QUAST: MORPHOMETRIC VARIATION ON PACIFIC OCEAN PERCH 



measurements, by sex and region, over which the 

 two methods were compared, the prediction of 

 character sizes in fish of 260 mm SL by nonUnear 

 least squares was close to the mean and within 

 the 957c confidence interval for prediction from 

 transformed data (Fig. 2) and did not change the 

 results. 



Graphical comparison of character measure- 

 ments, related to the standardized fish of 260 mm 

 SL, also disclosed that geographic and sexual 

 variation were frequent: 



Nape (Fig. 2A). — Geographic variation in dis- 

 tance between tip of snout and the dorsal fin 

 was significant. Napes averaged shortest in the 

 Atka-Gulf and Kodiak regions and lengthened 

 in a cline to the Vancouver region. Females 

 averaged smaller napes than males in all re- 

 gions. 



Spinous dorsal-fin length (Fig. 2B). — Geographic 

 variation in length of spinous dorsal fin was 

 significant, and the fin shortened in a cline 

 from the Bering Sea to the Southeastern and 

 Vancouver regions. Because geographic varia- 

 tion in length of the spinous dorsal fin was 

 nearly opposite to geographic variation in the 

 nape, changes in position of the anterior fin 

 insertion probably caused the reciprocal clines. 

 Sexual dimorphism was not important. 



Hind-trunk dorsal (Fig. 2C). — Distance between 

 the spinous dorsal fin and the tail changed sig- 

 nificantly between regions. It averaged largest 

 in specimens from the middle regions (Atka- 

 Gulf and Kodiak) and smallest in the Vancou- 

 ver region. Sexual dimorphism was not impor- 

 tant. 



Hind-trunk ventral (Fig. 2D). — Distance between 

 the second spine of the anal fin and the tail 

 changed significantly between regions. It aver- 

 aged largest in the Southeastern and Vancou- 

 ver regions and shortest in the Atka-Bering re- 

 gion. The data formed a geographic cline 

 opposite to that of the belly measurement. 

 From the Yakutat region westward, females 

 averaged significantly shorter in this measure- 

 ment than males. 



Belly (Fig. 2E). — Distance between the pelvic fins 

 and anal-fin spines varied significantly geo- 

 graphically and decreased in a cline from 

 northwest to southeast (Atka-Bering region to 

 the Southeastern and Vancouver regions). 

 Belly measurements averaged about 1.6 cm 

 smaller in the southeastern extreme of the 

 sampling range than in the northwestern. The 



cline apparently is caused by opposing rela- 

 tional movements of pelvic girdle and anal-fin 

 spines along the body axis because the pelvic 

 insertion and hind-trunk ventral measure- 

 ments decreased from southeast to northwest. 

 Sexual dimorphism was significant, with bel- 

 lies of males averaging about 4 mm smaller 

 than those of females. 



Pelvic insertion (Fig. 2F). — Geographic variation 

 in distance between snout and pelvic fins was 

 significant, and the distance increased from 

 northwest to southeast, from the Atka-Bering 

 region to the Vancouver region. The measure- 

 ment averaged shorter in females than males 

 in all regions, evidence for significant sexual 

 dimorphism. 



Head (Fig. 2G). — Geographic variation in head 

 length was significant, and heads averaged 

 shortest in the Kodiak region and longest in the 

 Vancouver region. Sexual dimorphism was 

 usually significant, and females averaged 

 smaller heads than males. Trends in geo- 

 graphic and sexual variation between the head 

 and pelvic insertion were similar, probably be- 

 cause both measurements include similar re- 

 gions of the head. 



Body -depth pelvic (Fig. 2H). — Geographic varia- 

 tion in body depth at the pelvic fins was signif- 

 icant and formed a broken cline. Deepest bodies 

 occurred at the extremes of the sampling range 

 (Atka-Bering and Vancouver regions), and 

 were shallowest in the Kodiak region. Sexual 

 dimorphism was inconsistent. 



Body-depth anal (Fig. 21). — Geographic variation 

 in body depth at the anal spines was signifi- 

 cant, and depth was shallowest in Kodiak spec- 

 imens and deepest in Atka-Bering specimens. 

 Body depths averaged smaller in females than 

 in males from the Yakutat region to the Atka- 

 Bering region, but the differences may not be 

 significant. 



Caudal peduncle (Fig. 2J). — Geographic varia- 

 tion in depth of caudal peduncle was signifi- 

 cant, but important geographic differences 

 were limited to regions west of Yakutat. Speci- 

 mens from the Kodiak region averaged narrow- 

 est peduncles, and specimens from the Atka- 

 Bering, Southeastern, and Vancouver regions 

 averaged widest peduncles. Caudal peduncles 

 averaged significantly narrower in females 

 than in males except in the Southeastern and 

 Vancouver regions. 



Orbit (Fig. 2K). — Geographic variation in orbit 

 diameter was significant. Diameters were 



669 



