Buckel et al.; A comparison of biomass harvested by Pomatomus saltatrix with that harvested by fisheries 



783 



population's consumption, but it was only IS'yf of to- 

 tal population consumption in 1995 (Figs. 2 and 3). 

 Similarly, Hartman and Brandt ( 1995b) found that rela- 

 tive prey consumption by bluefish at the population 

 level in Chesapeake Bay varied with age structure. 



The fact that age structure and ontogenetic diet 

 shifts interact provides further justification for ob- 

 taining bluefish diet information by age or size. For 

 example, the impact of the 1995 bluefish population 

 on long-finned squid and butterfish was underesti- 

 mated when pooled diet indices were used compared 

 with diet indices that were partitioned by age or size 

 (Fig. 5). 



Comparison of biomass consumed by bluefish with 

 that harvested by fisheries 



Edwards and Bowman (1979) and Sissenwine et al. 

 (1984) found that total piscivore consumption often 

 exceeded fishery landings on Georges Bank. We have 

 shown that bluefish alone consume an amount of 

 squid and butterfish far exceeding the harvest of 

 these species. How do our findings relate to the man- 

 agement of squid and butterfish? The current man- 

 agement plan for long-finned squid recommends a 

 target yield of 21,000 t (Mid- Atlantic Fishery Man- 

 agement Council^). From calculations presented here 

 bluefish consumed almost five times this amount of 

 long-finned squid. Consistent with our findings, es- 

 timates of natural mortality for butterfish, long- 

 finned squid, and boreal squid are high compared 

 with other species (Anonymous, 1995; Atlantic but- 

 terfish, M=0.80; long-finned squid, monthly M=0. 34; 

 boreal squid, M>1.0). Our analysis does not allow us 

 to determine what fraction of total prey mortality is 

 due to bluefish consumption. Nor do we have esti- 

 mates of impact of other predators on these prey. If 

 we knew that bluefish were the dominant contribu- 

 tor to natural mortality in these prey, then questions 

 regarding allocation could be dealt with more explic- 

 itly. For example, if the goal were to build larger squid 

 stocks, would increased fishing mortality on blue- 

 fish be more or less effective than reductions in fish- 

 ing mortality on squid? Future research should deter- 

 mine the magnitude of predation mortality resulting 

 ft"om bluefish and other predators. Until this informa- 

 tion is available we must assume that predation mor- 

 tality is similar to or below estimates of natural mor- 



5 Mid-Atlantic Fisheries Management Council. 1996. Amend- 

 ment 6 to the fishery management plan and the draft environ- 

 mental assessment for the Atlantic mackerel, squid, and but- 

 terfish fisheries. [Available from Mid-Atlantic Fishery Man- 

 agement Council, Room 2115, Federal Building, 300 South New 

 Street, Dover, DE 19904-16790, 16 p.] 



tality for squid and buttei-fish. The management plan 

 for these prey species assumes a high natural mortal- 

 ity (high predation mortality already taken into ac- 

 count) and fishing harvests are targeted accordingly. 



However, these stock assessments assume that 

 natural mortality is a fixed value. For prey of blue- 

 fish, this may not be true given the persistent changes 

 in bluefish abundance on multiyear to decadal time 

 scales. Natural mortality in squid and butterfish may 

 vary greatly as a function of bluefish abundance. This 

 has important management implications. For ex- 

 ample, if bluefish abundance is high and remains so 

 for extended periods and if this results in natural 

 mortality rates on the prey that exceed the baseline 

 natural mortality rates, then the biological reference 

 points used in squid and butterfish management 

 must be adjusted accordingly. Even if fishery remov- 

 als are low in relation to those for other species, they 

 may still contribute to population collapse by driv- 

 ing the prey population past the replacement level. 

 Much more attention needs to be focused on assess- 

 ment of the dynamics and predatory impact of blue- 

 fish if we wish to manage its prey, not to mention 

 bluefish. Our results illustrate the importance of in- 

 corporating interspecific interactions in the manage- 

 ment of marine fisheries (Sissenwine and Dann, 

 1991; Magnusson, 1995). 



Bluefish may have limited influence on menhaden 

 population dynamics because bluefish consumption 

 of Atlantic menhaden was substantially below the 

 fisheries landings of this species in our calculations. 

 Oviatt (1977) estimated the aggregate demand for 

 menhaden by bluefish in Narragansett Bay. She 

 found that menhaden abundance was sufficient to 

 meet the demands of bluefish even when the men- 

 haden stock was low. Hartman and Brandt (1995b) 

 found that combined predation by bluefish, weakfish, 

 and striped bass on menhaden in Chesapeake Bay was 

 low compared with fisheries harvest of menhaden. 



There are several limitations to this analysis. The 

 estimates of the biomass of prey consumed rely on 

 several other estimated parameters. These include 

 the VPA of bluefish biomass, the allometry of gross 

 growth efficiency as a function of bluefish weight, 

 diet estimates, and assumptions about the spatial 

 and temporal location of the bluefish population. We 

 therefore urge readers to view our estimates as rough 

 approximations of the true values. At the very least, 

 however, our analysis clearly suggests a need to 

 greatly improve our knowledge of the population 

 dynamics and foraging ecology of bluefish. In par- 

 ticular, more accurate data on bluefish abundance, 

 spatial distribution, and diet as a function of blue- 

 fish size, season of the year, and habitat (e.g. estu- 

 ary, coastal shoreline, or offshore) are required. 



