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



529 



the Tumon Bay preserve for four types of fish; 

 convict surgeonfish (Acanthurus triostegus), 

 spinefeet (Siganus spp., known regionally as 

 rabbitfish), juvenile jacks and trevallies iCaranx 

 spp.), and juvenile goatfishes (Mullidae). Cast 

 nets are sometimes used for convict surgeonfish 

 and spinefeet along the reef margin. 



At each site, visual census and tagging were 

 performed as near to the geometric center of 

 the MPA as possible. The closest distance to 

 the MPA boundary was measured, and census 

 sites on adjacent fished reefs were chosen at 

 this same distance outside the MPA. All tran- 

 sects were surveyed and tagging was conducted 

 in habitats as similar as possible, generally 

 on the seaward edge of the reef flat at a 2-3 

 m depth, where the substrate typically con- 

 sisted of sand, rubble, and scattered patch reefs 

 (mainly Porites spp.), and where seagrass beds 

 are nearer to shore (except in Tumon, where 

 seagrass is scarce). This experimental design 

 served to minimize bias due to among-site vari- 

 ation in distance from the reserve boundary 

 and habitat type. 



Estimation of biomass in MPAs and 

 adjacent fished areas 



Underwater visual census was used to esti- 

 mate biomass of five exploited reef fishes at 

 three MPAs and adjacent fished sites. These 

 included three herbivores: convict surgeonfish 

 {Acanthurus triostegus), orangespine unicorn- 

 fish (Naso lituratus), and little spinefoot iSiga- 

 nus spinus), and two carnivores: yellowstripe 

 goatfish (Mulloidichthys flavolineatus), and hon- 

 eycomb grouper (Epinephelus merra). These five 

 species were chosen because of their great abun- 

 dance and prevalence in the Guam nearshore 

 reef fishery and because they are relatively 

 easy to capture and tag. At each study site, 

 four 50 m X 5 m transects were laid haphazardly. 

 Divers swam each transect slowly, counting all 

 commercially important species and estimating 

 their lengths to the nearest cm. All divers were 

 trained in fish size estimation for one month before this 

 study. Biomass estimates were conducted biweekly from 

 May through August 2003. In total, 32 transects were 

 completed at each MPA and adjacent reference sites (8 

 census days x 4 transects at each site). Biomass was 

 estimated by substituting the length of each fish on the 

 transect into length-weight regressions for that species. 

 The total weight of all fish by species was then used as 

 an estimate of biomass. 



Philippines 



^'.^ 



Papua New Guinea 



■5. . ,C^ 

 Australia — ^  *'  ' 



Philippine Sea 



Asan Bay 

 Piti Bomb Holes y^ ^^ East Agana Bay 



Pacific Ocean 





Cocos Lagoon 



Achang Reef Flat 



Mark-recapture study 



The five study species were collected from three pre- 

 serves and from surrounding areas of reference sites. 

 Within each preserve, six permanent quadrats, each 



Figure 1 



Map of Guam showing the locations of the three marine protected 

 areas (MPAs) and three reference sites. MPAs (hatched areas) are 

 Achang Reef Flat Marine Preserve, Piti Bomb Holes Marine Preserve, 

 and Tumon Bay Marine Preserve. Reference sites are Cocos Lagoon. 

 Asan Bay, and Tanguisson. Inset shows the location of Guam within 

 the Western Pacific region. 



20 X 20 m in planar area, were located at a distance 

 of 800 m from the boundary of the MPA and reference 

 site. The perimeter of each study site was marked at 

 1-m intervals by tying flagging tape to a dead coral. 

 Another six quadrats were marked on fished reefs at 

 the reference site, also at a distance of 800 m from 

 the MPA boundary. This distance was chosen because 

 it allowed the location of the study sites to fall into 

 appropriate and similar habitats at all MPAs and fish 

 sites. This distance also ensured that all tagged fish 

 had an equal distance to move in order to enter or 

 leave the MPA. It also represented sufficient distance 

 from the MPAs to encompass a large area of fished 

 reef. A square area was chosen for the mark-recapture 

 study because it was easier to deploy surround nets 



