McKinnon: Feeding habits of Lagenorhynchus obscurus 



575 



Table 6 



Mean estimated total lengths of anchoveta, Engraulis ringens, consumed by dusky dolphins, 

 Lagenorhynchus obscurus, in central Peruvian coastal waters in 1985 and 1986, compared to mean 

 lengths of anchoveta taken in the purse-seine fishery. 



Collection source 



Summer 1985 Winter 1985 Summer 1986 Winter 1986 



Stomachs of dusky dolphins 



Mean estimated total length of 



anchoveta' (SE) 

 Number of otoliths measured 

 Number of stomachs from which otoliths 



were collected 

 Purse-seine fishery 

 Mean total length of anchoveta 2 



' Based upon measurements of otoliths in dusky dolphin stomachs. 



2 Mean total length of anchoveta taken in the fishery, calculated from data in Pauly and Palomares ( 1989); SE's and /Vs 



unavailable. 

 ' Means with different letters in the superscript differed significantly (Sidak adjusted (-tests; (>8.41, P<0.001 in all 



cases) while those with the same letter did not U=1.42, P>0.5). 



energy, usually considered the best measure of rela- 

 tive prey importance (La vigneetal., 1982). Few data 

 are available on feeding habits for other regions, but 

 in Argentina dusky dolphins also fed on a species of 

 anchovy (Wiirsig and Wiirsig, 1980). 



There were no consistent seasonal patterns in prey 

 consumption in the present study. Rather, anchoveta 

 was the most important prey species in both seasons 

 of both years, probably owing, in part, to its rela- 

 tively high abundance throughout the study period 

 (Pauly and Palomares, 1989). Consumption of T. 

 symmetricus was more variable and opportunistic. 

 In 1985 T. symmetricus was almost absent from stom- 

 ach samples, but it was a major prey item in 1986 

 when unusually large numbers of juveniles, similar 

 in size to other important dusky dolphin prey, were 

 observed in the coastal waters of central and north- 

 ern Peru (IMARPE et al., 1986). Other species, par- 

 ticularly L. gahi, varied greatly among collection 

 periods in their importance as prey, but did not ex- 

 hibit consistent seasonal or annual patterns. 



In examining the effects of season, year, and other 

 variables on diet, only frequency of occurrence was 

 analyzed statistically because analyses of other mea- 

 sures of prey importance involve excessive violations 

 of the assumptions underlying most statistical tests 

 (Recchia and Read, 1989). Percent-weight estimates 

 were used for qualitative comparisons among collec- 

 tion periods and reproductive classes, rather than 

 percent gross energy values, because the latter were 

 very gross approximations. 



Anchoveta was the main prey of all reproductive 

 classes of dusky dolphins. In contrast, in the eastern 

 tropical Pacific most spotted dolphins, Stcnella 

 attenuate, eat mainly ommastrephid squids whereas 



lactating females eat principally fish (Bernard and 

 Hohn, 1989). The greater energy and water require- 

 ments of lactating females may force them to feed on 

 fish, which contain more energy and water per unit 

 weight than squid or, alternatively, the presence of a 

 calf may prevent females from feeding at the depths 

 at which squid occur (Bernard and Hohn. 1989). In 

 Peru, lactating dusky dolphins were apparently able 

 to satisfy their energy and water requirements, as 

 did other females, males and juveniles, by feeding 

 on the abundant, high energy anchoveta ( Lam, 1968 ). 



Like the fishermen, fur seals, and seabirds of the 

 Peruvian coast, dusky dolphins were somewhat op- 

 portunistic in their feeding in respect to the sizes of 

 the anchoveta they preyed upon, taking more of the 

 more abundant size classes (Muck and Pauly, 1987; 

 Majluf, 1989). Anchoveta consumed by dusky dol- 

 phins were consistently 1.4-2.0 cm smaller, however, 

 than those taken in the fishery, perhaps because of 

 the fishery's bias towards larger anchoveta 

 (Palomares et al., 1987). Alternatively, the slight 

 degradation present in some of the otoliths from 

 which anchoveta sizes were estimated may have re- 

 sulted in underestimation of total lengths (Recchia 

 and Read, 1989). The observed discrepancy may also 

 be due to variation in the relationship between otolith 

 size and body size, which can differ among years be- 

 cause of variation ingrowth rates offish (McKinnon, 

 1988; Reznick et al., 1989; Secor and Dean, 1989; 

 Campana, 1990). 



Relative consumption of fishes with small otoliths, 

 such as anchoveta, may have been underestimated. 

 Large otoliths and squid beaks are less easily dis- 

 solved by stomach acids than are smaller otoliths 

 (Hawes, 1983; Bigg and Fawcett, 1985; da Silva and 



