Interplay between antipredator behavior, parasitism, and gut microbiome in wild stickleback populations

doi:10.21203/rs.3.rs-5263430/v1

Download PDF

Research Article

Interplay between antipredator behavior, parasitism, and gut microbiome in wild stickleback populations

https://doi.org/10.21203/rs.3.rs-5263430/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

The role of predation stress in driving behavioral and microbial-host interaction changes is recognized, but the impact of microbial composition in aquatic organisms and its relationship with stress-related behavior remains poorly understood. This study explored the relationship between antipredator behavior, parasitism, and the gut microbiome in wild stickleback fish populations from two different lakes: Galtaból, clear and spring-fed versus Þristikla, turbid and glacial-fed. We aimed to identify potential links between these factors by analyzing behavioral responses to simulated predation, comparing microbiomes between populations with or without parasite infection, and examining potential correlations between behavior and microbiome composition.

Results

Behavioral analysis revealed differences between populations, with each exhibiting unique baseline behaviors i.e., higher activity in Galtaból fish and higher angular velocity in Þristikla fish, and varied responses to the presence of predator i.e., increased boldness in Galtaból fish and higher activity in Þristikla fish. The response to the predator attack was similar between fish from both populations. Parasitism influenced behavior, with parasitized fish displaying increased boldness. Microbiome analysis showed that a small proportion of its variation was explained by population, likely reflecting differences in lake environments. Only the marine genus Pseudoalteromonas abundance differed between populations. Parasitism in Galtaból fish population was linked to decreased alpha diversity in the microbiome, with an increase in specific microbial taxa, including potential pathogens, and a decrease in commensal microbes.

Conclusions

Our findings suggest that behavior and microbiome correlations may primarily reflect environmental adaptations and parasite status rather than direct gut-brain interactions. However, some tentative evidence suggests a potential innate connection between swimming activity, stress levels, and specific microbes. The study highlights the complexity of the gut-brain axis in wild populations and suggests future research directions, including experimental manipulations to uncover causal relationships between microbiome composition and behavior.

microbiome

gut-brain axis

stress

predation

Sticklebacks

parasites

boldness

activity

Stress is pervasive in natural environments, shaping ecosystems and their inhabitants. Habitats globally experience constant change and encounter a variety of stressors, which in turn influence animals and select individuals capable of responding to and tolerating environmental changes (Schulte 2014). Stressors, originating from both anthropogenic and natural sources such as predation can impact phenotypic variation, reduce structural complexity, and diminish diversity in ecosystems (Jenkins et al. 2021; Fakan et al. 2023). Changes in environments might affect the response of organisms to predation stress and other environmental stress. Predation stress is recognized to drive the evolution of trait change patterns in organisms, including behavioral responses and microbial-host interactions. However, the impact of the composition and dynamics of the microbes that inhabit aquatic organisms remains inadequately understood (Fakan et al. 2023). Similarly, our understanding of the relationship between stress-related changes in behavior and the microbiome is limited(K. R. Foster et al. 2017; J. A. Foster, Rinaman, and Cryan 2017; Johnson and Foster 2018).

Predation is a key selective force driving the evolution of various antipredator adaptations in prey, such as morphology(Eklöv and Svanbäck 2006; Lapiedra et al. 2018) behavior, physiology(Rödl et al. 2007) and life history traits. These often co-vary(Endler 1983) and determine an individual's fitness. Chemical and visual cues from predators have been shown to prompt changes in the spatial behavior of juvenile fish, leading to decreased foraging and exploratory swimming. This has been observed in species such as sticklebacks (Gasterosteus aculeatus), perch (Perca fluviatilis), and rainbow trout (Oncorhynchus mykiss) (Milinski and Heller 1978; Fraser and Gilliam 1992; Diehl et al. 1995; Brown and Smith 1998; Lehtiniemi 2005; Kopack et al. 2015). A direct encounter with a predator usually includes a visual cue, which serves as a spatially and temporally reliable indicator of the presence of a risk, thereby indicating a high predation risk(Hall and Clark 2016). The antipredator behavior of fish depends on many factors, such as the distance to the predator, the behavior of the predator, the species of predator, group size, and various environmental factors (Cerri 1983; Lima and Dill 1980; Godin and Valdron Clark 1997), as well as evolutionary history of predation risk. Behavioral responses to predation have been extensively studied in stickleback fish. This species exhibits a range of antipredation behaviors, including freezing, adopting a defensive posture by swimming slowly with dorsal and pelvic fins or spines fully extended, and actively approaching the threat in a deliberate, slow manner to deter the predator or assess its intentions (Näslund, Pettersson, and Johnsson 2017). Alternatively, they may employ escape responses such as retreating, jumping, swimming away, or freezing(Huntingford and Giles 1987) and fast-start (Bergstrom 2002).

Parasites are considered environmental stressors that impose fitness costs on their hosts, and play an important role in the evolution and ecology of many fish phenotypic traits, such as morphology, physiology, and behavior (Bush et al. 2001). Behavioral changes, known as host manipulation, are extensively described in the literature (Poulin 1995; Thomas, Adamo, and Moore 2005). They include changes in foraging efficiency, time budget, habitat selection, competitive ability, predator-prey relationships, swimming performance, sexual behavior, and mate choice (reviewed in (Barber, Hoare, and Krause 2000)).

Parasites can impact fish swimming behavior in various ways, including quantifiable reductions in swimming performance, such as speed and stamina, or increases in conspicuous behaviors, such as erratic movements (Barber, Hoare, and Krause 2000). The underlying mechanisms may involve atrophy of musculature or pathology of the nervous system (Barber 2007). Additionally, parasites may increase the energetic cost of locomotion by affecting the hydrodynamic properties of fish, including buoyancy or buoyancy control (Barber 2007; Barber, Hoare, and Krause 2000; LoBue and Bell 1993).

The anti-predator response in parasitized fish can be affected by reducing spatio-temporal overlap with their predators (Barber, Hoare, and Krause 2000). For example, infected three-spined sticklebacks (Gasterosteus aculeatus) return to food patches more quickly than uninfected conspecifics following a simulated predator attack (N Giles 1983). Predator avoidance can be reduced through different mechanisms, such as a decreased propensity for infected fish to join and remain with shoals(Barber, Hoare, and Krause 2000) or a reduction in the efficiency of the escape response. For example, the effects of parasite load on the fast-start performance of the three-spined stickleback are associated with negative impacts on escape kinematics (Blake, Kwok, and Chan 2006), while also increasing cortisol levels, indicating a primary stress response (Demandt et al. 2020).

In addition to the above effects, the microbiome of the fish can also be affected by physiological changes in response to predation and parasite stress. The microbiome plays key roles in the associated host, including involvement in the immune response, development, production of key metabolites, and even function of vital organs such as the brain (Marchesi et al. 2015; Varg, Bergvall, and Svanbäck 2022). Recently, there has been an increase in the literature on the effect of gut microorganisms on the host’s brain and behavior. For example,Collins & Bercik, 2009 show that the nerve system, as well as the immune system, can play an important role in the regulation of the host microbial communities and vice versa. The main mechanism of microorganisms interacting with the brain is based on microorganism-derived neuroactive chemicals (Johnson and Foster 2018). The coined name for this interaction is microbiota-gut-brain-axis. Some studies have shown that the microbiome can affect host behavior, likely mediated through this microbiota-gut-brain-axis (Vuong et al. 2017). For example, in a zebrafish model, a probiotic supplement increased exploratory behaviour and protected fish against stress-induced changes in the microbiome gut (D. J. Davis et al. 2016). However, associations between microbiome changes and host behavior have been poorly studied in wild systems. Thus, while the importance of the microbiome in the host is unmistakable, more knowledge is needed regarding links between microbiome composition and host behavior in a variety of systems and under different stressors, such as predation or parasite infection.

The gut microbiome has been shown to play an important role during parasite infections. The gut microbiome can promote or inhibit parasite colonisation and reproduction (White et al. 2018; Plieskatt et al. 2013; Jaenike et al. 2010), while parasite exposure and infection can disrupt the normal gut microbiome composition (Hahn et al. 2022; Gaulke et al. 2019; Williams et al. 2021). Parasites can interact directly with gut microbes as they transverse the gut, but they can also indirectly modify the gut microbiome through interactions with the host immune system (Reynolds, Finlay, and Maizels 2015), even when residing in other body sites, e.g. the body cavity of intermediate hosts (Hahn et al. 2022). Parasites can also harbour their own microbes that may interact with the vertebrate host microbiome (Landmann et al. 2011; Jorge, Dheilly, and Poulin 2020; Brealey et al. 2022; Kashinskaya et al. 2023) and act as vectors for introducing novel, potentially pathogenic microbes to the host animal (Vaughan, Tkach, and Greiman 2012; Dessì et al. 2006; Deenonpoe et al. 2015). Together these studies highlight the importance of including parasite status in microbiome studies of wild populations, to disentangle interactions between parasite, host microbiome and the environment.

Substantial research shows that threespine sticklebacks show extensive geographic variation among populations in a multitude of phenotypic traits and are often genetically differentiated (M. A. Bell and Foster 1994).This includes variation among populations in predation regimes and anti-predator responses (Nick Giles and Huntingford 1984; Ålund et al. 2022), as well as the overall incidence of parasitism and which parasites typically infect them (Young and MacColl 2016). Recently, population variation in the microbiome and its role in adapting to distinct environments has received attention(Rennison, Rudman, and Schluter 2019; Härer et al. 2023) including adapting to varying parasite communities (C. C. R. Smith et al. 2015; Ling et al. 2020). Exposure to predators and parasites varies for populations that are ecologically differentiated, and such differences are likely to affect both anti-predator behavior and the microbiomes associated with hosts.

We studied two populations of threespine sticklebacks inhabiting ecologically differentiated lakes in Iceland. One lake, Galtaból, is fed by springs and is quite clear while the other lake, Þristikla, is fed by glacial melt and is quite turbid, reducing visibility. The sensory environment in these two lakes is thus quite distinct and expected to alter both the detection of predators and escape behavior (Ajemian, Sohel, and Mattila 2015; Ålund et al. 2022). Other aspects of ecology are also likely to differ between lakes (Natsopoulou, Pálsson, and Ólafsdóttir 2012; Laspoumaderes et al. 2013; Lucek et al. 2016); thus we reasoned that the two populations might differ in their anti-predator behavior, parasitism, and their microbiome (Shankregowda et al. 2023).

We explored a set of questions to further understand the association of behavior and microbiomes in populations adapted to diverse environments. First, we asked whether glacial and spring-fed populations differ in their response to predatory attack, including their escape behavior, activity and boldness. Next, we asked whether the composition or diversity of host microbiome differed for glacial and spring-fed populations. We also asked whether the microbiome for parasitised and non-parasitized fish differ. And last, given the important functional role of the microbiome for host stress response, we explored whether aspects of anti-predator behavior and the microbial community were related. To address these questions we simulated predatory attack on threespine sticklebacks from these two ecologically differentiated populations, observed their behavioral responses, recorded the presence of major parasites, and characterized the gut microbial community for each fish in our experiment. We tested a set of predictions relating to the questions posed above. First, we predicted that Galtaból and Þristikla fish would differ in their anti-predator behavior, including their personality traits, and that given the clear-water environment of Galtaból, those fish would show enhanced anti-predator responses. Next, given the differences in lake ecology we predicted that the two populations would differ in their microbial community, and also that microbiomes would differ for parasitised and non-parasitized fish. Lastly, we predicted that anti-predator behavior would be related to the gut microbial community in some manner.

2.1 Sampling and fieldwork

We collected stickleback with minnow traps from both lakes on 14 June 2019, and transported them to the fish facility at Verið aquaculture station of Hólar University (Saudárkrókur, Iceland), where they were housed by population, with 30 fish in each 19 l aerated bucket, at 12 degrees C , which is close to the temperature of the lakes. Fish were fed ad libitum once per day with frozen bloodworms. The experiment started on 21 June 2019 and ended on 13 July 2019. We selected subjects in the mornings before feeding and moved them from their home tanks to ensure their guts were free of digesting food before behavioral trials and gut dissection.

2.2 Behavioral experiments

Food-deprived adult sticklebacks for one day were recorded for behavior in tanks of 150 x 50 cm with a water depth of 21 cm and a distance from the camera of 155 cm. The experiment was divided into two treatments: predation stress and control treatment. For the predation stress group, the stickleback fish had two minutes of adjustment time to behave normally in the tank. After that, the activity of the fish was observed for 4 minutes. A Silicone model fish was used as a visual predator simulating an attack. The predation cue was triggered when the stickleback fish passed by in front of the silicone fish. After triggering the predation cue, the behavior was recorded for 8 minutes. In a similar way, for the control group, the stickleback fish had two minutes of adjustment to behave normally in the tank. Then, the activity was recorded and followed for 4 minutes. The behavior was recorded for an extra 8 minutes in the same way as the treatment group but there were no silicone fish in the tank therefore no predator cue to be triggered. For both treatments, the stickleback fish was measured and dissected to extract the gut for further analyses.

2.3 Parasite identification

Parasites were identified by visual inspection during the dissection for gut extraction, focusing on a major stickleback parasite in the body cavity, Schistocephalus solidus. Parasitism was higher in Galtaból compared to Þristikla (27.3% vs 15.4%), resulting in an unbalanced study design when testing for associations with parasite status. Final sample numbers for the behavioral analyses were as follows: 40 Galtaból fish in the control group (30 non-parasitized, 10 parasitized), 48 Galtaból fish in the predator-exposed group (34 non-parasitized, 14 parasitized), 38 Þristikla fish in the control group (38 non-parasitized, 0 parasitized) and 40 Þristikla fish in the predator-exposed group (28 non-parasitized, 12 parasitized).

2.3 Characterization of behaviour

2.3.1 Arena for control condition

The arena was divided into five virtual zones (Figure 1A) using the software EthoVision XT 15 (Noldus, The Netherlands) :

-Three equal-sized zones : Start (left part of the arena where the tested fish is placed at the beginning of experiment), Opstart (right part of the arena at the opposite side of start), Middle (zone located between Start and OpStart)

-Two additional zones : Center (The centre zone was considered a risky area, a common measure indicative of a high degree of boldness in such an apparatus), Border (this zone is associated with thigmotaxis that is staying close to the walls of an arena, a common measure indicative of a high degree of shyness in such an apparatus)(Dahlbom et al. 2011). The variables of interest extracted with EthoVision XT 15 were as follows: i)The time spent in each zone previously described, ii)The distance travelled by each fish in the device (Dtot in mm), iii)The absolute angular velocity of the fish expressed in degrees per second (Vang in ° s−1), and its mean velocity expressed in body length per second (Vel in BL s−1).

2.3.2 Arena for treatment condition

Two reversed arena settings were designed in Ethovision for the situations before and after triggering the fake predator. So, we only describe the situation before triggering the predator (Figure 1B).

The same five zones as described for the control were also used here. In addition to these zones, we defined Pred (hidden zone under the fake predator) and Entry (zone located all around the predator). The variables of interest extracted with EthoVision XT 15 were as follows: i)The time spent in each zone previously described, ii)The time spent in Pred and Entry, iii)The distance travelled by each fish in the device (Dtot in mm), iv)The absolute angular velocity of the fish expressed in degrees per second (Vang in ° s−1), and its mean velocity expressed in body length per second (Vel in BL s−1).

2.4 Nucleic Acids Extraction, Library Preparation and Sequencing

For nucleic acid extraction, whole guts previously stored in RNA later were homogenized and lysed with a TissueLyser for 1.5 min at 25–30 Hz in 1.6 ml Buffer RLT (Qiagen). DNA and RNA were then simultaneously extracted from 450 µl of lysate per sample using the Allprep® DNA/RNA Mini kit (Qiagen, ID/No:80204) following the manufacturer's protocol. The extractions were performed in six batches and each batch included a blank negative control containing only Buffer RLT, which were taken through all stages of library preparation and sequencing.

DNA extract concentrations and fragment length distributions were quantified. Based on DNA quantity and quality, 71 gut extracts were selected for sequencing, including 27 non-parasitized Galtaból individuals, 13 parasitized Galtaból individuals and 31 non-parasitized Þristikla individuals. The parasitized Þristikla group was excluded from sequencing due to low sample numbers (only four extracts were available). TruSeq Nano DNA libraries (Illumina) were prepared and sequenced by SciLifeLab Uppsala on two lanes of a Illumina NovaSeq 6000 SP flowcell using paired-end 150 bp read length v1 sequencing chemistry.

2.5 DNA Sequence data processing, taxonomic assignment and taxa filtering

Adapter and quality trimming of sequenced DNA reads was performed using AdapterRemoval v.2.2.4 (Schubert, Lindgreen, and Orlando 2016), trimming consecutive bases with quality scores of <30 and removing reads <50 bp after trimming. Paired reads from both lanes were concatenated into a single fasta file per sample. Reads with mean base quality < 30 and exact PCR duplicates (original or reverse complement) were filtered out with PrinSeq-Lite v0.20.4 (Schmieder and Edwards 2011). Reads resulting from the phage PhiX, a positive control spiked in during Illumina sequencing, were removed by mapping to the PhiX reference genome (GCF_000819615.1) with bwa mem v0.7.17 (Li 2013). The unmapped reads were retained with SAMTools v1.9 (Li et al. 2009) and BEDTools v2.27.1 (Quinlan and Hall 2010) downstream analyses. In the same manner, reads from human contamination were removed by mapping to the Homo sapiens reference genome (GCF_000001405.38) (Schneider et al. 2017) and reads from the host stickleback were removed by mapping to the G. aculeatus reference genome (GCF_016920845.1) (Nath, Shaw, and White 2021).

Microbial taxonomic assignment was then performed on the unmapped, non-host reads, using the k-mer based classifier Kraken2 v2.0.8 (Wood, Lu, and Langmead 2019) with the standard Kraken2 database (all archaea, bacteria, viruses and the human genome in RefSeq; built 2020-10-01) and default parameters. Bracken v2.0 (Lu et al. 2017) was used to estimate taxa abundances from the Kraken read assignments at the species level (-l S) using a read length of 150 bp (-r 150), a k-mer length of 35 bp (-k 35) and without an abundance threshold (-t 0). Kraken-biom (https://github.com/smdabdoub/kraken-biom) was used to extract the summarized abundances assigned at the species levels.

Quality control was then performed on the taxa abundances in RStudio v402 using R v4.3.1. To reduce noise, taxa present at < 0.05% relative abundance (Bracken abundance divided by sum of Bracken abundance in a sample) were filtered out. The community compositions of the blank control samples were then compared to the stickleback gut samples, with the aim to identify and remove putative laboratory contaminants. All six negative blank controls included in the DNA extractions had low numbers of microbial reads (mean: 403, range: 113–722). One blank had a microbial community more similar to the fish gut samples than the other blanks (Supplementary Figure S1). Since there was no clustering by extraction batch (Supplementary Figure S2), we determined that low-level cross-contamination between samples and the blank had occurred during this extraction batch. We therefore excluded this blank from contamination identification. Using the other five blanks, contaminants were identified using the decontam ‘prevalence’ function (N. M. Davis et al. 2018) with the default threshold of 0.1, resulting in 134 taxa identified as contaminants and removed from the dataset. Using the decontam ‘frequency’ function (N. M. Davis et al. 2018) with DNA extraction concentration, an additional 38 taxa were identified and removed. The final taxa abundances were summarized at the genus-level for subsequent statistical analysis.

2.6 Statistical analysis

2.6.1 Behavior

To test if there were differences between fish populations in the control treatment and between the control and the period before triggering the robotic predator, a linear model was performed. First, to test the swimming activity, we used Total Distance Swum, Velocity, and Angular Velocity as response variables. Second, to test tank use, Center Duration, and Border Duration were the response variables. For both swimming activity and tank use, the fixed factors included Population (Galtaból and Þristikla), Type (Control and Predator) and their interaction (Population by Type). Posthoc Tukey tests were used using the emmeans R package to observe differences between treatments (Lenth et al. 2020). This analysis is a post hoc of multiple comparison test on the population by type interaction (Hothorn et al. 2006) to assess only biologically meaningful pairwise differences between populations, i.e., Galtaból-Control vs Þristikla-Control; Galtaból-Predator vs Þristikla-Predator; Galtaból-Predator vs Galtaból -Predator; Þristikla -Predator vs Þristikla -Predator.

To further test the differences in behavior under the robotic fish stress, a linear mixed effect model was performed using Total Distance swum, Velocity, and Angular Velocity as response variables of Swimming activities parameters. Center duration and Border duration were used as well as response variables to test the fish boldness. Population (Galtaból and Þristikla), Period (Before and after triggering the robotic predator), and Parasite (presence-absence) and their interactions were used as fixed factors. The individual fish ID was used as a random effect factor. Diagnostics based on residuals of the model were performed to assess the adequacy of the reduced model and compliance with the underlying assumptions. A contrast list post hoc test was performed as previously described by using the R package emmeans (Lenth et al. 2020). This analysis is a multiple comparison post hoc test on the population by Period by Parasite interaction (Hothorn et al. 2006)to assess only biologically meaningful pairwise differences between populations, Period and Parasite (full results reported in Supplementary Tables S1-2).

The dependent variables were transformed whenever necessary to ensure that the residuals followed the assumed error distribution. Finally, the effects of the independent variables were estimated from the models and their significance was tested by likelihood ratio tests (LRT) between models, respecting the marginality of the effects that are supposed to follow a chi-2 distribution under the null hypothesis (type II tests; (Fox and Weisberg 2011)).

2.5.2 Microbiome

Alpha diversity metrics were calculated from Braken relative abundances using the Hill numbers framework with the R package hillR (Alberdi and Gilbert 2019), using q = 0 to estimate richness and q = 1 to estimate the Shannon index. Statistically significant differences (p < 0.05) in alpha diversity metrics between groups were evaluated using Wilcoxon rank-sum tests implemented in the R package ggsignif (Ahlmann-Eltze and Patil 2021). For unsupervised analyses, Bracken abundance counts were normalized by the center-log ratio (CLR) transformation, using a pseudocount of 1 added to all taxa in all samples to resolve the problem of zero values. Euclidean distance matrices were calculated from the CLR-transformed data with the phyloseq (McMurdie and Holmes 2013) function distance. Principal Coordinates Analysis (PCoA) was performed using the phyloseq function ordinate, using Euclidean distances. Permutational multivariate analysis of variance (PERMANOVA) was performed on the Euclidean distances using the adonis2 function in the R package vegan v.2.6–2 (https://github.com/vegandevs/vegan), including the following as covariates: sequencing depth (total Braken abundance counts in a sample), length of individual, population or parasite (depending on the comparison), type (control or predator-stimulated) and extraction date. As there was still an effect of extraction batch on our community composition data (Supplementary Figure S3, Supplementary Table S3), we used limma’s function removeBatchEffect (Ritchie et al. 2015) to regress out the DNA extraction batch effect from the normalized data and repeated the ordination and PERMANOVA as described.

General linear models implemented through MaAsLin2 (Mallick et al. 2021) were used to identify genera with significantly different relative abundance (calculated via total sum scaling) between 1) non-parasitized Galtaból and Þristikla individuals and 2) non-parasitized and parasitized Galtaból individuals. In both models, type (control vs predator) was included as a fixed effect while extraction batch was included as a random effect. MaAsLin2 was run without additional normalization or transformation steps and otherwise default parameters. Genera with adjusted p-values (MaAsLin2 q-value) < 0.05 were classed as significantly differentially abundant.

To integrate the behavioral and microbiome datasets, multi-omic factor analysis (MOFA) was performed using the R package MOFA2 (Argelaguet et al., 2018). MOFA is similar to PCA, where matrices of different types of data generated from the same individuals are reduced to a small number of latent factors representing the key contributors of variation across the datasets (Argelaguet et al. 2018). In this study, MOFA was performed on the control individuals and predator-triggered individuals separately. For the behavioral dataset, all variables were included, both before and after the trigger. The behavioural variables were converted to approximate a normal distribution using the inverse normal transformation. For the microbiome dataset, the 100 most variable genera in each MOFA run (control and predator) were included, using CLR-transformed abundances after regressing out the extraction batch effect (as described for PCoA). Both MOFA models were trained with 7 factors, using Gaussian distributions for both datasets, scaling each dataset to have similar variances and otherwise using default values. After the MOFA models were trained, the function correlate_factors_with_covariates in the MOFA2 package was used to identify factors that were significantly correlated (alpha = 0.05) with metadata variables (fish population, parasite status, length of individual and sample extraction batch). Factors that explained >1% of the variation in both datasets were investigated further. The top 10 features (behavioural variables and microbial genera) contributing to the variation captured by each factor (5 with positive weights and 5 with negative weights) were extracted and classified.

3.1 Behavior

3.1.1 Effect of the presence of the robotic predator before simulated predator attack

We compared ecologically different fish populations in the control to characterize baseline behavioral differences among populations. Those that differed among populations were aspects of activity including total distance traveled and angular velocity, with spring-fed Galtaból fish displaying higher distance travelled and glacial Þristikla fish showing higher angular velocity (Table 1, Fig. 2).

Table 1

Results of linear models, using data from control trials and the acclimation period before predator attack, to test population (glacial Þristikla and spring-fed Galtaból fish) and type (Control and Treatment) and their interaction on swimming activity (Total Travelled Distance, Angular Velocity, and Mean Velocity) and the tank use (Center duration, Border duration, Distance to Predator, and Predator Duration) in stickleback fish.
Variable	Effect	Estimate	s.e.	df	_F	p-value
Total Traveled Distance	Intercept	27.181	1.628	n.a.	n.a.	n.a.
	Population	-2.667	2.333	1	4.5095	0.036837
	Type	-3.649	2.205	1	7.2248	0.008766
	Population:Type	-1.394	3.21	1	0.1885	0.665356
Mean Velocity	Intercept	0.64736	0.03962	n.a.	n.a.	n.a.
	Population	-0.03569	0.05676	1	2.7802	0.099393
	Type	-0.08393	0.05364	1	7.9774	0.005993
	Population:Type	-0.05551	0.0781	1	0.5051	0.479349
Angular Velocity	Intercept	2.248	0.04027	n.a.	n.a.	n.a.
	Population	0.11437	0.0577	1	17.8079	< 0.0001
	Type	0.07676	0.05453	1	9.7867	0.002462
	Population:Type	0.10009	0.07939	1	1.5896	0.211098
Center Duration	Intercept	106.12	14.83	n.a.	n.a.	n.a.
	Population	39.77	21.25	1	0.7495	0.38927
	Type	61.58	20.09	1	6.5468	0.01242
	Population:Type	-51.36	29.24	1	3.0848	0.0829
Border Duration	Intercept	193.88	14.96	n.a.	n.a.	n.a.
	Population	-39.76	21.43	1	0.8758	0.352195
	Type	-62.21	20.25	1	7.0216	0.009724
	Population:Type	49.2	29.49	1	2.7843	0.099151

We compared the behavior of control fish to that of predator-exposed fish during the acclimation period for two purposes. Firstly, we aimed to determine if there was any effect of the presence of the robotic predator on activity-related behaviors. We observed differences in activity, with all the variables being impacted by the presence of the robotic predator regardless of the populations (i.e., control versus predator-exposed (type) significant effect in Fig. 2 and Table 1). The total distance traveled and mean velocity were lower in the presence of the robotic predator, whereas the angular velocity was higher (Fig. 2, Table 1). Both total distance traveled and angular velocity significantly differed between populations during the acclimation period. Specifically, total distance traveled was significantly higher in Galtaból compared with Þristikla (spring-fed and glacial lake respectively; Fig. 2A, Table 1) whereas angular velocity was significantly lower in Galtaból (Fig. 2C).

Secondly, we assessed boldness by comparing tank use of the control fish to that of the predator-exposed fish. An increase in center duration was identified as bold behavior (Fig. 3A), while an increase in border duration was indicative of shy behavior (Fig. 3B). We observed an effect of the robotic predator regardless of the populations: center duration increased in the presence of the predator whereas the border duration decreased (Table 1, Fig. 3).

3.1.2 Effect of the simulated predatory attack

In predator-exposed fish, we compared behavior during the acclimation period to that after the simulated predatory attack to examine behavioral responses to predation and to investigate potential differences in responses between glacial and spring-fed populations. We found a significant interaction between population and period for angular velocity: angular velocity tended to increase after simulated attack in Galtaból (spring-fed fish) whereas it was the opposite in Þristikla (glacial fish, Table 2, Fig. 4C). Furthermore, the simulated predatory attack influenced both center and border duration (Table 2), leading to a decrease in time spent in the tank center and an increase in time spent at the tank border after the attack for both glacial and spring-fed populations (Fig. 5).

Table 2

Results of mixed-effects models (Fish ID as random effect) testing for the effects of robotic predator attack (period), population (glacial Þristikla and spring-fed Galtaból fish), effect of parasites and all interactions on swimming activity (Total Travelled Distance, Angular Velocity, and Mean Velocity) and the tank use (Center Duration, Border Duration, Distance to Predator, and Predator Duration) in stickleback fish.
Variable	Effect	Estimate	s.e.	df		p-value
Total Traveled Distance	Intercept	18.2051	1.9748	n.a.	n.a.	n.a.
	Population	0.1233	2.9387	1	0.821	0.365
	Period	5.4564	2.0611	1	2.329	0.127
	Parasite	4.7571	3.6567	1	2.0467	0.152
	Population:Period	-5.515	3.067	1	2.8394	0.093
	Population:Parasite	0.4693	5.3997	1	0.2853	0.593
	Period:Parasite	-5.2034	3.8164	1	1.4465	0.230
	Population:Period:Parasite	3.9818	5.6356	1	0.4992	0.480
	Intercept	0.44038	0.04813	n.a.	n.a.	n.a.
Mean Velocity	Population	0.01209	0.07162	1	0.6842	0.408
	Period	0.12044	0.04865	1	1.9465	0.163
	Parasite	0.12466	0.08912	1	1.8804	0.170
	Population:Period	-0.12387	0.07239	1	2.5661	0.109
	Population:Parasite	-0.02137	0.1316	1	0.0428	0.836
	Period:Parasite	-0.11569	0.09008	1	1.2648	0.262
	Population:Period:Parasite	0.08973	0.13302	1	0.4551	0.5
	Intercept	2.4651	0.042	n.a.	n.a.	n.a.
Angular Velocity	Population	0.08237	0.0625	1	11.2397	0.0008
	Period	-0.14152	0.04675	1	3.2083	0.0733
	Parasite	-0.08738	0.07777	1	2.7778	0.095
	Population:Period	0.17588	0.06957	1	5.5803	0.018
	Population:Parasite	-0.01778	0.11484	1	0.7374	0.390
	Period:Parasite	0.09143	0.08657	1	0.262	0.609
	Population:Period:Parasite	-0.12829	0.12784	1	1.007	0.316
	Intercept	80.14	16.18	n.a.	n.a.	n.a.
Center Duration	Population	-18.16	24.08	1	1.0091	0.315
	Period	72.19	17.26	1	39.7622	< 0.0001
	Parasite	84.07	29.96	1	9.6593	0.0018
	Population:Period	15.46	25.69	1	0.2373	0.626
	Population:Parasite	-14.34	44.24	1	0.3687	0.543
	Period:Parasite	-31.4	31.96	1	2.7629	0.096
	Population:Period:Parasite	-16.76	47.2	1	0.1261	0.722
	Intercept	195.509	17.383	n.a.	n.a.	n.a.
Border Duration	Population	30.693	25.867	1	1.2493	0.263
	Period	-48.446	16.435	1	28.1934	< 0.0001
	Parasite	-68.786	32.188	1	6.1594	0.013
	Population:Period	-30.252	24.456	1	1.2828	0.257
	Population:Parasite	7.766	47.531	1	0.2192	0.640
	Period:Parasite	15.995	30.431	1	1.4386	0.230
	Population:Period:Parasite	23.685	44.937	1	0.2778	0.598

3.1.3 Effect of parasites on anti-predator behavior

We did not observe any significant interaction between parasite, population, and period, nor any significant interaction between parasite and population for any behavior (Table 2). The parasite effect overall was significant for border and center duration only (Table 2). Center duration (proxy for boldness) significantly increased in each population when parasitized (Table 2, Fig. 5A). Border duration (proxy for shyness) showed the opposite pattern and significantly decreased in parasitized populations (Table 2, Fig. 5B).

3.2 Microbiome

All samples for microbiome analysis were taken after the predatory trigger (one per individual for control and predator-exposed fish). After sequencing, sample and taxa quality filtering and taxa contamination filtering (see methods/Supp), 1673 microbial taxa at the species-level were recovered from 62 stickleback gut samples, representing 652 genera. We observed some batch effects of extraction date on microbiome composition (Supplementary Figures S2–S3, Supplementary Table S3), thus we took extraction date into account in all analyses (see methods). Due to low sample numbers for the parasitized Þristikla group, these individuals were not included in the microbiome analysis. Instead, we focus on comparisons between i) non-parasitized Galtaból (n = 24) and Þristikla (n = 25) individuals, and ii) non-parasitized (n = 24) and parasitized (n = 13) Galtaból individuals.

The sticklebacks generally had quite diverse gut microbiomes at the genus level, with the most abundant genera including Streptomyces and Kitasatospora (Fig. 6). While there were no differences in alpha diversity between non-parasitized Galtaból and Þristikla individuals (Fig. 7A–B), the populations had significantly different gut microbiome compositions after adjusting for batch effects (Fig. 7E; Supplementary Table S3). Population explained 4.2% of the variation in microbiome composition (F = 1.93, p = 0.003). However, sequencing depth contributed to 5.0% of the variation in the data (F = 2.30, p = 0.001) and thus could be confounding the results. Pseudoalteromonas had significantly higher relative abundance in the non-parasitized, spring-fed Galtaból individuals compared to Þristikla individuals (adjusted p < 0.001, Supplementary Table S4); however, no other taxa were associated with differential abundance between populations.

In comparison to non-parasitized individuals in the spring-fed Galtaból population, several parasitized Galtaból individuals were dominated by one bacterial genus, which varied between Aeromonas, Deefgea and Labrenzia depending on the individual (Fig. 6). Consequently, parasitized individuals had significantly lower alpha diversity compared to non-parasitized individuals (Fig. 7C–D). Parasite status explained 7.2% of the variation in gut microbiome composition (F = 2.42, p = 0.001, Fig. 7F, Supplementary Table S3). Again, sequencing depth explained 6.4% of the variation (F = 2.16, p = 0.002). In the differential abundance tests, Kitasatospora and Shewanella had significantly higher relative abundance in non-parasitized Galtaból individuals compared to parasitized ones (adjusted p < 0.05, Supplementary Table S4). No other taxa were associated with differential abundance between parasite status.

3.3 Links between fish behaviors and microbiomes

The microbiome composition did not differ significantly between control and predator-triggered fish, based on the PCoA (Fig. 7E–F, Supplementary Table S3). No microbial genera were significantly differentially abundant between control vs predator-triggered fish in pairwise comparisons with a linear mixed model.

As an alternative approach to explore associations between the microbiome and behavioral datasets, we used MOFA, an unsupervised, multivariate approach. Since the predator-exposed fish included behavioral variables not relevant for the control fish (distance to and duration around the predator stimulus), control fish and predator-exposed fish were analysed separately.

The MOFAs revealed some evidence for associations between behavioral traits and microbe abundances. Generally each factor in the MOFAs explained substantial variation in only one dataset (Fig. 8). However, the MOFA on control fish identified one factor explaining > 1% of variation in both datasets (Factor 2 in Fig. 8A) and the MOFA on predator-triggered fish identified two such factors (Factor 1 and 2 in Fig. 8C). We therefore focus on these three factors (see Supplementary Figures S4–14 for summary figures from the other factors).

In control fish, the behavioral traits contributing the most to Factor 2 included traits related to boldness and shyness (time spent in the center and borders of the enclosure, respectively) (Fig. 9A), while the most contributing microbial genera included those likely sourced from the environment (water, soil, etc) (Fig. 9B). This factor was correlated with fish population and parasite status, although the correlations were not significant (p > 0.05; Fig. 8B). As an example, center duration and the environmental microbe Massilia were particularly associated with this factor, with both tending to have higher values in glacial Þristikla, compared to spring-fed Galtaból individuals (Fig. 9C–D). The animal-associated microbe Propionibacterium showed a similar pattern (Fig. 9A–B).

In predator-exposed fish, Factor 2 was correlated with parasite status and population (Fig. 8D). Again, behavioral traits relating to boldness and shyness contributed most strongly to this factor (Fig. 10A), including center duration (Fig. 10C). In the microbiome dataset, putative animal- and fish-associated microbial genera were among the strongest contributors, including the documented fish pathogen Aeromonas and the fish microbe Deefgea (Fig. 10B). For example, center duration and Deefgea abundance was higher in parasitized individuals compared to non-parasitized individuals (Fig. 10C–D).

Factor 1 in predator-exposed fish was also of interest, although it was not significantly correlated with any fish phenotype (Fig. 8D). Traits relating to swimming activity and response to the predator stimulus contributed the most to this factor in the behavioral dataset (Fig. 11A), while environmental genera like Mesorhizobium and animal-associated bacteria like Clostridioides contributed the most in the microbial dataset (Fig. 11B). In particular, mean velocity and the genus Clostridioides were associated with this factor (Fig. 11C–D).

Our main objective in analyzing behavior and responses to simulated predation, along with exploring how parasitism may affect behavior, is to uncover possible links between antipredator behavior and the microbiome. We aim to understand whether there are associations between predation, parasitism, and the microbiome. We begin by discussing behavioral findings, followed by comparing the microbiome between populations and parasite status, and finish by discussing potential links between behavior and the microbiome suggested by our findings.

4.1. Differences between populations and effects of parasitism in behavior

The glacial and spring-fed populations differed in baseline behavior and some responses to predation. Baseline behavior revealed that fish from the two populations were active in different ways: glacial Þristikla fish showed higher angular velocity while spring-fed Galtaból fish traveled further. Predator responses were strong for both populations as well. The mere presence of the robotic predator influenced behavior and did so differently for the two populations, increasing time spent in the tank center (and therefore boldness) for spring-fed fish from Galtaból and increasing activity measures for glacial fish from Þristikla. Differences in activity and boldness may relate to the ability to detect predators in the distinct sensory environments of each lake. The clear waters in spring-fed Galtaból allow sticklebacks to detect predators from a distance, and unobstructed vision may foster high exploration and boldness in those open waters (Cerri 1983). In contrast, the turbid waters in glacial Þristikla make visual detection of predators difficult except at very close distances (Ålund et al. 2022), perhaps fostering more erratic movements and shyness, both of which may help fish escape capture by predators (Leahy et al. 2011; Ajemian, Sohel, and Mattila 2015).

In spite of these population differences in both baseline behavior and in the mere presence of the robotic predator, both populations responded strongly and similarly to simulated predatory attack. After the attack, both increased activity and moved towards the tank border(Sih, Bell, and Johnson 2004; Ward et al. 2004; A. M. Bell and Sih 2007; Laskowski et al. 2024). This similarity is perhaps surprising given that the lakes differ substantially in ecology and visibility, and that a prior study showed different escape behaviors for glacial and spring-fed fish (Ålund et al. 2022). That study analyzed a larger set of behaviors, including first response to predator attack, and some of the differences seen were in those first responses, which we did not measure. Our findings suggest that a robust and cautious response to predator attacks is favored in both types of populations.

Both populations showed similar responses when parasitized, although there was higher overall parasitism in spring-fed Galtaból fish. Parasitized fish appeared to be bolder compared with their unparasitized counterparts in response to the robotic predator attack. This finding is in accordance with the results of previous studies on the same species, showing that parasitized sticklebacks returned more quickly to food patches following disturbance by a model predator (N Giles 1983), or spent more time in the open water and reduced their flight responses to a predator attack from above (Milinski and Heller 1978; N Giles 1983; Huntingford and Giles 1987; Barber, Walker, and Svensson 2004). Both of these responses would be likely to enhance transmission of the parasite by increasing the likelihood of predation, as parasites act as strong selective agents on their host by altering behavior and habitat use, ultimately decreasing the hosts’ fitness (Blake, Kwok, and Chan 2006; Thorburn et al. 2022). However, we cannot exclude undetectable effects due to the low sample size of tested parasitized fish.

4.2 Differences between populations and effects of parasitism in the microbiome

Consistent with previous studies of wild-caught sticklebacks, a small but significant proportion of variation in the gut microbiome was explained by population, likely reflecting differences in lake environment and genetic background between the two populations (Bolnick et al. 2014; C. C. R. Smith et al. 2015). However, only one microbial taxon, the typically marine genus Pseudoalteromonas, was significantly differentially abundant between populations. Thus, the differences in gut microbiome in our populations appear to be driven by small changes in the abundance of many taxa, rather than large differences in a few taxa. To minimize the confounding effects of transient microbes derived from food or water sources (Bolnick et al. 2014), fish were co-housed by population and fed the same diet for several weeks before the experiments were performed. We therefore cannot rule out that this set-up had a homogenizing effect on the microbiome of both populations. However, this set-up did not mask the large microbiome variation observed in parasitized fish, which were co-housed with non-parasitized fish. Previous research has shown that transitioning fish from a wild to a laboratory setting, including a laboratory diet, does affect gut microbiome composition within 14 days (Restivo et al. 2021). However, a study in wild sticklebacks found that host genotype explains more variation in the gut microbiome than environmental effects like diet and habitat, suggesting that microbiome composition may depend more on host selection of microbes, rather than transient colonisation from environmental sources (C. C. R. Smith et al. 2015).

Our findings align with recent work by Viver et al., 2023, who showed that the transient microbiome in fish is highly dependent on the supplied food, and that fasting the fish for 80 hours effectively eliminated most transient microbes, leaving the resident microbiome. In our study, despite many days of a standardized diet, we observed significant differences between ecotypes, which may reflect the persistence of the resident-core microbiome. These findings suggest that both ecotype and environment shape the microbiome as the host adapts to its surroundings. Furthermore, Viver et al., 2023 found that the bacteria present in the supplied diet were unable to successfully colonize the gut and replace the resident bacteria naturally found in the wild. These findings indicate that signals from the resident microbiome, possibly influenced by feeding habits in the wild, may still persist even after an extended period on a laboratory diet.

The effect of parasitism on the microbiome was only investigated in the spring-fed Galtaból population, where it was associated with decreased alpha diversity and an increase in abundance of specific microbial taxa, such as the potential pathogen Aeromonas (Beaz-Hidalgo and Figueras 2013), the marine genus Labrenzia (Biebl et al. 2007) and the fish gut microbe Deefgea (Gim et al. 2022), and the decrease of putatively commensal microbes such as Shewanella and Kitasatospora (Gaulke et al. 2019). Some Kitasatospora species even produce secondary metabolites with anthelmintic properties (Ōmura et al. 1981; Takahashi 2017). Previous studies of fish microbiomes during parasite exposure or infection suggest that parasite infection promotes the growth of taxa that are likely not beneficial to the host, while inhibiting the growth of commensal microbes (Brealey et al. 2022; Gaulke et al. 2019; Hahn et al. 2022; Ling et al. 2020). For example, in zebrafish, Shewanella was negatively associated with nematode parasite load (Gaulke et al. 2019), while in sticklebacks, bacteria related to Labrenzia (of the Rhodobacteraceae family or Rhodobacterales order) were associated with exposure or infection with the cestode Schistocephalus solidus (Hahn et al. 2022; Ling et al. 2020).

4.3 Links between behavior and microbiome

To our knowledge, this is the first report on the relationship between the microbiome and behavior in these wild populations. We observed some correlations between behavioral traits and the microbiome, particularly related to variations among fish populations and parasite status. However, these interactions were generally weak, perhaps due to the inherent complexity of the gut-brain axis and its causal relationships, which have yet to be fully understood (K. R. Foster et al. 2017). However, we acknowledge that limitations in our study, including the sample size, reduced our ability to detect small effects. Nonetheless, our results suggest some intriguing links between the microbiome and behavior in wild populations, which we discuss in more detail below.

Since most of the links we observed between behavioral traits and the microbiome can be explained by fish population or parasite status, we cannot be certain whether these are true interactions along the gut-brain axis influencing behavior, or instead that they are indirect correlations driven primarily by these factors acting in a non-behavioral context. For example, in control fish, we observed correlations between population, bold/shy traits and environmental microbes that likely reflect baseline differences in fish adaptation to glacial or clear-water lake environments. The most correlated microbes particularly reflected these environmental differences, with the bacteria Massilia, which has been previously isolated from glacier permafrost (Wang et al. 2018), at higher abundance in the glacial Þristikla population, and Gemmatimonas, which has been previously found in high arctic streams (Zeng et al. 2021), at higher abundance in the spring-fed Galtaból population. Similarly, in predator-exposed fish, we observed correlations between parasite status, bold/shy traits and potential pathogens, reflecting the results identified in the separate analyses discussed above. However, some of our results suggested a more direct link between microbes and behavior. For example, in control fish, Propionibacterium was correlated with boldness (center duration, Fig. 9B). Related Propionibacteriaceae bacteria have been associated with boldness in zebrafish (Ayayee and Wong 2024).

We also observed correlations between swimming activity and putative environmental microbes in predator-exposed fish, which could not be associated with either population or parasite status. While most of the associated microbes are not well-documented, abundance of the bacterium Clostridioides was correlated with increased mean velocity, which can be an indicator of bolder and more active fish. In a selection experiment of red junglefowl (the ancestors of domestic chickens), the abundances of related Clostridiales bacteria were enriched in birds selected for a low-fear of humans, compared to high-fear animals (Puetz et al. 2021). Other studies have also found links between specific microbes and their products and how these affect metabolism and locomotion of the host (Schretter et al. 2018; Schretter 2020). While tentative, our results are an intriguing suggestion that there could be some innate link between swimming activity, stress levels, bold/shy traits, and the microbiome.

A number of other studies have found links between the gut microbiota and host behaviors, including stress-related behavior, social behavior, locomotion, and personality, and also point out that microbes potentially act as neuromodulators or affect neural development, all of which are relevant to our study (Archie and Tung 2015; Schretter 2020). The reciprocal relationships we found between the microbiome and anti-predator behavior are relatively weak as evidenced by the factor correlations and factor weightings and are seen for only a subset of microbes and behaviors. However, this is not surprising, considering that any mechanistic connections are likely to be indirect and mediated through host physiology and neural systems (Archie and Tung 2015; Ayayee and Wong 2024), host-microbiota crosstalk (Lyte et al. 2022), as well as coevolution between microbiota and the host (Archie & Tung CurrOpinionBehavSci 2015). These are complex systems, and thus, the connections are also likely to be multifaceted, given the many potential mechanisms influencing the microbiota-gut-brain axis (Awe et al. 2024). This complexity means that each connection might be of small effect, yet perhaps combine for a larger effect on the host and their behavior, and the microbial community. Large effects like those seen in lab-based manipulations on germ-free animals would be exciting but seem unlikely in our wild fish (Ayayee and Wong 2024). We pick up significant, albeit weak relationships, which we argue seems to be the most likely outcome if it is the case that host anti-predator behavior and population and parasite status, and the microbiome have reciprocal effects on each other. We would expect only a subset of microbes and behaviors to respond to each other, given that the microbiome plays multiple roles within hosts, and only some of these are likely to be functionally connected to host anti-predator behavior/responses to stress (the variables we measured), and these connections are probably mediated through various physiological mechanisms (Archie and Tung 2015).

Our study is one of the first to investigate microbiome–behavioural associations in a wild population. Most previous studies have relied on laboratory animals, where germ-free or reduced microbiome animals and/or probiotic supplements have been used to manipulate the microbiome, resulting in large effects on behaviour (reviewed in (Vuong et al. 2017)). For example, in a study on zebrafish, a probiotic supplement was found to protect conventionally-raised laboratory animals against stress-induced alteration of the gut microbiome (D. J. Davis et al. 2016). Similar results have been observed in mice and rats (Liang et al. 2015; C. J. Smith et al. 2014). However, laboratory-reared animals often have different, potentially maladapted microbiomes compared to wild populations, possibly contributing to the large effects seen (Restivo et al. 2021; Rosshart et al. 2017; Ayayee and Wong 2024). We hypothesize that our wild populations, with their “natural” microbiomes, may have adapted to the stresses of a natural environment, like predator attacks. Thus, their microbiomes might already be optimized to protect the fish against stress-related microbiome changes, perhaps helping to explain why we see few associations between behavioral traits and microbiome composition.

In the future, performing stress experiments in wild populations while manipulating the microbiome, e.g. through probiotic, antibiotic or phage therapy, might reveal stronger evidence for behavior–microbe interactions. Teasing apart causality will also require experimental manipulations. As a first step, a holo-omics approach, where multi-omics techniques like transcriptomics and metabolomics are used to simultaneously investigate both the host and the microbiome (Nyholm et al. 2020) could identify gene expression networks and key metabolites involved in modulating the coordinated microbial–host response to stress. Selection experiments could compare microbiome changes over generations selected for stress-resilience or stress-susceptibility (Puetz et al. 2021; Ayayee and Wong 2024). Potential microbial and/or host effects could then be followed up in validation experiments, to confirm causality and work out the mechanisms. For example, identifying the Clostridioides strain associated with bolder fish in our study and administering it as a “probiotic” before exposure to predator stress in a follow-up experiment could be used to confirm or reject its association with anti-predator response behaviour. To determine causal effects of a microbial community, fecal microbiome transplants could be performed by switching microbiomes between bold and shy animals, to see if the behavioral traits also switch based on microbiome composition, similarly as the experiment carried out by Collins et al., 2013on mice. Such microbiome transplant methods are well-established in zebrafish models and could be adapted to sticklebacks (Cornuault et al. 2022). Such future work, building on studies like ours, will improve our understanding of the complexity of the gut-brain axis across the animal clade.

Ethics approval and consent to participate

The animal treatments were conducted in accordance with the Icelandic Food and Veterinary Authority (Matvaelastofnum, MAST), and approved by Michigan State University Institutional Animal Care (IACUC, protocol numbers 05/18-077-99, 05/16-064-00 and 201900128). Permits to collect stickleback fish were granted by Fjállabak Nature Reserve and the Vantajökulsþjóðgarður National Park.

Competing interests

The authors declare no competing interests.

Funding

Royal Swedish Academy of Sciences, BS2018-0075, Helge Ax. SON johnsons stiftelse, F21-0293, US NSF-DEB Dimensions of Biodiversity, 1638778, Fulbright Fellowship in Arctic Research, and Liljewalch travel scholarships.

Author Contribution

JEV and JCB Conceived the project and designed the experiment with input from JWB and DB. JWB, RL, and JEV contribute to the fieldwork. RL and JEV Performed the behavioral experiment and dissection of the fish. JCB and JEV performed the DNA extraction. DB performed the fish-tracking behavior with input from JEV. JEV Performed the Behavioral statistical analysis with key help from DB and input from JCB and JWB. JCB performed the microbiome analyses and the figures with comments and input from JEV, DB, and JWB. JCB, JEV, DB, and JWB wrote equally the paper. All the authors read and approved the final version of the manuscript.

Acknowledgement

We thank Jason Keagy and Murielle Ålund for help collecting fish and fieldwork, Rakel Þorbjörnsdóttir for fish care, and Kári Heiðar Árnason and Bjarni Kristófer Kristjánsson for laboratory and field support. We acknowledge that the antipredator experiment was initially designed by Jason Keagy, with key input from Murielle Ålund, and modified by us to test our hypothesis. We thank Niki Chondrelli for assistance in the lab. We thank Richard Svanbäck for the comments and conversations about the data analysis. We also acknowledge Michael Martin from the Norwegian University of Science and Technology for providing additional computational resources and data storage for the later stages of this project. This work was supported by the following: Royal Swedish Academy of Sciences grant BS2018-0075 (JCB, JEV), US NSF-DEB Dimensions of Biodiversity #1638778 (JWB), Fulbright Fellowship in Arctic Research (JWB) and Helge Ax. SON johnsons stiftelse F21-0293 (JEV). Travel expenses were paid by Liljewalch travel scholarships (JEV). Sequencing was performed by the SNP&SEQ Technology Platform in Uppsala. The facility is part of the National Genomics Infrastructure (NGI) Sweden and Science for Life Laboratory. The SNP&SEQ Platform is also supported by the Swedish Research Council and the Knut and Alice Wallenberg Foundation. We acknowledge the Swedish National Bioinformatics Infrastructure for providing computational resources to this project.

Data Availability

Upon manuscript acceptance, raw metagenomic sequencing data will be available at the European Nucleotide archive (ENA) under project PRJEB52754. All sample ENA accessions are provided in Supplementary File S1. Sample metadata, including behavioral traits extracted from the videos, are also provided in Supplementary File S1. All R code for figures, microbiome and microbiome–behavior analysis is provided in Supplementary File S2. All R code for behavior statistical analysis is provided in Supplementary File S3.

Ahlmann-Eltze, Constantin, and Indrajeet Patil. 2021. ‘Ggsignif: R Package for Displaying Significance Brackets for “Ggplot2”’. PsyArxiv, February, doi:10.31234/osf.io/7awm6. https://doi.org/10.31234/OSF.IO/7AWM6.
Ajemian, M. J., S. Sohel, and J. Mattila. 2015. ‘Effects of Turbidity and Habitat Complexity on Antipredator Behavior of Three-Spined Sticklebacks (Gasterosteus Aculeatus): Antipredator Behavior in Sticklebacks’. Environmental Biology of Fishes 98 (1): 45–55. https://doi.org/10.1007/s10641-014-0235-x.
Alberdi, Antton, and M. Thomas P. Gilbert. 2019. ‘A Guide to the Application of Hill Numbers to DNA-Based Diversity Analyses’. Molecular Ecology Resources 19 (4): 804–17. https://doi.org/10.1111/1755-0998.13014.
Ålund, Murielle, Brooke Harper, Sigurlaug Kjærnested, Julian E. Ohl, John G. Phillips, Jessica Sattler, Jared Thompson, et al. 2022. ‘Sensory Environment Affects Icelandic Threespine Stickleback’s Anti-Predator Escape Behaviour’. Proceedings of the Royal Society B: Biological Sciences 289 (1972). https://doi.org/10.1098/rspb.2022.0044.
Archie, Elizabeth A., and Jenny Tung. 2015. ‘Social Behavior and the Microbiome’. Current Opinion in Behavioral Sciences. Elsevier Ltd. https://doi.org/10.1016/j.cobeha.2015.07.008.
Argelaguet, Ricard, Britta Velten, Damien Arnol, Sascha Dietrich, Thorsten Zenz, John C Marioni, Florian Buettner, Wolfgang Huber, and Oliver Stegle. 2018. ‘Multi‐Omics Factor Analysis—a Framework for Unsupervised Integration of Multi‐omics Data Sets’. Molecular Systems Biology 14 (6): 8124. https://doi.org/10.15252/MSB.20178124/SUPPL_FILE/MSB178124-SUP-0002-EVFIGS.PDF.
Awe, Temitope, Ayoola Fasawe, Caleb Sawe, Adedayo Ogunware, Abdullahi Temitope Jamiu, and Michael Allen. 2024. ‘The Modulatory Role of Gut Microbiota on Host Behavior: Exploring the Interaction between the Brain-Gut Axis and the Neuroendocrine System’. AIMS Neuroscience. AIMS Press. https://doi.org/10.3934/NEUROSCIENCE.2024004.
Ayayee, Paul A, and Ryan Y Wong. 2024. ‘Title: Zebrafish (Danio Rerio) Behavioral Phenotypes Not Underscored by Different Gut Microbiota’. BioRxiv. https://doi.org/10.1101/2024.05.29.596447.
Barber, Iain. 2007. ‘Parasites, Behaviour and Welfare in Fish’. Applied Animal Behaviour Science 104 (3–4): 251–64. https://doi.org/10.1016/j.applanim.2006.09.005.
Barber, Iain, Daniel Hoare, and Jens Krause. 2000. ‘Effects of Parasites on Fish Behaviour: A Review and Evolutionary Perspective’. Reviews in Fish Biology and Fisheries 10 (2): 131–65. https://doi.org/10.1023/A:1016658224470.
Barber, Iain, Peter Walker, and P Andreas Svensson. 2004. ‘BEHAVIOURAL RESPONSES TO SIMULATED AVIAN PREDATION IN FEMALE THREE SPINED STICKLEBACKS: THE EFFECT OF EXPERIMENTAL SCHISTOCEPHALUS SOLIDUS INFECTIONS’.
Beaz-Hidalgo, R., and M. J. Figueras. 2013. ‘Aeromonas Spp. Whole Genomes and Virulence Factors Implicated in Fish Disease’. Journal of Fish Diseases 36 (4): 371–88. https://doi.org/10.1111/JFD.12025.
Bell, Alison M., and Andrew Sih. 2007. ‘Exposure to Predation Generates Personality in Threespined Sticklebacks (Gasterosteus Aculeatus)’. Ecology Letters 10 (9): 828–34. https://doi.org/10.1111/j.1461-0248.2007.01081.x.
Bell, M.A., and S.A. Foster. 1994. ‘The Evolutionary Biology of the Threespine Stickleback’. The Journal of Animal Ecology 64 (3): 418. https://doi.org/10.2307/5902.
Bergstrom, C. A. 2002. ‘Fast-Start Swimming Performance and Reduction in Lateral Plate Number in Threespine Stickleback’. Canadian Journal of Zoology 80 (2): 207–13. https://doi.org/10.1139/z01-226.
Biebl, Hanno, Rüdiger Pukall, Heinrich Lünsdorf, Stefan Schulz, Martin Allgaier, Brian J. Tindall, and Irene Wagner-Döbler. 2007. ‘Description of Labrenzia Alexandrii Gen. Nov., Sp. Nov., a Novel Alphaproteobacterium Containing Bacteriochlorophyll a, and a Proposal Reclassification of Stappia Aggregata as Labrenzia Aggregata Comb. Nov., and of Stappia Alba as Labrenzia Alba Comb. Nov., and Emended Descriptions of the Genera Pannonibacter, Stappia and Roseibium, and of the Species Roseibium Denhamense and Roseibium Hamelinense’. International Journal of Systematic and Evolutionary Microbiology 57 (5): 1095–1107. https://doi.org/10.1099/IJS.0.64821-0/CITE/REFWORKS.
Blake, R W, P Y L Kwok, and K H S Chan. 2006. ‘Effects of Two Parasites, Schistocephalus Solidus (Cestoda) and Bunodera Spp. (Trematoda), on the Escape Fast-Start Performance of Three-Spined Sticklebacks’. Journal of Fish Biology 69 (5): 1345–55. https://doi.org/10.1111/j.1095-8649.2006.01193.x.
Bolnick, Daniel I., Lisa K. Snowberg, Philipp E. Hirsch, Christian L. Lauber, Rob Knight, J. Gregory Caporaso, and Richard Svanbäck. 2014. ‘Individuals’ Diet Diversity Influences Gut Microbial Diversity in Two Freshwater Fish (Threespine Stickleback and Eurasian Perch)’. Ecology Letters 17 (8): 979. https://doi.org/10.1111/ELE.12301.
Brealey, Jaelle C., Laurène A. Lecaudey, Miyako Kodama, Jacob A. Rasmussen, Harald Sveier, Nolwenn M. Dheilly, Michael D. Martin, and Morten T. Limborg. 2022. ‘Microbiome “Inception”: An Intestinal Cestode Shapes a Hierarchy of Microbial Communities Nested within the Host’. MBio 13 (3). https://doi.org/10.1128/mbio.00679-22.
Brown, Grant E, and R Jan F Smith. 1998. ‘Acquired Predator Recognition in Juvenile Rainbow Trout (Oncorhynchus Mykiss): Conditioning Hatchery-Reared Fish to Recognize Chemical Cues of a Predator’. Can J Fish Aquat Sci 55:611–17.
Bush, Albert O, Jacqueline C Fernández, Gerald W Esch, and J. Richard Seed. 2001. Parasitism : The Diversity and Ecology of Animal Parasites. Cambridge University Press.
Cerri, Richard D. 1983. ‘THE EFFECT OF LIGHT INTENSITY ON PREDATOR AND PREY BEHAVIOUR IN CYPRINID FISH: FACTORS THAT INFLUENCE PREY RISK’. Anlm. Behav 31:736–42.
Collins, Stephen M., and Premsyl Bercik. 2009. ‘The Relationship Between Intestinal Microbiota and the Central Nervous System in Normal Gastrointestinal Function and Disease’. Gastroenterology 136 (6): 2003–14. https://doi.org/10.1053/j.gastro.2009.01.075.
Collins, Stephen M., Zain Kassam, and Premysl Bercik. 2013. ‘The Adoptive Transfer of Behavioral Phenotype via the Intestinal Microbiota: Experimental Evidence and Clinical Implications’. Current Opinion in Microbiology. https://doi.org/10.1016/j.mib.2013.06.004.
Cornuault, Jeffrey K., Gabriel Byatt, Marie Eve Paquet, Paul De Koninck, and Sylvain Moineau. 2022. ‘Zebrafish: A Big Fish in the Study of the Gut Microbiota’. Current Opinion in Biotechnology 73 (February):308–13. https://doi.org/10.1016/J.COPBIO.2021.09.007.
Dahlbom, S. Josefin, David Lagman, Katrin Lundstedt-Enkel, L. Fredrik Sundström, and Svante Winberg. 2011. ‘Boldness Predicts Social Status in Zebrafish (Danio Rerio)’. PLoS ONE 6 (8). https://doi.org/10.1371/journal.pone.0023565.
Davis, Daniel J., Holly M. Doerr, Agata K. Grzelak, Susheel B. Busi, Eldin Jasarevic, Aaron C. Ericsson, and Elizabeth C. Bryda. 2016. ‘Lactobacillus Plantarum Attenuates Anxiety-Related Behavior and Protects against Stress-Induced Dysbiosis in Adult Zebrafish’. Scientific Reports 2016 6:1 6 (1): 1–11. https://doi.org/10.1038/srep33726.
Davis, Nicole M., DIana M. Proctor, Susan P. Holmes, David A. Relman, and Benjamin J. Callahan. 2018. ‘Simple Statistical Identification and Removal of Contaminant Sequences in Marker-Gene and Metagenomics Data’. Microbiome 6:226. https://doi.org/10.1186/s40168-018-0605-2.
Deenonpoe, Raksawan, Chariya Chomvarin, Chawalit Pairojkul, Yaowalux Chamgramol, Alex Loukas, Paul J. Brindley, and Banchob Sripa. 2015. ‘The Carcinogenic Liver Fluke Opisthorchis Viverrini Is a Reservoir for Species of Helicobacter’. Asian Pacific Journal of Cancer Prevention 16 (5): 1751–58. https://doi.org/10.7314/APJCP.2015.16.5.1751.
Demandt, Nicolle, Marit Praetz, Ralf H J M Kurvers, Jens Krause, Joachim Kurtz, and Jörn P Scharsack. 2020. ‘Parasite Infection Disrupts Escape Behaviours in Fish Shoals: Parasite Infection Disrupts Escape Waves’. Proceedings of the Royal Society B: Biological Sciences 287 (1938). https://doi.org/10.1098/rspb.2020.1158.
Dessì, Daniele, Paola Rappelli, Nicia Diaz, Piero Cappuccinelli, and Pier Luigi Fiori. 2006. ‘Mycoplasma Hominis and Trichomonas Vaginalis: A Unique Case of Symbiotic Relationship between Two Obligate Human Parasites’. Frontiers in Bioscience. Vol. 11.
Diehl, Sebastian, Peter Eklöv, Sebastian Diehl2, and Peter Eklov. 1995. ‘Effects of Piscivore-Mediated Habitat Use on Resources, Diet, and Growth of Perch’. Source: Ecology 76 (6): 1712–26.
Eklöv, Peter, and Richard Svanbäck. 2006. ‘Predation Risk Influences Adaptive Morphological Variation in Fish Populations’. American Naturalist 167:440–52.
Endler, John A. 1983. ‘Multiple-Trait Coevolution and Environmental Gradients in Guppies’. Trends Ecology & Evolution 10:22–29.
Fakan, Eric P., Bridie J.M. Allan, Björn Illing, Andrew S. Hoey, and Mark I. McCormick. 2023. ‘Habitat Complexity and Predator Odours Impact on the Stress Response and Antipredation Behaviour in Coral Reef Fish’. PLoS ONE 18 (6 June). https://doi.org/10.1371/journal.pone.0286570.
Foster, Jane A., Linda Rinaman, and John F. Cryan. 2017. ‘Stress & the Gut-Brain Axis: Regulation by the Microbiome’. Neurobiology of Stress. Elsevier Inc. https://doi.org/10.1016/j.ynstr.2017.03.001.
Foster, Kevin R., Jonas Schluter, Katharine Z. Coyte, and Seth Rakoff-Nahoum. 2017. ‘The Evolution of the Host Microbiome as an Ecosystem on a Leash’. Nature 548 (7665): 43–51. https://doi.org/10.1038/nature23292.
Fox, J, and S Weisberg. 2011. ‘An R Companion to Applied Regression.’ Sage Publishing, Thousand Oaks.
Fraser, Douglas F, and James F Gilliam. 1992. ‘This Content Downloaded from 74.221.174.247 on Wed’. Source: Ecology. Vol. 73.
Gaulke, Christopher A., Mauricio L. Martins, Virginia G. Watral, Ian R. Humphreys, Sean T. Spagnoli, Michael L. Kent, and Thomas J. Sharpton. 2019. ‘A Longitudinal Assessment of Host-Microbe-Parasite Interactions Resolves the Zebrafish Gut Microbiome’s Link to Pseudocapillaria Tomentosa Infection and Pathology’. Microbiome 7 (1): 10. https://doi.org/10.1186/s40168-019-0622-9.
Giles, N. 1983. ‘Behavioural Effects of the Parasite Schistocephalus Solidus (Cestoda) on an Intermediate Host, the Three-Spined Stickleback, Gasterosteus Aculeatus L.’ Animal Behaviour 31 (4): 1192–94. https://doi.org/10.1016/S0003-3472(83)80025-6.
Giles, Nick, and Felicity A Huntingford. 1984. ‘PREDATION RISK AND INTER-POPULATION VARIATION IN ANTI-PREDATOR BEHAVIOUR IN THE THREE-SPINED STICKLEBACK, GASTEROSTEUS ACULEATUS L’. Anim. Behav. Vol. 32.
Gim, Do Hun, So Yeon Lee, Jeong Eun Han, Jae Yun Lee, Seo Min Kang, and Jin Woo Bae. 2022. ‘Description of Deefgea Piscis Sp. Nov., and Deefgea Tanakiae Sp. Nov., Isolated from the Gut of Korean Indigenous Fish’. Journal of Microbiology (Seoul, Korea) 60 (11): 1061. https://doi.org/10.1007/S12275-022-2250-5.
Godin, Jean Guy J., and Kyla A. Valdron Clark. 1997. ‘Risk-Taking in Stickleback Fishes Faced with Different Predatory Threats 1’. Ecoscience 4 (2): 246–51. https://doi.org/10.1080/11956860.1997.11682403.
Hahn, Megan A., Agnes Piecyk, Fátima Jorge, Robert Cerrato, Martin Kalbe, and Nolwenn M. Dheilly. 2022. ‘Host Phenotype and Microbiome Vary with Infection Status, Parasite Genotype, and Parasite Microbiome Composition’. Molecular Ecology 31 (5): 1577–94. https://doi.org/10.1111/mec.16344.
Hall, April E., and Timothy D. Clark. 2016. ‘Seeing Is Believing: Metabolism Provides Insight into Threat Perception for a Prey Species of Coral Reef Fish’. Animal Behaviour 115 (May):117–26. https://doi.org/10.1016/j.anbehav.2016.03.008.
Härer, Andreas, Alexander A. Mauro, Telma G. Laurentino, Erica B. Rosenblum, and Diana J. Rennison. 2023. ‘Gut Microbiota Parallelism and Divergence Associated with Colonisation of Novel Habitats’. Molecular Ecology 32 (20): 5661–72. https://doi.org/10.1111/mec.17135.
Hothorn, Torsten, Kurt Hornik, Mark A. Van De Wiel, and Achim Zeileis. 2006. ‘A Lego System for Conditional Inference’. American Statistician 60 (3): 257–63. https://doi.org/10.1198/000313006X118430.
Huntingford, Felicity, and Nick Giles. 1987. ‘Individual Variation in Anti‐predator Responses in the Three‐spined Stickleback (Gasterosteus Aculeatus L.)’. Ethology 74 (3): 205–10. https://doi.org/10.1111/j.1439-0310.1987.tb00933.x.
Jaenike, John, Robert Unckless, Sarah N. Cockburn, Lisa M. Boelio, and Steve J. Perlman. 2010. ‘Adaptation via Symbiosis: Recent Spread of a Drosophila Defensive Symbiont’. Science 329 (5988): 212–15. https://doi.org/10.1126/science.1188235.
Jenkins, Matthew R., John M. Cummings, Alex R. Cabe, Kaj Hulthén, M. Nils Peterson, and R. Brian Langerhans. 2021. ‘Natural and Anthropogenic Sources of Habitat Variation Influence Exploration Behaviour, Stress Response, and Brain Morphology in a Coastal Fish’. Journal of Animal Ecology 90 (10): 2446–61. https://doi.org/10.1111/1365-2656.13557.
Johnson, Katerina V.A., and Kevin R. Foster. 2018. ‘Why Does the Microbiome Affect Behaviour?’ Nature Reviews Microbiology 16 (10): 647–55. https://doi.org/10.1038/s41579-018-0014-3.
Jorge, Fátima, Nolwenn M. Dheilly, and Robert Poulin. 2020. ‘Persistence of a Core Microbiome Through the Ontogeny of a Multi-Host Parasite’. Frontiers in Microbiology 11:954. https://doi.org/10.3389/fmicb.2020.00954.
Kashinskaya, Elena N., Evgeniy P. Simonov, Larisa G. Poddubnaya, Pavel G. Vlasenko, Anastasiya V. Shokurova, Aleksey N. Parshukov, Karl B. Andree, and Mikhail M. Solovyev. 2023. ‘Trophic Diversification and Parasitic Invasion as Ecological Niche Modulators for Gut Microbiota of Whitefish’. Frontiers in Microbiology 14 (March):1090899. https://doi.org/10.3389/FMICB.2023.1090899/BIBTEX.
Kopack, Christopher J., E. Dale Broder, Jesse M. Lepak, Eric R. Fetherman, and Lisa M. Angeloni. 2015. ‘Behavioral Responses of a Highly Domesticated, Predator Naïve Rainbow Trout to Chemical Cues of Predation’. Fisheries Research 169 (September):1–7. https://doi.org/10.1016/j.fishres.2015.04.005.
Landmann, Frederic, Denis Voronin, William Sullivan, and Mark J. Taylor. 2011. ‘Anti-Filarial Activity of Antibiotic Therapy Is Due to Extensive Apoptosis after Wolbachia Depletion from Filarial Nematodes’. PLoS Pathogens 7 (11): e1002351. https://doi.org/10.1371/journal.ppat.1002351.
Lapiedra, Oriol, Thomas W Schoener, Manuel Leal, Jonathan B Losos, and Jason J Kolbe. 2018. ‘Predator-Driven Natural Selection on Risk-Taking Behavior in Anole Lizards’. Science 360:1017–20. http://science.sciencemag.org/.
Laskowski, Kate L, Chia-Chen Chang, Kirsten Sheehy, and Jonathan Aguiñaga. 2024. ‘Consistent Individual Behavioral Variation: What Do We Know and Where Are We Going?’ Evolution, and Systematics Downloaded from Www.Annualreviews.Org. Guest (Guest. https://doi.org/10.1146/annurev-ecolsys-102220.
Laspoumaderes, Cecilia, Beatriz Modenutti, María Sol Souza, Marcela Bastidas Navarro, Florencia Cuassolo, and Esteban Balseiro. 2013. ‘Glacier Melting and Stoichiometric Implications for Lake Community Structure: Zooplankton Species Distributions across a Natural Light Gradient’. Global Change Biology 19 (1): 316–26. https://doi.org/10.1111/gcb.12040.
Leahy, Susannah M., Mark I. McCormick, Matthew D. Mitchell, and Maud C.O. Ferrari. 2011. ‘To Fear or to Feed: The Effects of Turbidity on Perception of Risk by a marine FIsh’. Biology Letters 7 (6): 811–13. https://doi.org/10.1098/rsbl.2011.0645.
Lehtiniemi, M. 2005. ‘Swim or Hide: Predator Cues Cause Species Specific Reactions in Young Fish Larvae’. J Fish Biology 66:1285–99. https://doi.org/10.1111/j.1095-8649.2005.00681.x.
Lenth, Russell V., Paul Buerkner, Maxime Herve, Jonathon Love, Hannes Riebl, and Henrik Singmann. 2020. ‘Emmeans: Estimated Marginal Means, Aka Least-Squares Means. R Package Version 1.4.6.’ https://cran.r-project.org/package=emmeans.
Li, Heng. 2013. ‘Aligning Sequence Reads, Clone Sequences and Assembly Contigs with BWA-MEM’. ArXiv Preprint, arXiv:1303.3997. https://doi.org/10.1186/s13756-018-0352-y.
Li, Heng, Bob Handsaker, Alec Wysoker, Tim Fennell, Jue Ruan, Nils Homer, Gabor Marth, Goncalo Abecasis, Richard Durbin, and 1000 Genome Project Data Processing Subgroup. 2009. ‘The Sequence Alignment/Map Format and SAMtools’. Bioinformatics 25 (16): 2078–79. https://doi.org/10.1093/bioinformatics/btp352.
Liang, S., T. Wang, X. Hu, J. Luo, W. Li, X. Wu, Y. Duan, and F. Jin. 2015. ‘Administration of Lactobacillus Helveticus NS8 Improves Behavioral, Cognitive, and Biochemical Aberrations Caused by Chronic Restraint Stress’. Neuroscience 310 (December):561–77. https://doi.org/10.1016/j.neuroscience.2015.09.033.
Lima, Steven L, and Lawrence M Dill. 1980. ‘Behavioral Decisions Made under the Risk of Predation: A Review and Prospectus’. Canadian Journal of Zoology 68:619–40. www.nrcresearchpress.com.
Ling, Fei, Natalie Steinel, Jesse Weber, Lei Ma, Chris Smith, Decio Correa, Bin Zhu, Daniel Bolnick, and Gaoxue Wang. 2020. ‘The Gut Microbiota Response to Helminth Infection Depends on Host Sex and Genotype’. ISME Journal 14 (5): 1141–53. https://doi.org/10.1038/s41396-020-0589-3.
LoBue, C P, and M A Bell. 1993. ‘PHENOTYPIC MANIPULATION BY THE CESTODE PARASITE SCHISTOCEPHALUS SOLIDUS OF ITS INTERMEDIATE HOST, GASTEROSTEUS ACULEATUS, THE THREESPINE STICKLEBACK’. The American Naturalist 142 (4): 725–35.
Lu, Jennifer, Florian P. Breitwieser, Peter Thielen, and Steven L. Salzberg. 2017. ‘Bracken: Estimating Species Abundance in Metagenomics Data’. PeerJ Computer Science 3:e104. https://doi.org/10.7717/peerj-cs.104.
Lucek, Kay, Bjarni K. Kristjánsson, Skúli Skúlason, and Ole Seehausen. 2016. ‘Ecosystem Size Matters: The Dimensionality of Intralacustrine Diversification in Icelandic Stickleback Is Predicted by Lake Size’. Ecology and Evolution 6 (15): 5256–72. https://doi.org/10.1002/ece3.2239.
Lyte, Joshua M., Lucas R. Koester, Karrie M. Daniels, and Mark Lyte. 2022. ‘Distinct Cecal and Fecal Microbiome Responses to Stress Are Accompanied by Sex- and Diet-Dependent Changes in Behavior and Gut Serotonin’. Frontiers in Neuroscience 16 (April). https://doi.org/10.3389/fnins.2022.827343.
Mallick, Himel, Ali Rahnavard, Lauren J. McIver, Siyuan Ma, Yancong Zhang, Long H. Nguyen, Timothy L. Tickle, et al. 2021. ‘Multivariable Association Discovery in Population-Scale Meta-Omics Studies’. PLOS Computational Biology 17 (11): e1009442. https://doi.org/10.1371/JOURNAL.PCBI.1009442.
Marchesi, Julian R, David H Adams, Francesca Fava, Gerben D a Hermes, Gideon M Hirschfield, Georgina Hold, Mohammed Nabil Quraishi, et al. 2015. ‘The Gut Microbiota and Host Health: A New Clinical Frontier’. Gut , 1–10. https://doi.org/10.1136/gutjnl-2015-309990.
McMurdie, Paul J., and Susan Holmes. 2013. ‘Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data’. PLOS ONE 8 (4): e61217. https://doi.org/10.1371/JOURNAL.PONE.0061217.
Milinski, and Heller. 1978. ‘Influence of a Predator on Optimal Foraging Behaviour of Sticklebacks’. Nature 275:642–44.
Näslund, Joacim, Leo Pettersson, and Jörgen I Johnsson. 2017. ‘Behavioural Reactions of Three-Spined Sticklebacks to Simulated Risk of Predation—Effects of Predator Distance and Movement’. FACETS 1 (1): 55–66. https://doi.org/10.1139/facets-2015-0015.
Nath, Shivangi, Daniel E. Shaw, and Michael A. White. 2021. ‘Improved Contiguity of the Threespine Stickleback Genome Using Long-Read Sequencing’. G3 Genes|Genomes|Genetics 11 (2). https://doi.org/10.1093/G3JOURNAL/JKAB007.
Natsopoulou, M. E., S. Pálsson, and G. Á Ólafsdóttir. 2012. ‘Parasites and Parallel Divergence of the Number of Individual MHC Alleles between Sympatric Three-Spined Stickleback Gasterosteus Aculeatus Morphs in Iceland’. Journal of Fish Biology 81 (5): 1696–1714. https://doi.org/10.1111/j.1095-8649.2012.03430.x.
Nyholm, Lasse, Adam Koziol, Sofia Marcos, Amanda Bolt Botnen, Ostaizka Aizpurua, Shyam Gopalakrishnan, Morten T. Limborg, M. Thomas P. Gilbert, and Antton Alberdi. 2020. ‘Holo-Omics: Integrated Host-Microbiota Multi-Omics for Basic and Applied Biological Research’. IScience 23 (8): 101414. https://doi.org/10.1016/j.isci.2020.101414.
Ōmura, Satoshi, Kazuhiko Otoguro, Takaaki Nishikiori, Ruiko Ōiwa, and Yuzuru Iwai. 1981. ‘SETAMYCIN, A NEW ANTIBIOTIC’. The Journal of Antibiotics 34 (10): 1253–56. https://doi.org/10.7164/ANTIBIOTICS.34.1253.
Plieskatt, Jordan L., Raksawan Deenonpoe, Jason P. Mulvenna, Lutz Krause, Banchob Sripa, Jeffrey M. Bethony, and Paul J. Brindley. 2013. ‘Infection with the Carcinogenic Liver Fluke Opisthorchis Viverrini Modifies Intestinal and Biliary Microbiome’. FASEB Journal 27 (11): 4572–84. https://doi.org/10.1096/fj.13-232751.
Poulin, Robert. 1995. ‘“Adaptive” Changes in the Behaviour of Parasitized Animals: A Critical Review’. International Journal for Parasitology 25 (12): 1371–83. https://doi.org/10.1016/0020-7519(95)00100-X.
Puetz, Lara C, Tom O Delmont, Ostaizka Aizpurua, Chunxue Guo, Guojie Zhang, Rebecca Katajamaa, Per Jensen, et al. 2021. ‘Gut Microbiota Linked with Reduced Fear of Humans in Red Junglefowl Has Implications for Early Domestication’. Advanced Genetics 2 (4): 2100018. https://doi.org/10.1002/GGN2.202100018.
Quinlan, Aaron R., and Ira M. Hall. 2010. ‘BEDTools: A Flexible Suite of Utilities for Comparing Genomic Features’. Bioinformatics 26 (6): 841–42. https://doi.org/10.1093/bioinformatics/btq033.
Rennison, Diana J., Seth M. Rudman, and Dolph Schluter. 2019. ‘Parallel Changes in Gut Microbiome Composition and Function during Colonization, Local Adaptation and Ecological Speciation’. Proceedings of the Royal Society B: Biological Sciences 286 (1916). https://doi.org/10.1098/rspb.2019.1911.
Restivo, Victoria E., Karen A. Kidd, Michael G. Surette, Carol Bucking, and Joanna Y. Wilson. 2021. ‘The Gut Content Microbiome of Wild-Caught Rainbow Darter Is Altered during Laboratory Acclimation’. Comparative Biochemistry and Physiology Part D: Genomics and Proteomics 39 (September):100835. https://doi.org/10.1016/J.CBD.2021.100835.
Reynolds, Lisa A., B. Brett Finlay, and Rick M. Maizels. 2015. ‘Cohabitation in the Intestine: Interactions among Helminth Parasites, Bacterial Microbiota, and Host Immunity’. The Journal of Immunology 195 (9): 4059–66. https://doi.org/10.4049/jimmunol.1501432.
Ritchie, Matthew E., Belinda Phipson, Di Wu, Yifang Hu, Charity W. Law, Wei Shi, and Gordon K. Smyth. 2015. ‘Limma Powers Differential Expression Analyses for RNA-Sequencing and Microarray Studies’. Nucleic Acids Research 43 (7): e47–e47. https://doi.org/10.1093/NAR/GKV007.
Rödl, Thomas, Silke Berger, L. Michael Romero, and Martin Wikelski. 2007. ‘Tameness and Stress Physiology in a Predator-Naive Island Species Confronted with Novel Predation Threat’. Proceedings of the Royal Society B: Biological Sciences 274 (1609): 577–82. https://doi.org/10.1098/rspb.2006.3755.
Rosshart, Stephan P., Brian G. Vassallo, Davide Angeletti, Diane S. Hutchinson, Andrew P. Morgan, Kazuyo Takeda, Heather D. Hickman, et al. 2017. ‘Wild Mouse Gut Microbiota Promotes Host Fitness and Improves Disease Resistance’. Cell 171 (5): 1015-1028.e13. https://doi.org/10.1016/J.CELL.2017.09.016.
Schmieder, Robert, and Robert Edwards. 2011. ‘Quality Control and Preprocessing of Metagenomic Datasets’. Bioinformatics 27 (6): 863–64. https://doi.org/10.1093/bioinformatics/btr026.
Schneider, Valerie A., Tina Graves-Lindsay, Kerstin Howe, Nathan Bouk, Hsiu-Chuan Chen, Paul A. Kitts, Terence D. Murphy, et al. 2017. ‘Evaluation of GRCh38 and de Novo Haploid Genome Assemblies Demonstrates the Enduring Quality of the Reference Assembly’. Genome Research 27 (5): 849–64. https://doi.org/10.1101/gr.213611.116.
Schretter, Catherine E. 2020. ‘Links between the Gut Microbiota, Metabolism, and Host Behavior’. Gut Microbes. NLM (Medline). https://doi.org/10.1080/19490976.2019.1643674.
Schretter, Catherine E., Jost Vielmetter, Imre Bartos, Zsuzsa Marka, Szabolcs Marka, Sulabha Argade, and Sarkis K. Mazmanian. 2018. ‘A Gut Microbial Factor Modulates Locomotor Behaviour in Drosophila’. Nature 563 (7731): 402–6. https://doi.org/10.1038/s41586-018-0634-9.
Schubert, Mikkel, Stinus Lindgreen, and Ludovic Orlando. 2016. ‘AdapterRemoval v2: Rapid Adapter Trimming, Identification, and Read Merging’. BMC Research Notes 9:88. https://doi.org/10.1186/s13104-016-1900-2.
Schulte, Patricia M. 2014. ‘What Is Environmental Stress? Insights from Fish Living in a Variable Environment’. Journal of Experimental Biology. Company of Biologists Ltd. https://doi.org/10.1242/jeb.089722.
Shankregowda, Aruna M., Prabhugouda Siriyappagouder, Marijn Kuizenga, Thijs M.P. Bal, Yousri Abdelhafiz, Christophe Eizaguirre, Jorge M.O. Fernandes, Viswanath Kiron, and Joost A.M. Raeymaekers. 2023. ‘Host Habitat Rather than Evolutionary History Explains Gut Microbiome Diversity in Sympatric Stickleback Species’. Frontiers in Microbiology 14. https://doi.org/10.3389/fmicb.2023.1232358.
Sih, Andrew, Alison Bell, and J. Chadwick Johnson. 2004. ‘Behavioral Syndromes: An Ecological and Evolutionary Overview’. Trends in Ecology and Evolution. https://doi.org/10.1016/j.tree.2004.04.009.
Smith, Carli J., Jacob R. Emge, Katrina Berzins, Lydia Lung, Rebecca Khamishon, Paarth Shah, Colin Reardon, et al. 2014. ‘Probiotics Normalize the Gut-Brain-Microbiota Axis in Immunodeficient Mice’. American Journal of Physiology - Gastrointestinal and Liver Physiology 307 (8): G793–802. https://doi.org/10.1152/ajpgi.00238.2014.
Smith, Chris C.R., Lisa K. Snowberg, J. Gregory Caporaso, Rob Knight, and Daniel I. Bolnick. 2015. ‘Dietary Input of Microbes and Host Genetic Variation Shape Among-Population Differences in Stickleback Gut Microbiota’. The ISME Journal 2015 9:11 9 (11): 2515–26. https://doi.org/10.1038/ismej.2015.64.
Takahashi, Yoko. 2017. ‘Genus Kitasatospora, Taxonomic Features and Diversity of Secondary Metabolites’. Journal of Antibiotics. Nature Publishing Group. https://doi.org/10.1038/ja.2017.8.
Thomas, Frédéric, Shelley Adamo, and Janice Moore. 2005. ‘Parasitic Manipulation: Where Are We and Where Should We Go?’ Behavioural Processes 68 (3 SPEC. ISS.): 185–99. https://doi.org/10.1016/j.beproc.2004.06.010.
Thorburn, Doko Miles J., Thijs M.P. Bal, Io S. Deflem, Filip A.M. Volckaert, Christophe Eizaguirre, and Joost A.M. Raeymaekers. 2022. ‘Context-Dependent Parasite Infection Affects Trophic Niche in Populations of Sympatric Stickleback Species’. Parasitology 149 (9): 1164–72. https://doi.org/10.1017/S0031182022000531.
Varg, Javier Edo, Christoffer Bergvall, and Richard Svanbäck. 2022. ‘The Stressful Effects of Microplastics Associated With Chromium (VI) on the Microbiota of Daphnia Magna’. Frontiers in Environmental Science 10 (May): 1–13. https://doi.org/10.3389/fenvs.2022.875512.
Vaughan, Jefferson A., Vasyl V. Tkach, and Stephen E. Greiman. 2012. ‘Neorickettsial Endosymbionts of the Digenea: Diversity, Transmission and Distribution’. In Advances in Parasitology, 79:253–97. Academic Press. https://doi.org/10.1016/B978-0-12-398457-9.00003-2.
Viver, Tomeu, Alberto Ruiz, Edgar Bertomeu, Martina Martorell-Barceló, Mercedes Urdiain, Amalia Grau, Marco Signaroli, et al. 2023. ‘Food Determines Ephemerous and Non-Stable Gut Microbiome Communities in Juvenile Wild and Farmed Mediterranean Fish’. Science of the Total Environment 889 (September). https://doi.org/10.1016/j.scitotenv.2023.164080.
Vuong, Helen E, Jessica M Yano, Thomas C Fung, and Elaine Y Hsiao. 2017. ‘The Microbiome and Host Behavior’. https://doi.org/10.1146/annurev-neuro-072116.
Wang, Haisheng, Xiaoxia Zhang, Shuangyan Wang, Baisuo Zhao, Kai Lou, and Xin Hui Xing. 2018. ‘Massilia Violaceinigra Sp. Nov., a Novel Purple-Pigmented Bacterium Isolated from Glacier Permafrost’. International Journal of Systematic and Evolutionary Microbiology 68 (7): 2271–78. https://doi.org/10.1099/IJSEM.0.002826/CITE/REFWORKS.
Ward, Ashley J.W., Philip Thomas, Paul J.B. Hart, and Jens Krause. 2004. ‘Correlates of Boldness in Three-Spined Sticklebacks (Gasterosteus Aculeatus)’. Behavioral Ecology and Sociobiology 55 (6): 561–68. https://doi.org/10.1007/s00265-003-0751-8.
White, Emily C., Ashley Houlden, Allison J. Bancroft, Kelly S. Hayes, Marie Goldrick, Richard K. Grencis, and Ian S. Roberts. 2018. ‘Manipulation of Host and Parasite Microbiotas: Survival Strategies during Chronic Nematode Infection’. Science Advances 4 (3): 1–10. https://doi.org/10.1126/sciadv.aap7399.
Williams, Andrew R., Laura J. Myhill, Sophie Stolzenbach, Peter Nejsum, Helena Mejer, Dennis S. Nielsen, and Stig M. Thamsborg. 2021. ‘Emerging Interactions between Diet, Gastrointestinal Helminth Infection, and the Gut Microbiota in Livestock’. BMC Veterinary Research. BioMed Central Ltd. https://doi.org/10.1186/s12917-021-02752-w.
Wood, Derrick E., Jennifer Lu, and Ben Langmead. 2019. ‘Improved Metagenomic Analysis with Kraken 2’. Genome Biology 20 (1): 1–13. https://doi.org/10.1186/S13059-019-1891-0/FIGURES/2.
Young, Rebecca E., and Andrew D.C. MacColl. 2016. ‘Spatial and Temporal Variation in Macroparasite Communities of Three-Spined Stickleback’. Parasitology 144 (4): 436–49. https://doi.org/10.1017/S0031182016001815.
Zeng, Yonghui, Yonghui Nupur, Naicheng Wu, Anne Mette Madsen, Xihan Chen, Alastair T. Gardiner, and Michal Koblížek. 2021. ‘Gemmatimonas Groenlandica Sp. Nov. Is an Aerobic Anoxygenic Phototroph in the Phylum Gemmatimonadetes’. Frontiers in Microbiology 11 (January). https://doi.org/10.3389/FMICB.2020.606612/FULL.

No competing interests reported.

BehaviormicrobiomeinsticklebacksSI.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Interplay between antipredator behavior, parasitism, and gut microbiome in wild stickleback populations

Status:

Version 1

Abstract

Background

Results

Conclusions

Figures

1. Introduction

2. Methods

2.1 Sampling and fieldwork

2.2 Behavioral experiments

2.3 Characterization of behaviour

2.3.1 Arena for control condition

2.3.2 Arena for treatment condition

2.5 DNA Sequence data processing, taxonomic assignment and taxa filtering

2.6 Statistical analysis

2.6.1 Behavior

2.5.2 Microbiome

3. Results

3.1 Behavior

3.1.1 Effect of the presence of the robotic predator before simulated predator attack

3.1.2 Effect of the simulated predatory attack

3.1.3 Effect of parasites on anti-predator behavior

3.2 Microbiome

3.3 Links between fish behaviors and microbiomes

4. Discussion

4.1. Differences between populations and effects of parasitism in behavior

4.2 Differences between populations and effects of parasitism in the microbiome

4.3 Links between behavior and microbiome

Declarations

Ethics approval and consent to participate

Competing interests

Funding

Author Contribution

Acknowledgement

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1