Genetic Mixed-Stock Analysis of American Shad in Two Atlantic Coast Fisheries: Delaware Bay, USA, and Inner Bay of Fundy, Canada

American Shad Alosa sapidissima in the Hudson River, New York, and coastwide have shown major long-term declines. A possible contributing factor is commercial fisheries that harvest this population outside of the Hudson River estuary. Using previously published and new reference microsatellite data from 33 baseline populations, our goals were (1) to estimate the proportion of Hudson River American Shad contributing to the two remaining major mixed-stock fisheries along the Atlantic coast in Delaware Bay and the Bay of Fundy and (2) to estimate the proportions of other American Shad stocks contributing to these two fisheries at the highest level of stock specificity. Stock composition estimates for 2009 and 2010 Delaware Bay collections were made using three models that ranged from the most simple question (Hudson River and Delaware Bay populations) to one with all 33 baseline populations included. In all cases, a Hudson River contribution nearly equal to that of the Delaware Bay contribution was observed, indicating a substantial take on the otherwise protected Hudson River population. When all baseline populations were included for the larger 2010 Delaware Bay collection, 19 showed nonzero contributions, largely drawn from mid-Atlantic U.S. rivers. The 2009 Bay of Fundy collection showed contributions from across most of the species’ range but was dominated by northern populations. Mixed-stock analyses of collections from the two sites together indicate that these estuarine fisheries harvested not only proximal populations but those originating from a wide latitudinal range. *Corresponding author: [email protected] Received April 3, 2014; accepted August 5, 2014 1190 North American Journal of Fisheries Management 34:1190–1198, 2014 American Fisheries Society 2014 ISSN: 0275-5947 print / 1548-8675 online DOI: 10.1080/02755947.2014.954067 D ow nl oa de d by [ N ew Y or k U ni ve rs ity ] at 0 7: 26 1 9 N ov em be r 20 14 Populations of American Shad Alosa sapidissima (hereafter referred to as shad) once occurred in nearly 140 rivers from the Pinware River, Labrador (Dadswell et al. 1987), to the St. John’s River, Florida (Limburg et al. 2003). However, this anadromous species has shown declines of two orders of magnitude over 2 centuries across its native range (ASMFC 2008). Limburg and Waldman (2009) estimated that shad have suffered extirpations of almost half of their original populations. Catches also have fallen drastically in natal rivers. Within the Hudson River, New York, shad landings reached as high as 1.8 million kg in 1889, declining to as low as 25,000 kg in 2007 (NYSDEC 2008). The Hudson River’s commercial shad fishery closed in 2010, 2 years after a moratorium began on its minor recreational fishery. Genetic (Bentzen et al. 1989; Nolan et al. 1991; Waters et al. 2000), otolith microchemistry (Walther et al. 2008), and tagging evidence (Melvin et al. 1986), in addition to independent fluctuations in abundance (Leggett and Whitney 1972), indicate that shad form discrete, philopatric populations. Much has been learned about the spatial and phenological aspects of shad life histories in rivers (e.g., the locations and timing of spawning, the out-migration of juveniles). However, this contrasts with a paucity of information about the major portion of their existence, which occurs at sea. The mark–recapture program of Dadswell et al. (1987) contributed much to the knowledge about the species’ broadscale marine migratory circuits, the annual timing of these movements, and the locations of offshore wintering regions. In their study, shad tagged in the Hudson River and New York Bight were mainly recaptured in the Hudson River itself, but recoveries stretched from slightly south of Cape Hatteras, North Carolina, to Halifax, Nova Scotia, and into Cumberland Basin in the upper Bay of Fundy, New Brunswick. Thus, Hudson River shad have potentially extensive geographic vulnerability to coastal fisheries and other anthropogenic stressors distant from their natal rivers. From a management perspective, anadromous fishes are best harvested during their spawning runs because (with high homing fidelity) harvests are nearly stock-specific and, thus, the harvests of individual stocks can be managed in accordance with their fluctuations in abundances (Wirgin et al. 1993). This important connection between harvest and its effects at the population level breaks down when that link between harvest and stock identity is lost, as occurs in mixedstock fisheries. Mixed-stock fisheries are fisheries that harvest two or more stocks simultaneously, usually at coastal or offshore locations instead of within rivers, where stock segregation would already have occurred. Traditionally, shad were harvested almost exclusively by in-river fisheries at spawning time, allowing fishing pressure to be adjusted to the demographic trajectories of individual populations. Indeed, the Atlantic States Marine Fisheries Commission (ASMFC) considers individual stocks (defined as populations herein) to be the appropriate management units for shad. However, in the 1980s, new “intercept” fisheries for shad developed in marine waters in which they were harvested as they followed coastal migrations (ASMFC 2007). These new shad fisheries had two deleterious consequences: (1) they considerably increased total harvest and (2) they eroded the connection between fishing and stock-specific demographics. These intercept fisheries grew, especially in U.S. mid-Atlantic and northern states, where they exceeded in-river landings (ASMFC 2007). Beginning in 2000, U.S. open-ocean intercept fishery landings were decreased and then closed over 5 years, and they remain closed. However, at least two intercept-type fisheries for Atlantic coast shad persist today. One of these occurs in U.S. waters in a possible staging area between open marine waters and spawning rivers, in Delaware Bay. Delaware Bay is a location that presumably includes a high proportion of Delaware River spawners but also may include substantial fractions of other populations (including the proximal Hudson River stock) as they move northward toward their own spawning rivers. A second mixed-stock harvest occurs in the brush weir and drift gill-net fisheries (Dadswell et al. 1984a) in the inner Bay of Fundy, Canada. Dadswell et al. (1987) showed that many U.S. stocks make subadult and adult movements that include looping passage through the Bay of Fundy. In this exploratory study, we estimated the stock composition of each of these mixed-stock fisheries using analysis of microsatellites under three scenarios that ranged in complexity from a simple two-stock problem to one that considered individually every included population. Microsatellites are a powerful tool in studies of molecular evolution, stock structure, and genetic mixed-stock analysis (Wirgin and Waldman 2005), primarily due to the codominant multiallelic variation often detected at individual microsatellite loci and the large number of loci that can be screened. We used microsatellite data for 33 baseline populations collected between 2003 and 2006 from across the species’ geographic range (Figure 1) and reported in Hasselman et al. (2013). Hasselman et al. (2013) found significant allelic-frequency differences among most populations, except for some pairwise comparisons within Chesapeake Bay and among some semelparous populations towards the southern extent of the species’ range. Genetic differentiation (FST) was significant among most rivers, and Bayesian inference suggested the existence of nine clusters of populations, with major differences in allelefrequency distributions among clusters (Hasselman et al. 2013). Cumulatively, these results suggest that this suite of microsatellite loci provides suitable statistical power to assign individuals captured in mixed-stock fisheries to one previously identified cluster and, possibly, even to specific rivers within clusters. GENETIC ANALYSIS OF AMERICAN SHAD IN TWO FISHERIES 1191 D ow nl oa de d by [ N ew Y or k U ni ve rs ity ] at 0 7: 26 1 9 N ov em be r 20 14