527 



Abstract — Does adult spillover (move- 

 ment out of marine protected areas 

 [MPAs)) offish create a net export of 

 fish biomass from MPAs to adjacent 

 fished reefs? Biomass of five commer- 

 cial reef fish species was estimated 

 by visual census within and outside 

 three MPAs in Guam, Micronesia. 

 For most species and sites, biomass 

 was significantly higher within the 

 MPAs than in adjacent fished sites. 

 Movement of fishes into and out of 

 the MPAs was determined by mark- 

 recapture experiments, in which 

 fishes were tagged both inside and 

 outside of MPAs. Four out of five 

 species studied showed little or no 

 net movement out of MPAs. However, 

 the orangespine surgeonfish (Naso 

 litiircitus) showed a net spillover of 

 biomass from all three MPAs; 21.5'^ 

 of tagged individuals and 29'7r of the 

 tagged biomass emigrated from MPAs. 

 Patterns of spillover were strongly 

 influenced by physical habitat barri- 

 ers, such as channels, headlands, or 

 other topographic features. MPAs that 

 are physically connected by contigu- 

 ous reef structures will likely provide 

 more spillover to adjacent fished sites 

 than those that are separated by habi- 

 tat barriers. This study demonstrates 

 that MPAs can enhance export of fish 

 biomass to fished areas, but spill- 

 over is species-specific and depends 

 on factors such as species size and 

 mobility. 



Spillover of commercially valuable reef fishes 

 from marine protected areas in Guam^ Micronesia 



Mark H. Tupper 



Email address: m.tupperacgiarorg 



The WorldFish Center (address for correspondence) 

 Jalan Batu Maung, Batu Maung 

 11960 Bayan Lepas 

 Penang, Malaysia 



and 



Marine Laboratory 



University of Guam, DOG Station 



Mangilao, Guam 96923 



Manuscript submitted 21 November 

 2005 to the Scientific Editor's Office. 



Manuscript approved for publication 

 2 July 20007 by the Scientific Editor. 



Fish. Bull. 105:527-537 (2007). 



Over the past two decades, marine 

 protected areas (MPAs) have been 

 increasingly adopted as an important 

 fisheries management tool. The pri- 

 mary goals for MPAs are to protect 

 critical habitat and biodiversity, to 

 sustain or enhance fisheries by pre- 

 venting spawning stock collapse, and 

 to provide recruitment to fished areas 

 (Gell and Roberts, 2003; Halpern, 

 2003). Enhancement of fished areas 

 may occur through transport of larvae 

 from spawning stock within the MPA 

 (Gerber et al., 2005) or by a net emi- 

 gration of adult fish to adjacent fished 

 areas — a movement that is commonly 

 known as "spillover" (e.g., Alcala et 

 al., 2005; Abesamis et al., 2006; Goni 

 et al., 2006). 



The role of spillover in determin- 

 ing MPA effectiveness has been ad- 

 dressed in both theoretical modeling 

 studies (e.g., DeMartini, 1993; Kram- 

 er and Chapman, 1999; Gerber et al., 

 2005) and in empirical studies. The 

 latter involve indirect documentation 

 of spillover inferred from increases in 

 fish biomass or catch per unit of ef- 

 fort (CPUE) in adjacent fished areas 

 (e.g., Russ and Alcala, 1996; Roberts 

 et al., 2001; Tupper and Rudd, 2002; 

 Alcala et al., 2005), and direct docu- 

 mentation of fish movement through 

 mark-recapture or sonic tracking ex- 

 periments (e.g.. Chapman and Kram- 

 er, 2000; Meyer et al., 2000; Kaun- 

 da-Arara and Rose, 2004; Meyer and 

 Holland, 2005). There are a number 

 of ways in which movement across 

 MPA boundaries may occur; these 



include random movements of fish 

 during their routine activities (some- 

 times referred to as diffusion), emi- 

 gration in response to density depen- 

 dence (e.g., Tupper and Juanes, 1999; 

 Abesamis and Russ, 2005), directed 

 dispersal due to migration, and onto- 

 genetic habitat shifts (Gerber et al., 

 2005). In order for spillover to effec- 

 tively enhance adjacent fisheries, the 

 net direction of these movements, and 

 the number and size of fishes mov- 

 ing, must result in a net flow of bio- 

 mass out of the MPA. Measuring the 

 movement of biomass into the MPA 

 should therefore be equally important 

 to measuring outward movements. 

 However, few studies have measured 

 immigration, and therefore net spill- 

 over, and those that do address bi- 

 directional movements have reported 

 conflicting results, depending on the 

 species or life history stage, the habi- 

 tat, and the size and placement of the 

 MPA (e.g., Kelly and MacDiarmid, 

 2003; Zeller et al., 2003; Tremain et 

 al., 2004; Goni et al., 2006). 



The degree of spillover from MPAs 

 depends on the rate of fish migration 

 across MPA boundaries (DeMartini, 

 1993; Gerber et al., 2005). Larger or 

 more mobile species with large home 

 ranges may spend too much time out- 

 side of the MPAs to be effectively pro- 

 tected, whereas smaller, more seden- 

 tary species with small home ranges 

 may not cross MPA boundaries in 

 sufficient numbers to enhance adja- 

 cent fisheries by spillover (DeMartini, 

 1993; Tupper and Rudd, 2002). If the 



