Temporal variation in winter flounder recruitment at the southern margin of their range: is the decline due to increasing temperatures?

ICES Journal of Marine Science ICES Journal of Marine Science (2014), 71(8), 2186– 2197. doi:10.1093/icesjms/fsu094 Contribution to the Special Issue: ‘Commemorating 100 years since Hjort’s 1914 treatise on fluctuations in the great fisheries of northern Europe’ Original Article Temporal variation in winter flounder recruitment at the southern margin of their range: is the decline due to increasing temperatures? K. W. Able*, T. M. Grothues, J. M. Morson, and K. E. Coleman Marine Field Station, Rutgers, the State University of New Jersey, 800 c/o 132 Great Bay Boulevard, Tuckerton, NJ 08087, USA *Corresponding author: tel: +1 609 296 5260 X230; fax: +1 609 296 1024; e-mail: Able, K. W., Grothues, T. M., Morson, J. M., and Coleman, K. E. Temporal variation in winter flounder recruitment at the southern margin of their range: is the decline due to increasing temperatures? – ICES Journal of Marine Science, 71: 2186 –2197. Received 27 August 2013; revised 29 April 2014; accepted 30 April 2014; advance access publication 10 June 2014. The southern-most stock of winter flounder (Pseudopleuronectes americanus), a cold temperate species of the Northwest Atlantic, has not recovered from overfishing despite continued restrictive measures, and appears to be contracting northward. We regressed larval and settled juvenile abundance (after accounting for adult and larval contribution to variation, respectively) on temperature over several decades from collections in New Jersey, the United States, at the southern edge of their range to determine if increasing temperatures during the first year of life were responsible for this contraction. A significant stock – recruitment relationship at both stages was moderate, explaining 27.5% of the variance for larvae on adults and 20.6% for juveniles on larvae. There was no significant effect of average monthly temperature in explaining variance of the residuals for larvae, or of degree day on explaining the abundance of residuals for juveniles over a months-long settlement period. However, in both cases, residuals were widely distributed at cold temperatures, while they were always low at warm temperatures. Thus, years in which spring temperatures were warm (5 –7oC for February, 7 –9 for March, and 11 –20 for May) always experienced poor recruitment. This threshold effect may result from an intersection with predators in response to temperature, and this may play a more important role than heat stress in determining recruitment success. Keywords: climate change, larvae, recruitment, temperature, temporal variation, winter flounder. Introduction With prompting by Hjort (1914, 1926), there has been extensive literature addressing our inability to understand the factors influencing fish recruitment (Browman, this volume; Chambers and Trippel, 1997; Houde, 2008; Doyle and Mier, 2012). An emerging consensus is that there are many factors, especially in the early life history stages, that contribute to this variability. These include natural mortality of the larvae and juveniles due to starvation, predation, adverse transport, and habitat degradation, potentially including climate change, and others (see other papers in this volume). It is also possible that adults may be overfished to the point that reproduction is hampered by depensation, (Liermann and Hilborn, 2001; Marato and Moran, 2014). Many of these factors may be contributing to the decline in spawning-stock biomass (SSB) and landings (Figure 1a and b) of winter flounder (Pseudopleuronectes americanus), from the Southern New England–Mid-Atlantic (SNE/MA) stock. The species is commercially and recreationally important along the western North Atlantic margin. Overfishing was credibly blamed for the precipitous decline of this stock in the 1980s, but poor recruitment rather than overfishing in the late 1990s and 2000s contributed to its failure to recover (Northeast Fisheries Science Center, 2011). This species has been reported from Georgia (the United States) to Labrador (Canada) and is cold adapted with antifreeze proteins (Collette and Klein-MacPhee, 2002). National Marine Fisheries Service (NMFS) trawl surveys indicate that the southern limit of the range is between Cape Hatteras, North Carolina, and Cape Cod, Massachusetts (the United States; Able and Fahay, 2010) and the centre of abundance is north of New Jersey (Perlmutter, 1947; Scott and Scott, 1988). This is supported by evaluations in New # International Council for the Exploration of the Sea 2014. All rights reserved. For Permissions, please email: Temporal variation in winter flounder recruitment 2187 Figure 1. Annual variation in SSB (line) and recruitment (bars) (a), and landings (b) for the Southern New England– Mid-Atlantic Bight stock, temperature behind Little Egg Inlet (c), and catch per unit effort (CPUE) for larval abundance at Little Sheepshead Creek behind Little Egg Inlet (d). See Figure 2 for New Jersey locations. 2188 Jersey, which have consistently indicated that the juveniles and adults are less abundant in the southern and more abundant in the northern part of the state (Scarlett, 1991; Scarlett and Allen, 1992; Sogard et al., 2001). A species-wide and local population shift in the distribution appears correlated with a decline in abundance for the SNE/MA stock (Collie et al., 2008; ASMFC, 2013). The decline is evident in the change in SSB, a frequently used index of population status (Danila, 2000; Brodziak et al., 2001), and landings from a long timeseries from NMFS (Figure 1a and b). The centre of the adult distribution has shifted further north with warming temperatures (Nye et al., 2009; Lucey and Nye, 2010; Pinsky and Fogarty, 2012). Larvae for populations in the SNE/MA stock are hatched from benthic eggs deposited in estuaries and juveniles continue to grow in these natal estuaries and be exposed to temperature fluctuations there (Jeffries and Johnson, 1974; Laurence, 1975; Rogers, 1976; Chant et al., 2000; Keller and Klein-MacPhee, 2000; Sogard et al., 2001; Curran and Able, 2002; Manderson et al., 2007). Thus, recruitment success is expected to be constrained by local (watershed scale) phenomenon; yet, the annual abundance of juveniles in southern New England has become synchronized by very large-scale events (Manderson, 2008). This possibly homogenizes an otherwise diverse recruitment portfolio (Secor, 2007; Kerr et al., 2010) and destabilizes population dynamics. Over the same period, estuarine temperatures have increased with fewer cold winters and increasing annual average water temperatures (Figure 1c, Able and Fahay, 2010) raising the suspicion that climatic factors may have played a role, at least in New Jersey, during 1989–1999 (Sogard et al., 2001). Despite these changes, there was no evidence of a decline in larval abundance at a site within the Great Bay –Barnegat Bay estuaries over the period from the late 1980s to 2006 (Figure 1d, Able and Fahay, 2010). In fact, larval abundance ha (...truncated)