ROGERS ET AL.: ANATOMICAL TRAUMA TO SPONGE-CORAL REEF FISHES 



A 



60 



50 



a. 40 



c 30- 



oj 



O 20^ 



10 

 



Trawl Caught 



i i i i i i 1 1 1 1 



10 20 30 40 50 60 70 80 90 100 



e. 



Ul 



S 



100 

 90 

 80 

 70 

 60 



C 50 



£ 40 



30 



20 



10- 



o- 



Angling Caught 



i i i i i i i 1 1 1 



10 20 30 40 50 60 70 80 90 100 



Bottom Depth (m) 



Figure 1.— Plots of the proportions of red snappers with everted 

 stomachs (PE) captured by (A) trawling and (B) angling as a func- 

 tion of bottom depth (data from Camber 1955; Moseley 1966; 

 Bradley and Bryan 1975; except this study). Ordinates were arc- 

 sine transformed (Snedecor and Cochran 1980). Abscissas are 

 plotted as actual depths or midpoints of ranges. The dashed line 

 (plot ^4) is the least-squares line including data from this study. 

 The only significant relationship was for trawl-caught fishes from 

 the literature (r = 0.90, df = 5, P < 0.01). 



ity of orange filefish were collected during periods 

 of the day when there was very little material in the 

 alimentary tract, which is likely responsible for her- 

 niation/intussesception rates at variance with plane- 

 head filefish and blue angelfish values. Sampling of 

 other species was more equitably distributed over 

 the 24-h period. 



Considerations for Feeding Studies 



Negligible biases in stomach and intestinal con- 

 tents are expected among trawl-caught black sea 

 bass, bank sea bass, tomtate, the three porgy 

 species, sand perch, and vermilion snapper at depths 

 of 37 m. Angling-caught red porgies and vermilion 

 snappers should be equally free of bias-producing 



gut displacements at these depths. However, cau- 

 tion is necessary in analyses of stomach contents for 

 trawl-caught red snappers, Mycteroperca groupers, 

 short bigeyes, and angling-caught black sea bass 

 from 37 m. Stomach content data for angling-caught 

 red snappers, groupers, bank sea bass, and sand 

 perch should also be interpreted with attention to 

 the likelihood of bias. These considerations have 

 been previously acknowledged for angling-caught 

 black sea bass and bank sea bass (Link 1980), trawl 

 and angling-caught red snappers (Stearns 1884; 

 Adams and Kendall 1891; Camber 1955; Moseley 

 1966; Bradley and Bryan 1975), angling-caught red 

 groupers, Epinephelus morio (Moe 1969), and 

 angling and longline-caught blueline tilefish, Caulo- 

 latilus microps (Ross 1982), from southeastern U.S. 

 shelf and slope waters. Moseley (1966) and Link 

 (1980) both stated that partial or full stomach ever- 

 sion renders quantification of consumed prey 

 suspect, particularly with respect to across-depth 

 comparisons (e.g., Godfriaux 1974). Studies of food 

 habits of fishes in the South Atlantic and Gulf of 

 Mexico shelf snapper-grouper complex have either 

 not discussed depth as a diet-determining variable 

 (Camber 1955; Moseley 1966; Moe 1969; Bradley 

 and Bryan 1975; Dixon 1975; Henwood et al. 1978; 

 Ross 1982; Steimle and Ogren 1982) or if depth was 

 considered, dealt with fishes not prone to stomach- 

 eversion bias (Manooch 1977; Grimes 1979; Sed- 

 berry 1985). 



Species and gear-specific considerations should 

 also be made for analyses of daily feeding chron- 

 ologies and rations based on stomach content 

 weights. Fishes with partially or completely everted 

 stomachs should be eliminated from the data set. 

 It is clear that trawl-caught specimens of most 

 species are more suited to such analyses than those 

 caught with angling gear. However, some species 

 cannot be efficiently collected with trawling gear at 

 certain times of day, over certain types of bottom, 

 or indeed at all. Extra angling effort (offsetting 

 eversion rates) and well-designed multigear ap- 

 proaches (including traps and longlines) can be used 

 to complete data sets for such fishes. 



Displacements of the posterior portion of the 

 alimentary tract can also have significant effects on 

 studies of feeding biology. Trawl-caught planehead 

 filefish, blue angelfish, and orange filefish are sub- 

 ject to such bias. Prey position data used to examine 

 the rate of movement and evacuation of material 

 through the gut (e.g., Klumpp and Nichols 1983) will 

 be affected by both herniations and intussusceptions. 

 During herniation, fecal material is either shifted 

 into the protruded portion of the intestine or the 



701 



