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



415 



rity, and harvesting schedules of Georges Bank cod 

 Gadus morhua and haddock Melanogrammus aeglefinus. 

 The growth and mortality dynamics of tropical coral 

 reef fishes, however, may be quite unlike those of higher- 

 latitude species (Munro and Williams, 1985; Longhurst 

 and Pauly, 1987). Further simulations using growth and 

 mortality data of other fishes are needed. 



Observations of traditional practices by marine is- 

 landers (Johannes, 1978) suggest that reserves can 

 augment the SSB, and perhaps the yield of tropical 

 reef fishes. To date, controlled empirical measurements 

 of fishery yields in an area adjacent to a MFR have 

 been published for only one study site (Sumilon Is- 

 land, central Philippines; Alcala and Russ, 1990). The 

 changes in catches that Alcala and Russ (1990) ob- 

 served, however, were based on yield, not yield per 

 recruit (Y/R), over a 1-year period and therefore may 

 not have represented equilibrium conditions. 



The present paper evaluates the effects of perma- 

 nently closed MFRs of different sizes on net changes 

 in SSB and yield for several types of tropical Pacific 

 reef fishes. The author simulated various combinations 

 of fishing mortality and emigration-immigration 

 ("transfer") rates for fishes having different but typical 

 natural growth and mortality schedules. Because tropi- 

 cal reef fishes have higher natural growth and death 

 rates than do temperate zone fishes, the focus is on 

 the relative sensitivity of spawning stock biomass per 

 recruit (SSB/R) and Y/R to MFR size, transfer rates, 

 variations in natural mortality and growth, fishing ef- 

 fort, and age at first capture. Another objective is to 

 compare the management potential of "single large 

 or several small" (SLOSS) (Simberloff, 1988) reserves 

 of equal total area. Finally, the potential of MFRs on 

 island reefs is discussed, with particular reference 

 to reef areas surrounding the island of Oahu. Hawaii. 



Methods 



Parameter values for von Bertalanffy growth rates, 

 derived mortality rates, and maturity and harvesting 

 schedules were chosen to bracket the spectrum of life 

 histories and exploitation characteristics of tropical 

 Pacific reef fishes. At one extreme, values were as- 

 sembled to describe a reef transient (e.g., a jack of the 

 family Carangidae) that is relatively slow-growing but 

 long-lived and large-bodied is likely to travel rapidly 

 over relatively large distances. Such species often sup- 

 port valuable commercial fisheries. Values used for 

 growth and mortality rates of the jack resemble those 



2 Ralston, S., and H. A. Williams. 1988. Age and growth of 

 Lutjanus kasmira, Lethrinus rubrioperculatus, Acanthurus 

 lineatus, and Ctenochaetus striatus from American Samoa. 



of many species of commercially important snappers, 

 groupers, and jacks from the South Pacific and other 

 tropical seas (Munro, 1983; Munro and Williams, 1985). 



A short-lived, fast-growing but small-bodied reef 

 damselfish (family Pomacentridae), with limited move- 

 ments after settlement, was used to represent the op- 

 posite extreme. Such small tropical reef species are 

 collected for the aquarium fish trade. 



In between these two extremes lies a broad con- 

 tinuum of fishes with intermediate longevities, body 

 sizes, and movement rates. These fishes probably rep- 

 resent most species targeted by recreational and 

 artisanal fisheries on tropical reefs. The few data avail- 

 able (Galzin, 1987; Ralston and Williams 2 ; Russ and 

 St. John, 1988; Dalzell, 1989) suggested a range of 

 moderate growth and mortality rates for a number of 

 Pacific parrotfishes (family Scaridae) and surgeonfishes 

 (family Acanthuridae). For convenience, these fishes 

 were labeled the "surgeonfish" type. 



The growth parameters used in this study were based 

 partly on published values for a particular species popu- 

 lation, complemented by data for other Pacific popula- 

 tions of the same species. For the pomacentrid, the au- 

 thor used a Moorean population of Stegastes nigricans 

 (Galzin, 1987). He selected Ctenochaetus striatus as the 

 surgeonfish; the length-weight relation was based on a 

 Moorean population (Galzin, 1987), and a Samoan popu- 

 lation provided the VBGF parameters (Ralston and Wil- 

 liams, 1988). Sudekum et al.'s (1991) data for NWHI 

 Caranx ignobilis were used to represent the jack. Thus 

 the degree to which values were population-specific var- 

 ied among the three fish types. Natural mortality rates 

 iM) were estimated using Pauly 's (1980) multiple re- 

 gression of M on maximum size, growth coefficient, and 

 mean water temperature (Pauly and Ingles, 1981); 25°C 

 was chosen as representative for shallow, tropical Pa- 

 cific waters. All parameter values used are listed by 

 fish type in Tables 1 and 2. 



Modeling closure effects 



Polacheck's (1990) model of the effects of closure size 

 (1-50%), rates offish transfer between closed and ex- 

 ploited areas, and fishing mortality rate on the bio- 

 mass and production of fishes in an adjacent exploited 

 area was used with one small but important differ- 

 ence: the use of a three- rather than a four-sided closed 

 area. (Closures on island reefs usually extend seaward 

 from the shoreline, so fishes can move across upcoast, 

 downcoast, and offshore boundaries only. Other fac- 

 tors being equal, dispersion rates out of and into shore- 



Dep. of Commer., Natl. Mar. Fish. Serv., Southwest Fish. Sci. 

 Cent., P.O. Box 271, La Jolla, CA 92038. Admin. Rep. 

 H-88-18, lip. 



