DeMartini: Potential of fishery reserves for managing Pacific coral reef fishes 



425 



growing, moderately vagile reef species such as 

 surgeonfish: at a fishing mortality rate (F c ) of 1.0 (equal 

 to M = 1.0), SSB/R can be increased by about 24% 

 over the no-closure case for an MFR of 10%, if the 

 lower bound of T u (0.10) is used as the fundamental 

 emigration rate. By using the higher bound for T u 

 (0.5) and an F c still at 1.0, SSB/R is increased by 12% 

 for the single 10% closure. 



The overall contribution of multiple, small closures 

 will be less, however, than that of one closure of equal 

 total size, in inverse proportion to the increase in the 

 total perimeter of (hence dispersal from) the multiple 

 closures. (This follows from the general rule that 

 smaller reserves have greater perimeter-to-area ratios 

 than larger reserves of equivalent shape [Schonewald- 

 Cox and Bayless, 1986].) In our example, the total 

 perimeter often 1% closures is about three times (VlO) 

 that of one 10% closure. Hence, for a given fundamen- 

 tal transfer rate, the actual transfer rate will be ap- 

 preciably greater, and the additive contribution of ten 

 1% closures to SSB/R will be several times less, than 

 that of one 10% closure. Multiple 1% closures might 

 translate to an actual emigration rate approaching 0.5 

 for a surgeonfish that is moderately vagile (T ls = 0.1- 

 0.25) relative to an MFR of 10%. 



For the preceding arguments to hold, we must fur- 

 ther assume that the multiple closures will not inter- 

 act spatially. It seems reasonable though that for 

 post-settlement stages of species with low fundamen- 

 tal transfer rates, ten 1% reserves may function inde- 

 pendently if well-spaced, perhaps even on a relatively 

 small (e.g., 100-km perimeter) island such as Oahu, 

 Hawaii. 



For simplicity, SLOSS arguments generally ignore 

 the interactive effects of refuge shape and size. How- 

 ever, habitat geometry (shape as well as size) impor- 

 tantly influences dispersal into and from a habitat, 

 regardless of whether an organism's movements are 

 home-ranging or free-ranging (Stamps et al., 1987). 

 Edge "permeability" (i.e., whether a reserve is sited 

 within a larger area of homogeneous "soft-edged" habi- 

 tat or is a functional island surrounded by discontinu- 

 ous "hard-edged" habitat) also significantly affects dis- 

 persal (Buechner, 1987; Stamps et al., 1987). Small 

 increases in edge permeability have disproportionately 

 large effects on dispersal when boundaries are hard, 

 whereas habitat shape exerts a controlling influence 

 on dispersal when boundaries are soft (Buechner, 1987; 

 Stamps et al., 1987). For shoreline-bounded reef clo- 

 sures, the shape of the MFR is clearly more important 

 to residential reef fishes if the MFR is sited within a 

 larger region of similar habitat than if different habi- 

 tats (e.g., extensive sand channels) are used to provide 

 its physical boundaries. 



Future research 



The overall net effect of the diverse compensatory and 

 depensatory factors that may influence the movement 

 patterns of fishes into and out of MFRs is beyond even 

 semiquantitative appraisal at present. Major advances 

 in our understanding of the function of MFRs await 

 development of techniques that describe the funda- 

 mental transfer rates of fishes (Polacheck, 1990) and 

 that further estimate the changes in movement rates 

 that may occur as densities change over time. The 

 results of my preliminary analyses suggest that future 

 studies should focus on fast-growing, moderately vag- 

 ile species such as many surgeonfishes, rather than 

 reef transients or philopatriots. My preliminary con- 

 clusions should be reevaluated as more data become 

 available for these and other types of reef fishes. 



Management potential 



MFRs have the potential to enhance the biomass and 

 spawning stock of species with rapid growth and mod- 

 erate, fundamental transfer rates (Polacheck, 1990; the 

 present paper). Many factors, however, contribute to 

 whether this potential is likely to be realized in prac- 

 tice: the age schedule of harvest and the magnitude of 

 fishing effort in the non-closed area; whether one ver- 

 sus a few or many MFRs of a given total area are 

 used, thereby promoting progressively larger increases 

 in fundamental transfer rates; and whether compen- 

 satory emigration acts to further inflate vulnerability 

 and deflate SSB/R. Decisions as to the number versus 

 size of MFRs that can be sited, given the total shore- 

 line extent of reef available, are social issues. Size 

 limits and fishing effort (bag limits) also are subject to 

 their own management politics, and MFRs will not 

 make them obsolete. Although the fast growth and 

 moderate movement rates of certain tropical reef fishes 

 predispose them to benefit from refuges, it is unlikely 

 that the establishment of MFRs alone, without comple- 

 mentary regulation of effort and the size composition 

 of catch in non-closed areas, can augment the SSB/R 

 of fishes on heavily exploited island reefs. 



Acknowledgments 



I thank M. Yong and especially D. Tagami for help 

 with programming code. J. Polovina and D. Somerton 

 offered stimulating discussion, and J. Bohnsack, 

 J. Polovina, T. Ragen, C. Roberts, G. Russ, D. Somerton, 

 and an anonymous reviewer provided constructive criti- 

 cisms of draft manuscripts. T Polacheck graciously 

 provided a copy of his computer program and assisted 



