156 



Fishery Bulletin 89(1), 199! 



Shark and cobia catches were highest while bottom 

 fishing in the spring near large platforms in shallow 

 water. The MRA with six significant predictor vari- 

 ables accounted for 11.5% of the variance in the shark- 

 cobia factor (Table 5). The negative regression coeffi- 

 cient for water depth and positive regression coefficient 

 for submerged surface area (Table 5) provide evidence 

 that shark and cobia abundances were highest in 

 shallow waters near large structures. Highest cajtches 

 of shark and cobia occurred while bottom fishing in the 

 spring, based on the negative regression coefficient for 

 fishing method and linear date (Table 5). Conflicting 

 results on the presence of electronic gear were found 

 with a positive regression coefficient for presence of 

 graph recorders and a negative regression coefficient 

 for the presence of echo sounders (Table 5). 



Discussion 



Anglers and charterboat operators utilized the entire 

 range of sizes and operational types of platforms 

 available off the coast of Louisiana (single-well caissons, 

 steel template platforms, and mobile semisubmersible 

 drilling platforms), although certain trends in platform 

 size utilization and fishing method were evident. Near- 

 shore fishermen most often fished at the small struc- 

 tures (i.e., caissons) in shallow water (i.e., <10m), while 

 offshore bottom-fishing and trolling fishermen fished 

 near much larger steel template platforms in deeper 

 water (i.e., >30m). Charterboat operators had larger 

 vessels and were able to fish in deeper waters and far- 

 ther offshore than anglers. 



Anglers and charterboat operators caught a total of 

 36,839 fish representing at least 46 different species, 

 providing evidence for the high diversity of fish around 

 the oil and gas platforms. Fishes caught ranged from 

 relatively common and highly desirable species such as 

 spotted seatrout and red snapper to relatively rare 

 fishes such as hake, bearded brotula, and squirrel fish. 

 Highly sought-after gamefish such as tarpon, blue 

 marlin, king mackerel, and yellowfin tuna were also 

 caught. Catches by angling are selective and biased 

 towards larger species because of the hook-and-line 

 gear utilized, and usually only carnivorous species are 

 susceptible to the gear. Consequently species not 

 susceptible to angling were not represented (Grimes 

 et al. 1982). 



The associations of fish identified by the reef fish and 

 pelagic fish PC's were in agreement with fish classifica- 

 tion systems using direct observation around natural 

 and artificial reefs in the Gulf of Mexico by Starck 

 (1968) and Lukens (1981). This confirms that these 

 groupings have an ecological basis, and were not an 

 artifact of the sampling or analysis techniques. 



Factors affecting 



the abundances of fish 



Physical platform variables Generally, the highest 

 abundances of spotted seatrout were found in shallow 

 water (i.e., <10m) around small, non-producing plat- 

 forms such as caissons. These results were expected, 

 as this estuarine-dependent species (Johnson and 

 Seaman 1982) would likely have its highest abundances 

 in shallow water near estuaries. 



Our results suggest that reef fish, Atlantic croaker, 

 and silver/sand seatrout abundances increased with size 

 and complexity of the artificial reef, agreeing with past 

 studies (Turner et al. 1969, Grove and Sonu 1983). 

 However, an optimal artificial reef size occurred for 

 reef fish based on the response surface plot of water 

 depth, volume enclosed, and fish abundance as highest 

 reef-fish abundances occurred at intermediate depths 

 (i.e., 70-100m) near relatively large platforms (i.e., 

 mean volume enclosed 150,000-250,000 m 3 ). The op- 

 timal size range of oil and gas platforms acting as 

 artificial reefs was significantly larger than the optimal 

 artificial reef sizes reported in past studies. This dif- 

 ference could be due to the open construction and lack 

 of interstitial spaces on oil and gas platforms which 

 may not be as efficient at attracting or increasing 

 secondary production of fish as the large rubble or 

 prefabricated artificial reef units on which past esti- 

 mates were calculated. Also, oil and gas platforms ex- 

 tend throughout the entire water column, and because 

 many reef fish are demersal, a large portion of the plat- 

 form may not be suitable reef-fish habitat. Reef-fish 

 abundances were lowest at the largest platforms in 

 extremely deep water, probably because the water 

 depths exceeded the preferred ranges for these species. 



Physical platform variables were not important 

 predictor variables of pelagic fish abundance. Our 

 results are consistent with Wickham et al. (1973) and 

 Grove and Sonu (1983) who concluded that pelagic 

 fishes respond to the visual attraction of artificial reefs 

 and not to the overall size or surface area. 



Bluefish, red drum, cobia, and shark abundances 

 were highest around large, structurally complex plat- 

 forms. These species were probably not trophically 

 dependent on the structures, but were attracted to plat- 

 forms by an optical stimulus as reported by Wickham 

 et al. (1973) for cobia. 



Age of the structure was not a significant factor in 

 explaining fish composition or abundance around oil 

 and gas platforms. Therefore it appears that the 

 species-equilibrium model, which suggests that the 

 number of species present and their abundances in- 

 crease rapidly over a colonization period eventually 

 stabilizing (MacArthur and Wilson 1967), was not ap- 

 plicable, or that the platforms may have been fully 



