Abstract Catches of European hake (Merluccius merluccius) in the North Sea have increased substantially during the last decade, even though there is no directed commercial fishery of hake in this area. We analysed the spatial distributions of hake in the northern the parts of its range, (where it is less well-studied), using ICES international bottom trawl survey data from 1997 to 2015. We examine length-frequency distributions for (i) distinct modes enabling the assignment of fish into categories which likely corresponded to the ages 1, 2, and 3+ and (ii) patterns of seasonal spatial distribution for the different groups. Age categories 1 and 2 fish were most abundant in the northern North Sea, and appear to remain in the North Sea until 2 years of age, when they move into deeper waters. Their distribution has expanded into the western-central North Sea in the last decade. Age category 3+ fish were most abundant in the northern and central North Sea during summer, indicating a seasonal influx of large individuals into this area likely associated with spawning activity. The distribution of these older fish has gradually expanded westward in both seasons. Introduction The European hake (Merluccius merluccius; hereafter only referred to as hake) is a demersal fish commercially exploited by many European coastal states (ICES, 2016a). Its distribution extends from the north-west African coast northwards to the Bay of Biscay, further north into Irish, Scottish, and Norwegian Sea waters, and covers most of the North Sea, Skagerrak, and Kattegat (Murua, 2010). Hake has frequently been caught as by-catch by various fisheries in the North Sea and the Skagerrak since the beginning of the 20th century (Hickling, 1927; ICES, 2017). Over the last decade, the species has become more prevalent, particularly in bottom trawl catches from the North Sea and waters west of Scotland (Baudron and Fernandes, 2015; Werner et al., 2016), with total landings from the North Sea having increased from 2718 t in 2006 to a record high 12 091 t in 2014 (ICES, 2016a). Earlier studies from the North Sea and the adjacent Norwegian Trench based on both commercial and scientific catches showed that hake vary seasonally in abundance, length composition and depth distribution (Hickling, 1927; Sahrhage, 1964, 1967; Bergstad, 1991; Baudron and Fernandes, 2015; Werner, 2015). Such temporal and spatial variations have also been observed in scientific catches of other Merluccius species (Burmeister, 2001; Orsi Relini et al., 2002; Wilhelm et al., 2015). Spatial and temporal variations in abundance of hake in the North Sea are hypothesized to be a result of the migration/influx of mature and spawning fish (Hickling, 1927; Bergstad, 1991). Recent analysis of gonad maturity data strongly supports these earlier observations by showing that migrations of mature/spawning hake into shallower parts of the northern and central North Sea occurs during late spring and summer months (Werner et al., 2016). Hake is a batch spawner, producing multiple batches within a spawning season (Murua, 2010). Spawning occurs year-round in the southern part of its range, with peak spawning early in the first half of the year. Around Scotland and in the North Sea, peak spawning occurs between May and October as sea surface temperature increase to 12–15°C (Alvarez et al., 2004; Werner et al., 2016). The main spawning area east of Shetland is located between 57° and 61°N, between the 80 and 120 m isobath on the shelf west of the Norwegian trench (Werner et al., 2016). Length at maturity (L50) increases with increasing latitude, and for fish in the North Sea is estimated at 37–45 cm for males and 50–59 cm for females (Werner et al., 2016). In the Bay of Biscay and Celtic Sea hake reach an estimated length of 23.8 cm at the end of the first year (Kacher and Amara, 2005). Growth estimates of recaptured tagged fish from the Bay of Biscay indicate that mean length at the end of the first and second year is ∼19 and 35 cm, respectively (de Pontual et al., 2013). Given the estimated L50 ranges of 36.8–45.4 cm for males and 50.5–59.2 cm for females (Werner et al., 2016), males mature when 2–3-years old and females when 3–4 years old. Despite a recent analysis of spatial and temporal changes in biomass of hake in the North Sea (Baudron and Fernandes, 2015), the temporal and spatial dynamics of different length/age groups remain unclear. We therefore analysed seasonal length-frequency distributions from IBTS catch data for 1997–2015 and calculated abundance and biomass estimates by putative age groups. Our objective was to examine the annual and seasonal changes in hake demography and size/age dependent temporal changes in the spatial distribution in the central and northern North Sea. Methods Sampling gear and method Scientific research data were downloaded from the ICES dataportal DATRAS (ICES, 2013) for the biannnual International Bottom Trawl Survey (IBTS) during the period 1997–2015, which covers the North Sea (ICES areas IVa, IVb, IVc, and VIId) and Skagerrak (ICES area IIIa) in late winter (quarter 1: January—March) and in summer (quarter 3: June—September). The surveyed area is divided into statistical rectangles measuring 1-degree latitude by 0.5-degree longitude (Figure 1), and ideally the fishing gear samples the demersal community 4, 5 m above the bottom from inshore waters to 200-m depth (Figure 1a). The gear deployed is a Grande Overture Verticale (GOV) trawl with a 20-mm mesh codend lining towed at 3.5–4.5 knots for 30 min (ICES, 2015). Although the IBTS protocol aims for at least two tows in each statistical rectangle, in some instances only one valid trawl was completed. We focussed only on the North Sea (from −2° W to 8°E, and 53.5°N to 62°N, i.e. ICES areas IVa, IVb, IVc, and VIId), given that recent genetic results indicate that there may be different populations in the North Sea and Skagerrak (Westgaard et al., 2017). Further, preliminary distribution maps of the data indicated that with the exception of one sampling station in one year, hake are not caught in the southernmost areas of the North Sea, i.e. ICES areas VIIc and VIId. Stations from these areas were therefore omitted from all analyses. The data products analysed include length measurements per haul (DATRAS file format CPUE (number*h−1) per length per haul), as well as station and catch information (DATRAS file format Haul-meta data which is part of the Exchange data). Missing trawl speed values—15, 7% of all stations with hake catch did not have trawl speeds—were replaced with the overall average trawl speed of 3.8 knots. Figure 1. View largeDownload slide Distribution of all IBTS stations in winter (January–March) and summer (June–September) in the period 1997–2015 (left panel), and stations with European hake in trawl catches in winter (yellow) and summer (orange) in the period 1997–2015 (right panel). The rectangles represent the ICES statistical rectangles used as the nesting units for calculating total abundance/biomass. ICES Subareas IVa (northern North Sea) and IVb (central North Sea), the Norwegian Trench and bottom depths are indicated. Figure 1. View largeDownload slide Distribution of all IBTS stations in winter (January–March) and summer (June–September) in the period 1997–2015 (left panel), and stations with European hake in trawl catches in winter (yellow) and summer (orange) in the period 1997–2015 (right panel). The rectangles represent the ICES statistical rectangles used as the nesting units for calculating total abundance/biomass. ICES Subareas IVa (northern North Sea) and IVb (central North Sea), the Norwegian Trench and bottom depths are indicated. Catch sampling Catch sampling was done according to the IBTS manual. The entire catch was fully sorted, and total length of all fish measured to 1 cm below, i.e. a fish measured to be 35.5-cm long would be given the length 35 cm (ICES, 2015). Swept area estimation Trawl sweeps seem to cause little or no herding behaviour in other hake species (Huse et al., 2001), implying that the trawl width (wing spread) defines the area swept by the trawl gear and therefore only fish between the wings are captured in the cod end. We used the formula presented in the final report for “Managing Fisheries to Conserve Groundfish and Benthic Invertebrate Species Diversity” (MAFCONS; Greenstreet, 2007), which describes the wing spread (WS; m) as a function of fishing depth D (equivalent to bottom depth for bottom trawls); WS=6.8515* log D+5.8931 (1) For a given station (i) the swept area (SAi; km2) could then be calculated as; SAi=Spi*Duri*WSi (2) where Sp equals vessel speed and Dur the duration of the trawl. Data analyses All statistical analyses were done with the R version 3.3.2 (R Core Team, 2016). Estimates of total abundance/biomass development and proportion of stations with catches Only daytime stations within 200-m bottom depth and with a trawl duration between 15 and 45 min were used. To estimate the abundance of fish we used the ICES rectangles (Figure 1) as our nesting unit, i.e. catches at a given trawl station were assumed to be representative of the ICES rectangle to which the trawl station belonged. For a given year, catches (TC; tonnes) were summed as: TC=∑y=1n((∑i=1nCatchiy∑i=1nSAiy)×Areay) (3) where Catchiy represents the catch at station i in ICES rectangle y; SAiy represents the corresponding swept area; and Areayrepresent the total area of the rectangle in question. Quarterly catches were calculated separately. To calculate biomass development, the same general approach (Equation 3) was used, but length measurements (L) were converted to whole round weight (W; gr). This conversion used a weight–length relationship calculated from individual weight and length measurements collected on surveys conducted by the Institute of Marine Research between 2010 and 2014 in the North Sea and Skagerrak; W=0.0043*L3.114 (4) Since the number of stations sampled was very similar in most years, we also examine whether hake became more available to the sampling gear by calculating the fraction of trawl stations producing positive catches through our time series. Changes in this fraction would provide an indication of the species expanding or contracting its range. Size distribution of hake in winter and summer We focussed our analyses on patterns indicative of seasonal migrations by comparing size distributions of fish caught in winter to those from summer, and on modes indicative of distinct length cohorts to identify possible age-dependent trends. If migration of young fish in and out of an area is negligible and spawning occurs during a relatively fixed period every year, different cohorts will have both different size ranges and average sizes resulting in identifiable/distinct modes (Pauly and Morgan, 1987). Hence, if modes are present then size can be used to assign fish to different age categories roughly corresponding to their actual age. Due to the low number of length samples, especially in the earlier years (1997–2003), an average distribution was calculated for both winter and summer, based on length data from all years. In doing this, the underlying assumption is that growth is constant between years. To avoid any confounding effects of cohort abundance on the observed pattern in size distribution, we used the average yearly proportion of fish in each 1-cm length class as the basis for both our assignment of fish into size-based aged categories and the spatial comparison of winter and summer distributions. First the combined average distributions for winter and summer were plotted to visualize peaks and identify length modes in the summer and winter distributions. Six peaks were evident in both the summer and winter size distributions, although patterns were less clear for larger sizes. The visualized peaks were confirmed/tested with a mixed distribution analysis, using the R-package mixDist (Macdonald, 2012). This package allows for the computation of gaussian distributions underlying mixed distribution curves and produces estimates of pi, mu, and sigma of the underlying curves. These summary outputs were used to calculate the proportion of fish in each 1-cm length class, belonging to each of the identified size-based age categories (Table 1). Table 1. Results of estimated Mixdist parameters for winter and summer. Season Cohort pi mu sigma Winter 1 0, 15282 112, 6 25, 24 2 0, 54874 256 36, 83 3 0, 11411 355, 7 30, 13 4 0, 13039 442 43, 45 5 0, 03162 561 31, 14 6 0, 02233 650, 1 87, 93 Summer 1 0, 240875 157, 8 33 2 0, 323835 329, 8 53, 61 3 0, 183349 471, 2 50, 87 4 0, 147579 588, 2 61, 6 5 0, 094533 688, 9 100, 94 6 0, 009829 770, 3 181, 72 Season Cohort pi mu sigma Winter 1 0, 15282 112, 6 25, 24 2 0, 54874 256 36, 83 3 0, 11411 355, 7 30, 13 4 0, 13039 442 43, 45 5 0, 03162 561 31, 14 6 0, 02233 650, 1 87, 93 Summer 1 0, 240875 157, 8 33 2 0, 323835 329, 8 53, 61 3 0, 183349 471, 2 50, 87 4 0, 147579 588, 2 61, 6 5 0, 094533 688, 9 100, 94 6 0, 009829 770, 3 181, 72 Pi, estimated proportion; mu, estimated average; and sigma, estimated SD). Table 1. Results of estimated Mixdist parameters for winter and summer. Season Cohort pi mu sigma Winter 1 0, 15282 112, 6 25, 24 2 0, 54874 256 36, 83 3 0, 11411 355, 7 30, 13 4 0, 13039 442 43, 45 5 0, 03162 561 31, 14 6 0, 02233 650, 1 87, 93 Summer 1 0, 240875 157, 8 33 2 0, 323835 329, 8 53, 61 3 0, 183349 471, 2 50, 87 4 0, 147579 588, 2 61, 6 5 0, 094533 688, 9 100, 94 6 0, 009829 770, 3 181, 72 Season Cohort pi mu sigma Winter 1 0, 15282 112, 6 25, 24 2 0, 54874 256 36, 83 3 0, 11411 355, 7 30, 13 4 0, 13039 442 43, 45 5 0, 03162 561 31, 14 6 0, 02233 650, 1 87, 93 Summer 1 0, 240875 157, 8 33 2 0, 323835 329, 8 53, 61 3 0, 183349 471, 2 50, 87 4 0, 147579 588, 2 61, 6 5 0, 094533 688, 9 100, 94 6 0, 009829 770, 3 181, 72 Pi, estimated proportion; mu, estimated average; and sigma, estimated SD). Spatial distribution of hake To estimate the area covered annually and seasonally by different age categories, we summed the areas of those ICES rectangles which contained hake catches of the respective age categories resulting from the assignment described earlier. The estimated area by age category included only length classes where the contribution of the respective age category was 90% or higher. Areas not covered in 2 consecutive years, or not sampled in at least 3 years, were excluded to ensure consistency in coverage. The average swept area abundance by age category was calculated for each ICES statistical rectangle, year and season. The size of each ICES rectangle was corrected for landcover and includes only the area within the 200-m isobath. Results Hake population trends Estimated abundance and biomass of hake has increased since 2005 (Figure 2). The trend in abundance is apparent in both summer and winter seasons, with lower abundances estimated in winter than in summer. Biomass has increased more rapidly in summer than in winter (Figure 2). Hake catches appeared stable prior to 2005, and the subsequent increase in catches has been accompanied by increases in both the proportion of stations with catches and the area occupied by hake (Figure 3). Though the trend was similar in the two seasons, both the proportion of stations with hake and the area occupied were lower in winter (Figure 3). Figure 2. View largeDownload slide (a) Estimated abundance and (b) biomass of European hake in the North Sea during the study period 1997–2015. Figure 2. View largeDownload slide (a) Estimated abundance and (b) biomass of European hake in the North Sea during the study period 1997–2015. Figure 3. View largeDownload slide (a) Proportion of stations with and area occupied by European hake during the study period in winter and (b) summer. The lines are loess smoothers fitted by using the geom_smooth function in the ggplot2 library (Wickham, 2009) in R (R Core Team, 2016). Area occupied is the sum of ICES rectangle areas where European hake was present in trawl catches. Figure 3. View largeDownload slide (a) Proportion of stations with and area occupied by European hake during the study period in winter and (b) summer. The lines are loess smoothers fitted by using the geom_smooth function in the ggplot2 library (Wickham, 2009) in R (R Core Team, 2016). Area occupied is the sum of ICES rectangle areas where European hake was present in trawl catches. Size distribution of hake The average length distributions for winter and summer, with cohort curves fitted, are shown in Figure 4. In both winter (Figure 4a) and summer (Figure 4b), two clear cohorts with distinct peaks (means) are visible in the left side of the distribution, and in both distributions peaks become less apparent with an increase in length (Figure 4a and b). The first two peaks were assigned to age categories 1 and 2, and all larger fish were assigned to a single category 3+. In winter, fish < 21 cm (mode at 11 cm) were assigned to age category 1, fish 14–36 cm (mode at 26 cm) to age category 2, and fish >29 cm to age category 3+. In summer, fish < 26 cm (mode at 16 cm) were assigned to age category 1, fish 18–46 cm (mode at 33 cm) were assigned to age category 2, and fish > 36 cm were assigned to category 3+ (Figure 4). The size ranges and corresponding age categories obtained from the averaged distributions corresponded well to the the annual size distributions, matching the peaks and modes found in the annual distributions very well (Figure 5). Winter and summer distributions were different, except in 2015, with a much higher proportion of large hake (age category 3+) in summer (Figure 5b). The development of the length cohorts was particularly clear in 2012–2015, and the age categories can be tracked easily. Large numbers of age category 1 fish (recruits) were observed during 2012 in both winter and summer. These fish appear as age category 2 in winter and summer 2013, and in 2014 and 2015 have grown to age category 3+ (Figure 5). Figure 4. View largeDownload slide (a) The average yearly proportion of European hake caught in each 1-cm length class in the study period during winter and (b) summer. Fitted length cohort curves (age categories) were estimated using the R package Mixdist (Macdonald, 2012) in R Core Team (2016). The green line is the sum of the individual cohorts (red lines), and triangles indicate the mean length of the estimated cohort (age-categories). Figure 4. View largeDownload slide (a) The average yearly proportion of European hake caught in each 1-cm length class in the study period during winter and (b) summer. Fitted length cohort curves (age categories) were estimated using the R package Mixdist (Macdonald, 2012) in R Core Team (2016). The green line is the sum of the individual cohorts (red lines), and triangles indicate the mean length of the estimated cohort (age-categories). Figure 5. View largeDownload slide View largeDownload slide Annual length distributions in winter and summer. Length classes assigned age categories are shown as red lines for age category 1, green lines for age category 2, and blue lines for age category 3+. Figure 5. View largeDownload slide View largeDownload slide Annual length distributions in winter and summer. Length classes assigned age categories are shown as red lines for age category 1, green lines for age category 2, and blue lines for age category 3+. Spatial distribution of hake Changes in areal extent Changes in the spatial distributions of hake represent three distinct periods: (i) from 1997 to 2002 hake abundance was low and the total area occupied by the different age classes was relatively stable, (ii) during the period 2003–2006, hake distributions expanded rapidly, and (iii) from 2007 to 2015 abundance was high and there were only small inter-annual changes in the area occupied (Figure 3a and b). The average area occupied was 1–171% larger in summer compared with winter, across periods and age categories (Figure 6). Across all periods, there was a smaller seasonal difference in the area occupied by fish in age categories 1 and 2 compared with fish in age category 3+ (Figure 6). The average area populated in summer, particularly in the period 2007–2015, increased with age category, whereas the areal extent in winter showed less age-related variation between periods. Figure 6. View largeDownload slide Average area occupied by assigned age categories in summer and winter during three identified periods for northern North Sea (ICES subarea IVa) and central North Sea (ICES subarea IVb). Boxes represent the lower 25 and upper 75 percentiles. Figure 6. View largeDownload slide Average area occupied by assigned age categories in summer and winter during three identified periods for northern North Sea (ICES subarea IVa) and central North Sea (ICES subarea IVb). Boxes represent the lower 25 and upper 75 percentiles. Changes in spatial distribution Not only have hake occupied a wider area during the time period covered by these data, but there has also been a shift in the distribution of the fish within the North Sea. The spatial distributions have changed over time, and differ between seasons and age categories (Figure 7a–c). During both winter and summer in the first period (1997–2002), age categories 1 and 2 fish were largely limited to the northern North Sea (north of 57°N and west of 3°E). By the third period (2007–2015) the fish in these age categories were also distributed further south into the central North Sea (to 54°N) and east (to west of 4°E), as well as shallower) into waters with bottom depth of 50-100 m (Figure 7a and b). The distribution of larger fish (age category 3 fish) was predominantly limited to the northern North Sea and the western parts of the central North Sea where abundance increased during the latter part of our study period (Figure 7c). Figure 7. View largeDownload slide View largeDownload slide Distribution of European hake abundance of (a) age category 1, (b) age category 2, and (c) age category 3+ in summer and winter during three analysed periods. Depth contours are indicated by red lines (50 m), green lines (100 m), and black lines (200 m). Figure 7. View largeDownload slide View largeDownload slide Distribution of European hake abundance of (a) age category 1, (b) age category 2, and (c) age category 3+ in summer and winter during three analysed periods. Depth contours are indicated by red lines (50 m), green lines (100 m), and black lines (200 m). In summer, during the first period (1997–2003), the older fish were mainly located off the Danish west coast (in waters < 50 m), the entrance of the Skagerrak and the south-western shelf of the Norwegian trench. The summer distribution of the age 3+ category fish expanded south and west during the two more recent periods, covering large parts of the 50–100 and 100–200 m shelf and extending into deeper water as a result (Figure 7c). The seasonal change in depth distribution was not apparent in the age categories 1 and 2 fish. However, across the three periods of the data, the summer distribution of the younger fish has increasingly overlapped the age category 3+ fish. The spatial overlap was less pronounced during the winter season, particularly between 1997 and 2006 (Figure 7). Regardless of period, only age category 3+ fish were present in the (shallower) eastern central North Sea off Denmark in summer, while the distribution of younger fish was limited to waters > 50 m (Figure 7). Discussion Our study shows that the seasonal and horizontal distribution of European hake in the northern and central North Sea has changed visibly since 1997. Higher biomass estimates, particularly in summer, are the result of the increased abundance of age category 3+ fish, and since 1997 these larger fish have expanded across the shallower shelf (50–100 m) of the central and northern North Sea. Younger fish (age categories 1 and 2) are also distributed over a wider area in the western central and northern North Sea. The area occupied by the younger fish appears to be similar regardless of season, suggesting that most immature fish remain in the central-western and northern North Sea after having settled there. Hake population development Fish abundance varies in response to environment (Rijnsdorp et al., 2009), fishing pressure (Casini et al., 2005) and migrations (Hilborn and Walters, 1992). Baudron and Fernandes (2015) showed substantial increase in hake SSB between 2004 and 2011, particularly in the summer. Our analysis, using a different method to calculate swept area, confirms the same trend and shows a continuous increase in total biomass since 2011. Norwegian hake by-catch landings have increased 1600% since 2004 (Fiskeridirektoratet, 2016), and Danish hake landings have increased 300% between 2007 and 2015 (Eurostat, 2017). As a by-catch, this seems likely to reflect overall hake abundance in the area rather than fishers switching to target hake. The IBTS surveys adequately cover the northern and central North Sea down to 200-m depth, and coverage and effort have been similar between years and seasons, except in 1997 and 1998. The rigging of the GOV bottom trawl has been the same during the study period, and though sweep lengths have been adapted in some years in the winter survey depending on fishing depth, we assume that this will not have changed the efficiency of the gear. As bottom trawl stations are selected randomly and independently of fish abundance, and survey effort per ICES rectangle has generally been very similar between years, changes in catch per unit of effort can be hoped to reflect actual changes in abundance rather than changes in catchability. It is possible, however, that seasonal changes in the observed abundance of age category 3+ fish are caused by aggregating or spawning behaviour, resulting in areas with lower and higher densities (Arreguin-Sanchez, 1996). We conclude that observed seasonal trends in abundance and biomass estimates are most likely a realistic reflection of the actual development of hake abundance in the North Sea. The seasonal differences in trends between abundance and biomass, particularly from 2004 onwards are a result of seasonal differences in size composition. In both seasons abundance increases. In winter this is mainly due to an increase in age categories 1 and 2 fish, while in summer abundance increases as greater numbers of age category 3+ move into the North Sea. This proportional increase of large fish in summer also explains why the increase in biomass in summer is much steeper than in winter. In addition to the increased migration of category 3+ fish, the observed increase is also a consequence of actual growth of the hake population, particularly since 2010 (Figure 2b). This suggests that (i) juvenile fish mortality is generally low and that recruitment has been high for several years, (ii) good growth conditions for all age classes prevail in the northern North Sea, and that (iii) fishing mortality of larger fish is relatively low. Although overall abundance decreased in the summer of 2014 and 2015, the biomass continued to increase, which suggests that the observed decrease in number of fish was more than compensated for by the presence of numerous large fish. Length modes and age assignments of hake The assignment of age categories to a multimodal length distribution, given the lack of a validated age-length key, assumes that length cohorts can be assigned to products of a fixed spawning period and followed over time (Pauly and Morgan, 1987; Gulland and Rosenberg, 1992). The spawning season for hake east of Scotland is believed to be from July to September (Werner et al., 2016), implying that in summer the first peak of the length distribution (16 cm) roughly corresponds to age category 1 fish and the second peak with a mode at 33 cm corresponds to age category 2 fish. The selectivity of the GOV bottom trawl leads to undersampling of fish < 10 cm, and thus there are relatively fewer age category 1 fish as 0-group compared with age category 2 fish in the winter surveys. Our observed length modes and assigned age-categories correspond closely to the estimated mean lengths for age 1 (18–20 cm) and age 2 (34–36 cm) fish, based on von Bertalanffy growth parameters for hake captured in the Bay of Biscay (de Pontual et al., 2013), assuming similar growth rates of fish in the North Sea and the Bay of Biscay. The length distribution for winter showed clearly defined modes for age categories 1 and 2, while the higher frequency of large (>40 cm) fish in summer resulted in less defined distributions assigned to age categories 1, 2, and 3+ fish. Spatial distribution of hake The spatial area occupied by fish of all ages has increased steadily over the course of the study period. Distributional changes of many North Sea demersal fish species observed in the last decades have been mainly attributed to climatic changes, i.e. an increase of both bottom and sea surface temperature (Perry et al., 2005; Dulvy et al., 2008), though fishing pressure has also been identified as a possible confounding factor for some species (Engelhard et al., 2014). The increase in temperature has led to demersal fish generally moving deeper (Dulvy et al., 2008), boreal fish species shifting northwards, and increases in the abundance of Lusitanian species, such as the hake, in the North Sea (Engelhard et al. 2011). An alternative hypothesis is that this increase in range could simply be the result of fish moving into marginal feeding areas as a result of competition pressure within the current very large hake biomass. The increased geographical coverage of hake also poses the question of whether growth conditions or survival in the larger part of the central and northern North Sea have improved during the last decade for both juvenile and adult fish. If they have, then this implies that prey availability (food abundance) in these areas would be adequate to support increased numbers of foraging fish. In the case of fish larvae, who feed mainly on various zooplankton (Ferraton et al., 2007), this would imply prey being available during and shortly after the spawning season in summer and late autumn. With increasing length, the proportion of fish in the hake diet increases (Mahe et al., 2007). In the North Sea this includes the mackerel and herring (Werner, 2015), both of which have increased in abundance in recent years (Bakkedeig, 2016). Greater abundance of large hake will very likely lead to increased competition with other piscivorous fish, e.g. saithe (Cormon et al., 2014) and potentially increased cannibalism (Mahe et al., 2007; Preciado et al., 2015). In a recent study Cormon et al. (2016) showed that an increase in hake in the North Sea could have a negative impact on the saithe biomass due to increased predation pressure on Norway pout. The increase in abundance of hake thus poses challenging ecological questions, which could be used to test the application of Ecosystem Based Management of the North Sea. Length- and age- related differences in geographical distribution by seasonal and by depth have also been observed among other hake stocks (Carpentieri et al., 2005; Bartolino et al., 2008) and different hake species (Burmeister, 2001; Wilhelm et al., 2015). Earlier studies found that in winter months predominantly juvenile hake (<50 cm) were present on the north-western North Sea shelf edge at depths >100 m (Sahrhage, 1967; Werner et al., 2016) and along the upper slope of the Norwegian trench and the entrance to the Skagerrak (Bergstad, 1991). The length composition changed in summer and autumn months when larger fish moved from deeper areas onto the shelf, particularly in the central North Sea (Sahrhage, 1964; Bergstad, 1991; Werner et al., 2016). Similarly, Recasens et al. (1998) showed that the depth distribution of adult hake in the north-west Mediterranean Sea changed during the spawning season. In the North Sea, fish in different age categories were unequally distributed across the northern and central regions. The spatial overlap between age categories has changed over time, especially as the oldest fish have become more abundant and spread westwards into shallower waters of the central North Sea. Their distribution has overlapped increasingly with that of recruits, which were initially mostly present in the northern North Sea, but gradually extended southwards into shallower waters (<100-m depth). Large and mature hake moved from deeper waters into shallower areas (100–150 m) in summer. A large proportion of these then spawn on the shelf west and south-west of the Norwegian trench (Werner et al., 2016). However, throughout the study period, age 3+ category fish predominate in the shallow south-eastern area of the central North Sea (<50-m depth) in early summer. This area may be the destination for adults who have “overwintered” in the southern part of the deep Norwegian trench and possibly the Skagerrak. These fish may move into the shallower areas in summer to either spawn, as suggested by the presence of spawning fish in this area (Werner et al., 2016), or to feed, although this would need to be verified. This seasonal migration is independent of the increase in abundance observed in the northern North Sea, since it is observed throughout the study period. The absence of juveniles in shallow waters (<50 m) as well as in areas with high adult abundance could be due to avoidance of increased predation risk by larger hake. Cannibalism in hake and other hake species is well documented, and the extent of it can be linked to juvenile-adult fish overlap and availability of other prey species (Macpherson and Gordoa, 1994; Mahe et al., 2007, Preciado et al., 2015). Conclusions We found a rapid increase in hake abundance and biomass since 2005, confirming patterns reported in previously published results. This increase was initially observed as hake distributed over larger areas of the North Sea, expanding south and partly westwards, occupying a wider range of depths in the process. However, since 2007 the rate of areal expansion has slowed, and the continued increase in hake catches has been caused by higher densities of hake within these areas. Our results also point to an influx of large mature hake into the North Sea in early summer, presumably migrating into the area to spawn. Our data suggest that the juvenile hake in the North Sea are likely products of spawning within the North Sea. Post—spawning, the mature fish leave the area whereas the younger fish likely remain in the North Sea until they reach ages 2–3 before the bulk of them begin seasonal migrations together with the older fish. The progressive increase in hake abundance and distribution has important implications for the ecosystem and for fisheries management. The hake are now included in the multispecies models that gives predation mortality for the North Sea single-species stock assessments (ICES, 2016b), and the presence of an increasing stock of highly predatory fish needs to be considered in any ecosystem-wide management advice or forecasts for the area. These models could be improved with better resolution of the spatial and temporal changes of hake in the North Sea, in particular where adult fish “overwinter” once they leave the area in autumn. Further, while one might readily suspect that fish that used the North Sea as their juvenile nursery area return there to spawn, this remains to be verified through a combination of tagging and genetic studies. Acknowledgements We would like to thank Alexander Beck for providing ICES rectangle areas corrected for land mass and according to the 200-m isoline. References Alvarez P. , Fives J. , Motos L. , Santos M. 2004 . Distribution and abundance of European hake Merluccius merluccius (L.), eggs and larvae in the North East Atlantic waters in 1995 and 1998 in relation to hydrographic conditions . Journal of Plankton Research , 26 : 811 – 826 . Google Scholar CrossRef Search ADS Arreguin-Sanchez F. 1996 . Catchability: a key parameter for fish stock assessment . Reviews in Fish Biology and Fisheries , 6 : 221 – 242 . Google Scholar CrossRef Search ADS Bakkedeig I. E. , Hauge M. , Kvamme C. , Sunnset B. H. , Toft K. Ø. 2016 . Havforskningsrapporten . Fisken og havet, særnr. I- 2016 : Bartolino V. , Ottavi A. , Colloca F. , Ardizzone G. D. , Stefansson G. 2008 . Bathymetric preferences of juvenile European hake (Merluccius merluccius) . ICES Journal of Marine Science , 65 : 963 – 969 . Google Scholar CrossRef Search ADS Baudron A. , Fernandes P. 2015 . Adverse consequences of stock recovery: European hake, a new “choke” species under a discard ban? Fish and Fisheries , 16 : 547 – 725 . Google Scholar CrossRef Search ADS Bergstad O. A. 1991 . Distribution and trophic ecology of some gadoid fish of the Norwegian Deep. 1. Accounts of individual species . Sarsia , 75 : 269 – 313 . Google Scholar CrossRef Search ADS Burmeister L.-M. 2001 . Depth-stratified density estimates and distribution of the Cape hake Merluccius capensis and M.paradoxus off Namibia deduced from survey data, 1990–1999 . South African Journal of Marine Science , 23 : 347 – 356 . Google Scholar CrossRef Search ADS Carpentieri P. , Colloca F. , Cardinale M. , Belluscio A. , Ardizzone G. D. 2005 . Feeding habits of European hake (Merluccius merluccius) in the central Mediterranean Sea . Fishery Bulletin , 103 : 411 – 416 . Casini M. , Cardinale M. , Hjelm J. , Vitale F. 2005 . Trends in cpue and related changes in spatial distribution of demersal fish species in the Kattegat and Skagerrak, eastern North Sea, between 1981 and 2003 . ICES Journal of Marine Science , 62 : 671 – 682 . Google Scholar CrossRef Search ADS Cormon X. , Loots C. , Vaz S. , Vermard Y. , Marchal P. 2014 . Spatial interactions between saithe (Pollachius virens) and hake (Merluccius merluccius) in the North Sea . ICES Journal of Marine Science , 71 : 1342 – 1355 . Google Scholar CrossRef Search ADS Cormon X. , Kempf A. , Vermard Y. , Vinther M. , Marchal P. 2016 . Emergence of a new predator in the North Sea: evaluation of potential trophic impacts focused on hake, saithe, and Norway pout . ICES Journal of Marine Science , 2016. 73 : 1370 – 1381 . Google Scholar CrossRef Search ADS de Pontual H. , Jolivet A. , Garren F. , Bertignac M. 2013 . New insights on European hake biology and population dynamics from a sustained tagging effort in the Bay of Biscay . ICES Journal of Marine Science , 70 : 1416 – 1428 . Google Scholar CrossRef Search ADS Dulvy N. K. , Rogers S. I. , Jennings S. , Stelzenmller V. , Dye S. R. , Skjoldal H. R. 2008 . Climate change and deepening of the North Sea fish assemblage: a biotic indicator of warming seas . Journal of Applied Ecology , 45 : 1029 – 1039 . Google Scholar CrossRef Search ADS Engelhard G. H. , Ellis J. R. , Payne M. R. , ter Hofstede R. , Pinnegar J. K. 2011 . Ecotypes as a concept for exploring responses to climate change in fish assemblages . ICES Journal of Marine Science , 68 : 580 – 591 . Google Scholar CrossRef Search ADS Engelhard G. H. , Righton D. A. , Pinnegar J. K. 2014 . Climate change and fishing: a century of shifting distribution in North Sea cod . Global Change Biology , 20 : 2473 – 2483 . Google Scholar CrossRef Search ADS PubMed Eurostat . 2017 Landings of fishery products. http://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction, do (last accessed February 2018). Ferraton F. , Harmelin-Vivien M. , Mellon-Duval C. , Souplet A. 2007 . Spatio-temporal variation in diet may affect condition and abundance of juvenile European hake in the Gulf of Lions (NW Mediterranean) . Marine Ecology Progress Series , 337 : 197 – 208 . Google Scholar CrossRef Search ADS Fiskeridirektoratet . 2016 . https://www.fiskeridir.no/Statistikk/Statistikkbank (last accessed 2016). Greenstreet S. 2007 . MAFCONS Managing Fisheries to Conserve Groundfish and Benthic Invertebrate Species Diversity: final Report. Fisheries Research Services Marine Laboratory, Aberdeen. 838 pp. Gulland J. A. , Rosenberg A. A. 1992 . A review of length-based approaches to assessing fish stocks. FAO Fisheries Technical Paper, 323. FAO, Rome. 100 pp. Hickling C. 1927 . The natural history of the hake. Parts I and II . Fishery Investigation, Series II , 10 : 1 – 112 . Hilborn R. , Walters C. J. 1992 . Quantitative Fisheries Stock Assessment: Choice, Dynamics, and Uncertainty . Chapman and Hall , New York Huse I. , Iilende T. , Strømme T. 2001 . Towards a catchability constant for trawl surveys of Namibian hake . South African Journal of Marine Science , 23 : 375 – 383 . Google Scholar CrossRef Search ADS ICES . 2013 . Report of the Workshop on Implementation in DATRAS of Confidence Limits Estimation of Abundance Indices from Bottom Trawl Survey Data , 10–12 May 2006, ICES Headquarters , Copenhagen . 39 pp. ICES . 2015 . Manual for the international bottom trawl surveys. Series of ICES Surveys Protocols SISP 10 – IBTS IX, 86 pp. ICES 2016a . Report of the Working Group for the Bay of Biscay and the Iberian waters Ecoregion (WGBIE). ICES Document ICES CM/ACOM: 12. ICES . 2016b . Report of the Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak (WGNSSK), 26 April–5 May 2016, Hamburg, Germany. ICES CM 2016/ACOM: 14.1096 pp. ICES . 2017 . Report of the Working Group on Fish Distribution Shifts (WKFISHDISH), 22–25 November 2016, ICES HQ, Copenhagen, Denmark. ICES CM 2016/ACOM: 55. 197 pp. Kacher M. , Amara R. 2005 . Distribution and growth of 0-group European hake in the Bay of Biscay and Celtic Sea: a spatial and inter-annual analyses . Fisheries Research , 71 : 373 – 378 . Google Scholar CrossRef Search ADS Macdonald P. 2012 . Mixdist: finite mixture Distribution models. R Package Version 0.5-4. http://CRAN.R-project.org/package=mixdist Macpherson E. , Gordoa A. 1994 . Effect of prey densities on cannibalism in Cape hake (Merluccius capensis) off Namibia . Marine Biology , 119 : 145 – 149 . Google Scholar CrossRef Search ADS Mahe K. , Amara R. , Bryckaert T. , Kacher M. , Brylinski J. M. 2007 . Ontogenetic and spatial variation in the diet of hake (Merluccius merluccius) in the Bay of Biscay and the Celtic Sea . ICES Journal of Marine Science , 64 : 1210 – 1219 . Murua H. 2010 . The biology and fisheries of European hake, Merluccius merluccius, in the North-East Atlantic . Advances in Marine Biology , 58 : 97 – 154 . Google Scholar CrossRef Search ADS PubMed Pauly D. , Morgan G. R. 1987 . Length-based methods in fisheries research. ICLARM Conference Proceedings 13, 468 p. International Center for Living Aquatic Resources Management, Manila, Philippines, and Kuwait Institute for Scientific Research, Safat, Kuwait. Perry A. L. , Low P. J. , Ellis J. R. , Reynolds J. D. 2005 . Climate change and distribution shifts in marine fishes . Science , 308 : 1912 – 1915 . Google Scholar CrossRef Search ADS PubMed Preciado I. , Punzón A. , Velasco F. 2015 . Spatio-temporal variability in the cannibalistic behaviour of European hake Merluccius merluccius: the influence of recruit abundance and prey availability . Journal of Fish Biology , 86 : 1319 – 1334 . Google Scholar CrossRef Search ADS PubMed R Core Team . 2016 . R: A language and environment for statistical computing. R Foundation for Statistical Computing , Vienna , Austria . https://www.R-project.org/. Recasens L. , Lombarte A. , Morales-Nin B. , Tores G. J. 1998 . Spatiotemporal variation in the population structure of the European hake in the NW Mediterranean . Journal of Fish Biology , 53 : 387 – 401 . Google Scholar CrossRef Search ADS Orsi Relini L. , Papaconstantinou C. , Jukic-Peladic S. , Souplet A. , Gil de Sola L. , Piccinetti C. , Kavadas S. , et al. 2002 . Distribution of the Mediterranean hake populations (Merluccius merluccius smiridus Rafinesque, 1810) (Osteichthyes: gadiformes) based on six years monitoring by trawl-surveys: some implications for management . Scientia Marina , 66 : 21 – 38 . Google Scholar CrossRef Search ADS Rijnsdorp A. D. , Peck M. A. , Engelhard G. H. , Möllmann C. , Pinnegar J. K. 2009 . Resolving the effect of climate change on fish populations . ICES Journal of Marine Science , 66 : 1570 – 1583 . Google Scholar CrossRef Search ADS Sahrhage D. 1964 . Über die Verbreitung der Fischarten in der Nordsee I. Juni-Juli 1959 und Juli 1960 . Ber. Dt. Wiss. Komm. Meeresforsch , 17 : 165 – 278 . Sahrhage D. 1967 . Über die Verbreitung der Fischarten in der Nordsee II. Januar 1962 und 1963 . Ber. Dt. Wiss. Komm. Meeresforsch , 19 : 66 – 179 . Werner K.-M. 2015 . A first examination of its biology, ecology and fisheries: What is the role of European Hake (Merluccius merluccius) in the waters of the northern North Sea and along the Norwegian coast? MSc thesis, Department of Biology, University of Bergen, Bergen. 85 pp Werner K.-M. , Staby A. , Geffen A. J. 2016 . Temporal and spatial patterns of reproductive indices of European hake (Merluccius merluccius) in the northern North Sea and Norwegian coastal areas . Fisheries Research , 183 : 200 – 209 . Google Scholar CrossRef Search ADS Westgaard J.-W. , Staby A. , Godiksen J. , Geffen A. J. , Svensson A. , Charrier G. , Svedäng H. 2017 . Large and fine scale population structure in European hake (Merluccius merluccius) in the Northeast Atlantic . ICES Journal of Marine Science , 74 : 1300 – 1310 . Google Scholar CrossRef Search ADS Wilhelm M. R. , Jarre A. , Moloney C. L. 2015 . Spawning and nursery areas, longitudinal and cross-shelf migrations of the Merluccius capensis stock in the northern Benguela . Fisheries Oceanography , 24 : 31 – 45 . Google Scholar CrossRef Search ADS Wickham H. 2009 . ggplot2: Elegant Graphics for Data Analysis . Springer-Verlag , New York © International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: email@example.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
ICES Journal of Marine Science – Oxford University Press
Published: Aug 13, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera