386 



Fishery Bulletin 90(2), 1992 



30 1 



---•— fall 



10 



<93 93-183 >183 



Depth range (m) 

 Figure 10 



Mean fork length (cm) of river herring by depth strata from 

 spring, summer, and fall groundfish surveys, 1979-84, con- 

 ducted off the Atlantic coast of Nova Scotia. Vertical bars 

 represent 95% CI. The number offish in each category is ad- 

 jacent to symbols. 



outer Bay of Fundy and central Scotian Shelf coincide 

 with high secondary productivity and an abundance of 

 prey. River herring may move inshore in summer and 

 offshore in winter in order to exploit seasonally avail- 

 able food resources. 



The diurnal pattern of vertical migration by river her- 

 ring accounts for the higher catches and proportion of 

 sets with catches during daylight hours in summer and 

 fall. Alewives and blueback herring that are closer to 

 the bottom during the day are more susceptible to cap- 

 ture in bottom-trawling gear (Neves 1981, Loesch et 

 al. 1982). Diel migrations, involving an upward move- 

 ment at dusk followed by a downward movement at 

 dawn, occur in landlocked adult alewives (Janssen and 

 Brandt 1980), as well as anadromous juvenile (Jessop 

 1990) and adult river herring (Neves 1981). While 

 spring catches did not follow this pattern, the presence 

 of a cold (< 5°C) intermediate water mass over warmer, 

 deeper waters on the Scotian Shelf (Hatchey 1942), 

 where the largest catches occurred, may have re- 

 stricted the extent of vertical migration resulting in 

 more captures at night. Our study indicates that few 

 river herring are captured in areas where bottom 

 temperatures are <5°C during spring; therefore, ver- 

 tical migrations may be confined by a water temper- 

 ature inversion. 



Catch indices by number and weight from spring 

 groundfish surveys (when relative abundance is great- 

 est) both appear to be useful indicators of river her- 

 ring bycatch in foreign and domestic trawl fisheries on 



the eastern and central shelf regions (4VW). Poor cor- 

 relations, obtained when the Bay of Fundy-eastern 

 Gulf of Maine bycatch landings (NAFO area 4X) were 

 included in the analysis, were puzzling. Survey data in- 

 dicate the presence of river herring in this area from 

 spring through fall. Underreporting of river herring 

 bycatches in domestic fishing operations (i.e., bottom 

 trawl, gillnet, purse seine) may explain the poor cor- 

 relations when landings from 4X and 4VW were com- 

 bined. Catches from the central and eastern shelf 

 region (4VW) are largely bycatches from the silver 

 hake fishery. The frequent presence of DFO observers 

 aboard these vessels may reduce the incidence of 

 misreporting. Another possibility is the seasonal im- 

 migration of American-origin river herring into the Bay 

 of Fundy-eastern Gulf of Maine region (Rulifson et al. 

 1987) which would increase the variabOity in the NAFO 

 landings for area 4X. 



The small proportion of fish < 19 cm in summer col- 

 lections, relative to spring and fall collections, may 

 reflect their movement outside the survey area into 

 coastal embayments and estuarine habitats in Maine 

 and the inner Bay of Fundy. These areas serve as im- 

 portant summer feeding areas for river herring (Stone 

 and Daborn 1987) and other anadromous fish species 

 (Haedrich and Hall 1976). Summer resident river her- 

 ring generally leave the inner Bay of Fundy in autumn 

 when secondary production declines (Stone 1985). 



Both large (> 19 cm FL) and small (< 19 cm FL) river 

 herring occurred nearshore from spring through fall, 

 but were widely distributed offshore during spring (i.e., 

 southern Gulf of Maine, Scotian Gulf). Most river her- 

 ring <19cmFL are sexually immature while those 

 >19cmFL are generally mature fish which have 

 spawned previously or are maturing to spawn for the 

 first time. Smaller, immature river herring evidently 

 migrate offshore seasonally as do larger, mature fish. 

 Size-related differences in depth distribution were such 

 that small river herring occurred in shallow regions 

 (<93m) during spring and fall, while larger fish oc- 

 curred in deeper areas (> 93 m) in all seasons. Janssen 

 and Brandt (1980) reported that the nocturnal depth 

 distribution of adult landlocked alewife differed by size- 

 class, with the smaller fish at shallower depths. 



Both Canadian and American marine survey data 

 provide evidence of distinct seasonally and geograph- 

 ically separate aggregations of river herring. Off the 

 Atlantic coast of the United States, the Middle Atlan- 

 tic Bight is an important overwintering area for river 

 herring, while in summer they concentrate further 

 north in the Nantucket Shoals and on Georges Bank 

 (Neves 1981). Aggregations of river herring in spring 

 and fall on the central Scotian Shelf and in the eastern 

 Gulf of Maine-Bay of Fundy suggest that these areas 

 are important overwintering sites off Nova Scotia, 



