Examensarbete, 60hp
Master´s Degree Thesis in Biology | 60.0 hp | 5BI208
Vt 2020
Putting behavioral assays
on fish to the test: Are
sociality and scototaxis
trials relevant in the
wild?
Examensarbete, 60hp
Master´s Degree Thesis in Biology | 60.0 hp | 5BI208
Vt 2020
Title page
Putting behavioral assays on fish to the test: Are sociality and scototaxis trials relevant in the wild?
Status report: Final report Date of issue: 28-05-2020 Picture background: Arno Veenstra Picture salmon: Arno Veenstra
Abstract
Animal behavior has become a frequently used tool in modern ecology and ecotoxicology, where laboratory behavioral traits are recognized as sensitive endpoints for assessing natural behavior or non-lethal effects of pollutants on animals. Within these research fields, behavioral traits measured in laboratory environments have been used to formulate predictions of ecological consequences that accompany specific behavior. However, the predictive power of behavioral traits measured in simplified laboratory environments for complex natural aquatic ecosystems has been questioned. In this study, I have examined to what extent behavioral changes, noted in laboratory settings in response to chemical stressors (an anxiolytic drug) or visual cues of black and white bottom substrates, are also expressed in the wild. In my first experiment, I scrutinized whether reduced social behavior previously shown to occur in the lab for European perch (Perca fluvatilis) in response to oxazepam also occurs within a natural lake subjected to oxazepam. The in situ behavior was measured using high temporal resolution (3 sec) acoustic telemetry. In my second experiment, I assessed if the Atlantic salmon’s (Salmo
salar) preference for black bottom substrates (scototaxis) in laboratory assays could be
utilized for guiding migrating Atlantic salmon in situ. I show that: i) Oxazepam does not affect the social (association) behavior or the social network structure of perch in natural settings, in contrast to laboratory-based predictions; ii) Atlantic salmon show a
Table of contents
1. Introduction ... 5
2. Methods ... 8
2.1. Whole lake experiment ... 8
2.1.1. Perch association behavior and social network in a natural system ... 8
2.1.2. Data preparation for measuring perch social behavior ... 9
2.1.3. Analysis perch social (association) behavior ... 9
2.2. The use of scototaxis to guide migrating salmon ... 10
2.2.1. Data handling: anxiety- like behavior of salmon in the lab ... 10
2.2.2. Data analyses: anxiety- like behavior of salmon in the lab ... 10
2.3. The use of scototaxis to guide migrating salmon: the natural system ... 11
2.3.1. Data preparation: anxiety- like behavior of salmon in a natural system ... 12
2.3.2. Data analyses: anxiety- like behavior of salmon in a natural system ... 12
3. Results... 13
3.1. Whole lake experiment: perch association behavior and social network structure ... 13
3.1.1. Testing hypothesis one: the effect of oxazepam on the social behavior of perch ... 13
3.1.2. Testing hypothesis two: perch social network ... 14
3.2. The use of scototaxis to guide migrating salmon ... 17
3.2.1. Testing hypothesis three: salmon preference for black surfaces... 17
3.2.2. Testing hypothesis four: anxiety- like behavior of salmon in a natural system ... 18
4. Discussion ... 19
4.1. Anxiolytics and social perch behavior ... 19
4.2. The use of scototaxis to guide migrating salmon: behavior in the lab ... 20
4.3. The use of scototaxis to guide migrating salmon: behavior in the natural system ... 20
4.4. Laboratory behaviors and its relevance for behavior in nature ... 21
Acknowledgement...22
References ...23
5
1. Introduction
Anthropogenic development is currently changing environments at an unprecedented rate and scale. Human-induced environmental change is extremely fast and poses a major threat to global biodiversity (Butchart et al., 2010). Humans affect terrestrial and aquatic environments through, for example, increasing greenhouse gas emissions (Nagelkerken and Munday, 2016), renewable energy, (Renöfält et al., 2010) and environmental pollutants (aus der Beek et al., 2016; Scott and Sloman, 2013). Animals respond to a changing environment predominantly through modification of their behavior
(Tuomainen and Candolin, 2011). In fact, research has shown that environmental change has a direct effect on animal behavior and that it changes behavioral traits, such as, foraging, predator avoidance, dispersal, migration and reproduction (Traill et al., 2010; Tylianakis et al., 2008). Animal behavior is known to construct and shape ecological communities (Peacor and Werner, 2003; Rothley and Dutton, 2006) and addressing behavioral responses of animals to environmental changes is crucial in understanding and predicting effects on these ecological communities (Rothley and Dutton, 2006). The realization of the importance of animal behavior as a consequence of environmental change and the occurrence of systematic difference in the behavior repertoire of animals of the same species (i.e. personality) have stimulated numerous studied within ecology
(Brodin et al., 2018, 2017, 2014; Fahlman et al., 2020; Hellström et al., 2016; Klaminder et al., 2016; Saaristo et al., 2019; Scott and Sloman, 2013; Weis et al., 2001). Animal behavior has also become an increasingly important tool in modern ecotoxicology, where behavioral traits are recognized as sensitive endpoints for assessing non-lethal effects of pollutants on animals (Arnold et al., 2014; Brodin et al., 2014; Hellou, 2011; Hellström et al., 2016; Lagesson et al., 2018; Melvin and Wilson, 2013; Pyle and Ford, 2017; Weis et al., 2001). Over the past two decades, studies simulating aquatic pollution and
environmental change have used behavioral tests, originally advanced from research focusing on animal personality, for ecological predictions (Melvin and Wilson, 2013). Recently, due to ethical concerns, toxicity and exposure models have been developed to prevent the use of laboratory toxicity test on a substantial number of animals, but the uncertainty of using behavior as a measure within ecotoxicology has been highlighted
(Johnson et al., 2020). The uncertainty of behavioral endpoints for ecological predictions makes it urgent to further assess their predictive power.
Assessing animal behavior for ecological predictions is especially necessary in aquatic systems, with pharmaceutical contaminants being recognized threats that these ecosystems face (Brodin et al., 2018). This threat is mostly caused by the expanding human consumption of pharmaceuticals and low removal efficiency in sewage treatment plants (aus der Beek et al., 2016; Nikolaou et al., 2007; Verlicchi et al., 2012). Research has shown that pharmaceutical contaminants can have various effects on aquatic wildlife and their behavior (Brodin et al., 2014; Hellström et al., 2016; Kidd et al., 2007;
6
oxazepam that may play a central role behind these observed effects in nature, is an increased anti-social behavior. Although the potential impact of psychiatric drugs on the social behavior of fish is demonstrated and the consequences for ecosystem functioning have been addressed (Fahlman et al., 2020; Klaminder et al., 2016; Lagesson et al., 2018; Saaristo et al., 2019), The impact on social behavior by benzodiazepines of fish,
habituated in complex ecosystems, has so far not been tested. Thus, more research is needed to understand the effects of benzodiazepine contaminants on the social behavior of fish in natural ecosystems, especially since social behavior for wild is crucial for
shoaling, and the anti-predator response (Brodin et al., 2018). In other words, there is an urgent need to further assess how well laboratory predictions of social behavior describe behavior in more complex settings. In this study behavioral endpoints are used as a tool to bridge the gap between behavioral assays in the lab and behavioral in-situ studies. Another major anthropogenic threat to aquatic ecosystems, where behavior of fish is crucial for survival, is the hydropower dams blocking fish migration routes. Dams are distributed worldwide, predominantly to mitigate the high demand for renewable energy
(Bratrich et al., 2004; Calles, 2005), but dams also fundamentally transform rivers and their ecosystems (Nieminen et al., 2016; Renöfält et al., 2010). In fact, fragmentation of rivers by dams is one of the major threats to aquatic ecosystems worldwide (Dudgeon et al., 2006; Nilsson et al., 2005; van Puijenbroek et al., 2019). One of the most detrimental ecological effects, caused by hydropower dams, is the disruption of dispersal of riverine species (Calles, 2005). Many fish species rely on an intact longitudinal connection to be able to migrate between habitats to feed and spawn, and depend on vertical connectivity for development of embryos on a hard substrate (Calles, 2005; Lucas and Baras, 2001). Hydropower dams influence both connection types by blocking migration routes and by altering sediment dynamics (Fuller et al., 2015; Renöfält et al., 2010). To reduce the negative effects of hydropower production on fish, fishways (i.e. by-passes, fish-pass or fishladder) are commonly used as mitigation measures (Renöfält et al., 2010). However, optimizing the effectiveness of fishways remains a challenge, as the efficacy of passing fish through these structures remains low at locations around the world (Bunt et al., 2016; Nieminen et al., 2016; Noonan et al., 2012; Roscoe and Hinch, 2010; Silva et al., 2018). For example, salmon were observed to be reluctant to enter the fishway, and were even observed backing out of the fishway after entering (Karppinen et al., 2002). Not only is an efficiently working fishway, that facilitates bi-directional movement for anadromous fish, important for fish population continuity, it is also essential for ecosystem functioning
(Silva et al., 2018). As most anadromous fish, such as salmonids, provide an important nutrient subsidy to freshwater environments through migration, spawning and death
(Gresh et al., 2000; Wheeler and Kavanagh, 2017).Therefore, new methods that can affect in situ fish behavior and improve migration succession through fishways, would be a great advancement for the research field with positive implications for rivers world-wide. Previous laboratory based behavior assays have noted that fish show a preference for swimming on black substrates if subjected to black and white substrates (Maximino et al., 2010; McCallum et al., 2019), but if this behavior (referred to as scototaxis) can be used to guide migrating fish in situ has not been tested.
7
high temporal resolution (3 sec) acoustic telemetry, suitable to track fish behavior and performance, in two natural lakes. I also examined a novel method for steering
downstream migrating Atlantic salmon (Salmo salar). I hypothesize that: i) Oxazepam reduces the social (association) behavior of wild perch, i.e. in line with observation made in laboratory studies (Brodin et al., 2013; Klaminder et al., 2014); ii) Oxzepam alters the social network structure of perch; and iii) Atlantic salmon show a preference for the dark compartment in the dark/light preference protocol, i.e. in line within other fish species
(Maximino et al., 2010); and iv) the preference for most Atlantic salmon to follow black bottom substrates rather than white bottom substrates (i.e. anxiety-like behavior) can be used to guide fish in situ.
Figure 1: Conceptual model of the thesis. This flow model shows the common interests of knowing the importance of anxiety-like behaviour within both ecotoxicology and aquatic ecology. This thesis addresses the need of evaluating how fish behavior measured in laboratory trials can be used to predict fish behavior in nature. Ecotoxicology Aquatic ecology Chemical stressor Scototaxis Anxiety-like behavior of
fish in the lab
Behavior in nature Prediction s in nature Anxiety-like behavior of
8
2. Methods
My four hypotheses were tested using two different experiments. One whole-lake experiment where oxazepam was added to the lake water (hypothesis i and ii) and one stream experiment where I tested the use of scototaxis behavior for guiding salmon smolt (hypothesis iii and iv). The details of each experiment are described below.
2.1.
Whole lake experiment
2.1.1. Perch association behavior and social network in a natural system
The study was conducted in July-October of 2016. In July, Eurasian Perch (Perca
fluviatilis) (size range was 166±27 mm) were caught in lake Stöcksjön (63°45'45.9"N
20°11'54.1"E) close to Umeå, northern Sweden, using a beach seine net. The perch were then transported in oxygenated tanks to Umeå Marine Research Facility (UMF) and kept in a 1 m3 flow-through tank at 15±1.5˚C. Shortly after being moved to UMF, 44 perch were equipped with acoustic transmitters (VEMCO V4 180 khz, tag weight = 0.38 g), surgically inserted into the abdominal cavity. The tag burden was <1.5%, well below the lowest reported effect found in the literature (Brown et al., 2010). During the procedure, the fish were under anesthesia from MS-222. The incision was closed with a suture, and the fish was left to recover for a minimum of 14 days before any further handling and/or testing of the fish was performed. Every individual was eating chironomid larvae during the entirety of the laboratory phase of the study, and appeared healthy by the end of the recovery time
(Fahlman et al., 2020).
The field site consisted of two twin lakes close to Åmsele, Västerbotten, Sweden
(64°29'2.5"N 19°25'8.2"E). Both lakes (here named Upper and Lower Lake) are old dead-ice depressions, each with an area of about 4000 m2 and a maximum depth of roughly 6 m, and a border of quagmire covering the shoreline. Water temperature was 13±1.5 ˚C during the course of the study. The lakes were fishless prior to the experiment, which was confirmed by previous research done, as well as netting with no catch. Resource sampling showed an abundance of pelagic (Daphnia zooplankton and Chaoborus larvae) and benthic (mainly Ephemeoptera, Odonata, and Asellus) invertebrates. One day prior to the perch being introduced, each lake had been stocked with four Northern Pikes (Esox
Lucius) (length range 600±50 mm), caught in Lake Tavelsjön (64°0'2.4"N 20°3'5.1"E),
9
~14 ug/l, methods are available in Fahlman et al. 2018). The fish were once again left undisturbed, with the exception of pike reintroduction on the 23rd of September to compensate for observed mortality. The experiment was terminated on the 30th of September, as this was the battery life limit of the acoustic transmitters.
2.1.2. Data preparation for measuring perch social behavior
Within lake, fish positioning data was filtered using the methods of (Leander et al., 2019)
and then interpolated at a 60 sec level to ensure consistency and comparability over time for the analysis, avoiding differing number of data points for a given time period. From this data, association between individuals (dyads) was used to measure the effect of oxazepam on the social behavior of perch. To calculate “Association”, the software SOCPROG version 2.9 for MatLab was utilized to find association levels for each individual. Association levels was defined as being within a one-meter proximity of any given conspecific for more than 15 minutes, with one calculation performed every hour. The strength of pairwise associations was determined using the simple ratio index (SRI). The SRI is calculated as: 𝑆𝑅𝐼 = 𝑥
𝑥+ Yab + Ya + Yb where 𝑥 is the number of sampling periods
in which both perch a and b were associated, Yab is the number of sampling periods in which both perch were detected but not associated, and Ya and Yb are the number of sampling periods in which only a and b, respectively are identified (Ginsberg and Young, 1992). This gives individuals an average level between 0 and 1 for the entire study period, with 0 being no association at any point, and 1 being within 1 meter of a conspecific for the entire period. The time limitation was chosen as this is a common time limitation of laboratory assays, as well as in this study. Association distance is commonly considered body-length dependent (Hoare et al., 2000), and might in this case be shorter than the one-meter constraint of the model. However, due to system limitations, I cannot guarantee sub-meter precision of the acoustic array, and thus choose to be conservative when pre-determining the minimum association distance of the perch. Association for every individual was calculated over two periods, a control and a treatment period. In those periods, day and night were analyzed separately to control for the effect of sunlight. “Daytime” was defined as the time between sunrise and sunset and “Night-time” was defined as the time between sunset and sunrise. The control period ran from the 5th of September until the 15th of September 2016 and the treatment period ran from the 17th of September until the 30th of September 2016, the periods correspond with the exposure to oxazepam.
2.1.3. Analysis perch social (association) behavior
Generalized linear mixed models (GLMM) with binomial distribution (for non-normally distributed data) was used to test the difference in association of perch between the control and treatment period. The GLMM was performed with the glmer command in the package lme4 version 1.1-19 (Bates et al., 2015) within the R statistical environment (R Core Team 2016). The response variable was “Association” and “Treatment”, “Time of
day” and “Lake” were used as two-level fixed effects in the model. “ID-number” was used
as a random effect. In the initial candidate model, the two-way and three- interactions that were included were: “Treatment*Time of day”, “Treatment*Lake”, “Lake*Time of
day” and “Treatment*Time of day*Lake”. Stepwise backwards model selection was used
10
2.2. The use of scototaxis to guide migrating salmon
The study was conducted on 15-05-2019 until18-05-2019 in Stornorrfors’s fish breeding facility (63°52'50.0"N 20°01'22.0"E) close to Umeå, northern Sweden. During this period, the anxiety-like behavior (i.e. scototaxis) of 200 hatchery reared, smoltified, 2-year old, Atlantic salmon (Salmo salar) was measured in the lab (forklength range was 176 ± 13mm; mean ± sd). The dark/light preference protocol was used to measure the anxiety-like behavior (Maximino et al., 2010). In this test, the time spent in the black section (anxious) and the white section (none-anxious) of the tank was used a measure of the shyness-boldness continua for each individual. In this study, the proportion of time spent in the white section of the tank was
measured and used as a measure of boldness for each individual. In order to measure the
anxiety-like behavior in the lab, four half-black and half-white containers were used (L80 x W50 x H40 cm) (see Figure 2) . Each container was filled with river water up to approximately 15cm and each tank was refreshed every day before initiating the next cycle of experiments. On 15-05-2019 the tanks were accidently filled with tap water due to a miscommunication. The effect of this will be discussed in chapter 4, paragraph 4.2. While recording, every single
individual salmon was released in the center of the tank and was recorded for 20 minutes with a camera (SONY Handycam HDR-PJ50VE) that was mounted overhead of each container. After 20 minutes, the recording was stopped and the salmon were caught and released in 12 groups of four and one pair. On each day, a total of 50 salmon were released in the stream.
On 13-05-2019, the salmon were equipped with Passive Integrated Transponder tags (Biomark APT12™, 12.5 mm x 2.1 mm, 134.2 kHz), which were inserted into the
abdominal cavity (see Figure 3). During the procedure, the salmon were anesthetized with MS-222. The PIT-tags were inserted in order to measure the salmon’s anxiety-like
behavior in the natural system after the lab experiment, that ran from 15-05-2019 until 18-05-2019. The salmon were kept in a rearing tank to recover until 15-05-2019, which was fed with a constant flow of river water. The fish were not fed during the entirety of the study.
2.2.1. Data handling: anxiety- like behavior of salmon in the lab
The recorded behavior of each salmon was analyzed with Toxtrac and the “proportion of
time spent on the white” was extracted from the analysis. The “proportion of time spent on the black” for each salmon was then extracted using the formula:
"𝑃𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓 𝑡𝑖𝑚𝑒 𝑠𝑝𝑒𝑛𝑡 𝑜𝑛 𝑡ℎ𝑒 𝑏𝑙𝑎𝑐𝑘" = 1 − "𝑝𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓 𝑡𝑖𝑚𝑒 𝑠𝑝𝑒𝑛𝑡 𝑜𝑛 𝑡ℎ𝑒 𝑤ℎ𝑖𝑡𝑒"
2.2.2. Data analyses: anxiety- like behavior of salmon in the lab
Generalized linear models (GLM) with binomial distribution (for non-normally distributed data) was used to test the difference in anxiety-like behavior between
experiment days. The response variable was “Proportion of time spent on the black” and Figure 2: Setup of the four containers
to measure the anxiety-like behaviour in the lab.
11
“Experiment days” was used as fixed effect in the model. Two tailed 95% confidence intervals were constructed to visualize the difference in anxiety-like behavior between the four experiment days. All analyses were performed using R version 3.5.3 (R Core Team 2016).
2.3. The use of scototaxis to guide migrating salmon: the
natural system
The study was conducted on 15-05-2019 until 18-05-2019 in a stream, adjacent to
Stornorrfors’s fish breeding facility (63°52'50.0"N 20°01'22.0"E) close to Umeå, northern Sweden. The weather conditions on 15-05-2019 until 18-05-2019 were sunny with no precipitation and with temperatures during the day between 15 °C and 19 °C. During the night, however, the temperature dropped and stayed between -1 °C and 3 °C. The stream is approximately 250m long, 3m wide and 0.3m deep, and the average velocity of the water was 0.8m/s. The stream leads up to the river Umeälven, which the salmon smolt use to migrate to the Baltic sea. In total, 199 smoltified Atlantic salmon (same individuals were used during the scototaxis behavior study in the lab) were released in the stream. Approximately 145 meters downstream of the point of release, the bottom substrate and color of a five meter stretch of stream was altered to black and white to assess how migrating salmon reacted to the two colors and thus, test hypothesis iv during field conditions (see Figure 4).
Figure 4: Picture of the experimental setup in the stream at Stornorrfors. Visible on the left are the two antennas upstream and on the right are the two antennas downstream. The antennas connect to the HPRs that collect the data in the bottom of the picture. Also visible are the two moveable strokes of a black and white polymer that are held in place with granite rocks.
12
Four PIT-tag antennas were placed to measure the salmon’s point of entry and point of exit, this was done by placing two rectangular antennas (Biomark, 30.4cm x 81.2cm) upstream and two square antennas (Biomark, 60.9cm x 60.9cm) downstream. However, only two antennas were reading simultaneously, as reading three or four antennas simultaneously created too much interference. The two antennas that were not reading, functioned as dummy antennas. The two antennas were read with two Biomark HPR PLUS readers. The upper-right antenna and lower-left antenna were reading from 15-05-2019 until 18-05-15-05-2019. The data was extracted every day around 17.00 PM from both readers. The field experiment was terminated around 17.00 PM on 18-05-2019. To test if cover affects the salmon’s anxiety- like behavior in a natural system, a cover was placed in front of the antenna on the black polymer on the morning of 16-05-2019. On the morning of 17-05-2019, the cover was placed in front of the antenna on the white polymer. No cover was placed on 15-05-2019 and 18-05-2019. A camera was placed downstream, behind the downstream antennas, to capture the movement of the salmon within the experimental setup. The camera recorded the salmon in the experimental setup every day from 08.30 AM until 17.30 PM.
2.3.1. Data preparation: anxiety- like behavior of salmon in a natural system
In order to extract the behavior of the salmon, the last timestamp was selected for each individual that was read by an antenna. The last time stamp indicates when the individual passed the antenna. Each read individual was then given a code that resembled their behavior in the experimental setup. For example, if a fish passed the upstream antenna on the black substrate and the downstream antenna on the white substrate, it received the code: “BlackWhite”. Because it entered the experimental setup on the black substrate and it left the experimental setup on the white substrate. If the individual was only recorded on the upstream antenna on the black substrate, it would have received the code: “BlackBlack”. If the individual was only recorded on the downstream antenna on the white substrate, it would have received the code: “WhiteWhite”. The data from the camera was used to verify the number of diagonal crossing individuals, which neither antenna could record. This behavior according to the situation described in the example would be given the code: “WhiteBlack”. However, the code for “WhiteBlack” was not linked to individual fish, as these fish were not read by an antenna. The number of “WhiteBlack” swimming individuals were added to a second dataset as “unknown individuals” and were added to the analysis in a later stage. Besides, the number of passing individuals on the white substrate were counted to verify the number of “WhiteWhite”. If a fish was not recorded on the antenna, it received the code: “NA”. This code resembles a missing value. The “proportion of individuals exiting the black substrate” was calculated per day to extract the anxiety-like behavior of the group of salmon for that day. The “proportion of
individuals exiting the black substrate (P)” was calculated as:
P =𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙𝑠 𝑒𝑥𝑖𝑡𝑖𝑛𝑔 𝑡ℎ𝑒 𝑏𝑙𝑎𝑐𝑘 𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙𝑠
2.3.2. Data analyses: anxiety- like behavior of salmon in a natural system
13
3. Results
3.1. Whole lake experiment: perch association behavior and
social network structure
In the two lakes, all the tagged fish could be successfully detected, and the system displayed high coverage and detection precision (Figure S1). This was also true for the introduced pike, which were released into the lake three days prior to the perch. The total number of detections used for analysis was ~670,000. No detectable predation took place during the study period and all the transmitters were functional throughout the
experiment. Besides, no change in behaviour was detected after reintroduction of the pikes on 23-09-2016.
3.1.1. Testing hypothesis one: the effect of oxazepam on the social behavior of perch
Stepwise backwards model selection showed that the best fitted GLMM model was “Time
of day” as the two-level fixed effect and “ID-number” as the random effect. During the
model selection, none of the two-way and three-way interactions showed any significant effect on association (P>= 0.05). Indicating that exposure to oxazepam did not have a significant effect on the association between perch. This results is in contrast with my first hypothesis that oxazepam would reduce the social behavior of perch in a natural
environment. However, new insights were gained, as the results from the GLMM showed that the association behavior of the perch decreased significantly during the night
compared to their association behavior during the day (Z = -2.91, P= 0.004). This change was observed in both lakes (see Figure 5).
Figure 5: Graph on the left: 95% confidence interval of the mean association value of perch for day and night in Upper lake for both study periods (n=22). Graph on the right: 95% confidence interval of the mean association value of perch for day and night in Lower lake for both study periods (n=22).
During the study period in the Upper lake (treated with oxazepam), four individuals showed a significant change in their association behavior during the day between the control and treatment period (P<= 0.05). Three of these individuals significantly
14
0.05). This is in contrast to my first hypothesis, as this fortifies the results from the GLMM that the treatment with oxazepam did not affect the association behavior of the perch, since the observed changes on an individual level were similar in both lakes.
3.1.2. Testing hypothesis two: perch social network
In contrast to my second hypothesis, the social network of perch was not affected by oxazepam, as the social network structure was remarkably similar during both the control and the treatment period. The circadian rhythm, however, seems to be the most
influential factor in directing social structure of perch. The social network of perch illustrates that tie strength (association) was typically higher in both lakes during the day compared to the tie strength during the night. During the day in both lakes, most
individuals generally spent more than 60% of the sampled hours within one meter proximity of one or more conspecifics (see Figure 6 and 8). During the night, this decreased to approximately 35% of the sampled hours (see Figure 7 and 9).
The highest degree of association found between dyads (pairs of individuals) during the study period was 91% in Upper lake and 82% in Lower lake. This was substantially lower during the night in both lakes, as the highest degree of association found between dyads was 56% in Upper lake. In Lower lake, the highest degree of association was found to be 64% between dyads.
Figure 6: Social network of perch (n=22) in the Upper lake during the day. Every node represents an individual and the ties between every node represents the association between individuals. The number next to every node represents the number of ties each individual has within the network. Ties are only visualized when individuals associated >= 60% of the total study period during the day. The labels next to each node represent the degree of association. The individual in the upper left corner with zero ties to the network, has a tie strength lower than 60% and is thus not shown.
15
Figure 7: Social network of perch (n=22) in the Upper lake during the night. Every node represents an individual and the ties between every node represents the association between individuals. The number next to every node represents the number of ties each individual has within the network. Ties are only visualized when individuals associated >= 35% of the total study period during the night. The labels next to each node represent the degree of association.
Figure 8: Social network of perch (n=22) in the Lower lake during the day. Every node represents an individual and the ties between every node represents the association between individuals. The number next to every node represents the number of ties each individual has within the network. Ties are only visualized when individuals associated >= 60% of the total study period during the day. The labels next to each node represent the degree of association. The individuals in the upper left corner with zero ties to the network, have a tie strength lower than 60% and are thus not shown.
16
Figure 9: Social network of perch (n=22) in the Lower lake during the night. Every node represents an individual and the ties between every node represents the association between individuals. The number next to every node represents the number of ties each individual has within the network. Ties are only visualized when individuals associated >= 35% of the total study period during the night. The labels next to each node represent the degree of association. The individual in the upper left corner with zero ties to the network, has a tie strength lower than 35% and is thus not shown.
17
3.2. The use of scototaxis to guide migrating salmon
The anxiety-like behavior of 199 individuals was recorded successfully in the lab. One individual was euthanized before recording, as fungi had compromised the structure of the entire tailfin. All releases of salmon were completed successfully and the individuals appeared healthy when released into the stream.
3.2.1. Testing hypothesis three: salmon preference for black substrates
The results from the GLM showed that on 16-05-2019 until 18-05-2019, Atlantic salmon spent on average 78 ± 2 (mean ± s.d.) percent of the time on the black substrate. The results are in line with hypothesis three: hatchery reared, smoltified, 2-year old, Atlantic salmon have a distinct preference for the dark compartment in the dark/light preference protocol. The results from the GLM are confirmed when comparing the 95% confidence interval between days (see Figure 10). However, the results from the GLM also showed that “proportion of time spent on the black” on 15-05-2019 was significantly lower compared to 16-05-2019 (Z= 2.96, P = 0.003), 17-05-2019 (Z= 3.27, P = 0.001), 18-05-2019 (Z= 3.45, P = 0.001). Indicating that the salmon, on 15-05-18-05-2019, spent relatively more time on the white substrate. Which suggests that the anxiety-like behaviour of salmon, on 15-05-2019, was influenced by other abiotic factors besides the color of the substrate. The details of the behavior observed on 15-05-2019 will be further discussed in chapter 4.2.
18
3.2.2. Testing hypothesis four: anxiety- like behavior of salmon in a natural system
In total, 90 individuals passed through the experimental setup. The camera footage showed that the salmon passed both head-first and tail-first through the experimental setup. Most of the individuals passed through in seven seconds. Two individuals were found dead on the morning of 16-05-2019 and did not pass the experimental setup. The individuals had no puncture wounds, but one had compromised scales around the mid-section of its body (Figure S2). The other 107 individuals were not recovered and were most likely lost due to predation.
The two-sample test for equality of proportions showed that cover had no significant effect on the anxiety-like behavior of the salmon in the experimental setup (P = 0.666, df =1). Since the salmon showed a significant preference to the dark compartment in the lab, an one-tailed confidence interval was used to illustrate the results from the field
experiment. The results from the field experiment are in line with my fourth hypothesis and show that the salmon past slightly more over the black substrate compared to the white substrate (𝜇= 0.61 + 0.11). This is illustrated by Figure 11.
19
4. Discussion
4.1. Anxiolytics and perch social behavior
Previous studies have shown, that behavioral traits of fish are significantly affected when exposed to anxiolytics like oxazepam (Brodin et al., 2017, 2013; Hellström et al., 2016; Klaminder et al., 2014; Saaristo et al., 2019). To date, this benzodiazepine is known to alter several behavioral traits in perch, such as: feeding rate (Brodin et al., 2013), home range, activity and boldness (Klaminder et al., 2016). Research also suggested that, high concentrations of oxazepam also significantly reduces social behavior in 2- year old perch
(Brodin et al., 2013; Klaminder et al., 2014). However, in contrast to my first hypothesis I found no impact on social behavior by oxazepam in the studied lakes. Importantly, previously mentioned studies have measured effects of oxazepam on perch at an individual level, mostly during laboratory behavioral assays. My study, however, tested the effects of oxazepam on perch as a collective, which could elucidate why the results of this study are not in line with previous findings of other studies concerning the effects of oxazepam on perch. The notion is well grounded as previous research suggests that collective decision making of animals in groups could counter individual behavioral variance (Lü et al., 2008). The theory supports my findings that exposure to oxazepam does not significantly change the social behavior and social network structure of perch in
situ. This means that the hypotheses that oxazepam reduces social (association) behavior
(hypothesis i) and alters the social network structure of perch can both be rejected (hypothesis ii). Notice that the oxazepam concentration is about ten times higher than previous studies where effects on fish behavior has been observed (Brodin et al., 2018, 2013; Hellström et al., 2016), making it very unlikely that differences between the lake and previous laboratory trials were due to too low exposure in the lake. Especially since the uptake of oxazepam seems to become more efficient over time due to food-web transfer (Lagesson et al., 2016).
Previous behavioral studies that examined the effect of oxazepam on perch were
performed in the lab, where most abiotic and biotic factors remained constant during the study (Brodin et al., 2013; Klaminder et al., 2014; Saaristo et al., 2019). Saaristo et al. (2019) concluded that the impact on behavior by temperature was independent from the exposure to oxazepam, indicating that environmental stressors could overrule the effect of oxazepam. During this study, the perch were situated in two natural lakes and were thus exposed to all environmental stressors. This also supports why no difference was found in association behavior and social network structure, since any of the environmental
stressors could have overruled the effects of oxazepam.
20
These collective findings insinuate the need for a better understanding of the independent and interaction effects with environmental stressors of benzodiazepines in natural aquatic systems (reviewed in Saaristo et al. 2019). Moreover, future studies should decrease the resolution when exploring the effects of benzodiazepines on the social network structure, since the overall network structure does not seem to be affected.
4.2. The use of scototaxis to guide migrating salmon: behavior
in the lab
In line with my third hypothesis, the results show that hatchery reared, smoltified, 2-year old, Atlantic salmon have a distinct preference for black surfaces. The preference could be explained by the fact that Atlantic salmon smolt are cryptic (Donnelly and Whoriskey, 1993). Crypsis is a primary defensive adaptation, which relies upon sufficient
morphological resemblance of prey to their habitat in order to reduce the likelihood of predator visual recognition (Donnelly and Dill, 1984). The smolt adapt to their respective background by countershading and parr marks (melanin bars or blotches) on their lateral side. Their silvery sides and ventrums of parr reflect substrate spectral characteristics and adapt to the wavelength and intensity of gravel of a given colour. The melanin of their parr marks absorbs light and enhances the resemblance of parr to the gravel substrate, especially to resemble darker substrate particles that absorb more light (Donnelly and Whoriskey, 1993). Which indicates that smolts are better adapted to darker spectral characteristics. Other research on salmonids, that have used the dark/light protocol, found similar results. Sea trout (Salmo trutta) smolt, observed in similar dark/light containers, were seemingly unaffected by other pharmaceutical pollutants (in the category of oxazepam) and spent 60 ± 4 (mean ± s.d.) percent of the time on the black substrate (McCallum et al., 2019). McCallum et al. 2019 also argued that their findings, in the dark/light protocol, are linked to fact that Sea trout smolt school together and rely on their cryptic camouflage to avoid predation. Thus, explaining why scototaxis failed to fully capture their predicted anxiety effects. This mechanism, observed in salmonids, also circles back to the theory that suggests that collective decision making of animals in groups could counter individual behavioral variance (Lü et al., 2008). However, other findings in this study reveal that the mechanisms embedded in the DNA of animals respond in dissimilar ways. On 15-05-2019, the salmon smolt, in this study, did not show a distinct preference for the black substrate. In fact, the proportion of time spent on the black substrate was significantly lower, compared to other days, which indicates a much bolder behavior. The increase in boldness could be an effect of abiotic stress, since the anxiety-like behavior of the salmon on 15-05-2019 was tested in tap water. Tap water is inherently low in predator cues (and possibly oxygen) instead of river water that
constituted the water source during the other days. Further investigation of the different water properties is necessary to assess what environmental stressor(s) caused the
difference in anxiety-like behavior. These findings could improve the dark/light protocol and avoid potential false-positives.
4.3. The use of scototaxis to guide migrating salmon: behavior
in the natural system
The results suggest that Atlantic salmon smolt can be steered towards a more favorable direction by altering the color of the bottom substrate, which is in line with my fourth hypothesis. The results from the lab confirm that the salmon smolt prefer dark substrate and that it inherently links to their crypsis. However, the effect of the color of the bottom substrate was slightly nuanced in the natural system compared to the lab. On average, 61% (see 3.3.2.) of the smolts used the dark bottom substrate to migrate downstream. Foregoing research suggests that smolts actively select their path during their
21
their downstream migration. For example, Vattenfall releases 80,000 salmon smolt every year in the Vindelälven river(Vattenfall, 2020).Assuming that the path selection of the salmon smolt is non-random, then proper use of different colors of the bottom substrate can direct an extra 11 percent towards the safety of the fishway -saving 8800 salmon smolt in the process. If the more fish-friendly Kaplan turbines are used at the hydropower dams, then the mortality of Atlantic salmon smolt lies between the 5% - 20% (Castro-Santos and Haro, 2013; Larinier, 2008; Thorstad et al., 2017). Under favorable conditions (minimal mortality and with Kaplan turbines), it would translate to a reduced loss of 1560 (5%) salmon smolts with the use of different colors of the bottom substrate. Instead of the normal predicted loss of 2000 (5%), under favorable conditions and assuming that 50% travels downstream through turbines in the hydropower dams and the other 50% through the safety of the fishway. Note that I used favorable conditions in my calculations and that the real mortality number could be substantially higher depending on the turbines and the speed of the water in the turbines. In order to extrapolate these findings to key points during the downstream migration, the scale of the experimental setup of the natural system should be increased. In other words, the water depth and length of the
experimental setup should be enlarged. This is necessary in order to understand at what depth the effect of color of bottom structure still influences pelagically migrating Atlantic salmon smolt. In any case, the results of this novel method for steering fish in natural systems is promising and should be further explored in the future.
4.4. Laboratory behaviors and its relevance for behavior in
nature
In recent years, several studies have attempted to link behavioral assays in the lab to behavioral studies conducted in natural aquatic systems, combining ecotoxicology and behavioral ecology research. These studies have shown that the complexity of real aquatic ecosystems introduces additional sources of variation, which in turn decreases the
22
Acknowledgement
This research was supported by the Department of Ecology and Environmental Science. I would like to devote this part of my research report to thank my supervisor and other involved persons who helped me realize this research.
First of all, I would like to thank Jonatan Klaminder, who fulfilled the role as supervisor. He helped me coordinate this research, assisted me when needed and provided me with great insights for my writing. I would also like to thank Johan Fahlman and Johan Leander, who assisted me during the research and gave me the data that has built this research. Besides, I would like to thank Saúl Rodriguez and Micael Jonsson for their help and support during my thesis. And last but not least, I would like to thank Viktor
23
References
Arnold, K.E., Brown, A.R., Ankley, G.T., Sumpter, J.P., 2014. Medicating the
environment: Assessing risks of pharmaceuticals to wildlife and ecosystems. Philos. Trans. R. Soc. B Biol. Sci. 369. https://doi.org/10.1098/rstb.2013.0569
aus der Beek, T., Weber, F.A., Bergmann, A., Hickmann, S., Ebert, I., Hein, A., Küster, A., 2016. Pharmaceuticals in the environment-Global occurrences and perspectives. Environ. Toxicol. Chem. 35, 823–835. https://doi.org/10.1002/etc.3339
Bates, D., Mächler, M., Bolker, B.M., Walker, S.C., 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67. https://doi.org/10.18637/jss.v067.i01 Borgatti, S., 2002. NetDraw: Graph Visualization Software. Harvard: Analytic
Technologies.
Bratrich, C., Truffer, B., Jorde, K., Markard, J., Meier, W., Peter, A., Schneider, M., Wehrli, B., 2004. Green hydropower: A new assessment procedure for river management. River Res. Appl. 20, 865–882. https://doi.org/10.1002/rra.788 Brodin, T., Bertram, M.G., Saaristo, M., McCallum, E.S., Wong, B.B.M., Balshine, S.,
Arnold, K.E., Sundin, J., Ehlman, S.M., Brooks, B.W., Sih, A., 2018. Direct and indirect effects of chemical contaminants on the behaviour, ecology and evolution of wildlife. Proc. R. Soc. B Biol. Sci. 285, 20181297.
https://doi.org/10.1098/rspb.2018.1297
Brodin, T., Fick, J., Jonsson, M., Klaminder, J., 2013. Dilute concentrations of a
psychiatric drug alter behavior of fish from natural populations. Science (80-. ). 339, 814–815. https://doi.org/10.1126/science.1226850
Brodin, T., Klaminder, J., Piovano, S., Jonsson, M., Fick, J., Heynen, M., 2014. Ecological effects of pharmaceuticals in aquatic systems--impacts through behavioural
alterations. Philos. Trans. R. Soc. B Biol. Sci. 369, 20130580. https://doi.org/10.1098/rstb.2013.0580
Brodin, T., Nordling, J., Lagesson, A., Klaminder, J., Hellström, G., Christensen, B., Fick, J., 2017. Environmental relevant levels of a benzodiazepine (oxazepam) alters important behavioral traits in a common planktivorous fish, (Rutilus rutilus). J. Toxicol. Environ. Heal. - Part A Curr. Issues 80, 963–970.
https://doi.org/10.1080/15287394.2017.1352214
Brown, R.S., Harnish, R.A., Carter, K.M., Boyd, J.W., Deters, K.A., Eppard, M.B., 2010. An Evaluation of the Maximum Tag Burden for Implantation of Acoustic
Transmitters in Juvenile Chinook Salmon. North Am. J. Fish. Manag. 30, 499–505. https://doi.org/10.1577/m09-038.1
Bunt, C.M., Castro-Santos, T., Haro, A., 2016. Reinforcement and validation of the analyses and conclusions related to fishway evalution data from Bunt et al.: “Performance of fish passage structures at upstream barriers to migration.” River Res. Appl. 32, 2125–2137. https://doi.org/10.1002/rra.3095
Burnham, K.P., Anderson, D.R., 2004. Multimodel inference: Understanding AIC and BIC in model selection. Sociol. Methods Res. 33, 261–304.
https://doi.org/10.1177/0049124104268644
Butchart, S.H.M., Walpole, M., Collen, B., Strien, A. van, Scharlemann, J.P.W., Almond, R.E.A., Baillie, J.E.M., Bomhard, B., Brown, C., Bruno, J., Carpenter, K.E., Carr, G.M., Chanson, J., Chenery, A.M., Csirke, J., Davidson, N.C., Dentener, F., Foster, M., Galli, A., Galloway, J.N., Genovesi, P., Gregory, R.D., Hockings, M., Kapos, V., Lamarque, J.-F., Leverington, F., Loh, J., McGeoch, M.A., McRae, L., Minasyan, A., Morcillo, M.H., Oldfield, T.E.E., Pauly, D., Quader, S., Revenga, C., Sauer, J.R., Skolnik, B., Spear, D., Stanwell-Smith, D., Stuart, S.N., Symes, A., Tierney, M., Tyrrell, T.D., Vié, J.-C., Watson, R., 2010. Global biodiversity: indicators of recent declines. Science (80-. ). 328, 1164–1168.
https://doi.org/10.1111/j.1757-1707.2010.01088.x
Calles, E.O., 2005. Re-establishment of connectivity for fish populations in regulated rivers.
24
Estuaries and Coasts 38, 203–214. https://doi.org/10.1007/s12237-013-9680-6 Donnelly, W.A., Dill, L.M., 1984. Evidence for crypsis in coho salmon, Oncorhynchus
kisutch (Walbaum), parr: substrate colour preference and achromatic reflectance. J. Fish Biol. 25, 183–195. https://doi.org/10.1111/j.1095-8649.1984.tb04865.x
Donnelly, W.A., Whoriskey, F.G., 1993. Transplantation of Atlantic salmon (Salmo salar) and crypsis breakdown., in: Gibson, R.J., Cutting, R.E. (Eds.), Canadian Special Publication Fisheries and Aquatic Sciences. National Research Council of Canada 1993, St. John’s, pp. 25–34.
Dudgeon, D., Arthington, A.H., Gessner, M.O., Kawabata, Z.I., Knowler, D.J., Lévêque, C., Naiman, R.J., Prieur-Richard, A.H., Soto, D., Stiassny, M.L.J., Sullivan, C.A., 2006. Freshwater biodiversity: Importance, threats, status and conservation challenges. Biol. Rev. Camb. Philos. Soc. 81, 163–182.
https://doi.org/10.1017/S1464793105006950
Fahlman, J., Fick, J., Karlsson, J., Jonsson, M., Brodin, T., Klaminder, J., 2018. Using laboratory incubations to predict the fate of pharmaceuticals in aquatic ecosystems. Environ. Chem. 15, 463–471. https://doi.org/10.1071/EN18154
Fahlman, J., Hellström, G., Jonsson, M., Veenstra, A., Klaminder, J., 2020. Six common behavioral trials and their relevance for perch performance in natural lakes. Sci. Total Environ. Accepted.
Fuller, M.R., Doyle, M.W., Strayer, D.L., 2015. Causes and consequences of habitat fragmentation in river networks. Ann. N. Y. Acad. Sci. 1355, 31–51.
https://doi.org/10.1111/nyas.12853
Ginsberg, J.R., Young, T.P., 1992. Measuring association between individuals or groups in behavioural studies. Anim. Behav. 44, 377–379. https://doi.org/10.1016/0003-3472(92)90042-8
Gresh, T., Lichatowich, J., Schoonmaker, P., 2000. An Estimation of Historic and Current Levels of Salmon Production in the Northeast Pacific Ecosystem: Evidence of a Nutrient Deficit in the Freshwater Systems of the Pacific Northwest. Fisheries 25, 15–21. https://doi.org/10.1577/1548-8446(2000)025<0015:aeohac>2.0.co;2 Hellou, J., 2011. Behavioural ecotoxicology, an “early warning” signal to assess
environmental quality. Environ. Sci. Pollut. Res. 18, 1–11. https://doi.org/10.1007/s11356-010-0367-2
Hellström, G., Alanärä, A., Persson, L., Jonsson, M., Fick, J., Finn, F., Brodin, T., Klaminder, J., 2016. GABAergic anxiolytic drug in water increases migration behaviour in salmon. Nat. Commun. 7. https://doi.org/10.1038/ncomms13460 Hoare, D.J., Krause, J., Peuhkuri, N., Godin, J.G., 2000. Body size and shoaling in fish. J.
Fish Biol. 57, 1351–1366. https://doi.org/10.1006/jfbi.2000.1446
Johnson, A.C., Jin, X., Nakada, N., Sumpter, J.P., 2020. Learning from the past and considering the future of chemicals in the environment. Science (80-. ). 367, 384– 387.
Karppinen, P., Mäkinen, T.S., Erkinaro, J., Kostin, V. V., Sadkovskij, R. V., Lupandin, A.I., Kaukoranta, M., 2002. Migratory and route-seeking behaviour of ascending Atlantic salmon in the regulated River Tuloma. Hydrobiologia 483, 23–30.
https://doi.org/10.1023/A:1021386319633
Kidd, K.A., Blanchfield, P.J., Mills, K.H., Palace, V.P., Evans, R.E., Lazorchak, J.M., Flick, R.W., 2007. Collapse of a fish population after exposure to a synthetic estrogen. PNAS 104, 8897–8901. https://doi.org/10.1073/pnas.0609568104
Klaminder, J., Brodin, T., Fick, J., Fahlman, J., Jonsson, M., Hellström, G., Bergman, E., Lagesson, A., 2016. Drug-Induced Behavioral Changes: Using Laboratory
Observations to Predict Field Observations. Front. Environ. Sci. 4, 1–9. https://doi.org/10.3389/fenvs.2016.00081
Klaminder, J., Jonsson, M., Fick, J., Sundelin, A., Brodin, T., 2014. The conceptual imperfection of aquatic risk assessment tests: Highlighting the need for tests
designed to detect therapeutic effects of pharmaceutical contaminants. Environ. Res. Lett. 9. https://doi.org/10.1088/1748-9326/9/8/084003
25
https://doi.org/10.1016/j.scitotenv.2017.09.070
Lagesson, A., Fahlman, J., Brodin, T., Fick, J., Jonsson, M., Byström, P., Klaminder, J., 2016. Bioaccumulation of five pharmaceuticals at multiple trophic levels in an aquatic food web - Insights from a field experiment. Sci. Total Environ. 568, 208– 215. https://doi.org/10.1016/j.scitotenv.2016.05.206
Larinier, M., 2008. Fish passage experience at small-scale hydro-electric power plants in France. Hydrobiologia 609, 97–108. https://doi.org/10.1007/s10750-008-9398-9 Laundre, J.W., Hernandez, L., Ripple, W.J., 2010. The Landscape of Fear: Ecological
Implications of Being Afraid. Open Ecol. J. 3, 1–7. https://doi.org/10.2174/1874213001003030001
Leander, J., Klaminder, J., Jonsson, M., Brodin, T., Leonardsson, K., Hellström, G., 2019. The old and the new: Evaluating performance of acoustic telemetry systems in tracking migrating Atlantic salmon (Salmo salar) smolt and European eel (Anguilla anguilla) around hydropower facilities. Can. J. Fish. Aquat. Sci. 77, 177–187.
https://doi.org/10.1139/cjfas-2019-0058
Lü, J., Liu, J., Couzin, I.D., Levin, S.A., 2008. Emerging collective behaviors of animal groups, in: Proceedings of the World Congress on Intelligent Control and
Automation (WCICA). pp. 1060–1065.
https://doi.org/10.1109/WCICA.2008.4593068
Lucas, M.C., Baras, E., 2001. Migration of Freshwater Fishes, Oxford UK: Blackwell Science Ltd. https://doi.org/10.1643/0045-8511(2002)002[0878:]2.0.co;2 Maximino, C., Marques De Brito, T., De Mattos Dias, C.A.G., Gouveia, A., Morato, S.,
2010. Scototaxis as anxiety-like behavior in fish. Nat. Protoc. 5, 221–228. https://doi.org/10.1038/nprot.2009.225
McCallum, E.S., Sundelin, A., Fick, J., Alanärä, A., Klaminder, J., Hellström, G., Brodin, T., 2019. Investigating tissue bioconcentration and the behavioural effects of two pharmaceutical pollutants on sea trout (Salmo trutta) in the laboratory and field. Aquat. Toxicol. 207, 170–178. https://doi.org/10.1016/j.aquatox.2018.11.028 Melvin, S.D., Wilson, S.P., 2013. The utility of behavioral studies for aquatic toxicology
testing: A meta-analysis. Chemosphere 93, 2217–2223. https://doi.org/10.1016/j.chemosphere.2013.07.036
Nagelkerken, I., Munday, P.L., 2016. Animal behaviour shapes the ecological effects of ocean acidification and warming: Moving from individual to community-level responses. Glob. Chang. Biol. 22, 974–989. https://doi.org/10.1111/gcb.13167 Nieminen, E., Hyytiäinen, K., Lindroos, M., 2016. Economic and policy considerations
regarding hydropower and migratory fish. Fish Fish. 18, 54–78. https://doi.org/10.1111/faf.12167
Nikolaou, A., Meric, S., Fatta, D., 2007. Occurrence patterns of pharmaceuticals in water and wastewater environments. Anal. Bioanal. Chem. 387, 1225–1234.
https://doi.org/10.1007/s00216-006-1035-8
Nilsson, C., Reidy, C.A., Dynesius, M., Revenga, C., 2005. Fragmentation and flow regulation of the world’s large river systems. Science (80-. ). 308, 405–408. https://doi.org/10.1126/science.1107887
Noonan, M.J., Grant, J.W.A., Jackson, C.D., 2012. A quantitative assessment of fish passage efficiency. Fish Fish. 13, 450–464. https://doi.org/10.1111/j.1467-2979.2011.00445.x
Peacor, S.D., Werner, E.E., 2003. A Review of Trait-Mediated Indirect Interactions in Ecological Communities. Ecol. Soc. Am. 84, 1083–1100.
Peter, K.T., Tian, Z., Wu, C., Lin, P., White, S., Du, B., McIntyre, J.K., Scholz, N.L.,
Kolodziej, E.P., 2018. Using High-Resolution Mass Spectrometry to Identify Organic Contaminants Linked to Urban Stormwater Mortality Syndrome in Coho Salmon. Environ. Sci. Technol. 52, 10317–10327. https://doi.org/10.1021/acs.est.8b03287 Pierce, R.B., Tomcko, C.M., 2005. Density and Biomass of Native Northern Pike
Populations in Relation to Basin-Scale Characteristics of North-Central Minnesota Lakes. Trans. Am. Fish. Soc. 134, 231–241. https://doi.org/10.1577/t03-211.1 Pyle, G., Ford, A.T., 2017. Behaviour revised: Contaminant effects on aquatic animal
26
Renöfält, B.M., Jansson, R., Nilsson, C., 2010. Effects of hydropower generation and opportunities for environmental flow management in Swedish riverine ecosystems. Freshw. Biol. 55, 49–67. https://doi.org/10.1111/j.1365-2427.2009.02241.x
Roscoe, D.W., Hinch, S.G., 2010. Effectiveness monitoring of fish passage facilities: Historical trends, geographic patterns and future directions. Fish Fish. 11, 12–33. https://doi.org/10.1111/j.1467-2979.2009.00333.x
Rothley, K.D., Dutton, G., 2006. Behavioral responses to environmental change alter direct and indirect trait-mediated interactions. Can. J. Zool. 84, 1053–1058. https://doi.org/10.1139/Z06-076
Saaristo, M., Lagesson, A., Bertram, M.G., Fick, J., Klaminder, J., Johnstone, C.P., Wong, B.B.M., Brodin, T., 2019. Behavioural effects of psychoactive pharmaceutical
exposure on European perch (Perca fluviatilis) in a multi-stressor environment. Sci. Total Environ. 655, 1311–1320. https://doi.org/10.1016/j.scitotenv.2018.11.228 Scott, G.R., Sloman, K.A., 2013. The effects of environmental pollutants on complex fish
behaviour: Integrating behavioural and physiological indicators of toxicity. Aquat. Toxicol. 68, 369–392. https://doi.org/10.1016/j.aquatox.2004.03.016
Silva, A.T., Lucas, M.C., Castro-Santos, T., Katopodis, C., Baumgartner, L.J., Thiem, J.D., Aarestrup, K., Pompeu, P.S., O’Brien, G.C., Braun, D.C., Burnett, N.J., Zhu, D.Z., Fjeldstad, H.P., Forseth, T., Rajaratnam, N., Williams, J.G., Cooke, S.J., 2018. The future of fish passage science, engineering, and practice. Fish Fish. 19, 340–362. https://doi.org/10.1111/faf.12258
Svendsen, J.C., Eskesen, A.O., Aarestrup, K., Koed, A., Jordan, A.D., 2007. Evidence for non-random spatial positioning of migrating smolts (Salmonidae) in a small lowland stream. Freshw. Biol. 52, 1147–1158.
https://doi.org/10.1111/j.1365-2427.2007.01743.x
Thorstad, E.B., Havn, T.B., Sæther, S.A., Heermann, L., Teichert, M.A.K., Diserud, O.H., Tambets, M., Borcherding, J., Økland, F., 2017. Survival and behaviour of Atlantic salmon smolts passing a run-of-river hydropower facility with a movable bulb turbine. Fish. Manag. Ecol. 24, 199–207. https://doi.org/10.1111/fme.12216 Traill, L.W., Lim, M.L.M., Sodhi, N.S., Bradshaw, C.J.A., 2010. Mechanisms driving
change: Altered species interactions and ecosystem function through global warming. J. Anim. Ecol. 79, 937–947. https://doi.org/10.1111/j.1365-2656.2010.01695.x
Tuomainen, U., Candolin, U., 2011. Behavioural responses to human-induced environmental change. Biol. Rev. 86, 640–657. https://doi.org/10.1111/j.1469-185X.2010.00164.x
Tylianakis, J.M., Didham, R.K., Bascompte, J., Wardle, D.A., 2008. Global change and species interactions in terrestrial ecosystems. Ecol. Lett. 11, 1351–1363.
https://doi.org/10.1111/j.1461-0248.2008.01250.x
van Puijenbroek, P.J.T.M., Buijse, A.D., Kraak, M.H.S., Verdonschot, P.F.M., 2019. Species and river specific effects of river fragmentation on European anadromous fish species. River Res. Appl. 35, 68–77. https://doi.org/10.1002/rra.3386
Vattenfall, 2020. powerplants.vattenfall.com [WWW Document]. Stornorrfors. URL https://powerplants.vattenfall.com/stornorrfors (accessed 4.30.20).
Verlicchi, P., Al Aukidy, M., Zambello, E., 2012. Occurrence of pharmaceutical
compounds in urban wastewater: Removal, mass load and environmental risk after a secondary treatment-A review. Sci. Total Environ. 429, 123–155.
https://doi.org/10.1016/j.scitotenv.2012.04.028
Weis, J.S., Smith, G., Zhou, T., Santiago-Bass, C., Weis, P., 2001. Effects of Contaminants on Behavior: Biochemical Mechanisms and Ecological Consequences. Bioscience 51, 209. https://doi.org/10.1641/0006-3568(2001)051[0209:eocobb]2.0.co;2
27
Appendix
Figure S1: Spatial variation in accuracy of the positions generated by the acoustic telemetry system used in the
study. The numbers represent the median distance (m) between estimated positions and known locations of fixed sync- and reference transmitters, and can be used to make inference on the positional accuracy of fish positions (for which true positions are not known).
