342 



Fishery Bulletin 92(2), 1994 



depth) show trends from the east to the west, south- 

 west, and south. Not unexpectedly, a mosaic of varia- 

 tion patterns is present in the suite of morphologic 

 characters we assessed. Some, like width of tempo- 

 ral fossa (Fig. 10A), align closely with environmen- 

 tal variables — such as sea surface temperature 

 (July) (Fig. 10B) — subsumed under environmental 

 component I (Fig. 9A). Others, like skull width (at 

 parietals) (Fig. 11A), display patterns similar to 

 those of thermocline depth (winter) (Fig. 11B) and 

 other environmental measures summarized by en- 

 vironmental component II (Fig. 10B). However, the 

 overall, general morphological trend is reflected best 



by projections onto the first canonical variable based 

 on morphologic data (Fig. 4), which has a relatively 

 strong negative correlation with environmental com- 

 ponent I and a weaker positive one with environ- 

 mental component II. 



By adding the more westerly blocks to the analy- 

 sis, environmental-morphological covariation pat- 

 terns that emerged, in some cases, were different 

 from those reported by Schnell et al. (1986). For 

 example, the previous statistically significant corre- 

 lations they found for several morphologic charac- 

 ters with sea current (N., winter) and sea surface 

 temperature (annual variation) were not repeated 



