342 



Fishery Bulletin 102(2) 



where the species is the target of commercial fisheries in 

 the Gulf of Alaska. During trawl surveys, it is most com- 

 monly associated with S. alutus and S. polyspinis, and 

 at greater depths with S. aleutianus (rougheye rockfish) 

 throughout its range in Alaskan waters (Reuter, 1999; 

 Ackley and Heifetz, 2001). 



Females and males captured during summer ( May- July ) 

 trawl surveys ranged widely in maturity state. Occasional 

 ripe females were observed, although most females were 

 maturing (Orr, personal observ. ). A high percentage of 

 females caught in trawl surveys during early April off 

 southeast Alaska were releasing larvae, indicating that 

 parturition occurs in the spring (Lunsford 7 ). During July 

 in shallower waters (ca. 40 m) of southeast Alaska, all S. 

 variabilis collected were immature. 



Etymology 



The specific name variabilis is presumed to be a reference 

 by Pallas (1814) to the wide range of body color in the 

 species. 



Remarks 



Pallas ( 1814 ) described Perca variabilis from at least three 

 specimens probably collected by Merck during the 1786-94 

 Billings expedition to the Russian Far East, including the 

 Aleutian Islands, eastern Bering Sea, and northern Gulf 

 of Alaska (Schmidt, 1950; Svetovidov, 1978, 1981; Pierce, 

 1990). One specimen was more completely described and 

 used by him to obtain a set of counts and measurements. 

 The other specimens were used to describe variation in 

 the species, as in the following excerpt translated by the 

 authors from the Latin text of Pallas (1814): "Body colored 

 according to life and sex, varied, sometimes dark blue, 

 belly white, fins blackish; female red below; those older 

 wholly red or even purplish...." Pallas (1814) ultimately 

 based the name P. variabilis on the supposed variability 

 in color in this species. 



Jordan and Evermann (1898) examined an individual 

 from the Pallas collection, recognized by him as the "sum- 

 mer variety" of P. variabilis (ZMB 8145). They identified 

 this "summer variety" as Sebastes aleutianus Jordan and 

 Evermann (Jordan, 1884, 1885; Jordan and Evermann, 

 1898), a species easily distinguished from both S. varia- 

 bilis and S. ciliatus by its full complement of eight pairs 

 of strong cranial spines. These specimens have since been 

 lost, probably during the destruction of the Berlin Zoo- 

 logical Museum during World War II (Paepke and Fricke, 

 1992). Although Jordan and Gilbert (1883) wrote that 

 S. proriger was also confounded with S. ciliatus, Jordan 

 (1885) and Jordan and Evermann (1898) corrected this 

 statement, noting that only S. ciliatus and S. aleutianus 

 were included within the material described as E. ciliatus 

 and P. variabilis by Tilesius and Pallas. 



7 Lunsford, C. 2002. Personal commun. National Marine 

 Fisheries Service, Aukc Bay Laboratory, Alaska Fisheries 

 Science Center, 11305 Glacier Highway, Juneau, AK 99801- 

 8626. 



Although MNHN 8670 (Fig. 2A) is from the Pallas col- 

 lection (Svetovidov, 1978, 1981), it is apparently not the 

 specimen used for the complete description. In his original 

 account, Pallas (1814) listed the following meristic data 

 (modified to standard notation): D XIII, 15; A III, 7; PI 17 

 (8 simple). Although the dorsal-fin ray count is identical 

 with that of the MNHN 8670 specimen, both anal- and 

 pectoral-fin ray counts differ. All elements are well pre- 

 served and easily counted. 



Comparisons of proportions are more difficult to inter- 

 pret because measurements had not been standardized 

 at the time of the original description. However, of those 

 measurements that can be readily compared, the following 

 significant differences were found, providing additional 

 evidence that this individual was not the specimen used 

 for the primary description: total length (391.6 mm in 

 Pallas [1814; "longitudo majoris speciminis"] vs. 414.0 mm 

 taken from MNHN 8670), head length (101.6 mm ["capi- 

 tis a summa maxilla ad operculum angulum"] vs. either 

 96.6 mm [standard head length] or 110.2 mm [head length 

 to tip of lower jaw]). The specimen used by Pallas for the 

 detailed meristics and morphometries is presumed lost 

 (Svetovidov, 1978). 



Ayres ( 1854) misidentified S. melanops from the vicinity 

 of San Francisco Bay as S. variabilis. Giinther (I860) and 

 Ayres (1862, 1863) placed S. variabilis of Ayres into the 

 synonymy of S. melanops. 



Comparisons 



Sebastes variabilis is most similar to S. ciliatus; the latter 

 is distinguished by its uniformly dark-blue to black color. 

 Sebastes ciliatus is invariably dark at the base of the anal 

 fin and on the lower pectoral rays, areas of lighter color in 

 S. variabilis even in those individuals that have an overall 

 dark body. The peritoneum of S. ciliatus is always jet black, 

 unlike the usual gray peritoneum of S. variabilis, which 

 however may often be dark or occasionally jet black. In 

 combination with these color differences, a low count of 

 39-42 lateral-line pores will distinguish S. ciliatus from 

 S. variabilis, although the total range of counts overlaps 

 considerably. 



The extrinsic morphological features of the swimblad- 

 der of both S. ciliatus and S. variabilis are of type I (a-z) 

 of Hallacher (1974) in which the anterior muscle mass 

 originates from the occipital region of the neurocranium. 

 attaches to the pectoral girdle near the insertion of Baude- 

 lot's ligament, passes between the epineural and pleural 

 ribs of vertebrae 3 and 4, passes ventral to the pleural rib 

 of vertebrae 5, and continues posteriorly as three tendons 

 that insert on the pleural ribs of vertebrae 8, 9, and 10. 

 In S. ciliatus and not S. variabilis, the anterior striated 

 muscle mass is separated into two sections by a thin fascia, 

 similar to the condition reported in S. paucispinis alone 

 among species of Sebastes (Hallacher, 1974). The morpho- 

 logical features of the complex differ significantly in S. 

 paucispinis, however, in that the striated muscle does not 

 attach to the pectoral girdle but bypasses it to insert by a 

 single tendon into the posterior portion of the swimblad- 

 der. Only five specimens each of S. ciliatus and S. varia- 



