Climate-related changes in recruitment of the bivalve Macoma balthica

Population dynamics of common intertidal bivalves (Cerastoderma edule, Macoma balthica, Mya arenaria, Mytilus edulis) are strongly related to seawater temperatures. In northwestern European estuaries, series of mild winters followed by low bivalve recruit densities lead to small adult stocks. In this study, we examine temperature-induced effects on reproductive output (eggs m22), onset of spawning (day of the year), and the juvenile instantaneous mortality rate (per day) of M. balthica. Data analysis was based on an extensive long-term data set (1973–2001) originating from the western Wadden Sea. Our results strongly suggest that rising seawater temperatures affect recruitment by a decrease in reproductive output and by spring advancement of bivalve spawning. Apparently, global warming upsets the evolved reproductive strategy of this marine bivalve to tune its reproduction to the most optimal environmental conditions for the first vulnerable life stages, most importantly the match/mismatch of time of spawning with that of the phytoplankton bloom and the settlement of juvenile shrimps on the tidal flats. It is hypothesized that the observed density-dependent mortality of juvenile bivalves may act via competition for food, a behavioral response of shrimp to low spat densities, or be the result of the response of age and size at metamorphosis of marine bivalves to resource variability. It is to be expected that prolonged periods of lowered bivalve recruitment and stocks will lead to a reformulation of estuarine food webs and possibly a reduction of the resilience of the system to additional disturbances, such as the depletion and disturbance by shellfish fisheries. In northwestern European estuaries, severe winters are often followed by high densities of intertidal bivalve recruits, whereas densities after mild winters are usually low. This negative relationship between winter temperature and bivalve recruitment has been reported for cockles (Cerastoderma edule), Baltic tellins (Macoma balthica), gaper clams (Mya arenaria), and blue mussels (Mytilus edulis) (e.g., Beukema 1992a; Young et al. 1996). Large and covarying fluctuations in bivalve recruit densities largely determine the year-to-year variation in the size of the adult stocks (e.g., Beukema et al. 2001; Van der Meer et al. 2001). If global warming shifts the winter character toward higher temperatures, bivalve stocks as well as their predators may be affected (Beukema 1992a). Although species have responded to climatic changes throughout their evolutionary history, a primary concern for wild species living today is the current rapid rate of the temperature increase (Mac et al. 1998). Two recent articles show that present biological trends, i.e., shifts in range boundaries, phenological (timing) shifts, and changes in species abundances, are consistent with predicted effects of climate warming (Parmesan and Yohe 2003; Root et al. 2003). The synergism of rapid temperature rise and other stresses 1 Corresponding author ([email protected]). Acknowledgments We thank W. de Bruin, J. Hegeman, D. Waasdorp, and J. Zuidewind for their help in collecting the phytoplankton and macrozoobenthos samples during the past decades. Furthermore, we acknowledge J. Drent, M. F. Leopold, J. van der Meer, I. Williams, and two anonymous reviewers for their critical comments on earlier versions of this manuscript. Funding for this study was provided by research grants from the Netherlands Organisation for Scientific Research (Priority Program on Sustainable Use and Conservation of Living Marine Resources), the Netherlands Bremen Oceanography (NEBROC), and Rijkswaterstaat (Rijksinstituut voor Kust en Zee). could easily disrupt the connectedness among species and lead to a reformulation of species communities and possibly extinctions (Root et al. 2003). In the Wadden Sea, e.g., a series of mild winters (1988–1990) led to a prolonged period of low bivalve recruitment, which ultimately resulted in mass emigration, starvation, and mortality of bivalve-eating birds (Beukema and Cadée 1996; Camphuysen et al. 2002). Knowledge of the underlying mechanisms of failure of bivalve recruitment following mild winters is necessary to predict and manage the consequences of global warming for marine ecosystems. Under the assumption that recruitment takes place in a closed system, i.e., the pelagic larvae are retained in proximity to their natal population (see reviews in Warner and Cowen 2002) and that the instantaneous mortality rate (z; per day) is constant during the development from egg to postlarval recruit, the relationship between reproductive output (N0; eggs m22) and recruitment (Nt; recruits m22) densities at a particular age of the juvenile bivalve (t) can be described as Nt 5 N0e. Under these assumptions, low recruitment levels (Nt) following mild winters must be caused by either relatively low egg densities (small N0), a relatively long period between spawning and assessment of recruitment levels (large t) and/or high instantaneous mortality rates (high z) during the development from egg to recruit. Juvenile mortality rates may, however, differ for different phases, e.g., the pelagic larval and the bottom-dwelling recruit phase (Thorson 1966). If so, the phase mortality rates and the duration of the successive phases must be taken into account. During cold winters, bivalves show a lower metabolism, resulting in a higher preservational biomass (Zwarts 1971; Honkoop and Beukema 1997; Beukema 1992a) and consequently in the production of more eggs in early spring (Honkoop and Van der Meer 1997, 1998). Similarly, well-fed bi2172 Philippart et al. Fig. 1. Location of study area showing sampling stations for wind direction and speed (KNMI weather station), water temperature and chlorophyll a (NIOZ sampling jetty), and marine bivalves and shrimps (Balgzand). Dark-hatched areas indicate tidal flats, i.e., depth ,0 m (relative to mean lower low-water springs); lighterhatched areas indicate a water depth between 0 and 22 m. Table 1. List of variables included in the linear regression analyses of instantaneous mortality rate of Macoma balthica (per day) in the western Wadden Sea between 1973 and 2001 (Period I) and between 1983 and 2001 (Period II).