378 



Fishery Bulletin 90(2), 1992 



Table 1 



Summary of river herring (Alosa spp.) catch (in numbers) and effort for spring, summer, and fall groundfish 

 surveys conducted off Nova Scotia by the Canadian Department of Fisheries and Oceans, 1970-89. Paren- 

 theses enclose percentages of total numbers of sets containing catch. SD= sample standard deviation. 



Season 



Year 



No. of 

 surveys 



No. of 

 sets 



Catch-per-set 



X 



SD 



Total no. 

 of fish 



Sets with 

 catch 



Spring (Feb-Apr) 

 Summer (June-July) 

 Fall (Sept-Dec) 



Total 



1979-89 

 1970-89 

 1978-84 



1970-89 



11 

 20 



7 



38 



1231 



1892 



982 



4105 



6.6 

 2.0 

 1.6 



3.3 



73.85 



18.05 



9.47 



42.53 



8117 

 3703 

 1537 



13357 



268 (21.8) 



214 (11.3) 



120 (12.3) 



602 (14.7) 



predominantly within the water column and provides 

 forage to pelagic fish species (Emerson et al. 1986). 

 On Georges Bank, frontal regions generated by tidal 

 mixing along the northern and southern edges are 

 highly productive due to advection of nutrients from 

 deeper waters on both sides of the bank (Cohen et al. 

 1982). Intrusions of warm, saline slope water occur into 

 the southern Gulf of Maine through the Great South 

 Channel (Mountain et al. 1989) and the Northeast 

 Channel (Ramp and Wright 1979). 



Materials and methods 



Survey design and sampling 



River herring catch and length-frequency data were 

 obtained from 38 bottom-trawl surveys conducted 

 between 1970 and 1989 (Table 1). The main study area 

 (Scotian Shelf, eastern Gulf of Maine, and Bay of 

 Fundy; Fig. 1) was surveyed at least once annually 

 during this period. A single summer (June-July) sur- 

 vey was conducted annually between 1970 and 1977; 

 spring (February-April), summer, and fall (Septem- 

 ber-December) surveys were made from 1978 to 1984; 

 and spring and summer surveys from 1985 to 1989. 

 Changing research requirements shifted spring survey 

 effort to Georges Bank and the eastern Scotian Shelf 

 in 1986, thereby excluding the western shelf region and 

 the Bay of Fundy (Fig. 1). Recent (1987-89) summer 

 surveys also included some coverage of northeastern 

 Georges Bank. 



All surveys used a stratified random design with 

 trawl stations allocated to depth strata in proportion 

 to stratum area and randomly positioned within strata. 

 Stratum depth ranges were 0-92 m (0-50 fm), 93- 183 m 

 (51-lOOfm), and 184-366 m (101-200 fm). Before 1981, 

 summer cruises used a No. 36 Yankee bottom trawl 

 with a 10mm stretched-mesh liner in the cod end; all 

 other cruises used a Western IIA trawl with a 10- or 



20 mm stretched-mesh liner. Tows at each sampling 

 station were for 30 minutes. Halliday and Koeller 

 (1981) and Smith (1988) give further details of Cana- 

 dian groundfish survey methods. 



Total number and weight (kg) of river herring in the 

 catch were recorded for each tow as was bottom-water 

 temperature (°C), tow deployment time (Atlantic Stan- 

 dard Time), latitude, longitude, and bottom depth (m). 

 Fork lengths (FL) were measui-ed (to nearest cm) for 

 all fish in catches of < 250 fish; otherwise, catches were 

 subsampled (no fixed procedure) for length. Ale wife 

 and blueback herring were not differentiated and sex 

 was not determined. 



Catches of river herring in Canadian surveys prob- 

 ably consist mainly of alewives since both species tend 

 to be vertically separated by depth. Blueback herring 

 frequent upper levels; alewives frequent mid-depths 

 (Neves 1981) where they are more available to capture 

 in bottom-trawl gear. All of the river herring catch 

 from 1990 spring and summer groundfish surveys 

 (n 1048) were confirmed by examination to be ale- 

 wives. However, because river herring were not iden- 

 tified to species during the 1970-89 surveys, upon 

 which our analysis is based, we cannot exclude the 

 possibility that the catch included a small proportion 

 of blueback herring captured incidentally when setting 

 or hauling the gear. 



Data analysis 



Catch-per-set was standardized to a tow length of 

 1.75nm with no adjustment for differences between 

 gear types. Only cruises with catches of one or more 

 fish were analyzed; 17 cruises on the eastern Scotian 

 Shelf with no catches were omitted. River herring 

 >33cmFL were excluded from the data set (<0.5% of 

 all fish measured), since they exceed the maximum fork 

 lengths recorded for alewife and blueback herring 

 (Loesch 1987) and are believed to be incorrectly iden- 

 tified American shad Alosa sapidissima. 



