Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea

SEARES-01386; No of Pages 9 Journal of Sea Research xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

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

Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea D. Ustups a,⁎, U. Bergström b, A.B. Florin b, E. Kruze a, D. Zilniece a, D. Elferts a, E. Knospina a, D. Uzars a a b

Fish Resources Research Department, Institute of Food Safety, Animal Health and Environment “BIOR”, Daugavgrivas 8, Riga, Latvia Department of Aquatic Resources, Institute of Coastal Research, Swedish University of Agricultural Sciences, Skolgatan 6, Öregrund, Sweden

a r t i c l e

i n f o

Article history: Received 31 January 2015 Received in revised form 26 June 2015 Accepted 29 June 2015 Available online xxxx Keywords: Flounder Turbot Non-indigenous species Introduced Feeding

a b s t r a c t The present study offers a comprehensive analysis of changes in the abundance and diet composition of juvenile flounder (Platichthys flesus) and turbot (Psetta maxima), along with other dominant coastal fish species, before and after the establishment of the alien round goby off an exposed stretch of coast in the eastern Baltic Sea. In the study area, the round goby (Neogobius melanostomus) was recorded for the first time in 2009. After a few years of low abundance, a sharp increase in the population occurred. After the round goby invasion, flatfish juveniles exhibited an increased diet overlap with other species and had a lower feeding success, reflecting an increase in resource competition. For juvenile turbot, the increase was mainly caused by the round goby, while for flounder it was due to both the round goby and the lesser sandeel (Ammodytes tobianus). Juvenile turbot, whose dominant food item before the round goby establishment had been mysids, shifted their diet towards Crangon crangon, reflecting a decrease in mysid abundance by three orders of magnitude and a concurrent doubling in C. crangon abundance in the habitat. At the same time a significant decrease in turbot recruitment was observed. Juvenile flounder had the widest food spectrum of the studied species. When the availability of the primary food item, Bathyporeia pilosa, decreased, flounder juveniles adapted by increasing the share of zooplankton in their diets. No changes in flounder feeding success and recruitment were observed. However, the recruitment estimates of flounder and turbot show an increasing co-variation after the round goby invasion, suggesting that recruitment of the species may currently be regulated by processes in the common nursery habitat. © 2015 Elsevier B.V. All rights reserved.

1. Introduction In the central part of the Baltic Sea two common, ecologically and economically important flatfish species are the flounder (Platichthys flesus) and the turbot (Psetta maxima). Juveniles of both species inhabit shallow, nearshore sandy bottoms in the Baltic Sea and have very specific habitat requirements during their early juvenile phase (Florin et al., 2009; Martinsson and Nissling, 2011; Ustups et al., 2007; Vitinsh, 1989). The juveniles of flatfish remain in these coastal nurseries from the post-metamorphosis stage until they reach maturity (Gibson, 1994; Vitinsh, 1989). During their first year, juvenile flounder and turbot inhabit waters up to one metre deep (Florin et al., 2009; Martinsson and Nissling, 2011), while older juveniles gradually move to deeper coastal waters. The shallow coastal waters warm up early in the season and provide favourable conditions for the feeding and growth of juvenile flatfish as well as other fish species, both as juveniles and as small adults (Florin and Lavados, 2010; Lappalainen and Urho, 2006; Martinsson and Nissling, 2011; Stankus, 2006). Previous ⁎ Corresponding author at: Daugavgrivas 8, Riga, LV-1048, Latvia. E-mail address: [email protected] (D. Ustups).

investigations have identified several dietary guilds in these shallow waters (Ustups et al., 2007). Flounder juveniles are generalists with a wide food spectrum, while turbot juveniles are mainly mysid feeders (Aarnio et al., 1996; Florin and Lavados, 2010; Nissling et al., 2007, Ustups et al., 2007). The round goby (Neogobius melanostomus) is a demersal benthivorous invasive fish species originating from the Ponto-Caspian region. In the Baltic Sea, this species was first reported in 1990 from the Gulf of Gdansk off the Polish coast in the southern part of the sea (Skora and Stolarski, 1993). Since its introduction, the round goby has successfully established itself in the coastal waters of the central Baltic Sea. The first records of round goby occurrence in Lithuania were reported in 2002 (Zolubas, 2003), in Estonia in 2002 (Ojaveer, 2006), and in Latvia in 2004 (Minde, 2007). Currently, one of the highest round goby concentrations and one of the biggest commercial landings in the central Baltic Sea are observed in the coastal area close to the Latvian–Lithuanian border (Knospina and Putnis, 2014). Round gobies breed and feed in shallow water during summer (Kornis et al., 2012). At this time they have a very restricted home range with migrations mostly restricted to distances of a few hundred metres (Ray and Corkum, 2001). Longer migrations, probably up to

http://dx.doi.org/10.1016/j.seares.2015.06.021 1385-1101/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

2

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

several kilometres, take place to and from deeper waters in early spring and late autumn (Sapota, 2012). Round gobies prefer rocky bottom habitats, where blue mussels (Mytilus trossulus) are the main food item (Järv et al., 2011; Karlson et al., 2007; Strāķe et al., 2013). Although round gobies will probably colonize hard before soft bottoms, muddy and sandy habitats are not resistant to invasion (Kornis et al., 2012). In the Baltic Sea, these shallow areas, especially sandy bottoms, simultaneously constitute the main nursery areas for turbot and flounder. Diet analyses in previous studies suggest that the round goby is an opportunistic feeder and will feed on prey according to availability (Kornis et al., 2012; Skora and Rzeznik, 2001). Previous studies have shown negative interactions between the invasive round goby and native fish fauna due to resource and interference competition (Järv et al., 2011; Karlson et al., 2007; Kornis et al., 2012). Thus, there is a concern that the dispersal of the round goby, which is territorial and highly competitive, may have a vast impact on the resident fish communities via several channels: through interference competition, through resource competition via heavy predation on certain benthic prey species, and through direct predation on fish in the early life stages. Although knowledge is scarce, there are also indications that round gobies may predate on newly-settled flatfish (Institute of Food Safety, Animal Health and Environment (BIOR), Riga, Latvia, unpublished data). So far, knowledge on these interspecific interactions is limited in the Baltic Sea, especially concerning potential population-level effects. The objective of this study was to assess the diet overlap between the round goby and other common species of the nearshore fish community, focusing mainly on juvenile turbot and flounder. Previous studies have indicated that the feeding spectrum of turbot juveniles is relative narrow (Ustups et al., 2007), making it vulnerable to resource competition, while flounder juveniles are opportunistic feeding generalists (Nissling et al., 2007; Ustups et al., 2007). However, in the areas subject to round goby invasion, the flounder is a specialist feeder with a small niche width (Järv et al., 2011), which suggests that the presence of the round goby might influence the feeding strategy of the flounder. Utilizing time-series data on the nearshore fish community and the

diets of different fish species, we identify changes in feeding conditions before and after the invasion of round gobies and discuss implications for the recruitment of turbot and flounder.

2. Materials and methods 2.1. Sampling The study was conducted in the coastal zone of the eastern central Baltic Sea, at two locations off the Latvian coast, Pape (56.15° N latitude, 21.03° E longitude) and Jurmalciems (56.30° N latitude, 20.98° E longitude) (Fig. 1). In the study area, bottoms down to a depth of 7 m are mainly dominated by sandy substrates, while below 8 m bottom substrates are composed of boulder, cobble, and gravel (Strāķe et al., 2013). A beach-seine study was conducted annually in the late spring or early summer (May or the beginning of June), to capture the period when the round goby is most abundant in shallow waters. The available information from the Latvian scientific gillnet survey clearly indicated a peak in the abundance of the round goby in shallow waters from April to June. The seine data were collected annually from 1998 to 2014. On each sampling occasion, five hauls were made per location. The seine had a mesh size of 10 mm in the wings and 5 mm in the cod-end. The length of the wings was 12.5 m, and the vertical opening was 1.5 m. The hauls were carried out perpendicularly to the shoreline, starting at a distance of 130 m (up to 3 m deep) and hauling towards the shore. Each haul covered an area of approximately 0.4 ha (Vitinsh, 1989). All hauls were made during daytime under calm wind conditions (b 5 m/s). Fish were sorted by species, counted, weighed (total weight per species) and immediately preserved in 80% ethanol. For all fish species that were included in the diet analyses, the weight and total length of each individual were determined later in the laboratory. In total, 143 hauls were made from 1998 through 2014.

Fig. 1. Study area (black diamonds) off the Latvian coast of the central Baltic Sea. The black dot (Liepaja) indicates the first round goby finding in Latvian waters in 2004.

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

2.2. Flatfish recruitment To describe flatfish recruitment, the flounder and turbot recruitment index produced annually by the Latvian institute BIOR was used (BIOR, 2014; Vitinsh, 1989). The index is based on the average abundance of flounder juveniles in the above-described beach seine surveys in the spring (May–June) and in an additional survey in the same study area in the summer (July–August). 2.3. Zoobenthos Changes in food abundance were analysed using data on two crustacean groups, mysid shrimps and decapods, which constitute the most important juvenile turbot prey taxa and may also be a major food item for juvenile flounder (Florin and Lavados, 2010). In these shallow habitats the mysid group is mainly dominated by the species Neomysis integer, while Crangon crangon dominates the decapod group. Data on mysid and decapod abundance were obtained from beach seine surveys in the study area, where both groups were encountered as bycatch. The available data were divided into two time periods: before (1999–2005) and after (2013–2014) the invasion of the round goby. No mysid or decapod data were available from before 1999 or for the period between 2006 and 2012.

3

5.5–9.5 cm (FL3), and 10–15 cm (FL4)) and three length groups were used for turbot juveniles (b4 cm (TU1), 4–8 cm (TU2), and 9–18 cm (TU3)). For the round goby, no previous information on ontogenetic shifts and feeding of the smallest size group in this study was available. Instead, chronological clustering and the broken stick model were used (R Core Team, 2013) to determine the number of significant groups in the analysis (Neto et al., 2005). Eigenvalues from principal component analysis were plotted on the scree plot to identify the number of significant components. To identify the exact number of components, the broken stick model was plotted. 2.5.2. Comparisons between periods Fish abundance data were divided into two periods: before (1998–2008) and after (2009–2014) round goby colonization in the study area. Changes in the abundance of native fish species between the periods were assessed using a Poisson generalized linear mixed model (GLMM) where the year was added as a random factor. A negative binomial linear mixed model was used to identify annual changes in the abundance of the round goby in the years following its first appearance in the study area. A binomial GLMM was used to test for the significance of changes in the proportion of empty stomachs between the periods for flounder and turbot. Linear regression was used to analyse correlations between the recruitment indices of the juvenile flounder and turbot.

2.4. Diet analysis The analysis of fish diet composition was performed for two time periods: before the invasion of round gobies in the area (1998–2004) and after this invasion (2013–2014). In total, 2793 stomachs of all frequently-occurring fish species in the study area were analysed (Table 1). The stomach content was examined under a stereo microscope, and food items were determined to the lowest possible taxon. Food composition was expressed as frequency of occurrence (% of stomachs) per haul (Hyslop, 1980). For flounder, turbot, and round gobies, feeding success was assessed by the proportion of empty stomachs.

2.5.3. Comparisons of food composition Wilcoxon rank-sum test (WRS) was used to analyse the differences in food composition among length groups of flounder and turbot. Quantile–quantile plots were used for checking the normality assumption of residuals. Due to non-normal distribution of the residuals, a non-parametric test was chosen. As multiple tests per group were performed, p-values were adjusted using the Bonferroni correction. 3. All of the analysis and visualisation of data was performed using R (R Core Team, 2013) 3.1. Diet overlap

2.5. Data analysis 2.5.1. Diet based length groups Prey items were aggregated into nine categories: (Aarnio et al., 1996) Bathyporeia pilosa, (Aarnio and Bonsdorff, 1997) other amphipods, (Aarnio and Mattila, 2000) decapods, (Andersen et al., 2005) fish, (BIOR, 2014) molluscs, (de Gouveia, 2011) mysids, (Florin et al., 2009) polychaetes, (Florin and Lavados, 2010) zooplankton, and (Gibson, 1994) other food items. Fish were assigned to different length groups that correspond to ontogenetic shifts and associated changes in their feeding patterns. Flatfish juveniles were divided into diet-based length groups based on feeding information from studies conducted before the invasion of the round goby (Ustups et al., 2013). Four length groups were used for flounder (b 3.5 cm (FL1), 3.5–5 cm (FL2), Table 1 Number of stomachs examined in the study. Fish species

Years 1998–2004

Flounder Turbot Round goby Greater sandeel Lesser sandeel Perch Sand goby Smelt Three-spined stickleback Total

939 137 33 69 18 15 33 30 1274

2013–2014 280 20 320 169 329 1 31 221 5 1376

Diet overlap, expressed as the percentage similarity between diets, was calculated according to the formula of Shorygin (Ivlev 1961, cited in Langton, 1982):

C xy ¼

Xn i¼1

minðpi ; qi Þ:

ð1Þ

Where Cxy n pi qi min

is the diet overlap index; is the number of food items considered; is the proportion of food item i in the diet of species x; is the proportion of food item i in the diet of species y; and is the minimum operator.

The value of the index ranges from 0 to 100%, with 0% representing no diet overlap and 100% representing identical diets. Because of the sensitivity of this overlap measure to the taxonomic breakdown of prey, statistical methods of evaluating absolute overlap values are not practical. Instead, for the purpose of discussion, the values will be classified as low, 0–29%; medium, 30–60%; or high, N 60% (Langton, 1982). For each flatfish length group in each period, pairwise feeding overlap was measured between this length group and each of the other species and length groups. Pearson's chi-square test was used to test for between-periods differences in the frequency of pairs with high feeding overlap.

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

4

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

addition, in the beginning of the time series, relatively high abundance occurred every 2–3 years, but, during the latter period, only the 2009 year class could be classified as moderate, while all other generations were relatively weak (Fig. 3a). For flounder, no significant difference was found between the two time periods (GLMM, p = 0.441; Fig. 3b). There was no correlation between the two flatfish juvenile recruitment indices for the study period as a whole (linear regression, R2 = 0.039, p = 0.22). However, in the last five years (2010–2014) there was a strong positive correlation (R2 = 0.83, p = 0.019). 4.2. Feeding, diet composition, and food availability

Fig. 2. Average biomass (grams/haul) of flatfish, round gobies, and other fish species in the study area along the Latvian coast from 1998 to 2014.

To assess the aggregate feeding competition faced by flatfish, weighted overlap between each flatfish length group and all other species was calculated, where period-average annual catch biomass was used as the weighting factor. For flounder, turbot and round goby, the total fish species biomass was split by length group. 4. Results 4.1. Fish community The total biomass of fish in the beach-seine data fluctuated without any trend, and the two periods, before and after the invasion of round goby, were not significantly different (GLMM, p = 0.597). The dominant fish species in most of the years was the flounder (representing on average 53.8% of the total biomass), except in the very last year, 2014, when a drastic increase in the abundance of the round goby was observed (Fig. 2). Round goby biomass has significantly increased since this species first appeared in the study area in 2009 (Negative Binomial GLM, p b 0.001). The turbot was the second or third most common fish species in all years (representing on average 9.0% of total biomass over the years). Other abundant fish species were lesser sandeel (Ammodytes tobianus), smelt (Osmerus eperlanus), vimba (Vimba vimba), roach (Rutilus rutilus), and greater sandeel (Hyperoplus lanceolatus), representing on average 8.7%, 4.7%, 4.4% 1.8%, and 1.6%, respectively, of the total fish biomass (Appendix 1). No other fish species had relative abundance above 1% of total fish biomass. The abundance of juvenile turbot decreased significantly after the invasion of round gobies in the study area (GLMM, p = 0.003). In

Based on food composition, three clusters of round gobies were identified using the broken stick and clustering method (Fig. 4), corresponding to length groups 3–5 cm (RG1), 7–14 (RG2), and 15–21 cm (RG3). The smallest round goby length group (RG1) was feeding only on zooplankton and was clearly separated from the other length groups. For the second group (RG2), the principal food items were mysids and decapod crustaceans. Decapods were gradually replaced in the largest round goby group (RG3) by molluscs. The proportion of mysids was stable for the two largest length groups. These two groups also had fish in their diet. Feeding success was lower for turbot in the recent time period relative to the rates observed prior to the round goby invasion (GLMM, p = 0.064, Table 2). No significant difference in feeding success between periods was found for flounder (GLMM, p = 0.316). In the first time period, all stomachs of the smallest flounder (FL1) contained food, as did all stomachs of the smallest round gobies (RG1) in the second period. The smallest flounder FL1 fed purely on zooplankton in both time periods (Fig. 5). For FL2 and FL3 the most important food item in the first period was B. pilosa. In 2013–2014, this food item was significantly less frequent (WRS, FL2—p b 0.001, FL3—p b 0.001). In fact, this item almost disappeared from the diet of flounders and was replaced by zooplankton (WRS, FL3—p = 0.02). For turbot, the most important food item in the first period was mysids (92% and 67% for TU2 and TU3, respectively, Fig. 5). In the second period, the occurrence of this food item decreased to 67% for TU2 (WRS test, p = 0.08) and to 50% for TU3 (p = 0.06). The importance of decapods (mostly C. crangon) increased significantly in the second period (WRS, TU2—p = 0.005). For a general comparison, other dominant fish species inhabiting the coastal zone were included in the analysis. In the first time period, mysids were the most important food item for several fish species (three-spined stickleback (Gasterosteus aculeatus), perch (Perca fluviatilis), sand goby (Pomatoschistus minutus), lesser sandeel, and smelt; Fig. 5). Zooplankton was a secondary food item for lesser sandeel, three-spined stickleback, and greater sandeel. When comparing the two periods, the

Fig. 3. Number of 1-year-old turbot (a) and flounder (b) per 10 hauls of beach-seine. The solid bars denote the period before the invasion of the round goby (1998–2008) and the open bars represent the period after the invasion (2009–2014). The calculated abundances are means of aggregated samples formed by combining spring and summer surveys. The dotted lines represent the long-term averages.

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

5

Fig. 4. Coniss cluster plot (upper panel) and feeding groups (lower panel) of round gobies according to the broken stick model.

importance of zooplankton increased for greater sand goby, lesser sand goby, and smelt. The abundance of mysids in the beach seine hauls was variable among years, with the highest abundance in 2002, when an average of 6956 individuals per haul were observed (Table 3). The mean abundance for the first period was 1392 mysids per seine haul. In the second time period, the mean abundance decreased by three orders of magnitude, to 1.4 individuals per haul, with mysids only found in 2013. Mysid abundance was significantly lower in the latter period compared to the first (GLMM, p b 0.001). The abundance of C. crangon in the latter period was significantly higher than in the period before the round goby invasion (GLMM, p = 0.01), with a mean abundance of 269 individuals per haul compared to a mean of 49 in the first period (Table 3). 4.3. Feeding overlap For flounder, the feeding overlap with other fish species was low to intermediate in the first time period (Table 4). Intraspecifically, the

Table 2 The proportion of empty stomachs for the different length groups of flounder, turbot, and round gobies during the time periods 1998–2004 (period 1) and 2013–2014 (period 2). Fish group

Period 1

Period 2

Number of stomachs

FL1 FL2 FL3 FL4 TU2 TU3 RG1 RG2 RG3

0±0 0.25 ± 0.05 0.33 ± 0.04 0.30 ± 0.04 0.13 ± 0.04 0.14 ± 0.05

0.33 ± 0.21 0.30 ± 0.06 0.29 ± 0.07 0.40 ± 0.07 0.60 ± 0.24 0.19 ± 0.13 0±0 0.35 ± 0.01 0.35 ± 0.02

9 349 472 389 80 77 3 121 196

second and third flounder length groups (FL2–3) had a high overlap, with a wide feeding spectrum for both groups. The diets of turbot groups TU2 and TU3 had high overlaps with each other and with those of perch, sand gobies, and, in the case of TU3, also smelt. For all of these fish groups mysids dominated the diets in the first period. In the second period (2013–2014), the smallest flounder had a high diet overlap with the second smallest flounder group, with greater sandeels, and with lesser sandeels (Table 4). Furthermore, there was a full overlap between FL1 and the smallest round gobies RG1, both of which consumed only zooplankton. FL2 showed a high overlap with RG1, with greater sandeels, and with smelt. For FL3, the feeding overlap was high only with lesser sandeels and smelt. FL4 had high overlap with TU2 and TU3, with RG2 and RG3, with three-spined sticklebacks, and with perch (note the small sample size for the stickleback (n = 5) and perch (n = 1) during the second period, which makes comparisons uncertain). The primary food item of these groups was mysids. TU3 had a high overlap only with RG2 and RG3. The common food item for this group was also mysids. The number of occasions with high feeding overlap increased significantly between periods (Pearson χ2 test, p = 0.03). The average weighted food overlap more than doubled for all flatfish feeding groups (Table 4). For juvenile turbot, the increase in feeding overlap was mainly caused by the round goby, while for flounder the increase in overlap was due to both the round goby and the lesser sandeel. 5. Discussion The present study offers a comprehensive analysis of changes in both the abundance and the diet composition of juvenile flounder and turbot, as well as of other dominant coastal fish species, before and after the establishment of the invasive round goby off an exposed stretch of coast in the eastern Baltic Sea. The time-series data show,

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

6

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

Fig. 5. Food composition for the main coastal fish species in the study area during 1998–2004 (top) and 2013–2014 (bottom). RG1–3 are three length groups of round goby (in increasing order of length), FL1–4 are four length groups of flounder juveniles, and TU2–3 are two length groups of turbot juveniles. The species not disaggregated by length group are three-spined stickleback (3ST), greater sandeel (GSA), lesser sandeel (LSA), perch (PER), sand goby (SGA), and smelt (SMA).

that with the expansion of the round goby, the feeding overlap of juvenile flatfish with other species increased, resulting in a lower feeding success for both flounder and turbot and a decreased recruitment for turbot. These rapid changes illustrate the potential impact on coastal fish communities of this powerful invasive species, which through ontogenetic niche shifts may have effects on several different taxa in the food webs. Overall, this shallow, nearshore habitat constitutes an important nursery and feeding ground for flatfish and a few other species, while many other species utilize the area occasionally. In total, 25 fish species were sampled from 1998 to 2014. Flounder, turbot and lesser sandeel were represented in all years, whereas round gobies were caught every year after their establishment in 2009. Flounder dominated the total biomass throughout the study, except for the very last year, when round goby catches were much greater than those of all other species combined. Across the whole time period, the total biomass of coastal fish did not change significantly, although the biomass in the last year was the highest ever observed. Turbot was the second most common species in the period before the round goby became established in the area. After the establishment of the round goby, the abundance of juvenile turbot decreased, while no change was evident for flounder. While turbot has a local population of benthic spawners Table 3 Abundance of macro zoobenthos in beach-seine hauls. Year

Period

C. crangon

Mysid

1999 2000 2001 2002 2003 2004 2005 2013 2014

1 1 1 1 1 1 1 2 2

78.4 ± 32 35 ± 12 32 ± 19 19 ± 13 25 ± 5 115 ± 28 13 ± 7 70 ± 11 351 ± 206

1120 ± 351 469 ± 157 69 ± 53 6956 ± 6777 161 ± 47 1219 ± 973 3.0 ± 1 2.9 ± 1.9 0±0

in the study area, flounder juveniles are a mix of two populations: a local population of benthic spawners and a larger offshore pelagic spawning population (ICES, 2014). This mixture of two populations of flounder might potentially obscure any effect on the local population. Even though the round goby in the study area was first found in the shallow sandy habitat (with beach seine), the increase in abundance in deeper waters (with rocky bottoms) has been remarkably faster, with a doubling in abundance from one year to the next since the first observation (Knospina and Putnis, 2014). A 2012 study in the area reported that the highest round goby density was found at a depth of 10–15 m, in habitats dominated by boulder and gravel (Strāķe et al., 2013). At the time of that study, only three years after the first observation of the species in the area, the round goby was the third most abundant species. In 2014, the abundance of the round goby increased sharply in another habitat, shallow sandy bottoms, as demonstrated in the current study. This resembles the dispersal pattern in the southern part of the Baltic Sea, where the round goby initially inhabited stony and rocky habitats, but later also occupied sandy bottoms (Sapota, 2004). The current study area is close to the Lithuanian border, where the round goby was first observed in 2002, seven years earlier than in our study area. Assuming that the round goby in the area originates from the Lithuanian population or the Liepaja harbour area, the approximate dispersal rate of round gobies can be estimated to have been approximately 7 km per year. This dispersal rate is similar to that observed in the Curonian Lagoon, Lithuania (Rakauskas et al., 2013). The round goby is an opportunistic feeder, feeding on prey within a preferred size range depending on availability, and shifting diet with habitat and through its ontogeny (Skora and Rzeznik, 2001; Karlson et al., 2007; Rakauskas et al., 2008; Järv et al., 2011). In rocky habitats, the round goby is mainly a mollusc feeder (Skora and Rzeznik, 2001; Strāķe et al., 2013), while in shallower waters the diet range is wider and consists of zooplankton, larger crustaceans, and molluscs (Rakauskas et al., 2008). The round goby competes with other species through resource competition and interference, whereby its aggressive behaviour may

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

7

Table 4 Feeding overlap (Shorygin index) of different length groups of flounder, turbot, and round gobies, and of other common coastal fish species, in period 1 (1998 to 2004) and period 2 (2013–2014). Numbers in bold indicate high overlap between groups. RG1–3 are three length groups of round goby (in increasing order of length), FL1–4 are four length groups of flounder juveniles, and TU2–3 are two length groups of turbot juveniles. The species not disaggregated by length group are three-spined stickleback (3ST), greater sandeel (GSA), lesser sandeel (LSA), perch (PER), sand goby (SGA), and smelt (SMA). Period

1

2

Fish group

FL1 FL2 FL3 FL4 TU2 TU3 FL1 FL2 FL3 FL4 TU2 TU3

Weighted overlap

FL2

FL3

FL4

TU2

TU3

31

19 86

0 48 58

0.9 14 19 16

0 13 20 36 75

69

54 75

0 27 36

0 19 35 58

0 19 35 71 83

RG1

100 69 51 0 0 0

RG2

0 24 35 73 75 82

displace native species to sub-optimal habitats (Kornis et al., 2012). In the studies in the southern part of the Baltic Sea, the round goby has been found to primarily compete with flounder (longer than 10 cm), as evidenced by strong similarities in diet and a negative correlation between flounder and round goby abundance (Karlson et al., 2007). The round goby also restricts flounder habitat utilization and therefore food availability (Karlson et al., 2007). The earlier studies did not cover flatfish nursery grounds and most were performed in deeper water. The current study may thus provide a rare insight into the interactions between round goby and recruiting flatfish. The diet composition of flatfish juveniles in the Baltic Sea is relatively well known, regarding both flounder (for example, Aarnio and Bonsdorff, 1997; Aarnio and Mattila, 2000; Andersen et al., 2005) and turbot (Kostrzewska-Szlakowska and Szlakowski, 1990; Aarnio et al., 1996; Nissling et al., 2007; Florin and Lavados, 2010). The shallow, sandy areas functioning as flatfish nursery grounds also host a large number of other fish species (Ustups et al., 2003; Stankus, 2006; Ustups et al., 2007). To our knowledge, there are no previous studies covering the diet composition of all major species of this shallowwater fish community. This unique dataset offers an opportunity to study feeding overlaps among the species of this community, identifying potential competition. Further, as the diet data span a period of several years, the material also allows an analysis of possible changes in the feeding strategies of flatfish juveniles before and after the invasion of the round goby. Round gobies in the study area preyed mainly upon planktonic copepods, the hyperbenthic N. integer and the larger benthic decapod C. crangon. Three different feeding guilds of the round goby, defined by the ontogeny of the species, were identified in the study area. The smallest round gobies, up to 5 cm length, preyed only on copepods, while larger individuals, which were the most abundant in the study area, consumed mainly mysids and decapods, supplemented by fishes and molluscs. Small burrowing amphipods, such as B. pilosa or Monoporeia affinis, important prey items of other species in the area, especially before the invasion, were not found in the stomachs of round gobies. In the first period, small-sized flounder mainly preyed on planktonic copepods, while from a length of 5 cm, their diet changed from zooplankton to B. pilosa and N. integer. From the size of 9 cm, prey selection shifted again, to larger amphipods, decapods and fishes. In 2013–2014 substantial changes in the food composition of juvenile flounder occurred. The two medium-sized flounder groups (3.5–9.5 cm) now included a large proportion of planktonic copepods in their diet. The share of the previously-important prey taxa, mysids, decapods, and polychaetes, was now low. Only the largest juveniles (from 10 cm) had shifted to decapods and mysids. Further, there was a sharp decrease of B. pilosa in the diet, from about 50% in 1998–2004 to less than 10% in 2013–2014. The decrease of B. pilosa was not a direct effect of round

RG3

3ST

GSA

LSA

PER

SGA

SMA

0 28 42 65 60 60

43 40 32 10 58 57 0 14 32 25 67 50

27 36 32 11 18 20 80 80 57 13 5.1 5.1

48 42 34 13 51 50 92 77 58 7.5 7.2 7.2

0 8.2 13 10 93 73 0 14 32 25 67 50

10 31 39 33 62 65 13 30 50 25 60 50

0 31 41 57 52 69 46 69 87 35 31 31

5 10.7 11.4 10 11.5 12.4 10.6 20 24.1 25.8 24.7 24.9

goby predation, since the latter did not feed on this amphipod. Instead, the increase of C. crangon, a highly potent predator on both B. pilosa and flatfish juveniles (Gibson et al., 1995; Oh et al., 2001; de Gouveia, 2011), seen over the years, might have resulted in a decrease of B. pilosa. Juvenile turbot, whose dominant food item before the round goby invasion was mysids, shifted their diet towards C. crangon, indicating a shift in food availability. Juvenile turbot are highly selective of both prey size and species (Aarnio et al., 1996; Nissling et al., 2007; Ustups et al., 2007). The abundance of the primary food item of turbot juveniles in the study area, the mysid N. integer, decreased by three orders of magnitude in the catches of the beach seine study, suggesting a drastic decrease in their population densities. The decrease in the abundance of N. integer, which is a shallow water species, might be a direct effect of exploitation by the round goby. The diet of round gobies in this study differed markedly compared to other Baltic Sea studies, in that the proportion of molluscs was much lower and instead the proportion of mysids and decapods was higher (Karlson et al., 2007; Järv et al., 2011). This difference may be due to differences in the studied habitats. Previous studies mainly investigated hard bottom or mixed substrate habitats, while here we explore shallow, sandy bottoms, offering much less bivalves, but instead high densities of epibenthic crustaceans. The predation on mysids by the round goby results in a high feeding overlap with both flounder and turbot juveniles, accompanied by a sharp decline in mysid abundance in the habitat. This may be a result of exploitation by round gobies, while the increase of C. crangon may have also contributed to the decline (Oh et al., 2001). A decrease in the occurrence of mysids and a concurrent increase in the presence of zooplankton in the diet was also observed for the lesser sandeel and smelt. Thus, a shift away from mysids due to a decrease of this prey species likely increases the competition for zooplankton prey among juveniles that depend upon this food source, such as turbot and flounder in the early stages of life. After the round goby invasion, a decrease in feeding success was observed for turbot juveniles. Although potential changes in the larval supply are unknown, the large decline in major prey taxa suggests that food shortage may have contributed to the lower recruitment that was observed in the last years. After the establishment of the round goby in our study area no strong year classes of juvenile turbot were observed, whereas in the earlier years strong year classes appeared every two to three years. Turbot juveniles had a very narrow food spectrum in the first period, with mysids as their major prey item. During that period mysids were also highly abundant in the environment. The new food items in the second period, the decapods C. crangon and Palaemon elegans, are more mobile than mysids, grow to a bigger body size, and probably are harder to catch and ingest for juvenile turbot (Janas and Mańkucka, 2010; Grzeszczyk-Kowalska et al., 2014). Of these two, C. crangon dominated in the environment,

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

8

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx

In conclusion, this study identifies changes in the feeding conditions for flatfish, which coincide with, and are likely to have been affected by, a massive invasion of the round goby into the shallow, sandy flatfish nurseries in the eastern Baltic Sea. We also provide evidence that these changes may have affected flatfish recruitment. The study also underscores the importance of maintaining long-term environmental monitoring in order to track changes in the food webs caused by different perturbations due to human activity, noting in particular that the introduction of alien species is currently one of the major impacts on the species of the Baltic Sea.

with an abundance that was estimated to be 15 times higher than P. elegans in the beach seine catches (BIOR unpublished data). The weighted food overlap between turbot and the entire analysed fish guild doubled in the second period. The species that determined this overlap in the second period was primarily the round goby. This implies that round goby might have a strong negative effect on turbot recruitment and possibly also on the local turbot stock. Juvenile flounder had the widest food spectrum of the studied species. When the availability of the primary food item B. pilosa decreased, flounder juveniles adapted by increasing the share of zooplankton in their diet. The productive generations and high feeding success suggest that flounder juveniles have so far been able to adapt to the new feeding conditions and that the round goby so far has not impaired recruitment, unless recruits of the large offshore spawning stock have masked any effect on the local population of demersal spawners. Interestingly, the recruitment estimates of flounder and turbot show an increasing covariation after the round goby invasion, suggesting that flounder and turbot abundance may currently be regulated by the same biotic or environmental factors. Given that the spawning time and habitat of eggs and larval stages differ between the populations to some extent, this pattern suggests that recruitment of both species may currently be determined in the common nursery habitat.

Acknowledgements The research leading to these results has received funding from the Latvian National Research Programme 2014–2017 “EVIDEnT” and BONUS (INSPIRE project), the joint Baltic Sea research and development programme (Art 185), funded jointly from the European Union's Seventh Framework Programme for Research, Technological Development and Demonstration, and from the Latvian Academy of Science. The authors express their gratitude to Dr. Maris Goldmanis for his constructive comments, which helped to improve the manuscript.

Appendix 1. The biomass (grams/haul) of fish species sampled in the study area from 1998 to 2014

Common name

Latin name

Year 1998

Flounder Round goby Turbot Lesser sandeel Smelt Vimba Roach Greater sandeel Sprat Bleak Bream Herring Perch White bream Pikeperch Three-spined stickleback Sea trout Sand goby Sabrefish Ide Eelpout Plaice Ninespine stickleback Black goby Dace Garfish

1999

Platichthys flesus 2181 1586 Neogobius melanostomus Scophthalmus maximus 407 312 Ammodytes tobianus 892 13 Osmerus eperlanus 243 520 Vimba vimba 57 2 Rutilus rutilus 0.7 84 Hyperoplus lanceolatus 203 2 Sprattus sprattus 130 0.8 Alburnus alburnus 7 0.4 Abramis brama 7 29 Clupea harengus 7 Perca fluviatilis Blicca bjoerkna 6 Sander lucioperca 4 Gasterosteus aculeatus 14 Salmo trutta Pomatoschistus minutus Pelecus cultratus Leuciscus idus Zoarces viviparus Pleuronectes platessa Pungitius pungitius Gobius niger Leuciscus leuciscus Belone belone

2000

2001 2002

2003 2004

2005

2006 2007 2008

910

1017 2163

2038 675

802

3972 509

130 261 151 34 44 12 18

67 6 156 121 14 7

0.9 6

2 2 4 0.6

453 49 69 27 2 8 19 50 1 68

159 51 298 363 121 163 8 25 143 70 13 28 25 3

45 4

5

2 1 3

3

1

9

4 5 0.8 769 73

2 4 19

4 2

0.6 2

1.3

3 2

2

477 204 13 96 6 20 86 14 8 13 2 36 5 26 30

6 3

2009 2010

2011 2012 2013

2689

1038 1018 599 2 156 2 672 322 460 55 517 208 373 68 422 19 78 356 187 321 2 46 96 8 22 5 0.6 21 2 7 3 35 128 1 1 52 0.8 28 6 34 5 0.5 3

2014

1358 1465 809 229 225 5455 108 156 132 194 396 283 2 170 38 300 58 45 110 44 46 24 95 33 23 0.8 13 77 5 2 10 6 8 29 1 14 5 1 5 0.6

2

8 0.6

2

3 4

0.6 0.7 0.2 0.1

References Aarnio, K., Bonsdorff, E., 1997. Passing the gut of juvenile flounder, Platichthys flesus: differential survival of zoobenthic prey species. Mar. Biol. 129, 11–14. Aarnio, K., Mattila, J., 2000. Predation by juvenile Platichthys flesus (L.) on shelled prey species in a bare sand and a drift algae habitat. Hydrobiologia 440, 347–355. Aarnio, K., Bonsdorff, E., Rosenback, N., 1996. Food and feeding habits of juvenile flounder, Platichthys flesus (L.), and turbot, Scophthalmus maximus L., in the Aland archipelago, northern Baltic Sea. J. Sea Res. 36, 311–320. Andersen, A.K., Schou, J., Sparrevohn, C.R., Nicolajsen, H., Stottrup, J.G., 2005. The quality of release habitat for reared juvenile flounder, Platichthys flesus, with respect to salinity and depth. Fish. Manag. Ecol. 12, 211–219.

BIOR, 2014. Fish Resources Research Department 2013 Annual Report. Annual Health and Environment, BIOR. Institute of Food Safety, Riga, p. 223 (in Latvian). Core Team, R., 2013. A language and environment for statistical computing. Foundation for Statistical Computing, Vienna, Austria (ISBN 3–900051–07–0, URL http://www.R-project.org/.). de Gouveia, M., 2011. Predation av sandräka (Crangon crangon) på juvenil piggvar (Psetta maxima) och juvenol skrubbskädda (Platichtys flesus): betydelse av yngelstorlek för överlevnad hos piggvar och skrubbskädda efter bottenfällning. Student paper. Högskolan på Gotland, pp. 1–17. Florin, A.-B., Lavados, G., 2010. Feeding habits of juvenile flatfish in relation to habitat characteristics in the Baltic Sea. Estuar. Coast. Shelf Sci. 86, 607–612. Florin, A.-B., Sundblad, G., Bergström, U., 2009. Characterisation of juvenile flatfish habitats in the Baltic Sea. Estuar. Coast. Shelf Sci. 82, 294–300.

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021

D. Ustups et al. / Journal of Sea Research xxx (2015) xxx–xxx Gibson, R.N., 1994. Impact of habitat quality and quantity on the recruitment of juvenile flatfishes. J. Sea Res. 32, 191–206. Gibson, R.N., Yin, M.C., Robb, L., 1995. The behavioural basis of predator–prey size relationships between shrimp (Crangon Crangon) and juvenile plaice (Pleuronectes Platessa). J. Mar. Biol. Assoc. U. K. 75, 337–349. Grzeszczyk-Kowalska, A., Chojnacki, J.C., Raczyńska, M., Raczyński, M., 2014. Biological characterization of Neomysis integer (Lech, 1815) from Pomeranian bay in 2006–2007. J. Ecol. Eng. 2, 40–52. Hyslop, E.J., 1980. Stomach content analysis: a review of methods and their applications. J. Fish Biol. 17–4, 411–429. ICES, 2014. Report of the Benchmark Workshop on Baltic Flatfish Stocks (WKBALFLAT), 27–31 January 2014, Copenhagen, Denmark (ICES CM 2014/ACOM:39: 320). Ivlev, V.S., 1961. Experimental ecology of the feeding of fishes. Yale Univ. Press, New Haven, p. 302. Janas, U., Mańkucka, A., 2010. Body size and reproductive traits of Palaemon elegans Rathke, 1837 (Crustacea, Decapoda), a recent colonizer of the Baltic Sea. Oceanol. Hydrobiol. Stud. 39, 3–24. Järv, L., Kotta, I., Raid, T., 2011. Linking the structure of benthic invertebrate communities and the diet of native and invasive fish species in a brackish water ecosystem. Ann. Zool. Fenn. 49–3, 129–141. Karlson, A.M.L., Almqvist, G., Skora, K.E., Appelberg, M., 2007. Indications of competition between non-indigenous round goby and native flounder in the Baltic Sea. ICES J. Mar. Sci. 64, 479–486. Knospina, E., Putnis, I., 2014. Round Goby — Increasingly Guest in the Latvian Coastal Area (Apaļais jūrasgrundulis - aizvien biežāks viesis Latvijas piekrastē). Latvijas zivsaimniecība 2014, Rīga, pp. 46–51 (in Latvian). Kornis, M.S., Mercado-Silva, N., Vander Zanden, M.J., 2012. Twenty years of invasion: a review of round goby Neogobius melanostomus biology, spread and ecological implications. J. Fish Biol. 80–2, 235–285. Kostrzewska-Szlakowska, I., Szlakowski, J., 1990. Summer food of juvenile turbot, Psetta maxima (L.) and flounder, Platichthys flesus (L.), Świętouść, Pomeranian Bay. Acta Ichthyol. Piscat. 20, 73–87. Langton, R.W., 1982. Diet overlap between Atlantic cod, Gadus morhua, silver hake, Merluccius bilinearis, and fifteen other Northwest Atlantic finfish. Fish. Bull. 80, 4. Lappalainen, A., Urho, L., 2006. Young-of-the-year fish species composition in small coastal bays in the northern Baltic Sea, surveyed with beach seine and small underwater detonation. Boreal Environ. Res. 11, 431–440. Martinsson, J., Nissling, A., 2011. Nursery area utilization by turbot (Psetta maxima) and flounder (Platichthys flesus) at Gotland, central Baltic Sea. Boreal Environ. Res. 16, 60–70. Minde, A., 2007. Attention — Round Goby! (Uzmanību - apaļais jūrasgrundulis!). Latvijas Zivsaimniecības gadagrāmata 2007, Rīga, pp. 89–92 (in Latvian). Neto, P., Jackson, D., Somers, K., 2005. How many principal components? Stopping rules for determining the number of non-trivial axes revisited. Comput. Stat. Data Anal. 49, 974–997. Nissling, A., Jacobsson, M., Hallberg, N., 2007. Feeding ecology of juvenile turbot Scophthalmus maximus and flounder Pleuronectes flesus at Gotland, Central Baltic Sea. J. Fish Biol. 70, 1877–1897.

9

Oh, C.-W., Hartnoll, R.G., Nash, R.,.D.M., 2001. Feeding ecology of the common shrimp Crangon crangon in Port Erin Bay, Isle of Man, Irish Sea. Mar. Ecol. Prog. Ser. 214, 211–223. Ojaveer, H., 2006. The round goby Neogobius melanostomus is colonizing the NE Baltic Sea. Aquat. Invasions 1, 44–45. Rakauskas, V., Bacevičius, E., Pūtys, Ž., Ložys, L., Arbačiauskas, K., 2008. Expansion, feeding and parasites of the round goby, Neogobius melanostomus (Pallas, 1811), a recent invader in the Curonian Lagoon, Lithuania. Acta Zool. Lit. 18–3, 180–190. Rakauskas, V., Pūtys, Ž., Dainys, J., Lesutiene, J., Ložys, L., Arbačiauskas, K., 2013. Increasing population of the invader Neogobius menanostomus (Actinopterygii: Perciformes: Gobiidae), and its trophic role in the Curonian lagoon, SE Baltic sea. Acta Ichthyol. Piscat. 43–2, 95–108. Ray, W.J., Corkum, L.D., 2001. Habitat and site affinity of the round goby. J. Great Lakes Res. 27, 329–334. Sapota, M.R., 2004. The round goby (Neogobius melanostomus) in the Gulf of Gdansk—a species introduction into the Baltic Sea. Hydrobiologia 514, 219–224. Sapota, M.R., 2012. NOBANIS–Invasive Alien Species Fact Sheet–Neogobius melanostomus.– From: Online Database of the European Network on Invasive Alien Species–NOBANIS (www.nobanis.org, Date of access 26/01/2015). Skora, K.E., Rzeznik, J., 2001. Observations on diet composition of Neogobius melanostomus Pallas 1811 (Gobiidae, Pisces) in the Gulf of Gdansk (Baltic Sea). J. Great Lakes Res. 27–3, 290–299. Skora, K.E., Stolarski, J., 1993. New fish species in the Gulf of Gdansk, Neogobius sp. [cf. Neogobius melanostomus (Pallas 1811)]. Bull. Sea Fish. Inst. 1–128, 83–84. Stankus, S., 2006. The role of turbot Psetta maxima and flounder Platichthys flesus trachurus Duncker juveniles in fish communities in the Lithuanian coastal zone of the Baltic Sea. Ekologija 4, 51–57. Strāķe, S., Perkons, V., Yermakov, V., Minde, A., Kazmers, I., Kruze, E., Romanovich, S., Medne, R., Zagars, M., Ojaveer, H., Kotta, I., Jaanus, A., Põllumäe, A., 2013. Nonindigenous species field study report. GES-REGp. 47. Ustups, D., Urtans, E., Minde, A., Uzars, D., 2003. The structure and dynamics of fish communities in the Latvian coastal zone (Pape -Pērkone), Baltic Sea. Acta Univ. Latv. 662, 33–44. Ustups, D., Uzars, D., Müller-Karulis, B., 2007. Structure and feeding ecology of the fish community in the surf zone of the eastern Baltic Latvian coast. Proc. Lat. Acad. Sci. B 61, 20–30. Ustups, D., Müller–Karulis, B., Bergstrom, U., Makarchouk, A., Sics, I., 2013. The influence of environmental conditions on early life stages of flounder (Platichthys flesus) in the central Baltic Sea. J. Sea Res. 75, 77–84. Vitinsh, M., 1989. Methods and results of surveys of the juvenile flounder and turbot in Easter Baltic in 1986 and 1987. Fish. Fors. 27, 59–61 (in Russian). Zolubas, T., 2003. Nauja invazinė žuvų rūšis Lietuvos vandenyse (New Invasive Fish Species in Lithuanian Waters). Jūra ir aplinka. 2–9 pp. 52–56 (in Lithuanian).

Please cite this article as: Ustups, D., et al., Diet overlap between juvenile flatfish and the invasive round goby in the central Baltic Sea, J. Sea Res. (2015), http://dx.doi.org/10.1016/j.seares.2015.06.021