ALLEN and DeMARTINl: NEARSHORE DISTRIBUTION AND ABUNDANCE OF PELAGIC FISHES 



sible explanation for the virtual absence of Sardc 

 chiliensis from night collections is that they may 

 migrate offshore of 30 m depth before or at dusk on a 

 daily basis. Net avoidance can be effectively discount- 

 ed, since this fast swimming scombrid can pre- 

 sumably see the net better during the day than at 

 night. 



Although location differences and temporal 

 changes were evident for some species within this 

 assemblage, the dominant pattern shown by the most 

 abundant species was one of a general dispersal 

 offshore at night from nearshore diurnal schools. Diel 

 interactions with depth were found for total in- 

 dividuals, total individuals minus Engraulis mordax, 

 species counts, numbers of E. mordax, and numbers 

 of Seriphus politus. Various diel and/or depth effects 

 were also found for other taxa, including Peprilus 

 simillimus, Genyonemus lineotus, atherinid spp., and 

 Scomber japonicus. These results plus the significant 

 correlations between species abundances and time 

 of collection and depth underscore the general im- 

 portance of diel and depth factors to the abun- 

 dance and distributions of fishes in this 

 assemblage. 



Greater net avoidance under conditions of increas- 

 ing water clarities, such as occur at greater distances 

 offshore, potentially confounds our evaluation of 

 depth (onshore/offshore) patterns. Correlations be- 

 tween the CPUE of major taxa and water clarity indi- 

 cate persistent, negative relationships (Table 5). 

 Variations in water clarity, however, were never ob- 

 served to explain > 167< of the variance in catch; and 

 among the major taxa analyzed in detail for CPUE- 

 water clarity relations, mean CPUE differed by more 

 than a factor of 10 to over four orders of magnitude 

 between the 5-11 mand 18-27 m depth blocks during 

 the day. For this reason, we strongly feel that the 

 potentially greater net avoidance in clearer waters 

 farther offshore is insufficient to explain the ob- 

 served daytime onshore/offshore stratification of 

 these fishes. 



Net avoidance also is a possible complicating factor 

 to our interpretation of day-night differences in catch 

 (i.e., higher night catches). Our test results in fact in- 

 dicate an approximately twofold greater average 

 catch efficiency for Seriphus politus at night versus 

 during the day. However, CPUE of S. politus differed 

 by a factor of four and by over three orders of 

 magnitude between diel periods at the 12-16 m and 

 18-27 m depths, respectively. As above for the poten- 

 tial confounding of daytime depth patterns, we feel 

 that diel differences in catch efficiencies alone can- 

 not explain the marked diel patterns that we have 

 observed. 



Dispersal of schooling (especially pelagic) fishes 

 under low light conditions has been noted repeatedly 

 in the literature (Shaw 1961; Loukashkin and Grant 

 1965; Woodhead 1966; Hobson 1968). Four possible 

 hypotheses (or a combination thereof) can be pro- 

 posed to explain the phenomenon of nocturnal 

 dispersal. 



1. Schools disperse because light levels are insuf- 

 ficient for fish to maintain visual contact. Vision is an 

 important factor in the maintenance of a polarized 

 state and parallel swimming in schooling fishes 

 (Shaw 1978). Some fishes can maintain polarized 

 schools in light intensities as low as 0.5-0.01 foot- 

 candle (Shaw 1961), but disperse in total darkness 

 (Shaw 1961; Loukashkin and Grant 1959, 1965). 

 Some species have the ability to form polarized 

 schools by the light of the full moon alone (Shaw 

 1961, 1978). The acousticolateralis system may also 

 play an important role in polarization and spacing 

 within schools (Shaw 1978) and is not dependent on 

 light levels. Highly sensitive visual and acoustic sys- 

 tems may allow nearshore pelagic fishes to school 

 even under minimal light intensities. If the visual and 

 acoustic systems of these fishes prove to be sufficient- 

 ly sensitive, an explanation for the observed noctur- 

 nal dispersal probably lies elsewhere. 



2. Schools disperse because predation pressure is 

 less intense at night. Hobson (1978) has argued that 

 the threat from predators is a major force behind 

 aggregating (i.e., schooling) behavior in fishes. At low 

 light intensities pressure from visual predators such 

 as those found in this nearshore habitat should be 

 less. However, for this hypothesis to be valid there 

 must be some disadvantage to schooling at night. 

 Theoretically (Eggers 1976) and empirically (Koslow 

 1981), schooling occurs at the expense of prey con- 

 sumption. With the threat from predators di- 

 minished at night, dispersal of fishes within a school 

 may allow greater food consumption by lessening 

 visual-field overlap (Eggers 1976). Dispersal under 

 these conditions, however, would only be advan- 

 tageous if the species feeds at night (see hypo- 

 thesis 4 below). 



3. Schools disperse offshore at night to facilitate 

 reproduction among members. DeMartini and Foun- 

 tain (1981) presented evidence for dusk spawning in 

 Seriphus politus during March-August along the 

 same stretch of southern California coastline. 

 Crepuscular spawning probably helps conceal adults 

 and planktonic eggs from visual predators. Unbal- 

 anced, day-night sex ratios at shallow depths 6 and egg 



"DeMartini, E. E., and L. G. Allen. Diel and seasonal shifts in the 



583 



