FISHERY BULLETIN: VOL. 75, NO. 1 



of larvae from each of these samples were stan- 

 dardized to numbers per 1,000 m 3 of water filtered. 



TAXONOMIC PROBLEMS 



The 287 samples yielded 23,578 fish larvae in 27 

 families and 1 order (Table 2). To date 90 

 taxonomic groups have been identified, 78 at the 

 species level, although 17 of these, primarily in 

 the Cottidae and Stichaeidae, are still only 

 numbered "larval types" 12 which are considered to 

 be identified at the level of distinct species. These 

 larval types have not yet been named because 

 large specimens needed for positive identification 

 were absent from the collections. This is the 

 greatest number of species recorded from a larval 

 fish study in the northeast Pacific which reflects, 

 in part, refinements in larval fish identification as 

 well as the intensity of the sampling effort which 

 yielded many complete developmental series. 

 Many of these larvae, particularly the coastal 

 forms, have not yet been described in detail in the 

 literature. 



While identification of many of the abundant 

 larvae, particularly the pleuronectids and 

 myctophids, has been accomplished with cer- 

 tainty, a few major taxonomic problems remain, 

 most notably with the osmerids and the scor- 

 paenids, primarily Sebastes spp. We have not yet 

 been able to identify the larval osmerids ( <30 mm) 

 to species, of which there are five possibilities: 

 Allosmerus elongatus, Hypomesus pretiosus, 

 Spirinchus starksi, Spirinchus thaleichthys, and 

 Thaleichthys pacificus. Available descriptions 

 (Morris 13 ; Yap-Chiongco 1941; DeLacy and 

 Batts 14 ; Dryfoos 1965; Moulton 1970) are in- 

 adequate to distinguish all five species. We have 

 not even established "larval types" below the 

 family level. 



No attempt was made to separate Sebastes spp., 

 another problem group, into "larval types" 

 (species or species groups) although a few distinct 

 kinds appeared to be present. Samples from Ore- 

 gon waters may contain some 35 species and 



12 The term larval type used in this paper refers to a particular 

 kind of larva which may be distinguished from other larvae on 

 the basis of larval characters but which has not yet been named. 

 The term does not necessarily denote identification to the species 

 level and is not intended to have any taxonomic implications. 



13 Morris, R. Some notes on the early life history of the night 

 surf smelt, Spirinchus starski (Fisk) 1913. Unpubl. manuscr., 

 37 p. 



14 DeLacy, A. C, and B. S. Batts. 1963. A search for racial 

 characteristics in the Columbia River smelt. Res. Fish., Fish. 

 Res. Inst. Univ. Wash. Contrib. 147:30-32. 



identification of the larvae is difficult (Moser 1967, 

 1972; Moser et al. in press). 



One other problem group is the Cyclopteridae. 

 Based on its broad distribution pattern, our 

 Cyclopteridae spp. 1 probably represents a 

 multispecies group, perhaps Liparis spp., but we 

 have not yet been able to subdivide it on the basis 

 of larval characters. 



These identification problems impose limita- 

 tions on analysis of ichthyoplankton data. Caution 

 must be exercised in interpretation of results 

 when multispecies groups constitute a major 

 proportion of larvae taken, such as Sebastes spp. 

 and osmerids off Oregon. 



SAMPLING VARIABILITY 



A series of replicate oblique tows (four day and 

 four night samples at stations 2, 6, 9; two day and 

 two night samples at stations 46, 56, 65, 74) made 

 in June 1971 was examined to assess sampling 

 variability. Species composition of day and night 

 tows at a station was similar, based on common 

 larvae collected and their relative rank abun- 

 dance. Total larvae in night catches exceeded 

 those in day catches at all stations except 65 and 

 74 (Figure 2). Large day-night differences oc- 

 curred at stations 6 and 9. This was primarily due 

 to increased catches of large (>23 mm) osmerid 

 larvae at night (Figure 3), which presumably 

 avoided the net by day or were deeper, although 76 

 to 87% of the water column was sampled in 

 daytime. Even so, osmerids were the most 

 abundant larvae captured in all samples from 

 these two stations. At station 2, the increased 

 night catches were due to an increase in the 

 numbers of large larvae (including osmerids), as 

 well as an increase in the number of species 

 captured (7-10 in daytime vs. 13-14 at night). Both 

 Isopsetta isolepsis (most >16.5 mm) and Micro- 

 gadus proximus (most >29 mm), species com- 

 mon at stations 6 and 9 during day and night, 

 were collected only at night at station 2. At sta- 

 tions 46 and 56, night catches yielded increased 

 numbers of Engraulis mordax (4-10 mm) and 

 Stenobrachius leucopsarus (4-15 mm) while night 

 catches of Sebastes spp. (3-9 mm) were half the 

 daytime numbers (3-12 mm). At station 65, E. 

 mordax (6-10 mm) was again more abundant in 

 night tows while Stenobrachius leucopsarus was 

 much less abundant at night, composing only 10 

 and 34% of the numbers of larvae in the two night- 

 time tows (6-13 mm) but 61 and 54% in the two 



128 



