Past and Present Processes Influencing Genetic Diversity and Effective Population Size in a Natural Population of Atlantic Sturgeon

Threats such as habitat loss, invasive species, and overexploitation cause species extinctions; however, stochastic processes can accelerate extinction rates as census sizes decline. Using molecular and ecological data, we explored the influence of these processes on the demography of a candidate species under the U.S. Endangered Species Act—the Atlantic sturgeon Acipenser oxyrinchus oxyrinchus. We used molecular microsatellite markers to estimate the effective population size (Ne) and effective number of breeders (Nb) and we used mark–recapture data to estimate the number of spawners (Na) for Atlantic sturgeon of the Altamaha River, Georgia. We found that estimates of Nb were 7–45% less than the estimated Na over four consecutive cohorts and that skewed sex ratios could explain the relative decrease of Nb to Na. Our estimate of contemporary Ne was 125 (95% confidence interval = 75–348) and was at least an order of magnitude less than our estimate of historical Ne. To explain the large discrepancy between these estimates, we tested several alternative evolutionary scenarios that might explain the observed pattern of genetic diversity. Our results indicated that the observed genetic data were indeed best explained (i.e., 0.998 posterior probability of the data given the hypothesis) by overexploitation during the last half of the 20th century. Population census size (N) is an important parameter in the conservation and management of threatened and endangered species because variability of N determines extinction risk. As N becomes small owing to deterministic events (e.g., natural selection, habitat loss or modification, or overharvest), stochastic factors such as inbreeding and loss of genetic variation (genetic factors), skewed sex ratio or variance in family size (demographic factors), and variation in environmental conditions (environmental factors) can accelerate population decline (Fagan and Holmes 2006). For example, inbreeding and loss of neutral *Corresponding author: greg [email protected] Received October 4, 2010; accepted June 28, 2011 genetic variation associated with a small population size can further reduce the fitness of the population (Reed and Frankham 2003). The rate of these genetic processes, however, is not contingent on N but rather on the effective population size (Ne) of the population. The parameter Ne, which refers to the size of an ideal population experiencing the same rate of random genetic change over time as the real population under consideration (Wright 1938), is typically much smaller than N because of various life history and reproductive biology aspects, such as fluctuating population size, unequal sex ratio, and variance in reproductive 56 D ow nl oa de d by [ U S Fi sh & W ild lif e Se rv ic e] , [ G re g M oy er ] at 1 1: 13 3 0 Ja nu ar y 20 12 ATLANTIC STURGEON HISTORICAL DEMOGRAPHICS 57 success (Frankham 1995; Palstra and Ruzzante 2008). Populations, therefore, may become susceptible to stochastic factors at levels where census estimates would provide little indication of a high risk of extinction or extirpation (Turner et al. 2002, 2006). While Ne is important in evolutionary and conservation biology, it is often difficult to measure in natural populations. Estimates of contemporary Ne (roughly, the Ne that applies to the time period encompassed by the sampling effort) are typically derived from a single sample (Hill 1981) or two samples (Nei and Tajima 1981). The two-sample (temporal) method, which depends on random changes in allele frequency over time, has been widely applied; however, in addition to other simplifying assumptions, the standard temporal method assumes that generations are discrete. This method is thus difficult to apply to iteroparous, age-structured species unless temporal samples are taken several generations apart (Waples and Yokota 2007). An age-structured population presents a problem for the estimation of Ne because it does not constitute a homogeneous breeding unit; therefore, there will not necessarily be a direct relationship between Ne and temporal allele frequency fluctuations. Instead, variances and covariances in allele frequencies within and among age-classes depend on the age-specific birth and survival rates of each population and will tend to bias estimation of Ne (Waples and Yokota 2007) unless these sampling effects are corrected (Jorde and Ryman 1995). In such situations, demographic information must be accounted for when estimating Ne for observed short-term fluctuations in allele frequency. Conservation and management of age-structured, iteroparous species thus depend on detailed knowledge of interactions between life history and population dynamics. For example, theoretical findings suggest that species characterized by overlapping generations and multiple mating opportunities are more resistant to the detrimental genetic consequences of years with poor recruitment and low numbers of breeders (Warner and Chesson 1985; Nunney 1993; Ellner and Hairston 1994) and that these characteristics render a population less sensitive to environmental variance (Gaggiotti and Vetter 1999). This resilience, however, may not protect against other random processes or against declines driven by deterministic factors. Therefore, while theoretical connections between Ne and life history, behavioral ecology, and demography are becoming better understood, few empirical studies on long-lived iteroparous species are available to support these findings (however, see Gaggiotti and Vetter 1999). One such long-lived species is the Atlantic sturgeon Acipenser oxyrinchus oxyrinchus. During the mid-1800s, this species was very abundant in many of the major river systems along the Atlantic coast of North America (Armstrong and Hightower 2002); however, in less than 10 years after the caviar market was established, annual landings had declined to less than 10% of their former peak (Secor and Waldman 1999). In 1990, after several decades of continued decline, the Atlantic sturgeon fishery was closed; shortly thereafter, this species was petitioned for listing under the U.S. Endangered Species Act (USFWS 1997) and is currently listed as a candidate species (USFWS 2006). Atlantic sturgeon spend most of their adult life in the marine environment but migrate to freshwater to spawn (in the Altamaha River, which is the focus of our study, the spawning run typically occurs in February–March). The mating system of Atlantic sturgeon is unknown but presumed to be polygamous like that of other sturgeon species (Schueller and Hayes 2010). Available evidence indicates that eggs, which are highly adhesive, are broadcast into flowing water and subsequently settle on bottom substrate, usually on hard surfaces such as cobble (Smith and Clungston 1997). Juvenile Atlantic sturgeon move downstream into brackish waters and eventually become residents in estuarine waters at 2–5 years of age. As subadults (76–92 cm total length), Atlantic sturgeon typically move to coastal waters (Smith 1985), where they migrate among coastal and estuarine habitats. Despite extensive mixing in coastal waters, Atlantic sturgeon appear to return to their natal river to spawn, as indicated by tagging records (Collins et al. 2000) and relatively low rates of gene flow (King et al. 2001). Life history characteristics of Atlantic sturgeon show clinal variation, with faster growth and earlier age at maturation in more southerly systems (e.g., Altamaha River) compared with more northerly systems (e.g., Hudson River; Van Den Avyle 1984). The estimated age structure of Atlantic sturgeon in the Altamaha River consists of 19 year-classes (Peterson et al. 2008). Males mature at age 5, whereas females typically mature at 11 years of age (Peterson et al. 2008). Spawning intervals of Atlantic sturgeon range from 1 to 5 years for males (Smith 1985; Collins et al. 2000; Caron et al. 2002) and from 2 to 5 years for females (Van Eenennaam et al. 1996; Stevenson and Secor 2000). Fecundity of Atlantic sturgeon (ranging from 400,000 eggs to 8 million eggs) has been correlated with weight (Van Den Avyle 1984). The main purpose of this study was to estimate Ne for Atlantic sturgeon from the Altamaha River by using genetic and life history data and, in doing so, to provide empirical evidence for the various processes influencing this value.