 and were densely populated by a variety of hard and 

 soft coral species. They host rich soft coral communities 

 and fish and invertebrate assemblages not often found 

 within the reef margin. The main sinkhole is occupied 

 by an 11-m in diameter underwater observatory that 

 was completed in 1996. It is also frequented by com- 

 mercial scuba divers during diving tours (up to 200 

 divers a day). Fish feeding is a common practice around 

 the observatory; therefore, our survey sites were located 

 away from the sinkholes in order to minimize possible 

 confounding effects of fish feeding and diver presence. 



The Tumon preserve lies adjacent to the central tour- 

 ist district on Guam. It is 4.5 km- in total area and con- 

 sists of extensive reef flats (2.7 km-), a gently sloping 

 fore reef slope, and a broad shelf habitat. The reef flat 

 is dominated by coral patch reefs. The Tumon MPA is 

 bordered by headlands to the east and west, which in- 

 terrupt the reef flat, although there is contiguous fring- 

 ing reef which links Tumon with Tanguisson to the 

 east. However, there is a sewage outfall at Tanguisson 

 which may reduce movement of fish to the east. Because 

 of the distance from the reef flat at the western bound- 

 ary of the Tumon MPA to the reef flat at East Agana 

 Bay (approximately 3 km), Tanguisson was the location 

 chosen for the fished sites. Limited traditional fishing 

 with hook-and-line or cast net from shore is allowed in 



