176 



Fishery Bulletin 93(1). 1995 



Patchiness-at-age curves for six species (Engraulis 

 mordax and Trachurus symmetricus, Hewitt, 1981; 

 Clupea harengus pallasi, McGurk, 1987; Sardinella 

 brasiliensis and Harengula jaguana, Spach, 1990; 

 and Scomber japonicus, reported here) describe a 

 similar sequence: a high index is observed at the 

 youngest larval ages, a low index is observed at one 

 or two weeks after spawn, and thereafter the index 

 increases suggesting the onset of schooling behavior 

 (Fig. 3). The highest index of patchiness at early lar- 

 val age was observed for S. brasiliensis (P=14.5). This 

 can be attributed to intensive spawning behavior of 

 adult sardine, short incubation time (Matsuura, 

 1983), and fast larval growth (Yoneda, 1987) rela- 

 tive to the other species. The lowest index of patchi- 

 ness was observed for C. harengus pallasi (P=3.5) 

 collected in a small inlet on the west coast of 

 Vancouver Island, British Columbia; McGurk (1987) 

 noted that this may be a reflection-dispersed prey. 

 Houde and Lovdal (1985) reported that fish larvae 

 in Biscayne Bay, Florida, were only slightly more 

 patchy (P=1.3) than their prey, which was abundant 

 and not aggregated (P=1.06). Henri et al. (1985) also 



Time since spawn (days) 



Figure 3 



Patchiness-at-age curves for three species from the California Current (Engraulis 

 mordax, Trachurus symmetricus, and Scomber japonicus), two species from southern 

 Brazil (Sardinella brasiliensis and Harengula jaguana), and one species from British 

 Columbia (Clupea harengus pallasi). 



reported low patchiness values (P=1.63-3.52) for 

 Clupea harengus harengus larvae collected in the St. 

 Lawrence estuary, Quebec. Hewitt (1981) discussed 

 differences in patchiness-at-age curves forE. mordax 

 and T. symmetricus in terms of their prey availabil- 

 ity and foraging strategies. In contrast to T. sym- 

 metricus, E. mordax exhibited an initial high degree 

 of patchiness and slowly dispersed before showing a 

 rapid increase in patchiness at about 18 days of age. 

 Trachurus symmetricus larvae were approximately 

 1/10 as abundant, exhibited lower initial patchiness, 

 and achieved maximum dispersion at an earlier age. 

 E. mordax depend on small, but abundant, prey; they 

 have poorly developed swimming capabilities and can 

 effectively forage only through a small volume of 

 water. In contrast, T. symmetricus depend on large, 

 but rare, prey items; they have well-developed swim- 

 ming capabilities and are able to search through rela- 

 tively large volumes of water. 



In comparison to the four clupeoid species, the in- 

 crease in patchiness was observed to occur at an early 

 age for both S. japonicus and T. symmetricus. Scom- 

 ber japonicus and T. symmetricus larvae share simi- 

 lar morphologies and life his- 

 tory traits. Hunter and 

 Kimbrell (1980) noted that 

 Pacific mackerel larvae may 

 be characterized as having 

 fast growth, rapid swimming 

 abilities, high metabolism, a 

 dependence on increasingly 

 larger prey, and a tendency 

 for cannibalism. Sibling can- 

 nibalism may be an impor- 

 tant survival strategy for 

 mackerel larvae, where 

 larger individuals prey on 

 smaller ones. Grave (1981) 

 reported that by the time 

 Atlantic mackerel, Scomber 

 scombrus, larvae were 12 mm 

 long, 83% of the food items in 

 their diet were other mack- 

 erel larvae. High initial dis- 

 persal, followed by aggrega- 

 tion of similar-sized larvae 

 may be mechanisms for re- 

 ducing sibling cannibalism. 

 Although the patchiness-at- 

 age curves for S. japonicus 

 and T. symmetricus are dis- 

 tinct (Fig. 1), the patchiness- 

 at-size curves are almost co- 

 incident (Fig. 4), suggesting 

 that change in patchiness 



Scomber japonicus 

 Trachurus symmetricus 

 Engraulus mordax 

 Clupea harengus pallasi 

 Sardinella brasiliensis 

 Harengula jaguana 



