Ecology of Juvenile Turbot and Flounder in the Central Baltic Sea Implications for Recruitment

Ecology of juvenile turbot and flounder in the central Baltic Sea Implications for recruitment

Jesper Martinsson © Jesper Martinsson, Stockholm 2011

ISBN 978-91-7447-402-2

Printed in Sweden by US-AB, Stockholm 2011 Distributor: Department of Systems Ecology, Stockholm University Cover: Juvenile flounder and turbot, photo by Marie Jacobsson

To my son Samuel, thanks for your patience

Abstract

Our understanding of turbot and flounder ecology in the Baltic Sea is insuf- ficient for sound management decisions. This thesis aims to fill some gaps in current knowledge by providing information of the ecology of turbot and flounder within their juvenile habitat, and to relate these findings to issues assumed relevant for recruitment variation. Main focus is on turbot due to its relatively low abundance and high variability in recruitment. The distribution of both species was studied on different scales, as was environmental effects on food consumption in 0-group turbot. The 0-group turbot display a relatively restricted spatial distribution compared to flounder. This is possibly due to a more specialized diet, which may make them more vulnerable to habitat degradation, especially eutrophication as a strong negative correlation was found with the organic content in the sediment. The species show high temporal and spatial overlap when settling in July-September, with peak abundances in August, and at depths <1 m. Both species display sedentary behavior within the nursery ground. Compared to flounder, turbot was more mobile potentially due to its restricted diet calling for extended searches. For turbot, feeding conditions appear to vary between size groups, which potentially could cause variations in survival between years through size-selective mortality. But, the predation may be low in central Baltic Sea as the abundance of the main predator, brown shrimp are comparatively low during flatfish occupancy. A significant positive relationship was found between the recruitment of turbot and flounder, which suggests that no inter- specific interactions during the juvenile stage affect recruitment. This co- variation also suggests that the recruitment of the species is determined by the same phenomena, potentially by large scale abiotic factors during the egg- and larval stage. For turbot, additional variability is potentially generated during the juvenile stage due to its relatively restricted food and habitat requirements. The specific habitat demands of turbot revealed in this thesis may be used to protect and restore essentially nursery grounds.

Keywords : distribution, habitat use, consumption, recruitment, Psetta maxima , Platichthys flesus , nursery ground, juvenile, Baltic Sea

Sammanfattning

Förståelse för vilka faktorer som påverkar rekryteringen av en art fungerar är nödvändigt för en effektiv förvaltning. Kunskapen kring piggvarens och flundrans ekologi i Östersjön är emellertid otillräcklig, vilket hindrar sådana ambitioner, och syftet med denna avhandling är att fylla i några av dessa kunskapsluckor.

Fördelningen av piggvarens och flundrans årsyngel och användning av upp- växtområden studerades på olika skalor runt Gotlands kust. Vidare studerades piggvarens fysiologiska respons på rådande förhållanden genom upp- skattning av storleksspecifik konsumtion.

Piggvarens årsyngel är relativt begränsade i sin utbredning jämfört med flundra, vilket medför att piggvaren är mer känslig mot habitatförstöring. Arterna uppvisar dock stort överlapp i både temporal och rumslig fördelning då de båda bottenfäller i kohorter under perioden juli-september med högst tätheter i augusti. Yngel av båda arterna återfanns i huvudsak på bottnar grundare än 1 m och var relativt stationära inom uppväxtområdet. Piggvaren var dock något mer mobil än flundran, vilket skulle kunna bero på piggvarens mer specialiserade diet som sannolikt kräver ett större födosöksområde. Piggvarens födokonsumtion var relativt hög men något begränsad vissa år för storlekar under 40 mm, vilket kan bero på variation i födotillgång. Ge- nom storleksspecifik predation på ynglen skulle detta kunna påverka variation i överlevnad mellan år. Predationstrycket kan dock vara lågt då den huvudsakliga predatorn, sandräka, uppvisar låga tätheter under den tid piggvaren och flundran är närvarande i uppväxtområdet. Ett positivt samband mellan piggvarens och flundrans rekrytering kan betyda att denna bestäms av samma faktorer under ägg- och larvstadiet och att eventuella mellan- artsinteraktioner inte har någon negativ effekt på rekryteringen för någon av arterna. För piggvar skulle ytterligare variation i rekrytering kunna genereras under det juvenila stadiet på grund av artens restriktiva födo- och habitatval. Resultaten över arternas habitatval skulle kunna användas för skydd och restaurering av viktiga uppväxtområden.

List of Papers

Below follows a list of the papers supporting the thesis and are referred to in the text by their roman numerals:

I Martinsson, J . and Sällebrant, J. Habitat associations of 0-group turbot ( Psetta maxima ) and flounder ( Platichthys flesus ) at Got- land, Central Baltic Sea. Submitted manuscript .

II Martinsson, J . and Nissling, A., 2011. Nursery area utilization by 0-group turbot ( Psetta maxima ) and flounder ( Platichthys flesus ) at Gotland, Baltic Sea. Boreal Environment Research 16: 60-70.

III Martinsson, J . The potential of Visible Implant Elastomer as a nursery role evaluation tool for 0-group turbot ( Psetta maxima ) and flounder ( Platichthys flesus ) habitats. Submitted manuscript .

IV Martinsson, J ., Nissling, A., Jacobsson, M and Retz, R. Poten- tial food limitation of 0-group turbot: A bioenergetics approach. Submitted manuscript

Paper II is reprinted by the kind permission by Boreal Environment Re- search.

My contribution to the papers was: ( I) carried out half of the field work, analyzed field samples, performed all statistical analyses and conducted main part of writing – (II) compiled and processed data, performed all statistical analyses and wrote major part of the manuscript – (III) – planned the study, carried out field work and modeling and wrote the manuscript – (IV ) contributed to a major part of the planning process, field work and data processing, conducted all modeling and main part of writing.

The thesis is the result of a collaboration between Stockholm University and Gotland University.

Contents

Abstract ...... v Sammanfattning ...... vi List of Papers ...... vii Introduction ...... 11 Recruitment of fishes ...... 11 Recruitment of flatfishes ...... 13 Turbot and flounder in the Baltic Sea ...... 14 Aims of the thesis ...... 15 Methods ...... 16 Study area ...... 16 Field studies ...... 16 Laboratory experiments ...... 18 Quantitative approaches ...... 19 Results and discussion ...... 21 The ecology of 0-group turbot and flounder ...... 21 Potential implications for recruitment of turbot and flounder in the Baltic Sea ...... 23 Final Remarks ...... 26 Acknowledgements ...... 27 References ...... 29

Introduction

Recruitment of fishes The factors determining recruitment of young fish to the adult population has been under the scope of fisheries research for about a century. The first hypothesis of what causes variation in recruitment was proposed by Hjort (1914). He suggested that mechanisms such as advection of larvae from the nursery grounds and starvation at time of first feeding were important in generating recruitment variability. This was followed up by the match- mismatch hypothesis by Cushing (1975), which suggests that fishes in tem- perate waters, where the timing of spawning is about the same every year, can hit or miss the period of abundant plankton since the timing of the production of phytoplankton and subsequent production of zooplankton show higher inter-annual variability. Hence, high mortality rates are to be expected at a mismatch between the timing of spawning and the production cycle. The member-vagrant hypothesis (Sinclair 1988) further emphasis the timing of spawning and favorable oceanographic conditions that ensures that the larvae are retained in or transported to suitable environments. The role of factors affecting transportation of larvae have been shown to be important for the survival of several species (Van der Veer et al. 1998, Hinrichsen 2001, Hinrichsen et al. 2003, Abella et al. 2008). Other hypotheses that account for the survival during the early life stages are the “growth-mortality” hypothesis (Anderson 1988) and the “Bigger is better ” hypothesis (Houde 1987). Both suggest that survival is a function of growth rate. For example, individuals growing fast are susceptible for predation during a shorter period of time compared to slow growing individuals. Furthermore, low growth rate could decrease the survival of individuals during the first winter due to a lower amount of energy reserves and consequently higher starvation rates (e.g. MacLean et al. 1981).

Obviously a combination of physical and biotic factors is important in determining the level of recruitment. Some of these factors may also be correlated to each other and/or act in concert. For example, temperature has a direct effect on several processes that could affect recruitment. As for hake in the Mediterranean (Bartolino et al. 2008), high temperature was shown to be negatively correlated to recruitment. The temperature was suggested to have direct effects on egg development and larval growth, as well as indirect

11 effects by increasing water stratification thereby preventing vertical mixing, and reducing the amount of nutrients needed for the production of phyto- and zooplankton. Further, egg and larval mortality of cod in the Baltic Sea increases at low salinity and oxygen concentrations, as a direct effect and as the overlap between the cod eggs and clupeid predators may increase as an indirect effect (Nissling 1995, Köster et al. 2005). The correlation between factors affecting the survival and subsequent level of recruitment makes it difficult to reveal the causal relationship between important mechanisms and recruitment, as well as the actual effect of specific mechanisms on recruitment.

A body of literature support the hypothesis proposed by Gulland (1965), i.e. that recruitment is determined at the egg- and larval stages under the control of abiotic factors (Legget and DeBlois 1994). The impact of these controlling factors on variability in recruitment is strongest for pelagic species, as they remain in a 3-dimensional environment throughout their lives (Iles and Beverton 1998, Rothschild 2000). In contrast, demersal species such as flatfishes shift from a pelagic to a demersal environment as early juveniles, thus being influenced only by two dimensions. This shift is often associated with concentrations in restricted habitats, which initiates density-dependent mortality due to e.g. competition and predation (Bailey 1994, Iles and Beverton 1998, 2000). Hence, the ultimate number of larvae that is determined under the control of abiotic factors is regulated by density-dependent processes during the juvenile stage, and thereby dampens variability in recruitment (Iles and Beverton 1998, 2000, Van der Veer et al. 2000). Density-dependent processes also regulate pelagic fish populations (Zheng 1996, Nash et al. 2009), but not to the same extent (Jennings et al. 2001). Populations expe- riencing density-dependent effects may compensate for low stock levels since mortality decreases due to the relaxation of these density-dependent effects. Hence, the higher the density-dependent mortality experienced by a population, the higher the compensation will be at lower population levels. Moreover, the number of recruits in stocks subjected to strong density- dependent effects would be more or less independent on spawning stock biomass as the mortality increases with the number of individuals produced. This may explain why there appear to be no relationship between spawning stock biomass and recruitment of several flatfish stocks (Leggett and Frank 1997, Iles and Beverton 2000).

12 Recruitment of flatfishes As for fish in general, factors affecting the survival during the egg and larval stages of flatfish have been shown to generate variability in recruitment. The most important factors acting on these stages are proposed to be large scale abiotic factors such as temperature, hydrodynamic circulation and wind (Riley et al. 1981, Van der Veer 1986, Van der Veer et al. 1998, 2000).

Flatfishes in general show relatively low variability in recruitment compared to pelagic species due to density-dependent processes when the juveniles settle within the nursery grounds (Bailey 1994, Leggett and Frank 1997, Iles and Beverton 1998, 2000). Studies of plaice in the North Sea and the west coast of Sweden have shown that this regulation is primariliy due to the predation by shore crab ( Carcinus maenas ) and brown shrimp ( Crangon crangon ) in the beginning of the juvenile stage (Pihl 1990, Bailey 1994, Modin and Pihl 1996). However, other predators may also be responsible for the regulation occurring during the juvenile stage (Nash and Geffen 2000). But, reduced growth causing higher mortality due to competition is not likely to occur as the carrying capacity rarely is reached (Van der Veer et al. 2000). For species relying predominantly on hyperbenthic food, such as mysids (e.g. Neomysis integer ) and juvenile fish, variation in recruitment might even be generated during the juvenile stage as these food items show high intra- and inter annual variability (Bonsdorff and Blomqvist 1993, Van der Veer 2000).

According to the nursery size hypothesis (Rijnsdorp et al. 1992) the ultimate number of recruits to the adult stock may be positively correlated to the surface area of the nursery grounds. This may explain differences in level of recruitment to the adult population both within and between species (Gibson 1994, Van der Veer et al. 2000). Species being more specific in their habitat requirements might have a lower number of recruits to the adult population compared to generalists in respect to habitat, as the surface area of the nursery grounds would be smaller (Gibson 1994). It is not necessarily the size of the nursery grounds per se that would determine the level of recruitment, but the distribution over a large area. A wide distribution of suitable nursery grounds would thus increase the probability of survival of the settling larvae.

According to the species range hypothesis (Miller et al. 1991), the factors affecting recruitment might vary for species in different areas due differences in life history traits and prevailing abiotic conditions. In the North Sea, variation in hydrodynamic- and wind conditions have been shown to control the transportation of plaice larvae to nursery areas and therefore works as factors generating variability in recruitment (Van der Veer et al. 1998). In the Baltic Sea, however, this factor might not be as important for controlling

13 the recruitment for those flatfish populations that spawn at the coast with demersal eggs. Further, irregular inflows of saline water might act as a factor generating variability in recruitment in the Baltic Sea affecting egg- and larval survival (Nissling et al. 2002, 2006).

Turbot and flounder in the Baltic Sea In contrast to species in the North Sea, flounder (Platichthys flesus ) and turbot ( Psetta maxima ) in the Baltic Sea have evolved demersal eggs in contrast to pelagic, and spawn at offshore banks and at the coast (Bagge 1981). This is an evolutionary response to the brackish water of the Baltic Sea that comprise a salinity gradient ranging from about 3-8 psu in the surface water and between 10-20 psu in the deep basins below the halo- cline (Voipio 1981). Although the higher salinity in the deep basins provides opportunity for successful spawning of flatfishes with pelagic eggs, such as flounder, plaice and dab ( Limanda limanda ). The feature of pelagic eggs is completely lost in turbot while flounder spawn with pelagic eggs in the deep basins (ICES Sub-division (SD) 24-26 and 28-2) (Figure 1). For both species spawning is successful at about 6 psu (Nissling et al. 2002, Figure 1. The Baltic Sea and ICES Sub- 2006), but the viable hatch of divisions (SD, downloaded from turbot is significantly reduced be- www.ices.dk). low 7 psu (Nissling et al. 2006). Thus, flounder is best adapted to the low salinity and may spawn as far north as the Bothnian Sea (SD 30). Spawning of turbot on the other hand is rare north of Åland (SD 29) (Bagge 1981). Hence, reproductive success is influenced by irregular inflows of saline water from the Kattegat affecting both species abundance and distribution (Nissling et al. 2002, 2006). As the viable hatch is reduced when spawning in a brackish environment, the fecundity is higher for both turbot and flounder with demersal eggs compared to other areas where pelagic eggs are used (Nissling and Dahlman 2010, Nissling et al. , in prep.). After spawning in spring, flounder and turbot feed in shallow water at the coast, followed by an offshore migration in autumn to larger depths (Molander 1964, Bagge 1987). The larvae of both species settle in cohorts during summer, from mid-June

14 and late July for flounder and turbot respectively in shallow sandy bays (Nissling et al. 2007, Florin et al. 2009). The juveniles migrate offshore in autumn (Molander 1964).

In the Baltic, both flounder and turbot have been subjected to high fishing pressure in the past, resulting in population declines (Ojaveer et al. 1985, Temming 1989). Flounder yield the highest catches compared to the other flatfish species in the Baltic Sea and is today predominantly caught in the cod fisheries as by-catch. A high exploitation rate of flounder resulted in a decline of the flounder stock in the 1930s. The stock recovered during the 1940s-50s and has been relatively stable since then in the southern Baltic Sea, whereas an increase was observed in the 1990s of the northern stock (Florin 2005). Exploitation of turbot increased in the 1950s, with highest catches in the 1990s due to increased fishing effort (Florin 2005). The land- ings decreased during the next decade despite high fishing effort and turbot was listed as near threatened in the Swedish Red List (Gärdenfors 2005). Turbot has, however, recovered in recent years and has consequently been removed from the list.

Overall, turbot is less abundant and display higher variation in recruitment compared to flounder in the Baltic Sea (Molander 1964, Draganik et al. 2005, Florin 2005). The reason for this is yet unknown as the knowledge of the mechanism affecting the survival during the early life stages is limited. Baltic turbot and flounder populations have received fairly little attention compared to species of more commercial importance, which hinders sustain- able management of these two species. At present, no management options can be presented by the International Council for the Exploration of the Sea (ICES) as knowledge about stock structure and recruitment is limited (ICES 2011a, 2011b).

Aims of the thesis The aim of the thesis is to increase the knowledge of juvenile flatfish ecology in the central Baltic Sea with special emphasis on 0-group turbot, and to relate these findings to issues assumed relevant for recruitment variation.The thesis focuses on the distribution of the species on different scales, a regional scale (paper I) and within typical nursery grounds (paper II & III), as well as the functioning of 0-group turbot within nursery grounds in terms of size specific consumption (paper IV). Examination of both temporal and spatial distribution and how species function within a habitat is essential to grasp the ecological context of any organism, and hence for highlighting possible factors affecting survival and subsequent fluctuations in the population.

15 Methods

Study area The studies in the thesis were either conducted in (paper I-III) or applied to (paper IV) shallow bays or coastal areas at Gotland, central Baltic Sea (N 57° 28' 32", E 18° 32' 19"), between 2003 and 2008. The spatial distribution of 0-group turbot and flounder on a regional scale was examined in a total of 82 locations (paper 1) classified according to the dominating substrate; sand, soft sand, gravel or flat rock (Figure 2a). Further, a set of six sandy bays (Figure 2b), differing in exposure to the sea and organic content, was used to explore the spatial and temporal distribution of 0-group turbot and flounder on a smaller scale, i.e. within bays (paper II & III). The sediment of bay A, C and E consist of plain sand whereas bay B, D and F comprise in full or partly muddy sand (Nissling et al. , 2007). Bay F was used to assess the movement patterns of the species (paper III) as it is a quite characteristic nursery ground for both species and easy to sample due to its small size (only about 300 m across). Bay A and B were also used in the estimation of size specific consumption of 0-group turbot (paper IV).

Field studies The field studies were conducted in August 2006 - 2008 (paper I & III) and mid-July to mid-September 2003- 2005 (paper II). Juvenile 0-group turbot and flounder were sampled using either a beach seine (paper II & III) or a push net (paper II). The beach seine (Figure 3) is very effective on sandy beaches but its use become very limited in habitats containing structures such as stones and boulders. In these areas the push net is more effective due to its maneuvering capabilities. The density of each species was calculated using the length of each haul and the width of the gear. Each individual was measured to the nearest millimeter ( Lt), which enabled the separation of 0- group turbot and flounder from older individuals (paper I & II). Individuals of both species were preserved in 96 % ethanol for stomach analyses (Nissling et al. 2007) and otolith extraction (paper IV).

Figure 2. The island of Gotland (ICES SD 27 and 28-2) and locations sampled to a) reveal the spatial distribution in relation to habitat characteristics ofdistribution within nursery grounds (paper II-III). Observed weight at age data and thermal regime in bay A and B was used to estimate size specific consumption of 0-group turbot (paper IV).

Temperature and salinity was recorded at each location and sampling occa- sion in paper I & II. In paper I, water- and sediment samples were also col- lected for estimating turbidity, grain size, grain sorting and organic content. Also, the presence/absence of different types of vegetation was recorded. In paper III, a more detailed habitat classification was carried out in bay F according to substrate, vegetation and depth to aid the discussion of estimated movement patterns of the species.

The lateral movement patterns, i.e. along the shore, of the species in bay F was assessed by a mark – recapture experiment (paper III). Juvenile turbot and flounder were caught in beach seine hauls in the offshore-onshore direc- tion in both ends of the bay. The fishes were tagged with different colours of Visible Implant Elastomer, which corresponded to tagging date and each of four treatment groups; released at the same site as captured or transplanted to the other end of the bay. Fish were recaptured at seven occasions along a total of 20 transect evenly distributed over four sites within the bay.

Figure 3. Sampling of juvenile flatfish using a beach seine in bay C

Laboratory experiments Juvenile turbot (20-81 mm) and flounder (23-40 mm) were caught in bay F in August 2007 and brought to the Ar Research Station for examination of the initial effect of Visible Implant Elastomer on mortality and growth (paper III). The fish of each species were measured, divided into two size classes, tagged and placed in small aerated trays with sand on the bottom to resemble natural conditions and fed frozen chironomids once a day during the experiment period, i.e. six days, where after the fishes were measured again.

The maximum growth was estimated for 0-group turbot at four temperatures (14-27 ˚C) in 2009 and 2010 (paper IV). Individuals 20-29 mm in size were used in the 2009 experiment whereas the growth was estimated for three different size classes (20-29, 30-39 and 40-40 mm) in 2010. The size classes were chosen according to ontogenetic differences in diet. Turbot <30 mm mainly feed on calanoid copepods, chironomids and amphipods (mainly B. pilosa ). Mysids ( N. integer ) are found in diet of fish >27 mm and is an important part in the diet of fish >30 mm. The share of juvenile fish (gobies and sticklebacks) in the diet increases with size and are the main prey item found in stomachs of fish ≥55 mm (Aarnio et al. 1996, Nissling et al. 2007, Florin and Lavados 2010). Turbots were fed ad libitum for 7-8 days according to the size specific food preferences. The dry weight (g) was estimated from the observed length at the end of the experiment by an established weight-length relationship (paper IV). Also, the specific growth rate (g g -1 d- 1) was calculated for each individual.

The sagitta otolith was extracted from a total of 187 turbot individuals caught in bay A and B, grinded and digitalized at varying magnifications (paper IV). The number of rings was counted from the accessory growth centres (Morales-Nin 2000, Figure 4) to the edge of the otolith. These rings represent the number of days from metamorphosis to the day of capture, which provided a relationship between fish weight and age to be used in the estimation of size specific consumption of 0-group turbot (IV).

Figure 4. Picture of a 0-group turbot sagitta otolith. Ex- amples of accessory growth centres are highlighted with arrows (Photo: Marie Jacobsson).

Quantitative approaches Numerical analyses and statistical tools were used to draw conclusions about ecological patterns. These include generalized linear models (paper I), permutation tests (paper II), non-parametric tests (paper III) and ecological models (paper III & IV).

A common problem when working with species distribution is that the data seldom meet assumptions of normality and homogenous variances, which is needed to proceed with conventional parametric tests such as ANOVA or linear regression (Zuur et al. 2007). However, generalized linear models are extensions of ordinary linear regression, which allows for the inclusion of

19 any family of distribution by the use of a link function. Ordinary linear regression or ANOVA are thus special cases of a generalized linear model using the Gaussian probability distribution and the identity link function (Lindsey 1997). The binomial probability distribution and the logit link function were used to model the probability of presence of the species in relation to a set of environmental variables (paper I).

Permutation tests could also be used relaxing the assumption of normality. The basis of permutation tests is that if the null-hypothesis holds true, the origin ( i.e. group) of each observation does not matter. Any observation obtained in group A could just as well have been obtained in group B. The test therefore shuffles the observations between the groups ( i.e. depths) and cal- culates a statistic at each shuffle. In the case of PERMANOVA (Anderson 2001, McArdle and Anderson 2001), which was used to analyze the spatial distribution of the species in relation to depth (paper II), a distribution of F- values are obtained, when the null-hypothesis holds true. The F- value that was obtained using the original dataset is compared to the distribution of F- values leading to rejection or acceptance of the null-hypothesis. The test requires no assumptions of normality, but variances should be equal as significant results may be due to differences in either the mean or the variance between the groups (Anderson 2001).

The movement patterns of 0-group turbot and flounder was analyzed using a random walk model (Burrows et al. 2004), which is based on three parame- ters that provide a set of assumptions about dispersal by constraining them in different combinations (paper III). Further, a bioenergetics approach was used to estimate the consumption of 0-group turbot in the field (paper IV). Bioenergetics models are effective tools to provide a mechanistic understanding of habitat mediated effects on the performance of an organism (Hansen et al. 1993), and thus also survival and population dynamics. Bio- energetics serves as a link between external factors, such as temperature and food availability, and responses by the organism in the form of consumption and growth.

20 Results and discussion

The ecology of 0-group turbot and flounder The spatial and temporal distribution of a species is the result of numerous causal events affected by inherited characteristics through numerous genera- tions. These characteristics determine the physical context and boundaries of a species appearance as well as responses to the surrounding environment (Begon et al. 1990). Such responses may affect the distribution by changes in location to remain within optimal conditions to maximize fitness (Gibson 1997) or by mortality in unsuitable habitats.

0-group turbot displayed restricted habitat requirements on the regional scale mainly occupying sandy locations of low organic content (paper I). 0-group flounder, on the other hand, expressed a more general distribution with relatively high probability of presence at all types of substrates (sand, soft sand, gravel and flat rock) although preferring plain sand and gravel. The probability of presence of flounder was also negatively correlated to organic content, but the relationship was not as strong as for turbot. Further, the probability of presence of turbot and flounder were correlated to other environmental variables. However, these were correlated to substrate in the same manner as the species. Thus, the substrate itself or perhaps some other unmeasured variable correlated to substrate may be responsible for the observed distribution patterns. The restricted habitat preferences of turbot could possibly reflect its relatively specialized diet, mainly feeding on a number of epi- and hyperbenthic prey. 0-group flounder on the other hand is more general, feeding on a variety of endobenthic species such as copepods, chironomids and oligochaetes (Aarnio et al. 1996, Nissling et al. 2007, Florin and Lavados 2010) and may thus potentially feed successfully on several types of habitats. The specific demands of turbot may imply that it is more vulnerable to habitat degradation, especially due to eutrophication leading to higher organic content in the sediment (Cederwall and Elmgren 1990). The suitability of different habitats should, however, be evaluated also in respect to growth, mortality, habitat use and connectivity to the adult habitat in addition to abundance as to make a full assessment. In the case of estimating growth, mortality and habitat use within nursery grounds, tagging with Visible Implant Elastomer may be useful (paper III).

21 0-group turbot and flounder show considerable overlap in both temporal and spatial distribution (paper I & II). Both species settle in sandy bays in cohorts during July-September with peak abundances in August and mainly occupy depths <1 m). The depth distribution may reflect thermal preferences as larger individuals, generally displaying lower optimum for maximum growth (e.g. Imsland et al. 1996, paper IV), were found at increasing depths. The depth distribution may also be affected by the distribution of food and predators as suggested for plaice (Burrows et al. 1994, Gibson et al. 1998). Mysids, which are important prey items for turbot >30 mm, mainly occupy shallow depths and may therefore affect the vertical distribution of turbot. The main predator, brown shrimp on the other hand are mainly found at larger depths (paper II). The overlap in distribution of 0-group turbot and flounder enables the possibility of inter-specific interactions, which may explain the observed negative relationship between feeding incidence of turbot <30 mm and 0-group flounder density (Nissling et al. 2007). Howev- er, no apparent negative relationship as could be expected was displayed between the proportion of maximum consumption of turbot <30 mm and the density of 0-group flounder (paper IV). Thus, according to these results, the presence of flounder does not have an effect on the consumption of turbot. This relationship was not tested statistically, though, as only three data points were available (2003-2005). Flounder may probably utilize the nursery ground to a high extent also as 1-group (paper II) as it continuous to prey on endobenthic species whereas turbot needs to search for appropriate food items on larger depths. The differing strategies of the species can also explain why turbot was more mobile compared to flounder within the juvenile habitat (paper III) as predictions from optimal foraging models suggest that specialists should adopt a wider home range compared to generalists (e.g. Ford 1983). However, both species were quite sedentary moving on average less than 100 m from the release site in 22 days.

The conditions within a habitat also affect growth, which may have an effect on survival and thus the number of offspring reaching age of maturity. Ac- cording to the “growth-mortality” hypothesis (Anderson 1988), restricted growth during ontogeny due to low food availability, competition or the presence of predators can affect the mortality rate through size selective mortality (Van der Veer et al. 1994). In flatfishes, size-selective mortality may operate by predation by brown shrimp, which targets individuals <30 mm (Van der Veer and Bergman 1987). The proportion of maximum consumption (the P-value) of 0-group turbot was in general relatively high (about 0.9) compared to an investigation on 0 - group flounder in Scotland where the proportion of maximum consumption was estimated to 0.54 (Ste- vens et al. , 2006). Studies on other species report on values ranging between 0.2 and 0.7 (Luo and Brandt, 1993; Tabor et al. , 1996; Chipps and Wahl, 2004; Mazur et al. , 2007). Studies have previously reported that fishes re-

22 strict foraging activities (e.g. Werner and Hall, 1988) or occupy sub-optimal habitats at high predation risk (e.g. Power, 1984, Mittelbach, 1986), which consequently results in limited consumption rates. Hence, potentially the relatively higher proportion of maximum consumption in the Baltic Sea reflect lower predation pressure as the main predator, brown shrimp, show quite low abundances during the period turbot and flounder juveniles occupy the nursery grounds (A. Nissling, unpublished). But, when estimating different P-values for different size groups, turbot individuals <40 mm displayed quite low values some years, i.e. about 0.6 (paper IV). Potentially variations in food supply could be responsible for these patterns. The hyper- and epi- benthic species preferred by turbot display variation both within and between years, which may affect the availability of suitable prey (Bonsdorff and Blomqvist 1993, Nissling et al. 2007). Intra-specific competition cannot be excluded, but a positive rather than a negative relationship as expected in such a case was indicated between the proportion of maximum consumption and density of 0-group turbot (paper IV).

Potential implications for recruitment of turbot and flounder in the Baltic Sea As turbot is restricted to plain sand habitats, variability in recruitment may be generated if settling in this type of habitat is dependent on specific hydrodynamic and wind conditions responsible for the transportation of larvae. According to the nursery size hypothesis (Rijnsdorp et al. 1992), restricted habitat requirements results in a relatively low surface area that can support the 0-group population, which restricts the upper limits of recruitment, and may thus also explain why turbot is rarer compared to flounder displaying a more general distribution. Moreover, as shown in paper IV, turbot may be food limited some years possibly affecting stage duration and thus mortality rate due to size selective mortality. However, the existence of size selective mortality needs further investigation since there seems to be a mismatch between the peak abundance of brown shrimp and 0-group flatfish in the central Baltic Sea (A. Nissling, unpublished). Possibly, food limitation may also affect mortality rate during winter due to low fat reserves (Anderson 1988) or susceptibility to predation by other species than brown shrimp when leaving the nursery ground.

Further, as shown in figure 5, the recruitment of turbot and flounder off the coast of Gotland display a significant positive correlation (Spearman, r = 0.95, n = 10, p <0.01). Recruitment estimates are based on catches by the Swedish National Board of Fisheries incorporating females of 3-9 years and 3-12 years of turbot and flounder, respectively (Florin et al. 2011). This

23 relationship suggests that inter-specific interactions between turbot and flounder affecting recruitment are not likely, which contradicts the observation by Molander (1964) where strong year classes of turbot seemed to occur at weak year classes of flounder.

The co-variation of turbot and flounder recruitment suggests that variation in recruitment of the species is affected by the same factors; potentially during the egg or larval stage. Salinity may affect the year class strength by influencing fertilization rates and viable hatch for the demersal flounder as well as the survival probabilities in relation to oxygen levels for the pelagic flounder (Nissling et al. 2002, 2006). Temperature could also have an effect as egg survival for turbot have been found to be lower at 9 and 21 °C compared to 12-18 °C (Nissling et al. 2006). Thus, up-welling occasions (Myrberg and Andrejev 2003), or unusual cold or warm springs could potentially lead to low recruitment in some years. Similarly, a negative correlation was reported between year class strength of plaice and the prevailing temperature at time of spawning (Van der Veer 1986). Also temperature can have an effect on the composition of zooplankton species, which may affect feeding success of the larvae and thus also survival (Alheit et al. 2005).

Figure 5. The relationship between flounder and turbot relative recruitment in central Baltic Sea. Data obtained from Florin et al . (2011).

24 Abiotic factors such as winds and currents affecting transportation of larvae to the coast (Riley et al. 1981, Van der Veer et al. 1998) could also be a common factor affecting the level of recruitment for turbot and flounder. But, the conditions needed for successful transportation of larvae originating from the flounder with pelagic eggs, spawning in the deep basins far from the coast, may differ compared to the flounder with demersal eggs and turbot, which spawn in the vicinity of the nursery grounds. No data is, however, available of the location of the larvae for any of the species.

Further, an apparent positive relationship was revealed for both species when plotting the density of 0-group fish during peak abundance (obtained in paper II) and the relative recruitment of the species. However, the relationship could not be tested statistically as only four data points are available. Such a correlation would suggest that factors acting during the period before peak abundance determine the level of recruitment, i.e. factors determining the amount of settling larvae as well as mortality from settlement until peak abundance.

25 Final Remarks

At Gotland, central Baltic Sea, 0-group turbot and flounder share shallow sandy bays as nursery grounds during summer and autumn. Flounder, however, is also relatively abundant at other habitats. Thus, turbot is more restricted in terms of habitats, which potentially reflects its more specialized diet. This suggests that turbot is more vulnerable to habitat degradation reducing the number of available habitats. The observed habitat preferences of the species could be used for the protection and restoration of essential habitats, although management could benefit from further evaluations of the nursery role of different types of habitats in respect to growth, mortality and connectivity to the adult habitat. The restricted requirements of 0-group turbot could potentially infer variability in recruitment during the juvenile stage in addition to events during the egg- and larval stage. To elucidate the existence of density dependent mortality intense sampling of 0-group individuals at settling and at the end of the season should be performed. Moreover, the apparent positive relationship between 0-group density at peak abundance and recruitment suggest that continued monitoring of juveniles as well as adults should be conducted as to establish a significant relationship that could be used to predict the recruitment several years in advance. Increased knowledge of the larval biology of these species is also likely to help understanding the causes of their recruitment variation in the Baltic Sea.

26 Acknowledgements

“No man is an island” John Donne (1572-1631)

It is impossible to overcome great challenges in life without the help and support from others. In the case of the work as a Ph.d. student vital contribu- tions can range between anything from scientific discussions to feeling the warmth from the one next beside you in the morning. Below I acknowledge all people and institutions that have played a significant role in the process of realizing this thesis:

First, I would like to thank Magnus Appelberg , my supervisor throughout the main part of the thesis, for valuable discussions, advice and comments. Further, my assistant supervisor, Anders Nissling , has been of tremendous help and support in my everyday work at Ar Research Station. I also acknowledge Sture Hansson for his role as supervisor, even though for a brief period of time, and advice on bioenergetic models as well as his overall scientific input during the entire process. I also want to thank Göran Sundblad and Olle Hjerne for advice, discussions and support. I sincerely thank Michael Burrows for help on the random walk model and Ben Bolk- er as well as Safi Ahlberg for advice on programming in R. Safi, without your help, I would still be stuck with all for loops. I thank Ulf Bergström and Anders Sjögren for help with analyzes of field samples. I further want to thank all people involved in the field and laboratory work at Ar Research Station that all have contributed by hard work and great company; Bertil Widbom , Rebecca Retz , Marie Jacobsson , Jens Bardtrum , Manuela de Gouveia , Örjan Jacobsson , Daniel Nygren , Nina Hallberg , Siw Jansson , John Sällebrant , Peter Sandqvist and Emma Westerlund , and once again last but not least Anders Nissling . Thank you all colleagues at the former department of biology at Gotland University for your support. Also, thanks to all Ph.d. students at the department of Systems ecology and elsewhere that have enriched my work by advice, discussions and just plain company, I acknowledge in particular Ida Ahlbeck , Ben Jaeschke , Antonia Nyström Sandman , Anders Wallin, Josefin Sundin and Malin Jonell . My family, specially my son, Samuel Sandgren , and parents Lars-Martin Martinsson and Eva-Marie Elingmark , has also played a crucial role in my work by great support and role as refuge enabling requisition of energy. This as well

27 as great assistance with analyzes of field samples also holds true for Emily Sahlsten being next to me for a major part of the period. Also, thank you Henk van der Jeugd for raising my interest for ecological modeling. The thesis was financially supported by Gotland University , the DBW society and the Lars Hierta Memorial Foundation .

28 References

Aarnio, K., Bonsdorff, E. and Rosenback, N., 1996. Food and feeding habits of juvenile flounder Platichthys flesus (L.), and turbot Scophthalmus maximus L. in the åland archipelago, northern Baltic Sea. Journal of Sea Research 36:311-320. Abella, A., Fiorentino, F., Mannini, A. and Orsirelini, L., 2008. Exploring relationships between recruitment of European hake ( Merluccius merluccius L. 1758) and environmental factors in the Ligurian Sea and the Strait of Sicily (Central Mediterranean). Journal of Marine Systems 71:279-293. Alheit, J., Mollmann, C., Dutz, J., Kornilovs, G., Loewe, P., Mohrholz, V. and Wasmund, N., 2005. Synchronous ecological regime shifts in the central Baltic and the North Sea in the late 1980s. ICES Journal of Marine Science 62:1205- 1215. Anderson, J. T., 1988. A review of size dependent survival during pre-recruit stages of fishes in relation to recruitment. Journal of Northwest Atlantic fishery science 8:55-66. Anderson, M. J., 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32-46. Bagge, O., 1981. Demersal fishes. In A. Voipio, editor. The Baltic Sea. Elsevier Oceano-graphic Series No. 30. Elsevier Scientific Company, Amsterdam. pp. 320-327. Bagge, O., 1987. Tagging of Turbot and Brill in the Kattegat 1965-70. ICES C.M. 1987/G:10. Bailey, K. M., 1994. Predation on juvenile flatfish and recruitment variability. Neth- erlands Journal of Sea Research 32:175-189. Bartolino, V., Colloca, F., Sartor, P. and Ardizzone, G., 2008. Modelling recruitment dynamics of hake, Merluccius merluccius , in the central Mediterranean in relation to key environmental variables. Fisheries Research 92:277-288. Begon, M., Harper, J. L. and Townsend, C. R., 1990. Ecology: individuals populations, and communities, 2nd edition. Blackwell Scientific Publications, Massa- chusetts. Bonsdorff, E., and Blomqvist, E. M., 1993. Biotic couplings on shallow water soft bottoms - examples from the northern Baltic Sea. Oceanography and Marine Biology: An Annual Review 31:153-176. Burrows, M. T., Gibson, R. N., Robb, L. and Comely, C. A., 1994. Temporal patterns of movement in juvenile flatfishes and their predators: underwater televi- sion observations. Journal of Experimental Marine Biology and Ecology 177:251-268. Burrows, M. T., Gibson, R., Robb, L. and Maclean, A., 2004. Alongshore dispersal and site fidelity of juvenile plaice from tagging and transplants. Journal of Fish Biology 44:620-634. Cederwall, H., and Elmgren, R., 1990. Biological effects of eutrophication in the baltic Sea, particularly the coastal zone. Ambio 19:109-112.

29 Chipps, S.R. and Wahl, D.H., 2004. Development and evaluation of a western mos- quitofish bioenergetics model. Transactions of the American Fisheries Society 133: 1150–1162. Cushing, D. H., 1975. Marine Ecology and Fisheries. Cambridge University Press, Cambridge. Draganik, B., Maksimov, Y., Ivanov, S. and Psuty-Lipska, I., 2005. The status of the turbot Psetta maxima (L.) stock supporting the Baltic fishery. Bulletin of the sea fisheries institute 164:23-53. Florin, A.-B., 2005. Flatfishes in the Baltic Sea - a review of biology and fishery with a focus on Swedish conditions. Finfo 2005:14, Swedish Board of Fisheries, Öregrund. Florin, A.-B., Bergström, U., Ustups, D., Lundström, K., Nissling, A. and Jonsson, P., 2011. Uppföljning av fredningsområdet vid Gotska Sandön 2006-2010. Finfo 2011:8, Swedish Board of Fisheries, Öregrund. Florin, A.-B., and Lavados G., 2010. Feeding habits of juvenile flatfish in relation to habitat characteristics in the Baltic Sea. Estuarine, Coastal and Shelf Science 86:607-612. Florin, A.-B., Sundblad, G. and Bergström, U., 2009. Characterisation of juvenile flatfish habitats in the Baltic Sea. Estuarine, Coastal and Shelf Science 82:294- 300. Ford, R. G., 1983. Home Range in a Patchy Environment: Optimal Foraging Predic- tions. American Zoologist 23:315-326. Gibson, R. N., 1994. Impact of habitat quality and quantity on the recruitment of juvenile flatfishes. Journal of Sea Research 32:191-206. Gibson, R. N., 1997. Behaviour and the distribution of flatfishes. Journal of Sea Research 37:241-256. Gibson, R., Pihl, L., Burrows, M., Modin, J., Wennhage, H. and Nickell, L., 1998. Diel movements of juvenile plaice Pleuronectes platessa in relation to predators, competitors, food availability and abiotic factors on a microtidal nursery ground. Marine Ecology Progress Series 165:145-159. Gulland, J. A., 1965. Survival of the youngest stages of fish and its relation to year- class strenght. International Commission for the Northwest Atlantic Fisheries 6:365-371. Gärdenfors, U. 2005. The 2005 Red list of Swedish species. ArtDatabanken, Uppsa- la. Hansen, M. Boisclair, J., D., Brandt, S. B., Hewett, S. W., Kitchell, J. F., Lucas, M. C. and Ney, J. J., 1993. Applications of bioenergetics models to fish ecology and management: Where do we go from here? Transactions of the American Fi- sheries Society 122:1019-1030. Hinrichsen, H., 2001. Testing the larval drift hypothesis in the Baltic Sea: retention versus dispersion caused by wind-driven circulation. ICES Journal of Marine Science 58:973-984. Hinrichsen, H. H., Bottcher, U., Köster, F. W., Lehmann, A. and St.John, M. A., 2003. Modelling the influences of atmospheric forcing conditions on Baltic cod early life stages: distribution and drift. Journal of Sea Research 49:187-201. Hjort, J., 1914. Fluctuations in the great fisheries of Northern Europe viewed in the light of biological research. Rapports et Procés-Verbaux des Réunions du Con- seil International pour l’Exploration de la Mer 20:1-228. Houde, E. D., 1987. Early life dynamics and recruitment variability. American fisheries society symposium 2:17-29. ICES., 2011a. ICES WGBFAS Report, book 8. pp 80-84. ICES. 2011b. ICES WGBFAS Report, book 8. pp 59-69.

30 Iles, T. C. and Beverton, R. J. H., 1998. Stock, recruitment and moderating processes in flatfish. Journal of Sea Research 39:41-55. Iles, T. C. and Beverton, R. J. H., 2000. The concentration hypothesis: the statistical evidence. ICES Journal of Marine Science 57:216-227. Imsland, A. K., Sunde, L. M., Folkvord, A. and Stefansson, S. O., 1996. The interac- tion of temperature and fish size on growth of juvenile turbot. Journal of Fish Biology 49:926-940. Jennings, S., Kaiser, M. J. and Reynolds, J. D., 2001. Marine Fisheries Ecology. Blackwell Science, Oxford. Köster, F. W., Mollmann, C., Hinrichsen, H.-H., Wieland, K., Tomkiewicz, J., Kraus, G., Voss, R., Makarchouk, A., MacKenzie, B. R., St. John, M. A., Schnack, D., Rohlf, N., Linkowski, T. and Beyer, J. E., 2005. Baltic cod recruitment - the impact of climate variability on key processes. ICES J. Mar. Sci. 62:1408-1425. Legget, W. C. and DeBlois, E., 1994. Recruitment in marine fishes: is it regulated by starvation and predation in the egg and larval stages? Netherlands Journal of Sea Research 32:119–134. Leggett, W. C. and Frank, K. T., 1997. A comparative analysis of recruitment variability in North Atlantic flatfishes - testing the species range hypothesis. Journal of Sea Research 37. Lindsey, J. K., 1997. Generalized linear modelling. In G. Casella, S. Fienberg, and I. Olkin, editors. Applying Generalized Linear Models. Springer, Secaucus, NJ. pp 1-26. Luo, J., Brandt and S.B., 1993. Bay anchovy Anchoa mitchilli production and consumption in mid-Chesapeake Bay based on a bioenergetics model and acoustic measures of fish abundance. Marine ecology progress series 98: 223-236. MacLean, J. A., Shuter, B. J., Regier, H. A. and MacLeod, J. C., 1981. Temperature and year-class strenght of smallmouth bass. ICES Rapports et Procés_Verbaux des Réunions du Conceil International pour l’Exploration de la Mer 178:30-40. Mazur, M.M., Wilson, M.T., Dougherty, A.B., Buchheister, A. and Beauchamp, D.A., 2007. Temperature and prey quality effects on growth of juvenile walleye pollock Theragra chalcogramma (Pallas): a spatially explicit bioenergetics approach. Journal of fish biology 70: 816–836. McArdle, B. H. and Anderson, M. J., 2001. Fitting Multivariate Models to Commu- nity Data: A Comment on Distance-Based Redundancy Analysis. Ecology 82:290-297. The Ecological Society of America. Miller, J. M., Burke, J. S. and Fitzhugh, G. R., 1991. Early life history patterns of Atlantic North American flatfish: likely and (unlikely) factors controlling recruitment. Netherlands Journal of Sea Research 27:261-275. Mittelbach, G., 1986. Predator-mediated habitat use: some consequences for species interactions. Environmental Biology of Fishes 16 (1-3): 159-169. Modin, J. and Pihl, L., 1996. Small-scale distribution of juvenile plaice and flounder in relation to predatory shrimp in a shallow Swedish bay. Journal of Fish Biolo- gy 49:1070-1085. Molander, A. R., 1964. Underordning plattfiskar. In K. A. Andersson, editor. Fiskar och fiske i norden. Natur och kultur, Stockholm. pp 120-122. Morales-Nin, B., 2000. Review of the growth regulation processes of otolith daily increment formation. Fisheries Research 46: 53-67. Myrberg, K. and Andrejev, O., 2003. Main upwelling regions in the Baltic Sea — a statistical analysis based on three-dimensional modeling. Boreal Environment Research 8:97-112.

31 Nash, R. D. M. and Geffen, A. J., 2000. The influence of nursery ground processes in the determination of year-class strength in juvenile plaice Pleuronectes platessa L. in Port Erin Bay, Irish Sea. Journal of Sea Research 44:101-110. Nash, R. D. M, Dickeycollas, M. and Kell L., 2009. Stock and recruitment in North Sea herring ( Clupea harengus ); compensation and depensation in the population dynamics. Fisheries Research 95:88-97. Nissling, A., 1995. Salinity and oxygen requirements for successful spawning of Baltic cod, Gadus morhua . Doctoral thesis, Stockholm University, Stockholm. Nissling, A. and Dahlman, G., 2010. Fecundity of flounder, Pleuronectes flesus, in the Baltic Sea - reproductive strategies in two sympatric populations. Journal of Sea Research 64:190-198. Nissling, A., Jacobsson, M. and Hallberg, N., 2007. Feeding ecology of juvenile turbot Scophthalmus maximus and flounder Pleuronectes flesus at Gotland, Central Baltic Sea. Journal of fish biology 70:1877-1897. Nissling, A., Johansson, U. and Jacobsson, M., 2006. Effects of salinity and temperature conditions on the reproductive success of turbot ( Scophthalmus maximus ) in the Baltic Sea. Fisheries Research 80:230-238. Nissling, A., Florin, A.-B., Thorsen, A. and Bergström, U., in prep. Egg production of turbot ( Scophthalmus maximus ) in the Baltic Sea. Nissling, A., Westin, L. and Hjerne, O., 2002. Reproductive success in relation to salinity for three flatfish species, dab ( Limanda limanda ), plaice ( Pleuronectes platessa ), and flounder ( Pleuronectes flesus ), in the brackish water Baltic Sea. ICES Journal of Marine Science 59:93-108. Ojaveer, E., Kaleis, M., Aps, R., Lablaika, I. and Vitins, M., 1985. The impact of recent environmental changes on the main commercial fish stocks in the Gulf of Finland. Finnish Fisheries Research 6:1-14. Pihl, L., 1990. Year-class strength regulation in plaice (Pleuronectes platessa L.) on the Swedish west coast. Hydrobiologia 195:79-88. Power, M.E., 1984. Depth distribution of armored catfish: Predator-induced resource avoidance? Ecology 65: 523-528. Rijnsdorp, A. D., van Beek, F. A., Flatman, S., Millner, R. M. , Riley, J. D., Giret, M. and de Clerck, R., 1992. Recruitment of Sole stocks, Solea solea (L.), in the northeast atlantic. Netherlands Journal of Sea Research 29:173-192. Riley, J. D., Symonds, D. J. and Woolner, L., 1981. On the factors influencing the distribution of 0-group demersal fish in coastal waters. Rapports et Procés- Verbaux des Réunions du Conseil International pour l’Exploration de la Mer 178:223-228. Rothschild, B. J., 2000. “Fish stocks and recruitment”: the past thirty years. ICES Journal of Marine Science 57:191-201. Sinclair, M., 1988. Marine populations: an Essay on population Regulation and Speciation. University of Washington, Seattle. Stevens, M., Maes, J. and Ollevier, F., 2006. A bioenergetics model for juvenile flounder Platichthys flesus . Journal of applied ichthyology 22: 79-84. Tabor, R., Luecke, C. and Wurtsbaugh, W.A., 1996. Effects of daphnia availability on growth and food consumption of rainbow trout in two Utah reservoirs. North american journal of fisheries management 16: 591-599. Temming, A., 1989. Long-term changes in stock abundance of the common dab (Limanda limanda L.) in the Baltic Proper. Rapports et Procés-Verbaux des Réunions du Conseil International pour l’Exploration de la Mer 190:39-50. Van der Veer, H., 1986. Immigration, settlement, and density-dependent mortality of a larval and early postlarval 0-group plaice (Pleuronectes platessa) population in the western Wadden Sea. Marine Ecology Progress Series 29:223-236.

32 Van der Veer, H. W., Berghahn, R., Miller, J. M. and Rijnsdorp, A. D., 2000. Re- cruitment in flatfish, with special emphasis on North Atlantic species: Progress made by the Flatfish Symposia. ICES Journal of Marine Science 57:202-215. Van der Veer, H. W. and Bergman, M. J. N., 1987. Predation by crustaceans on a newly settled 0-group plaice Pleuronectes platessa populations in the western Wadden Sea. Marine ecology progress series 35:203-215. Van der Veer, H. W., Berghahn, R. and Rijnsdorp, A. D., 1994. Impact of juvenile growth on recruitment in flatfish. Netherlands Journal of Sea Research 32:153- 173. Van der Veer, H. W., Ruardij, P., Van den Berg, A. J. and Ridderinkhof, H., 1998. Impact of interannual variability in hydrodynamic circulation on egg and larval transport of plaice Pleuronectes platessa L. in the southern North Sea. Journal of Sea Research 39:29-40. Voipio, A., 1981. The Baltic Sea. Elsevier Ocenanographic Series No. 30. Elsevier Scientific Company, Amsterdam. Werner, E.E., Hall, D.J., 1988. Ontogenetic habitat shifts in bluegill: The foraging rate-predation risk trade – off. Ecology 69 (5): 1352-1366. Zheng, J., 1996. Herring stock-recruitment relationships and recruitment patterns in the North Atlantic and Northeast Pacific oceans. Fisheries Research 26:257-277. Zuur, A. F., Ieno, E. N. and Smith, G. M., 2007. Analysing ecological data. M. Gail, J. Krickeberg, J. Samet, A. Tsiatis, and W. Wong, Editors. Springer, New York.