244 



Fishery Bulletin 104(2) 



Although the USNM catalog ledger has 

 "cotype" handwritten next to each catalog 

 number (all 19 CAS-SU and USNM speci- 

 mens are labeled as such) and Smithson- 

 ian Institution accession file no. 54484, 

 dated 6 September 1912, lists the mate- 

 rial as "co-types described by Dr. Gilbert" 

 (PalmerM, a description was never pub- 

 lished. 



Why Gilbert, on closer inspection, chose 

 not to formally recognize T. dorothy is 

 probably because of its extreme similar- 

 ity to T. pingeli. Triglops dorothy is easily 

 confused with the latter species, but can 

 be distinguished satisfactorily on the ba- 

 sis of only three meristic characters that 

 overlap slightly between the two species 

 in the Sea of Okhotsk (Tables 1-2): the 

 number of oblique, scaled dermal folds 

 below the lateral line (67-99 in T. dorothy 

 versus 50-87 in T. pingeli), the number 

 of dorsolateral scales (18-31 in T. doro- 

 thy versus 28-37 in T. pingeli). and the 

 number of gill rakers (7-12 in T. dorothy 

 versus 5-9 in T. pingeli). A significant 

 difference in the length of the first infra- 

 orbital pore (Fig. 2A) reflects differences 

 in the overall shape of the snout of the 

 two species. In T. dorothy. the snout is 

 slightly shorter and slightly more ven- 

 trally oriented than in T. pingeli. 



Triglops dorothy is known only from the 

 shallow continental shelf of the southern 

 Sea of Okhotsk between Sakhalin Island 

 and Hokkaido, an area that is thought 

 to have served as a refugia for other spe- 

 cies during the Quaternary glacial maximum (see Noll 

 et al., 2001; Brykov et al., 2003). During this period, 

 glaciers extended to just north of Terpeniya Bay on 

 Sakhalin Island and to central Iturup Island in the 

 southern Kurils, leaving the southern and central part 

 of the Sea of Okhotsk ice-free. This region is thought 

 to have been relatively isolated during this time from 

 the Pacific Ocean and the Sea of Japan because of lower 

 sea levels that connected the southern coastlines from 

 Sakhalin to Hokkaido and the southern Kuril Islands 

 (Kryvolutskaya, 1973; Pietsch et al., 2001; 2003). These 

 land connections may have served to split and isolate 

 populations of benthic marine fishes, such as T. pingeli. 

 providing the opportunity for allopatric speciation of 

 T. dorothy. 



Figure 4 



Distribution of Triglopx dorothy n. sp. (O) and T. pingeli (•) in the 

 Sea of Okhotsk based on material e.xamined. Each symbol may 

 represent more than one capture. 



Acknowledgments 



We thank the many people who made material avail- 

 able, especially W. N. Eschmeyer, T. Iwamoto, and D. 

 Catania (CAS); S. L. Jewett, L. F. Palmer, and J. T. Wil- 

 liams (USNM); and M. Yabe (HUMZ). K. P. Maslenikov 

 (UW) provided curatorial assistance. We are grateful 

 also to two anonymous reviewers who greatly improved 

 the final draft of the paper. A presubmission draft was 

 critically read by J. R. Dunn (UW) and D. E. Stevenson 

 (Alaska Fisheries Science Center, Seattle). The work 

 was supported by the National Science Foundation under 

 contract number DPP74-13988 to the Smithsonian Insti- 

 tution, F. D. Ferrari, principal investigator. 



Literature cited 



' Palmer, L. F. 1991. Personal commun. Smithsonian Institu- 

 tion, U.S. National Museum of Natural History, Division of 

 Fishes, 10th Street and Constitution Ave. NW, Washington, 

 D.C. 20560. 



Andriashev, A. P. 



1949. On species composition and distribution of scul- 

 pins of the genus Triglops Reinh. in the northern seas. 

 Trudy Vsesoyuznogo Gidrobiolgicheskogo Obshchestra, 

 1:194-209. ITranslated from the Russian by L. Penny, 

 United States Bureau of Commercial Fisheries, Ichthyo- 

 logical Laboratory, National Museum of Natural History, 

 Washington, D.C. 20560, translation no. 30, 1965. J 



