FISHERY BULLETIN: VOL. 87, NO. 1 



DISCUSSION 



Net Escapement: Considerations 



In the present study, the extent of bias due to net 

 avoidance is unknown, but our trawling program 

 was designed to minimize the problem as much as 

 possible. All of our samples were taken at night 

 when net avoidance is supposedly mitigated (e.g., 

 Pearcy and Laurs 1966). Observations from submer- 

 sibles suggest that small fishes (<60 to 70 mm) are 

 not as successful in avoiding nets as are larger ones 

 (B. H. Robison pers. commun.""). The myctophid 

 faunas in low latitude ecosystems (upwelling regions 

 excepted) tend towards the small end of the size 

 range (Clarke 1973; Gartner et al. 1987). Also, the 

 detection of nets by mechanical stimuli and the ef- 

 fective mouth area of a net have been suggested to 

 enhance avoidance by midwater fishes (Harrisson 

 1967; Clarke 1973; Pearcy 1980; Stein 1985). Based 

 on these factors, a smaller catch per unit volume 

 would have been predicted for the 1.6 mm mesh net, 

 which was white and had one-half the effective fish- 

 ing area of the dark green 4 mm net. However, at 

 size ranges smaller than 30 mm SL, it outcaught the 

 larger mesh net by a factor of 2.7, whereas among 

 larger fishes, it collected 80% to 97% (x = 90%) 

 of the number of specimens of the large mesh. 



Since the introduction of the Isaacs-Kidd mid- 

 water trawl (Isaacs and Kidd 1953), most midwater 

 fish surveys have used gear with a net mesh size of 

 4 to 6 mm bar length (e.g., Badcock 1970; Hulley 

 1972; Clarke 1973; Gartner et al. 1987; Karnella 

 1987). While such gear is necessary for general 

 surveys, our data suggest that they may be inade- 

 quate for obtaining certain quantitative estimates 

 for small micronekton because of net escapement. 

 It is clear that underestimates of myctophid abun- 

 dance are marked in the 4 mm mesh (Figs. 1, 2). 

 These underestimates may apply only to certain size 

 ranges of the population, as in D. dumerilii or L. 

 guentheri, or may include the entire size spectrum 

 of a species, as in N. valdimae. Similar trends of 

 escapement have been noted among dominant spe- 

 cies of sergestid shrimps examined from the same 

 collections used in the present study (M. E. Flock 

 pers. commun.^). 



Many myctophid species, especially strong vertical 



*B. H. Robison. Monterey Bay Aquarium Research Institute, 160 

 Central Avenue, Pacific Grove, CA 93940, pers. commun. June 

 1988. 



'M. E. Flock, Department of Marine Science, University of South 

 Florida, 140 Seventh Avenue S.E., St. Petersburg, FL 33701, pers. 

 commun. June 1988. 



migrators, are generally muscular and slender and 

 possess very small teeth. They present a relatively 

 small cross section that does not appear to be ef- 

 fectively retained by the large net meshes until some 

 critical threshold of body thickness is reached. 

 Among species with relatively pointed heads, i.e., 

 those whose maximum body dimensions lie behind 

 the head, the ability of large mesh nets to hold in- 

 dividuals is as much a function of the lateral thick- 

 ness as it is of the dorsoventral dimension of the 

 body. Even though dorsoventral thickness may 

 greatly exceed mesh size, lateral measurements 

 must be close to or exceed the mesh bar length in 

 order for the 4 mm mesh net to sample these spe- 

 cies as efficiently as the 1.6 mm mesh net (Table 3, 

 Fig. 3). It appears that until both body axes are equal 

 or greater in size than the mesh diameter, if a 

 "pointed head" fish succeeds in getting its head 

 through the mesh, it can readily escape. In contrast, 

 species with maximum body dimensions on the head 

 (blunt heads) show reduced escapement from the 

 large mesh when the dorsoventral aspect alone 

 reaches a critical threshold, i.e., they are not able 

 to push their head through the mesh. 



Implications of Escapement for 

 Ecological Data 



Collections of mesopelagic fishes from many re- 

 gions are now extensive enough to provide a good 

 representation of species composition and distri- 

 bution, especially with respect to the families Myc- 

 tophidae and Gonostomatidae (Gjosaeter and 

 Kawaguchi 1980; Hulley 1981; Gartner et al. 1987, 

 Karnella 1987). There has been increasing emphasis 

 on quantitative assessment of various aspects of 

 myctophid ecology, such as population dynamics 

 (J. Gjosaeter 1973a, 1981; Clarke 1983b, 1984; Lin- 

 kowski 1985; H. Gjosaeter 1987), trophodynamics 

 (Clarke 1978, 1980; Baird and Hopkins 1981a, b; 

 Hopkins and Baird 1981, 1985) and fishery poten- 

 tials for midwater fish species (Gjosaeter and Kawa- 

 guchi 1980). In virtually all of these studies, data 

 from mesh sizes of 4 mm or greater were used. 



Our findings indicate that escapement among size 

 classes and species smaller than 30 mm SL is pro- 

 nounced and that midwater trawls with mesh of <2 

 mm diameter should be used in order to obtain ac- 

 curate estimates of fishes in this size range. It is not 

 enough to assume that catch efficiencies are propor- 

 tional over all size ranges and that some factor can 

 be applied to catches with larger mesh nets to ac- 

 count for escapement. It is clear that some species 

 do appear to have proportional catch rates between 



218 



