3. Introduction

During their phylogenesis, all species became adapted to their distinct habitats by means of complex, long-term,self-regulation processes which enable them to react to naturally-occurring environmental changes to maintain their physiological and ecological balance. All traits and adaptive mechanisms of recent species evolved along these historical pathways driven by interacting changes of proximate and ultimate environmental factors. If these changes take place within phylogenetic short periods, the evolved adaptive mechanisms to maintain the homeostasis can fail and furthermore new mechanisms of adaptation can often not develop in an appropriate time.

Anthropogenic activity can extend the range of natural stressors, such as increasing UV-irradiation caused by substances that deplete the ozone layer or secondary effects of man-made eutrophication leading to high concentrations ofalgal toxins, e.g., cyanotoxins. Furthermore anthropogenic activities rapidly produce, and introduce, new substances such as xenobiotic chemicals into the environment that may act as stressors.

Biological systems respond to chemical stresses on various levels of aggregation, from molecule to organisms and subsequently to populations and biocoenosis. Within the hierarchy of biological organisation the behaviour is one toxicological endpoint reflecting whole organism-level effects. Behaviour is the product of the interaction(s) of an organism with its external environment. It represents the integration of underlying physiological processes with the environmental stimuli that trigger them and the evolutionary forces that have and continue to shape them (Grue et al., 2002). Adaptations of behaviour to changes in the physical or social environment are common in the animal world, either as short-term or as long-term modifications in behavioural or physiology properties (Hofmann, 2003). The resulting behavioural plasticity is the ability of a single genotype to produce more than one alternative, potentially adaptive, behaviour in response to environmental conditions (West-Eberhard, 1989). This way behaviour turns into a major regulative mechanism to overcome exogenous disturbances and stabilise the endogenous milieu (Tembrock, 1987). Activities such as food acquisition, predator avoidance, prey capture, migration, and habitat preference are critical to the survival of the organism and thus the population, and commonly used as indicators of environmental stressors (Little, 2002).

External stressful factors may influence internal physiological and biochemical processes resulting in behavioural changes. Generic pathways of stressor induced behavioural changes are shown in Figure 1.

Behavioural reactions can occur if the contaminants directly impair physiological and biochemical processes, internal sensors assess a status of inherent disorder and reactive physiological processes initialise the effectors. If the contaminants are detectable by sense organs they can lead to preference-avoidance behaviour that can significantly influence the level or duration of exposure. The impact of natural and anthropogenic stressors on the avoidance behaviour of aquatic organisms was reported e.g. by Beitinger et al. (1990), Richardson et al. (2001) and Hölker and Stief (2005).

Since behaviour is the outcome of many complex developmental and physiological processes, it should provide a more comprehensive measure than one or a few biochemical or physiological parameters (Warner et al., 1966; Zala and Penn, 2004).

	Fig.1 The integrative role of behaviour. Behaviour is influenced by a variety of abiotic, biotic and internal factors.

A broad range of chemicals can act as behavioural toxicants through different modes of actions those are compiled in a classification scheme (Table 1) by Barron (2002). This scheme should be viewed as a generalisation of the behavioural effects and is not intended to substitute chemical-specific and species-specific evaluations of behavioural alterations. Effects of naturally occurring chemical stressors (like cyanobacteria toxins) and combined effects of pollutants are not considered therein.

Tab. 1. Classification scheme for behavioural toxicants compiled by Barron (2002).

In evaluating the impact of stressors in freshwater systems, fish have a special importance because they are situated at the end of the aquatic food chain, and thus may also indicate a contamination with persistent pollutants at lower trophic levels. Fish are immersed in their physical and chemical environment, and therefore, in continuous interactions with potential environmental pollutants.

Pollutants may alter a wide variety of fish behaviours, including e.g., the sexual and reproductive behaviour, the schooling behaviour, the avoidance/preference behaviour, the chemosensory communication and the social behaviour.

Furthermore it was shown that the swimming performance of fish is affected by a range of chemical stressors including e.g., metals, CNS seizure agents, ChE inhibitors and endocrine disrupters (e.g., Spieler et al., 1977; Thomas and Rice, 1987; Reide and Siegmund, 1989; Boujard and Leatherland, 1992; Al-Kahlem et al., 1994; Steinberg et al., 1995; Saglio et al., 1996; Paul and Simonin, 1996; Spieser et al., 2000; Campbell et al., 2002; Schmidt et al., 2004).

Because all behavioural reactions of organisms are essentially coupled with natural processes and physiological reactions that mainly have rhythmic components even behavioural parameters normally occur in rhythmic structures. Biological rhythms are mainly classified according to the length of the period of oscillation (τ). The rhythms whose period of oscillation is 24 ± 4 hours are defined as circadian (τ ~24h). Rhythms with a period of less than 20 hours or more than 28 hours are defined respectively as ultradian and infradian.

Internal mechanisms of self-sustaining oscillators which generate biological rhythms of organisms (e.g., at the gene level in individual cells) are called biological clocks. A defining feature of clocks is that they can synchronise themselves using environmental time signals called time triggers (zeitgeber). Circadian clocks can be entrained by any zeitgeber that varies during a day; the dominant and therefore, physiologically most important one comes from the environmental light cycle (e.g., Aschoff et al., 1972; Pando et al., 2001; Kobayashi et al., 2003; Dekens et al., 2003).

The vertebrate clock is based on a complex hierarchy consisting of a small number of specialised central and multiple peripheral pacemakers (Schibler and Sassone-Corsi, 2002). The zebrafish Danio rerio is a model organism in which the molecular mechanisms of the vertebrate circadian clock have been investigated. The zebrafish pineal gland contains both the circadian pacemaker that drives rhythms of melatonin synthesis, as well as photoreceptive molecules responsible for the entrainment of the clock phase (Cahill, 1996).

In contrast to mammals, isolated organs and cell lines from zebrafish exhibit circadian oscillations in clock gene expression that can be entrained to a 24 hours light/dark cycle (Foster et al., 2003), showing that endogenous oscillators exist in peripheral organs (Whitmore et al., 2000, Foster et al., 2003), e.g., in heart and kidney (Pando et al., 2001; Iigo et al., 2003), and in the retina (Cahill, 1996; Rajendran et al., 1996). Most peripheral cells of zebrafish (Danio rerio) appear to contain photoreceptors, that respond to light fluctuations in the environment by regulating oscillations in the gene expressions (Kobayashi et al., 2003). Using a cell line that derives from zebrafish embryos, differential light-dependent gene activation for several central clock components was shown by Pando et al. (2001).

Circadian clocks are involved in the persistence of circadian rhythmicity after transfer of fish from light cycles to constant conditions, these so called free running rhythms of activity were observed for several fish species including Danio rerio and Leucaspius delineatus (Hurd et al., 1998; Siegmund and Wolff, 1972). Rhythms of circadian period are evident in the locomotor activity of a number of fish species (e.g., Siegmund, 1981; Kadri et al., 1991; Siegmund and Biermann, 1992; Sims et al., 1993; Sanchez-Vazquez and Tabata, 1998; Plaut, 2000; Campbell et al., 2002). The analysis of rhythmicity allows an unspecific indication of contaminants whereby changes of the circadian rhythm in swimming activity of fish were shown for substances of different chemical categories e.g. copper, lead, formaldehyde, chlorine and chloramphenicol (Spieler et al., 1977; Steele, 1989; Reide and Siegmund, 1989; Campbell et al., 2002).

For analysing pollutant-induced effects on fish behaviour, the cyanobacteria toxin microcystin-LR (MC-LR) and a trichlorobiphenyl (PCB 28) were investigated at sublethal levels. MC-LR was chosen as an example for a naturally occurring toxin, whereas PCB 28 was chosen as a typical xenobiotic chemical and in this respect both of them were compared. They are widespread in the aquatic environment, but there is rather few knowledge about their impact on fish behaviour.

Despite the direct impact by pollutants from human usage, such as heavy metals or persistent organic pollutants, eutrophication is still the major cause of indirect deterioration of freshwater systems (Chorus, 2001). Especially under eutrophic and hypertrophic conditions cyanobacteria are known for mass developments and blooms worldwide (Paerl 1996).

Many cyanobacterial strains are known to produce a range of toxins (Codd and Poon, 1988; Carmichael, 1992), functionally classified into the acute lethal poisonous hepatotoxins and neurotoxins and the less lethal cytotoxins (Dow and Swoboda, 2000).

Microcystins are widespread cyanobacterial hepatotoxins that are produced by some cyanobacterial genera, e.g., Microcystis, Anabaena, Oscillatoria, Nostoc, and Anabaenopsis.

	Fig.2 Chemical structure of microcystin-LR.

Microcystins exist in more than 60 variants (Codd, 2000), and toxic strains of cyanobacteria usually produce a mixture of different microcystins (Rinehart et al., 1994; Sivonen and Jones, 1999). In this study microcystin-LR (MC-LR) was investigated which is one of the most frequently found microcystins.

Structurally, microcystins are monocyclic heptapeptides with the general structure of cyclo (-D-Ala-X-D-erythromethylaspartic acid -L-Z-Adda-D-Glu-N methylde-hydroalanine-); X and Z are variable amino acids, for example microcystin-LR contains leucine (L) and arginine (R) (Fig. 2). Adda is the β-C20 amino acid (2S, 3S, 8S, 9S)-3-amino-9-methoxy-2,6,8-trimethyl-10-phenyldeca-4,6-dienoic acid. The key component for biological activity appears to be linked with the Adda side chain, as cleavage of the ADDA side chain from the cyclic peptide renders both components non-toxic (Carmichael, 1992). Microcystin-LR has a molecular weight of about 1000 Daltons (WHO, 1998) and the octanol/water partitioning coefficients (log K_ow) of MC-LR lies at 2.16 (Ward and Codd, 1999).

Microcystins were released into the water body mainly through the lysis of cyanobacterial cells (Watanabe et al., 1992; Sivonen and Jones, 1999).

Generally, microcystin concentrations well in excess of the WHO guideline of 1 µg l^-1 (WHO, 1998) would be expected at sites where toxic cyanobacteria accumulate, rather than in the open water (Welker et al., 2001). Using the microcystin to chlorophyll-a ratios and the chlorophyll-a concentrations in pelagic water, the concentrations of particulate (cell-bound) microcystin in lake water were calculated by Fastner et al. (1999); in over 70% of the samples of 55 German water bodies, total particulate microcystin concentrations were below 10 µg l^-1, however the spatial and temporal concentrations varied by 4 orders of magnitude (2-25,000 µg l^-1) at bathing sites during a mass development of Microcystis spp.. Welker et al. (2001) found the highest microcystin concentration in the lake Müggelsee (Berlin, Germany) at a site within the fringing aquatic reed belt on the downwind side of the lake; samples yielded a concentration of 120 µg l^-1 dissolved microcystins which was, however, only a minor part (about 10%) of the total microcystins (cell-bound and dissolved) found at concentrations of up to 1,200 µg l^-1.

Two main mechanisms for decomposition of MC-LR in water bodies are microbial activity and degradation photosensitised by natural organic matter (NOM) (Cousins et al., 1996; Welker and Steinberg, 1999, 2000; Welker et al., 2001) that may lead to a low persistence of microcystins in aquatic systems. For instance the concentration of dissolved microcystins in the lake Müggelsee (Berlin, Germany) increased from non-detectable to over 70 µg l^-1 and dropped to non-detectable values again within a few days (Welker et al., 2001). Studies in Australia have shown that dissolved microcystins were present up to 21 days following treatment of a Microcystin aeruginosa bloom with an organic copper algaecide (Jones and Orr, 1994). Currently there are no explanations for these differences in degradation time of MC-LR indicating the lack of the knowledge about underlying elimination mechanisms.

The acute hepatotoxic symptoms that result from microcystin exposure are generally caused by binding and inhibiting the serine/threonine protein phosphatase types 1 and 2A (Solter et al., 1998; MacKintosh et al., 1990, 1995). The inhibition of these enzymes results in a massive structural disorganisation of the liver cells, followed by massive internal haemorrhage, often causing mortality (Råbergh et al., 1991; Thompson and Pace, 1992; Wickstrom et al., 1996; Dawson, 1998). In addition, microcystins have been identified as potent tumour promoters (Fujiki et al., 1996; Humpage and Falconer, 1999) and were found to induce DNA damage in mouse liver (Rao and Bhattacharya, 1996).

Although it is well known that 50% of Microcystis blooms show hepatotoxicity (Namikoshi and Rinehart, 1996), little is known about influences of cyanobacteria on aquatic organisms. Microcystis caused damages of zooplankton by inhibition of population growth and feeding activity, reductions in body size, filtering rate and survival time (Jungmann and Benndorf, 1994; Smith and Gilbert, 1995; Rohrlack et al., 1999, 2001). Smith and Gilbert (1995) described the potential of M i crocystis to alter the zooplankton community structure by differentially impacting individual species. Furthermore microcystins may also affect plants (Yamasaki, 1993; Pflugmacher et al., 1998).

The prevailing focus of investigations of the impact of cyanobacterial toxins on fish has been on the acute toxic effects of intraperitoneal injection and oral application of microcystin. The main effects of MC-LR found in these studies were damage to the liver, kidneys, or gills; disturbances of the ion balance; changes in cardiac function; growth inhibition; and mortality (Phillips et al., 1985; Sugaya et al., 1990; Tencalla et al., 1994; Råbergh et al., 1991; Carbis et al., 1997; Bury et al., 1995, 1997; Gaete et al., 1994;Rodger et al., 1994; Kotak et al., 1996; Fischer and Dietrich, 2000; Zimba et al., 2001; Zambrano and Canelo, 1996; Zhao et al., 2004; Li and Xie, 2004). The LD₅₀ of MC-LR (550 μg kg^-1) in common carp (C y prinus carpio L.) caused total loss of the parenchymal structure of the liver and degeneration of kidney tubuli (Råbergh et al., 1991).

In contrast to studies of acute toxicity, an ecological vague toxicity parameter, only a few findings on sublethal and chronic influences of microcystin have been reported. For instance, Råbergh et al. (1991) found several types of liver damage in fish that had been injected sublethal doses of the toxin. The so called “netpen liver disease” of Atlantic salmon was also associated with microcystins (Kent et al., 1990; Andersen et al., 1993; Williams et al., 1995).

The main uptake route of microcystin in trout was the gastrointestinal tract and toxicity was manifested as massive hepatic necrosis (Tencalla et al., 1994). In contrast no microcystin related effects on the growth of Rutilis rutilus fed with Aphanizomenon and Microcystis were found by Kamjunke et al. (2002).

With respect to ecological issues, it is necessary to study the effects of microcystins diluted in water because aquatic organisms also may absorb the toxins directly by their body or cell surfaces. It is reasonable to assume uptake via gills because direct effects of MC-LR on gills were reported by Zambrano and Canelo (1996).

Wiegand et al. (1999) showed there was uptake of MC-LR by zebrafish (Danio rerio) in early life stages, whereby MC-LR accumulated in the embryo by the factor 1.2. Recent evidence indicates that fishes and invertebrates can bioaccumulate cyanotoxins, and the ingestion of contaminated food represents one clear human health risk (Magalhães et al., 2003).

Polychlorinated biphenyls (PCBs) belong to persistent organic pollutants (POPs). POPs are stable, long-lived chemicals that are environmentally persistent, prone to global atmospheric transport and in some cases (which also applies for PCBs), accumulate in the food chain to levels that are potentially toxic to aquatic and terrestrial life (UNECE, 1998; Vallack et al., 1998). PCBs are human-manufactured chemicals produced during the mid-twentieth century and extensively used in a wide variety of industrial applications including hydraulic oils, solvent extenders, plasticisers, flame-retardants, lubricants, organic diluents, and dielectric fluids due to their desirable physical and chemical properties (dielectric and flame resistance, chemical and thermal stability). Unfortunately these properties of PCBs also contribute to their ability to lead to environmental problems. Although banned from further production (in North America and western Europe in the 1970s), PCBs can be found in almost every compartment of terrestrial and aquatic ecosystems (Tanabe, 1988; Simonich and Hites, 1995). In 1977 it was estimated that approximately 68 million kg (68,000 t) of PCBs had been released into the environment, and that additional 340,000 t were still in use and a possible source of future contamination (Cohen et al., 1993).

PCBs are chlorinated aromatic hydrocarbons with the general chemical formula C₁₂H_{10 n}Cl_n where n is the number of chlorine atoms ranging from 1 to 10. Depending on the number and position of chlorine atoms substituted on the biphenyl moiety, 209 possible individual PCB congeners can be formed (Ballschmiter and Zell, 1980).

PCB mixtures found in the environment are different from the commercially produced PCB mixtures because of differences in physicochemical properties, persistence, and bioaccumulation among the different congeners (EPA 1999). Specific PCB products are sometimes given specific names often completed by numbers. Aroclor was the brand name for Monsanto PCBs, Phenochlor and Pyralene were brand names for PCBs manufactured by the French company Prodelec.

The log K_ow values of PCB congeners range from 4.5 to 8.1 (Steinberg et al., 1992). For different PCB congeners the octanol/water partitioning coefficients were directly proportional to the chlorine content of the congener (Alkhatib and Weigand, 2002). Furthermore the water solubility and vapour pressure decrease as the degree of substitution increases, and the lipid solubility increases with increasing chlorine substitution (Schwarzenbach et al., 2002).

Differences between ortho- and non-ortho substituted PCBs were found, e.g., in binding of PCBs to aquatic humic substances (Uhle et al., 1999) and in toxic effects (see 3.4.2.4). It is hypothesised that the lack of chlorine substitution at opposing ortho positions allows the two phenyl rings to rotate into the same plane (Giesy and Kannan, 2002) and so these congeners are commonly referred to as coplanar PCBs, whereas congeners with substitution in the ortho positions are referred to as non-coplanar PCBs.

In the present study the ortho-substituted congener (non-coplanar PCB) 2,4,4` trichlorobiphenyl (PCB 28; Ballschmiter and Zell, 1980) was tested (Fig. 3, Tab. 2). PCB 28 (C₁₂H₇Cl₃) is one of the 6 reference congeners of the PCB/PCT/VC ordinance on the ban of PCB and PCT (1989).

Tab. 2. Physico-chemical properties of PCB 28 (from Paasivirta et al. 1999).

	Fig.3 Chemical structure of 2.4.4`-trichlorobiphenyl (C12H7Cl3).

The global cycling of PCBs results from their evaporation from soils and surface waters to the atmosphere and their redeposition back to land and surface water (EPA, 1999). In aquatic environments PCBs become mainly adsorbed to sediments and were found in concentrations of 9.6 to 567 µg g^-1dw in freshwater sediments by Kannan et al. (1997). High PCB concentrations of 1,000 µg g^-1 (suspended matter) in the river Niederrhein were described by Hellmann (1986). PCBs may be mobilized from sediments if disturbed, e.g., by flooding, dredging (EPA, 1999).

Moreover the persistence of PCBs and their high lipophilicity results in their bioaccumulation in fatty tissues and their biomagnification in the food chain. Generally bioaccumulation factors increase with chlorine content from the trichlorobiphenyls up through the hexachlorobiphenyls and then generally decrease with higher chlorine content of hepta- and octachlorobiphenyls (EPA, 1999). For instance bichlorobiphenyls display an approximately 450-fold decrease in the tendency to bioaccumulate in fish compared with tri- and tetrachlorinated PCBs (Abramowiscz and Olson, 1995). Furthermore it was found that the chlorine atoms in position 2, 4 and 5 at least one phenyl ring of the PCB molecule were a dominant factor causing accumulation of PCBs in aquatic organisms e.g., in smelt (Osmerus mordax) (Gagnon et al., 1990).

There is a considerable controversy about the relative contribution from food versus direct uptake from water in determining organochlorine levels in aquatic biota (Moriarty, 1988). Bruggeman et al. (1981) described the diet as primary route by which fish accumulate PCB compounds with log K_ow values greater than 5, whereas according to Borlakoglu and Haegele (1991) the concentration of PCBs in fish depends primarily on the PCB concentration in the sediments and particulate matter in the ambient water. High bioconcentration factors of 28 different PCB congeners (with n ranging from 2 to 10) between 7,710 and 940,000 (wet weight) were found in zebrafish (Danio rerio) by Fox et al. (1994). PCBs have been detected at µg g^-1 levels in fish from contaminated areas which were not appreciable below the current FDA action level of 2 µg g^-1 (e.g., Elskus et al., 1994; Stow et al., 1995).

PCB exposure is associated with a wide array of acute toxic effects on fish including liver damages, impairment of osmoregulation, reduction of immune functions, reproductive dysfunctions, impairment of sexual maturation, developmental disturbances, apoptosis ATPase inhibition, altered retinoid homeostasis and mortality (Merkens and Kinter, 1971; Koch et al., 1972; Hansen et al., 1974; Nebeker et al., 1974; Svoboda et al., 1994; Monosson et al., 1994; Rice and Schlenk, 1995; Billsson et al., 1998; Kim and Cooper, 1999; Piechotta et al., 1999).

PCBs can act as endocrine-disrupting chemicals that alter the behaviour of vertebrates as reviewed by Zala and Penn (2004). Hydroxylated metabolites of PCBs (OH-PCBs) have been shown to have agonist or antagonist interactions with estrogen receptors (Carlson and Williams, 2001).

The congeners appear to act by a variety of mechanisms. One proposed mechanism is based on the high affinity of coplanar PCBs (like dioxins) for the aryl hydrocarbon receptor (AhR) where the ligand-AhR complex induces the synthesis of the cytochrome enzyme P4501A1 (CYP 1A1) which was obtained by in vivo and in vitro experiments of fish (Hermens et al., 1990; Clemons et al., 1996; Stegeman and Lech, 1991; Monosson and Stegeman, 1991; Koponen et al., 2000). These coplanar PCBs may exert, thus, dioxin-like effects in addition to AhR independent effects which they share with non-coplanar PCBs that have no or only slight AhR agonist activity.

Based on studies that indicated the role of the AhR in mediating toxic and biochemical effects induced by PCBs, a TCDD equivalency factor (TEF) approach was developed. This approach allows the expression of toxic potential of a complex mixture of individual congeners as one integrated parameter, the toxic equivalency value, in which the toxic potency of the mixture corresponds to the potency of the most toxic congener, TCDD (2,3,7,8-tetrachloro dibenzo-p-dioxin). However, for every PCB congener tested, the TEF values are response- and species dependent (Safe, 1990).

Another mechanism of PCB toxicity may be related to the genetic level, e.g., the bioactivated form of a coplanar PCB led to an increase of the recombination rate in somatic cells of Drosophila;in contrast, PCB was not genotoxic in bacterial systems (Butterworth et al., 1995; McGowen et al., 2000).

Most of the acute toxic effects on fish are related to commercial PCBs mixtures and to coplanar PCBs. In contrast, only a few findings exist concerning effects of non-coplanar PCBs. The development of scientifically based regulations for the risk assessment of PCBs requires analytical and toxicological data on the individual PCB congeners present in any technical mixture (Giesy and Kannan, 2002), since the degree of chlorination and the chlorine positions on the molecule will greatly influence the fate and toxicity of each congener (Tang et al., 1991).

Particularly, with the recent developments in the understanding of non-dioxin-like effects of PCBs, it is pertinent to examine critically the effective doses at which non-coplanar PCBs could elicit non-dioxin-like effects in animals (Giesy and Kannan, 2002). Non-coplanar congeners may be responsible for the neurobehavioural effects of PCBs, but most of these effects were related to mammals including humans (e.g., Schanz et al., 1991; Seegal, 1996). In contrast, little information is available on behavioural toxic potencies of PCBs on aquatic organisms. For that reason in the present study the behaviour of fish was investigated under the influence of a single non-coplanar PCB congener (PCB 28).

The main focus of this thesis is the analysis of specific behavioural aspects of two fish species under the influence of chemical stressors and therefore, a contribution to the behavioural approach for ecotoxicological issues. The chosen fish species both belonging to the family Cyprinidae were:

a) the tropical species zebrafish Danio rerio,

b) the temperate species sunbleak Leucaspius deline a tus.

Effects of chemical stressors on Danio rerio that is a model system for integrative physiology and toxicology were compared with those of the native Eurasian species Leucaspius deline a tus.

For analysing pollutant-induced effects on fish, chemicals were selected which may serve as model substances for investigating toxic effects of chemical stressors on behaviour:

a) the naturally occurring cyanobacteria toxin microcystin-LR (MC-LR) and

b) a typical man-made substance 2.4.4`-trichlorobiphenyl (PCB 28).

Both substances were tested on both species in different concentrations on a sublethal level for investigating their influences on behaviour and to answer the following questions:

Do MC-LR and PCB 28 act as stressors for fish and if they do, which stress symptoms on the behavioural and chronobiological level can be evaluated?

Do fish show similar reactions when exposed to different pollutants?

Are dose-related effects detectable and can they theoretically be described?

How consistent are the reactions among different species?

Which is the temporal development of mean motilities over the whole period of exposure to the stressors?

A second but not less important objective was to apply different analytical methods to the data derived from the tests. So the basic behavioural analyses were combined with chronobiological procedures such as cosinor analysis and power spectral analysis which are not commonly used in estimating the risks of aquatic contaminants. Questions to be answered in this respect were:

Which chronobiological methods are useful to enhance behavioural investigations in the field of ecotoxicology?

Do the analyses of rhythmical changes provide novel information on the stress potential of the investigated contaminants?

Is the application of time series investigation valuable for studying harmful environmental factors?

Do the analysis of the relation between motility versus number of turns by using regression analysis indicate exposure-induced changes in swimming mode?

Last but not least, the widening of the basic knowledge for applying behavioural tests on fish as standardized methods for biomonitoring was considered to be of practical importance and raised these questions:

How are the experiences gained in this study applicable for biomonitoring?

Which standards of the experimental design are applicable for biomonitoring using behavioural toxicity endpoints?

What importance has chronobiological aspects in the field of biomonitoring?

Furthermore the results are discussed regarding potential ecological effects of the induced changes in behavioural pattern including cycling aspects (e.g., swimming activity levels and circadian rhythms) which may lead to some adverse consequences for fish populations or communities.