Research articlesFeeding in murky waters: acclimatization and landmarks improve foraging efficiency of zebrafish (Danio rerio) in turbid waters
Abstract
Fish inhabiting human-dominated ecosystems are prone to altered sensory environments in which they must live and function. Increased turbidity is one such change that they must deal with. We tested whether an increase in water turbidity and the presence of visual landmarks (coloured stones) affect the foraging efficiency of wild zebrafish. We also tested the influence of extended exposure to differing turbidity levels on the subsequent foraging efficiency of acclimatized individuals. Feeding latency (time taken to find food) increased significantly with increase in turbidity levels from a minimum of 4 s to ca 300 s. However, extended exposure of fish to varying levels of turbidity decreased feeding latencies in acclimatized conditions, indicating that acclimatization to the immediate visual environment plays an important role in determining foraging success. Most significantly, we found that feeding latencies in turbid conditions decreased significantly if visual landmarks were present. This demonstrates that zebrafish use visual landmark cues to navigate to foraging sites when visibility is impaired. This study has important implications on the role of behavioural plasticity and spatial learning in animals that allow them to cope with altered sensory environments such as episodes of enhanced turbidity that could be natural or anthropogenic.
1. Introduction
Fish rely heavily on vision for acquiring information related to communication and foraging [1]. Increased turbidity due to natural and anthropogenic processes can significantly alter the visual environment of aquatic ecosystems [2–6]. In such environments, foraging efficiency can reduce due to poor prey detection [7] and prey capture [8,9]. This can occur when the maximum distance at which a fish reacts to its prey (reactive distance) is reduced, giving additional time for prey to escape [10,11]. Therefore, energetic costs to foragers increase with increased time spent in searching for food [12]. Furthermore, a study on cichlid fish species in Lake Victoria showed that high turbidity impaired mate choice decisions by limiting colour vision in cichlid fishes, resulting in disruption of reproductive isolation between closely related species [13]. Thus, turbidity can impact both survival and prospects of reproduction of fishes.
Yet, several aquatic systems are prone to a periodic or sustained rise in turbidity and dispersing from such environments, though the easiest solution, may not always be feasible. Behavioural plasticity provides a rescue mechanism for animals under such circumstances. It is a type of phenotypic plasticity by virtue of which an organism exhibits an altered behavioural response to changed environmental conditions, thereby increasing the survival advantage of the animal [14–17]. In altered visual environments, animals may simply acclimatize to the change over time, use additional cues such as landmarks [18] or switch to a different sensory modality to operate in changed conditions [19]. The use of visual landmarks for spatial orientation in clear water has been demonstrated in several aquatic animals [20–25]. However, their role in improving foraging efficiency in turbid environments has seldom been tested (but see [18]).
Zebrafish (Danio rerio) is a tropical freshwater fish that relies on vision for foraging and predator avoidance [26]. It is widely distributed in the Indian subcontinent and occurs naturally in a variety of habitats differing in water clarity ranging from clear streams to waterbodies in human-dominated landscapes [27,28]. Thus, to investigate how turbidity impedes visual tasks in fishes and how they cope with it, zebrafish serves as a good model system. In the present study, we examined the impact of turbidity on the foraging efficiency of zebrafish and whether acclimatization to turbid conditions mitigates the impact. We also examined the effect of the presence of familiar visual landmarks on their ability to find food in high turbid conditions.
2. Methods
(a) Collection and training
A total of 160 wild zebrafish were collected using drag fishing from a clear stream near Kolkata and transported to IISER Mohali. They were initially housed for a month in a collection tank filled with clear water (CW) up to 25 cm (66 l glass tank; 63 × 33 × 32.5 cm3; pH 8.42; temperature approx. 24°C; 12 : 12 light-dark cycle) to eliminate the possibility of stress due to relocation. Fish were fed on fish pellets and artemia, constant aeration was provided by a pump and tank walls were covered with white sheets to keep fish visually isolated from external visual cues. At the end of this period, fish were trained for two weeks in tanks filled up to 10 cm (47.5 × 27.5 × 27.5 cm3; water and ambient conditions same as collection tank) in batches of 25 individuals. Tanks were partitioned into two equal halves using a removable transparent plexiglass (electronic supplementary material, figure S1). Food was provided inside a feeder (white plastic ring, 3 cm in diameter) on one side of the partition while fish were released on the other side. A training session commenced when the temporary partition was removed, allowing fish to navigate to the feeder and extract the food pellets.
(b) Experiments
Experiments were carried out between September 2015 and February 2016 between 11.00 to 16.00.
| (1) | Effect of turbidity and acclimatization: after training, 75 similar-sized adult fish were acclimatized for a month in three glass tanks (25 fish per tank; tank dimensions and ambient conditions same as collection tank) having different turbidity levels: CW, low turbid (LT) and high turbid (HT). Water was made turbid using kaolin powder (Sigma-Aldrich). Beam attenuation value of CW, LT and HT were maintained at 0.5, 3 and 8 m–1, respectively. TLT condition mimicked the turbidity levels of coastal water, while the HT condition mimicked turbidity of an estuary mouth. After a month of acclimatization, fish from each of the three different acclimatizing conditions were exposed to each of the three test conditions: CW, LT and HT to test the effect of acclimatizing environment on foraging efficiency (N = 30 trials/case). Initially, fish that were acclimatized to CW were tested in three different test conditions to examine whether turbidity affected foraging efficiency at all. This was followed by the other two sets of experiments. During experimental trials, five fish were released together into the tank and were given 5 min to acclimatize, followed by dropping of five pellets inside the feeder and removing the plexiglass partition, thereby marking the start of a trial. Foraging efficiency was determined by measuring feeding latency (time lapsed before at least one fish reached the feeder and captured at least one pellet of food). The cut-off time for each trial was set at 5 min. If the fish failed to forage within 5 min, feeding latency for that trial was recorded as 300 s (to assign a numerical value to the latency). Additionally, the frequency of successful and unsuccessful foraging attempts was recorded. Responses of fish were videoed (Sony Cyber-shot DSC-HX400 V, Sony Corporation, Japan). After every trial, the used fish were released into a holding tank containing the same water conditions as what the fish were acclimatized to. Once all 25 individuals were tested, they were released back into their respective tanks until subsequent tests were conducted. Five trials were carried out per day and subsequent trials using previously tested fish were carried out after a break of at least 48 h. To control for any learning due to the reuse of individuals, a series of tests were carried out to examine the effect of reuse of fish on their subsequent foraging efficiency. No effect of reuse of individuals on subsequent foraging efficiency was found (electronic supplementary material, figures S2 and S3). | ||||
| (2) | Effect of landmarks: 50 adult fish (naive individuals not used in previous experiments) were kept for a month in two different conditions (25 each), CW and HT, with five coloured stones (three red and two white, kept in the same configuration for all trials) placed below the feeder. The rationale for using coloured stones was that such stones occur naturally in aquatic habitats. Moreover, red colour is likely to be visible from a distance even in high turbid conditions, thereby serving as a reliable visual landmark. After training, fish were released into test tanks with and without landmarks with the same water condition (CW or HT) as what they were acclimatized to (N = 15 trials/case) and feeding latency was measured. | ||||
(c) Statistical analysis
Statistical tests were performed using Statistica 64 (Dell Inc.2015, v.12). Data were checked for normality using the Shapiro–Wilk test. Data for experiment 2 were log-transformed for normality before conducting Student's t-test. Data for experiment 1 did not follow normal distribution even after transformation. Hence, non-parametric Kruskal–Wallis test followed by Mann–Whitney U-test for pairwise comparisons were done applying Bonferroni correction adjusted to p < 0.017 (a/k = 0.017, where a = 0.05, k = 3 comparisons).
3. Results
(a) Effect of turbidity and acclimatization
There was a significant effect of increased turbidity on foraging efficiency (Kruskal–Wallis, H = 59.16, d.f. = 2, N = 90, p < 0.001; electronic supplementary material, table S1a). Pairwise comparisons showed significant differences in foraging efficiency between all three test conditions wherein fish had significantly lower feeding latency in CW (condition to which they were acclimatized) than in LT and HT conditions (Mann–Whitney U-test, p < 0.001, table 1a and figure 1a). Furthermore, there was a significant effect of acclimatization on the feeding latency (electronic supplementary material, table S1a,b,c). Pairwise comparisons revealed that the performance of fish was significantly better in the water conditions to which they were acclimatized (Mann–Whitney U-test, p < 0.001, table 1a, b & c; figure 1a; electronic supplementary material, figure S4).
Figure 1. Box plots (median; box 25–75%, whiskers: non-outlier range). (a) Effect of acclimatizing environment on feeding latency of zebrafish acclimatized to three conditions (clear water, low turbid and high turbid condition) and tested in each of these three conditions (N = 30 trials/case). Asterisks signify differences between treatments (Mann–Whitney U-test for pairwise comparisons with Bonferroni correction, p < 0.017). (b) Effect of landmark on feeding latency of zebrafish in clear water and turbid water conditions in the presence and absence of landmarks, N = 15 trials/case. Student's t-test, * signifies p < 0.05. NS indicates no significant difference.
| acclimatizing condition | test condition pairs | Mann–Whitney |
p-value |
|---|---|---|---|
| (a) clear water (CW) | CW versus LT | U = 82 | <0.0001 |
| CW versus HT | U = 12 | <0.0001 | |
| LT versus HT | U = 165 | <0.0001 | |
| (b) low turbid water (LT) | CW versus LT | U = 48 | <0.0001 |
| CW versus HT | U = 403 | NS | |
| LT versus HT | U = 75 | <0.0001 | |
| (c) high turbid water (HT) | CW versus LT | U = 273 | <0.01 |
| CW versus HT | U = 247 | <0.01 | |
| LT versus HT | U = 442 | NS |
(b) Effect of landmarks
Presence of visual landmarks significantly improved performance of fish that were acclimatized and tested in the HT condition (t-test: t = 2.54, d.f. = 28, p < 0.01; figure 1b) but not for fish that were acclimatized and tested in the CW condition (t-test: t = 1·29, d.f. = 28, p = 0.20; figure 1b).
4. Discussion
Our findings show a substantial decrease in foraging efficiency in zebrafish with an increase in turbidity levels. While a few studies have shown that increased turbidity does not alter foraging behaviour [29,30], or in some cases, improves foraging efficiency [31,32], the bulk of findings in the literature are in support of our results. Studies on a variety of fish species have demonstrated that even a slight rise in turbidity can significantly impact their foraging behaviour. This includes decreased prey consumption in fountain darters [33] and pinfish [8], decreased prey searching activity in juvenile cod [34], reduced prey detection in lake trout [7] and decreased foraging success in mandarin fish [9].
A study on guppies demonstrated developmental plasticity in turbid conditions, wherein, fish reared to adulthood in turbid water were more active when tested in turbid compared to CW [35]. In this study too, turbidity had an immediate negative effect on foraging efficiency in zebrafish. However, once acclimatized to turbid conditions for just one month, fish performed significantly better in turbid environments even compared to CW. This signifies behavioural and not developmental plasticity as they belonged to the same population. Behavioural plasticity in zebrafish has been demonstrated in an earlier study in laboratory-bred, pond and lake populations that were subjected to altered aquatic environments with differing flow and vegetation regimes [17].
Our study also gives the first evidence of the use of familiar visual landmarks by zebrafish to forage in turbid conditions. The improved performance of fish in the presence of landmarks in turbid, but not in CW demonstrates the utilization of additional cues to deal with disrupted sensory information without switching to a different sensory modality. In Amarillo fish, it was demonstrated that fish caught from both turbid water and CW used visual landmarks for spatial navigation. However, individuals from turbid water showed better navigational performance than those from CW, in the presence of landmarks [18]. This could be indicative of population-level differences that may have evolved as a response to altered sensory environments which they inhabit. Our study is notably different from this in two ways. First, in our study, all fish were caught from the same population of a stream and only acclimatized to different conditions for a month prior to the tests. This rules out any evolved difference in response due to long-term genetic changes. Second, in our study, the performance of fish acclimatized to CW did not change in the presence of cues. Both these put together give strong evidence for behavioural plasticity under altered environmental conditions. Significantly, a one-month period in an altered environment was sufficient to evoke these behavioural differences in zebrafish.
It is evident that zebrafish learnt to use the coloured stones as a proxy for feeder location during the training. However, this learning was not of much consequence in CW conditions where the feeder was clearly visible. In high turbid conditions, however, we expected the white plastic ring feeder to be difficult to locate by the fish. Yet, in the presence of landmarks, they located the feeder faster than when landmarks were missing. This clearly demonstrates the significance of the visual landmarks that mediated spatial learning in zebrafish.
Our findings open new avenues of investigation to examine behavioural plasticity in wild populations of fish that inherently possess broad fundamental niches with respect to visual environments, but are driven to narrow realized niches due to anthropogenic impact on their sensory environment.
Ethics
Zebrafish is designated as ‘Least Concern’ by IUCN's Red List of Threatened Species. The study complied with existing rules/guidelines outlined by the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA), Government of India. No animals were euthanized or sacrificed or subjected to any chemical treatment during the study. At the end of the experiments, all animals were returned to stock tanks and continued to be maintained in the laboratory.
Data accessibility
Data for this study are available from the Dryad Digital Repository at: https://doi.org/10.5061/dryad.hb3f25t [36]. Additional data are made available as electronic supplementary material.
Authors' contributions
M.J. and A.B. conceived the study and designed the experiments. M.A.S. carried out all the experiments. M.A.S. and R.S. carried out all analyses. R.S. and M.J. wrote the paper, A.B. and M.A.S. helped with editing and finalizing it. All authors approved the final version of the manuscript and agree to be accountable for the content therein.
Competing interests
We declare we have no competing interests.
Funding
This work was funded by IISER-Mohali.
Acknowledgements
We thank local fishermen for providing zebrafish, IISER-Kolkata for hosting M.A.S. for training and Danita Daniel for conducting additional control experiments. We thank Rittik and the anonymous reviewers for their comments that greatly improved the manuscript.
