Marine sublittoral benthos fails to track temperature in response to climate change in a biogeographical transition zone

ions from gridded SST datasets (Hawkins et al., 2003; Saulquin and Gohin, 2010; Smyth et al., 2010). Covering different periods and scales, these methods were more or less suitable to assess the effects of climate change on the benthic macrofauna at a regional scale. Although very useful to depict the long-term trend in climatic conditions, in situ measurements only provide local observations. Since the mid-1980s, remote sensing have provided monthly climatologies at a larger spatial scale (Saulquin and Gohin, 2010; Smyth et al., 2010), but have also displayed systematic differences from in situ observations (Parker et al., 1995). Therefore, interpolated SST data on a 1 geographic grid from the HadISST1 dataset (Rayner et al., 2003) have often been used in studies on the effects of climate variability on benthic communities at regional scales in the NE Atlantic (Hawkins et al., 2003, 2008; Southward et al., 2005; Hinz et al., 2011; but see Hiddink et al., 2015). Mainly based on quality controlled in situ SST observations, SST datasets provide widely available data used as a proxy for bottom temperature over larger geographical areas than the extremely sparse and access limited long-term in situ timeseries (Hughes et al., 2009). However, they only offer a relatively coarse resolution when working at the regional or local scale and may overestimate the temperature in waters that can experience stratification during summer months as reported in the westernmost part of the English Channel (Smyth et al., 2010; Marrec et al., 2013). Reflecting the climate change really experienced by benthic organisms at a regional scale and with a rather fine resolution, our results showed a clear warming trend of SBT between 1985 and 2012 for both minimum and maximum temperatures, in broad agreement with results obtained for SST using previously mentioned methods. A non-uniform warming was also highlighted, with September temperature increases ranging from 0.07 C per decade in the western end of the Channel to 0.50 C per decade in the eastern part. This difference between the most western areas and the eastern half of the Channel has previously been noted by Saulquin and Gohin (2010) who attributed it to the interplay between local physical and hydrodynamic conditions on the observed warming of the water masses. Generally in the English Channel, September and March isotherms longitudinally shifted at mean rates of 3.2 [0.2–8.4] km.year 1 to the west and 2.5 [1.4–3.4] km.year 1 to the east respectively which is more than the value reported by Hiddink et al. (2015) in the North Sea for the maximum temperature (mean 1.2 km.year ) and slightly less than the one reported for the minimum temperature (mean 3.3 km.year ). These changes in temperature could partly be associated with the changes in climatic conditions reported in the NE Atlantic over the last decades (Dippner et al., 2014). These include the climate regime shift that occurred in 1988–1989 in response to a shift in the North Atlantic Oscillation (NAO) from a persistent negative phase, characterized by cold winters, starting in 1977, to a persistent positive Table 1. continued Species 1959–1976 2012–2014 Observed change Total T A T B T C Total T A T B T C Total T A T B T C Alpheus macrocheles 20 11 5 4 29 12 10 7 Unclear trend " " " Echinocyamus pusillus 49 26 17 6 66 35 20 11 " " " Pagurus bernhardus 46 21 13 12 31 13 11 7 Psammechinus miliaris 49 29 14 6 34 16 15 3 # " # Species not classified A. prismatica 4 0 0 4 12 1 3 8 50% increase " " " Eurynome spinosa 2 2 0 0 17 11 0 6 " $ " Inachus leptochirus 9 8 1 0 15 10 4 1 " " " L. marmoreus 13 1 5 7 " " " Steromphala cineraria 20 7 8 5 22 7 9 6 <20% change $ " " T. ovata 64 33 21 10 61 38 14 9 " # # O. erinaceus 49 26 12 11 35 17 11 7 Unclear trend Pisa armata 13 12 0 1 16 7 5 4 # " " CW, Cold-water species; WW, Warm-water species; T, transect. 1902 F. Gaudin et al.