AVOIDANCE OF TOWED NETS BY THE EUPHAUSIID 

 NEMATOSCEL1S MEGALOPS l 



P. H. Wiebe, 2 S. H. Boyd, 2 B. M. Davis, 2 and J. L. Cox :i 



ABSTRACT 



Avoidance of towed nets by the common oceanic euphausiid crustacean Nematoscelis megalops was 

 studied by comparing aspects of its sampling distribution as revealed by day and night catches of 

 two nets of different size, one with a 1 m 2 mouth opening and one with a 10 m- opening. Both nets 

 yield essentially the same pattern in vertical distribution. Paired tows yield a highly significant 

 agreement in nighttime abundance estimates, but do not give comparable daytime estimates. Night 

 catches, especially with the smaller net, exceed day catches, an effect which is interpreted as result- 

 ing from greater avoidance during the day. Comparisons between nets show that neither size net has 

 a superior catch rate, day or night. No particular size group of the species is caught with greater effi- 

 ciency by either net. When X. megalops' center of distribution is shallower, differences between day 

 and night catches can be substantially enhanced. 



Application of Barkley's avoidance theory indicates that the potential advantage of greater mouth 

 area of the larger net is effectively cancelled by individuals reacting to the approach of the net at a 

 greater distance. Other theoretical predictions which depend upon the assumption of increasing 

 escape velocities as a function of body size are not corroborated by the field data. Thus, field popula- 

 tion size-frequency distributions are probably not materially affected by avoidance. 



The evidence suggests that N. megalops uses vision to detect the net approach. Net contrast with 

 the background due to down-welling light during the day and bioluminescence produced in and 

 around the net both day and night appear to be the most likely stimuli. Future efforts to reduce net 

 avoidance by species like N. megalops must focus on reduction of these signals. 



Avoidance of capture by towed nets is a major 

 source of underestimation bias associated with 

 zooplankton abundance measurements (Clutter 

 and Anraku 1968; Wiebe and Holland 1968; 

 Wiebe 1971). This factor is perhaps the most im- 

 portant determinant of the accuracy of abun- 

 dance estimates for some of the larger zooplank- 

 ton species. Patchiness of zooplankton may cause 

 large differences between successive tows taken 

 at a single station (Wiebe 1971; Wiebe et al. 

 1973), but the error induced by this factor is not 

 comparable with avoidance error, since patchi- 

 ness "error" is essentially unbiased. The preci- 

 sion of the estimate of abundance for a particular 

 station location will improve with greater sam- 

 ple numbers, but avoidance error will persist as 

 an underestimation bias. Since patchiness of zoo- 

 plankton exists on scales from the microscale 

 (centimeters to meters) to the mesoscale (hun- 



'Contribution No. 4796 from the Woods Hole Oceanographic 

 Institution, Woods Hole, Mass. 



2 Woods Hole Oceanographic Institution, Woods Hole, MA 

 02543. 



Oceanic Biology Group, Marine Science Institute, Univer- 

 sity of California, Santa Barbara, Santa Barbara, CA 93106. 



Manuscript accepted August 1981. 

 FISHERY BULLETIN: VOL. 80, NO. 1, 1982. 



dreds of kilometers) (Mackas and Boyd 1979; 

 Haury et al. 1978), patchiness itself can be 

 viewed not as a sampling problem, but rather as 

 a reflection of natural distributions. Avoidance 

 and technical problems such as clogging and 

 escapement ( Vannucci 1968) represent sampling 

 biases which tend to obscure our picture of these 

 natural distributions. Although clogging and 

 escapement are still important sources of error, 

 improved net design can, in many instances, 

 eliminate these as major problems (Smith et al. 

 1968). Zooplankton avoidance of nets, at least for 

 some species, remains as an important sampling 

 problem. 



Avoidance is variable, depending upon such 

 factors as time of day; light regime; size, shape, 

 and color of the net; speed of tow; species; sex or 

 developmental stage of the organisms; their 

 physiological state; and absolute density (Flem- 

 inger and Clutter 1965; Isaacs 1965; McGowan 

 and Fraundorf 1966; Brinton 1967; Clutter and 

 Anraku 1968; Laval 1974; Boyd et al. 1978). 

 Almost certainly this diversity of factors is one of 

 the major reasons for a lack of consensus regard- 

 ing the extent or magnitude of avoidance bias for 



75 



