Warning This Web page has been archived on the Web.

Archived Content

Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards, as per the Policy on Communications and Federal Identity.

Skip booklet index and go to page content

COSEWIC assessment and status report on the American Eel in Canada

FEA5 - Eastern Arctic (Labrador)

Data regarding eels in Labrador are scarce. Electrofishing and fyke net studies have been conducted throughout the English River watershed over the past six years (1999-2005). This watershed flows in an easterly direction into Kaipokok Bay. The lower portion of the catchment is dominated by English River Pond (Clarke et al. 2004). The only eels captured since 1999 (n = 3) were in the lower main stem of the system in 2004. This section of the river is characterized by braided channels, created by islands, with a predominance of cobble/gravel substrate (Clarke et al. 2004). Given this, while eels certainly can and do inhabit this area, they would appear to be rare (K. Clarke, DFO, pers. comm.). These captures extend the known Canadian distribution of the species further north (Figures 3 and 23).

Figure 23.  Northern eel captures in Canadawithin the ecological freshwater area FEA5, Eastern Arctic.

Figure 23.  Northern eel captures in Canadawithin the ecological freshwater area FEA5, Eastern Arctic.

 

Canadian Eel Components in the North American Context

The collapse of the American eel in the upper St. Lawrence River and Lake Ontario prompts questions about the relative importance of this component to the species as a whole.  Specifically, it would be useful to know, for the period prior to the St. Lawrence - Lake Ontario collapse, what proportion of the total number of American eel eggs released on the spawning grounds were from eels that had used the upper St. Lawrence and Lake Ontario as rearing habitat.  It would also be useful to know the relative contribution of all Canadian eel components to total egg release on the spawning grounds.

There is no rigorous way to calculate these proportions.  However, two methods are available to give a first approximation.  One is based on river discharge (Castonguay 1994a) and the other is based on commercial landings.

Contribution of the St. Lawrence Eel Component - Discharge Method

Laboratory studies have demonstrated attraction by glass eels to low salinities and to natural organic chemicals that are contained in river discharge (Miles 1968, Tosi et al. 1988, Sola 1995).  Given the acute chemo-sensitivity of eels, it seems likely that young eels coming in from the ocean use chemical features of river discharge plumes to locate rivers.  Since rivers with larger discharge produce larger plumes, it is plausible that the number of young eels attracted to a river is a function of its discharge.  The discharge method assumes that the number of young eels recruiting to river systems is a linear function of water discharge at the river's mouth.

Studies on the European eel have shown that some glass eels caught off the coastline prefer to remain in salt water, when given a choice of salinities (Édeline and Élie 2004).  These are presumably the eels that colonize salt water bays and the lower parts of estuaries.  However, even eels that prefer salt water need a mechanism to find the coast, because open marine waters are not suitable growth habitat.  It is plausible that such eels locate the coastline by chemically detecting river discharge plumes, and the greater the discharge, the more eels will be attracted.  Hence we expand the basic assumption of the discharge method to say that recruitment of young eels to a continental area is a linear function of the total river discharge of that area.

The discharge method assumes that recruitment is unrelated to the amount of accessible upstream habitat, because young eels attracted to a river mouth have no way of knowing what proportion of habitat in the system is accessible.  The method further assumes that cumulative survivorship between recruitment to continental habitat and return to the spawning grounds does not vary among regions.

The availability of numerous historical reports of American eels in the Mississippi drainage basin suggests that the Mississippi basin was once a significant part of the species' historical range (Casselman 2003).  However, this may no longer be the case. The discharge method was therefore applied under two differing assumptions; first, that the American eel's range excludes the Mississippi basin, and second, that the American eel's range includes the Mississippi basin. 

Table 3 shows estimated total discharges for major eel rearing areas in North America. The discharge value for the St. Lawrence River is from Baie Comeau.  Water flowing past this point includes discharge from tributaries in the western part of FEA2. The Baie Comeau discharge is used because this location is near the river's mouth, and hence represents the attraction water that draws glass eels and elvers into the St. Lawrence River system.  Thus in the discharge analysis the St. Lawrence River basin includes FEA1, the western part of FEA2 which drains into the St. Lawrence estuary, and US drainage into the St. Lawrence.

Spawn output depends on number of eels, sex ratio, and fecundity.  The number of silver eels produced, relative to production in the US eastern seaboard, is shown in Table 3.  For this purpose production in the US eastern seaboard is arbitrarily set at 1,000 eels.  Relative production for each region is calculated by comparing that region's discharge to the discharge of the US eastern seaboard.  For example, mean discharge of the St. Lawrence basin is 16,800 m3sec-1and mean discharge of the US eastern seaboard is 11,186 m3 sec-1.  Number of silver eels produced by the St. Lawrence, relative to production of the US eastern seaboard, is 1000 x 16,800/11,186 = 1,502.

Table 3.  Relative contribution of American eel rearing regions to total egg production, based on the assumption that the number of silver eels produced is proportional to mean freshwater discharge.

Table 3.  Relative contribution of American eel rearing regions to total egg production, based on the assumption that the number of silver eels produced is proportional to  mean freshwater discharge.

The relative number of silver eels produced is multiplied by the proportion of silver eels that are female in the various regions to produce the relative number of female silver eels.  Eels from the St. Lawrence River system are virtually all female. Based on mean regional values from the sex-ratio compilation of Nilo and Fortin (2001), 97.8% of silver eels in FEA2, FEA3, FEA4, and FEA5 are female and 65.5% of silver eels in the U.S. eastern seaboard are female (Table 3). 

The relative number of eggs produced is determined by multiplying the relative number of female silver eels by mean fecundities.  Fecundities of FEA1 eels, measured at two locations, and fecundities from the Sud-Ouest River, which flows into the south side of the St. Lawrence Estuary in FEA2, are derived from large eels whose size is typical of the St. Lawrence system (Tremblay 2004). The mean of these values is taken as the St. Lawrence River fecundity (Table 3). Silver eels measured in two north shore tributaries in FEA2 are typical of sizes found in Prince Edward Island, Nova Scotia, and Newfoundland (Table 2). The great majority of FEA2 is on the north shore of the St. Lawrence Estuary and Gulf (Figure 3). Hence fecundity in central and eastern FEA2, FEA3, FEA4, and FEA5 is assumed to be the mean of two fecundity measurements in the Gulf of St. Lawrence (Tremblay 2004).  Fecundity for the U.S. is derived from the fecundity-length relation of Barbin and McCleave (1997), applied to the mean length of female silver eels in the U.S. eastern seaboard as compiled by Nilo and Fortin (2001). 

The discharge method estimates that the St. Lawrence River basin contributes 67.2% of the spawn output of the species if the species’ range is considered to exclude the Mississippi Basin, and 60.3% of the spawn output if range includes the Mississippi (Table 3).

There are a number of factors which limit the confidence that can be placed in this result.  These are summarized below.

The assumption that recruitment of young eels from the sea is a direct function of discharge is based on the premise that young eels in the ocean can access all continental rearing areas with equal ease.  Continental rearing areas that are close to the spawning grounds and that directly face the Atlantic Ocean (e.g. the US eastern seaboard) require only a short and direct trip from the spawning ground.  Habitats that drain into semi-enclosed gulfs which are distant from the spawning ground (e.g. the St. Lawrence and Mississippi basins) require a long and circuitous trip.  It is unlikely that a linear relation between discharge and recruitment applies across habitats which differ so greatly in their distance and routing from the spawning ground.  European glass eel influx is intense in the central part of the species' continental range, but is much lighter toward the edge of the species' range (Dekker 2000, Knights 2003).  On a smaller scale, Jessop (1998b) found that discharge did not explain elver run size in two rivers on the Atlantic coast of Nova Scotia.

If the pool of recruiting eels available to colonize the St. Lawrence River is diminished by the long and circuitous route, then the discharge method will overestimate the relative importance of the St. Lawrence River to the species' egg production.  The potential error can be illustrated by assuming that St. Lawrence River recruitment is only 50% of what it would be if recruitment is linearly related to discharge.  Under this scenario, the St. Lawrence would produce 50.6% of total eggs if range is assumed to exclude the Mississippi basin, and 43.2% of total eggs if the Mississippi is included.  If St. Lawrence recruitment is 25% of what it would be if recruitment is linearly related to discharge, St. Lawrence egg production would be 33.9% of the total when Mississippi is excluded, and 27.5% of total when the Mississippi is included in the species’ range.

Low salinity and natural organic chemicals appear to attract young eels coming in from the sea (Miles 1968, Tosi et al. 1988, Sola 1995).  However, the relative importance of these two types of attractants is not known.  Miles (1968) showed that attractiveness to eels of natural organic chemicals varied among Nova Scotia rivers.  If natural organic chemicals are the main attractant for eels at sea, variability in the natural chemical properties of river plumes may invalidate the assumed linear relation between discharge and recruitment.

The discharge method assumes that cumulative natural mortality is similar across all regions.  Eels that rear in fresh water take much longer to reach the silver stage than those that do so in brackish or salt water (Lamson et al. submitted).  If natural mortality on an annual basis is similar across habitats, then cumulative mortality to the silver stage would be much greater in fresh habitats.  Alternately, eels in salt water might be trading faster growth rates against higher mortality, so that cumulative mortality between the habitat types is similar.  Rearing habitat in the St. Lawrence and Mississippi basins is fresh.  Rearing habitat in other regions is a mix of fresh and marine.  Hence any difference between cumulative mortality with respect to salinity of rearing habitat would lead to error in the estimate of the relative egg production of the St. Lawrence River basin.

Silver eels escaping from continental waters must travel to the Sargasso Sea in order to spawn.  Differential survival rates on this journey would introduce error in the estimates of relative egg production.  Silver eels leaving the St. Lawrence River might have lower survival rates than those of the US eastern seaboard because of the longer distance they must travel.  On the other hand, St. Lawrence River eels might have higher survival rates because they are large, and large fish typically have lower mortality rates than small fish.

The discharge method assumes that there is no inter-regional variation in the effect that habitat quality and accessibility have on silver eel production.  Accessibility can be blocked by natural and artificial obstacles.  It is also affected by distance, as eel density in a system typically declines with distance from the river mouth (Moriarty 1987). The assumption that habitat effects are similar across regions seems unlikely.  In particular, production in the St. Lawrence River basin, a long system with obstacles on its mainstem and tributaries and no marine habitat, may be more affected by habitat issues than production in regions that include readily-accessible marine habitat.  The discharge method also ignores the effects of commercial fisheries on silver eel production.  Commercial fishing effort varies geographically, and substantial areas in both Canada and the US are unexploited. 

Reliability of the analysis is affected by limited data on sex ratio and fecundity. Eel sex ratios may be biased because of under-representation of males due to size selectivity of gear. Fecundity data from the Gulf of St. Lawrence are used for US waters, because no published fecundities are available there.