VALENTINE, SOULE, and SAMOLLOW: ASYMMETRY ANALYSIS IN FISHES 



Table 7. — CK^ values and character means (x^^j) of grunion by locality. 



Table 8. — CV^ values and character means (-f^^^) of barred surfperch by locality. 



closely approximate the previously presented 

 asymmetry trends for both barred sand bass 

 and grunion. The two oldest collections, Bahia 

 Sebastian Vizcaino and San Simeon, uniformly 

 have low CV^ values. The other relatively old 

 collection, Santa Rosa Island, has moderately 

 high pectoral fin ray asymmetry. The majority 

 of specimens in this collection (37 of 56) are 

 small, ranging between 46 and 77 mm. Barred 

 surfperch, at parturition, are between 42 and 

 53 mm (Carlisle et al., 1960). A possible expla- 

 nation for the higher Santa Rosa value is that 

 these small individuals have not yet been 

 exposed to prolonged stabilizing selection. The 

 only other young barred surfperch available for 

 comparison are 18 from Bahia Sebastian 

 Vizcaino (50.1-77.5 mm). Of these 18 speci- 

 mens, 5 are asymmetric for pectoral fin rays. 

 For our Santa Rosa Island specimens 11 of 

 37 are asymmetric. Differences are not statis- 

 tically significant. 



To examine the association of asymmetry and 

 size we grouped our three oldest collections 

 (Bahia Sebastian Vizcaino, Santa Rosa Island, 

 San Simeon) and analyzed these collections 



by size class. Table 9 shows the results of this 

 analysis. Neither gill raker nor pectoral fin 

 ray asymmetry appear to increase with size 

 (age). In Table 8. Figure 3, gill raker data 

 points from Bahia Sebastian Vizcaino, Santa 

 Rosa Island, and San Simeon are connected 

 (solid line) because we feel that these three 

 populations probably represent what might be 

 termed "unperturbed" populations. All three 

 collections are old, all were taken at least 100 

 miles from Los Angeles and all are probably 

 little exploited by other than natural predators. 

 Also in Figure 3, all relatively recent col- 

 lections (San Quintin, 1970-71, Belmont Shore, 

 1969-70, and Carpinteria, 1970) are connected 

 by a dashed line. The extremely elevated CV^ 

 value for the Belmont Shore collection of 1969-70 

 should be noted, as well as the high value at 

 Carpinteria. 



DISCUSSION 



In the discussion of these results we find it 

 necessary to distinguish between two levels of 

 causation. The first is the origin of the asym- 



363 



