Nizinski et al.: Separation of Ammodytes amencanus and A dubius in western North Atlantic 



247 



for specimens collected south of the Scotian Shelf (New 

 York-North Carolina, x 30.6; Maine-Massachusetts, 

 X 30.8) while specimens collected further north had 

 mean values of 33.4 (Quebec-Nova Scotia) and 32.1 

 (Labrador). 



The number of pectoral rays varied little geograph- 

 ically (Table 5) and was one of the least successful char- 

 acters for separating the species. Modally, A. ameri- 

 canus had 13 rays (11-15, x 13.2) while A. dubius 

 had one more ray, 14 (12-16, x 14.0). Only 78.0% of 

 the specimens could be separated solely on this char- 

 acter (at a line of separation between 13 and 14 pec- 

 toral rays). 



Gill rakers did not show pronounced geographic 

 trends as was evident for vertebral counts (Table 6). 

 For A. americanus, counts varied between 21 and 28, 

 with means ranging from 23.7 (Labrador) to 24.9 (Mas- 

 sachusetts-Maine). Mean values for .4. duhiu^s clustered 

 around 26 or 27 gill rakers (x 26.6, range 23-31 over- 



all). At a line of separation between 25 and 26 gill 

 rakers. 81.4% of the specimens could be separated into 

 species using only this character. 



Ammodytes dubius showed geographic variation in 

 all the meristic characters examined. Meristic values 

 increased northward with the exception of Labrador. 

 Offshore Labrador collections were sparse; therefore, 

 sample size is only half as large as other regions. Ex- 

 amination of additional material is necessary to deter- 

 mine if meristic values for A. dubius collected off 

 Labrador are consistently lower. 



The meristic characters used in this study separate 

 the western North Atlantic sand lances into low- and 

 high-count species. Each character, however, exhibited 

 some overlap (ranging from approximately 3'^i for 

 plicae counts to 49% for anal rays) and did not separate 

 100% of the specimens. Combinations of characters 

 then became important in verifying identifications, par- 

 ticularly for individuals with intermediate counts. 



