534 



Fisher/ Bulletin 105(4) 



12 3 4 5 6 



Biomass ratio (biomass Inside MPA/biomass outside MPA) 



Figure 3 



Effect of biomass ratio (biomass in marine protected 

 area/ biomass in reference site) on spillover of reef fishes 

 from marine protected areas in Guam, Micronesia. Points 

 represent the level of spillover for a given biomass ratio. 

 Data are pooled for five species and three MPAs. Study 

 species are convict surgeonfish (Acanthurus triostegus), 

 honeycomb grouper {Epinephelus merra), yellowstripe 

 goatfish (Mulloidichthys flavolineatus), orangespine uni- 

 cornfish (Naso Uturatus). and little spinefoot iSiganus 

 spinus). 



pomus spp. after <10 years of protection (reviewed in 

 Evans and Russ, 2004). In Kenyan MPAs, complete 

 recovery of fish populations was estimated to take 22 

 years (McClanahan and Graham, 2005). Other studies 

 have reported more rapid buildup of biomass. Biomass of 

 five commercially exploited fish families tripled within 

 three years of implementation of the Soufriere Marine 

 Management Area in St. Lucia (Roberts et al., 2001). 

 Biomass of reef fish in another MPA in St. Lucia, Anse 

 Chastanet Reserve, doubled within two years of effective 

 protection (Roberts and Hawkins, 1997) and a simi- 

 lar, rapid buildup of biomass was observed in a small 

 MPA in Saba (Roberts, 1995). Rates of biomass buildup 

 inside MPAs should generally be consistent with life 

 history characteristics of the fish (Russ and Alcala, 

 1996; Mosquera et al., 2000; Alcala et al., 2005). Large 

 predators (e.g., epinepheline serranids [large groupersl, 

 Lutjanidae, Lethrinidae, and Carangidae) and many 

 Acanthuridae (surgeonfish) are long-lived, often with 

 low rates of natural mortality and recruitment. Such 

 characteristics would indicate that recovery rates would 

 be gradual, as observed by Evans and Russ (2004) and 

 Alcala et al. (2005). 



In this study, it was not possible to determine the 

 rate of biomass buildup because data were not collected 

 at the initial implementation of the MPAs or at the 

 initiation of full enforcement. After approximately 2.5 

 years of protection, biomass of all five study species of 

 reef fish was higher within the MPAs than in fished 

 sites, although the difference was not statistically sig- 



nificant at the 95% confidence level for yellowstripe 

 goatfish. A longer period of protection may result in 

 greater biomass differentials between the marine pre- 

 serves and fished sites. Biomass of all three species of 

 herbivores was significantly higher within the MPAs, 

 indicating that fishing pressure on herbivores in Guam 

 is sufficient to show a biomass increase within no-fish- 

 ing zones. Thus, increasing herbivore biomass on ex- 

 ploited reefs through spillover from MPAs may have the 

 potential to reduce algal overgrowth, at least within a 

 limited area adjacent to the MPA. 



Given rapid population turnover, coupled with high 

 fishing effort in the reference areas, significantly higher 

 biomass in the MPAs may be evident after only 2-3 

 years of protection. All the species in this study are 

 small to medium-size fishes with rapid growth and 

 maturation rates (Choat and Robertson, 20021. What 

 is somewhat surprising is that the densities of convict 

 surgeonfish, yellowstripe goatfish, and little spinefoot 

 were all significantly higher in the Tumon MPA than 

 at the adjacent fished site, given that these species were 

 legally targeted by subsistence fishermen within the 

 MPA. However, the regulations stipulate that fishing 

 with hook-and-line or cast net only and from shore or 

 the exposed reef margin only. Moreover, fishermen in 

 Tumon would have to contend with large numbers of 

 tourists in the immediate vicinity, and fishing effort 

 appeared low throughout the course of this study. 



Several studies indicate that MPAs connected to 

 fished areas by continuous reef will have higher rates 

 of spillover (e.g., Kaunda-Arara and Rose, 2004). In 

 this study, the highest overall spillover to surrounding 

 fished reefs occurred from Piti, where the MPA and 

 fished site are connected by a continuous reef flat. This 

 was the only site at which no net inward movement of 

 any species occurred. Fishermen were often seen along 

 the boundary of the marine preserve, no doubt "fishing 

 the line" in hopes of catching larger fish emigrating 

 from the Piti MPA. When species were combined, the 

 lowest rate of net flux occurred at Tumon. The fished 

 site to the east of Tumon was at Tanguisson. These 

 two bays are divided by a high, rocky headland (Punta 

 Dos Amantes) with no reef flat. A sewage outfall just 

 north of Punta Dos Amantes also separates the MPA 

 and fished site. However, three of the five study species 

 appeared to move freely between Tumon and Tanguis- 

 son — only yellowstripe goatfish and honeycomb grou- 

 per did not. Thus, the low overall rate of net flux was 

 caused by the net import of convict surgeonfish and 

 little spinefoot that balanced the net export of orang- 

 espine unicornfish. The overall spillover from Achang 

 to Cocos Lagoon was also low, perhaps partly because 

 these areas are separated by a wide tidal channel to 

 the west of the Achang reef flat. However, the low mean 

 spillover from Achang occurred because of the large 

 number of adult yellowstripe goatfish moving into the 

 MPA. These results demonstrate the importance of 

 determining spillover at the species level. Because 

 different species can vary in their market value, the 

 mean spillover of all fishes from an MPA may not be 



