NET FEEDING IN MESOPELAGIC FISHES 



Thomas L. Hopkins and Ronald C. Baird' 



ABSTRACT 



In an investigation of net feeding, 11 species of fish (5 gonostomatids, 6 myctophids) captured in a 

 double-net Tucker trawl were examined. Stomach contents of fish retained by a coarse mesh "fish- 

 catcher" in one net were compared to contents of fish which had accumulated with plankton in the cod 

 end of the adjacent net. Out of 19 species-collection pairs (700 fish), there were significant (P<0.05) 

 differences in number of prey items in stomachs of only three species in five collections. Two pairs, 

 fish from the cod end and in three pairs, those from the "fish-catcher," contained significantly more prey 

 than fish from the adjacent trawl. There were little or no significant differences between trawls in 

 number of fish scales, prey diversity, or prey"size. These results suggest that literature data on diet of 

 mesopelagic fishes is not heavily biased from net feeding and that existing collections can be used for 

 feeding investigations. 



Diet studies of mesopelagic fishes taken from 

 plankton net cod ends of mid-water trawls could be 

 seriously biased if fish feed extensively in the cod 

 end. Indirect evidence for net feeding in pelagic 

 shrimp has been presented by Judkins and 

 Fleminger (1972) and in myctophid fish by Ander- 

 son (1967). DeWitt and Hopkins (in press) also 

 suggest the possibility of net feeding in 

 Pleuragramma antarcticnm, a mid-water no- 

 tothenioid fish. The problem of net feeding, 

 though recognized by the above authors and others 

 (e.g., Holton 1969; Collard 1970; Hopkins and Baird 

 1973), is largely unresolved; consequently the 

 validity of published data on the diet of 

 mesopelagic fishes is questionable. The present 

 study was initiated to estimate the nature and 

 degree of net feeding in the cod end of plankton 

 net trawls by mid-water fishes to better judge the 

 reliability of published information on diets of 

 these fishes, and to determine if existing collec- 

 tions and present methods of collecting are 

 adequate for trophic studies. 



METHODS 



Most of the material examined was collected 

 from the eastern Gulf of Mexico with a double 

 (side-by-side) closing Tucker trawl (Figure 1). One 

 side of the trawl had an unmodified plankton net 

 at the cod end. The mouth of the cod end plankton 

 net of the adjacent trawl was fitted with a coarse 

 mesh (1.1 cm stretched) conical "fish-catcher." In 



principal, the conventional trawl allowed the pas- 

 sage of fish into the cod end where plankton was 

 concentrated; in the adjacent trawl fish were 

 prevented from accumulating with plankton in the 

 cod end by the fish-catcher. The body of the trawl 

 was constructed of 1.1-cm stretched, knotless 

 mesh. The cod ends for most collections were 333- 

 jum mesh, 0.5-m diameter plankton nets. In collec- 

 tions subsequent to tow 152 (see Table 1), l,050-ju,m 

 mesh nets were substituted for the finer 333-ju.m 

 mesh cod ends to improve the internal flow 

 characteristics of the trawl. Other details of trawl 

 design are in Hopkins et al. (1973). We have also 

 included data from tow 98, a Caribbean sample, 

 made with a single-net closing Tucker trawl 

 (Hopkins et al. 1973). Fish gilled in the body of the 

 trawl in this tow were compared with specimens 

 from the cod end. Trawl hauls represented discrete 

 depth samples which ranged from horizontal tows 

 (±10 m) to stepped oblique tows which sampled 

 over a specified segment of the water column. 



Fish were preserved in 10^ Formalin- and sub- 

 sequently transferred to 40% isopropyl alcohol. 

 Specimens selected for analysis (370 from cod end; 

 332 from fish-catcher) were measured to the 

 nearest millimeter (standard length, SL) prior to 

 stomach removal. Contents of the pigmented dis- 

 tensible region posterior to the esophagus and 

 anterior to the intestine were identified to genus 

 when possible, measured, and counted (see 

 Hopkins and Baird 1973; Baird et al. 1975). Prey 



'Department of Marine Science, University of South Florida, 

 St. Petersburg, FL 33701. 



Manuscript accepted February 197.'5. 

 FISHERY BL'LLETIN: VOL! 7.3, NO. 4, 197.5. 



908 



-Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



