688 



Fishery Bulletin 88(4), 1990 



the Gulf of Maine (CeTAP 1982, Watkins and Schevill 

 1982, Kraus 1985, Schevill et al. 1986, Weinrich et al. 

 1986, Hamilton and Mayo 1990). Both of these species 

 have been characterized principally as "skim feeders" 

 (Nemoto 1959, Kawamura 1974, Nemoto and Kaw^a- 

 mura 1977) which subsist primarily on dense swarms 

 of calanoid copepods, notably Calanusfinmarchicus in 

 the North Atlantic (Mitchefl 1975a, 1975b; Jonsgard 

 and Darling 1977; Mitchell and Chapman 1977; Winn 

 et al. 1986; Wishner et al. 1988; Mayo and Marx 1990), 

 in preference to schooling fish (Watkins and Schevill 

 1979). 



This similarity in prey preference led Mitchell (1975a) 

 to hypothesize that competition between these two 

 whale species might have adversely affected the recov- 

 ery of the right whale in the western North Atlantic. 

 However, sandlance are also known to play an impor- 

 tant trophic role as a major predator of copepods, 

 especially C. finmarchicus (Monteleone and Peterson 

 1986, Meyer et al. 1979). Given their abundance and 

 prey selectivity towards copepods, Kenney et al. (1986) 

 proposed that sandlance represented a more significant 

 competitor to right whales for copepodite prey than sei 

 whales (as suggested by Mitchell 1975a). 



The existence of some degree of intra- and inter- 

 specific competition between right/sei whales and sand- 

 lance (ecologically similar species) for a common prey 

 is generally accepted, but is difficult to demonstrate. 

 An overlap in preferred prey is at best a qualitative 

 phenomenon, and offers only "soft corroboration" 

 (from Strong et al. 1984) as evidence for competition 

 between these species. Another approach (used by 

 Smith 1968, 1970; Eck and Wells 1987), which we have 

 adopted for the purpose of this study, is to look for a 

 demonstrable shift in the ecological balance between 

 competing species, or the complete replacement of one 

 competing species by another. In this paper we ex- 

 amine the ecological relationships between sandlance 

 and two of its major predators, the humpback and fin 

 whale; between sandlance and its major prey, the 

 copepod Calanusfinmarchicus; and between sandlance 

 and two potential competitors for C. finmarchicus in 

 this region, the right and sei whale. 



Methods 



Study area 



In this paper we have focused on Stellwagen Bank (Fig. 

 1), a glacial deposit of sand and gravel in the southwest 

 Gulf of Maine that is associated with high levels of 

 biological productivity and a rapidly developing com- 

 mercial whalewatching industry. The margins of the 

 bank are defined by the 40-m isobath with depths as 



low as 18 m. Since the early 1980s whalewatching 

 vessels have provided intensive survey coverage of 

 cetaceans in the area from June through September 

 of each year. 



Collection and treatment of data 



Fisheries data Data on the abundance of sandlance 

 used here were collected on Stellwagen Bank and im- 

 mediately adjacent waters by National Marine Fish- 

 eries Service, Northeast Fisheries Center (NMFS/ 

 NEFC) biologists and technicians during standard- 

 ized NMFS/NEFC spring bottom-trawl surveys. The 

 NMFS/NEFC survey area has been spatially stratified 

 into approximate ecological units based principally 

 on depth (see Grosslein 1969). Stellwagen Bank is a 

 principal bathymetric feature in the Gulf of Maine- 

 NMFS/NEFC stratum #26. The number of sandlance 

 caught per tow in stratum #26 (for each spring survey) 

 was transformed into logarithmic values, and the mean 

 number of sandlance (transformed data) was con- 

 sidered representative of Stellwagen Bank and used 

 in all further analyses, a procedure which follows that 

 described in Payne et al. (1986). 



The Calanusfinmarchicus data were also collected 

 by NMFS/NEFC personnel. A transect across the Gulf 

 of Maine has been sampled with the Hardy Continuous 

 Plankton Recorder (CPR) since 1961 (Jossi and Smith 

 1990). Since the early 1980s a desired sampling fre- 

 quency of one transect per month has been generally 

 achieved. Data used in this study came from the 10- 

 nautical-mile section of the transect centered on Stell- 

 wagen Bank. Water passing through the CPR is 

 filtered with bolting silk having mean aperture dimen- 

 sions of 225 X 234 \jl. All large zooplankton (^2 mm) in 

 the sample are identified and enumerated. Counts of 

 smaller (< 2 mm) zooplankton are made from an aliquot 

 (~l/45) of the sample. Counts were converted to num- 

 ber of organisms per 100 m^, and then transformed to 

 log base 10 for subsequent calculations. Further details 

 of this procedure are given in Colebrook (1975) and in 

 Jossi and Smith (1990). All reference to copepod data 

 and abundance used in this study exclusively concern 

 C. finmarchicus. 



The abundance of sandlance (expressed as the mean 

 of the log-transformed number of sandlance per tow/ 

 year) and of the copepod C. finmarchicus (log-trans- 

 formed number of individuals/100 m-Vyear) were com- 

 pared by first ranking the data by year, 1982-88 (A'' 

 = 7), then calculating Spearman's coefficient of rank 

 correlation (rj (Zar 1984). The data are therefore 

 compared on an ordinal, rather than on an absolute, 

 scale; this is because of the magnitude of the difference 

 in the scale of the data that were brought together for 

 this study. 



