MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 514: 105–118, 2014doi: 10.3354/meps10972

Published November 6

INTRODUCTION

Understanding the mechanisms shaping the fluctu-ations of marine populations has been the core offisheries research since its beginnings in the last cen-tury (Hjort 1914). However, in the last few decades,there has been increasing evidence that changesin marine species abundance over time are often

accompanied by equally large-scale variations inspatial distribution patterns (Ciannelli et al. 2013 andreferences therein). This has significant implicationsfor marine ecology (e.g. trophic interactions) andfisheries management. Additionally, the understand-ing of the complex relationships between abiotic andbiotic processes is essential for a complete descrip-tion and interpretation of the distributional changes

© Inter-Research 2014 · www.int-res.com*Corresponding author: patricia.puerta@ba.ieo.es,patrix.puerta@gmail.com

Role of hydro-climatic and demographic processeson the spatio-temporal distribution of cephalopods

in the western Mediterranean

Patricia Puerta1,*, Manuel Hidalgo1, María González2, Antonio Esteban3, Antoni Quetglas1

1Instituto Español de Oceanografía, Centre Oceanogràfic de les Balears, Moll de Ponent s/n, Apdo. 291, 07015 Palma de Mallorca, Spain

2Instituto Español de Oceanografía, Centro Oceanográfico de Málaga, Puerto Pesquero s/n, 29640 Fuengirola, Málaga, Spain3Instituto Español de Oceanografía, Centro Oceanográfico de Murcia, Varadero 1, 30740 San Pedro del Pinatar, Murcia, Spain

ABSTRACT: Fluctuations in marine populations occur both in terms of abundance and distribu-tional ranges, and this has major implications for marine ecology and fisheries management. How-ever, little is known about the variability in, and factors influencing, spatial dynamics in marinegroups other than fish. Using time series data from trawl surveys conducted in the western Medi-terranean Sea from 1994 to 2012, we analysed the variability in population distribution (latitude,longitude and depth) of 2 cephalopod species with contrasting life histories, viz. the nektobenthicsquid Illex coindetii and the benthic octopus Eledone cirrhosa. We investigated the influence ofdemographic information (population density and mean individual size) together with environ-mental variables (including chlorophyll a concentration, runoff, precipitation, temperature andclimate) to identify the main drivers shaping the cephalopod distributions in 4 shelf regions withcontrasting oceanographic and geographic conditions. Marked inter-annual fluctuations werefound in the distribution of the 2 species. In general, squid and octopus populations were affectedby the same variables, but the effect of each variable depended on the species and study region.Different life strategies resulted in contrasting responses to the environment, with I. coindetiiassociated with variables contemporary to spring surveys. In contrast, E. cirrhosa was primarilyaffected by conditions during the previous winter. Vertical distribution was mainly explainedby individual body length, while chlorophyll, temperature or local climate proved to be moreimportant for geographical displacements.

KEY WORDS: Spatio-temporal distribution · Spatial ecology · Illex coindetii · Eledone cirrhosa ·Demography · Hydro-climatic conditions

Resale or republication not permitted without written consent of the publisher

Mar Ecol Prog Ser 514: 105–118, 2014

in populations (Swain 1999). Thus, spatial distribu-tion analyses (e.g. geostatistics, variograms, dispersalindices, spatially-explicit regression analyses) havebeen demonstrated to be efficient tools to identifypatterns of variation in marine populations at differ-ent spatio-temporal scales, due to their direct link tohabitat characteristics, environmental variables andbiological processes such as recruitment, spawning,trophic migrations and species demography (Spencer2008, Bacheler et al. 2009, Tamdrari et al. 2010).

Although spatial ecology studies have increasedsignificantly for marine fish populations over recentyears (e.g. Bacheler et al. 2009, Tamdrari et al. 2010,Engelhard et al. 2011), there are comparatively fewstudies on other harvested taxonomic groups such ascephalopods (see for instance Waluda & Pierce 1998,Wang et al. 2003, Smith et al. 2013). Cephalo podsplay a pivotal ecological role in marine eco systems,being opportunistic predators feeding on a greatvariety of fishes and crustaceans, and are import -ant prey of fishes, marine mammals and seabirds(Piatkowski et al. 2001 and references therein). Theimportance of cephalopods in fisheries has risen inrecent decades (Hunsicker et al. 2010), with landingsincreasing steadily worldwide from the 1950s (FAO2012). Owing to their rapid growth rates, short lifespan (most cephalopods are single-season breedersthat die soon after reproduction takes place) and fastpopulation turnover, cephalopods show high sen -sitivity to environmental changes, which result inimportant inter-annual fluctuations in populationabundances (Boyle & Rodhouse 2005) that mightaffect entire ecosystem dynamics (André et al. 2010).In addition to this sensitivity to environmental condi-tions, other effects such as fishing impact, trophicinteractions or migratory behaviour have been re -ported to affect the distribution and abundance ofcephalopod populations (Caddy & Rodhouse 1998,Semmens et al. 2007, Hunsicker & Essington 2008).

Changes in spatial distribution are often related tomigratory movements, which are very common incephalopods either by vertical nycthemeral move-ments or spatial migrations related to feeding orspawning over extensive areas (Boyle & Rodhouse2005). Currently, most available studies dealing withspatial movements in cephalopods have analysed theeffects of environmental parameters on populationabundance (Pierce et al. 1998, Balguerías et al. 2002,Denis et al. 2002) and structure (Ceriola et al. 2006,Lefkaditou et al. 2008). However, little attention hasbeen devoted to investigate the effect of environ-mental and demographic processes on horizontal andvertical movements of cephalopod populations. To

shed light on these aspects, we analysed the popula-tions of 2 cephalopod species from the Mediterran-ean Sea, the short-fin squid Illex coindetii (Ommas-trephidae) and the horned octopus Eledone cirrhosa(Octopodidae). In that area, these species are themost frequent and abundant cephalopods (González& Sánchez 2002, Pertierra & Sánchez 2005, Pierce etal. 2010), having an important commercial value asby-catch (Sartor et al. 1998). Both species are mainlyfound on the lower continental shelf and upper slopebetween 50 and 400 m depth (Boyle & Rodhouse2005), although they show contrasting life histories.I. coindetii is a fast-swimming nektobenthic squid,but is usually associated with the bottom (Bakun &Csirke 1998, Sánchez et al. 1998, Quetglas et al.2014). Spawning takes place all year, with seasonalpeaks in spring and summer. Generalist diet and ver-tical nycthemeral movements and spring migrationstowards the continental shelf are characteristic ofpopulations from the Mediterranean Sea (Sánchez etal. 1998). In contrast, E. cirrhosa is a benthic octopusthat crawls along the sea bottom (Boyle & Rodhouse2005), with a diet based on decapod crustaceans, andstrictly seasonal spawning occurs in spring after amigration shoreward. Recruits can also appear dur-ing spring, since E. cirrhosa has a maximum life spanof 2 yr (Sánchez et al. 2004). Due to such contrastinglife histories and habitats, we expect that these spe-cies would respond differently to environmental anddemographic drivers, resulting in different inter-annual changes in their distributions.

The objective of the present study was to determinethe key drivers that shape the spatio-temporal distri-bution of the 2 cephalopod species in the wes ternMediterranean. This is an oligotrohic and oceano-graphically heterogeneous area, highly influenced bythe interchange of Atlantic and Mediterra nean watermasses through the Strait of Gibraltar (Millot 1999).Specifically, we aimed to analyse the effects of de-mography (population density and mean individualsize) together with environmental variables (includ -ing chl a concentration, runoff, precipitation, hydro -graphy and climate) on bathymetric and geographicchanges in population distribution.

MATERIALS AND METHODS

Study area

The study area covers the entire Mediterraneancoast of the Iberian Peninsula, from its southernmost(Strait of Gibraltar) to the northernmost (Cape of

106

Puerta et al.: Drivers shaping cephalopod distributions

Creus) points, including the Balearic Islands (Fig. 1a).For data analysis, however, we considered 4 differentregions characterized by contrasting oceanographicconditions. The first region comprises the easternIberian coast, from the Cape of Palos to the Cape ofCreus, including the Ibiza Channel (Fig. 1b). Thisarea is mainly influenced by the northern current ofMediterranean waters flowing southwards along thecoast on their way back to the Strait of Gibraltar(Pinot et al. 2002, Monserrat et al. 2008). This currentbifurcates in the Ibiza Channel, with a first branchcontinuing southward and a second branch flowingnorth-eastwards along the north-western slope of theBalearic Islands (Monserrat et al. 2008, Fig. 1a).

The second region covers the waters off the Balea -ric Islands, excluding the Ibiza Channel (Fig. 1c). We

additionally subdivided this region into north andsouth subareas, an operational subdivision which isfurther supported by known contrasting oceano-graphic conditions (López-Jurado et al. 2008, Mon-serrat et al. 2008). The northern subarea correspondsto the Balearic sub-basin and is surrounded by theBalearic Current, which is mainly influenced byatmospheric forcing and is characterized by cold andhighly saline Mediterranean waters (Fig. 1a). By con-trast, the southern subarea, included in the Algeriansub-basin, receives warmer and less saline Atlanticwaters (Fig. 1a) and is mainly driven by density gra-dient forcing (Pinot et al. 2002).

The last region comprises the Alboran Sea, fromthe Strait of Gibraltar to the Cape of Gata (Fig. 1d).This region receives the input flow of Atlantic waters,

107

Fig. 1. (a) Western Mediterranean and the main study area, showing themajor marine currents characterizing the regional circulation (modifiedfrom Millot 1999). ‘Alg’ and ‘Bal’ refer to Algerian and Balearic sub-basins, respectively. The 4 study regions and the locations of sampling sta-tions during annual MEDITS surveys are also shown: (b) eastern Iberian

coast; (c) northern and southern Balearic Sea; (d) Alboran Sea

Mar Ecol Prog Ser 514: 105–118, 2014

which are trapped by 2 quasi-permanent anticyclo -nic gyres and a series of anticyclonic eddies that aredeveloped due to the topography of the area (Millot1999, 2005). The presence of fronts, eddies, up -welling systems and other oceanographic processeswould act as a barrier for the dispersal or exchangebetween populations, especially for planktonic earlylife stages (Bouchet & Taviani 1992, Galarza et al.2009).

Sampling

Squid and octopus abundance data were collectedin the framework of the International Bottom TrawlSurveys in the Mediterranean (MEDITS program)following a common protocol (Bertrand et al. 2002),during the Spanish surveys carried out annuallysince 1994. The time series available depends on thestudy region: 1994 to 2012 in the Alboran Sea and theeastern Iberian coast, and 2001 to 2012 in the Balea -ric Sea. Sampling was done during late spring orearly summer at the same predefined stations, fol-lowing a stratified random sampling design based ondifferent bathymetric strata (10−50, 50−100, 100−200,200−500 and 500−800 m), with the number of haulsper stratum proportional to its area. An average of151 hauls was carried out each year, with proportion-ally few differences between regions. An experimen-tal trawl net (GOC 73) with a stretched mesh size inthe cod-end of 20 mm was used for sampling, with ca.16.4 and 2.8 m of horizontal and vertical openings,respectively (estimated using the SCANMAR sys-tem). Trawling duration was 30 and 60 min at depthsshallower and deeper than 200 m, respectively. Foreach haul, sampling information (date, time, initialand final latitude, longitude and depth, duration,trawled distance, vertical and wing opening of thenet) was recorded, together with the number of indi-viduals and size (dorsal mantle length, DML; to thenearest 0.5 cm) of the 2 cephalopod species. Abun-dances were standardized to number of individualsper km2, and mean DML of a population (weightedby number of individuals) was calculated as a proxyof the relative contribution of juvenile or adult indi-viduals in the population.

Environmental data

To investigate putative environmental drivers af -fecting the spatial distribution of the 2 cephalopods,different oceanographic and climatic related param-

eters were tested. Two series of monthly records, i.e.spring data contemporary to the survey and datafrom the previous winter to account for delayed res -ponses, were used as suggested by the results of relevant studies in the NW Mediterranean (e.g.Lloret et al. 2001, Quetglas et al. 2011). Chl a concen-tration (downloaded from http://disc.sci.gsfc.nasa.gov/ giovanni) as a proxy of food availability, togetherwith sea surface temperature (SST) and precipitationdata from the NCEP/ NCAR reanalysis fields pro-vided by the NOAA/ OAR/ ESRL PSD (Kalnay et al.1996) were obtained for each of the 4 regions. Owingto the important role played by the Ebro River in themarine productivity (Estrada 1996, Lloret et al. 2004)and on the abundance of demersal cephalopods(Lloret et al. 2001, Pertierra & Sánchez 2005) in theeastern Iberian coastal region, runoff information(available at www. chebro. es) was also included inthe analyses for this specific area.

The North Atlantic Oscillation (NAO) was in -cluded as a large-scale climatic index (both winterand spring NAO indices based on principal compo-nent analysis, PCA, were available at https://cli-mate data guide.ucar.edu/climate-data/hurrell-north-atlantic-oscillation-nao-index-pc-based). The NAOaverages out meteorological conditions across timeand space and integrates different aspects of climatesuch as air temperature, wind speed and precipita-tion. This has been demonstrated to influence theinter-annual variability of very different ecologicaltraits of marine populations (growth and mortalityrates, recruitment success) together with abundanceand distribution of different species (Drinkwater etal. 2003, Stenseth et al. 2003, Straile & Stenseth2007), even in our study area (Massutí et al. 2008).At a meso-scale level, a local climatic index (LCI)already used in the western Mediterranean (Fer-nández de Puelles & Molinero 2007, 2013, Hidalgoet al. 2011) was also calculated for each studyregion. This index quantifies the meso- scale hydro-climatic variability, synthesizing the following vari-ables by means of the first axis of a PCA (for details,see Molinero et al. 2005): monthly anomaly fields ofsurface air temperature, SST, atmospheric sea levelpressure, 500 hPa geopotential height and precipi-tation records (Kalnay et al. 1996). High LCI valuesare associated with higher temperatures, drier conditions and weaker winds compared to low LCIvalues. This regional climate variability has beendemonstrated to have a direct effect on zooplanktonbiomass in the Balearic Sea, with critical influenceon its variability during winter conditions (Fernán-dez de Puelles & Molinero 2007, 2008).

108

Puerta et al.: Drivers shaping cephalopod distributions

Spatial distribution descriptors and data analysis

Spatial indices were calculated for each species,region and year, to characterize the location (centreof gravity, CG) and occupation of space (inertia, I )in longitude, latitude and depth, as described byWoillez et al. (2007). The average location of the pop-ulation was estimated using the CG, the mean loca-tion of an individual taken at random in the field. I, ormean square distance between such an individualand the CG, describes the spatial dispersion of thepopulation around its CG. Where z(x) is the densityof a population at location x, the CG is estimatedfrom the data through discrete summations over sam-ple locations as

(1)

and I is calculated as its variance:

(2)

These statistics only depend on positive densityvalues but not on 0 values; that is, in the case of 0, itsnumerical contribution to the CG is also 0 (Woillez etal. 2007). When sampling is geographically heteroge-neous, as in our case, it is recommended to use areasof influence around each sample (sum of points of thespace that are closer to this sample than the others)as weighting factors of CG and I to ensure that dis-placements of CG are not due to unbalanced sam-pling areas or changes in sampling design over time(Woillez et al. 2007). However, we did not use thembecause no remarkable inter-annual variations wereobserved in our results. Similarly, it is worth notingthat isotropy and anisotropy indices, which are oftenused to describe whether the spatial direction of pop-ulation dispersion around its CG is identical or not(Woillez et al. 2007), were not calculated in our study.They proved uninformative in all our study regionsbecause the coastal morphology and orientationrestricted the spatial distribution of the species.

Spatial distribution studies of marine populationsaround islands imply having average estimated posi-tions on land in some cases, as well as buffering ofpotential changes in distribution on opposed sides ofthe islands. This geographical limitation was circum-vented with the additional split of the Balearic Searegion into northern and southern sub-areas, whichis also consistent with the oceanographic differencesmentioned above.

For each cephalopod species and region, general-ized additive models (GAMs) were fitted on the timeseries of longitude, latitude and depth CG to describe

the putative temporal linear or non-linear trends overtime. Secondly, we applied generalized linear mod-els (GLMs) to analyse the potential relationships ofenvironmental (described above) and demographicpredictors with the time series of CG. As demo-graphic information, we used the mean populationdensity (ind. km−2) and the mean DML (weighted bynumber of individuals) to account for density-depen-dent and size-dependent variations, respectively, inspatial distribution. GLMs were selected in this case,as non-linear relationships were not expected be -tween location descriptors and the covariates, butalso because they would be difficult to interpret.Prior to the GLM analysis, we applied variance in -flation factor (VIF) analysis to detect collinear vari-ables. A cut-off VIF value of 5 was applied to vari-ables, in which case they were removed for the nextVIF analyses until getting the final subset of covari-ates to include in the GLMs (Zuur et al. 2009). Modelselection was based on a stepwise approach obtainedusing Akai ke’s information criterion (AIC) as a meas-ure of goodness of fit, with the best model havingthe smallest AIC value. Finally, model residuals werechecked to fulfil the normality assumption and ab -sence of temporal autocorrelation in both GAMs andGLMs. All analyses were developed using R statis -tical software (R Development Core Team 2013).

RESULTS

Time series of distribution descriptors

The CG of both cephalopod populations showedmarked inter-annual fluctuations in longitude, lati-tude and depth (Fig. 2). Variation in latitudinal andlongitudinal changes was proportional to the areacovered in each region, being smoother in theBalearic Sea compared to the Alboran Sea and east-ern Iberian coast. The most important movements,which were obviously restricted by the shape of thecontinental shelf, were those parallel to the coast,with perpendicular displacements being compara-tively reduced (Fig. S1 in the Supplement at www.int-res. com/articles/suppl/m514p105_supp. pdf). In thecase of Eledone cirrhosa from the Alboran Sea, therewas a significant positive linear trend in both latitudeand longitude (p < 0.05 in both cases) indicating a netdisplacement northeastwards.

Depth changes between both cephalopod specieswere correlated in the Alboran Sea (correlation co -efficient, r = 0.39) and the eastern Iberian coast (r =0.43). In these 2 regions, depth was also correlated

xz x xz x x

= ∫∫

CG( )d

( )d

I xx z x x

z x x= = ∫

∫Var( )

( – CG) ( )d( )d

2

109

Mar Ecol Prog Ser 514: 105–118, 2014

with latitudinal and/or longitudinal movements (p <0.05), suggesting that changes in depth were partial -ly determined by the continental shelf topo graphy.This synchrony between species was absent in theBalearic Sea, but there was instead a synchronybetween the northern and southern areas in bothcephalopods (Illex coindetii: r = 0.64; E. cirrhosa: r =0.48). With the only exception of E. cirrhosa from theAlboran Sea, mean depth in the 2 species was alwayshigher in the Balearic Sea than in the 2 mainlandregions. There was only a significant non-lineartrend in inter-annual variation of depth distributionin the case of I. co in detii from the Alboran Sea (p <

0.05), with depth increasing from 75 m in 1994 to210 m in 2004 and decreasing subsequently down to100 m in 2012 (Fig. 2).

I-values of the 2 cephalopod populations also dis-played marked inter-annual fluctuations in latitude,longitude and depth (Fig. S2 in the Supplement).Latitu dinal and longitudinal I-values were also pro-portional to the area covered in each region, beinghigher in the Alboran Sea and the eastern Iberiancoast compared to the Balearic Sea. By contrast, I-values of depth were comparatively higher in theBale aric Sea, owing to the wider depth range of bothspecies in the archipelago.

110

Fig. 2. Time series of changes in centres of gravity of Illex coindetii (solid black line) and Eledone cirrhosa (solid grey line) inlongitude, latitude and depth for each study region. Significant linear and non-linear trends (p < 0.05) are described by dashed

lines. Scales of latitude and longitude are proportional to the area covered by each region

Puerta et al.: Drivers shaping cephalopod distributions

Time series of demographic and environmentalcovariates

Temporal variation of the demographic covariatesdepended on the species and the region considered(Fig. 3). In the Balearic Sea, for instance, the 2 speciesshowed lower fluctuations in both population densityand individual size, compared to the other 2 regions.In all cases, inter-annual changes in population den-sity were higher in squids than in octopuses. Therewere no clear patterns for density and size, with theonly exception being the mean size of E. cirrhosa in

the Alboran Sea, which increased progressively from70 to 119 mm between 2004 and 2011, reaching thehighest values of the whole series. Other generalitiesincluded the higher squid sizes in the Balearic Seathroughout the time series and the increase in squiddensity in all areas since 2009. Finally, it was ob -served that high densities of squid on the easternIberian coast were negatively correlated with size(r = −0.47), which might be explained by the inter-annual variation in the strength of recruitment.

The time series of environmental covariates (chl a,precipitation and LCI) during spring and the pre -

111

Fig. 3. Changes in annual population mean density and dorsal mantle length (DML) of Illex coindetii (black line) and Eledone cirrhosa (grey line) for each study region

Mar Ecol Prog Ser 514: 105–118, 2014

vious winter displayed saw-toothed inter-annualfluctuations without any clear general trends (Fig. 4).In agreement with northwestern Mediterranean phe-nology, winter chl a values were higher than springvalues. The opposite pattern occurred for precipita-tion, corresponding with the rain and drought peri-ods of the Mediterranean climate. As expected, theLCI also showed important dissimilarities betweenthe spring and previous winter time series.

Demographic and environmental effects

For each species and region, Table 1 displays thebest GLM for each response variable (latitude, longi-

tude, depth), with the statistically significant co -variates included in the model. In general, the 2cephalopod species responded to the same set ofvariables, indicating sensitivity to the same demo-graphic (DML and density) and environmental (pre-cipitation, LCI, chl a) drivers. However, the responseswere not homogeneous, evidencing clear contrastingeffects depending on the species and region. Themain contrast between the 2 species was the generalassociation of I. coindetii distribution with contempo-rary spring environmental variables, while E. cirrho -sa was more influenced by the previous winter pro-cesses. Different explanatory variables influenceddistributional changes in each region. However, wedid not find significant influences of covariates for

112

Fig. 4. Interannual variability in chl a concentration, precipitation and local climatic index during winter (black line) and spring (grey line) for each study region

Puerta et al.: Drivers shaping cephalopod distributions

some combinations (Table 1, no reported models),such as changes in latitude and longitude for bothspecies in the northern Balearic Sea or for I. coindetiion the eastern Iberian coast.

The influence of individual size was the most fre-quent explanatory variable related to distributionalchanges. Although it showed no influence in theAlbo ran Sea, it gave rise to contrasting responsesbetween species in the other regions. Size effects onI. coindetii, for instance, were always positive in alldescriptors (latitude, longitude and depth) from the

Balearic Sea, indicating a northeastdisplacement of the population anda movement of large individuals todeeper waters. By contrast, the effectson E. cirrhosa depended on the dis -tributional parameter analysed, beingnegative for latitude and longitude butpositive for depth both on the easternIberian Coast and in the southernBalearic Sea. Contrary to size, popula-tion density only affected the squiddistribution from the eastern Iberiancoast and that of octopus from theAlboran Sea. In the first case, negativeeffects of depth indicated higher densities of squid at shallower waters,while negative effects of longitudefor octopus indicated increased popu-lation densities related to westwardmovements.

Independently of the distributionalparameter considered, the effect offreshwater inputs (precipitation andEbro run-off) was always positive,whereas the effect of chl a was alwaysnegative. Precipitation in duced dis-placements of I. coindetii to deeperwaters in the Alboran Sea, while a similar response was observed on theeastern Iberian coast with the riverrun-off impact. In the southern BalearicSea, precipitation conditioned latitudi-nal and longitudinal movements, caus-ing a slightly net movement of thesquid population to the northeast. Inthe case of E. cirrhosa, precipitation in-duced longitudinal and depth chan gesin the Alboran Sea and southernBalearic Sea, respectively. Concerningchl a, I. coindetii populations from thenorthern Balearic Sea showed a highlynegative effect with depth. The same

negative res ponse was observed in latitudinal changesin the southern Balearic Sea for the 2 species.

The SST and LCI showed clear contrasting influ-ences depending on the region and species consid-ered. The LCI affected exclusively the Balearic Searegions, which in turn were not affected by the SST.In contrast, SST did affect both the Alboran Sea andthe eastern Iberian coast. In the Alboran Sea, SSThad negative effects on latitude and longitude inI. coindetii, but positive longitudinal effects in E. cirrhosa. Except longitudinal effects on the octopus

113

Descrip- Illex coindetii Eledone cirrhosator Variable Estimate SE R2 Variable Estimate SE R2

Alboran SeaLatitude (Int) 37.38 0.34 0.34

NAOspr* −0.03 0.01SSTspr* −0.06 0.02

Longitude (Int) 1.87 2.36 0.23 (Int) −12.55 4.33 0.40CHLwin· −2.09 1.12 PPspr* 1.20 0.37SSTwin· −0.39 0.18 SSTspr* 0.64 0.30

DEN* −0.01 0.00Depth (Int) 100.18 20.90 0.03

PPspr* 54.59 24.84

Eastern Iberian coastLatitude (Int) 41.30 0.60 0.24

DML* −0.02 0.01Longitude (Int) 1.76 0.35 0.28

DML* −0.01 0.00Depth (Int) −615.15 248.75 0.55 (Int) 65.36 37.82 0.12

EBRspr* 0.14 0.04 DML· 0.83 0.45DEN* −0.03 0.01SSTspr* 47.62 16.57

Northern Balearic SeaDepth (Int) −12.85 161.31 0.61 (Int) 189.23 7.36 0.32

CHLspr** −670.60 180.41 LCIwin* 16.48 6.63LCIwin** −41.02 11.82DML* 2.72 1.02

Southern Balearic SeaLatitude (Int) 37.41 0.38 0.77 (Int) 39.58 0.05 0.28

CHLspr* −1.62 0.52 CHLwin* −0.34 0.15PPspr*** 0.29 0.05DML*** 0.01 0.00

Longitude (Int) −1.54 1.46 0.44 (Int) 3.84 0.23 0.40PPspr* 0.42 0.16 LCIwin* 0.12 0.04DML* 0.03 0.01 DML* −0.01 0.00

Depth (Int) −572.07 260.89 0.42 (Int) 44.52 29.58 0.55LCIspr· 46.02 22.26 LCIwin* −13.84 5.66DML* 4.75 1.54 PPwin* 26.56 9.97

DML· 0.72 0.38

Table 1. Summary of variability in centres of gravity of Illex coindetii and Eledonecirrhosa obtained with the best generalized linear model (GLM) including signifi-cant explanatory variables. Note that model results are not shown for those withno significant variables. Explanatory variables are CHL: chlorophyll; DEN: den-sity; DML: dorsal mantle length; EBR: Ebro run-off; LCI: local climate index; NAO:North Atlantic Oscillation Index; PP: precipitation; SST: sea surface temperature;spr: spring; win: winter; Int: GLM-intercept. Level of significance: ·p < 0.1, *p <

0.05, **p < 0.01, ***p < 0.001

Mar Ecol Prog Ser 514: 105–118, 2014

populations from the southern Balearic Sea, the LCIin duced depth-related changes in all other cases. Inthe northern Balearic Sea, however, such effectswere negative in the squid but positive in the octo-pus. In the southern Balearic Sea, by contrast, theeffects on depth were positive in the squid but nega-tive in the octopus.

DISCUSSION

Our study reports clear inter-annual changes in thespatial distribution of 2 of the main cephalopod spe-cies harvested in the western Mediterranean Sea,viz. the squid Illex coindetii and the octopus Eledonecirrhosa, and provides a comprehensive overview oftheir space occupation in the study area. We de -monstrate that these distributional active movementswere associated with both the population demo -graphy (density- and size-dependent variations) andenvironmental processes (e.g. productivity, hydro -graphy and climate). In general, squid and octopuspopulations were affected by the same variables, buttheir responses differed depending on the speciesand regions. Besides the inherent complexity ofinteractions and synergies present in most ecologicalprocesses, some of the environmental variablesappeared to be principally associated with specificstudy regions (e.g. LCI to the Balearic Sea). This rein-forces our initial reasoning of splitting those regionsaccording to their different oceanographic condi-tions. Also, the contrasting responses in distribu-tional changes exhibited by the squid and octopusare a consequence of their contrasting life historystrategies. As a demersal, fast-swimming squid, I.coindetii presents a strong sensitivity to oceano-graphic features (Boyle & Rodhouse 2005, Pierce etal. 2008), which was reflected in our results by a clearassociation with the contemporary (spring) environ-mental variables. By contrast, the benthic octopus E.cirrhosa was preferentially affected by conditionsduring the previous winter, which is in accordancewith previous work reporting lag-time responses toenvironmental drivers on this species (Lloret et al.2001, Quetglas et al. 2011) and other octopuses (So -brino et al. 2002, Vargas-Yáñez et al. 2009).

Demographic factors, such as population density orindividual size, have been widely studied in ce pha lo -pods, but are usually considered as res pon ses insteadof drivers (e.g. Waluda & Pierce 1998, Chen et al.2006, Perdichizzi et al. 2011). Despite the importanceof population density in spatial distribution analysis,this was only associated with longitude in E. cirrhosa

and depth in I. coindetii from the Alboran Sea andeastern Iberian coast, respectively. The mean indi-vidual size, which was the main factor driving popu-lation displacements, indu ced homogeneous positivebathymetric responses in the 2 species. This suggeststhat populations with predominantly small-sizedindividuals are often distributed in shallow waters,whereas populations dominated by adults are distrib-uted through deeper habitats. This is in accordancewith the life history of the 2 cephalopods in the west-ern Mediterranean, with small individuals of E.cirrho sa mainly located between 100 and 200 mdepth and larger individuals at greater depths(Sánchez et al. 2004). Juveniles of I. coindetii are dis-persed in the water column during spring, displayingdiel vertical migrations, whereas adults are foundclose to the bottom (Sán chez et al. 1998, Quetglas etal. in press). Furthermore, the population density ofI. coindetii was negatively correlated with depth onthe eastern Iberian coast and also with the meanDML of the pop ula tion, suggesting the influence ofstrong re cruitment events at shallow depths. Sizeeffects on latitude− longitude were negative for theoctopus pop ulation from the eastern Iberian coast.This indicates a displacement of populations domi-nated by small individuals of E. cirrhosa towards themouth of the Ebro River, likely related to years ofstrong recruitment. We did not find significant effectsof Ebro runoff on these movements, although resultsfrom previous authors (Lloret et al. 2001, Quetglas etal. 2011) support a strong relationship between foodavailability around river mouths and recruitmentsuccess of this octopus in the western Mediterranean.

SST is the most studied environmental driver andhas been shown to have considerable effects oncephalopod abundance and distribution (Pierce et al.2008 and references therein). In our case, SST onlyaffected populations from the mainland regions butnot those from the Balearic Sea, where the LCImay account for thermal variability (see below). SSTmainly influences the Alboran Sea region, where con-trasting responses were observed in the 2 species.While I. coindetii tended to displace to cooler wa ters,E. cirrhosa displacements seemed to be more asso -ciated with warmer temperatures. De pen ding on thespecies considered, cephalopod res ponses to tem -perature can be highly variable (Vargas-Yáñez et al.2009 and references therein). How ever, temperaturevariations might also act as a proxy for other relatedoceanographic processes at different spatial andtemporal scales other than those investigated here.Particularly, the Alboran Sea is characterized by anintense mixing of Atlantic and Mediterranean waters

114

Puerta et al.: Drivers shaping cephalopod distributions

(Millot 1999), the presence of several up welling pro-cesses and fronts that favour nutrient enrichment andprimary production (Estrada 1996), and advection ofwater masses from distant places that trigger reten-tion/dispersion of early life stages (Bouchet & Taviani1992, Galarza et al. 2009). These processes havebeen found to highly influence the distribution andabundance of cephalopods worldwide (Semmens etal. 2007 and references therein). Thus, geographicpopulation displacements in the Alboran Sea mayalso respond to this intense oceanographic variabil-ity, also explaining the influence of spring variablesin the benthic species E. cirrhosa (an opposite patternto that observed in the other regions).

Food availability has also traditionally been ana-lysed in many cephalopod studies, with chl a as themost used proxy for marine productivity (Pierce et al.2008 and references therein). Chl a was associatedwith several population displacements in all regions,except the eastern Iberian coast, despite being themost productive region (Estrada 1996). However, ourresults suggest that local-scale processes on the east-ern Iberian coast, such as river input pulses, can bemore relevant for local primary production than theregional enrichment processes associated with theinfluence of water masses coming from the Gulf ofLion (Estrada 1996). Several studies suggest thatrainfall also indirectly affects food availability forcephalo pods due to changes in water quality andsalinity (Lefkaditou et al. 1998, Sobrino et al. 2002).In the Alboran Sea, rainfall can affect the availabilityof cephalopod prey as a consequence of changes inwater quality (e.g. turbidity, pollutants; Sobrino et al.2002), which remains consistent with our re sults. Inthe case of the Balearic Sea, our results agree withprevious studies reporting the strong ef fect of precip-itation on marine food webs through important varia-tions in nutrient concentrations and phytoplanktonand zooplankton biomass (Fernández de Puelles &Molinero 2007, 2008). Thus, in highly oligotrophicregions such as the southern Balearic Sea, the dis -tribution of cephalo pods seems to be especially sen-sitive to trophic-related drivers, such as chl a andprecipitation.

Large-scale climatic variability, including NAO,has been reported to influence population dynamicsof many cephalopods, especially ommastrephids(Zuur & Pierce 2004, Waluda et al. 2006). However, inagreement with more recent studies (Keller et al.2012, Quetglas et al. 2013), we found negligibleeffects of NAO on the distribution of cephalopodsfrom the western Mediterranean. By contrast, theLCI was a recurrent driver in the Balearic Sea. The

relevance of LCI in the Balearic Sea has already beenreported for different taxonomic groups (Fernándezde Puelles & Molinero 2007, Hidalgo et al. 2011),including cephalopods (Keller et al. 2012). Winterlocal climate determines the regional hydrographicpattern around the Balearic Islands, affecting pro-ductivity and food availability in the upper trophiclevels (Hidalgo et al. 2011). During low LCI winters,cold winds reinforce the mixing of the water column,strengthen trophic food webs and increase spatialand temporal coverage of favourable trophic condi-tions under a general oligotrophic system (Fernándezde Puelles & Molinero 2007, 2008). In our results, theLCI strongly influenced the bathymetric distributionof both cephalopods and the geographic variability ofE. cirrhosa in the Balearic Sea. However, the sign ofthis effect in the archipelago varies between speciesas well as between subareas, a pattern already ob -served for other species (Guijarro et al. 2008, 2009,Hidalgo et al. 2008, 2009) that might be related to dif-ferences in local hydrography between the north andsouth of the archipelago.

Finally, the lack of environmental influences in theI. coindetii populations from the northern BalearicSea and the eastern Iberian coast suggests 2 non-exclusive causes. Firstly, connectivity between popu-lations from these 2 regions may occur for a highlymobile species such as I. coindetii, as has been docu-mented for some fishes (Galarza et al. 2009, Hidalgoet al. 2009). Therefore, squid responses to environ-mental drivers might be masked by occasional arrivalof individuals from neighbouring subpopulations.Secondly, we could not identify the precise spatialscale at which a key environmental driver is affectingthe population distribution (Ciannelli et al. 2013),particularly in the case of the eastern Iberian coast.Besides the importance of the spatial scale, we alsorecognize that population displacements for bothspecies may be associated with drivers other than theones analysed here. For instance, we did not considerseabed information (e.g. substrata, shelf shape) thatcould be pivotal for understanding the mechanismsaffecting the distribution of benthic species such asE. cirrhosa (Boyle & Rodhouse 2005). Factors such aspredator−prey relationships or fishing effort may alsoexert a relevant influence on distributional patternsof marine species, as has already been observed infish populations (Bartolino et al. 2012, Hunsicker etal. 2013). Additionally, available data based only onspring trawl surveys presents some limitations. Thespawning peak of both species occurs in spring witha migration to shallower areas (Sánchez et al. 1998,2004). Inter-annual variability in the spawning time

115

Mar Ecol Prog Ser 514: 105–118, 2014

may thus bias the estimations of mean DML andinfluence the displacement of populations.

In summary, our study shed new light on thescarcely investigated issue of determining the driversshaping the spatio-temporal displacement of cepha -lopod populations. Despite the common sensitivity tothe same drivers in the 2 species studied, effects ofdemographic and environmental variables varied be -tween species and regions. The mean DML mainlyexplained bathymetric changes, while SST, local cli-mate or chl a was more relevant for geographical dis-placements. Our study improves our knowledge onthe population dynamics of cephalopods and enhan -ces the need for a deeper understanding of the com-plex interactions between physical and biologicalprocesses. This constitutes an important step to im -prove cephalopod assessment and management inthe framework of current climate change, which isexpected to disrupt the distribution patterns of marine populations in areas considered especiallysensitive, such as the Mediterranean Sea (Coma et al.2009). Displacements in spatial distribution andtheir association with environmental variables canhelp to forecast population dynamics to implementsustainable fishery management programmes andidentify essential habitats or potential marine re -serves. These applications can have special impor-tance for cephalopod stocks, which in most cases stillremain unmanaged.

Acknowledgements. We are grateful to all participants inthe MEDITS scientific surveys and funding support for thisresearch through the ECLIPSAME Project (CTM2012-37701) and FPI grant BES-2010-030315 from the SpanishMinistry of Economy and Competitiveness. M.H. is fundedby the European Community’s Seventh Framework Pro-gramme (FP7/2007-2013) under grant agreement No.289257 (MYFISH).

LITERATURE CITED

André J, Haddon M, Pecl GT (2010) Modelling climate-change-induced nonlinear thresholds in cephalopodpopulation dynamics. Glob Change Biol 16: 2866−2875

Bacheler NM, Bailey KM, Ciannelli L, Bartolino V, Chan KS(2009) Density-dependent, landscape, and climate ef -fects on spawning distribution of walleye pollock Thera-gra chalcogramma. Mar Ecol Prog Ser 391: 1−12

Bakun A, Csirke J (1998) Environmental processes and re -cruitment variability. In: Rodhouse PG, Dawe EG, O’DorRK (eds) Squid recruitment dynamics. The genus Illex asa model, the commercial Illex species and influences onvariability. FAO Fish Tech Pap No. 376. FAO, Rome,p 105−120

Balguerías E, Hernández-González C, Perales-Raya C(2002) On the identity of Octopus vulgaris Cuvier, 1797stocks in the Saharan bank (northwest Africa) and their

spatio-temporal variations in abundance in relation tosome environmental factors. Bull Mar Sci 71: 147−163

Bartolino V, Ciannelli L, Spencer P, Wilderbuer TK, Chan KS(2012) Scale-dependent detection of the effects of har-vesting a marine fish population. Mar Ecol Prog Ser 444: 251−261

Bertrand JA, Gil de Sola L, Papaconstantinou C, Relini G,Souplet A (2002) The general specifications of the MED-ITS surveys. Sci Mar 66: 9−17

Bouchet P, Taviani M (1992) The Mediterranean deep-seafauna: pseudopopulations of Atlantic species? Deep-SeaRes A 39: 169−184

Boyle P, Rodhouse P (2005) Cephalopods: ecology and fish-eries. Blackwell Science, London

Caddy JF, Rodhouse P (1998) Cephalopod and groundfishlandings: evidence for ecological change in global fish-eries? Rev Fish Biol Fish 8: 431−444

Ceriola L, Ungaro N, Toteda F (2006) Some information onthe biology of Illex coindetii Verany, 1839 (Cephalopoda,Ommastrephidae) in the south-western Adriatic Sea(Central Mediterranean). Fish Res 82: 41−49

Chen CS, Pierce GJ, Wang J, Robin JP and others (2006) Theapparent disappearance of Loligo forbesi from the southof its range in the 1990s: trends in Loligo spp. abundancein the northeast Atlantic and possible environmentalinfluences. Fish Res 78: 44−54

Ciannelli L, Fisher JAD, Skern-Mauritzen M, HunsickerME, Hidalgo M, Frank KT, Bailey KM (2013) Theory,consequences and evidence of eroding population spa-tial structure in harvested marine fishes: a review. MarEcol Prog Ser 480: 227−243

Coma R, Ribes M, Serrano E, Jiménez E, Salat J, Pascual J(2009) Global warming-enhanced stratification and massmortality events in the Mediterranean. Proc Natl AcadSci USA 106: 6176−6181

Denis V, Lejeune J, Robin JP (2002) Spatio-temporal analy-sis of commercial trawler data using general additivemodels: patterns of loliginid squid abundance in thenorth-east Atlantic. ICES J Mar Sci 59: 633−648

Drinkwater KF, Belgrano A, Borja A, Conversi A and others(2003) The response of marine ecosystems to climatevariability associated with the North Atlantic Oscillation.In: Hurrell JW, Kushnir Y, Ottersen G, Visbeck M (eds)The North Atlantic Oscillation: climatic significance andenvironmental impact. American Geophysical Union,Washington, DC, p 211−234

Engelhard GH, Pinnegar JK, Kell LT, Rijnsdorp AD (2011)Nine decades of North Sea sole and plaice distribution.ICES J Mar Sci 68: 1090−1104

Estrada M (1996) Primary production in the northwesternMediterranean. Sci Mar 60: 55−64

FAO (Food and Agriculture Organization of the UnitedNations) (2012) The state of world fisheries and aqua -culture. FAO Fisheries and Aquaculture Department,FAO, Rome

Fernández de Puelles ML, Molinero JC (2007) North Atlan -tic climate control on plankton variability in the BalearicSea, western Mediterranean. Geophys Res Lett 34: 1−5,doi: 10.1029/2006GL028354

Fernández de Puelles ML, Molinero JC (2008) Decadalchanges in hydrographic and ecological time-series inthe Balearic Sea (western Mediterranean), identifyinglinks between climate and zooplankton. ICES J Mar Sci65: 311−317

Fernández de Puelles ML, Molinero JC (2013) Increasing

116

Puerta et al.: Drivers shaping cephalopod distributions

zooplankton variance in the late 1990s unveils hydro -climate modifications in the Balearic Sea, Western Medi-terranean. Mar Environ Res 86: 76−80

Galarza JA, Carreras-Carbonell J, Macpherson E, PascualM, Roques S, Turner GF, Rico C (2009) The influence ofoceanographic fronts and early-life-history traits on con-nectivity among littoral fish species. Proc Natl Acad SciUSA 106: 1473−1478

González M, Sánchez P (2002) Cephalopod assemblagescaught by trawling along the Iberian Peninsula Mediter-ranean coast. Sci Mar 66: 199−208

Guijarro B, Massutí E, Moranta J, Díaz P (2008) Populationdynamics of the red shrimp Aristeus antennatus in theBalearic Islands (western Mediterranean): short spatio-temporal differences and influence of environmental fac-tors. J Mar Syst 71: 385−402

Guijarro B, Massutí E, Moranta J, Cartes JE (2009) Shortspatio-temporal variations in the population dynamicsand biology of the deep-water rose shrimp Parapenaeuslongirostris (Decapoda: Crustacea) in the western Medi-terranean. Sci Mar 73: 183−197

Hidalgo M, Massutí E, Moranta J, Cartes J, Lloret J, OliverP, Morales-Nin B (2008) Seasonal and short spatial pat-terns in European hake (Merluccius merluccius L.) re -cruitment process at the Balearic Islands (western Medi-terranean): the role of environment on distribution andcondition. J Mar Syst 71: 367−384

Hidalgo M, Tomás J, Moranta J, Morales-Nin B (2009) Intra-annual recruitment events of a shelf species around anisland system in the NW Mediterranean. Estuar CoastShelf Sci 83: 227−238

Hidalgo M, Rouyer T, Molinero JC, Massutí E, Moranta J,Guijarro B, Stenseth NC (2011) Synergistic effects offishing-induced demographic changes and climate vari-ation on fish population dynamics. Mar Ecol Prog Ser426: 1−12

Hjort J (1914) Fluctuations in the great fisheries of northernEurope viewed in the light of biological research. RappP-V Reùn Cons Int Explor Mer 20: 1−228

Hunsicker ME, Essington TE (2008) Evaluating the potentialfor trophodynamic control of fish by the longfin inshoresquid (Loligo pealeii) in the Northwest Atlantic Ocean.Can J Fish Aquat Sci 65: 2524−2535

Hunsicker ME, Essington TE, Watson R, Sumaila UR (2010)The contribution of cephalopods to global marine fish-eries: Can we have our squid and eat them too? Fish Fish11: 421−438

Hunsicker ME, Ciannelli L, Bailey KM, Zador S, Stige LC(2013) Climate and demography dictate the strength ofpredator-prey overlap in a subarctic marine ecosystem.PLoS ONE 8: e66025

Kalnay E, Kanamitsu M, Kistler R, Collins W and others(1996) The NCEP/NCAR 40-Year Reanalysis Project. BullAm Meteorol Soc 77: 437−471

Keller S, Valls M, Quetglas A (2012) Life-history, ecologyand fishery of Sepia officinalis in the western Medi -terranean. P69-Cephalopod International AdvisoryCouncil Symposium, 27 Oct–02 Nov 2012, Florianópolis,Santa Catarina, Brazil

Lefkaditou E, Sánchez P, Tsangridis A, Adamidou A (1998)A preliminary investigation on how meteorologicalchanges may affect beach-seine catches of Loligo vul-garis in the Thracian Sea (Eastern Mediterranean). Afr JMar Sci 20: 453−461

Lefkaditou E, Politou C, Palialexis A, Dokos J, Cosmopoulos

P, Valavanis VD (2008) Influences of environmental vari-ability on the population structure and distribution pat-terns of the short-fin squid Illex coindetii (Cephalopoda: Ommastrephidae) in the Eastern Ionian Sea. Hydrobio -logia 612: 71−90

Lloret J, Lleonart J, Sole I, Fromentin JM (2001) Fluctuationsof landings and environmental conditions in the north-western Mediterranean Sea. Fish Oceanogr 10: 33−50

Lloret J, Palomera I, Salat J, Sole I (2004) Impact of fresh-water input and wind on landings of anchovy (Engraulisencrasicolus) and sardine (Sardina pilchardus) in shelfwaters surrounding the Ebre (Ebro) River delta (north-western Mediterranean). Fish Oceanogr 13: 102−110

López-Jurado JL, Marcos M, Monserrat S (2008) Hydro-graphic conditions affecting two fishing grounds of Mal-lorca island (Western Mediterranean) during the IDEAProject (2003−2004). J Mar Syst 71: 303−315

Massutí E, Monserrat S, Oliver P, Moranta J and others(2008) The influence of oceanographic scenarios on thepopulation dynamics of demersal resources in the west-ern Mediterranean: hypotheses for hake and red shrimpoff the Balearic Islands. J Mar Syst 71: 421−438

Millot C (1999) Circulation in the Western MediterraneanSea. J Mar Syst 20: 423−442

Millot C (2005) Circulation in the Mediterranean Sea: evi-dences, debates and unanswered questions. Sci Mar 69: 5−21

Molinero JC, Ibanez F, Nival P, Buecher E, Souissi S (2005)The North Atlantic climate and the northwestern Medi-terranean plankton variability. Limnol Oceanogr 50: 1213−1220

Monserrat S, López-Jurado JL, Marcos M (2008) Amesoscale index to describe the regional circulationaround the Balearic Islands. J Mar Syst 71: 413−420

Perdichizzi A, Pirrera L, Giordano D, Perdichizzi F and oth-ers (2011) Distribution patterns and population structureof Illex coindetii (Cephalopoda: Ommastrephidae) in theSouthern Tyrrhenian Sea: historical series of 14 yearstrawl survey. Fish Res 109: 342−350

Pertierra JP, Sánchez P (2005) Distribution of four Cephalo -poda species along the Catalan coast (NW Mediterra -nean) using GIS techniques. Phuket Mar Biol Cent ResBull 66: 283−289

Piatkowski U, Pierce GJ, Morais da Cunha M (2001) Impactof cephalopods in the food chain and their interactionwith the environment and fisheries: an overview. FishRes 52: 5−10

Pierce GJ, Bailey N, Stratoudakis Y, Newton A (1998) Distri-bution and abundance of the fished population of Loligoforbesi in Scottish waters: analysis of research cruisedata. ICES J Mar Sci 55: 14−33

Pierce GJ, Valavanis VD, Guerra A, Jereb P and others(2008) A review of cephalopod−environment interactionsin European Seas. Hydrobiologia 612: 49−70

Pierce GJ, Allcock L, Bruno I, Bustamante P and others(2010) Cephalopod biology and fisheries in Europe. ICESCoop Res Rep No. 303. ICES, Copenhagen

Pinot JM, López-Jurado JL, Riera M (2002) The CANALESexperiment (1996–1998). Interannual, seasonal, andmeso scale variability of the circulation in the BalearicChannels. Prog Oceanogr 55: 335−370

Quetglas A, Ordines F, Valls M (2011) What drives seasonalfluctuations of body condition in a semelparous incomebreeder octopus? Acta Oecol 37: 476−483

Quetglas A, Ordines F, Hidalgo M, Monserrat S and others

117

Mar Ecol Prog Ser 514: 105–118, 2014118

(2013) Synchronous combined effects of fishing and cli-mate within a demersal community. ICES J Mar Sci 70: 319−328

Quetglas A, Valls M, Ordines F, de Mesa A, Olivar MP,Keller S, Massutí E (2014) Structure and dynamics ofcephalopod assemblages in the water column on shelf-break and slope from the western Mediterranean. J MarSyst 138:150–159

Sánchez P, González ÁF, Jereb P, Laptikhovsky V, MangoldKM, Nigmatullin CM, Ramos F (1998) Illex coindetii. In: Rodhouse PG, Dawe EG, O’Dor RK (eds) Squid recruit-ment dynamics. The genus Illex as a model, the commer-cial Illex species and influences on variability. FAO FishTech Pap No. 376. FAO, Rome, p 59−76

Sánchez P, Maynou F, Demestre M (2004) Modelling catch,effort and price in a juvenile Eledone cirrhosa fisheryover a 10-year period. Fish Res 68: 319−327

Sartor P, Belcari P, Carbonell A, González M, Quetglas A,Sánchez P (1998) The importance of cephalopods to trawlfisheries in the western Mediterranean. S Afr J Mar Sci20: 67−72

Semmens JM, Pecl GT, Gillanders BM, Waluda CM and others (2007) Approaches to resolving cephalopod move-ment and migration patterns. Rev Fish Biol Fish 17: 401−423

Smith JM, Macleod CD, Valavanis V, Hastie L and others(2013) Habitat and distribution of post-recruit life stagesof the squid Loligo forbesii. Deep-Sea Res II 95: 145−159

Sobrino I, Silva L, Bellido JM, Ramos F (2002) Rainfall,river discharges and sea temperature as factors affect-ing abundance of two coastal benthic cephalopod spe-cies in the Gulf of Cadiz (SW Spain). Bull Mar Sci 71: 851−865

Spencer PD (2008) Density-independent and density-dependent factors affecting temporal changes in spatialdistributions of eastern Bering Sea flatfish. FishOceanogr 17: 396−410

Stenseth NC, Ottersen G, Hurrell JW, Mysterud A and oth-ers (2003) Studying climate effects on ecology throughthe use of climate indices: the North Atlantic Oscillation,El Niño Southern Oscillation and beyond. Proc R SocLond B Biol Sci 270: 2087−2096

Straile D, Stenseth NC (2007) The North Atlantic Oscillation

and ecology: links between historical time-series, andlessons regarding future climate warming. Clim Res 34: 259−262

Swain DP (1999) Changes in the distribution of Atlantic cod(Gadus morhua) in the southern Gulf of St Lawrence —effects of environmental change or change in environ-mental preferences? Fish Oceanogr 8: 1−17

Tamdrari H, Castonguay M, Brêthes J, Duplisea D (2010)Density-independent and -dependent habitat selectionof Atlantic cod (Gadus morhua) based on geostatisticalaggregation curves in the northern Gulf of St Lawrence.ICES J Mar Sci 67: 1676−1686

Vargas-Yáñez M, Moya F, García-Martínez M, Rey J,González M, Zunino P (2009) Relationships betweenOctopus vulgaris landings and environmental factors inthe northern Alboran Sea (Southwestern Mediterra -nean). Fish Res 99: 159−167

Waluda CM, Pierce GJ (1998) Temporal and spatial patternsin the distribution of squid Loligo spp. in United King-dom waters. S Afr J Mar Sci 20: 323−336

Waluda CM, Yamashiro C, Rodhouse PG (2006) Influence ofthe ENSO cycle on the light-fishery for Dosidicus gigasin the Peru Current: an analysis of remotely sensed data.Fish Res 79: 56−63

Wang J, Pierce GJ, Boyle PR, Denis V, Robin J, Bellido JM(2003) Spatial and temporal patterns of cuttlefish (Sepiaofficinalis) abundance and environmental influences — acase study using trawl fishery data in French Atlanticcoastal, English Channel, and adjacent waters. ICESJ Mar Sci 60: 1149−1158

Woillez M, Poulard JC, Rivoirard J, Petitgas P, Bez N (2007)Indices for capturing spatial patterns and their evolutionin time, with application to European hake (Merlucciusmerluccius) in the Bay of Biscay. ICES J Mar Sci 64: 537−550

Zuur AF, Pierce GJ (2004) Common trends in northeastAtlantic squid time series. J Sea Res 52: 57−72

Zuur AF, Mira A, Carvalho F, Ieno EN, Saveliev AA, SmithGM, Walker NJ (2009) Negative binomial GAM andGAMM to analyse amphibian roadkills. In: Zuur AF, IenoEN, Walker NJ, Savelieu AA, Smith GM (eds) Mixedeffects models and extensions in ecology with R. Springer,New York, NY, p 383−397

Editorial responsibility: Konstantinos Stergiou, Thessaloniki, Greece

Submitted: December 5, 2013; Accepted: July 25, 2014Proofs received from author(s): September 30, 2014