Ecology of Freshwater Fish 2010: 19: 586–594 Ó 2010 John Wiley & Sons A/S Printed in Malaysia Æ All rights reserved ECOLOGY OF FRESHWATER FISH Spawning dynamics of American shad (Alosa sapidissima) in the St. Lawrence River, Canada–USA Maltais E, Daigle G, Colbeck G, Dodson JJ. Spawning dynamics of E. Maltais1,*, G. Daigle2, American shad (Alosa sapidissima) in the St. Lawrence River, G. Colbeck1, J. J. Dodson1 Canada–USA. 1De´partement de Biologie, Universite´ Laval, Ecology of Freshwater Fish 2010: 19: 586–594. Ó 2010 John Wiley & Que´bec, QC, Canada, 2De´partement de Sons A ⁄ S Mathe´matique et de Statistique, Universite´ Laval, Que´bec, QC, Canada Abstract – The most northerly population of American shad (Alosa sapidissima), located in the St. Lawrence River, is considered vulnerable because of low population abundance and limited spawning habitat located at the upstream extent of the population’s anadromous migration. Here, we aimed to establish the temporal and spatial extent of spawning based on a novel hatch-date analysis of juveniles. Spawning activity Key words: alosine clupeid; hatch-date analysis; lasted from early May to early July. We found that juveniles captured multiple spawning events downstream during the summer hatched later in the year than those captured further upstream. As a result, younger juveniles were distributed Julian J. Dodson, De´partement de Biologie, 1045 Avenue de la Me´decine, Universite´ Laval, Que´bec, somewhat further downstream. In addition, we found significant QC, Canada G1V 0A6; e-mail: julian.dodson@ multimodality in hatch-date distributions at midstream and downstream bio.ulaval.ca sampling stations. Together, these results provide evidence that the 2-month spawning period involved numerous spawning events that *Present address: AECOM, 231, Boulevard La progressed in a downstream direction as the season advanced, rather than Salle, Baie-Comeau, QC, Canada G4Z 1S7. being restricted to upstream sites over the spawning season. Accepted for publication May 28, 2010 Lawrence estuary and found immediately downstream Introduction of dams. One is located in the Ottawa River at Carillon The anadromous American shad (Alosa sapidissima) (Point Fortune) on the southwest shore at the Ontario– ranges from the St. Lawrence River, Canada, to the Quebec border downstream of a dam built in 1882 St. Johns River, Florida (Limburg et al. 2003). From (Provost et al. 1984). A second spawning area is their overwintering grounds on the Scotian shelf located in the vicinity of Montreal in the Rivie`re-des- (Dadswell et al. 1987), the St. Lawrence population Prairies below a dam built in 1928 (Bilodeau & Masse undertakes an annual spring migration of approxi- 2005). Prior to dam construction, American shad were mately 2000 km to reach its spawning grounds in the potentially capable of migrating as far upstream as vicinity of Montreal, Que´bec (Fig. 1). This population Ottawa on the Ottawa River, 125 km beyond the has suffered severe reductions, already recognised as Carillon dam (Provost et al. 1984). The historical early as the late 19th century (Montpetit 1897) and extent of upstream migration in the St. Lawrence continuing through the 20th century, with no sign of River, beyond Montreal and up to the outlet of Lake recovery (Provost et al. 1984). The reasons for this Ontario, a total distance of 290 km, is unknown. Given decline are probably numerous, but exploitation during the low abundance of the population through much of the spring migration and the construction of multiple the 20th century, the loss of potentially significant hydroelectric dams in the vicinity of Montreal and in spawning grounds upstream of contemporary dams the Ottawa River probably played a major role (Provost and the fact that only two spawning areas are presently et al. 1984). There are presently two known spawning known, American shad is one of 11 species listed as areas in the St. Lawrence River and its tributaries, both vulnerable under Que´bec legislation on threatened and located approximately 260 km upstream from the St. endangered species (MRNF, 2009). 586 doi: 10.1111/j.1600-0633.2010.00439.x Spawning dynamics of American shad Fig. 1. Location map of the St. Lawrence River, illustrating the major sectors included in the sampling of juvenile American shad. American shad are repeat spawners in the northern exist in the Sainte-Anne, Batiscan and Richelieu part (>32°N) of their distribution along the east coast rivers, as well as in the St. Lawrence fluvial lakes of North America (Leggett & Carscadden 1978). (Fig 1) (Provost et al. 1984). Provost (1987) proposed Under high levels of variance in mortality of the the existence of distinct populations among the youngest life-history stages, such iteroparity forms tributaries of the St. Lawrence River based on the part of the bet-hedging reproductive strategy that phenotypic characteristics of migrating adult fish. trades off a higher mean fitness in favour of a lower To achieve this objective, we captured young- variance in fitness under environmental uncertainty of-the-year shad throughout the St. Lawrence River to (Philippi & Seger 1989). Another manifestation of bet- obtain otoliths for aging and hatch-date analysis. We hedging is batch or serial spawning, whereby females tested the null hypothesis that all shad spawn uniquely spawn on several occasions and at different sites in the vicinity of Montreal. Given this scenario and during the same reproductive season. In American conservatively assuming downstream movement of shad, batch spawning has been observed in both larvae and juveniles (Limburg 1996), we predicted our semelparous (Edisto River) and iteroparous (York and first captures to include the youngest larvae hatched in Connecticut Rivers) populations (Olney et al. 2001). the vicinity of Montreal, with age increasing over time Evidence for batch spawning is based on the simul- in a downstream direction in concert with the down- taneous observation of developing and postovulatory stream displacement, either active or passive, of follicles in females captured at different points in their juveniles. As such, we expected juveniles collected migration. These authors proposed that batch spawn- downstream to have hatched earlier in the season than ing increases the probability of finding favourable juveniles captured further upstream. The alternative conditions for the survival of the most vulnerable life- observation, whereby juveniles collected downstream history stages by dispersing them over greater spatial hatched later in the season than juveniles collected and temporal scales (Olney et al. 2001). Alternatively, upstream, would indicate the presence of multiple different groups of fish, associated with different spawning events occurring downstream of known spawning events in space or in time, may also exist spawning sites in the vicinity of Montreal. The within the same river system (Hendry & Day 2005). existence of multiple spawning events would also be Our objective in this study is to establish the expected to generate bi- or multimodal hatch-date temporal and spatial extent of spawning by American distributions of juvenile shad at specific sampling shad in the St. Lawrence River. Although only two stations. spawning areas have been identified, there are numer- ous sites throughout the St. Lawrence River where Materials and methods spawning by American shad is suspected to occur based on the physical and biological characteristics of Study area and sampling the sites. Bouchard (1976) identified 16 potential spawning sites for American shad over the lower We sampled three to six sites (22 total) in each of five 120 km of the river. Putative spawning sites may also areas between Montreal and Ile d’Orleans, a distance 587 Maltais et al. of about 260 km (Figs 1 and 2). Each site was Laboratory methods sampled at several stations (Fig. 2). Sampling sites were clustered in three major regions for some Species identification and otolith extraction subsequent analyses (the upstream region encompass- We distinguished American shad from other clupeids ing the Montreal and Richelieu sampling sites, the (mostly alewife Alosa pseudoharengus) with identifi- downstream region in the vicinity of Iˆle d’Orle´ans and cation keys (Jones et al. 1978 and Provost et al. 1984) the midstream region encompassing the Trois-Rivie`res and genetic analysis in the case of very small and Batiscan sampling sites). Samples were obtained individuals. We used a restriction fragment length during seven sampling occasions that took place from polymorphism technique to discriminate between shad 14 June to 6 September 2006. One field team sampled and alewife. We extracted DNA with the salt technique upstream sites and a second team sampled midstream described by Aljanabi & Martinez (1997) and ampli- and downstream sites. Teams worked simultaneously fied a segment of 609 base pairs of the mitochondrial and the weekly sampling occasions lasted on average cytb gene. The primers used for the PCR amplification 3 days. were (L1-600) 5¢-GGG TTG TTT GAT CCT GTT TCG TG-3¢ and (H2-600) 5¢-AAA ACC ACC GTT GTT ATT CAA CTA CA-3¢. The amplified segment Larval and juvenile collections was digested with the HaeIII (5¢-GG # CC-3¢) restric- Larval American shad were collected with a 75-cm tion enzyme and afterward the fragments were counted diameter pelagic trawl (500-lm mesh) and juveniles on 2% agarose gels to discriminate between American (postmetamorphic) were caught with a 10-m beach shad and alewife. The American shad has two seine (2-mm mesh). Trawling took place between 14 restriction sites in this segment that produce three June and 5 July and seine fishing was conducted fragments and the alewife has one restriction site that during the entire sampling period. Trawling consisted produces two fragments (F. Martin, F. Colombani and of a 10-min tow in the upper 2 m of the water column. J.J. Dodson, unpublished data). All specimens were preserved in 95% ethanol for Altogether, we captured 3526 clupeids including subsequent analyses.