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



535 



indicative of the net value of fish moving from MPAs 

 to fished reefs. 



The results of this study demonstrate that large, 

 mobile herbivores like orangespine unieornfish may be 

 exported from MPAs. The high rates of spillover for 

 this species may result in part from its larger size in 

 relation to the other species. In general larger fish have 

 larger home ranges (Kramer and Chapman, 1999). In 

 this study, the percentage of biomass exported from a 

 given MPA tended to be higher than the percentage 

 of individuals exported, indicating that spillover was 

 primarily accomplished by larger fish. However, some 

 acanthurids are known to be very site attached (e.g. 

 Bell and Kramer, 2000; Meyer and Holland, 2005). 

 For example, the larger congener (N. unicornis) is very 

 site attached, and Meyer and Holland (2005) found 

 little evidence for spillover of this species from a small 

 (0.34 km^) no-take marine reserve in Waikiki, Hawaii. 

 Movements of adult A^. vlamingii across the boundaries 

 of Apo Island marine reserve are rare, but density-de- 

 pendent interactions within the reserve are sufficient 

 to displace smaller fish from the reserve (Abesamis and 

 Russ, 2005). In a separate study, no cross-boundary 

 migration was found for three other acanthurids: Acan- 

 thurus nigricans, Ctenochaetus striatus, and A^. uni- 

 cornis (M. Tupper, unpubl. data). Convict surgeonfish 

 showed notable spillover (14-22% of tagged biomass) 

 at Achang and Piti, but showed a net import of 20% 

 of tagged biomass into Tumon. Similarly, yellowstripe 

 goatfish showed net import to Achang and net spillover 

 from Piti. These species tended to form large, mobile 

 foraging schools at all locations. The spatial variation 

 in movement of these two species may be a function of 

 foraging or spawning movements constrained or modi- 

 fied by natural physical barriers (channels or head- 

 lands) and possibly anthropogenic barriers (e.g., the 

 sewer outfall south of Tanguisson). For example, net 

 inward movement of yellowstripe goatfish to Achang 

 may be related to a spawning aggregation of this spe- 

 cies located in Asgadao Channel, in the center of the 

 Achang MPA (M. Tupper, unpubl. data). Alternatively, 

 the disparity in direction of net movement at different 

 sites may be simply explained by large ranging schools 

 that happened to be tagged inside but recaptured out- 

 side the MPA at one location and vice versa at another. 

 As might be expected, the honeycomb grouper, a small, 

 sedentary, ambush predator, showed very low rates of 

 spillover; no net movement in either direction was found 

 at any of the study sites. 



Movement across MPA boundaries may occur as a 

 result of random dispersal of fish during their routine 

 activities, directed dispersal due to migration or onto- 

 genetic habitat shifts (Gerber et al., 2005), or emigra- 

 tion in response to density dependence. High densities 

 of conspecifics in MPAs may lead to increased juvenile 

 mortality (Goeden, 1979; Tupper and Juanes, 1999), 

 decreased growth (Bene and Tewfik, 2003; Tewfik and 

 Bene, 2003), or increased emigration, or to a combina- 

 tion of all three (Tupper and Juanes, 1999; Abesamis 

 and Russ, 2005). In this study, there was no relation 



between density and spillover of reef fish. This may 

 have been due to fact that the density of fish in Guam's 

 MPAs has not yet reached carrying capacity, i.e., the 

 biomass within the MPAs is not yet representative of 

 virgin, unfished stocks. It should be noted that pooled 

 species and locations were used in the regression analy- 

 sis. More data on individual species and locations would 

 result in a more powerful test. 



In conclusion, rates of adult import or export from 

 MPAs appear to result from a combination of foraging 

 behavior, potential spawning movements, and random 

 daily movements across MPA boundaries. These move- 

 ments were influenced by reef topography. Spillover 

 was highest in areas joined by continuous fringing reef 

 systems and lowest where reefs where separated by a 

 headland barrier. Knowledge of fish movement patterns 

 with respect to reef topography may be useful for choos- 

 ing MPA boundaries in order to maximize the spillover 

 of target species. The herbivorous orangespine unicorn- 

 fish showed the highest rate of spillover from MPAs, 

 which indicates that MPAs have the potential to provide 

 herbivore biomass to adjacent fished areas which may 

 be suffering from algal overgrowth due to fishing of 

 herbivores and from nutrient input due to agricultural 

 activities. However, given the declines in density of 

 exploited fishes at the fished sites since the implementa- 

 tion of the MPAs (Gutierrez^), it is evident that overall 

 spillover has not yet been sufficient to increase fish bio- 

 mass on adjacent reefs. This is not surprising, given the 

 relatively short time since the implementation of these 

 MPAs and the displacement of fishing effort from the 

 MPAs to adjacent fished areas. Although spillover rates 

 of four out of five study species were quite low, adult 

 migration is only one process that may benefit fisher- 

 ies. Further research is needed to determine the role of 

 MPAs in enhancing larval supply and the transport of 

 recruits from Guam's MPAs to adjacent fished areas. 



Acknowledgments 



I am grateful to J. Castro and J. Cruz for their invalu- 

 able assistance in collecting and tagging fishes. I thank 

 G. Davis, C. Leberer, and J. Gutierrez of the Guam Divi- 

 sion of Aquatic and Wildlife Resources for permitting 

 me to collect fishes at the marine preserves. I am also 

 grateful to D. Burdick and S. J. Teoh for assistance with 

 the map shown as Figure 1. Funding for data collection 

 was provided by grants to the University of Guam from 

 the National Oceanic and Atmospheric Administration's 

 Coral Reef Ecosystems Studies Program, the National 

 Oceanic and Atmospheric Administration State and 

 Territorial Coral Reef Management program and from 

 the U.S. Department of Fish and Wildlife. Funding for 

 data analysis and for manuscript writing was provided 

 by the WorldFish Center. 



' Gutierrez, J. 2004. Personal commun. Division of Aquatic 

 and WildUfe Resources, 163 Dairy Road, Mangilao, Guam 

 96910. 



