Seasonal refugia, shoreward thermal amplification, and metapopulation dynamics of the ctenophore Mnemiopsis leidyi in Narragansett Bay, Rhode Island

The lobate ctenophore Mnemiopsis leidyi occurs throughout Narragansett Bay, Rhode Island, during warm summer months but is often undetectable in the central portion of the bay during winter months. During 2 yr of weekly sampling, we found that M. leidyi populations in a shallow embayment, Greenwich Cove, either overwintered or were only briefly absent during winter. The Greenwich Cove population reproduced weeks earlier and reached higher average and peak population concentrations than open-bay populations. Shallow embayment populations such as that in Greenwich Cove probably serve as source populations that inoculate the main region of the bay by advective transport in the spring months. We propose that earlier occurrences of M. leidyi during recent years are due to amplification of pulsed spring warming events that permit early reproduction in the shallow embayments that serve as source regions for M. leidyi in Narragansett Bay. We further suggest that the source-sink perspective we describe is relevant not only to Narragansett Bay but other temperate regions of the world persistently occupied by M. leidyi. The importance of Mnemiopsis leidyi as a planktonic predator has been documented by a large body of research on its feeding capabilities (Kremer 1975; Reeve et al. 1978; Waggett and Costello 1999) and trophic impacts (Kremer 1979; Shiganova et al. 2001; Sullivan et al. 2001). These predatory capabilities underlie the importance of recent range expansion patterns for M. leidyi. Invasion of regions outside its historical distributions have resulted in dramatic planktonic community alterations in regions such as the Black Sea (Shiganova et al. 2003) and Sea of Azov (Studenikina et al. 1991). Although perhaps less acclaimed than these spatial range expansions, records of temporal range expansion within its endemic range can also cause important changes in planktonic community dynamics (Sullivan et al. unpubl. data). For example, within Narragansett Bay, peak occurrence of M. leidyi has shifted approximately 2 months earlier than the historic mean (Sullivan et al. 2001). However, the historically dominant summer copepod, Acartia tonsa, has not experienced a similar phenological shift, with the result that the seasonal timing of predator (M. leidyi) and prey (A. tonsa) overlap differently than during the past. One result of this change is that A. tonsa has been almost eliminated from the plankton during recent summers in Narragansett Bay as a result of predation pressure from M. leidyi (Sullivan et al. unpubl. data). The long-term trophic consequences of near removal of copepods from Narragansett Bay during summer months, historically a period of high copepod abundance, are not yet clear. However, there is evidence of reduction in numbers of some species of larval fish in recent years (Keller et al. 1999) and increases in summer values of chlorophyll a (Chl a) (Sullivan et al. unpubl. data). Despite the potentially important consequences of phenological shifts by M. leidyi, the mechanisms underlying M. leidyi seasonality are not well documented. Metabolism and growth of M. leidyi are clearly influenced by temperature (Kremer 1977), and climatic change has been suggested (Sullivan et al. 2001) to underlie recent alterations in M. leidyi seasonality. The relationship is supported by a positive correlation between periods of seasonal advance in M. leidyi abundance and average sea surface temperature increases. However, the overall change in average annual temperature of 2uC over the last 50 yr in Narragansett Bay (Hawk 1998) appears relatively small to result in such dramatic seasonality alterations. Likewise, it remains unclear why this level of warming has substantially altered M. leidyi’s seasonality but not that of other cooccurring species. We designed this program in order to describe seasonal patterns in abundance and to clarify the mechanisms enabling temporal range expansion by M. leidyi. Previous research indicated that a general inshore–offshore gradient characterized the summer increase in M. leidyi abundance and that seasonal population growth occurred first at interior regions of Narragansett Bay (Kremer and Nixon 1976). Our choice of sample locations reflected this Acknowledgments We thank A. Allen, C. Blackwelder, H. Chang, P. Ritacco, and P. Schilling for field assistance during this research. P. Norton allowed use of Norton’s Marina dock for sampling in Greenwich Cove. J. Frietag provided a meter readout; URI’s Graduate School of Oceanography Marine Technical Service provided Niskin bottles. Tom Puckett operated the Capn’ Bert weekly through all conditions for the entire duration of sampling. We are grateful to L. Mullineaux and two anonymous reviewers for their comments and suggestions. Financial support for this research was provided by the National Science Foundation (OCE 9103309 to J.H.C., OCE 0115177 to B.K.S., D.J.G.). 1 Corresponding author ([email protected]). Limnol. Oceanogr., 51(4), 2006, 1819–1831 E 2006, by the American Society of Limnology and Oceanography, Inc.