Movement Patterns of American Eels in an Impounded Watercourse, as Indicated by Otolith Microchemistry

—Otolith Sr:Ca ratios were used to determine migrational histories of American eels Anguilla rostrata in an eastern Canadian coastal stream that is impounded at the head of tide. American eels sampled from the estuary were smaller (41.0 6 9.3 cm [mean 6 SD], N 5 324) and younger (ages 1–4) than those in a freshwater pond (65.0 6 7.5 cm, N 5 265; ages 11–26) 2.5 km upstream of the head of tide. The Sr:Ca ratios permitted identification of three migrational contingents. In 7 of 13 (54%) American eels taken in the estuary, Sr:Ca ratios declined to 4.76 6 1.22 3 1023, indicating direct migration from the sea to estuarine settlement sites. In the 6 other American eels (46%), Sr:Ca ratios indicated an approach to freshwater soon after arrival in continental waters, followed by settlement in the estuary where ratios stabilized at 4.50 6 1.23 3 1023. In all 15 American eels sampled from the freshwater pond, Sr:Ca ratios indicated settlement in freshwater in the elver year and no subsequent change of habitat salinity (postelver ratio, 0.64 6 0.66 3 1023). The lack of evidence for postelver movements between freshwater and brackish water contrasts with literature findings that such movements are common in systems where dams do not impede access to freshwater. Dams may hamper normal American eel movements between rivers and estuaries, even in watercourses in which some American eels are able to colonize freshwaters. Recent declines of American eels Anguilla rostrata and European A. anguilla eel have focused attention on factors that limit numbers of these species. Possible causes of the declines include changes in ocean conditions, overfishing, pollution, and dams (Richkus and Whalen 2000). Busch et al. (1998) estimated that up to 84% of riverine habitat in U.S. eastern seaboard and Lake Ontario drainages lies upstream of dams. Although natural barriers would have impeded American eel access to some of this habitat, even if dams had not been * Corresponding author: [email protected] Received March 14, 2003; accepted July 2, 2003 built (Haro et al. 2000), it is clear that dams affect a substantial portion of potential eel habitat in North American fresh waters. American eels are weak swimmers and are unable to leap over waterfalls or swim directly against strong currents. Eels ascending rivers attempt to circumvent obstacles by seeking lowcurrent routes, usually along banks that have a rough substrate, which permits the animal to advance with a serpentine motion (Tesch 1977). Eels smaller than about 11 cm can creep up vertical walls, but larger eels require a sloping substrate (Legault 1988). Thus some obstacles impose a size-dependent barrier to migration. Although eels are known to perform remarkable feats of migra453 EEL MOVEMENTS IN AN IMPOUNDED WATERCOURSE FIGURE 1.—Map of the Boughton River, Prince Edward Island, where movement patterns of American eels were examined with otolith microchemistry. Triangles indicate dams and the arrow indicates the study area. tion, such as surmounting waterfalls and invading landlocked lakes (Eales 1968), the general case appears to be that physical obstacles substantially constrain upstream movements (Porcher 1982; White and Knights 1997; Feunteun et al. 1998). However, factors enabling an eel to overcome obstacles in a river have not been quantitatively investigated. This means that the effects of dams and other structures on eel ascent of a river cannot be predicted from the nature of the obstacle. Evaluations of the effects of dams on eel populations emphasize the classic catadromous paradigm that American eels ingress into freshwater, remain in freshwater during the yellow (growth) stage, and then return to the sea as silver eels (McCleave 2001). However, eels may exhibit other migratory patterns. Secor (1999) pointed out that diadromous fish commonly show intrapopulation variation in migration, some maintaining longterm residence in one habitat, others shifting among habitats. The development of microprobe analysis of Sr:Ca ratios in fish otoliths permits determination of the migratory history of individual fish through salt, brackish, and freshwater because Sr:Ca ratios are empirically linked to ambient salinity (Secor et al. 1995; Secor and Rooker 2000). Otolith Sr:Ca ratios decline sharply during the glass eel stage because of metabolic processes unrelated to ambient habitat (Tzeng et al. 1997). For subsequent stages, laboratory validation studies have confirmed that otolith Sr:Ca ratios reflect ambient salinity, although temperature and diet also play a role (Tzeng 1996; Bath et al. 2000; Kraus and Secor 2003). In many systems American eels exhibit seasonal migratory runs, especially in spring and fall (Smith and Saunders 1955; Medcof 1969; Jessop 1987). However, the occurrence of such runs does not necessarily mean that individual eels undertake regular seasonal movements. Application of the Sr:Ca method to American eels in an unimpounded stream in Nova Scotia showed that some individuals chose habitat of a certain salinity and then stayed there, whereas others showed irregular, often nonannual movements between the river and the estuary (Jessop et al. 2002). In the Hudson River estuary, Sr:Ca ratios of yellow American eels captured in brackish waters indicated that some animals had stayed in brackish water since their arrival as elvers, whereas others had spent varying periods in freshwater before returning to brackish water (Morrison et al. 2003). However, eels sampled in the freshwater upper reach of the estuary showed no evidence of residency in brackish water. Tzeng et al. (2003) found that the majority of Japanese eels Anguilla japonica sampled as silver eels in an estuary had shifted residence between freshwater and seawater, or vice versa, during their yellow eel phase. These studies suggest that movements between fresh and brackish waters are common in yellow-phase eels and that such movements are often irregular and nonannual. Barriers to fish migration could potentially reduce American eel production in a watercourse (estuary and river) by hindering colonization of freshwater rearing areas or by suppressing movements, either annual or irregular, between river and estuary. This study uses otolith Sr:Ca ratios to determine patterns of river–estuary movements in American eels in a small system that is impounded at head of tide. These data are used to evaluate potential effects of dams on eel movements between habitats of varying salinity.