Why
are
fish
in
the
Baltic
Sea
so
small?
A
study
of
somatic
and
gonad
growth
in
relation
to
salinity
in
turbot
(Scophthalmus
maximus)
Isa
Wallin
Degree project inbiology, Master ofscience (2years), 2014 Examensarbete ibiologi 30 hp tillmasterexamen, 2014 Biology Education Centre
Supervisor: Anders Nissling
1
ABSTRACT
2
CONTENTS
1. INTRODUCTION ...3
1.1 Fish growth ...3
1.2 Fish growth in relation to salinity ...3
1.3 Reproduction ...5
1.4 Aims ...6
2. MATERIAL AND METHODS ...7
2.1 Study species...7
2.2 Growth experiment ...9
2.2.1 Experiment design and implementation ...9
2.2.2 Statistical analysis of growth experiment ... 11
2.3 Literature review ... 12
2.3.1 Procedure of data standardization and calculations ... 13
2.3.2 Calculation of energy content ... 15
2.3.3 Statistical treatment of literature review ... 16
3
1. INTRODUCTION
1.1 Fish growth
Different fish species have evolved a diversity of life history patterns in response to environmental factors. There are differences, both between and within species, in traits such as timing of sexual maturity, reproductive strategies and life span, which all affect fish growth. Separate populations within the same species may display different life-history patterns due to differently constituted environments (Wootton 1990), as variations in life history traits are expected to be associated with significant variation in fitness (Hutchings 1997).
The somatic growth pattern of a fish is a combination of genetic factors, i.e. the potential for growth that depends on the genotype, and the environmental factors the fish encounters (Wootton 1990). The genetic factors include nervous, endocrine and neuroendocrinological mechanisms and systems. The environmental factors can be divided into two groups: abiotic and biotic. Abiotic factors are determining, for example temperature, salinity and photoperiod, or limiting, which operate above, for example ammonia, or below, for example oxygen, a specific threshold or within a tolerance range, such as pH (Bœuf & Payan 2001). Also biotic factors, such as availability and quality of food, social interactions and reproduction, influence fish growth (Wootton 1990).
1.2 Fish growth in relation to salinity
For fish species living in an environment with fluctuating salinity, energy loss is associated with osmoregulation and ionic regulation, which implies that less energy may be allocated to, for instance, growth (Wootton 1990 and references therein). Salinity may act as an inhibiting factor concerning growth in fish, and such an effect would be expressed as a decrease in growth efficiency at a given food availability. At some salinities, the extra energy cost for osmoregulation may reduce the energy amount available for growth or reproduction, if the fish cannot compensate for the loss by an increased feeding rate. Salinity seems, however, to have limited effect on food consumption, which is mainly regulated by temperature (Wootton 1990). Marine teleost fish drink water, partly to compensate for the water loss through diffusion via the gills. It is not clear to what extent fish in the Baltic Sea drink water, but it is reasonable to believe that they do drink water, although probably less than fish in fully marine environments as the osmotic challenges in such areas should be greater. It is not established whether or not there is a link between drinking mechanisms and growth, but it cannot be ruled out since drinking has an energetic cost. It is also possible that the water concentration in the stomach has an impact on digestive enzymes, which may affect food processing and as a result possibly affect growth rates in fish (Bœuf & Payan 2001).
4
Nissling 1991). In theory, inhabiting intermediate salinities may therefore result in a lower standard metabolism, which might in turn enhance growth. This is a consequence of that the energetic cost of osmoregulation increases with alteration of the salinity away from iso-osmotic conditions. The energetic cost of osmoregulation can be expected to be lower in an iso-osmotic medium, where the differences between the water and the fish’s internal environment are minimal (Bœuf & Payan 2001). For example, for the euryhaline cichlid
Oreochromis niloticus, the metabolic rate is higher at salinities of 0, 7.5 and 22.5 ‰ than at
11.6 ‰. This implies that the salinity of 11.6 ‰ is close to iso-osmotic with the blood plasma, meaning that the energetic costs of osmoregulation should be lower (Wootton 1990 and references therein). As a result, the energy savings from living in an iso-osmotic medium may be substantial enough to make a contribution to growth. The estimated level of energy used for such metabolic activity is debated: previous studies claim that between 20 and > 50 % of a fish’s energy budget may be allocated to osmoregulation, while more recent studies suggest that it may only be approximately 10 % (reviewed in Bœuf & Payan 2001).
Bœuf & Payan (2001) found that juveniles and adults of many different fish species, both of marine and limnetic origin, show increased growth at intermediate salinities, i.e. between 8-20 ‰. In larval sea bream (Sparus aurata), growth was studied at salinities between 15 and 40 ‰. It was concluded that the highest growth rates were achieved at 25 ‰ (Tandler et al. 1995). In summer flounder (Paralichthys dentatus), optimum salinities for early development and larval growth was shown to be between 8 and 14 ‰, and it was suggested that the larvae grow and develop better in environments more similar to their internal environment (Specker
et al. 1999). However, Moustakas et al. (2004) showed that larvae of southern flounder
(Paralichthys lethostigma) grew significantly better at 34 than 25 ‰. On the other hand, Gutt (1985) found that juvenile flounder (Pleuronectes flesus) displayed higher growth rates at 5 and 15 ‰ salinity than at 0 and 35 ‰ salinity. It was also concluded that growth at 15 ‰ was more uniform. The reason for the lower growth rates at low salinities was suggested to be due to lower food intake, and decreased growth at higher salinity was supposed to be the result of inferior food conversion (Gutt 1985). Lambert et al. (1994) showed that food conversion rates and growth rates in Atlantic cod (Gadus morhua) were higher at intermediate salinities of 14 ‰, compared to 7 and 28 ‰, respectively. However, the higher growth rates were not found to correlate to neither an increase in food consumption, nor changes in diet composition or relative allocation of energy to tissues or gonads.
For freshwater species such as carp (Cyprinus carpio), white amur (Ctenopharyngodon idella) and juvenile Russian sturgeon (Acipenser guldenstaedti), an increase in salinity to 2 ‰ significantly increased the growth rate, which was suggested to be the result of a more efficient food conversion (Bœuf & Payan 2001 and references therein). On the other hand, pike (Esox lucius) reared in fresh water grew significantly better compared to pike reared in brackish water. The authors suggested this to be the result of the pike not yet being fully adapted to the osmotic conditions in brackish water (Engström-Öst et al. 2005).
5
conversion efficiency. Dutil et al. (1997) concluded that salinities of 14 ‰ resulted in high growth rate in cod, but that the food assimilation was not related to salinity. Also turbot (Scophthalmus maximus) displays higher growth rates at intermediate salinities, and it has been shown that routine metabolic rate is at its minimum at isotonic salinities (between 8-10 ‰; Gaumet et al. 1995, Bœuf et al. 1999). Imsland et al. (2001) reared juvenile turbots in salinities of 15, 25 and 33.5 ‰ and in temperatures between 10 and 22°C during a three month experiment. They found that optimum growth in juvenile turbot occurs at 18.5 ± 0.8 ‰, and optimum food conversion efficiency at 19.0 ± 1.0 ‰, indicating that turbot growth can be improved when reared at intermediate salinities.
The fact that most species show optimum growth between 5 and 18 ‰ implies that estuaries, tidal coastal areas and coastal systems with brackish water (8-16 ‰) should provide suitable conditions for increased growth (Bœuf & Payan 2001).
1.3 Reproduction
Most fish species show indeterminate growth patterns, which means that a fish grows throughout its entire life (Jennings et al. 2001). After maturation, the proportion of energy available for reproduction increases at the expense of energy allocated to somatic growth (Rijnsdorp & Witthames 2005). Consequently, a cost of reproduction in fish may manifest itself as reduced growth, a reduction of future fecundity, an increase in the time between consecutive matings or increased mortality (Wootton 1990). Thus, there is a trade-off in energy investment between growth and reproduction.
The amount of resources allocated to ovary growth can, for example, be studied in experiments where food availability is limited (Wootton 1990). E.g. in threespined stickleback (Gasterosteus aculeatus), females fed less food were shown to invest a larger proportion of energy in egg production, compared to females fed more food (Wootton 1977). It was implied that the soma subsidizes egg production, since energy content of the soma decreased significantly when food availability was limited (Wootton 1977).
6
population. Fecundity in 1980 was much higher (Horwood et al. 1986 in Rijnsdorp & Witthames 2005) and close to the fecundity estimates for the years around 1900 (re-analyzed by Rijnsdorp 1991, see Rijnsdorp & Witthames 2005). However, the fecundity of fish in the Baltic Sea appears to support the theory of Bagenal (1966), since fecundity is much higher in this area (Rijnsdorp & Witthames 2005).
Higher fecundity in Baltic Sea populations compared to Atlantic populations has been found in a number of fish species, as described below. Fecundity of flounder, dab (Limanda
limanda) and turbot is higher in the Baltic than in the North Sea, and this is also true for
herring (Clupea harengus) (Bagenal 1966 and references therein). Additionally, it has been reported that plaice from the southern Kattegat display fecundity at intermediate levels, between the Baltic and North Sea (Bagenal 1966 and references therein).
Nissling & Dahlman (2010) concluded that there are two different populations of flounder in the Baltic Sea: one population with pelagic eggs, spawning off-shore at approximately 10-20 ‰, and another population with demersal eggs spawning close to the coast at around 6-7 ‰. The latter population displays significantly higher fecundity. Nissling & Dahlman (2010) suggested that differences in egg survival in relation to salinity between the populations result in a trade-off between growth and reproduction.
It has been found that low egg and larval survival is associated with low salinity (Nissling et
al. 2002 and references therein), such as the salinities prevailing in the Baltic Sea (Bernes
2005). Adaptations concerning egg buoyancy and sperm motility have been promoted in Baltic Sea populations of dab, plaice and flounder in order to cope with the difficulties of egg and larval survival in low salinities (Nissling et al. 2002). Despite such adjustments, fish egg fertilization and larval survival in the Baltic Sea may still be adversely affected by the low salinities (Nissling et al. 2002 and references therein).
The specific reproductive features, such as egg buoyancy and sperm motility, of Baltic Sea fish populations (Bagenal 1966, Nissling et al. 2002, Rijnsdorp & Witthames 2005, Nissling
et al. 2006, Nissling & Dahlman 2010) suggest that high fecundity compensates for
difficulties of reproduction related to salinity. It is therefore possible that low somatic growth of Baltic Sea fish is a consequence of high fecundity, and not the result of inferior growth rates.
1.4 Aims
7
My overall hypothesis was that the lesser size of turbot in the Baltic is an indirect effect of salinity related to poor egg survival compensated by higher energy investment in gonads and thus consequently lower energy investment in somatic growth.
To test this hypothesis, I predicted that no difference in growth rates between the juveniles reared in salinities of 6, 10.5, 15 and 30 ‰ would be detected. In the present study, 10.5 ‰ represents an isotonic environment (Fletcher 1978, Mangor-Jensen 1987, Westin & Nissling 1991), at which level the energetic costs of osmoregulation should be lower (Wootton 1990) and growth, in theory, higher. The salinities of 6 and 15 ‰ are considerably closer to 10.5 ‰ than is 30 ‰, and theoretically, the osmoregulatory levels should be lower at these salinities compared to at 30 ‰. As a consequence, the fish reared in 6, 10.5 and 15 ‰ were not expected to display lower growth rates compared to the fish reared in 30 ‰. Consequently, the Baltic Sea turbot was not expected to display any direct negative effects, i.e. inferior growth, resulting from low salinity.
Also, I predicted that turbot from the Baltic Sea would invest similar amounts of energy annually, compared to fish from populations outside the Baltic Sea, when both somatic and gonad energy investment was considered.
2. MATERIAL AND METHODS
2.1 Study species
The turbot (Scophthalmus maximus) is found in marine and brackish waters in the Northeast Atlantic, throughout the Mediterranean and along the European coasts to the Arctic Circle; it also occurs in most of the Baltic Sea (FishBase n.d.) where it is found in coastal waters from Skagerrak up to the Sea of Åland (Florin 2005). It is also found in the Black Sea, and opinions are divided if that population should be considered a subspecies, Psetta maxima maeotica (FishBase n.d.), or not (Zengin et al. 2006).
8
No consensus values of the lengths at which 50 % of the females in a population have become sexually mature (L50) have been found in the literature for the different populations
investigated in the present study. According to available data, 50 % of the turbot females have reached sexual maturity at 4.5 years in the North Sea (Jones 1974), while sexual maturity in the Black Sea is reached at between 3-5 years (STECF 2008). In the Baltic Sea, it has been shown that some females mature at the age of 3 years, part of the females have become sexually mature at the age of 4 years and that most turbot females have reached sexual maturity at the age of 5 years (Stankus 2003). If it is assumed that approximately 50 % of the females have become mature at the age of 4 years in all populations reviewed, it is evident that length at maturity differs between populations (Table 1), if age assessments are accurate (see section 4.2).
Table 1. Length at maturity for different populations of turbot.
Mean length (mm) at maturity (4 years)
North Sea (Jones 1974) 422
Black Sea (Zengin et al. 2006) 447
Baltic Sea, EEZ of Lithuania (Stankus 2003) 265 Baltic Sea, Karlskrona (Nissling et al. 2013) 278 Baltic Sea, Gotland (Nissling et al. 2013) 300
Spawning of the Baltic Sea turbot takes place in June-July (Martinsson & Nissling 2011 and references therein) in coastal areas and on off-shore banks (Nissling et al. 2006). Providing data from International Council for the Exploration of the Sea (ICES) subdivision (SD) 28 (mid-Baltic, Fig. 1), Nissling et al. (2006) established that turbot spermatozoa motility, fertilization rates and egg survival was affected by salinity, with significantly higher values of spermatozoa motility, fertilization and hatching success at salinities of ≥ 7 ‰. It was suggested that the reproductive success of turbot is expected to decrease towards the north and eastern parts of the Baltic Sea where salinity is lower, and to vary over years depending on occurrence of saline water inflows (Nissling et al. 2006). Also, the eggs of turbot in the Baltic Sea are demersal (Nissling et al. 2006 and references therein), unlike in the Atlantic and in the Black Sea where turbot eggs are pelagic (Jones 1974, Spectorova et al. 1974). Demersal eggs are more vulnerable to predation than are pelagic eggs (Dahlgren 1979 and references therein), a fact that may also have a detrimental impact on turbot egg survival rates in the Baltic Sea.
9
that are up to half (Aarnio et al. 1996) or even two thirds their own length (Nygren 2009). Both juvenile and adult turbot prefer sand bottoms, but depth distribution varies with age; juveniles are found mostly in shallower waters while adult fish dwell at greater depths (Florin 2005).
Figure 1. ICES subdivisions in the Baltic Sea. The northernmost limit for the Baltic Sea turbot appears to be in SD 29 (modified from Florin 2005).
2.2 Growth experiment
2.2.1 Experiment design and implementation
Experiments took place between August 7th and September 5th 2013. The growth experiment setup consisted of four salinities: 6, 10.5, 15 and 30 ‰, chosen because salinities of 6, 10.5 and 15 ‰ prevail in the Baltic Sea, and 30 ‰ occurs in the Kattegatt and the Skagerrak on the Swedish west coast (Bernes 2005). In this study, 6 ‰ represents a hypotonic environment, while 10.5 ‰ represents an isotonic environment and 15 and 30 ‰ represent hypertonic environments, respectively (Fletcher 1978, Mangor-Jensen 1987, Westin & Nissling 1991). 0-group turbot were captured in Fröjel on the west coast of central Gotland (salinity approximately 7 ‰) on three different occasions: August 5th
, 9th and 13th, using a beach seine. As a result of difficulties to catch enough fish on the same occasion, the experiment had to start in three different sets: on August 7th, 11th and 15th.
10
aquaria (18 x 37 x 26 cm) filled with 15 liters of brackish water from the Baltic Sea (6.7 ‰, temperature approximately 22° C), filtered through a plankton sieve to sift out potential prey items. The aquaria were also provided with airstones and a layer of sand (1-2 cm) allowing the fish to bury themselves, an arrangement used throughout the experiment.
The fish were held in the acclimation aquaria for at least three hours until they were tagged with Visible Implant Elastomer (VIE-tags) (Northwest Marine Technology) to make it possible to discern individual fish in each aquarium. The VIE-tags were injected subcutaneously with a 0.3 mm syringe on the unpigmented side of the body. Post tagging, the fish were released back into the acclimation aquaria to recover from the procedure.
After at least three hours, the fish were transferred to aquaria with salinities of 6 and 10.5 ‰ for adaptation to experimental salinities. From the aquaria of 10.5 ‰ fish were subsequently transferred to aquaria of 15 and 20 ‰, respectively, and later on from 20 ‰ to 30 ‰. The fish were given at least three hours to adapt to each salinity, and were held starving from the time of capturing and during the adaptation phase until the onset of the experiment, which meant a minimum of 27 hours and a maximum of 48 hours of starvation.
Water of salinities 10.5, 15 and 30 ‰ used in the experiment was prepared by filtering Baltic Sea surface water (salinity approximately 6.5 ‰) through a plankton sieve and adding proper amounts of Tropic Marine Sea Salt. To prepare 75 liters of water of 10.5, 15 and 30 ‰, 300, 637.5 and 1762.5 grams of salt were added, respectively. Water of salinity 6 ‰ was prepared by mixing filtered Baltic Sea surface water with filtered freshwater. The prepared water was kept in barrels in the thermal room, where the experiment was performed, during the experiment to ensure that the water temperature was held as constant as possible.
A total of 105 fish were kept in 15 liter plastic aquaria (18 x 37 x 26 cm) in the thermal room throughout the experiment, and the order of the aquaria was produced by a random number generator. 24 fish were kept in five aquaria in 6 ‰, 28 fish were kept in six aquaria in 10.5 ‰ and 27 fish were kept in six aquaria in 15 and 30 ‰, respectively (Table 2). Temperature and salinity in the aquaria was surveyed daily. Water temperature during the experiment ranged between 20-24° C, which is considered optimal for juvenile 0-group turbot (Retz 2011). To ensure even more homogenous temperatures in future experiments, the aquaria could be provided with thermostats or be placed in a thermo constancy chamber. The aquaria should also be provided with lids, which was not the case during the present study, to prevent fish from jumping out of the aquaria. Salinity was adjusted when needed to ensure it was kept constant. The oxygen level was surveyed once during the experiment and found to be 8.3-8.7 ml/l, which is considered to be within the normal range for Baltic Sea surface water (SMHI n.d.). The light regime during the acclimation and experiment phases mimicked natural conditions by simulating night between 8 pm and 5 am, and removal of debris and change of water was done at least every second day.
11
Thus, dividing according to size ensured that the fish were provided with prey of as suitable size as possible. The fish of two size classes were equally abundant.
Table 2. Number of fish in each replicate of each set at the beginning of the experiment, divided in size classes of > 25 mm and < 25 mm (total length). Set 1 started on August 7th, set 2 on August 11th and set 3 on August 15th.
Number of fish, 1 Number of fish, set 2 Number of fish, set 3 Salinity > 25 mm < 25 mm > 25 mm < 25 mm > 25 mm < 25 mm 6 ‰ 5 5 4 4 5 10.5 ‰ 5 5 4 4 6 4 15 ‰ 5 5 4 4 5 4 30 ‰ 5 5 4 4 5 4 Total 105
As a result of varying supply of prey caught in the field, the fish were fed mysid shrimps (Mysidae sp.), juvenile threespined stickleback (Gasterosteus aculeatus aculeatus) and gobies (Pomatoschistus sp.) depending on accessibility. The fish < 25 mm in series 1 were also fed
Bathyporeia pilosa, and the fish < 25 mm in series 3 were, for a few days, fed industrially
reared juvenile mysid shrimps.
Food was provided in excess. The fish were fed 15 prey items in the morning, and in the afternoon additional prey were added to maintain an abundance of approximately 15 prey of suitable size in the aquaria, which was in line with experiment implementations of Martinsson (2011) and Retz (2011). Abundance of prey was kept as constant as possible and was not adjusted to the number of turbots in each aquarium. The ambition was to provide prey of suitable sizes, but the varying supply of prey caught in the field caused problems. Thus, it was not possible to fully adhere to the predetermined feeding regime. The difficulties of rearing juvenile turbot on food of unsuitable sizes can be avoided by buying industrially reared prey during future experiments.
Fish length was measured on day 15, when survival rate amounted to 48 %, i.e. a total of 50 fish. Several factors contributed to the high mortality; inferior condition of individuals at start (series 2), starvation or death from trying to gulp oversized prey due to poor availability of preferred food items, bacterial fin disease and lack of oxygen due to a technical failure on one occasion. The rather small sample size may have been complicating possibilities to detect potential differences between salinities.
2.2.2 Statistical analysis of growth experiment
12
values at the respective salinities), indicating that almost no growth had occurred during the study period. It could not be ruled out that this unusually low growth resulted from insufficient access to food items of preferred sizes (see section 2.2.1). As a consequence, it was possible that salinity was not the primary factor affecting growth in this particular case, and the juvenile was therefore excluded from the analyses. For the second juvenile, reared in 15 ‰, length measurements appeared to be inadequate, and I decided not to include that specific juvenile in my analyses.
The dry weight was used to calculate the instantaneous rate of growth per unit of weight (SGR) (Wootton 1990), with the formula of Houde & Schekter (1981): SGR = (e g – 1) * 100, where g = (ln(W2) – ln(W1))/(t2 – t1), W2 andW1 referring to dry weight on day t2 and t1, i.e.
day 15 and 1. When SGR had been calculated, regression tests were performed for all salinities in Excel (Microsoft Office, version 12.0.4518.1014) to examine whether there was any relationship between original length (converted to dry weight at t1) and SGR during the
experiment.
I went on to test whether SGR differed between the different salinities. Because the regressions indicated that original length affected SGR, a regular ANOVA would not be sufficient and instead I proceeded with a GLM, General Linear Model (univariate ANOVA). I did this in IBM SPSS Statistics software (version 22) with SGR as dependent variable, salinity as fixed factor and original length as covariate. Also, an interaction between salinity and length was included.
Additionally, I wanted to investigate possible differences in growth of fish in the present study with growth of fish in a previous study. In the previous study, the fish were reared under optimum conditions and thus values of what could be considered maximum possible SGR had been obtained (Martinsson 2011). In order to detect and compare possible differences in SGR between the studies, I performed a GLM (univariate ANOVA) with SGR as dependent variable, experiment as fixed factor and original length as covariate.
For the GLM tests, the models were run stepwise with subsequent removal of non-significant (p > 0.05; starting with the highest) variables until only significant predictors were included. To test whether there were any differences in mortality between the salinities up to day 15, a chi-square test (Fowler et al. 1998) was performed.
2.3 Literature review
13 Data for six different populations were collected:
i) Karlskrona in the Hanö Bight (ICES SD 25; Fig. 1) (Nissling et al. 2013) ii) Eastern Gotland (ICES SD 28; Fig. 1) (Nissling et al. 2013)
iii) Gotska Sandön (ICES SD 28; Fig. 1) (Nissling et al. 2013)
iv) The exclusive economic zone (EEZ) of Lithuania (ICES SD 26; Fig. 1) (Stankus 2003)
v) North Sea (Jones 1974)
vi) Southeastern Black Sea (Zengin et al. 2006, Samsun 2004) The first four localities are from within the Baltic Sea.
Data from the different sources needed to be standardized in order to make a comparison of energy content at age of different populations. The different papers presented data differently: for example, fecundity can be estimated in various ways and be presented as both relative fecundity (i.e. number of eggs / g somatic weight) and absolute fecundity (total number of eggs). Also, weight can be specified in different ways and sometimes needed to be recalculated. A particularly important feature was if weight was specified as total weight, gutted weight or somatic weight. Total weight (WT) refers to the weight of the whole fish.
Gutted weight (WG) refers to total weight minus the weight of the gut (WG = WT - weight of
gut) and somatic weight (WS) is the total weight minus the weight of the gut and gonads (WS
= WT - weight of gut and gonads). In cases where it was necessary to achieve estimates of WG
from values of WT, a formula from Bedford et al. (1986) was used: (1/1.07) * w, where w
refers to WT.
2.3.1 Procedure of data standardization and calculations
For each population I set out to derive somatic and gonad weight for each age class. To do this, I used a series of calculations using formulae that were based on the specific information given for each population (see below). The somatic and gonad weights were then used to calculate energy (calorific) content in the respective tissues.
Two of the populations in the Baltic Sea, Eastern Gotland and Gotska Sandön, did not differ in fecundity (Nissling et al. 2013), compared to the population of Karlskrona. The populations of Eastern Gotland and Gotska Sandön were therefore combined and treated as one single population, hereafter referred to as Gotland. I calculated WS at age using the equations WS =
0.259a0.521 for Karlskrona and WS = 0.181a0.647 for Gotland, respectively, where a = age
(Anders Nissling, pers. comm.).
For both the populations of Karlskrona and Gotland, values of relative fecundity were given (Nissling et al. 2013). According to Nissling et al. (2013), relative fecundity is 2343 eggs/g WS for Karlskrona, and 3426 eggs/g WS for Gotland. Mean egg wet weight was calculated
14
600 – 18.058). Because egg weights were not given for the other populations, the egg weight estimated here (49.98 µg) was used to estimate gonad weight for the other populations (see below). Relative fecundity and WS were used to calculate absolute fecundity at age (absolute
fecundity (total number of eggs) = relative fecundity * WS). Gonad weight at age for the
populations Karlskrona and Gotland was calculated by multiplying egg weight with absolute fecundity at age. Age classes included in the analysis were 4-10 for both Karlskrona and Gotland.
For the EEZ of Lithuania, average turbot weight (assumed to be WT) at age and relative
fecundity were given (Stankus 2003). To calculate WS, WG was estimated using the formula
(1/1.07) * w, where w is WT (Bedford et al. 1986). Because WS is the same as WG minus
gonad weight, I could then calculate WS by deducting the estimated gonad weight from the
WG. To estimate gonad weight, absolute fecundity at age was calculated by multiplying
relative fecundity (2034 eggs/g total weight; Stankus 2003) by WT at age. Gonad weight at
age was then estimated by multiplying absolute fecundity at age by the egg weight (49.98 µg; Anders Nissling, pers. comm.). In this manner, somatic and gonad weight was calculated for each age class within the 3-15 year range. I chose to include these age classes because 3 is the age when the first females are estimated to reach sexual maturity (Stankus 2003).
For the North Sea turbot population, formulae for deriving length at age, weight in relation to length and for fecundity at age were given (Jones 1974). The references to weights in Jones (1974) only mentioned WG, and therefore I assumed that the weights obtained from the
formula for length/weight relationship also referred to WG. In order to estimate WS, I
calculated length at age according to the von Bertalanffy growth equation (lt = L ͚ (1-e-K(t-t0));
von Bertalanffy 1938) with parameters given in the article (Jones 1974). Then, I obtained WG
at age using a formula for length/weight relationships (lnW = -3.012 + 2.769 lnL). WS was
then obtained by withdrawing gonad weight from the values of WG. To obtain gonad weight
at age, absolute fecundity at age was estimated by using a formula given in the article: (1.0293 * W) – 0.02873, where W = weight. Gonad weight at age was calculated by multiplying absolute fecundity at age with egg weight (49.98 µg; Anders Nissling, pers. comm.) as described above. Age classes 3-15 were included in the analysis, chosen because 3 is the age when the first females reach sexual maturity (Jones 1974).
For the Black Sea turbot population, formulae for deriving length at age and weight in relation to length were given (Zengin et al. 2006), and relative fecundity was specified (1241 eggs/g body weight; Samsun 2004). Since no distinctions were made in the articles, I assumed that weights referred to WT. I derived length at age according to the von Bertalanffy growth
equation (lt = L ͚ (1-e-K(t-t0)); von Bertalanffy 1938), with parameters provided by Zengin et al.
(2006). Then, I obtained WT at age by using an equation for length/weight relationships, W =
aTLb, where W = weight in grams, TL = total length and regression parameter values for a and b were presented in the article (Zengin et al. 2006). WG was estimated using the formula
(1/1.07) * w, where w is WT (Bedford et al. 1986). In order to achieve WS, I withdrew gonad
weight from the values of WG. To calculate gonad weight at age I firstly estimated absolute
15
at age with mean egg weight (49.98 µg; Anders Nissling, pers. comm.). Egg weight based on values from the Baltic Sea turbot was a potential source of error for the North and Black Sea populations, having pelagic eggs (Jones 1974, Spectorova et al. 1974). In theory, the pelagic eggs of the North and Black Sea populations should be lighter in comparison to the demersal eggs of the Baltic Sea turbot. The length and weight parameters presented in Zengin et al. (2006) were based on age classes 1-7. To limit sources of errors that may occur from extrapolation, only values of age classes 3-7 were considered.
Above, all calculations of fecundity were based on potential fecundity, i.e. the number of oocytes in the ovary prior to spawning. The values of potential fecundity may be compromised by the finding that a rather large portion of eggs may be aborted during spawning, a process known as atresia. The overall condition of the fish affects the level of atresia and as a result, realized fecundity may differ significantly from potential fecundity in turbot. Down-regulation of fecundity is also found to vary slightly between populations within the Baltic Sea (A. Nissling et al., unpublished observations). The fact that varying amounts of oocytes may be aborted could to some degree impact the results of this study if the differences in atresia levels between populations would be extensive.
2.3.2 Calculation of energy content
In order to achieve values of total energy content at age in turbot, data on somatic and gonad tissue were combined using data from L. Grahn (unpublished observations), based on four turbot females caught outside Askö in the Baltic Sea (ICES SD 27). The data was presented as total fish weight with gonads removed (hereafter called body weight), as gonad weight and as calorific content (kcal) of the body weight and the gonad weight respectively.
To obtain the amount of kcal/g WS, I divided the body calorific content (kcal) with body
weight (g) for each of the four females. Subsequently, I calculated the mean energy content/g WS, based on the four females altogether. When calculating the calorific content of 1 g of
gonad tissue, data for calorific content of gonads was divided with gonad wet weight for each of the four specimens, and a mean value was derived. It was established that mean calorific content in turbot somatic tissue is 1.13 kcal/g, and 1.59 kcal/g in gonad tissue.
To validate the values of somatic calorific content, I performed an additional calculation, based on data for plaice given in Rijnsdorp & Ibelings (1989). The values obtained were rather similar compared to those based on Grahns data; the energy content of plaice was found to be 1.29 kcal/g WS. Since the values of Rijnsdorp & Ibelings (1989) and Grahn were
similar, I decided to proceed using the values of Grahn.
16
2.3.3 Statistical treatment of literature review
Total energy content was plotted against age for the respective populations. Since I detected tendencies of non-linear correlations, I performed an ln-transformation of age data as suggested in Sokal & Rohlf (1981) to optimize requirements for regression (on average r2 = 0.983 ± 0.016). To test whether the populations differed in total energy investment, I used a GLM (univariate ANOVA) with total energy as dependent variable, population as fixed factor and ln age as covariate.
To elucidate possible differences in growth rates between populations, I analyzed percent growth rate (PGR) in relation to energy content, used as an equivalent to size. This analysis was performed in order to overcome effects of anomalies in energy content at age among populations, potentially the result of differences in age-reading methods. With this analysis, it was possible to disengage the values of age and solely focus on PGR in relation to energy content. Plotting of PGR against energy content suggested ln-transformation of energy-data (Sokal & Rohlf 1981; on average r2 = 0.983 ± 0.012). A GLM (univariate ANOVA) was then used with PGR as dependent variable, population as fixed factor and ln energy as covariate to evaluate discrepancies in growth between populations.
PGR = Energy year tx+1 − energy year txEnergy year tx ∗ 100
For the GLM tests, the models were run stepwise with subsequent removal of non-significant (p > 0.05; starting with the highest) variables until only significant predictors were included.
3. RESULTS
3.1 Growth experiment
17
Figure 2. SGR, instantaneous growth rate per unit of weight, in relation to original length (mm), apportioned on salinity.
Table 3. SGR, instantaneous growth rate per unit of weight, in relation to fish size (mm) at different incubation salinities: tests of between-subjects effects.
df F p
Intercept 1 47.683 < 0.0001
Salinity 3 0.426 > 0.05
Original length 1 8.439 < 0.05
Salinity*original length 3 0.436 > 0.05
Table 4. Values of SGR at the respective salinities ± standard deviation in the growth experiment.
Incubation salinity, ‰ Mean SGR
6 15.33 ± 2.79
10.5 13.03 ± 2.78
15 12.68 ± 2.34
30 13.83 ± 2.54
According to the GLM investigating potential discrepancies in SGR between the present study and the previous experiment performed by Martinsson (2011), where maximum SGR of turbot juveniles was examined, there were no differences in SGR (Table 5). This implies that the juveniles of the present study displayed optimum growth at all salinities.
18
Table 5. Comparison between SGR at different salinities of the present study and maximum SGR (Martinsson 2011). Incubation salinity, ‰ df F p 6 1 0.335 > 0.05 10.5 1 1.921 > 0.05 15 1 2.201 > 0.05 30 1 0.805 > 0.05
Additionally, I found no difference in fish mortality between the different salinities (Chi square = 6.75, df = 3, p = 0.081)
3.2 Literature review
A compilation of average somatic and gonad energy investment for the different populations showed an increasing investment in gonads along with decreasing salinities (Table 6). Salinities presented for the North and Black Seas are the salinities in the open water, where the eggs of these populations develop, while salinities presented for the Baltic Sea populations are the salinities in coastal waters, where the demersal eggs of the Baltic Sea turbot populations develop.
Table 6. Energy investment in somatic and gonad tissue in percent for the different populations.
Approximate salinity, ‰ Somatic tissue Gonad tissue
North Sea (Jones 1974) 35 93.02 6.98
Black Sea (Zengin et al. 2006 & Samsun 2004) 18 90.97 9.03
Baltic Sea, EEZ of Lithuania (Stankus 2003) 7-8 85.44 14.56
Baltic Sea, Karlskrona (Nissling et al. 2013) 7-8 85.95 14.05
Baltic Sea, Gotland (Nissling et al. 2013) 6-7 80.71 19.29
19
Figure 3. Energy content (kcal) in relation to age for different populations. Note that energy contents at age are similar for populations Karlskrona and Gotland, respectively.
Table 7. Energy content (kcal) in relation to age: tests of between-subjects effects for all populations.
df F p
Intercept 1 152,89 < 0.0001 Population 4 22,31 < 0.0001
Age 1 723,78 < 0.0001
Population*age 4 78,1 < 0.0001
Table 8. Energy content (kcal) in relation to age: tests of between-subjects effects of populations of the North and Black Seas.
df F p
Intercept 1 133.65 < 0.0001 Population 1 13.55 < 0.05
Age 1 515.29 < 0.0001
Population*age 1 20.57 < 0.0001
Table 9. Energy content (kcal) in relation to age: tests of between-subjects effects of populations in the Baltic Sea. df F p Intercept 1 33.52 < 0.0001 Population 2 42.6 < 0.0001 Age 1 188.73 < 0.0001 Population*age 2 69.56 < 0.0001 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 0 0,5 1 1,5 2 2,5 3 En e r gy (k c al ) Year, ln transformed North Sea Black Sea
20
In the GLM with PGR, percent growth rate, as dependent variable (Fig. 4), a significant effect of energy content was revealed, showing that initial size affects PGR. It was also established that there was a non-significant interaction between population*energy (Table 10), i.e. there was no difference in PGR with increased energy content among populations. However, a small but significant effect of population occurred (Table 10). The population effect was mainly due to differences between Karlskrona and Gotland vs. the other three populations. When testing Karlskrona and Gotland on one hand and the populations from the Baltic Sea (data from the EEZ of Lithuania), the North Sea and the Black Sea on the other hand, there was no effect neither of interaction between population*energy nor of populations (Tables 11, 12). The non-significant interaction between population and energy content implies that the change in PGR in relation to energy content is the same, regardless of population. This indicates that fish from different populations display proportional changes in PGR, and that turbot from different populations allocate equal amounts of energy annually.
Figure 4. PGR in relation to energy content (kcal) for different populations. Note that PGR is similar for populations Karlskrona and Gotland, respectively.
Table 10. PGR in relation to energy content (kcal) for all populations: tests of between-subjects effects.
df F p Intercept 1 220.29 < 0.0001 Population 4 3.26 < 0.05 Energy content 1 73.48 < 0.0001 Population*energy content 4 2.133 > 0.05 0 20 40 60 80 100 120 140 4 5 6 7 8 9 P G R
Energy content (kcal), ln transformed
North Sea Black Sea
21
Table 11. PGR in relation to energy content (kcal) for populations from the Baltic Sea (data from the EEZ of Lithuania), the North Sea and the Black Sea.
df F p
Intercept 1 370.22 < 0.0001
Population 2 0.144 > 0.05
Energy 1 303.96 < 0.0001
Population*energy 2 0.388 > 0.05
Table 12. PGR in relation to energy content (kcal) for populations Karlskrona and Gotland from the Baltic Sea.
df F p Intercept 1 718.2 < 0.0001 Population 1 0.143 > 0.05 Energy 1 609.32 < 0.0001 Population*energy 1 0.064 > 0.05
4. DISCUSSION
4.1 Growth experimentThe results described above indicated that turbot juveniles grow equally well in all salinities investigated, i.e. salinities of 6, 10.5, 15 and 30 ‰ (Fig. 2, Table 3). No differences in growth rates were found between the present study and a previous study of maximum SGR (Martinsson 2011). Additionally, no differences in mortality between salinities were detected in the present study.
22
The data of the growth experiment in comparison to the data of the experiment of Martinsson (2011) show that the fish in the present study grew equally well, since no differences in SGR were detected (Table 5). N.B. the fish in Martinsson’s experiment were supposed to show maximum growth. As mentioned above, the ambition was to provide prey of suitable sizes, but the varying supply of prey caught in the field was problematic. The fish in the present study may, as a result, have been offered prey items of non-optimal sizes. Thus, it appears that if prey of suitable sizes was offered, growth was virtually optimal in all salinities.
Additionally, mortality did not differ between salinities. This is in line with the findings of Smith et al. (1999), establishing that southern flounder showed no differences in survival rates at salinities ranging between 5-30 ‰ at three weeks post-metamorphosis, while fish held at 0 ‰ displayed a statistically lower survival (20 %).
Because neither mortality nor growth differed between salinities, I concluded that the surviving juveniles did not experience any direct effects of salinity during the experiment.
4.2 Literature review
The compilation of the means of somatic and gonad energy investment for the different populations showed an increasing investment in gonad tissue along with decreasing salinities (Table 6).
Fecundity appears to be lower in populations having pelagic eggs (Nissling & Dahlman 2010). In the present study, the values of energy investment in gonad tissue were very low in populations of the North and Black Seas. It is assumed that conditions for egg and larval development are more favorable in the North Sea in comparison to the Baltic Sea with respect to salinity, because the eggs can maintain buoyancy in waters of higher salinities. The ability to remain buoyant and therefore avoid the low oxygen levels at greater depths is essential for the development of pelagic eggs in the Baltic Sea (Nissling et al. 2002 and references therein). In Baltic Sea flounder it has been shown that off-shore spawning flounder, spawning in waters of higher salinity, produce fewer eggs because its eggs can remain buoyant. In comparison coastal-spawning flounder with demersal eggs, spawning in low salinity waters, produce more eggs. It is believed that the implications for reproduction, i.e. high egg mortality, in low salinity need to be compensated for by producing more eggs (Nissling & Dahlman 2010). Also, there are indications that the structure of the cell wall of turbot eggs does not develop properly when cell cleavage takes place in low salinities (Karås & Klingsheim 1997). These factors indicate that additional energy needs to be invested in reproduction in Baltic Sea turbot as a result of inferior environmental conditions for eggs and larvae.
23
present literature study (Table 6), despite the potentially low egg and larval survival in the Black Sea. A possible explanation may be that the Black Sea data of relative fecundity is based mostly on 4- and 5-year old females (Samsun 2004), and then applied to weight at age derived from Zengin et al. (2006). Since young females produce fewer eggs than older ones (Wootton 1990), the age of the females is likely to affect the values of relative fecundity of Black Sea turbot used in the present study. The data of relative fecundity (Samsun 2004) is therefore not fully applicable to the data source of Black Sea turbot weight at age used in this study (Zengin et al. 2006). Potentially, a more evenly distributed sample of females, age-wise (Samsun 2004), would have yielded higher values of relative fecundity for the Black Sea turbot and thus shown a higher gonad energy investment in the compilation (Table 6).
The statement made by Bagenal (1966) that fish from the Baltic Sea invest more energy in reproduction, compared to their conspecifics in the Atlantic and the Black Sea, is supported by the numbers given in Table 6, where energy investment in gonad tissue increases with decreasing salinity compared to somatic tissue. According to my compilation (Table 6) and Nissling et al. (2013), it is obvious that fish from the Baltic Sea, and particularly from the northern population around Gotland, invest more energy in gonad growth compared to other populations. Thus, the results are in line with the theory of Bagenal (1966), stating that energy investment in gonad tissue increases towards the edges of a species’ distribution range. In the analysis of energy content in relation to age, it was established that there was a significant interaction between population and age, indicating that growth (increase in energy) differ between populations, and also a significant effect of population indicated that the origin of the fish affected its energy content at age. Further, the significant difference in intercept between populations implies that growth of individuals from different populations differs from the very beginning. Up until time of sexual maturity, all the available energy is invested in somatic growth. The difference in intercept is probably the result of a difference in timing of sexual maturation between populations.
Size at sexual maturity differs between populations (Table 1) and there are three possible explanations for this. Firstly, a potentially important factor affecting timing of sexual maturity is fishing pressure. A decrease in L50 may be the response to a situation where fish, which
invest large amounts of energy in somatic growth early in life and therefore start reproducing late in life, are caught before they have had the chance to spawn. In such a situation, fish which are genetically adapted to reproduce early in life are favored. For example, L50 has
24
Thus, it is highly likely that features of growth and sexual maturity of flatfish in general and turbot in particular have changed as a result of fishing pressure. This could affect the results from this analysis. For instance, it is likely that the premises have changed since Jones (1974) presented data collected in the late 1960’s. Possible effects are that the North Sea turbot in the 1960’s would have been bigger at the time of sexual maturity than they would have been if the data was sampled today, or the opposite, that Baltic Sea turbot of today are smaller than they were 40 years ago. This is definitely possible, because fishing pressure on Baltic Sea turbot has been intense (HELCOM 2013a). If sexual maturity at age has decreased in the Baltic Sea as well as in the Atlantic, it has done so without having been monitored since there are no known data from the 1960’s for the Baltic Sea. If this is the case, the data from the North Sea is not directly comparable to the more recently sampled data from the Baltic Sea (Stankus 2003, Nissling et al. 2013) or the Black Sea (Zengin et al. 2006, Samsun 2004). In conclusion, this would imply that turbot representing the North Sea in this study are too large in relation to their age, which complicates the comparison.
25
populations, and differences in migratory patterns between those areas, which could impact the length of the feeding season. Additionally, differences in time of hatching between populations may prevail, which could potentially affect the length of the first growth season. A third explanation to the discrepancies of size at sexual maturity is differences in otolith reading for age estimation. Age assessments can be performed by reading otoliths that are whole, burnt or sectioned, and discrepancies in age reading are both the results of different age reading methods and of differences in reading habits between readers. The methods for otolith readings are the same for flounder and turbot, and it has been established that methods of age reading in flounder vary significantly among the countries around the Baltic Sea (ICES 2007). Also for turbot discrepancies between readers prevail, mostly about the interpretation of the first growth ring (ICES 2008). For example, the logical outcome will be that length and weight would increase more slowly if a large number of growth rings are detected, compared to if fewer rings are found. In support of this argument is the difference between the turbot of the EEZ of Lithuania (Stankus 2003) and Karlskrona (Nissling et al. 2013) (Fig. 3). The two data sets represent turbot from the southern Baltic Sea (ICES SD 25 and 26) inhabiting waters of similar salinity. In theory, the turbot of the EEZ of Lithuania and Karlskrona should belong to the same population, and consequently it would be fair to assume that the fish would contain equal amounts of energy at the same age. However, as this is not the case, it is believed that differences in age reading methods are behind this discrepancy. As mentioned above, age assessments can be performed by reading whole, burnt or sectioned otoliths. The method commonly used in Lithuania is reading of whole otoliths, while Swedish researchers use sectioned otoliths. The method of using sectioned otoliths has now been recommended, as it yields the highest agreement among readers compared to other methods (ICES 2007). Since age reading methods differ between Sweden and Lithuania, it is plausible that the significant effects viewed in Table 9 have emerged from differences in age assessment. Additionally, because I have reason to believe that there are discrepancies in age reading between the fish of the EEZ of Lithuania and Karlskrona, there may be differences prevailing between populations of the North and Black Seas versus the Baltic Sea as well.
26
The analysis of PGR in relation to energy content (Table 10) was performed in order to partly avoid the difficulties related to age reading by substituting age with energy content (size). The age reading-related issues discussed above in the analysis of energy content in relation to age may be of importance for the difference in intercept in this analysis as well. If, for example, the first growth ring is neglected by some readers, the outcome will be that an assumed 1-year old fish is actually 2 years old, and the fish is considered to be larger from the very beginning. It is very likely that the difference in intercept between the fish from the EEZ of Lithuania and Karlskrona is the result of inconsistent age-reading in this analysis as well as in the analysis of energy content in relation age described above.
The significant value for population, when testing all populations against each other, suggested that the origin of the fish affects its PGR. However, when analyzing Karlskrona and Gotland together, and the other three populations were tested as a separate group, it became evident that there was no effect of population (Tables 11, 12). In fact, no detectable difference in PGR was found between populations of Karlskrona and Gotland when they were tested solely against each other (Table 12), which could probably be ascribed to the fact that age reading was performed in exactly the same way for these two populations. If a large number of growth rings are found in otoliths, which the method used by Swedish researchers implies, a lower growth rate between years will be the expected outcome (ICES 2007). It also supports the assumption that the difference in intercept between Karlskrona and the data from the EEZ of Lithuania (Stankus 2003) is the result of inconsistent age-reading. Similar to this, no effect of population occurred when comparing the North Sea, Black Sea and the EEZ of Lithuania (Stankus 2003), which indicates that age readings are more equal for these populations (see ICES 2007 for comparison of between whole vs. burnt otoliths).
27
4.3 Hypotheses evaluation
I wished to answer the questions 1) To what extent does somatic growth in turbot differ between juveniles reared in hypotonic, isotonic and hypertonic environments? and 2) To what extent does energy allocation concerning somatic and gonad growth in turbot differ between individuals inhabiting different salinity conditions?
My overall hypothesis was that the lesser sizes of turbot in the Baltic Sea was an indirect effect of salinity related to poor egg survival compensated by higher energy investment in gonads.
The prediction that no difference in growth would be detected between juveniles reared in salinities of 6, 10.5, 15 and 30 ‰ has received further support, since it was established that the juveniles in this experiment grew equally well in all salinities. This prediction has to be retained, since no effect of salinity or interaction was detected. It is concluded that ambient water of salinities 6, 10.5, 15 and 30 ‰ has no impact on somatic growth for turbot juveniles from Gotland. In the light of these findings, I find it unlikely that the osmoregulatory costs at different salinities are substantial enough to have an impact on turbot juvenile growth. Instead, it is possible that differences in energy content between populations arise later in life, when the turbot females become sexually mature. Then, the Baltic Sea turbot probably “switch mode” from investing basically all available energy in somatic growth to investing great amounts of energy into gonad growth, and as a result, differences in somatic energy content appear between populations.
Even though energy at age varied with population, it was shown that equal amounts of energy were invested between populations when total energy amounts at age were analyzed in the test of PGR in relation to energy content. Therefore, the prediction that turbot from the Baltic Sea annually invest total amounts of energy similar to those of turbot from populations outside the Baltic Sea has to be retained.
My overall hypothesis, that the lesser size of turbot in the Baltic Sea is an indirect effect of salinity related to poor egg survival compensated by higher energy investment in gonads, appears to have received further support.
28
4.4 Further research
Considering the results of the growth experiment, where the juveniles displayed similar growth in 6 ‰ compared to in 10.5, 15 and 30 ‰, it would be interesting to perform a similar experiment but with turbot from the west coast of Sweden. Such an experiment could elucidate genetic adaptations to the turbot’s environments and their possible impact on turbot juvenile growth.
In order to refine the results of the literature review and minimize the sources of error concerning fishing pressure, new data based on the contemporary turbot population of the North Sea would be highly utilizable. Also, further investigations of factors affecting the difference in size at sexual maturity need to be conducted in order to provide a hopefully seamless explanation to why fish in the Baltic Sea are smaller than fish outside the Baltic Sea.
29
5. ACKNOWLEDGEMENTS
I would like to thank my supervisor Anders Nissling for guidance, feedback and patience during my work on this manuscript. I would also like to thank Jennie Ljungberg, Manuela de Gouveia, Bertil Widbom and Karl Kihlberg for valuable assistance in the field. My sincere thanks also go to Jesper Martinsson for help with my fieldwork as well as access to data from previous experiments, and to Ann-Britt Florin for help with age-reading related issues.
6. REFERENCES
Aarnio K, Bonsdorff E & Rosenback N. 1996. Food and feeding habits of juvenile flounder, Platichtys
flesus, (L.) and turbot, Scophthalmus maximus (L.) in the Åland archipelago, northern Baltic Sea. J. Sea Res., 36: 311-320.
Bagenal TB. 1966. The ecological and geographical aspects of the fecundity of the plaice. J. Mar.
Biol. Assoc. UK, 46: 161-186.
Bedford BC, Woolner LE & Jones BW. 1986. Length-weight relationships for commercial fish species and conversion factors for various presentations. Fisheries Research Data Report No. 10. Lowestoft.
Bernes C. 2005. Förändringar under ytan – Sveriges havsmiljö granskad på djupet. Fälth & Hässler, Värnamo.
Black Sea Commission. 2009. Marine Litter Report. http://www.blacksea-commission.org/_publ-ML-CH1.asp, visited December 6th 2013.
Bœuf G, Boujard D & Person-Le Ruyet J. 1999. Control of the somatic growth in turbot. J. Fish Biol.,
55: 128-147.
Bœuf G & Payan P. 2001. How should salinity influence fish growth? Comp. Biochem. Phys. C., 130: 411-423.
Dahlgren MD. 1979. A review of survival rates of fish eggs and larvae in relation to impact assessments. Mar. Fish. Rev., 41: 1-12.
Dutil JD, Lambert Y & Boucher E. 1997. Does higher growth rate in Atlantic cod (Gadus morhua) at low salinity result from lower standard metabolic rate or increased protein digestibility? Can. J.
Fish. Aquat. Sci., 54: 99-103.
Engström-Öst J, Lehtiniemi M, Jónasdóttir SH & Viitasalo M. 2005. Growth of pike larvae (Esox
lucius) under different conditions of food quality and salinity. Ecol. Freshw. Fish, 14: 385-393.
FishBase. No date. Scophthalmus maximus (Linnaeus, 1758) – Turbot. http://fishbase.org/Summary/SpeciesSummary.php?ID=1348&AT=turbot. Downloaded 2013-08-25.
Fletcher CR. 1978. Osmotic and ionic regulation in the cod (Gadus callarias L.). J. Comp. Physiol.,
124: 149-155.
Florin A-B. 2005. Flatfishes in the Baltic Sea – a review of biology and fisheries with a focus on Swedish conditions. Finfo: 2005:14.
Florin A-B & Höglund J. 2007. Absence of population structure of turbot (Psetta maxima) in the Baltic Sea. Mol. Ecol., 16: 115-126.
Fowler J, Cohen L & Jarvis P. 1998. Practical Statistics for Field Biology. 2nd ed. John Wiley and Sons Ltd, Chichester, UK.
Gaumet F, Bœuf G, Severe A, Le Roux A & Mayer-Gostan N. 1995. Effects of salinity on the ionic balance and growth of juvenile turbot. J. Fish Biol., 47: 865-876.
Gutt J. 1985. The growth of juvenile flounders Platichthys flesus L. at salinities of 0, 5, 15 and 35‰.
J. Appl. Ichthyol., 1: 17-26.
HELCOM 2013a. Helsinki Commission. HELCOM Red List Species Information Sheets (SIS) Fish. http://helcom.fi/Documents/Ministerial2013/Associated%20documents/Background/HELCOM%2 0RedList%20All%20SIS_Fish.pdf, downloaded November 13th 2013.
30
Houde ED & Schekter RC. 1981. Growth rates, rations and cohort consumption of marine fish larvae in relation to prey concentrations. Rapp. P.-V. Re´un.-Cons. Int. Explor. Mer., 178: 441-453. Hutchings JA. 1997. Life history responses to environmental variability in early life. In: R. C.
Chambers and E. A. Trippel (ed.). Early Life History and Recruitment in Fish Populations, pp. 139-168. Chapman and Hall, London.
ICES 2007. International Council for the Exploration of the Sea. Report of the Workshop on Age Reading of Flounder (WKARFLO), 20–23 March 2007, Öregrund, Sweden.
ICES 2008. International Council for the Exploration of the Sea. Report of the Workshop on Age Reading of Turbot (WKART), 24–27 June 2008, Oostende, Belgium.
ICES 2012. International Council for the Exploration of the Sea. Report of the Inter-Benchmark Protocol on New Species (Turbot and Sea bass; IBPNew 2012), 1–5 October 2012, Copenhagen, Denmark.
Imsland AK, Foss A, Gunnarsson S, Berntssen MHG, FitzGerald R, Bonga SW, von Ham E, Nævdal G & Stefansson SO. 2001. The interaction of temperature and salinity on growth and food conversion in juvenile turbot (Scophthalmus maximus). Aquaculture, 198: 353-367.
Jennings S, Kaiser MJ & Reynolds JD. 2001. Marine fisheries ecology. Blackwell Science, Oxford. Jones, A. 1974. Sexual maturity, fecundity and growth of the turbot Scophthalmus maximus L. J. mar.
biol. Ass. U.K., 54: 109-125.
Karås P & Klingsheim V. 1997. Effects of temperature and salinity on embryonic development of turbot (Scophthalmus maximus L.) from the North Sea, and comparisons with Baltic populations.
Helgoländer meeresun., 51: 241-247.
Lambert Y, Dutil JD & Munro J. 1994. Effects of intermediate and low salinity conditions on growth rate and food conversion of Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci., 51: 1569-1576. Mangor-Jensen A. 1987. Water balance in developing eggs of the cod Gadus morhua L. Fish Physiol.
Biochem., 3: 17-24.
Martinsson J. 2011. Ecology of juvenile turbot and flounder in the Central Baltic Sea: Implications for recruitment. Diss. Dept. of Systems Ecology, Stockholm University.
Martinsson J & Nissling A. 2011. Nursery area utilization by turbot (Psetta maxima) and flounder (Platichthys flesus) at Gotland, central Baltic Sea. Boreal Env. Res., 16: 60-70.
Moustakas CT, Watanabe WO & Copeland KA. 2004. Combined effects of photoperiod and salinity on growth, survival, and osmoregulatory ability of larval southern flounder Paralichthys
lethostigma. Aquaculture, 229: 159-179.
NE 2014. Svarta havet. http://www.ne.se/lang/svarta-havet. Downloaded 2014-03-14.
Nielsen EE, Nielsen PH, Meldrup D & Hansen MM. 2004. Genetic population structure of turbot (Scophthalmus maximus L.) supports the presence of multiple hybrid zones for marine fishes in the transition zone between the Baltic Sea and the North Sea. Mol. Ecol., 13: 585-595.
Nissling A, Westin L & 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 J. Mar. Sci., 59: 93-108.
Nissling A, Johansson U & Jacobsson M. 2006. Effects of salinity and temperature conditions on the reproductive success of turbot (Scophthalmus maximus) in the Baltic Sea. Fish. Res., 80: 230-238. 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.
Nissling A & Dahlman G. 2010. Fecundity of flounder, Pleuronectes flesus, in the Baltic Sea – Reproductive strategies in two sympatric populations. J. Sea Res., 64: 190-198.
Nissling A, Florin A-B, Thorsen A & Bergström U. 2013. Egg production of turbot, Scophthalmus
maximus, in the Baltic Sea, J. Sea Res., 84: 77-86.
Niţă V, Diaconescu Ş, Zaharia T, Maximov V, Nicolae C & Micu D. 2008. The characterization of the main habitat types populated by the Black Sea Turbot in its different stages of development. AACL
Bioflux, 4: 552-570.
