Correlations between benthic habitats and demersal fish assemblages — A case study on the Dogger Bank (North Sea)

Journal of Sea Research 80 (2013) 12–24

Contents lists available at SciVerse ScienceDirect

Journal of Sea Research journal homepage: www.elsevier.com/locate/seares

Correlations between benthic habitats and demersal fish assemblages — A case study on the Dogger Bank (North Sea) Anne F. Sell a,⁎, Ingrid Kröncke b a b

Johann Heinrich von Thünen-Institut, Institute of Sea Fisheries, Palmaille 9, 22767 Hamburg, Germany Senckenberg am Meer, Marine Research Dept., 26382 Wilhelmshaven, Germany

a r t i c l e

i n f o

Article history: Received 16 July 2012 Received in revised form 23 December 2012 Accepted 28 January 2013 Available online 15 February 2013 Keywords: Groundfish Epifauna Infauna Biodiversity Community Ecology Cluster Analysis

a b s t r a c t The interdependence between groundfish assemblages and habitat properties was investigated on the Dogger Bank in the North Sea. Abiotic habitat parameters considered included topography, hydrographic conditions, sediment composition, and the biotic habitat variable the prevailing benthic invertebrates. Distinct epi- and infauna communities occurred at different locations on the Dogger Bank. Fish assemblages were clearly linked to both the biotic and abiotic habitat characteristics. Overall, fish and benthic communities revealed similar spatial distribution, represented in the respective clusters of characteristic and abundant species. Distribution patterns corresponded with the prevailing abiotic conditions such as depth and sediment composition, which appear to relate to autecological preferences of individual species. The apparently most generalist species, grey gurnard (Eutrigla gurnardus) and dab (Limanda limanda) occurred at all stations and dominated in terms of biomass in most cases. The absolute numbers of grey gurnards were related to the abundance of suitable prey, invertebrate and fish species, which stomach analyses revealed as part of the diet in an independent study during the same research cruise. Haddock (Melanogrammus aeglefinus) and whiting (Merlangius merlangus) were only abundant at deep stations along the flanks of the bank. The occurrence of lemon sole (Microstomus kitt), American plaice (Hippoglossoides platessoides) and cod (Gadus morhua) was also positively correlated with depth, whereas especially lesser weever (Echiichthys vipera), sandeel species and solenette (Buglossidium luteum) occurred predominantly at the shallower sites. At the same time, individual fish species such as solenette and lesser weever were associated with high densities of selected epi- or infauna species. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Distribution patterns of fish often depend on the spatial extent of appropriate habitat. This is particularly clear where complex habitats such as coral reefs or sea mounts lead to local accumulations of habitat-specific fish (e.g., Fock et al., 2002; Parrish and Boland, 2004). For the North Sea, habitat effects on small-scale distribution of fish have been studied around prominent physical structures, especially oil platforms (Løkkeberg et al., 2002; Soldal et al., 2002). However, for the more typical sand-bottom habitats of the North Sea, fewer investigations exist that report on the importance of particular in- and epifauna communities providing a biotic habitat for groundfish (Ellis et al., 2011; Kaiser et al., 2004; Reiss et al., 2010; Ryer et al., 2004). Sandbank habitats have attracted greater interest recently, when some EU member states designated Natura 2000 sites under the Habitats Directive (European Commission, 1992), aimed at the protection ⁎ Corresponding author. Tel.: +49 40 38905 246; fax: 49 40 38905 263. E-mail address: [email protected] (A.F. Sell). 1385-1101/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.seares.2013.01.007

of this particular type of benthic habitat with its associated marine communities. The Dogger Bank is the largest sandbank in the North Sea, covering sections of the Exclusive Economic Zones of four bordering states (UK, NL, GE, DK). Most of the bank's area has been declared as Natura 2000 sites, for which possible protection measures are presently being debated at the time of writing. The installation of appropriate management measures, geared to protect the structure itself and its associated typical fauna, calls for a detailed analysis of the interdependence of species and habitats, including invertebrate fauna and fish. The Dogger Bank is a sandbank with distinct ecological features (Kröncke and Knust, 1995; van Moorsel, 2011). Different salinities, temperatures and seasonal variability in both parameters distinguish the different water masses around the Dogger Bank. Waters in shallow areas on top of the bank are nearly permanently mixed, whereas seasonal stratification occurs in the deeper areas. Both, morphological heterogeneities as well as the co-occurrence of different water masses are known to support the development of fronts (Otto et al., 1990) with a related potential for enhanced primary and secondary production (Hill et al., 1994). Primary production on the Dogger Bank occurs throughout the year with higher values in winter (January and February)

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

than in any other area in the southern North Sea (Brockmann and Wegner, 1985; Howarth et al., 1994; Richardson and Olsen, 1987). The benthic infauna communities on the Dogger Bank have been studied since the 1920s by Danish, British and German scientists (e.g., Sell et al., 2007; Wieking and Kröncke, 2005; reviewed in Kröncke, 2011). In the present study, our objective was to build upon the initial analyses of fish and invertebrate communities presented in Sell et al. (2007) in order to evaluate how strongly the spatial structure of groundfish assemblages on the Dogger Bank depend on the different local habitat features. We characterize benthic habitats based on abiotic parameters as well as the biotic components and relate these habitat properties to the composition of groundfish communities. Two hypotheses are being tested: (1) Groundfish assemblages depend primarily on bank topography (association with abiotic habitat parameters), and (2) The distribution of groundfish assemblages is related to the composition of benthic in- and/or epifauna communities, which provide potential prey resources (association with biotic habitat parameters). 2. Methods 2.1. Sampling and sample treatment Bottom fish assemblages on the Dogger Bank (North Sea) were analyzed during a research cruise with FRV “Walther Herwig III” from April 28–May 9, 2006, using the GOV otter trawl and sorting the catch, both as described in the manual for the International Bottom Trawl Survey (IBTS) (ICES, 2006). All the sampling was conducted during daytime. During the cruise WH 287, 35 stations across the bank's area were sampled for the fish assemblages, benthic epifauna and abiotic habitat parameters; infauna was sampled at 24 of those stations (Fig. 1, Supplementary Table S1). The depth intervals sampled were: b30 m, 6 stations; 30 b 40 m, 11 stations; 40 b 50 m, 8 stations, and > 50 m, 10 stations. Abiotic habitat parameters measured included CTD profiles from a Seabird probe and sediment samples taken with a van Veen grab. In the lab, sediment fractions of gravel (grain size > 2 mm), sand (>63 μm–2 mm) and mud (b63 μm) were separated by wet-sieving over the respective mesh sizes. Two replicates per station of benthic infauna were taken with 0.2 m 2 van Veen grabs as well and were processed independently before averages of the results were taken. Epifauna and small bottom fish at the sea bed were sampled with a 2-m beam trawl, equipped with a mat of tickler chain and a net with 20-mm mesh size and a liner with 4-mm knotless mesh as described in Jennings et al. (1999). The net was applied at the same stations at which the GOV hauls were taken and towed for 5 min at a speed of 2 knots over ground. In the smaller fraction of fish caught with the 2-m beam trawl minimum length of fish caught was 3–4 cm and maximum length typically around 25 cm — outliers and elongated species like Entelurus aequoreus excepted. In the GOV, fishes with less than 4 cm length may occur, but in low numbers, and maximum lengths of fishes caught can reach 2 m, but were during this cruise around 60 cm. For easier phrasing, these two groups are in the following also called “small fish” from the 2-m beam trawl and “large fish” from the GOV. Epifauna species were counted and weighed on board ship, infauna samples were preserved in 4% formaldehyde and were post-processed in the laboratory. 2.2. Data analysis GOV haul target duration was 30 min, but in a few cases where haul duration differed by 1 min, count data of demersal fish were standardized to 30 min for comparison. For purposes of visualizing

13

the numbers of individuals caught, data were transformed to the swept area based on the net geometry (wingspread) of the individual GOV hauls and towed distance. Epifauna and fish data from catches with the 2-m beam trawl were also standardized to 1000 m 2 (roughly three times the area covered per haul), and for infauna data sampled with the van Veen grab to 1 m 2. For statistical analyses, abundance data were square-root transformed to reduce the impact of very abundant over the rarer species. The vast majority of the epifauna was identified to species level, only few taxa to a higher taxonomic level. Pelagic fish species present in the GOV hauls were excluded from the analyses (herring — Clupea harengus, sprat — Sprattus sprattus, Atlantic mackerel — Scomber scombrus, and European anchovy — Engraulis encrasicolus) because their abundance is not expected to be directly related to benthic habitats. None of these taxa was present in the beam trawl. Although sandeel species (Ammodytidae) feed in the pelagial, they were not excluded here, as they are a dominant element of sandbank communities (Ellis et al., 2011; Kaiser et al., 2004), while they are unlikely to be dependent on the benthic invertebrate fauna in their habitat, they can themselves be expected to form a critical habitat component as prey. Sandeel are known to be hunted by many predatory fish species, forming a significant part of their diet, and aggregative responses to sandeel have been observed (Engelhard et al., 2008; Weinert et al., 2010). Initial data exploration and canonical correspondence analyses (CCA) of the relations between fish assemblages and habitat parameters were performed with BRODGAR (Highland Statistics Ltd., version 2.6.6) and CANOCO, version 4.5 (ter Braak and Smilauer, 1998). Further statistical calculations were performed using the software package PRIMER — Plymouth Routines In Multivariate Ecological Research, version 6 (Clarke and Gorley, 2006). Cluster analyses were performed on the level of individual functional groups, e.g., benthic infauna, and with pre-treatment of the data in the form of Bray–Curtis similarity matrices of square-root transformed species abundances. Similarity profile permutation tests (SIMPROF) were applied to test for significant differences between groups identified through cluster analysis in the a priori unstructured set of samples (stations with their specific species assemblage). The SIMPER routine within the PRIMER program was used to examine the contribution of each species to the similarity within a group of assemblages, and to identify the dissimilarities between the groups. Relationships between the assemblages of larger groundfishes (GOV catches, clusters with 60% similarity) and the characteristics of their abiotic and biotic habitats were investigated. Analyses of similarities (ANOSIM) or RELATE routines were performed to reveal potential differences between the fish assemblages depending on a number of abiotic factors, particularly each station's depth, temperature, salinity and sediment properties. Similarity matrices used in these routines were based on Bray–Curtis ordination for abundance data and on Euclidean distance for (normalized) abiotic environmental data. ANOSIM was applied to test whether a particular parameter (e.g., sediment category) is significantly correlated to the composition of assemblages. It was used in analogy to a univariate ANOVA as a method of permutationbased hypothesis testing to look for differences between a priori defined groups of samples. RELATE measures how closely related two sets of multivariate data are, for a matching set of samples by calculating a rank correlation coefficient between all the elements of their respective (dis)similarity matrices. RELATE was used here for relation of species assemblages to multiple abiotic data, including both categorical and continuous data, or to compare two sets of abundance data from different taxa for the same stations. For all SIMPROF, ANOSIM or RELATE analyses, 999 permutations were conducted. BEST analyses employ rank correlation between two resemblance matrices and were used to investigate relationships between faunal communities and a suite of environmental data – here, the combination of the factors depth, temperature and salinity – for which with resemblance matrices were prepared based on Euclidean distance.

14

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

Fig. 1. Stations sampled during cruise WH 287 on the Dogger Bank, applying a GOV otter trawl, 2-m beam trawl, sediment grab and Seabird CTD. Figure adapted from Sell et al. (2007).

3. Results 3.1. Bottom fish assemblages 3.1.1. Large fish — GOV otter trawl A total of 32 species of demersal groundfish were present in the GOV otterboard trawl hauls on the Dogger Bank, of which grey gurnard (Eutrigla gurnardus; Fig. 2a) and dab (Limanda limanda) occurred in all hauls and also dominated the biomass of most catches. Plaice (Pleuronectes platessa) was included in 94% of all hauls, but ranked only seventh in average abundance. Typically, more fish were caught

at the Tail End (north-eastern area of the bank) and at the southwestern flank than at the shallow stations on the bank itself, with groundfish biomasses of often > 2 kg 1000 m −² at the Tail end and in the southwest (Fig. 3). Three significantly different clusters of groundfish species assemblages could be identified based on a SIMPROF analysis (Pi = 1.245, p b 0.003; Fig. 4a). Lesser sandeel (Ammodytes marinus) was only abundant in the northern and eastern regions of the bank (orange cluster in Fig. 4a), and sandeel contributed most to the dissimilarity between the groundfish assemblages here and elsewhere on the bank (Table 1). Consistent with the general distribution of ground

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

15

Fig. 2. Single-species distributions; symbol size represents number of individuals caught per swept area; (a) grey gurnard (Eutrigla gurnardus); (b) lesser weever (Echiichthys vipera). Crosses mark zero hauls. Individual abundance values [ind 1000 m−2] are given next to each symbol. Figure adapted from Sell et al. (2007).

fish, the abundance of grey gurnard as one of the dominant species was highest at the north-eastern “Tail End” of the Dogger Bank (Fig. 2a). The otherwise rare or – in most cases – absent lesser weever Echiichthys vipera (Fig. 2b) occurred in very high numbers of 34 individuals per 1000 m 2 swept area (52.5 kg per 30-min haul) at one of the shallowest stations, on the Southwest Patch (yellow in Fig. 4a). E. vipera was absent at the deeper areas around the bank. In contrast, whiting (Merlangius merlangus) and haddock (Melanogrammus aeglefinus) were only abundant in the cluster representing the deeper stations along the flanks of the bank (blue in Fig. 4a) and contributed significantly to its distinction from the other clusters on the bank itself (Table 1).

3.1.2. Small fish — 2-m beam trawl Overall, 21 fish species occurred in the 2-m beam trawl catches, with an average catch rate of 0.94 kg fish 1000 m −2, equivalent to 79 ind 1000 m −2. The five most abundant species were solenette — Buglossidium luteum, scaldfish — Arnoglossus laterna, dab — L. limanda, sand goby — Pomatoschistus minutus, and lesser sandeel — A. marinus. Cluster analyses revealed two main clusters, characteristic of the bank and of the flanks, respectively (Fig. 4b; Table 2). No significant difference existed between both clusters (SIMPROF analysis, Pi = 3.077, p b 0.003) and was primarily due to the higher abundances of flatfish on the bank, especially solenette, which contributed 36% to the total dissimilarity, but also dab and scaldfish. The snake pipefish

16

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

Fig. 3. Total fish biomass in GOV hauls at individual stations on the Dogger Bank, standardized to swept area.

E. aequoreus was more abundant in the deeper areas around the bank, whereas A. marinus, Callionymus lyra and E. vipera only occurred in the beam trawl catches on top of the bank. The distribution patterns of assemblages were similar between the fishes caught with the two different nets (RELATE, rho = 0.197, p b 0.01). 3.2. Abiotic habitat parameters 3.2.1. Hydrography The cruise took place at the end of spring, with a still mostly mixed water column around the bank. First signs of the onset of stratification were visible in the hydrographic profiles, but with a maximum temperature difference of about 2 °C, none of the off-bank stations was yet fully stratified. Bottom temperatures on the bank ranged from 5.8 to 8.1 °C (Table S1). Central North Sea water to the north of the Dogger Bank was more strongly saline in April/May 2006 than in the long-term average (Gerd Wegner, pers. comm.). Highest salinities were recorded at the stations along the northern flank of the bank, with transect II in the northeast showing a stronger gradient of increasing salinities toward the northern flank than transect I (Fig. 1). 3.2.2. Sediments Each of the sampled sediments consisted of at least 85% sand, most of them >95%. All sediments were classified as sand (“s”), unless they contained more than 2% of mud (stations labeled “sm”) or gravel (“sg”), respectively. Most of the stations sampled (22 out of 35) were assigned the sediment class ‘sand’ and were mainly located on the bank itself. Sediments at ten deeper stations along the flanks were classified as “muddy sand”, and three stations in the shallowest regions were classified as “gravelly sand”. 3.3. Biotic habitat parameters 3.3.1. Benthic infauna In total, 151 infauna taxa were observed and included in the analysis. The dominant taxonomic groups were polychaetes and crustaceans (53 taxa, each), followed by molluscs (29 taxa) and echinoderms (9 taxa). Cluster analysis of infauna communities, based on the square

root-transformed abundance data, revealed six significantly different clusters, of which three were more widely distributed across specific sections of Dogger Bank (SIMPROF analysis: Pi = 3.89, p b 0.001; Fig. 5a). Dominant species of the “Bank Community” (Table 3; orange in Fig. 5a) were the polychaete Spiophanes bombyx, the amphipod Bathyporeia elegans, the brittle star Acrocnida brachiata and another polychaete, Magelona johnstoni. On the shallowest section of the bank, the infauna formed a separate cluster (“South-west patch Community”) at two stations (399, 400). This cluster was distinguished by the highest observed numbers of the polychaete Nephtys cirrosa (34 ind m−2), as opposed to 7 ind m −2 in the general Bank Community, while it contained abundances of S. bombyx (215 ind m−2), which were similar to those of other stations on the greater bank area (255 ind m −2). Infauna of the deeper and higher saline stations at the northeast on the northern flank of the Dogger Bank, close to the 50-m depth contour, formed another cluster where S. bombyx and B. elegans were dominant again, but accompanied by the characterizing polychaete species Scoloplos armiger, Lanice conchilega, Ophelia limacina and Magelona filiformis. The infaunal assemblage along the southern flank belonged to a separate cluster (blue in Fig. 5a) based on dominance of the brittle star Amphiura filiformis with a mean number of [around] 500 ind m−2 (incl. 74 Amphiura spec. juv.), accompanied by the bivalves Kurtiella bidentata, Nucula nitidosa, Thyasira flexuosa and Abra nitida. The polychaete Pholoe baltica and the echiurid Echiurus echiurus were also more abundant in this community. At the different stations, only minor differences existed between sediment types, which did not have a significant impact on the infauna communities (ANOSIM: global R = 0.19, p = 0.074). In contrast, the combination of the prevailing environmental parameters depth, temperature and salinity did significantly influence the composition of infauna communities (BEST analysis: rho = 0.597, p b 0.01). 3.3.2. Epifauna communities — 2-m beam trawl Overall, 121 epifaunal taxa were observed. Note that, strictly speaking, not all species caught in the 2-m beam trawl are truly or exclusively epibenthic. However, the vast majority of the species is, and we hence use this nomenclature here for the entire assemblage caught

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

17

A

B

Fig. 4. Result of cluster analysis of demersal fish assemblages, based on square root-transformed abundance data and a Bray–Curtis similarity matrix. (a) Large-fish component, sampled with the GOV. Symbols of the same color represent stations where assemblages show at least 60% similarity. (b) Small-fish component, sampled with a 2-m beam trawl. Symbols of the same color represent stations where assemblages show at least 45% similarity.

with this net. Since samples were taken during the day only, diurnal migration of epifauna can be neglected. A cluster analysis with a SIMPROF test revealed seven significantly different types of epifauna communities (p = 2.058, p b 0.001). However, species compositions at most stations belonged to one of only two major clusters, a Bank Community and a Western/Northern Flank Community (Fig. 5b, Table 4). Dominant species on the bank in terms of abundance were the starfishes Asterias rubens and Astropecten irregularis, the hermit crab Pagurus bernhardus and the masked crab Corystes cassivelaunus, all with ≥ 20 ind 1000 m −2. On the bank, one of the shallowest stations (399) was exceptional due to the high numbers of Crangon crangon (210 ind 1000 m −2). The main distinction of the Western/Northern Flank Community from the Bank Community was the lack of the masked crab

C. cassivelaunus on the flanks. Furthermore, the anthozoans Epizoanthus incrustatus and Alcyonium digitatum as well the hermit crab Anapagurus laevis occurred specifically on the deeper flank areas and were characteristic of this community. Overall, the Western/Northern Flank Community was also dominated by A. rubens, A. irregularis and P. bernhardus, with A. rubens occurring at somewhat higher abundances compared with the Bank Community. Two stations on the flank in the northeast (red symbols in Fig. 5b) differed from other stations by higher abundance of all three generally dominant species (A. rubens, A. irregularis and P. bernhardus), by the presence of the amphipod Gammaropsis nitida and the absence of E. incrustatus. As for the infauna, the small differences in sediment composition between stations did not significantly impact the epifauna communities

18

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

Table 1 Bottom fish assemblages on the Dogger Bank: comparison of three major clusters, identified from square root-transformed abundance data from GOV ottertrawl catches. Breakdown of average dissimilarities between the assemblages into contributions from the dominant species (Contrib. %) and cumulative contributions (Cum. %; SIMPER analysis). Species listed by contribution in descending order; each cluster comparison limited to the 7 species which contribute most to the total dissimilarity. First two columns: untransformed abundance data standardized to [ind 30 min−1 haul duration]. Color-coding of clusters as in Fig. 4a. Species

Ammodytes marinus Eutrigla gurnardus Limanda limanda Echiichthys vipera Merlangius merlangus Hyperoplus lanceolatus Pleuronectes platessa

[ind 30 min−1] [ind 30 min−1] Contrib. % Cum. % Red

Orange

13.0 147.4 497.3 9.9 7.7 0.4 10.8

882.1 185.8 669.3 8.5 11.6 3.1 6.9

Blue Ammodytes marinus 0.1 Limanda limanda 1352.0 Merlangius merlangus 139.2 Eutrigla gurnardus 132.9 Melanogrammus aeglefinus 27.7 Echiichthys vipera 0.0 Microstomus kitt 8.2 Red Limanda limanda 497.3 Merlangius merlangus 7.7 Eutrigla gurnardus 147.4 Melanogrammus aeglefinus 0.5 Ammodytes marinus 13.0 Echiichthys vipera 9.9 Microstomus kitt 0.3

43.4 13.1 10.2 7.2 5.1 3.2 3.1

43.4 56.4 66.6 73.9 78.9 82.1 85.2

36.1 15.0 11.7 9.1 6.3 3.6 2.9

36.1 51.1 62.9 71.9 78.3 81.9 84.8

26.4 16.6 10.2 9.1 6.7 5.5 4.5

26.4 43.0 53.2 62.3 68.9 74.4 78.9

Orange 882.1 669.3 11.6 185.8 0 8.5 0.6 Blue 1352.0 139.2 132.9 27.7 0.1 0.0 8.2

Clusters red and orange, average dissimilarity = 41.79. Clusters blue and orange, average dissimilarity = 48.55. Clusters red and blue, average dissimilarity = 41.19.

Table 2 Bottom fish assemblages on Dogger Bank from catches with a 2-m beam trawl. Comparison of two main clusters, identification based on Bray–Curtis coefficients for square root-transformed abundance data. Breakdown of average dissimilarities between the assemblages into contributions from the individual species (SIMPER analysis). Table limited to the 7 species which contribute most to the total dissimilarity. First two columns: untransformed abundance data standardized to [ind 1000 m−2]. Compare with clusters in Fig. 4b (Flank Community = blue; Bank Community = orange). Species

Buglossidium luteum Limanda limanda Arnoglossus laterna Entelurus aequoreus Pomatoschistus minutus Ammodytes marinus Microstomus kitt

Flank Community

Bank Community

[ind 1000 m−2]

[ind 1000 m−2]

1.4 4.0 6.2 5.8 8.6 0.0 1.4

73.2 10.0 15.4 1.4 6.2 2.4 0.0

Contrib. %

Cum. %

35.9 9.2 8.5 7.9 7.4 6.3 6.0

35.9 45.1 53.6 61.6 68.9 75.2 81.2

Flank and bank communities, average dissimilarity = 55.78.

(ANOSIM: global R = 0.006, p = 0.435). In contrast, the combination of the factors depth, temperature and salinity did influence the benthic epifauna significantly (BEST analysis: rho = 0.315, p b 0.01). 3.4. Links between groundfish assemblages and habitat characteristics GOV fish assemblages were clearly different between depth intervals (ANOSIM: global R = 0.408, p b 0.1%). Pairwise comparisons reveal

that the neighboring depth intervals were not significantly different (b30 m versus 30 b 40 m, or 40 b 50 m versus > 50 m). Across greater depth differences however, species assemblages could be separated at significance levels of p b 0.1% (valid for all pairs: b30 m versus 40 b 50 m, b30 m versus > 50 m, or 30 b 40 m versus > 50 m). For the smaller groundfish caught in the 2-m beam trawl, a RELATE analysis confirms that assemblages were significantly different in terms of their dependence on bottom temperature (rho = 0.352, p b 0.001). Pairwise comparison of the fish assemblages in different depth intervals shows – as for the large fish caught in the GOV – that the neighboring intervals b30 m versus 30 b 40 m, or 40 b 50 m versus > 50 m could not be separated, but significant differences existed between all other pairs (ANOSIM: global R = 0.408, p b 0.001). The greatest dissimilarity of 69% occurred between species assemblages of the shallowest (b 30 m) and the deepest intervals (>50 m), mainly due to the high abundances of B. luteum, A. laterna, A. marinus, and L. limanda at depth b 30 m, as opposed to higher abundances of E. aequoreus, P. minutus and Microstomus kitt at stations of ≥50 m depth. The minor differences between sediment types did not have a significant impact on the assemblages of small groundfish in the beam trawl (ANOSIM: global R = 0.041, p = 0.28). The composition of groundfish communities was also correlated with biotic properties of their habitat. Both the assemblages of small fishes in the 2-m beam trawl and those of the larger fish in the GOV showed a significant correlation with the composition of the ambient epifauna (RELATE analysis: small fish: rho = 0.312, p b 0.001; large fish: rho = 0.287, p b 0.01). Additionally both, assemblages of small and large fish, were significantly related to the composition of the infauna (RELATE analyses: rho = 0.354, p b 0.01 and rho = 0.35, p b 0.001, respectively). These correlations could in principle just be a reflection of a dependence of benthic epi- and infauna on their own abiotic habitat properties, leading to a similarity of patterns, where both fish and benthic organism — composition is primarily structured by the prevailing abiotic conditions. Therefore, further analyses were performed to investigate the relative importance of various habitat characteristics for individual species of fish: Canonical correspondence analyses (CCA) were performed to simultaneously evaluate the relationships between several habitat parameters and groundfish assemblages. For the groundfish caught in the GOV, a Monte-Carlo test confirmed significant dependencies between the abiotic habitat parameters and species distribution patterns (test with 999 permutations; first canonical axis: F-ratio = 10.91, p = 0.0010; all canonical axes: F-ratio = 4.19, p = 0.0010). With respect to the abundance of the cluster-defining fish species in the GOV, depth was the abiotic habitat parameter with the strongest influence, followed by temperature. The group of characteristic groundfish species overall showed a significant positive correlation with depth (0.857) and a negative correlation with temperature (−0.689). Correlations of their abundance with salinity and sediment type were non-significant (− 0.121 and 0.494 respectively). Yet, individual species, particularly haddock (M. aeglefinus) and to a lesser degree whiting (M. merlangus), were linked to the deeper and muddier locations (Fig. 6). Lemon sole (M. kitt), and American plaice (Hippoglossoides platessoides) were also clearly positively correlated with depth, in contrast to sandeel species, solenette (B. luteum) and particularly lesser weever (E. vipera), which occurred predominantly at the shallower and warmer sites. A canonical correspondence analysis confirmed that the high abundance of haddock at the deepest, coldest and muddiest stations was also correlated with high abundance of the octocoral A. digitatum and the hermit crab A. laevis (Fig. 6). Nevertheless, its relationship to depth and sediment type was stronger than to the occurrence of the epifauna species mentioned. The numbers of solenette (B. luteum) and greater sandeel (Hyperoplus lanceolatus) were to a large degree explained by the reduced depths of the respective stations, and at the same time by the abundance of amphipods, especially B. elegans

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

19

A

B

Fig. 5. Result of cluster analysis of benthic invertebrate assemblages, based on square root-transformed abundance data and a Bray–Curtis similarity matrix. (a) Infauna; symbols of the same color represent stations where assemblages show at least 45% similarity (b) epifauna; symbols of the same color represent stations where assemblages show at least 40% similarity.

(Fig. 6), being a possible prey for these fish species. The high abundances of lesser weever (E. vipera) and greater sandeel (H. lanceolatus) at two shallow stations (399 and 414) were correlated with high temperatures. E. vipera was also stronger correlated with the high abundance of C. crangon in the epifauna as well as in the infauna. A corresponding CCA was performed for the small groundfish caught in the 2-m beam trawl. Again, abundances of the clusterdefining species were significantly related to depth (positively, 0.815) and temperature (negatively, − 0.809), and not with salinity or sediment type. M. kitt was the species most clearly increasing in abundance in response to increasing depth, explaining spatial differences in its occurrence almost entirely. In contrast, B. luteum was much more abundant at shallow stations. Scaldfish (A. laterna) was

least of all fish species from the 2-m beam trawl influenced by any of the abiotic habitat parameters. 4. Discussion 4.1. Abiotic and biotic properties of fish habitats In the analysis of Dogger Bank habitats, physical parameters such as temperature, depth, and to a lesser degree salinity, proved to be major factors structuring the communities of groundfish, as well as of the potential prey, belonging to the benthic infauna and epifauna. Cluster analysis of infauna communities revealed three main clusters, which broadly resembled the spatial distribution of communities on

20

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

Table 3 Benthic infauna on the Dogger Bank: comparison of three major clusters, based on square root-transformed abundance data from van Veen-grab samples. Breakdown of average dissimilarities between the assemblages into contributions from the 12 most dominant species (SIMPER analysis). Table limited to the 10 taxa each which contribute most to the total dissimilarity. First two columns: untransformed abundance data standardized to [ind m−2]. Color-coding of clusters as in Fig. 5a: orange = Bank Community; red = North-eastern Community; blue = Southern Community.

Amphiura filiformis Kurtiella bidentata Pholoe baltica Amphiura spec. Nucula nitidosa Thyasira flexuosa Bathyporeia elegans Echiurus echiurus Abra nitida Spiophanes bombyx

Acrocnida brachiata Spiophanes bombyx Magelona johnstoni Bathyporeia elegans Bathyporeia guilliamsoniana Urothoe poseidonis Echinocyamus pusillus Scoloplos armiger Lanice conchilega Edwardsia spec.

Amphiura filiformis Amphiura spec. Kurtiella bidentata Bathyporeia elegans Pholoe baltica Nucula nitidosa Spiophanes bombyx Thyasira flexuosa Magelona johnstoni Echiurus echiurus

[ind m−2]

[ind m−2]

North-eastern Community

Southern Community

2.7 0.2 1.8 1.0 1.8 0.0 66.6 0.6 0.3 113.0

436.0 100.2 87.8 73.6 66.6 46.0 2.7 39.3 20.5 100.2

North-eastern Community

Bank Community

0.7 113.0 1.4 66.6 0.6 0.6 8.8 48.6 43.4 19.4

69.9 255.0 55.8 95.1 35.2 30.4 24.9 6.4 22.3 3.6

Southern Community

Bank Community

436.0 73.6 100.2 2.7 87.8 66.6 100.2 46.0 0.9 39.3

5.5 0.0 3.6 95.1 1.6 0.0 255.0 0.2 55.8 0.0

Contrib. %

9.44 4.68 3.93 3.64 3.51 3.33 3.2 3.03 2.32 2.2

Cum. %

9.44 14.12 18.05 21.69 25.21 28.54 31.74 34.77 37.09 39.29

Table 4 Epibenthic invertebrates on Dogger Bank: comparison of two main clusters, identified from square root-transformed abundance data from catches with a 2-m beam trawl. Breakdown of average dissimilarities between the assemblages into contributions from the 12 most dominant species (SIMPER analysis). Table limited to the 10 species which contribute most to the total dissimilarity. (Individual species, e.g., A. irregularis may be considered semi-epibenthic as the burrow at certain time intervals.) First two columns: untransformed abundance data standardized to [ind 1000 m−2]. Color-coding of clusters as in Fig. 5b: orange = bank, blue = western and northern flank epifauna.

Corystes cassivelaunus Epizoanthus incrustatus Asterias rubens Pagurus bernhardus Alcyonium digitatum Astropecten irregularis Anapagurus laevis Psammechinus miliaris Ophiothrix fragilis Luidia sarsi

[ind 1000 m−2]

[ind 1000 m−2]

W/N Flank Community

Bank Community

1.0 12.0 32.8 26.0 7.0 27.0 6.0 0.8 1.8 1.6

20.4 0.0 22.0 26.0 0.0 25.4 0.0 6.0 1.2 0.1

Contrib. %

Cum. %

6.9 6.2 5.1 4.7 4.4 4.3 4.1 4.1 3.5 3.5

6.9 13.0 18.1 22.8 27.2 31.4 35.6 39.7 43.2 46.6

N/W Flank and Bank communities, average dissimilarity = 62.03. 3.96 3.72 3.44 3.05 2.82 2.59 2.53 2.37 2.17 2.05

3.96 7.68 11.12 14.16 16.99 19.58 22.1 24.47 26.65 28.7

7.59 3.4 3.37 3.34 3.32 3.25 3.16 2.69 2.62 2.55

7.59 10.99 14.36 17.7 21.02 24.28 27.43 30.13 32.75 35.29

North-eastern and southern communities, average dissimilarity = 61.63. North-eastern and bank communities, average dissimilarity = 56.14. Southern and bank communities, average dissimilarity = 64.15.

the Dogger Bank during the last century as described by Kröncke (2011). Characteristic species as well as the entire species composition of the separate communities were quite similar over the last 30 years, although Wieking and Kröncke (2001) and Kröncke (2011) found evidence for climate-driven changes in the abundance of warmand cold-temperate infaunal species, especially in the 1990s. The “Bank Community” is a Bathyporeia–Fabulina (Tellina) Community, the southern and western flanks are inhabited by an A. filiformis Community (Kröncke et al., 2011; Künitzer et al., 1992; Salzwedel et al., 1985; Wieking and Kröncke, 2003). Infauna of the stations on the northeastern flank of the Dogger Bank, close to the 50 m depth contour, formed a separate cluster, very similar to the previously described “North-eastern Community” (Table 3; Reiss and Kröncke, 2005; Wieking and Kröncke, 2001, 2003). The Bank Community in turn was probably affected by strong currents resulting in a mixed water column, year round. The spatial distribution and functional diversity of infauna communities are known to be structured by environmental parameters such as depth, temperature and sediments, which reflect the hydrodynamically mediated food availability (Dauwe et al., 1998; Kröncke et al., 2004; Rees et al., 2006).

In terms of epifauna diversity, the Dogger Bank is an intermediate area between the southern North Sea, where in general the diversity of epifauna is lower than in the species-rich central and northern North Sea beyond the 50-m depth contour (Callaway et al., 2002; Neumann et al., 2008; Sonnewald and Türkay, 2010; Zühlke et al., 2001). The Dogger Bank hosts a mixture of the sessile epifauna that is characteristic of the northern North Sea, and of free-living species typical of the south. A variety of studies have demonstrated that the distribution of epifauna communities on large spatial scales is mainly influenced by depth, bottom water temperature and salinity (Callaway et al., 2002; Jennings et al., 1999; Neumann et al., 2009; Reiss et al., 2010). Sediment composition plays a minor role on a large spatial scale because overall it is a reflection of the hydrodynamic energy. In contrast, on small spatial scales, Rees et al. (1999), Kaiser et al. (2000) and Hinz et al. (2004) found that substratum type appeared to be the main structuring force for epibenthic communities. In areas with a wide range of different substrate types including gravel and rocks, as studied by Rees et al. (1999), the relationship between epibenthic communities and substrate is more pronounced, due to an enhanced diversity of sessile species on coarser ground. However, as documented here, the variation in sediment types on the Dogger Bank is very low and did not appear to be responsible for differences between benthic assemblages of the separate bank regions. 4.2. Demersal fish assemblages and relation to habitat properties Species communities of both, fish and invertebrates are influenced by tabiotic habitat components on Dogger Bank, particularly depth and temperature. Furthermore, multivariate patterns in the communities of large or small demersal fish species and habitat parameters also showed clear relations with the in- and epifauna communities. The distribution patterns of assemblages were similar between the fishes caught with the 2-m beam trawl and the GOV, respectively. Although the catch efficiencies of both nets differ substantially for individual fish species – e.g., Buglossidium luteum is more efficiently caught in a beam trawl than in the GOV – a certain degree of overall similarity is expected as the assemblages are interrelated subsamples of the entire groundfish community. In contrast to an extensive analysis over the area of the entire North Sea by Reiss et al. (2010), who found no significant correlation between the large scale patterns of abundances of infauna, epifauna, and demersal fish, our study on an intermediate scale (kilometers to

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

21

Fig. 6. Canonical correspondence analysis, relating occurrence of characteristic fish species in the GOV to abiotic habitat parameters. Fish species names — blue; abundance data are square root-transformed. Abiotic parameters — red: station depth, bottom temperature, bottom salinity, and sediment type, classified by % mud. Selected epifauna and infauna species — grey, plotted to visualize occurrence with respect to abiotc habitat parameters (passive parameters in calculation of CCA). Species abbreviations — (fish): Bl: Buglossidium luteum; Ev: Echiichthys vipera; Ea: Entelurus aequoreus; Eg: Eutrigla gurnardus; Gm: Gadus morhua; Hp: Hippoglossoides platessoides; Hl: Hyperoplus lanceolatus; Ll = Limanda limanda; Ma: Melanogrammus aeglefinus; Mm: Merlangius merlangus; Mk: Microstomus kitt; Pp: Pleuronectes platessa; (epifauna and infauna): Af: Amphiura filiformis; Al: Anapagurus laevis; Be: Bathyporeia elegans; Bg: Bathyporeia guilliamsoniana; Ca: Crangon allmanni; Cc: Crangon crangon; Mj: Magelona johnstoni; Nn: Nucula nitidosa; Tf: Tellina fabula; Tyf: Thyasira flexuosa.

10th of kilometers) did reveal significant relationships between demersal fish caught with the GOV and the benthic habitat. This indicates close links between the different faunal communities, which are expected to reflect the importance of benthic invertebrates as prey for the respective ground fish species. However, the distribution patterns of individual fish species were influenced to different degrees by abiotic and biotic habitat properties. The distribution of haddock (M. aeglefinus) on the Dogger Bank was more strongly linked to depth, temperature and mud content, than to biotic factors. This relative independence from specific prey fields is consistent with studies of haddock stomach contents, which revealed a rather opportunistic feeding behavior (Mattson, 1992; Schückel et al., 2010). The snake pipefish E. aequoreus reached exceptionally high abundances in the North Sea around the year 2006 due to an extraordinary inflow of Atlantic water (Harris et al., 2007; Kloppmann and Ulleweit, 2007), and as a typically oceanic species was also more abundant in the deeper regions around the bank. In contrast, lesser weever (E. vipera) was associated with the shallowest areas on the bank and was absent elsewhere, which is consistent with other observations on habitat preferences (Kaiser et al., 2004). Ellis et al. (2011) found early life history stages of E. vipera on

sandbank crests in the North Sea, which suggested that sandbanks can provide suitable conditions as nursery grounds for this species. The positive correlation of E. vipera with temperature (Fig. 6) is supported by the findings of Perry et al. (2005), who showed that its distribution boundaries shifted in relation to both climate and time. Furthermore, Sonnewald and Türkay (2012), showed that its abundance can be directly affected by water temperature. In Portuguese waters, stomach content analyses of E. vipera have revealed a wide range of prey species (28 were found in ca. 250 stomachs), of which crustaceans – namely Mysidaceae, Amphipoda and Isopoda – were the most important component of the diet (Vasconcelos et al., 2004). We therefore suggest that on the Dogger Bank, related species such as C. crangon and B. elegans will have constituted an important element of its diet on the very shallow section of the bank. Plaice (P. platessa) also showed highest abundances on the bank itself (red cluster in Fig. 5a), where polychaetes (S. bombyx) were the numerically most dominant element of the infauna. The correlation of plaice occurrence with high numbers of polychaetes observed in the present study agrees with the analyses of Rijnsdorp and Vingerhoed (2001) and Schückel et al. (2011). The former authors documented that during the 20th century, dominance in prey composition for plaice has shifted from bivalves to polychaetes, which is in

22

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

line with the hypothesis that the abundance of small opportunistic benthic species such as polychaetes has increased due to fishing activities in heavily trawled areas of the North Sea. Thus, during the first half of the last century, extensive patches of the bivalve Spisula were present on Dogger Bank and Spisula were observed in very high numbers in plaice stomachs (E. Ursin, personal communication; Kröncke, 2011). Furthermore, plaice also feed on sandeels (Kaiser et al., 2004), which were also most numerous on the bank itself. Solenette (B. luteum) abundance was strongly correlated with shallower locations and with the abundance of the amphipod B. elegans, but less so with other infauna species such as the brittle star A. brachiata and the tube-building polychaete M. johnstoni. While the latter two are hardly preyed upon by solenette in the German Bight as shown by Schückel et al. (2011), the authors showed that amphipods belong to the preferred food. We therefore suggest that solenette on the Dogger Bank predominantly feeds on the abundant amphipods B. elegans, as well as on harpacticoid copepods. The latter form a typical component of the meiofauna in sand habitats such as the shallow Dogger Bank (Huys et al., 1992), but due to their small size (b 1 mm), harpacticoids are not included in the infauna fraction sampled for the present study. Based on the work of Schückel et al. (2011), harpacticoids can also be expected to contribute to the diet of solenette in the shallow sections of the bank with 100% sandy sediments. Dab (L. limanda) which was present at all stations sampled, is known to feed opportunistically, and adult dab to be quite independent of the composition of benthic prey species (Hinz et al., 2005; Kaiser and Ramsay, 1997). In contrast to the larger specimens, juvenile dab and plaice both feed preferentially on harpacticoid copepods and amphipods (Schückel et al., 2012). While sandeel species were also found in highest abundances on the bank itself, predator–prey interactions with the epifauna species within the “bank cluster” are not expected here, because sandeel feed in the water column. Ammodytes spp. prey on zooplankton and on some large diatoms, and H. lanceolatus also on small fish like clupeids and smaller ammodytids (Whitehead et al., 1984). Grey gurnard (E. gurnardus) is one of the most abundant ground fish species on the Dogger Bank. E. gurnardus consumes a multitude of prey taxa, including vertebrates as well as invertebrates (de Gee and Kikkert, 1993; Floeter et al., 2005). In parallel with the present study, Weinert et al. (2010) investigated feeding behavior and condition of grey gurnards at a subset of 17 stations. They found regional differences in stomach contents across and around the Dogger Bank, reflecting in part the spatial organization of fish and invertebrate communities described here. Identifiable fish prey consisted mainly of sandeel (Ammodytidae, including A. marinus and H. immaculatus). Single specimens of the following species were also found: sprat (S. sprattus), scaldfish (A. laterna), solenette (B. luteum), spotted dragonet (C. maculatus), and Nilsson's pipefish (Syngnathus rostellatus). In the shallower northeast of the Bank (mainly stations 422, 426 and 428), sandeel (Ammodytidae) were the dominant prey of grey gurnards in the size class 25–30 cm body length. In contrast, at the shallowest area on top of the bank, Crangon allmanni and unidentifiable Crangonidae comprised the main portion of the prey. Furthermore, the amphipod Hyperia galba was identified as the most abundant prey in grey gurnard stomachs overall, especially at a cluster of mostly deep stations in the northeast and south of the bank. In contrast to the obvious importance of this species in the prey spectrum of grey gurnard, H. galba rarely occurred in the in- and epifauna samples obtained. This under-representation in the grab as well as in the 2-m beam trawl is presumably the consequence of a migration into the water column, where amphipods often form large aggregations (Dittrich, 1988). H. galba is known to live commensally in various jellyfish species. Thus, gurnards might either ingest them together with jellyfish in the water column, or feed specifically on the hyperbenthic, free-living amphipods, analogous to the prey selection for caprellids observed in dab (Hinz et al., 2005). Weinert et al.

(2010) demonstrated that condition indices of E. gurnardus were correlated with the prey type found in the stomachs. Overall, condition indices in gurnards were – with few exceptions – most elevated where also their abundance was highest, particularly at the shallowest region on top of the bank and furthermore at the north-eastern section of the bank (Tail End), where the abundance of sandeel was highest. Therefore, sandeel may be seen as part of the biotic habitat properties (prey) for grey gurnards on Dogger Bank. Overall, dab and gurnard were the most generally abundant fish species on Dogger Bank, and due to their ability to exploit a wide spectrum of abiotic as well as biotic habitats, are represented at a central position in the CCA plot (Fig. 6).

4.3. Limitations of the methods applied Catches spectra from the GOV and from the 2-m beam trawl overlap for a portion of the species and size classes caught with the two gears, while catches of the largest or the smallest fishes are restricted to one of the two gears (Ehrich et al., 2004). Therefore, the catches have to be considered as two aspects of the same parent community. The purpose of taking both is not to compare between the two, but to cover the demersal fish community on Dogger Bank as completely as possible. Some demersal fish species are migratory and may hence not at all times be present at the Dogger Bank at the same relative abundance as more resident species. This aspect is not considered in this study. One particular species, the snake pipefish E. aequoreus, was present in the North Sea in high numbers during a few years – including 2006 – when it was transported through Atlantic inflow, but has since vanished again. The observed significant correlations between the occurrence of some fish species and the composition of benthic invertebrates have been interpreted as a sign of trophic linkages, especially where other studies have shown that particular species serve as suitable prey for the respective fish species. It cannot be precluded, that additional effects originate from the provision of habitat complexity through the physical presence of benthic epifauna. However, we expect that this kind of effect will be of minor importance on a sandbank, as opposed to a rocky sea bed. All sampling devices have limitations in their catch efficiency for individual species. In this case, the van Veen grab for infauna and the 2-m beam trawl for epifauna have been applied as we considered them as the best suited gears in many aspects, of which a decisive one is the application of these gears as standard tools in many studies in the North Sea, which makes the results of this study comparable to others and to existing long-term data sets. Nevertheless, limitations of these methods exist: Neither of the two gears appears to be efficient in catching amphipods like H. galba (this study) or caprellids (Hinz et al., 2005). Thus, these taxa would be underrepresented in the samples, in any community analyses based upon them or in related interpretations on prey availability for demersal fish. Sampling would ideally be done with multiple replicated per station. For fishing hauls, this not practicable because each haul covers a large area and a “replicate” haul would necessarily cover at least partly different benthic habitats. Limiting the replication of catches makes it more likely that outliers can influence the results. For sampling of infauna with a sediment grab, the limitation is less due to the ability to resample on a small spatial scale, but due to the time-consuming analyses of the infauna samples. We therefore – as generally common – limited the van Veen samples to two replicates per station. Future studies could use the results of this analysis to identify areas on the bank within which the composition of benthic and fish communities can be expected to be very similar, versus other areas that can be expected to be significantly different. A comparative study or monitoring building upon this work could hence concentrate their sampling

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

on a few stations for each different habitat or community type, and hence free resources for extra replication of samples. 4.4. Relevance of links between fish and habitat for marine policy The Dogger Bank is an area of high commercial and political interests for various reasons, including fishing opportunities, utilization of oil reserves, and the potential to install wind farms (Pedersen et al., 2009). Simultaneously, several EU member states are, at the time of writing, identifying conservation measures to be implemented under the Habitats Directive for their Natura 2000 sites (Fock, 2011). Marine spatial planning aims to balance multiple interests of potential and actual users of the seas (Foley et al., 2010; Katsanevakis et al., 2011; Stelzenmüller et al., 2010). This requires reasoning for the selection of sites for activities, or even a formal environmental impact assessment of potential effects on marine habitats and associated communities (European Commission, 2003). Planning of further human activities on the Dogger Bank could take into account the links between abiotic habitat parameters as well as benthic infauna or epifauna and demersal fish in order to predict possible impacts on the community or ecosystem level. Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.seares.2013.01.007. Acknowledgments We thank the ship's crew and captain H.-O. Janßen of the Walther Herwig III for their continuous support. The science crew aboard WH 287 is acknowledged for their dedication, and so are J. Appel and I. Wilhelms for data handling, and also S. and U. Schückel and S. Preuß for support with sample analysis of the in- and epifauna on shore. S. Ehrich provided helpful suggestions for improvement of the manuscript. The research presented here has received funding from the European Community’s Seventh Framework Programme (FP7/2007-2013) under Grant Agreement No. 266445 for the project Vectors of Change in Oceans and Seas Marine Life, Impact on Economic Sectors (VECTORS). References Brockmann, U., Wegner, G., 1985. Hydrography, nutrient and chlorophyll distribution in the North Sea in February 1984. Archiv für Fischereiwissenschaft 36, 27–45. Callaway, R., Alvsvåg, J., de Boois, I., Cotter, J., Ford, A., Hinz, H., Jennings, S., Kröncke, I., Lancaster, J., Piet, G.J., Prince, P., Ehrich, S., 2002. Diversity and community structure of epibenthic invertebrates and fish in the North Sea. ICES Journal of Marine Science 59, 1199–1214. Clarke, K.R., Gorley, R.N., 2006. PRIMER V.6: User Manual/Tutorial. PRIMER-E, Plymouth. Dauwe, B., Herman, P., Heip, C., 1998. Community structure and bioturbation potential of macrofauna at four North Sea stations with contrasting food supply. Marine Ecology Progress Series 173, 67–83. de Gee, A., Kikkert, A.H., 1993. Analysis of the grey gurnard (Eutrigla gurnardus) samples collected during the 1991 international stomach sampling project. International Council for the Exploration of the Sea, ICES CM 1993/G14 (25 pp. 1993/G14, 25 pp.). Dittrich, B., 1988. Studies on the life cycle and reproduction of the parasitic amphipod Hyperia galba in the North Sea. Helgoland Marine Research 42, 79–98. Ehrich, S., Reiss, H., Damm, U., Kröncke, I., 2004. Vulnerability of Bottom Fish Species to the Standard GOV. ICES CM 2004/D:05 . Ellis, J.R., Maxwell, T., Schratzberger, M., Rogers, S.I., 2011. The benthos and fish of offshore sandbank habitats in the southern North Sea. Journal of the Marine Biological Association of the United Kingdom 91, 1319–1335. Engelhard, G.H., van der Kooij, J., Bell, E.D., Pinnegar, J.K., Blanchard, J.L., Mackinson, S., Righton, D.A., 2008. Fishing mortality versus natural predation on diurnally migrating sandeels Ammodytes marinus. Marine Ecology Progress Series 369, 213–227. European Commission, 1992. Council Directive 92/43/EEC on the conservation of natural habitats and of wild fauna and flora. European Community Gazette, 206, pp. 1–50. European Commission, 2003. Report from the Commission to the European Parliament and the Council on the Application and Effectiveness of the EIA Directive (Directive 85/337/EEC as amended by Directive 97/11/EC) — How Successful are the Member States in Implementing the EIA Directive, p. 334 (COM 2003). Floeter, J., Kempf, A., Vinther, M., Schrum, C., Temming, A., 2005. Grey gurnard (Eutrigla gurnardus) in the North Sea: an emerging key predator? Canadian Journal of Fisheries and Aquatic Sciences 62, 1853–1864. Fock, H.O., 2011. Natura 2000 and the European Common Fisheries Policy. Marine Policy 35, 181–188.

23

Fock, H., Uiblein, F., Köster, F., von Westernhagen, H., 2002. Biodiversity and species– environment relationships of the demersal fish assemblage at the Great Meteor Seamount (subtropical NE Atlantic), sampled by different trawls. Marine Biology 141, 185–199. Foley, M.M., Halpern, B.S., Micheli, F., Armsby, M.H., Caldwell, M.R., Crain, C.M., Prahler, E., Rohr, N., Sivas, D., Beck, M.W., 2010. Guiding ecological principles for marine spatial planning. Marine Policy 34, 955–966. Harris, M.P., Beare, D., Toresen, R., Nøttestad, L., Kloppmann, M., Dörner, H., Peach, K., Rushton, D.R.A., Foster-Smith, J., Wanless, S., 2007. A major increase in snake pipefish (Entelurus aequoreus) in northern European seas since 2003: potential implications for seabird breeding success. Marine Biology 151, 973–983. Hill, A.E., James, I.D., Linden, P.F., Matthews, J.P., Prandle, D., Simpson, J.H., Gmitrowicz, E.M., Smeed, D.A., Lwiza, K.M.M., Durazo, R., Fox, A.D., Bowers, D.G., 1994. Dynamics of tidal mixing fronts in the North Sea. In: Charnock, H., Dyer, K.R., Huthnance, J.M., Liss, P.S., Simpson, J.H., Tett, P.B. (Eds.), Understanding the North Sea System. The Royal Society, Chapman and Hall, London, pp. 53–68. Hinz, H., Kröncke, I., Ehrich, S., 2004. Seasonal and annual mesoscale variability of an epifaunal community in the German Bight. Marine Biology 144, 735–745. Hinz, H., Kröncke, I., Ehrich, S., 2005. The feeding strategy of dab Limanda limanda in the southern North Sea: linking stomach contents to prey availability in the environment. Journal of Fish Biology 67 (Suppl. B), 125–145. Howarth, M.J., Dyer, K.R., Joint, I.R., Hydes, D.J., Purdie, D.A., Edmunds, H., Jones, J.E., Lowry, R.K., Moffat, T.J., Pomroy, A.J., Proctor, R., 1994. Seasonal cycles and their spatial variability. In: Charnock, H., Dyer, K.R., Huthnance, J.M., Liss, P.S., Simpson, J.H., Tett, P.B. (Eds.), Understanding the North Sea System. The Royal Society, Chapman and Hall, London, pp. 5–26. Huys, R., Herman, P.M.J., Heip, C.H.R., Soetaert, K., 1992. The meiobenthos of the North Sea: density, biomass trends and distribution of copepod communities. ICES Journal of Marine Science 49, 23–44. ICES, 2006. Manual for the international bottom trawl surveys — revision VII. Report of the International Bottom Trawl Survey Working Group (IBTSWG) Annex 1. (53 pp.). Jennings, S., Lancaster, J., Woolmer, A., Cotter, J., 1999. Distribution, diversity and abundance of epibenthic fauna in the North Sea. Journal of the Marine Biological Association of the United Kingdom 79, 385–399. Kaiser, M.J., Ramsay, K., 1997. Opportunistic feeding by dabs within areas of trawl disturbance: possible implications for increased survival. Marine Ecology Progress Series 152, 307–310. Kaiser, M.J., Spence, F.B., Hart, P.J.B., 2000. Fishing-gear restrictions and conservation of benthic habitat complexity. Conservation Biology 14, 1512–1525. Kaiser, M.J., Bergmann, M., Hinz, H., Galanidi, M., Shucksmith, R., Rees, E.I.S., Darbyshire, T., Ramsay, K., 2004. Demersal fish and epifauna associated with sandbank habitats. Estuarine, Coastal and Shelf Science 60, 445–456. Katsanevakis, S., Stelzenmüller, V., South, A., Sørensen, T.K., Jones, P.J.S., Kerr, S., Badalamenti, F., Anagnostou, C., Breen, P., Chust, G., D'anna, G., Duijn, M., Filatova, T., Fiorentino, F., Hulsman, H., Karageorgis, A., Kröncke, I., Mirto, S., Pipitone, C., Portelli, S., Qiu, W., Reiss, H., Sakellariou, D., Salomidi, M., Van Hoof, L., Vassilopoulou, V., Vega, T., Vöge, S., Weber, A., Zenetos, A., ter Hofstede, R., 2011. Ecosystem-based marine spatial management: review of concepts, policies, tools, and critical issues. Ocean and Coastal Management 54, 807–820. Kloppmann, M.H.F., Ulleweit, J., 2007. Off-shelf distribution of pelagic snake pipefish, Entelurus aequoreus (Linnaeus, 1758), west of the British Isles. Marine Biology 151, 271–275. Kröncke, I., 2011. Changes in Dogger Bank macrofauna communities in the 20th century caused by fishing and climate. Estuarine, Coastal and Shelf Science 94, 234–245. Kröncke, I., Knust, R., 1995. The Dogger Bank: a special ecological region in the central North Sea. Helgoländer Meeresuntersuchungen 49, 335–353. Kröncke, I., Stoeck, T., Wieking, G., Palojärvi, A., 2004. Relationship between structural and functional aspects of microbial and macrofaunal communities in different areas of the North Sea. Marine Ecology Progress Series 282, 13–31. Kröncke, I., Reiss, H., Eggleton, J.D., Aldridge, J., Bergman, M.J.N., Cochrane, S., Craeymeersch, J., Degraer, S., Desroy, N., Dewarumez, J.-M., Duineveld, G., Essink, K., Hillewaert, H., Lavaleye, M.S.S., Moll, A., Nehring, S., Newell, J., Oug, E., Pohlmann, T., Rachor, E., Robertson, M.R., Rumohr, H., Schratzberger, M., Smith, R., Vanden Berghe, E., van Dalfsen, J., van Hoey, G., Vincx, M., Willems, W., Rees, H.L., 2011. Changes in North Sea macrofauna communities and species distribution between 1986 and 2000. Estuarine, Coastal and Shelf Science 94, 1–15. Künitzer, A., Basford, D., Craeymeersch, J.A., Dewarumez, J.M., Dörjes, J., Duineveld, G.C.A., Eleftheriou, A., Heip, C., Herman, P., Kingston, P., Niermann, U., Rachor, E., Rumohr, H., de Wilde, P.A.J., 1992. The benthic infauna of the North Sea: species distribution and assemblages. ICES Journal of Marine Science 49, 127–143. Løkkeberg, S., Humborstad, O.-B., Jørgensen, T., Soldal, A.V., 2002. Spatio-temporal variations in gillnet catch rates in the vicinity of North Sea oil platforms. ICES Journal of Marine Science 59, 294–299 (Suppl.). Mattson, S., 1992. Food and feeding habits of fish species over a soft sublittoral bottom in the northeast Atlantic. 3. Haddock (Melanogrammus aeglefinus (L.)) (Gadidae). Sarsia 77, 33–45. Neumann, H., Ehrich, S., Kröncke, I., 2008. Spatial variability of epifaunal communities in the North Sea in relation to sampling effort. Hegoland Marine Research 62, 215–225. Neumann, H., Reiss, H., Rakers, S., Ehrich, S., Kröncke, I., 2009. Temporal variability of southern North Sea epifauna communities after the cold winter 1995/1996. ICES Journal of Marine Science 66, 2233–2243. Otto, L., Zimmermann, I.T.F., Furnes, G.K., Mork, M., Saetre, R., Becker, G., 1990. Review of the physical oceanography of the North Sea. Netherlands Journal of Sea Research 25, 161–238. Parrish, F.A., Boland, R.C., 2004. Habitat and reef-fish assemblages of banks in the Northwestern Hawaiian Islands. Marine Biology 144, 1065–1073.

24

A.F. Sell, I. Kröncke / Journal of Sea Research 80 (2013) 12–24

Pedersen, S.A., Fock, H.O., Sell, A.F., 2009. Mapping fisheries in the German Exclusive Economic Zone with special reference to offshore Natura 2000 sites. Marine Policy 33, 571–590. Perry, A.L., Low, P.J., Ellis, J.R., Reynolds, J.D., 2005. Climate change and distribution shifts in marine fishes. Science (New York, N.Y.) 308, 1912–1915. Rees, H.L., Pendle, M.A., Waldock, R., Limpenny, D.S., Boyd, S.E., 1999. A comparison of benthic biodiversity in the North Sea, English Channel, and Celtic Seas. ICES Journal of Marine Science 56, 228–246. Rees, H.L., Pendle, M.A., Limpenny, D.S., Mason, C.E., Boyd, S.E., Birchenough, S., Vivian, C.M.G., 2006. Benthic responses to organic enrichment and climatic events in the western North Sea. Journal of the Marine Biological Association of the United Kingdom 86, 1–18. Reiss, H., Kröncke, I., 2005. Seasonal variability of infaunal community structures in three areas of the North Sea under different environmental conditions. Estuarine, Coastal and Shelf Science 65, 253–274. Reiss, H., Degraer, S., Duineveld, G.C.A., Kröncke, I., Aldridge, J., Craeymeersch, J.A., Eggleton, J.D., Hillewaert, H., Lavaleye, M.S.S., Moll, A., Pohlmann, T., Rachor, E., Robertson, M., Vanden Berghe, E., van Hoey, G., Rees, H.L., 2010. Spatial patterns of infauna, epifauna, and demersal fish communities in the North Sea. ICES Journal of Marine Science 67, 278–293. Richardson, K., Olsen, O.V., 1987. Winter nutrient concentrations and primary production in the eastern North Sea. International Council for the Exploration of the Sea, ICES CM 1993/G14. (25 pp. 1987/C 23, 1–14). Rijnsdorp, A.D., Vingerhoed, B., 2001. Feeding of plaice Pleuronectes platessa (L.) and sole Solea solea (L.) in relation to the effects of bottom trawling. Journal of Sea Research 45, 219–229. Ryer, C.H., Stoner, A.W., Titgen, R.H., 2004. Behavioral mechanisms underlying the refuge value of benthic habitat structure for two flatfishes with differing anti-predator strategies. Marine Ecology Progress Series 268, 231–243. Salzwedel, H., Rachor, E., Gerdes, D., 1985. Benthic macrofauna communities in the German Bight. Verifflithungen des Institut fur Meeresforschung in Bremerhaven 20, 199–267. Schückel, S., Ehrich, S., Kröncke, I., Reiss, H., 2010. Linking prey composition of haddock Melanogrammus aeglefinus to benthic prey availability in three different areas of the North Sea. Journal of Fish Biology 77, 98–118. Schückel, S., Sell, A., Kröncke, I., Reiss, H., 2011. Diet composition and resource partitioning in two small flatfish species in the German Bight. Journal of Sea Research 66, 195–204.

Schückel, S., Sell, A.F., Kröncke, I., Reiss, H., 2012. Diet overlap between flatfish species in the southern North Sea. Journal of Fish Biology 80, 2571–2594. Sell, A.F., Kröncke, I., Ehrich, S., 2007. Linking the Diversity of Fish Assemblages to Habitat Structure: A study on Dogger Bank (North Sea). ICES CM 2007/E:18. Soldal, A.V., Svellingen, I., Jørgensen, T., Løkkeberg, S., 2002. Rigs-to-reefs in the North Sea: hydroacoustic quantification of fish in the vicinity of a “semi-cold” platform. ICES Journal of Marine Science 59, 281–287 (Suppl.). Sonnewald, M., Türkay, M., 2010. The megaepifauna of the Dogger Bank (North Sea): species composition and faunal characteristics 1991–2008. Helgoland Marine Research 1–13. Sonnewald, M., Türkay, M., 2012. Environmental influence on the bottom and nearbottom megafauna communities of the Dogger Bank: a long-term survey. Helgoland Marine Research 66, 503–511. Stelzenmüller, V., Lee, J., South, A., Rogers, S., 2010. Quantifying cumulative impacts of human pressures on the marine environment: a geospatial modelling framework. Marine Ecology Progress Series 398, 19–32. ter Braak, C.J.F., Smilauer, P., 1998. CANOCO for Windows. Centre for Biometry. van Moorsel, G.W.N.M., 2011. Species and Habitats of the International Dogger Bank. Ecosub, Doorn 74. Vasconcelos, R., Prista, N., Cabral, H., Costa, M.J., 2004. Feeding ecology of the lesser weever, Echiichthys vipera (Cuvier, 1829), on the western coast of Portugal. Journal of Applied Ichthyology 20, 211–216. Weinert, M., Floeter, J., Kröncke, I., Sell, A.F., 2010. The role of prey composition for the condition of grey gurnard (Eutrigla gurnardus). Journal of Applied Ichthyology 26, 75–84. Whitehead, P.J.P., Bauchot, M.-L., Hureau, J.-C., Nielsen, J., Tortonese, E., 1984. Fishes of the North-eastern Atlantic and the Mediterranean. UNESCO, Paris, p. 1473. Wieking, G., Kröncke, I., 2001. Decadal changes in macrofaunal communities on the Dogger Bank caused by large-scale climate variability. Senckenbergiana Maritima 31, 125–141. Wieking, G., Kröncke, I., 2003. Macrofauna communities of the Dogger Bank (central North Sea) in the late 1990s: spatial distribution, species composition and trophic structure. Helgoland Marine Research 57, 34–46. Wieking, G., Kröncke, I., 2005. Is benthic trophic structure affected by food quality? The Dogger Bank example. Marine Biology 146, 387–400. Zühlke, R., Alvsvåg, J., de Boois, I., Cotter, J., Ehrich, S., Ford, A., Hinz, H., JarreTeichmann, A., Jennings, S., Kröncke, I., Lancaster, J., Piet, G.J., Prince, P., 2001. Epibenthic diversity in the North Sea. Senckenbergiana Maritima 31, 269–281.