Orr and Matarese Revison of the genus Lepldopsetta Gill, 1862 



545 



line (Table 6). and ocular-side suborbital pores (Table 7 1. 

 Lepidopsetta polyxystra n. sp. also differed from L. bilin- 

 eata in number of gill rakers of the upper and lower first 

 arch and lower second arch (Table 3). dorsal-fin rays (Table 

 4), ocular-side and blind-side pectoral-fin rays (Table 4), 

 anterior ADB pores (Table 5), blind-side suborbital pores 

 (Table 7), and supraorbital pores (Table 7), and differed 

 from L. mochigarei in number of gill rakers of the upper 

 and lower first arch and lower second arch (Table 3), blind- 

 side pectoral-fin rays (Table 4), anterior and posterior ADB 

 pores (Table 5), and cheek scales (Table 6). 



All three species were found to differ significantly in the 

 means of two morphometric characters with either AN OVA 

 or ANCOVA: interorbital width and ocular-side mandible 

 length (Table 8). Lepidopsetta polyxystra n. sp. also dif- 

 fered from L. bilineata in seven additional characters: 

 head length, blind-side maxilla length, cheek length, body 

 depth, ocular-side pectoral-fin length, caudal-fin length, 

 and caudal peduncle depth; and differed in medians in 

 ocular-side maxilla length, blind-side pectoral-fin length, 

 and body depth. Lepidopsetta polyxystra n. sp. also differed 

 from L. mochigarei in means of two characters: dorsal orbit 

 length and cheek length; and in medians of head length. 



body depth, depth at caudal base, and caudal length. Lep- 

 idopsetta bilineata also differed from L. mochigarei in 

 means of two additional characters: blind-side maxilla 

 length and dorsal orbit length; and differed in medians 

 in head length, ocular-side maxilla length, cheek length, 

 body depth, caudal peduncle depth, and depth at caudal 

 base. No differences in means were found in snout length 

 or ocular-side maxilla length. 



In the standard PCA of all three species together, load- 

 ings of morphometric PCI were all positive and thus exhib- 

 ited a strong size effect (Table 9), accounting for 957^ of 

 morphometric variation. Principal components 2 and 3 

 accounted for 56rf of the remaining morphometric vari- 

 ation. Principal component 2 loadings described a gradi- 

 ent based on interorbital width and dorsal orbit length; 

 PC3 loadings, a gradient on caudal peduncle depth, cheek 

 depth, caudal depth at hypural margins, and body depth. 

 The plot of PC2 versus PC3 revealed extensive overlap in 

 morphometric characters (Fig. 2A). Principal components 

 1 and 2 of the meristics analysis accounted for 46% of 

 the meristic variance; PCI strongly loaded on lateral line 

 pores and scales above and below the lateral line, pre- 

 opercle pores, and suborbital pores, whereas PC2 strongly 



