242 



Fishery Bulletin 88(2), 1990 



of the family is completed these names should be 

 used. 



Taxonomic confusion results mainly from the mor- 

 phological similarity and large variability in characters 

 traditionally used to separate and identify the different 

 species ofAmmodytes. Generally, distinguishing char- 

 acters have been limited to meristic ones, especially 

 numbers of vertebrae and dorsal and anal fin rays. 

 Meristic overlap and variability are further complicated 

 by the trend for more northerly populations of both 

 species to have higher counts. Meristic characters show 

 clinal variation, increasing both with latitude and 

 distance offshore (Richards et al. 1963, Scott 1972, 

 Richards 1982). Body depth and maximum total length 

 have also been used to separate species (Richards et al. 

 1963, Winters and Dalley 1988), although this char- 

 acter combination has been shown to vary greatly with 

 stage of maturity and age. In fact, intraspecific varia- 

 tion is gi'eater than interspecific variation in some cases 

 (Scott 1972). Additionally, the phase of the reproduc- 

 tive cycle and the amount of food present in the 

 digestive tract also affect body depth in these fish. 



With numerous hypotheses and species names cir- 

 culating in the literature (see Richards et al. 1963 and 

 Winters and Dalley 1988 for additional reviews), a need 

 to alleviate some of the confusion associated with this 

 genus is obvious. The western North Atlantic sand 

 lances were studied in detail to determine appropriate 

 species definitions, delimit geographic distributions, 

 and describe meristic variation. 



Historical background 



Systematic problems involving Annnodyfes are preva- 

 lent at all levels of taxonomic complexity. Researchers 

 still are not sure of the phylogenetic relationships be- 

 tween genera of sand lances and the systematic place- 

 ment of the Ammodytidae among perciforms. Pietsch 

 and Zabetian (1990), however, believe ammodytids to 

 be trachinoids with the family Ammodytidae the sister 

 group of the Trachinidae plus Uranoscopidae. 



At the alpha taxonomy level, 23 nominal species of 

 the genus Arnmodytes have been described. However, 

 only the following six species have been consistently 

 recognized in the literature: A. americanus DeKay 

 1842 and A. dubius Reinhardt 1838 in the western 

 North Atlantic, A. mariyiiix Raitt 1934 and /I. tohianxs 

 Linnaeus 1758 in the eastern North Atlantic, and A. 

 hexapterus Pallas 1811 and A. personatus Girard 1857 

 in the North Pacific (Reay 1970). However, Reay's 

 (1970) synopsis of valid species names and delimitations 

 of geographic ranges has not always been accepted; 

 synonymies and geographical range adjustments are 

 abundant. For instance, several workers including 



Figure 1 



Abundance (10" metric tons) of mackerel Scomber scombrus, herring 

 Clufn'ii hart'ngu^, and sand lance Arnmodytes dubiun populations. 

 1963-87, in the western North Atlantic (Northeast Fi.sh. Cent., Natl. 

 Mar. Fish. Serv., NOAA, Woods Hole, MA). 



Andriashev (1954), Walters (1955), McAllister (1960), 

 and Richards et al. (1963) have proposed that A. hex- 

 apterus is circumpolar and synonymous with A. ameri- 

 canus and/or /I. marinus. Regarding the sand lances 

 that occur in the extreme northern oceans (particularly 

 Greenland), some investigators suggest that the low- 

 count inshore form should be regarded as A. ameri- 

 canus (Reay 1970). Winters and Dalley (1988) proposed 

 that these northern populations were A. wnrinus. 

 Thus, these two nominal species may be conspecific, 

 but researchers disagree on which name to use for the 

 species in the western North Atlantic. Ammodyfes 

 duhius also occurs off the coast of Greenland but 

 generally is considered to be a distinct species (e.g., 

 Backus 1957, Reay 1970, Winters and Dalley 1988). 

 Additionally, Jensen (1941, 1944) suggested that the 

 broad overlap in characters used to separate Arn- 

 modytes species is so great that there may be only one 

 extremely variable, polymorphic Atlantic species of 

 sand lance. 



Furthermore, clinal variation adds to the taxonomic 

 confusion since some investigators recognize this varia- 

 tion as worthy of species designation while others pro- 

 pose the use of subspecies or only recognize isolated 

 populations. Richards et al. (1963) suggested that clinal 

 variation may be due to effects of environmental fac- 

 tors such as temperature, a phenomenon that has been 

 documented in many other marine fishes. These trends 

 have been reported in all species oi Arnmodytes, again 

 leading to the suggestion of a single heterogeneous 

 species with meristic and morphological differences 

 attributable to environmental variables related to 

 distribution. 



