ICES Journal o f M arin e Science (2011), 68(6), 1 0 9 0 -1 1 0 4 . doi:10.1093/icesjm s/fsr031 Nine decades of North Sea sole and plaice distribution G eorg H. E n gelh ard 1* John K. P in n eg a r1, Laurence T. K eli1'2, an d Adriaan D. R ijnsdorp3'4 l Centre fo r Environment, Fisheries and Aquaculture Science (Cefas), Pakefield Road, Lowestoft NR33 OHT, UK 2International Commission fo r the Conservation o f Atlantic Tunas (ICCAT), C/Corazón de Maria, 8, 28002 Madrid, Spain ^Wageningen Institute fo r Marine Resources and Ecosystem Studies (IMARES), IJmuiden, The Netherlands 4Aquaculture and Fisheries Croup, Wageningen University, PO Box 338, 6700 AH Wageningen, The Netherlands *Corresponding Author: tel: + 44 IS02 527747; fax: + 44 IS02 513865; e-mail: georg.engelhard@cefas.co.uk. Engelhard, C. H., Pinnegar, J. K., Keli, L. T., and Rijnsdorp, A. D. 2011. Nine decades of N orth Sea sole and plaice distribution. - ICES Journal of Marine Science, 68: 1090-1104. R eceived 1 July 2010; a c c e p te d 13 Feb ru ary 2011; a d v a n c e a ccess p u b lic a tio n 11 A pril 2011. R ecent stu d ies based m ainly o n research survey d a ta suggest th a t w ithin th e N orth Sea, sole Solea solea a n d plaice Pleuronectes platessa have e xhibited d istrib u tio n shifts in re c e n t d e c a d e s— on average so u th w a rd for sole a n d n o rth w a rd to d e e p e r w aters for plaice. V arious h y p o th eses m ay a c c o u n t for such shifts, including clim ate ch an g e effects a n d m o re intensive fishing in so u th e rn a n d shallow er w aters; have m ad e use o f a un iq u e d a ta se t o f c atch a n d effort d a ta for British N o rth Sea traw lers; th ese cover nine d e ca d es (sp an n in g th e p eriod 1 9 1 3 -2 0 0 7 ) a n d are spatially d etailed by ICES rectangle (0.5° latitu d e by 1° longitude). W e quantify, for th e first tim e, long-term d istrib u tio n changes o f N o rth Sea sole a n d plaice over a p eriod ap p ro ac h in g a century, a n d d e m o n s tra te th a t th e distri­ b u tio n shift in plaice w as a ttrib u ta b le to clim ate ch an g e ra th e r th a n to fishing, b u t th a t b o th clim ate a n d fishing played a role in th e d istrib u tio n shift o f sole. T he discussion also highlights th e p o ten tial im p a c t o f additio n al factors, including e u tro p h ica tio n , prey availability, a n d h a b ita t m odification. Keywords; climate, depth, fishing, latitude, longitude, plaice, sole. Introduction T em perature is o n e o f th e p rim a ry factors, along w ith food avail­ ability a n d th e pro v isio n o f suitable spaw ning a n d n u rsery grounds, th a t d e te rm in e fish d istrib u tio n p atterns, a n d m o st fish species te n d to prefer a specific te m p e ra tu re range (C o û tan t, 1977; R ijnsdorp et al., 2009). C onsequently, shifts in lo n g -term tem p e ra tu re resulting fro m clim ate change are expected to result in contractio n s, expansions, o r shifts in fish d istrib u tio n , especially near the p e rip h e ry o f a species’ range (review: P innegar et al., 2010). In the N o rth Sea, a w a rm ing tre n d has h a p p en e d over the past th ree decades, w ith a p articu larly steep tem p e ra tu re rise in 1 9 8 8 -1 9 8 9 , a lth o u g h th ere w ere also som e relatively cold years; this has coincided w ith a n o rth w a rd shift in th e d istrib u tio n o f m any, b u t n o t all, fish species in the N o rth Sea (Beare et al., 2004; H edger et al., 2004; P erry et al., 2005), som etim es b y as m u c h as 400 k m . D ulvy et al. (2008) d em o n strated , however, th a t there has b e en n o n e t shift in the m ea n latitu d e o f th e fish assem blage as a w hole (because som e w ere m oving n o rth a n d som e w ere m o ving so u th ), b u t th ere has been a d eepening a n d this has h a p p en e d at a rate o f a ro u n d —3.6 m per decade (D ulvy et al., 2008). In th e N o rth Sea, plaice Pleuronectes platessa a n d sole Solea solea are principally cau g h t in a m ixed flatfish beam -traw l fishery. L andings o f plaice b y w eight are appro x im ately five tim es greater th a n those o f sole; how ever, because sole are c o n ­ siderably m o re valuable, landings o f the tw o species are roughly o f equal value overall (Pilling et al., 2008). Plaice a n d sole have © C row n c opyright [2011] b een targeted b y fishing fleets since th e early n in ete en th c en tu ry a n d conseq u en tly good fishery statistics exist for these species, unlike for m an y others. B oth species have displayed d istin ct d istri­ b u tio n shifts since th e 1980s. F or plaice, a shift to m o re offshore w aters has b e en re p o rte d (van Keeken et al., 2007), at a “deep en ­ in g ” rate o f - 3 .9 6 m p e r decade (D ulvy et al., 2008). For sole, P erry et al. (2005) re p o rte d a generally so u th w ard d istrib u tio n shift, i.e. o p p o site to the an tic ip a te d response to clim ate change in th e n o rth e rn hem isphere, assum ing a latitu d in a l g radient o f sea­ w ater tem p eratu res, a n d th e a u th o rs suggested th a t this m ig h t relate to im p ro v ed e n v iro n m en ta l co n d itio n s in th e so u th e rn N o rth Sea, because o f ru n -o ff fro m rivers (i.e. h igher p ro d u c tiv ity in recent years). T here has also b een a c o n c o m ita n t shift to shal­ low er w aters, a t a rate o f + 7 .6 4 m per decade (D ulvy et al., 2008). H ow ever, in te rn atio n al fisheries landings d a ta for the m o st recent years tentatively suggest a n o rth w a rd range expansion for sole (R ijnsdorp, 2010), so th e re lationship betw een d istrib u tio n a n d clim ate in this w arm -w ater species rem ains p o o rly u n d e rsto o d a n d u n c ertain (D ulvy et al., 2008). M ost studies o n N o rth Sea fish d istrib u tio n have so far been based o n fishery -in d ep en d en t survey d a ta from go v ern m en t in sti­ tutes a n d have been lim ited to the m o st recent th ree to four decades (H edger et al., 2004; P erry et al., 2005; H id d in k a n d ter H ofstede, 2008), because o f scarcity o r absence o f older, consist­ ently collected survey d a ta w ith reasonable spatial coverage. O ver this p e rio d (1 9 7 0 -2 0 1 0 ), how ever, sea tem p eratu res in the N o rth Sea generally have increased in th e so u th a n d at th e sam e tim e fishing pressure has been consistently h igher in th e so u th e rn Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 b u t th e relatively sh o rt tim e-sp an o f d a ta se ts analysed so far ( ~ 3 decades) has c o m p lica te d th e sep aratio n o f th ese tw o effects. W e N in e decades o f N orth Sea sole a n d plaice distribution p a rt th a n farth er n o rth (Jennings et al., 1999). T his m akes it p a r­ ticularly h a rd to disentangle the p o ten tial effects o f clim ate change a n d fishing pressure o n d istrib u tio n shifts (and these hypotheses are n o t m u tu a lly exclusive), because a greater rate o f fishery-induced d e p le tio n in the so u th e rn N o rth Sea th a n the n o rth w o u ld look like a n a p p are n t “n o rth w a rd ” shift, as m ight be an ticip ated to have h a p p en e d u n d e r clim ate change. T his diffi­ culty in in te rp re ta tio n has m o tiv ated th e c u rre n t stu d y b ased o n a u n iq u e d ataset o f B ritish co m m ercial catch per u n it effort (cpue) records sp a n n in g a far longer tim e -sp a n — th e past n in e decades— covering b o th w arm in g a n d cooling periods a n d inclu d in g periods o f c o n trastin g levels o f fishing effort. H ere we quantify, for th e first tim e, lo n g -term d istrib u tio n changes o f N o rth Sea sole a n d plaice over a p e rio d ap p ro ach in g a century. W e in te rp re t the findings in th e light o f clim ate change a n d lo n g -te rm changes in fishing pressure. W e test the hypotheses that: 2. L onger tim e-series o f fish d istrib u tio n d a ta offer greater insight in to likely m ec h an ism s/in flu e n c es th a n sh o rte r tim e-series derived from fish ery -in d ep en d en t surveys. were calculated b y rectangle for a large area encom passing m o st o f th e N o rth Sea (shaded in Figure 1). As a n ap p ro ac h to quan tify lo n g -te rm changes in p o p u latio n d istrib u tio n , we calculated the “centres o f gravity” o f the la titu d i­ nal, lo n gitudinal, a n d d e p th d istrib u tio n s o f the tw o species (as in H ein o et al., 2003; see also R in d o rf a n d Lewy, 2006). T his analysis was b ased o n a n area w ith in th e N o rth Sea th a t inclu d ed o nly those rectangles w ith cpue d a ta for th e m o st years in the tim e-series (see th e polygon line in Figure 1). W ith in this polygon, th e latitu d in al (o r lo n g itu d in al) centre o f gravity o f d istrib u tio n in a given year was calculated as the average o f the latitudes (o r longitudes) o f all rectangle centres, w eighted b y the cpue value in each rectangle. W eighted sta n d ard deviations a n d sta n d ard erro rs o f th e w eighted m ea n latitudes w ere calculated (Bevington, 1969). T he centres o f gravity o f d e p th d istrib u tio n s w ere calculated analogously and, given th a t b o th flatfish species are bottom -dw ellers, based o n the m ea n sea d e p th in a n y given rectangle. W e exam ined sole a n d plaice d istrib u tio n in relatio n to (i) clim ate variables a n d (ii) fishing pressure. As a broad-scale clim ate in dicator, the N o rth A tlantic O scillation (N A O ) w inter in d ex (D ecem ber o f th e previous year to M arch o f th e focal year) for 1 9 1 3 -2 0 0 7 was tak en from Jones et al. (1997), w ith 3. D ata fro m co m m ercial fisheries can yield useful insight in to the lo n g -te rm im plications o f clim ate change. Methods For the p e rio d 1 9 1 3 -1 9 8 0 , d a ta w ere o b tain e d fro m “statistical ch arts” (catalogued in E ngelhard, 2005) th a t w ere p ro d u c ed by th e UK M in istry o f A griculture, Fisheries a n d Food [MAFF; no w th e UK D e p a rtm e n t for E n vironm ent, F ood a n d R ural Affairs (D efra)]. T hese display fishing effort (h o u rs fished) a n d fish la n d ­ ings b y B ritish o tte r traw lers (either steam - o r m o to r-d riv en ) for each ICES rectangle (0.5° latitu d e b y I o lo ngitude) in the N o rth Sea. These d a ta reco rd all fish th a t w ere lan d ed b y the o tte r traw l fleet in to E ngland a n d W ales (1913, 1 9 6 8 -1 9 8 0 ) o r in to E ngland, Scotland, a n d W ales (1 9 2 0 -1 9 6 7 ). For the p erio d 1 9 6 8 -2 0 0 7 , d a ta o n o tte r traw ler landings in to S cotland were o b tain e d fro m th e Fisheries M an ag em en t D atabase o f M arine S cotland (cf. G reenstreet et a í, 1999). F or 1 9 8 2 -2 0 0 7 , data o n o tte r traw l landings in to E ngland a n d W ales w ere o b tain ed from th e Fisheries A ctivity D atabase o f D efra/C efas. O ver th e tim e -sp a n exam ined, im p o rta n t im p ro v em en ts have h a p p en e d in th e flatfish catching pow er, o r technical efficiency, o f o tte r traw lers (W im penny, 1953; R obinson, 2000), a lth o u g h a recent lo n g -te rm analysis suggested th a t plaice fishing pow er o f m o to r o tte r traw lers a ro u n d the M ille n n iu m was n o t m arkedly h igher th a n th a t o f steam o tte r traw lers o f the 1920s (b u t an o rd e r o f m ag n itu d e low er th a n in b eam traw lers; Engelhard, 2008). O u r a im was n o t to analyse tem p o ral changes in absolute cpue values, b u t ra th e r to look at tren d s in spatial d istrib u tio n o f catches. W e therefore no rm alized th e cpue values in any given year (divided b y th e a n n u al m ean ), to overcom e th e confusing effect o f a n increase in fishing pow er. W e assum ed th a t relative cpue b y the co m m ercial fleet gives an a p p ro p ria te in d ic a tio n o f the spatial d istribution o f the species, b u t acknowledge th at p otential bias m ight arise from uneven spatial d istribution o f effort b y m ore o r less pow erful vessels w ithin the N o rth Sea. By decade, cpue values Figure 1. M ap of th e N orth Sea displaying, for each rectangle, the num ber of years in th e 77-year time-series w here British o tte r trawler cpue data were available (based on th e rule th a t a t least 10 h of fishing effort had occurred in the rectangle in the given year). The grey-shaded area encom passes those rectangles where we calculated cpue by year or decade for spatial distribution maps. The thick-lined polygon encom passes those rectangles included in analyses on centres of gravity of cpue distributions. Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 1. L ong-term shifts in sole a n d plaice d istrib u tio n are b e tte r explained b y clim ate variables th a n b y indices o f fishing pressure. 1091 1092 G. H . Engelhard et al. u p d a te d values p ro v id ed o n lin e b y th e C lim atic Research U nit, U niversity o f East A nglia, N orw ich, U K (w w w .cru.uea.ac.uk/ ~ tim o /p ro jp a g e s/n a o _ u p d a te .h tm ; Figure 2a). T he NA O is associated w ith speed a n d d irec tio n o f w esterly w inds across the N o rth A tlantic a n d is particu larly im p o rta n t in w in ter w h en it exerts a stro n g influence o n E u ro p ean w eather p a tte rn s a n d o n A tlantic w ater inflow in to th e N o rth Sea; a positive NA O is gen er­ ally linked w ith stro n g w in d c irculation a n d h igher atm ospheric a n d sea tem p eratu res in w estern E urope (H urrell, 1995; Jones et al., 1997; O ttersen et al., 2001). W e also considered the A tlantic M ultidecadal O scillation (A M O ), a clim ate m o d e th at m anifests itself as a 2 0 -3 0 -y e a r cycle in d e -tre n d ed sea surface tem p e ra tu re series for th e N o rth A tlantic. These d a ta (back to 1871) w ere o b tain e d fro m NOAA (G lobal C hange M aster D irectory). A n u m b e r o f lo n g -te rm coastal tem p e ra tu re tim eseries w ere exam ined fro m sites a ro u n d th e N o rth Sea, including those fro m M arsdiep (D u tc h coast; o b tain e d fro m M ackenzie a n d Schiedek, 2007), H elgoland R oads (G erm an coast; o b tain e d fro m W iltshire a n d M anly, 2004), a n d D over (U K coast; o b tain e d fro m w w w .cefas.co.uk/data.aspx). As a m o re generic in d ic a to r o f sea te m p e ra tu re v ariations w ith in th e N o rth Sea, the H adley in te rp o lated sea surface tem p e ra tu re (H adlSST ) tim e-series was used (Figure 2b). A n n u a l m ea n o f sea surface tem peratures, in te rp o lated to I o latitu d e b y I o longitude, was used as described b y R ayner et al. (2003), w ith u p d a te d values p rovided online by th e UK M eteorological Office. For seabed te m p e ra tu re (SBT), n o com prehensive tim e-series covering th e en tire span o f this stu d y was available, b u t fo r a sh o rte r p e rio d (1980 o n ), w in ter SBT d a ta w ere calculated as average SBT in 80 ICES rectangles sam pled d u rin g the a n n u a l w in ter In tern a tio n a l B o tto m Trawl Surveys (IBTS) o f ICES (o b tain ed from D ulvy et a í , 2008). To describe the effects o f fishing pressure, estim ates o f fishing m o rta lity (F ) o n 4 -8 -y e a r-o ld plaice for the years 1 9 2 9 -1 9 3 7 a n d 1 9 4 7 -1 9 9 2 w ere tak en from R ijnsdorp a n d M illn er’s (1996) lo n g -te rm stu d y o n N o rth Sea plaice stock dynam ics. Estim ates o f F for 4 -8 -y e a r-o ld sole for the years 1 9 5 7 -1 9 9 3 w ere taken 05 1920 1940 1960 Y ear 1980 1920 2000 1940 1960 1980 2000 Year (C) (d) o CO to 00 * t/> 0) o> jo u. 05 Ç5 JO a. o C\J 1920 1940 1960 Y ear 1980 2000 1920 1940 1960 1980 Year Figure 2. Time-series of environm ental variables and fishing pressure exam ined here for possible relationships with N orth Sea sole and plaice distributions, (a) NAO w inter index; (b) Hadley interpolated annual average sea surface tem perature for the N orth Sea; (c) N orth Sea sole fishing mortality, averaged over ages 4 - 8 years (from Millner and Whiting, 1996); and (d) N orth Sea plaice fishing mortality, averaged over ages 4 - 8 years (from Rijnsdorp and Millner, 1996). Long-term variability is illustrated by heavy solid lines, representing values sm oothed with a low-pass filter with five weights (1, 3, 4, 3, and 1) to remove fluctuations with periods < 3 years (following Hurrell, 1995). Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 (b) 1093 N in e decades o f N orth Sea sole a n d plaice distribution 7 - 1920s 1930s L a te 1 1950S 1960s 1970S 1990s 2000s 7 1980s Figure 3. Long-term changes in relative sole cpue within th e N orth Sea. For each decade, spatial distribution of sole cpue by British trawlers within th e grey-shaded area is indicated by th e area sizes of the black circles (proportional to cpue). In rectangles where no cpue data were available in a given decade (no effort), white circles represent th e long-term average cpue. For each decade, the w hite cross indicates the centre of gravity of sole distribution, with its standard error (shorter, thick w hite lines) and standard deviation (longer, thin white lines) in the longitudinal and latitudinal directions. Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 • • • 1094 G. H . Engelhard et al. (a) Z o ^ in <D ■ö 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 1920 1930 1940 1950 1960 1970 1980 1990 2000 ■sr - CM - 1910 (C) Q<X> o n ---------------------------1-------------------------- 1-------------------------- 1---------------------------1-------------------------- 1---------------------------1-------------------------- 1---------------------------1-------------------------- 1— 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Y ear Figure 4. Long-term changes in (a) latitudinal and (b) longitudinal centre of gravity of N orth Sea sole distribution (calculated as th e m ean latitude and longitude weighted with sole cpue; bars indicate s.e. of weighted means); and (c) long-term changes in m ean depth distribution of N orth Sea sole (with s.e.). from M illner a n d W h itin g (1996). B oth F tim e-series were extended w ith recent d a ta from ICES W o rking G roup rep o rts (ICES, 2009). These fishing m o rta lity tim e-series are illustrated in Figure 2c a n d d. W e used c o rre latio n s as a first a p p ro a c h to explore w hich e n v iro n m e n ta l a n d / o r fishing pre ssu re variables m ig h t be associated w ith d e sc rip to rs o f sole a n d plaice d is trib u tio n (latitu d in a l, lo n g itu d in al, a n d d e p th ). P e a rso n ’s c ro ss-m o m e n t co rre latio n s (>p) w ere used, because n o n e o f th e variables displayed d istrib u tio n s significantly differen t fro m n o rm a lity (o n e -sam p le K o lm o g o ro v -S m irn o v tests, p > 0.05). T here was, how ever, a m o d e ra te o r w eak a u to c o rre la tio n b etw een several tim e-series variables. To a cc o u n t fo r this, th e test p r o ­ c ed u re for significance o f c o rre latio n s was a d ju ste d follow ing P yper a n d P e te rm a n [1998, E q u a tio n (1 )], b y re d u cin g the “effective” degrees o f free d o m (a n d th erefo re p -v alu es) a cc o rd ­ ing to th e degree o f a u to c o rre la tio n (a d ju ste d p-v alu es are referred to here as p a¿ß . M ultiple linear regressions w ere th e n used as a n ap p ro ach to exam ine th e relative im p o rta n ce o f e n v iro n m en ta l a n d fishing pressure variables as d e te rm in a n ts o f sole a n d plaice d istrib u tio n . As explan ato ry variables, these m odels inclu d ed fishing pressure (sole F o r plaice F ), th e NAO w in ter index, A M O , a n d H adley SST (o th er sea tem p e ra tu re tim e-series w ere closely correlated w ith H ad ley SST, a n d w ere n o t inclu d ed to avoid th e p ro b lem o f m u lticollinearity). Initially, we started w ith a m o d el w ith all c o n ­ sidered e x planatory variables, b u t n o interactions. F or exam ple, for sole latitu d in a l d istrib u tio n , th e startin g m o d el was S o le la titu d e ~ S o le i7 + N A O + A M O + H adley SST. (1) H ow ever, o nly param eters th a t significantly c o n trib u te d to the fit (p < 0.05) w ere retain ed in th e m o d el fo rm u latio n s. Finally, we checked w h eth er any se c o n d -o rd e r interactio n s betw een clim ate a n d fishing pressure w o u ld significantly im p ro v e th e fit (p < 0.05). F or the final m odel, we give th e stan d ard ized ß coefficients, Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 CT> 1095 N in e decades o f N orth Sea sole a n d plaice distribution 1930s 1950s 1960s 1980s 1990s Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 1920s 2000s Figure 5. Long-term changes in relative plaice cpue within the N orth Sea. For each decade, spatial distribution of plaice cpue by British trawlers within the grey-shaded area is indicated by th e area sizes of th e black circles (proportional to cpue). In rectangles where no cpue data were available in a given decade (no effort), w hite circles represent the long-term average cpue. For each decade, the white cross indicates the centre of gravity of plaice distribution, with its standard error (shorter, thick w hite lines) and standard deviation (longer, thin w hite lines) in th e longitudinal and latitudinal directions. G. H . Engelhard et al. 1096 0) in ^ <0 Z o 1910 LU 1920 1930 1940 1950 1960 1970 1980 1990 2000 1920 1930 1940 1950 1960 1970 1980 1990 2000 CO o b tain e d after sta n d ard iza tio n (su b tractin g th e m ea n a n d dividing b y th e sta n d ard deviation) o f th e e x planatory variables. T he sta n ­ dard ized ß coefficients are indicative o f th e relative stren g th o f the relationships betw een e n v iro n m en t a n d fishing pressure w ith each fish d istrib u tio n m etric; it sh o u ld be n o ted , how ever, th a t they can n o t be directly in te rp rete d as su b stan tiatin g causative effects. Results Shifts in fish distributions O ver the past n in e decades, b o th sole a n d plaice displayed long­ term , m ultid ecad al changes in spatial d istrib u tio n . M agnitudes o f each shift regarding “centres o f gravity” o f latitu d e a n d longi­ tu d e w ere ~ l ° - 2 ° fo r b o th species (co rresp o n d in g to 142 km for plaice a n d 93 k m fo r sole); those o f d e p th d istrib u tio n s were ~ 2 0 m for plaice, < 1 0 m for sole. T he d irec tio n a n d tim in g o f shifts were, however, very different for the tw o species. Sole distribution shifts Sole have always b e en d istrib u te d in th e shallow so u th e rn a n d so u th eastern N o rth Sea, w ith o nly very sm all n u m b ers in the n o rth e rn N o rth Sea, w here they are absent fro m m o st rectangles. F igure 3 show s the d istrib u tio n o f sole cpue over the N o rth Sea b y decade, fro m the 1920s to th e 2000s. In the 1920s a n d 1930s, high sole cpue values w ere d istrib u te d evenly betw een the S o u th e rn B ight a n d G erm an Bight, b u t th e relative im p o rta n ce o f th e G erm an B ight increased d u rin g th e 1950s a n d 1960s. Since th e 1970s, th e im p o rta n ce o f th e G erm an B ight has decreased, a n d fro m th e 1980s on, sole cpue has been considerably h igher in th e S o u th e rn Bight, especially close to th e T ham es E stuary a n d so u th e aste rn E ngland. U nfortunately, w e lack B ritish o tte r traw ler cpue d a ta for th e so u th e aste rn m o st p a rt o f th e G erm an B ight for m u c h o f the 1990-2000s. N o rth Sea sole have exhibited m ark ed m ultid ecad al flu ctu ­ atio n s in th e ir latitu d in a l centre o f gravity (Figure 4a), w hich shifted fro m ~ 5 3 ° N in th e 1920s, n o rth w a rd s to ~ 5 4 ° N in the Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 Figure 6. Long-term changes in (a) latitudinal and (b) longitudinal centre of gravity of N orth Sea plaice distribution (calculated as the m ean latitude and longitude weighted with plaice cpue; bars indicate s.e. of weighted means); and (c) long-term changes in m ean dep th distribution of N orth Sea plaice (with s.e.). 1097 N in e decades o f N orth Sea sole a n d plaice distribution 1960s, th e n southw ards again to ~ 5 3 ° N in the 1990s (and to a so u th e rn m o st latitu d e o f 52.5°N in 2002). T here have b e en less sm o o th a n d m o re irregular shifts in lo n g itu d in al centre o f gravity (Figure 4b), n o tab ly a w estw ard shift fro m ca. 1960 to 2000 (reflecting increased relative im p o rta n ce o f th e T ham es E stuary region). T here has b e en a m o re lim ited shift in d e p th d is­ trib u tio n (Figure 4c), b u t this b ro ad ly parallels the latitu d in al p attern . N o rth Sea. F ro m th e 1950s to the 1980s, th ere was a n eastw ard lo n g itu d in al shift (Figure 6b) related to th e te m p o ra ry m u ch greater im p o rta n ce o f the e a s t-c e n tra l N o rth Sea in this period. T he centre o f gravity “ju m p e d ” w estw ards b y 3° lo n g itu d e from 1988 to 1995 (a n d b y > 1 ° in 1 9 9 3 -1 9 9 4 alone), the latter reflected a n a p p are n t p o p u la tio n collapse in th e e a s t-c e n tra l N o rth Sea, b u t also th e recently increased a b u n d an c e o ff Scotland. Correlations betw een environm ental and fisheries variables Plaice distribution shifts Plaice are generally m o re a b u n d a n t in th e so u th e rn a n d central pa rts o f the N o rth Sea th a n in the n o rth . T heir d istrib u tio n differs fro m sole in being o n average m u ch m o re n o rth e rn a n d encom passing m o st o f th e n o rth e rn N o rth Sea, w here, a lthough at low er cpue values, th ey are typically caught b y o tte r traw lers in m o st rectangles, except for th o se th a t com prise the deepest w aters in th e no rth w est a n d the N orw egian T rench (Figure 5). Plaice have over th e past n in e decades also exhibited m ajo r d is­ trib u tio n shifts, b u t q u ite different from th o se observed for sole (Figure 5). D u rin g th e 1920-1940s (a n d in 1913), the highest plaice cpue values w ere observed in the S o u th e rn B ight a n d the G erm an Bight, in rectangles closest to th e coast o f so u th eastern E ngland, th e N etherlands a n d G erm any, a n d w ith cpue values rap id ly decreasing n o rth w a rd s a n d northw estw ards. D u rin g the 1960-1980s, the e a s t-c e n tra l N o rth Sea becam e far m o re im p o r­ tan t, in clu d in g a n offshore area stretching fro m th e D ogger B ank n o rth to th e G reat a n d T ittle Fisher Bank, b u t w ith relatively low c pue along m o st o f th e E nglish a n d Scottish coasts. D u rin g the 1990-2000S, plaice cpue d ro p p e d m arkedly in the previously im p o rta n t e a s t-c e n tra l N o rth Sea, especially in m an y rectangles close to D e n m ark a n d G erm any, b u t rem ain ed high in the c entral N o rth Sea. In the (late) 1990-2000s, th ere was a sudden, m ark ed increase in relative plaice cpue o ff Scotland, O rkney, a n d Shetland. F ro m the sta rt o f o u r tim e-series in 1913 to W orld W ar II, the centre o f gravity o f N o rth Sea plaice d istrib u tio n rem ain ed c o n ­ sta n t regarding latitude, longitude, a n d d ep th . Since then, how ever, th e centre o f gravity has alm ost c o n tin u o u sly shifted n o rth w a rd s b y m o re th a n 2° latitu d e fro m th e late 1940s to the 2000s (Figure 6a). A ccordingly, plaice have m oved offshore to greater d ep th s b y ~ 2 0 m since 1947 (Figure 6c), tracking th e lati­ tu d in a l shift a n d reflecting th e n o r t h - s o u t h d e p th grad ien t in the Before p roceeding to exam ine co rrelations betw een clim ate v a ri­ ables a n d /o r fishing pressure vs. sole a n d plaice d istrib u tio n , it is necessary to de te rm in e w h eth er such variables are them selves correlated. A n analysis revealed th a t all sea surface a n d b o tto m tem p e ra tu re datasets (H adley, M arsdiep, H elgoland, a n d D over SST; IBTS w in ter SBT) w ere highly correlated (all rp> 0.75, all p a d j IO- 5 , adjusted fo r a u to c o rrela tio n ). Similarly, th e NAO w in ter in d ex exhibited a stro n g positive co rrelatio n w ith H adley SST (rp = 0.462, p¡¡¿j < IO- 7 ), b u t b y c ontrast, the A M O index was o nly weakly co rrelated w ith H ad ley SST (rp = 0.227, p adj = 0.04), despite itself being derived from the c o m b in ed H adley SST a n d N O AA O I SST datasets. Sole a n d plaice fishing m o rta lity w ere strongly correlated w ith each o th er (rp = 0. 678, p¡¡¿j< 0.0002), b u t n e ith e r was signifi­ cantly co rrelated w ith a n y o f the clim ate variables (all rp< 0.24, a ll P a d j> 0 .1 ) . Because all sea surface tem p e ra tu re datasets w ere highly c o rre ­ lated, fu rth e r analysis o n relationships betw een clim ate a n d sole o r plaice d istrib u tio n s was o nly carried o u t using the H ad ley SST d ataset (along w ith NA O a n d A M O ), because this was ju d g ed to provide a representative in d ex o f regional hydroclim ate. Distribution shifts in relation to clim ate and fishing All th ree m easures o f N o rth Sea sole a n d plaice d istrib u tio n — m ea n latitude, longitude, a n d d e p th — w ere strongly correlated w ith the H adley a n n u al in d ex o f sea surface tem p e ra tu re in the N o rth Sea (Table 1). T he d irec tio n o f co rrelations was such th a t for sole, w arm er tem p eratu res were associated w ith m o re so u th e rn , w estern, a n d shallow er d istrib u tio n p a tte rn s (Figure 7, left panels). Conversely, for plaice, h igher tem p eratu res were associated w ith a m o re n o rth w e stern a n d a deeper d istrib u tio n Table 1. Correlations (rp) betw een sole and plaice distribution in the N orth Sea (latitude, longitude, and d epth) and variables related to clim ate and fishing pressure. Latitudinal shift Variable Sole NAO w inter index AM O index Hadley SST Sole F (ages 4 - 8 ) Plaice NAO w inter index AM O index Hadley SST Plaice F (ages 4 - 8 ) rP - 0.246 -0 .0 5 2 0.355 0.781 Longitudinal shift Depth shift P adj rP P adj rP P adj 0.035 - 0 .1 7 4 0.017 0.138 0.887 -0 .2 2 5 -0 .1 0 7 0.054 0.374 0.393 0.631 <0.005 <0.001 -0.022 -0.0 5 4 0.850 0.652 0.666 <0.005 <0.0001 -0.015 -0.085 0.900 0.478 0.431 -0.098 0.348 0.826 <0.005 <0.0001 -0.085 0.467 <0.001 -0.530 -0.443 <0.0002 <0.0005 0.449 -0.115 0.375 0.455 -0.046 <0.0005 0.718 There was weak or moderate autocorrelation within these variables. To account for this, the test procedure for significance of correlations was adjusted according to autocorrelation following Pyper and Peterman [1998; Equation (1)]. Correlations different from zero at adjusted p < 0.005 are shown emboldened. G. H . Engelhard et al. 1098 (b). 0) m 'S s CO - (d) 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.4 0.5 0.6 0.7 0.8 o if o cô CM " 7» • y •• o O. . 9.5 10.0 10.5 11.0 11.5 Hadley SST (°C) Sole F (ages 4 -8 ) Figure 7. Relationships of Hadley SST (left) and sole fishing pressure (right) with m etrics of N orth Sea sole distribution: (a and b) latitudinal centre of gravity; (c and d) longitudinal centre of gravity; and (e and f) m ean d ep th distribution. Regression lines indicate significant relationships. p a tte rn (Figure 8, left panels). Each o f th e co rrelations w ith H adley SST was significant beyond th e p a¿, < 0.005 level. T he NAO w in ter in d ex was significantly correlated o nly w ith the latitu d in a l d istrib u tio n o f sole a n d n o t very strongly (p a¿j = 0.035); the co rrelatio n w ith the d e p th d istrib u tio n o f sole a p p ro ach ed sig­ nificance (pa¿j = 0.054), such th a t a m o re negative phase o f the NAO w o u ld be associated w ith m o re so u th e rn a n d p erhaps deeper d istrib u tio n o f sole (Table 1). T he NAO w in ter in d ex was n o t co rrelated w ith sole lo n g itu d in al d istrib u tio n a n d w ith n o n e o f th e m easures o f plaice d istrib u tio n . How ever, plaice lo n g itu d in al d istrib u tio n was strongly co rrelated w ith th e A M O in d ex (p¡¡dj< 0.0002), such th a t a m o re positive phase o f A M O was associated w ith a m o re w estern d istrib u tio n o f this species. Fishing m o rta lity o n sole (averaged over ages 4 - 8 years) was correlated highly significantly w ith th e latitudinal, lo n gitudinal, as well as d e p th d istrib u tio n (Table 2 a n d Figure 7, right panels). H ig h er fishing m o rta lity o n sole was associated w ith a n o n average m o re so u th e rn , w estern, a n d shallow er d istrib u tio n o f th e species. In c ontrast, fishing m o rta lity o n N o rth Sea plaice (Figure 7, rig h t panels) was n o t co rrelated significantly w ith either th e latitudinal, longitu d in al, o r d e p th d istrib u tio n o f the species. M ultiple regression analyses w ere co n d u cted to explore th e rela­ tive stren g th o f relationships o f clim ate a n d fishing pressure v ari­ ables w ith d istrib u tio n al responses o f sole a n d plaice (Table 2). R egression m odels co n firm ed th a t H adley SST was a h ighly signifi­ can t p re d ic to r o f th e latitudinal, longitu d in al, a n d d e p th responses o f b o th fish species (all p < 0.0005); also th a t fishing m o rta lity was a significant p re d ic to r for sole (p < IO- 5 ), b u t n o t plaice d istri­ b u tio n . T he NAO w in ter index, however, was rejected as a p re d ic ­ to r for sole o r plaice d istrib u tio n (p > 0.05) b y any regression m o d el th a t also inclu d ed H adley SST, the latter displaying far greater e x planatory pow er. In contrast, the A M O was retained as a significant p re d ic to r o f plaice lo n g itu d in al d istrib u tio n (p < IO- 5 ), also w ith th e in clu sio n o f H adley SST (a h igh A M O being associated w ith a w estw ard d isplacem ent). T he A M O was also a sig­ nificant p re d ic to r o f plaice latitu d in a l a n d d e p th d istrib u tio n , b u t o nly in m odels th a t inclu d ed H ad ley SST as covariate. T here was n o evidence th a t interactio n s betw een clim ate a n d fishing m o rta lity p ro v id ed a d d itio n al e x planatory pow er. For Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 (f) 1099 N in e decades o f N orth Sea sole a n d plaice distribution (a) (b) 9.5 10.0 10.5 11.0 0.6 0.4 0.2 11.5 0.8 1.0 (d) Tj. - LU • • CM - 10.0 10.5 11.0 0.2 11.5 0.6 0.4 0.8 1.0 (f) o o ♦♦ • i o in . ♦♦ o . co 9.5 10.0 10.5 11.0 11.5 Hadley SST (°C) 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Plaice F (ages 4-8) Figure 8. Relationships of Hadley SST (left) and plaice fishing pressure (right) with metrics of North Sea plaice distribution: (a and b) latitudinal centre of gravity; (c and d) longitudinal centre of gravity; and (e and f) m ean dep th distribution. Regression lines are included only w here correlations are significant (p < 0.05, adjusted for autocorrelation). sole, th e in te ra c tio n H adley SST x sole F d id n o t explain la titu d i­ nal, lo ngitudinal, o r d e p th responses significantly (a llp > 0.2). For plaice, the interactio n s H ad ley SST x plaice F (all p > 0.6) a n d A M O x plaice F ( all p > 0.4) d id n o t explain any o f the responses in d istrib u tio n m etrics. A c o m p a riso n o f th e stan d ard ized ß coefficients in final m odels (Table 2) lent su p p o rt to the n o tio n th a t betw een fishing m o rta lity a n d H adley SST, th e fo rm er was th e stronger p re d ic to r o f sole latitudinal, lo n gitudinal, a n d d e p th d istrib u tio n responses, a lth o u g h b o th variables w ere highly significant predictors. For plaice, w here th e effect o f fishing m o rta lity o n d istrib u tio n was insignificant, the H ad ley SST a n d A M O c o n trib u te d ap p ro x i­ m ately equally to th e d istrib u tio n responses. Discussion O ver th e past n in e decades, b o th sole a n d plaice exhibited long­ term , m ultid ecad al changes in spatial d istrib u tio n w ith in the N o rth Sea. M agnitudes o f each shift regarding “centres o f gravity” o f latitu d e a n d lo n g itu d e w ere ~ l ° - 2 ° for b o th species; th o se o f d e p th d istrib u tio n s w ere ~ 2 0 m for plaice, < 1 0 m for sole. T he d irec tio n a n d tim in g o f shifts w ere, however, very differ­ e n t in th e tw o species: since the 1950s m ain ly northeastw ards and to d eeper w aters in N o rth Sea plaice, b u t m ain ly southw estw ards a n d to shallow er w aters in N o rth Sea sole (by 142 a n d 93 km , respectively). B oth species displayed co rrelations w ith th e H adley SST, b u t in o p p o site directions; for sole, b u t n o t plaice, a close c o r­ relatio n w ith fishing m o rta lity was also observed. U nlike in recent studies based o n IBTS (P erry et al., 2005; D ulvy et al., 2008), w e used d a ta derived from co m m ercial catch records, so it is necessary to co n sid er w h eth er o r n o t these provide a representative a n d n o n -b iased p ictu re o f fish d istri­ b u tio n in th e past o r sim ply a reflection o f favoured fishing localities. Im p ro v em en ts in fishing pow er o f U K o tte r traw lers have clearly h a p p en e d (E ngelhard, 2008), b u t this p ro b lem was Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 o> G. H . Engelhard et al. 1100 Table 2. Regression models on distribution responses of N orth Sea sole and plaice (latitudinal, longitudinal, and d epth) to variables related to clim ate and fishing pressure. Response variable Predictor Sole latitudinal shift Hadley SST Sole F (ages 4 - 8 ) Hadley SST Sole F (ages 4 - 8 ) Hadley SST Sole F (ages 4 - 8 ) A M O Index Hadley SST A M O Index Hadley SST A M O Index Hadley SST Sole longitudinal shift Sole d ep th shift Plaice latitudinal shift Plaice longitudinal shift Plaice d ep th shift //standardized s.e. -0 .1 6 1 - 0.350 - 0 .1 4 8 - 0.473 - 1 .3 4 8 - 1 .9 9 3 - 0.209 0.322 - 0.423 - 0.298 - 1 .7 7 8 3.502 0.036 0.039 0.045 0.047 0.353 0.376 0.074 0.071 0.082 0.079 0.788 0.759 t p-value 4.426 9.045 3.330 9.974 3.821 5.303 2.828 4.529 -5 .1 8 1 - 3.786 - 2.256 4.614 0.00006 <0.00001 0.002 <0.00001 0.0004 <0.00001 0.006 0.00003 <0.00001 0.0003 0.028 0.00002 - Final models are presented following a backward selection procedure, from a full model th at included the NAO winter index, AMO, Hadley SST, and fishing mortality as main effects. Only the terms th at significantly explained the distribution metric (p < 0.05) were retained. None of the interactions between environmental variables and fishing pressure was significant (p > 0.05). Standardized ß coefficients are displayed here, obtained after standardization of the explanatory variables and indicative of the relative strength of these in explaining the response variable. responses. A lthough n o su p p o rt for a n interactive effect was fo u n d in this study, the im p o rta n ce o f n o t o nly exam ining clim ate a n d fishing effects separately, b u t also th eir p o ten tial in te r­ action, has b een em phasized recently (P erry et al., 2010). A lthough clear relationships w ere fo u n d betw een SST a n d flatfish d istri­ b u tio n s, the effects o f SBT co u ld n o t be fully explored, because o f th e lack o f data for m o st o f th e earlier decades; w here tim e-series overlapped, SST a n d SBT w ere closely correlated. N ote th at a lth o u g h sole a n d plaice are b e n th ic as a d u lts/la te juveniles, th eir eggs a n d early juveniles are fo u n d in the pelagic zone, close to th e sea surface. It is c u rre n tly unclear w h eth er the d o m in a n t clim ate-related im p a c t o n fish d istrib u tio n p a tte rn s h appens th ro u g h processes th a t occu r d u rin g th e early life-history stages (as suggested b y R in d o rf a n d Lewy, 2006), o r th ro u g h tem p e ra tu re toleran ces/p referen ces d u rin g th e a d u lt phase, a n d consequently w h eth er SST o r SBT is the m o st a p p ro p ria te e n v iro n m en ta l v ari­ able to consider. Plaice seem to have resp o n d ed largely to clim ate changes w ith a shift to the n o rth , the west, a n d to deeper w aters, fully in line w ith expectations assum ing a n o r t h - s o u t h SST g radient in the n o r th ­ e rn h em isphere (P erry et al., 2005; D ulvy et al., 2008; R ijnsdorp et al., 2009). T he lack o f a relationship w ith fishing m o rta lity suggests th a t clim ate is th e d o m in a tin g driver o n plaice d istri­ b u tio n , a n d n o t fishing. In the N o rth Sea, juvenile plaice are ty p i­ cally co n ce n tra ted in shallow inshore w aters a n d m ove gradually offshore as th ey b ecom e larger. Surveys in the W ad d e n Sea have d e m o n stra ted th a t 1-group plaice are n o w alm ost absent from th e area w here th ey w ere o nce very a b u n d a n t. T he “plaice b o x ” (an area o f the D u tc h a n d G erm an coast closed to m o st plaice fishing u n d e r th e EU C o m m o n Fisheries Policy since 1989: P astoors et al., 2000) is n o w considered m u ch less effective as a m an ag em en t m easure th a n 10 o r 15 years ago, a n d this has been a ttrib u te d to th e d istrib u tio n shift resulting fro m lo n g -term clim ate change (van Keeken et al., 2007). M arin e p ro tec te d area b o u n d aries, such as those associated w ith th e “plaice b o x ”, m ay have to b e “adaptive” in fu tu re a n d m ove w ith th e fish th ey are try in g to p ro tect. It is interesting to n o te th a t th e a p p are n t b im o d a l d istrib u tio n p a tte rn o f plaice d istrib u tio n in recent years (1990s a n d 2000s in Figure 5), such th a t high cpue values were reco rd ed in the central N o rth Sea, b u t also o ff n o rth ea st Scotland. T his su d d e n o ccurrence o f large n u m b ers o f plaice off Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 overcom e, because n o a tte m p t was m ad e to estim ate the absolute biom ass in any p a rticu la r area. Instead, relative spatial d istri­ b u tio n s o f cpue were used. Spatial d istrib u tio n s o f cpue co u ld be influenced b y the level o f targeting o f a species b y th e fleet (Gillis et al., 2008; Q uirijns et al., 2008). H ow ever, alth o u g h N o rth Sea plaice a n d particu larly sole are targeted extensively by b eam traw lers, th ey are m erely im p o rta n t bycatch species for the UK o tte r traw l fleet used here to provide cpue indices; this fleet p rim a rily catches ro u n d fish (e.g. cod a n d h ad d o ck ) in a m ixed fishery (B annister, 2004). F u rth e r evidence o f representativeness o f o u r cpue m ap s for “tru e ” sole a n d plaice a b u n d an c e d istrib u tio n s is p rovided by great sim ilarity w ith spatial d istrib u tio n s based o n IBTS available for recent decades (P erry et al., 2005; D ulvy et al., 2008). This being said, th e p a tte rn s in recent decades m ay be affected b y the c o m p e titio n betw een o tte r traw lers a n d b e am traw lers (R ijnsdorp et al., 2008), because a n analysis o f th e D u tc h dem ersal fleet since th e 1950s has revealed th a t th e o tte r traw l fleet has been o u tco m p e te d a n d largely displaced to d eeper w aters b y th e highly efficient b eam traw l fleet th a t o perates in this region. A lthough this w ould b e expected to affect the to ta l effort expended b y th e o tte r traw l fleet in a n area, w e w o u ld n o t generally an ticip ate a n im pact o n catch rates (i.e. cpue). H ow ever, co m m ercial fishers m ay in te r­ act o n local g ro u n d s a n d “interference c o m p e titio n ” betw een vessels has b e en observed. In such cases, this is k n o w n to affect the re lationship betw een cpue a n d fish stock biom ass (Gillis a n d P eterm an, 1998; Poos a n d R ijnsdorp, 2007). It has also b een im p o rta n t to establish th a t clim ate variables a n d fishing pressure are n o t significantly correlated, i.e. th a t the tw o factors are n o t c o n fo u n d ed , because this has b e en o n e o f the m ajo r concerns voiced a b o u t earlier studies (P erry et al., 2005; D ulvy et al., 2008). It was n o t surp risin g th a t th e various N o rth Sea te m p e ra tu re tim e-series were highly correlated, because this was also observed b y M ackenzie a n d Schiedek (2007). H ow ever, it is im p o rta n t a n d useful to n o te th a t in the long term , there w ere n o significant co rrelations betw een fishing m o rta lity (in eith er species) a n d a n y clim ate variable. T his offers the h o p e th a t a lo n g er-term analysis o f fish d istrib u tio n s could start to p in d o w n th e p rim a ry influences th a t are c o n fo u n d ed in the sh o rte r te rm a n d also the possibility o f a n y effects o f the in te r­ actio n betw een clim ate a n d fishing pressure o n d istrib u tio n N in e decades o f N orth Sea sole a n d plaice distribution affected plaice d istrib u tio n (Nye et al., 2009). Elsew here in the N o rth Sea, A ttrill a n d Pow er (2002) d e m o n stra ted th a t p attern s in the NAO coincided w ith v a ria tio n in th e stru c tu re o f th e fish assem blage, explaining 54% o f v ariation, a n d th a t th e grow th o f m an y juvenile fish, in clu d in g sole, w ere also influenced b y the NAO. A stu d y o f th e changing fish c o m m u n ity in th e B ristol C hannel (H enderson, 2007) identified tw o p e riods o f discrete change in the fish c o m m u n ity over th e past 25 years. T he first change h a p p en e d in the late 1980s a n d involved a n a b ru p t shift in th e relative a b u n d an c e o f the “p e rm a n e n t” m em b ers o f the co m m u n ity , coinciding w ith observed changes in the p lan k to n o f the n o rth east, a n d was co rrelated w ith the w in ter NAO. A second discrete change, affecting the to ta l species assem blage, h a p ­ p en ed in th e early 1990s. T his was m ark ed b y a su d d e n a lteratio n in th e set o f “occasionally o c cu rrin g ” species. T his change was c o r­ related w ith average seaw ater te m p e ra tu re (and possibly th e A M O ) ra th e r th a n NAO. A p a rt fro m clim ate a n d fishing, o th er factors m ay also have h ad a n im p o rta n t influence o n plaice a n d sole d istrib u tio n s th ro u g h ­ o u t the tw en tieth century. These m ay have in clu d e d h a b ita t m o d i­ fication, changes in prey resources a n d /o r overall system productivity, a n d changes in p re cip ita tio n a n d ru n -o ff patterns. P erhaps fish p o p u latio n s have been affected b y successive loss o f im p o rta n t so u th e rn plaice (a n d sole) n u rse ry g rounds. T he D u tc h a n d Belgian coasts w ere previously very good n u rse ry g ro u n d s for plaice (W im penny, 1953), b u t several an th ro p o g e n ic changes have strongly altered this. In p articular, th e closure o f th e Z uiderzee in th e early 1930s a n d th e closure o f estuaries o f th e R hine, M euse, a n d Scheldt in the 1960s a n d 1970s (H eip, 1989) will have likely reduced the re cru itm e n t fro m these n u rse ry g ro u n d s a n d hence m ay have affected th e a b u n d an c e a n d d istrib u tio n o f o lder age groups in the N o rth Sea (R ijnsdorp et al., 1992; R ochette et al., 2010). In the coastal w aters o f the w estern M e d ite rran e a n a n d the Bay o f Biscay, w hich are affected b y th e ru n -o ff o f the R hone a n d the Loire, respectively, sole re cru itm e n t a n d grow th has b een positively affected b y th e river ru n -o ff subsidising th e b e n th ic foodw eb w ith a d d itio n o f n u trie n ts a n d terrestrial organic m a tte r (Salen-P icard et al., 2002; Le Pape et al., 2003). Plaice exp an sio n n o rth w a rd s in the N o rth Sea d u rin g th e 1 9 5 0 - 1980s m ig h t have hap p en ed because o f n u trie n t e n ric h m e n t a n d e u tro p h ic a tio n o f the coastal zone, because th e plaice grow th rate is k n o w n to correlate strongly w ith various e u tro p h ic a tio n param eters (e.g. p h o sp h o ru s in p u ts), a n d th ere is evidence o f increased b e n th ic p ro d u c tiv ity in large areas o f the N o rth Sea d u rin g this period. Sole a n d plaice in th e G erm an B ight have b e en particu larly heavily influenced b y river ru n -o ff a n d n u trie n t in p u ts th a t increased in the 1960s and 1970s, b u t w hich have subsequently declined (C olijn et al., 2002; P h ilip p a rt et al., 2007). T he decrease in relative a b u n d an c e in th e G erm an B ight in recent decades m ay b e d u e to th e increase in sm all-bodied, sm all­ m o u th e d flatfish, such as solenette Buglossidium luteum a n d scaldfish Arnoglossus laterna, w hich m ay co m p ete w ith plaice a n d sole in this region (Jennings et al., 2008; v a n H al et al., 2010), in c o n ­ ju n c tio n w ith changes in th e p ro d u c tiv ity o f th e b e n th ic ecosys­ tem . A n analysis o f plaice a n d sole diets in 1996 c o m p a red w ith th e begin n in g o f th e tw en tieth c en tu ry has revealed th a t p oly­ chaetes have increased a n d bivalves decreased (R ijnsdorp and V ingerhoed, 2001). These results m ig h t reflect a change in system productivity, b u t equally they are consistent w ith the hypothesis th a t b eam traw ling has im p ro v ed the feeding Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 S cotland co u ld result from a n invasion fro m th e west (this m ight be a n explan atio n for th e su d d e n “ju m p ” in longitude; Figure 6), as well as a g radual shift in th e centre o f gravity fro m the southeast. ICES stock assessm ents a n d fish ery -in d ep en d en t surveys provide little evidence o f a shift in plaice p o p u latio n s fro m the west to th e east side o f Scotland, because plaice p o p u latio n s have only ever been relatively sm all o ff th e Scottish H ighlands a n d islands. F ro m o u r analyses, sole seem to have re sp o n d ed to clim ate change a n d fishing, a n d the relationship w ith fishing seem s p a r­ ticularly strong. T his is in line w ith extensive targ etin g o f this highly priced species b y th e large N o rth Sea b eam traw l fleet, o ften in preference to plaice (Pilling et al., 2008). C o n trary to sim ple expectations regarding th e m an ifestatio n o f clim ate change, sole have exhibited a southw estw ard shift to shallow er w aters w ith w arm ing, in line w ith P erry et al. (2005). In the N o rth Sea, th e n o rth tends to be colder th a n th e so u th in su m m er, b u t th e so u th tends to be colder th a n th e n o rth in w inter. Som e so u th e rn N o rth Sea species, such as sole, have been previously excluded fro m large areas o f shallow inshore ha b ita t along th e c o n tin e n ta l coast in w inter, because these w aters cool d o w n to well below 3°C, a critical tem p e ra tu re below w hich sole suffer im p aired physiological fu n c tio n a n d increased m o rta lity (W oodhead, 1964a, b ). C onsequently, in the past, sole ten d e d to overw inter in d eeper w aters before re tu rn in g to th e shallow s in spring, to avoid th e lethally cold w in ter tem p e ra ­ tures (H e n d erso n a n d Seaby, 2005). H ow ever, th ere is evidence th a t sole have sta rte d spaw ning in coastal w aters earlier, because o f the ra p id ly w arm in g seas in w in ter since the 1980s (Teal et al., 2008). D u rin g severe w in ter cond itio n s, such as those experienced in the 1960s a n d in 1996, sole w ere excluded fro m the shallow s to th e extent th a t m ass m o rta lity events w ere re p o rte d (W oodhead, 1964a; H o rw o o d a n d M illner, 1998). A sim ilar m ass m o rta lity event for sole was re p o rte d in early 1929 (L um by a n d A tkinson, 1929), n o tab ly o n th e T erschelling g ro u n d o ff th e D u tc h coast. T his coincided w ith seaw ater tem p eratu res th a t were 5°C colder th a n w ere considered “n o rm a l” for th a t tim e o f year. B etter acces­ sibility to shallow, inshore w aters in w in ter m ig h t also explain w hy several o th e r sm all, w a rm -to lera n t so u th e rn species have recently displayed southw ard, “shallow ing” d istrib u tio n shifts (D ulvy et al., 2008). T his is equally a m an ifestatio n o f changing clim ate c o n d itio n s in th e N o rth Sea as the m o re w idely k n o w n (and re p o rte d ) “n o rth w a rd ” shift o f m arin e organism s in recent years. T here are m an y exam ples in th e literature w here fish d istri­ b u tio n a n d /o r a b u n d an c e has been related to the NAO (O ttersen et al., 2001; A ttrill a n d Power, 2002; H e nderson, 2007). H ere, a c o rre latio n o f th e NAO w ith sole latitu d in a l (and p erhaps d e p th ) d istrib u tio n was fo u n d , b u t n o significant relatio n ­ ships in a regression m o d el th a t also inclu d ed th e H adley SST. It is p erhaps n o t surp risin g th a t stronger correlations w ere identified w ith sea tem p eratu res w ith in the N o rth Sea, because tem p e ra tu re is the physical variable experienced directly b y the fish a n d kn o w n to affect th e ir g row th a n d behaviour. T he NAO w ould have a m u ch m o re indirect influence (see also E ngelhard a n d H eino, 2006). In ad d itio n , this stu d y p ro v id ed evidence indicatin g the im p o rta n ce o f the A M O , a n in d ex th a t is b e co m in g increasingly p o p u la r in studies o f fish a n d clim ate in th e N o rth A tlantic (Nye et al., 2009). T he A M O was significantly related to plaice (b u t n o t sole) d istrib u tio n , especially if th e effect o f sea te m p e ra tu re was acco u n ted fo r in regression m odels; this suggested th a t a p art fro m im m ed iate tem p e ra tu re effects, lo n g -te rm fluctu atio n s in th erm o h alin e circu latio n a n d hence shifts in tem p e ra tu re regim e 1101 1102 co n d itio n s for the tw o flatfish species, b y enhan cin g th e a b u n ­ dance o f sm all o p p o rtu n istic b e n th ic species, such as polychaetes, in th e heavily traw led areas. H ow ever, th e changes in d iet m ay also be related to e u tro p h ic a tio n a n d p o llu tio n . A recent co m p a riso n o f plaice diet in the central N o rth Sea (JKP, u n p u b lish ed d a ta ) d u rin g the early 1900s (1 9 0 2 -1 9 0 9 ), 1950s (1 9 5 0 -1 9 5 9 ), a n d early tw enty-first c e n tu ry (2 0 0 4 -2 0 0 9 ) has revealed sim ilar changes in diet, w ith a sw itch away fro m bivalves such as Spisida a n d Ensis tow ards polychaetes, crabs, a n d sandeels. This suggests th a t it is n o t e u tro p h ic a tio n o r p o llu tio n th a t has affected diets (a n d th ere ­ fore grow th a n d possibly d istrib u tio n ) in th e central N o rth sea, a n d th a t th e cause is m o re likely to be increased traw ling pressure, allied w ith lo n g -te rm clim ate change. Conclusions Acknowledgements GH E, JKP, a n d A D R w ere su p p o rte d th ro u g h th e EU FP6 project RECLAIM; GH E, JKP, a n d LTK w ere also su p p o rte d th ro u g h the U K ’s D e p artm en t for F ood, E n v iro n m e n t a n d R ural Affairs (c o n ­ tracts M F1201, M A010, a n d M F1108), a n d LTK was also su p ­ p o rte d b y th e EU U N C O V E R p roject (co n tract 022717). JKP thanks th e conveners o f the In tern a tio n a l S ym posium o n C lim ate C hange Effects o n Fish a n d Fisheries for inviting h im to p resen t this m aterial a n d CLISAP (an excellence cluster at the U niversity o f H a m b u rg ) in su p p o rtin g his a tte n d an c e at th e sym ­ p o siu m . W illiam C heung, N ick D ulvy, C h risto p h er Lynam , D avid M axwell, a n d o n e a n o n y m o u s referee gave valuable feedback a n d advice o n th e statistical analysis. References Attrill, M. J., and Power, M. 2002. Climatic influence o n a m arine fish assemblage. N ature, 417: 275-278. Bannister, R. C. A. 2004. The rise and fall o f cod in the N orth Sea. In M anagem ent o f Shared Fish Stocks, pp. 316-338. Ed. by A. I. L. Payne, C. M. O ’Brien, and S. I. Rogers. Blackwell Publishing, Oxford, UK. Beare, D. J., Burns, F., Greig, A., Jones, E. C., Peach, K., Kienzle, M., McKenzie, E., et a l 2004. Long-term increases in prevalence o f N orth Sea fishes having southern biogeographic affinities. M arine Ecology Progress Series, 284: 269-278. Bevington, P. R. 1969. Data Reduction and E rror Analysis for the Physical Sciences, 2nd edn. McGraw-Hill, New York. Colijn, F., Hesse, K. J., Ladwig, N., and Tillm ann, U. 2002. Effects o f the large-scale uncontrolled fertilisation process along the conti­ nental coastal N orth Sea. Hydrobiologia, 484: 133-148. C oûtant, C. C. 1977. C om pilation o f tem perature preference data. Journal o f the Fisheries Research Board o f Canada, 34: 739-745. Dulvy, N. K., Rogers, S. I., Jennings, S., Stelzenmüller, V., Dye, S. R., and Skjoldal, H. R. 2008. Climate change and deepening o f the N orth Sea fish assemblage: a biotic indicator o f warm ing seas. Journal o f Applied Ecology, 45: 1029-1039. Engelhard, G. H. 2005. Catalogue o f Defra historical catch and effort charts: six decades o f detailed spatial statistics for British fisheries. Science Series Technical Report, 128. Cefas, Lowestoft. 42 pp. Engelhard, G. H. 2008. One hundred and twenty years o f change in fishing power o f English N orth Sea trawlers. In Advances in Fisheries Science: 50 Years o n from Beverton and Holt, pp. 1-25. Ed. by A. Payne, J. Cotter, and T. Potter. Blackwell Publishing, Oxford, UK. Engelhard, G. H., and Heino, M. 2006. Climate change and condition o f herring ( Clupea harengus) explain long-term trends in extent o f skipped reproduction. Oecologia, 149: 593-603. Gillis, D. M., and Peterm an, R. M. 1998. Im plications o f interference am ong fishing vessels and the ideal free distribution to the interpretation o f CPUE. Canadian Journal o f Fisheries and Aquatic Sciences, 55: 37-46. Gillis, D. M., Rijnsdorp, A. D., and Poos, J. J. 2008. Behavioral infer­ ences from the statistical distribution o f comm ercial catch: patterns o f targeting in the landings o f the D utch beam trawler fleet. Canadian Journal o f Fisheries and Aquatic Sciences, 65: 27-37. Greenstreet, S. P. R., Spence, F. B., Shanks, A. M., and McMillan, J. A. 1999. Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and com m unity structure. II. Trends in fishing effort in the N orth Sea by UK registered vessels landing in Scotland. Fisheries Research, 40: 107-124. Hedger, R., McKenzie, E., H eath, M., W right, P., Scott, B., Gallego, A., and Bridson, J. 2004. Analysis o f the spatial distributions o f m ature cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) abundance in the N orth Sea ( 1980-1999) using generalised addi­ tive models. Fisheries Research, 70: 17-25. Heino, M., Engelhard, G. H., and Godo, O. R. 2003. Variations in the distribution o f blue whiting in the Barents Sea: climatic influences o r year class effects ? ICES D ocum ent CM 2003/Q : 03. 11 pp. Heip, C. 1989. The ecology o f the estuaries o f Rhine, Meuse and Scheldt in the Netherlands. Scientia M arina, 53: 457-463. H enderson, P. A. 2007. Discrete and continuous change in the fish com m unity o f the Bristol Channel in response to climate change. Journal o f the M arine Biological Association o f the U nited Kingdom , 87: 589-598. H enderson, P. A., and Seaby, R. M. 2005. The role o f climate in deter­ m ining the tem poral variation in abundance and growth o f sole Solea solea in the Bristol Channel. Journal o f the M arine Biological Association o f the U nited Kingdom, 85: 197-204. H iddink, J. C., and ter Hofstede, R. 2008. Climate induced increases in species richness o f m arine fishes. Global Change Biology, 14: 453-460. H orw ood, J. W., and Millner, R. S. 1998. Cold induced abnorm al catches o f sole. Journal o f the M arine Biological Association o f the United Kingdom, 78: 345-347. Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 The analyses carried o u t here su p p o rt hypothesis 1 th a t lo n g -term shifts a t least in plaice d istrib u tio n are b e tte r explained b y clim ate variables th a n b y fishing pressure. H ow ever, th e situ atio n for sole is m o re com plicated; sole have re sp o n d ed to clim ate change a n d fishing, a n d th eir te m p e ra tu re response is o p p o site to th a t o f plaice. T he results also provide su p p o rt for hypothesis 2, a b o u t the usefulness o f long tim e-series for d isentangling clim ate a n d fishing effects. In the sh o rt term , clim ate a n d fishing variables are c o n fo u n d ed , b u t over a longer p e riod, these factors are clearly separable, a n d we have been able to a ttrib u te the shift in plaice d istrib u tio n to clim ate ra th e r th a n to fishing (although o th e r factors m ay have played a n im p o rta n t role). W e have also co n firm ed hypothesis 3, b y d e m o n stra tin g th at d a ta fro m co m m ercial fisheries can yield useful insight in to the lo n g -te rm im plications o f clim ate change, especially w h en the d a ta are “n o rm alize d ” a n d n o t used to provide absolute biom ass o r a b u n d an c e estim ates. C om m ercial catch d a ta have th e benefit o f being collected over far greater tem p o ral a n d spatial scales th a n fish ery -in d ep en d en t survey data. T hey are usually collected system atically a n d consistently over long periods, a n d th e data ten d to be less n o isy th a n survey d a ta because o f th e m u ch greater sam pling effort. H ow ever, p ro p e r acknow ledgem ent o f p o ten tial biases (e.g. m isreporting, in te rac tio n betw een fishing fleets, technological im provem ents, etc.) is a p rerequisite in such analyses. G. H . Engelhard et al. N in e decades o f N orth Sea sole a n d plaice distribution Hurrell, J. W. 1995. Decadal trends in the N orth Atlantic Oscillation: regional tem peratures and precipitation. Science, 269: 676-679. ICES. 2009. Report o f the W orking G roup o n the Assessment o f Demersal Stocks in the N orth Sea and Skagerrak— Com bined Spring and A utum n (WGNSSK). ICES D ocum ent CM 2009/ ACOM: 10. 1028 pp. Jennings, S., Alvsvâg, J., Cotter, A. J. R., Ehrich, S., Greenstreet, S. P. R., Jarre-Teichm ann, A., M ergardt, N., et al. 1999. Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and com m unity structure. III. International trawling effort in the N orth Sea: an analysis o f spatial and tem poral trends. Fisheries Research, 40: 125-134. Jennings, S., van Hal, R., H iddink, J. C., and Maxwell, T. A. D. 2008. Fishing effects on energy use by N orth Sea fishes. Journal o f Sea Research, 60: 74-88. Jones, P. D., Jonsson, T., and Wheeler, D. 1997. Extension to the N orth Atlantic Oscillation using early instrum ental pressure observations from Gibraltar and South-W est Iceland. International Journal o f Climatology, 17: 1433-1450. Lumby, J. R., and Atkinson, G. T. 1929. O n the unusual m ortality am ongst fish during M arch and April 1929, in the N orth Sea. Journal d u Conseil Perm anent International p our l’Exploration de la Mer, 4: 309-332. Mackenzie, B. R., and Schiedek, D. 2007. Long-term sea surface tem p­ erature baselines-tim e series, spatial covariation and implications for biological processes. Journal o f M arine Systems, 68: 405-420. Millner, R. S., and W hiting, C. L. 1996. Long-term changes in growth and population abundance o f sole in the N orth Sea from 1940 to the present. ICES Journal o f M arine Science, 53: 1185-1195. Nye, J. A., Link, J. S., Hare, J. A., and Overholtz, W. J. 2009. Changing spatial distribution o f fish stocks in relation to climate and p o p u ­ lation size o n the N ortheast U nited States continental shelf. M arine Ecology Progress Series, 393: 111-129. Ottersen, G., Planque, B., Belgrano, A., Post, E., Reid, P. C., and Stenseth, N. Ch. 2001. Ecological effects o f the N orth Atlantic Oscillation. Oecologia, 128: 1-14. Pastoors, M. A., Rijnsdorp, A. D., and van Beek, F. A. 2000. Effects o f a partially closed area in the N orth Sea (“plaice box” ) on stock devel­ opm ent o f plaice. ICES Journal o f M arine Science, 57: 1014-1022. Poos, J. J., and Rijnsdorp, A. D. 2007. An “experim ent” o n effort allo­ cation o f fishing vessels: the role o f interference com petition and area specialization. Canadian Journal o f Fisheries and Aquatic Sciences, 64: 304-313. Pyper, B. J., and Peterm an, R. M. 1998. C om parison o f m ethods to account for autocorrelation in correlation analyses o f fish data. Canadian Journal o f Fisheries and Aquatic Sciences, 55: 2127-2140. Quirijns, F. J., Poos, J. J., and Rijnsdorp, A. D. 2008. Standardizing comm ercial CPUE data in m onitoring stock dynamics: accounting for targeting behaviour in mixed fisheries. Fisheries Research, 89: 1- 8. Rayner, N. A., Parker, D. E., H orton, E. B., Folland, C. K., Alexander, L. V., Rowell, D. P., Kent, E. C., et al. 2003. Global analysis o f sea surface tem perature, sea ice, and night m arine air tem perature since the late nineteenth century. Journal o f Geophysical Research, 108: ACL2.1-ACL22.9. Rijnsdorp, A. D. 2010. Sole. In Resolving Climatic Im pacts on Fish Stocks, pp. 193-198. Ed. by A. D. Rijnsdorp, M. A. Peck, G. H. Engelhard, C. M öllm ann, and J. K. Pinnegar. ICES Cooperative Research Report, 301. Rijnsdorp, A. D., and Millner, R. S. 1996. Trends in population dynamics and exploitation o f N orth Sea plaice (Pleuronectes pla­ tessa L.) since the late 1800s. ICES Journal o f M arine Science, 53: 1170-1184. Rijnsdorp, A. D., Peck, M. A., Engelhard, G. H., M öllm ann, C., and Pinnegar, J. K. 2009. Resolving the effect o f climate change on fish populations. ICES Journal o f M arine Science, 66: 1570-1583. Rijnsdorp, A. D., Poos, J. J., Quirijns, F. J., HilleRisLambers, R., de W ilde, J. W., and den Heijer, W. M. 2008. The arm s race between fishers. Journal o f Sea Research, 60: 126-138. Rijnsdorp, A. D., van Beek, F. A., Flatman, S., Miller, J. M., Riley, J. D., Giret, M., and de Clerk, R. 1992. Recruitm ent o f sole, Solea solea (L.), in the N ortheast Atlantic. Netherlands Journal o f Sea Research, 29: 173-192. Rijnsdorp, A. D., and Vingerhoed, B. 2001. Feeding o f plaice Pleuronectes platessa L., and sole Solea solea (L.) in relation to the effects o f bottom trawling. Journal o f Sea Research, 45: 219-229. Rindorf, A., and Lewy, P. 2006. W arm , w indy winters drive cod north and hom ing o f spawners keeps them there. Journal o f Applied Ecology, 43: 445-453. Perry, A. L., Low, P. J., Ellis, J. R., and Reynolds, J. D. 2005. Climate change and distribution shifts in m arine fishes. Science, 308: 1912-1915. Robinson, R. 2000. Steam power and distant-w ater trawling. In England’s Sea Fisheries: the Com mercial Sea Fisheries o f England and Wales since 1300, pp. 206-216. Ed. by D. J. Starkey, C. Reid, and N. Ashcroft. C hatham Publishing, London. Perry, R. I., Curry, P., Brander, K., Jennings, S., M öllm ann, C., and Planque, B. 2010. Sensitivity o f m arine systems to climate and fishing: concepts, issues and m anagem ent responses. Journal o f M arine Systems, 79: 427-435. Rochette, S., Rivot, E., M orin, J., M ackinson, S., Riou, P., and Le Pape, O. 2010. Effect o f nursery habitat degradation on flatfish p o p u ­ lation: application to Solea solea in the eastern Channel (western Europe). Journal o f Sea Research, 64: 34-44. Philippart, C. J. M., Beukema, J. J., Cadee, G. C., Dekker, R., Goedhart, P. W., van Iperen, J. M., Leopold, M. F., et al. 2007. Im pacts o f nu trient reduction o n coastal com m unities. Ecosystems, 10: 9 5-118. Salen-Picard, C., Darnaude, A. M., Arlhac, D., and Harm elin-Vivien, M. L. 2002. Fluctuations o f m acrobenthic populations: a link between clim ate-driven river ru n -o ff and sole fishery yields in the Gulf o f Lions. Oecologia, 133: 380-388. Pilling, G. M., Keli, L. T., H utton, T., Bromley, P. J., Tidd, A. N., and Bolle, L. J. 2008. Can econom ic and biological m anagem ent objectives be achieved by the use o f MSY-based reference points? A N orth Sea plaice (Pleuronectes platessa) and sole (Solea solea) case study. ICES Journal o f M arine Science, 65: 1069-1080. Teal, L. R., de Leeuw, J. J., van der Veer, H. W., and Rijnsdorp, A. D. 2008. Effects o f climate change o n growth o f 0-group sole and plaice. M arine Ecology Progress Series, 358: 219-230. Pinnegar, J. K., Engelhard, G. H., and Daskalov, G. M. 2010. Changes in the distribution o f fish. In Resolving Climatic Im pacts on Fish Stocks, pp. 9 4-110. Ed. by A. D. Rijnsdorp, M. A. Peck, G. H. Engelhard, C. M öllm ann, and J. K. Pinnegar. ICES Cooperative Research Report, 301. van Hal, R., Smits, K., and Rijnsdorp, A. D. 2010. H ow climate warm ing impacts the distribution and abundance o f two small flatfish species in the N orth Sea. Journal o f Sea Research, 64: 76-84. van Keeken, O. A., van H oppe, M., Grift, R. E., and Rijnsdorp, A. D. 2007. Changes in the spatial distribution o f N orth Sea plaice (Pleuronectes platessa) and implications for fisheries m anagem ent. Journal o f Sea Research, 57: 187-197. Downloaded from http://icesjms.oxfordjournals.org/ by guest on June 21, 2013 Le Pape, O., Chauvet, F., Désaunay, Y., and Guérault, D. 2003. Relationship between interannual variations o f the river plum e and the extent o f nursery grounds for the com m on sole (Solea solea, L.) in Vilaine Bay. Effects on recruitm ent variability. Journal o f Sea Research, 50: 177-185. 1103 1104 W iltshire, K. H., and Manly, B. F. J. 2004. The warm ing trend at Helgoland Roads, N orth Sea: phytoplankton response. Helgoland M arine Research, 58: 269-273. W im penny, R. S. 1953. The Plaice (being the Buckland Lectures for 1949). Edward A rnold & Co., London. 145 pp. W oodhead, P. J. M. 1964a. The death o f N orth Sea fish during the w inter o f 1962/63 particularly with reference to the sole, Solea G. H . Engelhard et al. vulgaris. Helgoländer Wissenschaftliche M eeresuntersuchungen, 10: 283-300. W oodhead, P. J. M. 1964b. Changes in the behaviour o f sole, Solea vulgaris, during cold w inters and the relation between the winter catch and sea tem perature. Helgoländer Wissenschaftliche M eeresuntersuchungen, 10: 328-342.