Tupper: Spillover of commercially valuable reef fishes from marine protected areas in Guam 



531 



E 

 o 



E 

 o 

 CD 



Soulh Central Nort 



Acanthurus thostegus 

 (convict surgeonfisti) 



South Central North 



Naso lituratus 

 (orangespine unicornfish) 



Mulloidichthys flavolineatus 

 (yellowstriped goatfish) 



Siganus spinus 

 (little spinefoot) 



Epinephelus merra 

 (fioneycomb grouper) 



I I Marine preser\« 

 I I Fished site 



Figure 2 



Mean biomass of five exploited reef fishes in three marine protected areas (MPAs) and 

 three reference sites at Guam, Micronesia. South = Achang Marine Preserve and Cocos 

 Lagoon reference site. Central = Piti Bomb Holes Marine Preserve and Asan Bay refer- 

 ence site. North = Tumon Bay Marine Preserve and Tanguisson reference site. Error 

 bars represent ± 1 standard deviation; n = 32 censuses per site. Letters indicate groups 

 of no significant difference among MPAs and reference sites, according to pairwise 

 comparisons (Tukey's HSD). 



of orangespine unicornfish was significantly higher in 

 the Piti and Tumon MPAs than in adjacent fished sites 

 (Tukey's HSD, P<0.05), but there was no significant dif- 

 ference in biomass between Achang and Cocos Lagoon 

 (Tukey's HSD, P<0.05). Mean biomass of little spinefoot 

 differed significantly between locations (2-way ANOVA, 

 F=15.02, P<0.0001) and was significantly higher in the 

 north and central locations than in the south (Tukey's 

 HSD, P<0.05). Mean biomass of little spinefoot was 

 also significantly higher on protected reefs (P=16.01, 

 P<0.0001) at all locations (Tukey's HSD, P<0.05 for all 

 paired comparisons). 



For carnivorous species, mean biomass of yellowstripe 

 goatfish (Fig. 2) did not differ between locations (2- 

 way ANOVA, P=1.76, P=0.18). There was a nonsignifi- 

 cant higher abundance in MPAs than in reference sites 

 (F=3.30, P=0.076). There was a significant interaction 

 between location and protective status (F=3.32, P<0.05), 

 which occurred because biomass of yellowstripe goatfish 

 was higher in the Tumon MPA than at Tanguisson, but 

 did not differ between MPAs and reference sites at the 

 central and south locations. Mean biomass of honeycomb 

 grouper (Fig. 2) did not differ between locations (2-way 

 ANOVA, P=0.07, P=0.93) but was significantly higher 



in MPAs than at fished sites (F=11.33, P=0.002) at all 

 locations. 



Spillover 



A total of 2674 fishes were tagged; of these 935 (35%) 

 were recaptured (Table 1). The species with the high- 

 est recapture rate (78%) was the honeycomb grouper; a 

 solitary, benthic predator. Mobile, schooling species such 

 as yellowstripe goatfish and little spinefoot had lower 

 recapture rates (25-35%) but were abundant enough to 

 allow relatively large numbers of returns. 



Most recaptured fish did not cross the MPA bound- 

 aries. The overall grand mean spillover (i.e., over all 

 species at all sites) was 5.9% of individuals and 8.0% 

 of biomass from the MPAs. Table 2 shows biomass and 

 mean number of tagged fish remaining within each 

 MPA and fished site, immigrating to each MPA, and 

 emigrating to each fished site. The difference between 

 emigration and immigration is shown in Table 2 as 

 the net flux in biomass and as the net flux in numbers. 

 Spillover (the proportion of tagged fish exported to ad- 

 jacent fished areas minus the proportion of tagged fish 

 imported to the MPA) is also presented as spillover in 



