Abstract
Behavioural syndromes are composed of correlated suites of personality traits and can include traits related to the behaviour and ecology of free-ranging animals. We used captive little brown bats (Myotis lucifugus) to test the hypothesis that behaviours measured in standardized tests reflect personality traits and form behavioural syndromes with roosting behaviours. We predicted: (1) measured behaviours would be repeatable; (2) personality traits and roosting behaviours would form behavioural syndromes; and (3) individuals with similar personality scores would associate more strongly. We observed repeatability for some traits and evidence of behavioural syndromes. Activity was strongly repeatable across time and contexts. More central individuals roosted in larger groups, while individuals with high roost-fidelity roosted in larger groups. Individuals with similar activity scores were also more likely to associate in day roosts, suggesting some behavioural assortment. Our results have implications for how behavioural variation might influence transmission of white-nose syndrome.
1. Introduction
Animal personality is defined as consistent individual variation in a behavioural trait expressed across time and, like other behaviours (e.g., Roberts & Sherratt, 2002), animal personality traits can affect fitness (Sih et al., 2004; Réale et al., 2007). Personality traits have been identified for a range of vertebrates (Bell et al., 2009) and identifying these traits is important for understanding behavioural evolution (Sih et al., 2012). Some personality traits are better suited to certain environmental conditions than others and, therefore, environmental variation should select for a range of personality traits within populations (Sih et al., 2012; Nicolaus et al., 2016). For instance, temporal fluctuation in food abundance maintained population-level variation in personality of red squirrels (Tamiasciurus hudsonicus) because more aggressive females had higher fitness in years with abundant food but lower fitness in years of food scarcity, presumably because of greater resource acquisition ability but greater energetic costs (Boon et al., 2007). This kind of fluctuating selection should promote behavioural diversity within populations which, in turn, could lead to selection favouring correlated suites of behaviours (Sih et al., 2012). Such correlations among behavioural traits are known as ‘behavioural syndromes’. Behavioural syndromes can include behaviours not typically associated with animal personality and can have implications for fitness (Biro & Stamps, 2008; Sih et al., 2012).
Linking personality with other behaviours, such as sociality, may help explain the adaptive benefits of personality (Aplin et al., 2013; Korsten et al., 2013). For example, in sticklebacks (Gasterosteus aculeatus), bold individuals had many weak social connections while shy individuals had fewer but stronger social connections (Pike et al., 2008), while in common lizards (Lacerta vivipara) more social individuals had higher fitness (Cote et al., 2008). If personality influences associations among conspecifics via social assortment (i.e., individuals with similar combinations of traits are more likely to associate: Croft et al., 2009), personality could influence fitness depending on the behavioural phenotypes of associating individuals (Smith & Blumstein, 2008). Recent empirical evidence indicates that social behaviours are repeatable (Aplin et al., 2015; Vander Wal et al., 2015) and social behaviour could be related to some personality traits contributing to behavioural syndromes.
Temperate-zone bat species could provide a good study system to test for correlations among personality traits and social behaviour because females of many species form fission-fusion societies that, in summer, are characterized by frequent roost switching and daily changes in group composition through merging (i.e., fusion) and splitting (i.e., fission) of sub-groups (Kerth & Konig, 1999; Willis & Brigham, 2004; Garroway & Broders, 2007; Patriquin et al., 2010, 2016). In bats, fission-fusion dynamics and social association are non-randomly distributed (Silvis et al., 2014; Webber et al., 2016) and not necessarily mediated by kinship (Metheny et al., 2008; but see Patriquin et al., 2013) suggesting a possible role for alternative mechanisms, like animal personality, to influence roosting associations. For instance, as colonies merge and split, day-roosting group size can vary over time, while roost fidelity can also vary if individuals prefer or avoid certain roosts. Personality, or social assortment of individuals with similar combinations of traits, could influence this variation (Croft et al., 2003). For instance, in guppies (Poecilia reticulata), more social individuals were more likely to shoal together than with less social individuals (Croft et al., 2005), a finding which could extend to bat fission-fusion dynamics and help explain why some pairs of individuals in a given colony tend to roost together more often than others (Willis & Brigham, 2004). The time-scale associated with the rate of fission-fusion for bat maternity colonies (i.e., days to weeks) can influence a wide range of ecological processes, such as transmission of information or pathogens and parasites (Reckardt & Kerth, 2007) or access to food and other resources (Kerth & Reckardt, 2003).
Understanding the roosting and social behaviour of North American bat species has also become an important priority from a ‘conservation behaviour’ perspective (Sutherland, 1998) because of extraordinary population declines from the invasive fungal disease white-nose syndrome (WNS; Frick et al., 2010, 2017). Understanding the influence of social assortment on roost selection decisions of bats could be important for understanding transmission of the fungal pathogen that causes WNS, identifying critical summer habitats for protection and enhancement and the potential for evolutionary responses of bat populations to WNS. For example, individuals with highly sociable or explorative personality traits, or combinations of traits, may be more likely to transmit the fungus that causes WNS potentially acting as super-spreaders (Hoyt et al., 2018). Conversely, individuals with less sociable or exploratory personalities could be less likely to contract high fungal loads and, if these personality traits are heritable in bats, WNS could drive directional selection on personality, given high rates of mortality and strong selection differentials currently at play (Willis, 2015). Preliminary evidence suggests that WNS survivors may have reduced social aggregation during hibernation (Langwig et al., 2012) which is thought to reflect within-individual behavioural change (i.e., sickness behaviour; Brownlee-Bouboulis & Reeder, 2013; Bohn et al., 2016) but could also reflect increased survival of less sociable individuals and directional selection on sociality (Willis, 2015). This could have pronounced implications for the habitat requirements of bats. For example, if WNS drives the evolution of reduced sociability and smaller summer colony sizes, significantly more high quality roosting habitat may be required if populations are ever to recover to pre-WNS numbers of individuals. Quantifying natural variation in behavioural traits of healthy bats is therefore an important conservation priority (Willis, 2015) and generating predictions about whether individuals with certain combinations of traits may be more likely to acquire WNS could inform models projecting long-term outcomes of WNS (e.g., Maslo & Fefferman, 2015).
Long-term recapture rates for free-ranging bats are typically low and, while repeatability of some social behaviours has been demonstrated over timescales ranging from a few days to several years (e.g., contact calls of free-ranging Thyroptera tricolor: Chaverri & Gillam, 2015), to date, repeatability of putative personality traits has only been quantified in bats on a short-term basis (i.e., 24 h: Menzies et al. 2013). Quantifying behavioural variation for captive bats held in a semi-natural context could provide an opportunity to understand links between personality and social or roosting behaviours which may otherwise be extremely challenging to quantify for completely free-ranging animals in the wild.
We used captive members of a colony of adult female little brown bats (Myotis lucifugus), housed in a semi-natural flight enclosure, to test the hypothesis that behavioural traits, measured in standardized hole-board and Y-maze tests (hole-board: Martin & Réale, 2008; Y-maze: Kilgour et al., 2013; see methods for descriptions) reflect personality and form behavioural syndromes that mediate roosting and social behaviours. We predicted: (1) that putative personality traits quantified in standardized tests, and roosting behaviours quantified in the flight enclosure, would be repeatable over time and contexts; (2) that putative personality traits would correlate with each other as well as with social and roosting behaviours of bats held in the flight enclosure, thus forming behavioural syndromes; and (3) that individuals with similar personality scores for a given trait would have stronger social associations than individuals with dissimilar personality scores (Croft et al., 2009), thus displaying social assortment based on personality.
2. Methods
2.1. Study site
Although our study site was negative for Pseudogymnoascus destructans, the fungal pathogen that causes WNS, we followed U.S. Fish and Wildlife Service (USFWS) and Canadian Wildlife Health Cooperative (CWHC) guidelines for decontamination by researchers (Canadian Wildlife Health Cooperative, 2015; United States Fish and Wildlife Service, 2016).
We caught little brown bats (Myotis lucifugus) on 5 June 2014 from two multi-chambered bat boxes that were mounted back-to-back on a post on private property near Nutimik Lake, Manitoba, Canada (50.14°N, 95.69°W). Based on exit counts, these boxes housed approximately 250 bats during spring and early summer. We captured 43 bats at dusk using bucket traps attached to the bottom of each bat box (i.e., a tube with approximately 75 cm diameter consisting of 50 cm of plastic sheeting attached to an additional 50 cm of mesh screening with a blind bottom). Bats slid or flew down the plastic sheeting and remained in the blind mesh bottom for approximately 30 min. Immediately after capture, we identified non-reproductive adult females by gently palpating the abdomen and we released two detectably pregnant bats immediately upon capture. We implanted a uniquely coded passive transponder (PIT tag, ID 100-01, Trovan Douglas, UK) subcutaneously between the scapulae for permanent identification for each bat. PIT tags allowed us to record roost preferences and social contacts with other bats during captivity and reduced the need to handle individuals. The night of capture was therefore the first night of captivity (for detailed timeline see Figure A1 in the Appendix).
2.2. Transport and housing
We retained 40 non-pregnant bats for our behavioural experiments. All bats were transported from the capture site approximately 80 km to the Sandilands Forest Discovery Centre (49.40°N, 95.54°W). During transport, bats were suspended in cloth bags in groups of three or four inside a ventilated picnic cooler to dampen noise during transport. During captivity, bats were housed in a nylon mesh flight tent (2.75 × 2.75 × 2.75 m), previously described by Webber & Willis (2018) that had a shaded roof but was otherwise open to ambient conditions. We allowed bats to acclimate to the flight tent for the first five days and nights of captivity, a timescale similar to behavioural studies of other vertebrates (Jacoby et al., 2014; von Merten et al., 2017). An average of 10% (range 2–20%) of bats per night roosted on the mesh wall of the tent outside one of the four roost boxes in the flight tent during this 5-day period. However, after the acclimation period only 3% (range 0–15%) of bats per night roosted outside one of the four roost boxes, and no bats did this more than once, suggesting that 5 nights was long enough for bats to discover, investigate and reliably use the roost boxes.
On the second night of captivity we began training bats to eat mealworms independently. Mealworms were gut-loaded with beta-carotene multivitamins (Herptivite, Rep-Cal, Campbell, CA, USA) and nutrient supplement (Repashy Superfoods, Oceanside, CA, USA). Throughout captivity bats were provided water and mealworms (larval Tenebrio molitor) ad libitum. Bats that failed to learn to eat mealworms independently during the next three nights were released at the capture site (
The flight tent was outfitted with four roost boxes constructed from cleanable ‘vinyl plywood’ each with a volume of 3000 cm3. Each bat box was outfitted with a reptile heating mat (160 cm2, HMA-4, All Living Things™, Living Things, Orefield, PA, USA), which we used in a concurrent study (Webber & Willis, 2018) to manipulate roost temperature. In that study, we found that group size varied as a function of roost temperature and that larger groups tended to occupy heated roosts. However, all individuals were exposed to the same variation in available roost temperatures and we accounted for potential effects of roost temperature on individually-based behaviours in our analyses (see below).
Each roost box was mounted on a 1.5 m stand and was outfitted with a rectangular PIT-tag antenna connected to a decoder/data-logger (10 × 30 cm: LID650N, v727, Dorset Identification, Aalten, The Netherlands). Each antenna surrounded the roost entrance so that, whenever an individual entered or exited a roost box, its unique PIT-tag code was recorded and time-stamped by the data logger. Decoder/data-loggers were powered via an AC outlet from a nearby building and the system operated continuously for the duration of the study. This captive experimental set-up was artificial in the sense that potential for night-time movements of bats was significantly curtailed relative to normal home-range size for free-ranging little brown bats. Nevertheless, Webber & Willis (2018) found that bats acclimated to this roosting system within only a few days after capture from the wild and exhibited predictable roosting behaviour similar to that of free-ranging bats (Table 1). We argue that this roosting enclosure study system can provide a useful perspective on roosting and social behaviours of bats that would be extremely difficult to collect for entirely free-ranging individuals. The study system provides a range of roost options, and a range of roosting partners, that would have allowed bats to switch among roosts and roost-mates normally, reflecting typical patterns of fission-fusion behaviour observed for free-ranging individuals. Despite the artificial nature of the flight tent, bats behaved similarly to free-ranging individuals for a range of behaviours (see Table 1) such as group size (Willis & Brigham, 2007) roost switching frequency (Willis & Brigham, 2004), social network centrality (Silvis et al., 2014) and within-night movement and activity (Anthony et al., 1981; Barclay, 1982).
Comparison of social and roosting behaviours quantified for a captive colony of Myotis lucifugus with the same, or similar, behaviour quantified for free-ranging bats in North America.
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
2.3. Measuring personality: hole-board test
We quantified activity (i.e., an individual’s general activity patterns) and exploration (i.e., an individual’s reaction to a novel object or situation) using a hole-board test (Figure A2 in the Appendix). The hole-board test consisted of a test arena (57 cm wide by 42 cm tall by 14 cm deep), with a transparent plexiglass cover, and window screening on the back surface to facilitate climbing by the bats (Figure A2 in the Appendix). Four-blind holes (2 cm wide by 1 cm deep) were positioned in the backboard material and vertical and horizontal lines intersected the hole-board to delineate four sections of the test arena (Figure A2 in the Appendix). A start chamber (16 cm long by 8 cm diameter tube) was fastened to the base of the test, with a sliding door to separate the animal from the main arena. The test apparatus was hung vertically so bats could crawl on the backboard material and explore the blind holes.
Hole-board tests were ten minutes long and occurred at night during the active phase and were video-recorded under infrared illumination (AVCHD NightShot handycam HDR-XR550, Sony, Tokyo, Japan). At the start of each trial, a bat was placed in the start chamber and the sliding door was opened. Each bat was given 60 s to voluntarily enter the test chamber before being gently pushed in using a plastic plunger. For the 16% (
Ethogram of twelve behaviours quantified in video recordings of hole-board and Y-maze tests conducted on 34 little brown bats.
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Hole-board tests have commonly been used to assay personality traits for rodents (e.g., Martin & Réale, 2008) and, more recently, bats (Menzies et al., 2013; Webber et al., 2015). Behavioural responses of bats from the hole-board test appear to meet the criteria for personality because responses of individuals are repeatable, at least over the short-term (Menzies et al., 2013). However, because Menzies et al. (2013) were only able to assess short-term repeatability of traits assessed in the hole-board test, we measured each personality trait twice, with 11–14 days between measurements to assess repeatability over a slightly longer term.
2.4. Measuring personality: Y-maze test
We used a Y-maze test, modified from Kilgour et al. (2013), to assess sociability (i.e., an individual’s reaction to the presence or absence of a conspecific). The Y-maze test consisted of a Y-shaped plexiglass maze (long end: 37 cm long; forked ends: 20 cm long; 6 cm wide by 10 cm tall) (see Figure 1 from Kilgour et al., 2013). We randomly selected one bat from the capture site as the designated ‘stimulus’ bat for all Y-maze trials. We chose to use the same stimulus bat for all Y-maze trials for two reasons. First, it would have been impossible to capture enough of this endangered species to hold a different stimulus individual for each Y-maze test. Second, using the same stimulus bat for each test controlled for potential differences in the stimulus individual (e.g., acoustic, olfactory or behavioural cues produced by the stimulus bat) to other individuals from the colony. To account for potential bias resulting from contact between the focal bat (i.e., the individuals being tested) and the stimulus bat prior to the trial, we followed standard methods from social memory studies (Ferguson et al., 2000) and housed the stimulus bat separately from the main colony during captivity. Focal bats did not interact with the stimulus bat except during Y-maze tests. For the first round of trials the stimulus bat was at the end of the left arm of the Y-maze and for the second round it was on the right side. Each trial began when we placed the focal bat at the long end of the Y-maze and the trial concluded after 5 min. The stimulus bat was held in a small stainless-steel mesh cage (20 × 20 × 20 cm) located at the end of one of the arms of the Y, while an identical empty cage was placed at the end of the other arm. The stimulus bat was held in this cage for up to 4 h before being returned to its housing enclosure. For Y-maze tests we scored duration of locomotion, duration of grooming, number of faecal pellets, duration of echolocation, time spent within 10 cm of the stimulus bat divided by total trial time, and latency to approach within 10 cm of stimulus bat (see Table 2 for descriptions of all behaviours). To date repeatability of behavioural traits measured in the Y-maze test has not been assessed for bats. Y-maze tests were five minutes long and, as for hole-board, occurred at night and were video-recorded under infrared illumination (AVCHD NightShot handycam HDR-XR550, Sony).
After completion of hole-board and Y-maze tests we returned bats to the flight tent and cleaned the testing apparatus with unscented disinfecting wipes (Lysol, Reckitt Benckiser, Mississauga, ON, Canada) to comply with North American decontamination guidelines for P. destructans (Canadian Wildlife Health Cooperative, 2015; United States Fish and Wildlife Service, 2016) and to prevent residual scent from influencing the behaviour of subsequent bats. To improve efficiency, we used two identical test chambers (both hole-board and Y-maze tests) so that one could be used while the other was being cleaned. The order that individuals were tested in the hole-board and Y-maze tests were randomized within nights, therefore altering the order an individual was tested from the first to second trial to ensure behavioural responses we observed were not a function of the time of day. We accounted for the time of each trial in our repeatability analyses (see below). Videos were scored by QMRW for a range of behavioural traits that have been used to quantify personality in rodents (e.g., Martin & Réale, 2008) and bats (Menzies et al., 2013; Webber et al., 2015) and QMRW was blind to the identity of bats in the videos while scoring.
2.5. Quantifying personalit y
All statistical analyses were conducted using R (R Core Team, 2019). We used principal component analysis (PCA) to reduce the large number of behavioural variables into a smaller number of components reflective of personality traits. Prior to conducting PCA, we confirmed that correlations existed among behavioural variables using Bartlett’s test and confirmed sampling adequacy using the Kaiser–Meyer–Olkin (KMO) test (see Tables A1–S4 in the Appendix for results, Budaev, 2010). Budaev (2010) highlighted that including multiple measures from the same individuals in a single PCA represents pseudo-replication, so we conducted four separate PCAs, one for each behavioural test (hole-board or Y-maze) and trial (one or two). We scaled and centred raw data by subtracting variable mean values from each individual value and dividing by the variable standard deviation using the ‘prcomp’ function in R. This generates a dataset with mean values of zero, which ensures that the first component describes the most variance. We retained components based on the Kaiser-Guttman criterion (eigenvalues > 1, Kaiser, 1991) and the parallel analysis method (Morton & Altschul, 2019). Given recent criticism of the Kaiser criterion (Morton & Altschul, 2019), in cases where the Kaiser criterion and parallel analysis disagreed, we visually inspected PCA results using a scree plot and chose the number of components based on the most conservative result. For the hole-board PCAs we condensed behavioural variables to determine activity and exploration scores for each individual (Menzies et al., 2013). For the Y-maze PCAs we condensed behavioural variables to quantify sociability (Kilgour et al., 2013) and activity for each individual. PC scores were then used to assess repeatability within tests (see below).
2.6. Measuring roosting behaviours
We quantified five roosting behaviours of individuals in the flight tent based on detection of PIT-tagged bats entering or exiting roost boxes. We selected behaviours that could be readily quantified using the PIT-tag systems and which we assumed would likely contribute to fission-fusion dynamics for free-ranging bats: day-roosting group size, day-roost fidelity (hereafter, roost fidelity), within-night movement/activity, roost-immergence timing, and communal night-roosting behaviour (measured as social network centrality, which we define as an individual’s connectivity to the rest of the network with each individual’s centrality value proportional to the sum of centralities of connected individuals as well as all of that individual’s connections; see below).
Day-roosting group size can vary from day-to-day for bats living in fission fusion societies and variation in group size can influence social thermoregulation (Willis & Brigham, 2007; Kuepper et al., 2016). For individuals roosting in larger groups, predicted energy expenditure is significantly lower than for bats roosting alone or in small groups (Willis & Brigham, 2007). We quantified day-roosting group size for each individual on each day by summing the total number of unique bats it roosted with each day.
RFI can range from −1 (complete infidelity), if the bat switched roosts every day, to 2 (complete fidelity), if the bat never moved (Chaverri & Kunz, 2006). For our closed system, RFI could over-estimate roost fidelity because bats only had four options for roosting sites, limiting the chance they might show complete infidelity (i.e., RFI = −1). While the minimum score in a system with an infinite number of roosts is −1, our study was limited to four roosts and the mesh lining of the tent (which no bat used more than once), so the minimum possible RFI value was approximately zero. However, little brown bats in the wild switch roosts approximately every two days (Olson & Barclay, 2013). The duration of our experiment (i.e., eight days), combined with the number of roost boxes we used, allowed bats to switch roosts at a normal frequency of approximately every two days.
General activity and movement patterns of free-ranging bats at night could influence foraging success as well as predation, with more active individuals facing higher risk of predation (Rydell et al., 1996). We quantified overall activity as the total number of times each individual was detected entering any roost box each night. We summed this total number of detections and used this variable as an index of movement patterns during the night (hereafter, within-night movement).
Emergence timing at dusk can also influence foraging success for insectivorous bats (Rydell et al., 1996), but less is known about immergence timing in the morning. Variation in immergence timing, especially for individuals that immerge during the dawn twilight or after sunrise, could reflect variation in foraging success and/or energetic status, while also influencing risk of predation by crepuscular or diurnal birds. We quantified timing of roost immergence based on the last detection of each individual each morning. We then calculated the difference between immergence time and sunrise time to assess each individual’s propensity to remain active in the morning. Sunrise occurred between 04:19 and 04:20 during the captivity period.
2.7. Social network analysis
After generating a network based on the SRI, we quantified ‘eigenvector centrality’ (hereafter, centrality), which refers to an individual’s social connectivity and each individual’s centrality value is proportional to the sum of the centralities of connected individuals as well as all of that individual’s connections (Csárdi & Nepusz, 2006). Thus, individuals with high centrality are connected to many individuals that are, in turn, connected to many other individuals. Social centrality reflects connectivity among individuals and has been linked to social rank-order (Sueur et al., 2011), transmission of information about food patches (Aplin et al., 2012) and fitness (Vander Wal et al., 2015).
2.8. Statistical analyses
To estimate repeatability of personality traits and roosting behaviours (i.e., day-roosting group size, roost immergence, within-night movement, roost fidelity and centrality) over time, we used the Bayesian modeling package ‘MCMCglmm’ in R (Hadfield, 2010). Bayesian analysis requires the specification of prior distributions for unknown parameters (Wilson et al., 2010), so we coded variance (s2) in our priors as s2/2 and degree of belief (nu) as one. All models were fit with Gaussian error structure and we assessed normality prior to analyses. We assessed autocorrelation for Markov Chain Monte Carlo (MCMC) chains using the ‘autocorr’ function to ensure autocorrelation was <0.1 (Hadfield, 2010). To ensure MCMC chains were uncorrelated we ran conservative MCMC chains of 1 300 000 iterations, a thinning length of 1000, and a burn-in of 300 000. Model convergence was confirmed visually.
Based on the distribution of repeatability values synthesized by Bell et al. (2009) we defined repeatability values of < 0.2 as weak, R values ⩾ 0.2, but ⩽ 0.40 as moderate, and R values > 0.4 as strong. We fit nine repeatability models, one for each personality trait and each roosting behaviour, except roost fidelity because Equation (1) provides a single value for each individual throughout captivity as opposed to daily values. For models of personality traits measured using the hole-board, we included trial (1 or 2), time of trial, and number of days between the first and second trial (range 11–14 days) as fixed effects. For models using the Y-maze test, we included trial (1 or 2) and time of trial as fixed effects. For roosting behaviour models, we included sampling night (1–8) and daily roost box identification as fixed effects. In addition, we also included a single model that assessed repeatability of activity across tests. For this model, we included time of trial, trial (1 or 2) in an interaction with test (hole-board or Y-maze) as fixed effects. For all models, we included individual identity as a random factor. We extracted individual best linear unbiased predictors (BLUPs) for each model and considered BLUPs as reflective of personality scores for all subsequent analyses.
2.9. Quantifying behavioural syndromes
We tested for behavioural syndromes by constructing a correlation matrix among the BLUPs we extracted for all personality traits and roosting behaviours. We used Spearman’s rank correlation coefficient for matrices to assess the direction and magnitude of correlations between variables. We accounted for multiple comparisons using Holm’s correction. All results are presented as the mean ± standard deviation.
Due to the inherent non-independence of social networks, we generated 1000 random social networks and calculated eigenvector centrality for each network. To test whether correlations that included eigenvector centrality were non-random, we re-generated corrected correlation matrices for all pairwise correlations and extracted correlation coefficients between eigenvector centrality and each of the other eight behaviours in our analysis. We considered observed correlation coefficients non-random if the observed coefficient fell outside the 95% confidence interval of the random distribution (Figure A3 in the Appendix). These comparisons allowed us to assess whether correlations between eigenvector centrality and other behaviours represented biologically meaningful relationships or, alternatively, whether they were the outcome of random combinations of individual connections.
2.10. Quantifying behavioural assortment
To assess whether individuals with similar personality scores had stronger social associations we used multiple regression quadratic assignment procedures (MRQAP, Dekker et al., 2007) in asnipe (Farine, 2013). MRQAPs are similar to Mantel tests, but allow permutation of regression coefficients in a matrix format (Dekker, Krackhardt & Snijders, 2007) with a single dependent matrix being regressed against one or more independent matrices (Farine & Whitehead, 2015). We ran 10 000 iterations of our MRQAP. Using the procedure described above (see ‘social network analysis’ section) we generated a single social association matrix using the SRI for all bats throughout captivity as the dependent matrix. We calculated similarity matrices for personality traits using BLUPs as reflective personality scores, where a dyad with a shared score of 0 had identical personality scores and shared scores > 0 indicate differences in personality with a larger number indicating a greater difference in individual personality scores for a dyad. To reduce the number of independent matrices in our model we a priori selected activity from the hole-board test, exploration, and sociability as candidate personality matrices and ran a single MRQAP model testing the effects of these personality matrices as predictors of social association.
3. Results
3.1. Personality traits
Based on parallel analysis, we retained the first two principal components for the first trial of hole-board tests and the first three principal components of the second trial of hole-board tests. The first two components explained 57.4% of the variance in results for the first test (Table A1 in the Appendix) and the first three components explained 71.3% of the variance for the second test (Table A2 in the Appendix). The first component (PC1H) for both PCAs was associated with activity-based behaviours such as locomotion, so we interpreted PC1H as an index of activity. The second component (PC2H) for the first hole-board PCA was associated with exploration-based behaviours such as number of head dips so we interpreted PC2H1 as an index of exploration. For the second hole-board PCA, the third component was associated with exploration-based behaviours, so we interpreted PC3H2 as an index of exploration.
Based on agreement between parallel analysis and the Kaiser criterion, we retained the first two components for the Y-maze PCAs. These explained 77.9% of the variance in results for the first Y-maze test (Table A3 in the Appendix) and 71.7% of the variance for the second test (Table A4 in the Appendix). As for the hole-board test, the first component of both PCAs (PC1Y) was associated with activity-based behaviours such as locomotion so we interpreted PC1Y as an index of activity. The second component (PC2Y) for both PCAs was associated with sociability-based behaviours such as duration spent within 10 cm of the stimulus bat (divided by trial duration), so we interpreted PC2Y as an index of sociability.
3.2. Roosting and social behaviours
Mean day roosting group size during captivity was 11.7 ± 2.6 (range: 2–29) bats per roost and roost immergence time was 11 ± 73 min from sunrise (range: 147 min before to 715 min after sunrise). On average, individual bats were detected 12.3 ± 6.2 (range 1–62) times per night for the night roosting social network analysis. Despite the relatively small number of available roosts, individuals switched roosts every 1.6 ± 0.4 days and showed low roost fidelity (RFI = 0.62 ± 0.52, range −0.08 to 1.28). Mean night-roosting social network centrality was 0.72 ± 0.07.
3.3. Repeatability and behavioural syndromes
All personality traits (i.e., activity hole-board:
Repeatability estimates and 95% credible intervals for four personality traits (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Positive correlation between activity measured in the hole-board (PC1H) and activity measured in the Y-maze (PC1Y) (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
We observed variation in the strength of correlations among personality traits, and between personality traits and roosting behaviours, consistent with the existence of one or more behavioural syndromes (Figure 3). After accounting for multiple comparisons, the overall correlation matrix was not significant (
Correlation matrix of four personality traits (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Visual depiction of correlations between selected personality traits and roosting behaviours for a captive colony of little brown bats. (A) Positive correlation between night-roosting eigenvector centrality and day-roosting group size (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
3.4. Behavioural assortment
We observed a positive correlation between activity matrices and social association (
4. Discussion
We found some moderate to strong correlations between clusters of traits measured in the hole-board and Y-maze tests, as well as social behaviours relevant to the roosting ecology of bats measured in a semi-natural enclosure. This finding suggests the existence of a behavioural syndrome, mostly associated with social traits, in little brown bats. We found evidence for repeatable personality traits in bats, with most traits showing moderate repeatability (i.e.,
Matrix regression between activity (PC1H), exploration (PC2H1 and PC3H2), and sociability (PC2Y) and social association (calculated using the simple ratio index, see Methods) for 34 little brown bats (Myotis lucifugus).
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Animal personality traits are thought to be maintained because of temporal and spatial variation in selection, but selection will only affect population evolution if phenotypic variation is heritable (Dingemanse et al., 2002; Nicolaus et al., 2016). Although heritability of personality traits in bats has never been studied, a recent meta-analysis of personality traits across vertebrates quantified a heritability to repeatability ratio of 0.52. This suggests that about 52% of population-level variation in personality (range 33–70%) can be attributed to additive genetic variation and that it is appropriate to infer evolutionary potential of personality based on repeatability (Dochtermann et al., 2015; but for an opposing example see Bierbach et al., 2017). If personality is heritable in little brown bats, ecological processes (e.g., predation or food availability) could impose directional selection against a given personality trait, or suite of traits. In Siberian chipmunks (Tamias sibiricus), boldness was selected for in years of low acorn availability, but selected against in years of high food availability, suggesting that bold individuals are more efficient foragers, which the authors suggest translates into higher reproductive success (Le Coeur et al., 2015). More work is needed to determine the heritability of personality in bats but our results, which build on those of Menzies et al. (2013), demonstrating repeatability suggest potential for heritability and evolution of personality in bats. By contrast, traits with low repeatability in our study, such as day roosting group size or night roosting behaviour, may be relatively flexible so that variation in roost selection could be more strongly influenced other factors, such as ambient temperature (Willis & Brigham, 2007; Webber & Willis, 2018) or roost availability (Willis et al., 2006; Bondo et al., 2019).
In addition to repeatability of individual personality traits, we found support for a sociality-based behavioural syndrome that linked together functionally similar behavioural correlations. Importantly, these correlations included behaviours measured in a standardized artificial setting (i.e., Y-maze test) and roosting and social behaviours measured in a semi-natural setting. We observed a positive correlation between day-roosting group size and night-roosting social centrality, suggesting that social behaviour during roosting, in general, is repeatable across contexts (i.e., night-roosting vs. day-roosting; Figure 4a). This suggests a socially-mediated behavioural syndrome in little brown bats where individuals with high scores on this axis day-roost in larger groups and are more socially embedded during night-roosting. Day-roosting group size in bats is influenced by roost temperature (Willis & Brigham, 2007; Webber & Willis, 2018), but the role of temperature on night-roosting has not been explicitly tested. If correlations between day-roosting group size and night-roosting social centrality are heritable and influence fitness, then selection could favour suites of traits that promote sociality at the individual and group levels.
We also found some evidence of correlations between activity-based and socially-based behaviours, specifically, a negative correlation between day-roosting group size and day-roost fidelity (
We also found some evidence of social assortment based on personality traits but, in contrast to our prediction, individuals with similar activity scores had stronger social associations. Based on the concept of social niche specialization (Montiglio et al., 2013) we expected that relatively active individuals would tend to roost most often with inactive conspecifics. Similar to our results, in guppies (Poecilia reticulata), individuals tended to associate with other individuals having similar patterns of predator vigilance behaviour, which has an activity component (Croft et al., 2009). This tendency could also help explain the pattern we observed. An alternative explanation could be that individual bats assorted based on patterns of activity because individuals with similar personality traits select roosts for the day at approximately the same time or have similar preferences for specific roosts or roost characteristics. This is consistent with the trend we observed suggesting that bats with high activity scores also showed high roost fidelity (Figure 3). Alternatively, it is possible that the overall colony size in our experiment was too small to reveal social niche specialization and studies with larger colonies of free-ranging bats may be needed.
One potential limitation of our study is that all behaviours measured were quantified in either completely artificial (i.e., hole-board and Y-maze) or semi-artificial environments (i.e., flight tent). Measuring behaviour in an artificial environment in which movement or activity patterns of individuals are constrained could be problematic. For instance, we likely only detected a small portion of the overall range in activity or movement patterns exhibited by little brown bats. Although studying behavioural syndromes in free-ranging animals is ideal, logistical constraints can prevent collection of detailed behavioural data. Thus, using captive populations in a semi-natural setting provides a good alternative for taxa in which free-ranging populations are cryptic and/or difficult to track and recapture. In the case of our study, although we concede that our flight enclosure was an artificial environment, roosting and social behaviours we observed were similar to those observed for free-ranging bats in the wild (Table 1). For example, bats switched roosts at nearly the same frequency (every 1.6 days, see results) as typical free-ranging bats (e.g., every 1.7 days; Willis & Brigham, 2004). In addition, even though maximum group size was constrained in our study, average group sizes were similar to those observed for free-ranging bats of the same and similar species (Willis & Brigham, 2007; Olson & Barclay, 2013). Finally, although we used different metrics, we found similar levels of within-night movement and activity as reported in past studies of free-ranging little brown bats (Anthony et al., 1981; Barclay, 1982). In addition, natural variation in resource distribution, such as foraging opportunities or roost availability, could influence some of the behaviours we measured. We there cautiously interpret our findings and encourage future studies to examine the role of animal personality on the same behaviours for free-ranging bats.
Another potential limitation of our study is the use of a single individual bat as the stimulus for all Y-maze tests. The primary reason we chose to use a single individual was logistical; it would have been difficult, if not impossible, for us to house many additional bats separate from the main colony to use as stimulus bats in the Y-maze test. There are, however, potential consequences of only using a single stimulus bat. The behaviour of the stimulus individual may have influenced focal bats in different ways. While there is no way to test for this without the use of different stimulus bats, we recommend future studies address this concern by conducting trials using different stimulus bats, perhaps using a non-endangered species for which more individuals can be captured and potentially housed. Although use of a single stimulus bat may have led to variation in responses of some focal individuals, in many ways use of a single individual is preferable because it should control for variation in traits of the stimulus bat that might be important to focal individuals. Thus, we argue that using a single stimulus bat was a reasonable way to quantify the sociability axis of personality for this study in the face of the logistic constraints of working with an endangered species.
Our flight tent system allowed us to obtain the first evidence of behavioural syndromes for any of the more than 1400 species of bats but also allowed us to collect data important from a conservation behaviour perspective (Angeloni et al., 2008; Blumstein, 2010), on a species that is federally endangered in Canada. White-nose syndrome has caused devastating impacts on little brown bats and other North American bat species in the past decade (Frick et al., 2017; Langwig et al., 2015). A few studies have quantified behaviour of bats with WNS in the laboratory (Brownlee-Bouboulis & Reeder, 2013; Wilcox et al., 2014; Bohn et al., 2016) and field (Langwig et al., 2012), and one consistent pattern is a reduction in clustering behaviour of infected bats during hibernation. The disease is also associated with reduced overall activity during arousals from torpor, and elevated expression of some pyrogens which, combined with self-isolation, typically reflect so-called ‘sickness behaviour’ (Adelman & Martin, 2009; Rapin et al., 2014; Bohn et al., 2016; Frick et al., 2016). Thus, some of the reduction in clustering by WNS-positive bats appears to reflect a behavioural change within individuals. However, if sociability and the tendency to cluster with other bats are heritable traits, as our repeatability estimates suggest, WNS could drive the rapid evolution of reduced sociability in little brown bats. Bats in WNS-positive hibernation sites may benefit from reduced clustering and solitary roosting because of reduced risk of multiple exposures and multiple points of infection with P. destructans on the skin (Frick et al., 2016) reduced disturbance from conspecifics (Turner et al., 2015) and/or the ability to reach lower body temperatures during torpor, which will save energy and slow fungal growth (Willis et al., 2005; Verant et al., 2012). These benefits suggest the potential for strong directional selection against sociability because of WNS which could lead to dramatic changes in social structure and habitat requirements of remnant populations of little brown bats. For example, if WNS selects for reduced sociability, and remnant bats prefer to roost in smaller colonies as a result, then significantly more suitable roosting habitat might be required to support a population size equivalent to pre-WNS numbers. We recommend future conservation behaviour studies aimed at understanding the potential for behavioural evolution of bats impacted by WNS in the face of an enormous selective bottleneck.
In summary, we observed moderate repeatability for all personality traits, identified a socially-mediated behavioural syndrome, and found some support for the hypothesis that individuals socially assort based on their personality traits. These results represent a useful stepping-stone to inform future studies testing how personality and behavioural syndromes might influence proximate ecological processes, such as risk of parasitism or predation and, ultimately, population dynamics. Given the importance of sociality and clustering for pathogen transmission in WNS, and for activity and energy balance on survival, our results also have implications for understanding how WNS might influence the evolution of behaviour in bat populations. Future work should aim to quantify long-term repeatability and heritability of personality and behavioural syndromes while also assessing how personality influences fitness of free-ranging bats.
Corresponding author, current address: Cognitive and Behavioural Ecology Interdisciplinary Program, Memorial University of Newfoundland, St. John’s, NL, Canada, e-mail address: webber.quinn@gmail.com
Acknowledgements
We are grateful to D. Baloun, H. Mayberry, K. Muise and F. Borrell for help with fieldwork. We also thank The Manitoba Forestry Association and Glenn Petersen for facilitating lodging in the field, and Mitch and Diane Wasylnuk for allowing us to capture bats on their private property. We acknowledge that our study took place on and near the traditional territories of Sagkeeng, Shoal Lake 40 and Buffalo Point First Nations of Treaties 1 and 3. We also thank Q. Fletcher, C. Garroway, S. Forbes, G. Avila-Sakar and two anonymous reviewers for outstanding suggestions on earlier versions of this manuscript. Funding was provided by a Discovery Grant to CKRW from the Natural Sciences and Engineering Research Council (NSERC, Canada), a Manitoba Graduate Scholarship and a NSERC Vanier Canada Graduate Scholarship to QMRW.
References
Adelman, J.S. & Martin, L.B. (2009). Vertebrate sickness behaviors: adaptive and integrated neuroendocrine immune responses. — Integr. Comp. Biol. 49: 202-214.
Angeloni, L., Schlaepfer, M.A., Lawler, J.J. & Crooks, K.R. (2008). A reassessment of the interface between conservation and behaviour. — Anim. Behav. 75: 731-737.
Anthony, E.L.P., Stack, M.H. & Kunz, T.H. (1981). Night roosting and the nocturnal time budget of the little brown bat, Myotis lucifugus: effects of reproductive status, prey density, and environmental conditions. — Oecologia 51: 151-156.
Aplin, L.M., Farine, D.R., Morand-Ferron, J. & Sheldon, B.C. (2012). Social networks predict patch discovery in a wild population of songbirds. — Proc. Roy. Soc. Lond. B: Biol. Sci. 279: 4199-4205.
Aplin, L.M., Farine, D.R., Morand-Ferron, J., Cole, E.F., Cockburn, A. & Sheldon, B.C. (2013). Individual personalities predict social behaviour in wild networks of great tits (Parus major). — Ecol. Lett. 16: 1365-1372.
Aplin, L.M., Firth, J.A., Farine, D.R., Voelkl, B., Crates, R.A., Culina, A., Garroway, C.J., Hinde, C.A., Kidd, L.R., Psorakis, I., Milligan, N.D., Radersma, R., Verhelst, B.L. & Sheldon, B.C. (2015). Consistent individual differences in the social phenotypes of wild great tits, Parus major. — Anim. Behav. 108: 117-127.
Barclay, R.M.R. (1982). Night Roosting behaviour of the little brown bat, Myotis lucifugus. — J. Mammal. 63: 464-747.
Bell, A.M., Hankison, S.J. & Laskowski, K.L. (2009). The repeatability of behaviour: a meta-analysis. — Anim. Behav. 77: 771-783.
Bierbach, D., Laskowski, K.L. & Wolf, M. (2017). Behavioural individuality in clonal fish arises despite near-identical rearing conditions. — Nature Commun. 8: 15361.
Biro, P.A. & Stamps, J.A. (2008). Are animal personality traits linked to life-history productivity?. — Trends Ecol. Evol. 23: 361-368.
Blumstein, D.T. (2010). Social behaviour in conservation. — Soc. Behav.: 520-534.
Bohn, S.J., Turner, J.M., Warnecke, L., Mayo, C., Mcguire, L.P., Misra, V., Bollinger, T.K. & Willis, C.K.R. (2016). Evidence of ‘sickness behaviour’ in bats with white-nose syndrome. — Behaviour 153: 981-1003.
Bondo, K.J., Willis, C.K.R., Metheny, J.D., Kilgour, R.J., Gillam, E.H., Kalcounis-Rueppell, M.C. & Brigham, R.M. (2019). Relocation of a maternity colony of big brown bats corresponds with high loss of roost trees. — J. Wildl. Manage. 83: 1753-1761.
Boon, A.K., Réale, D. & Boutin, S. (2007). The interaction between personality, offspring fitness and food abundance in North American red squirrels. — Ecol. Lett. 10: 1094-1104.
Brownlee-Bouboulis, S.A. & Reeder, D.M. (2013). White-nose syndrome-affected little brown Myotis (Myotis Lucifugus) increase grooming and other active behaviors during arousals from hibernation. — J. Wildl. Dis. 49: 850-859.
Budaev, S.V. (2010). Using principal components and factor analysis in animal behaviour research: caveats and guidelines. — Ethology 116: 472-480.
Cairns, S.J. & Schwager, S.J. (1987). A comparison of association indices. — Anim. Behav. 35: 1454-1469.
Canadian Wildlife Health Cooperative (2015). White-nose syndrome decontamination protocol. Available online at http://www.cwhc-rcsf.ca/docs/WNS_Decontamination_Protocol-Jun2015.pdf.
Chaverri, G. & Gillam, E.H. (2015). Repeatability in the contact calling system of Spix’s disc-winged bat (Thyroptera tricolor). — Roy. Soc. Open Sci. 2: 140197.
Chaverri, G. & Kunz, T.H. (2006). Roosting ecology of the tent-roosting bat Artibeus watsoni (Chiroptera: Phyllostomidae) in southwestern Costa Rica. — Biotropica 38: 77-84.
Chaverri, G., Quirós, O.E., Gamba-Rios, M. & Kunz, T.H. (2007). Ecological correlates of roost fidelity in the tent-making bat Artibeus watsoni. — Ethology 113: 598-605.
Cote, J., Dreiss, A. & Clobert, J. (2008). Social personality trait and fitness. — Proc. Roy. Soc. Lond. B: Biol. Sci. 275: 2851-2858.
Croft, D.P., Arrowsmith, B.J., Bielby, J., Skinner, K., White, E., Couzin, I.D., Magurran, A.E., Ramnarine, I. & Krause, J. (2003). Mechanisms underlying shoal composition in the Trinidadian guppy, Poecilia reticulata. — Oikos 100: 429-438.
Croft, D.P., James, R., Ward, A.J.W., Botham, M.S., Mawdsley, D. & Krause, J. (2005). Assortative interactions and social networks in fish. — Oecologia 143: 211-219.
Croft, D.P., Krause, J., Darden, S.K., Ramnarine, I.W., Faria, J.J. & James, R. (2009). Behavioural trait assortment in a social network: patterns and implications. — Behav. Ecol. Sociobiol. 63: 1495-1503.
Csárdi, G. & Nepusz, T. (2006). The igraph software package for complex network research. — InterJ. Complex Syst. 1695: 1-9.
Dekker, D., Krackhardt, D. & Snijders, T.A.B. (2007). Sensitivity of MRQAP tests to collinearity and autocorrelation conditions. — Psychometrika 72: 563-581.
Dingemanse, N.J. & Dochtermann, N.A. (2013). Quantifying individual variation in behaviour: mixed-effect modelling approaches. — J. Anim. Ecol. 82: 39-54.
Dingemanse, N.J., Both, C., Drent, P.J., van Oers, K. & van Noordwijk, A. (2002). Repeatability and heritability of exploratory behaviour in great tits from the wild. — Anim. Behav. 64: 929-938.
Dochtermann, N.A., Schwab, T. & Sih, A. (2015). The contribution of additive genetic variation to personality variation: heritability of personality. — Proc. Roy. Soc. Lond. B: Biol. Sci. 282: 20142201.
Farine, D.R. (2013). Animal social network inference and permutations for ecologists in R using asnipe. — Methods Ecol. Evol. 4: 1187-1194.
Farine, D.R. & Whitehead, H. (2015). Constructing, conducting and interpreting animal social network analysis. — J. Anim. Ecol. 84: 1144-1163.
Fenton, M.B., Rautenbach, I.L., Smith, S.E., Swanepoel, C.M., Grosell, J. & van Jaarsveld, J. (1994). Raptors and bats: threats and opportunities. — Anim. Behav. 48: 9-18.
Ferguson, J.N., Young, L.J., Hearn, E.F., Matzuk, M.M., Insel, T.R. & Winslow, J.T. (2000). Social amnesia in mice lacking the oxytocin gene. — Nature Genet. 25: 284-288.
Frick, W.F., Pollock, J.F., Hicks, A.C., Langwig, K.E., Reynolds, D.S., Turner, G.G., Butchkoski, C.M. & Kunz, T.H. (2010). An emerging disease causes regional population collapse of a common North American bat species. — Science 329: 679-682.
Frick, W.F., Pueschmaille, S.J. & Willis, C.K.R. (2016). White-nose syndrome in bats. — In: Bats in the anthropocene: conservation of bats in a changing world (Voigt, C.C. & Kingston, T., eds). Springer, New York, NY, p. 245-262.
Frick, W.F., Cheng, T.L., Langwig, K.E., Hoyt, J.R., Janicki, A.F., Parise, K.L., Foster, J.T. & Kilpatrick, A.M. (2017). Pathogen dynamics during invasion and establishment of white-nose syndrome explain mechanisms of host persistence. — Ecology 98: 624-631.
Garroway, C.J. & Broders, H.G. (2007). Nonrandom association patterns at northern long-eared bat maternity roosts. — Can. J. Zool. 85: 956-964.
Hadfield, J.D. (2010). MCMC methods for multi-respoinse generalized linear mixed models: the MCMCglmm R package. — J. Stat. Softw. 33: 1-22.
Hoyt, J.R., Langwig, K.E., White, J.P., Kaarakka, H.M., Redell, J.A., Kurta, A., Depue, J.E., Scullon, W.H., Parise, K.L., Foster, J.T., Frick, W.F. & Kilpatrick, A.M. (2018). Cryptic connections illuminate pathogen transmission within community networks. — Nature 563: 710-713.
Jacoby, D.M.P., Fear, L.N., Sims, D.W. & Croft, D.P. (2014). Shark personalities? Repeatability of social network traits in a widely distributed predatory fish. — Behav. Ecol. Sociobiol. 68: 1995-2003.
Kaiser, H.F. (1991). Coefficient alpha for a principal component and the Kaiser-Guttman rule. — Psychol. Rep. 68: 855-858.
Kerth, G. & Konig, B. (1999). Fission, fusion and nonrandom associations in female Bechstein’s bats (Myotis bechsteinii). — Behaviour 136: 1187-1202.
Kerth, G. & Reckardt, K. (2003). Information transfer about roosts in female Bechstein’s bats: an experimental field study. — Proc. Roy. Soc. Lond. B: Biol. Sci. 270: 511-515.
Kilgour, R.J., Faure, P.A. & Brigham, R.M. (2013). Evidence of social preferences in big brown bats (Eptesicus fuscus). — Can. J. Zool. 760: 756-760.
Korsten, P., van Overveld, T., Adriaensen, F. & Matthysen, E. (2013). Genetic integration of local dispersal and exploratory behaviour in a wild bird. — Nature Commun. 4: 2362.
Kuepper, N.D., Melber, M. & Kerth, G. (2016). Nightly clustering in communal roosts and the regular presence of adult females at night provide thermal benefits for juvenile Bechstein’s bats. — Mamm. Biol. 81: 201-204.
Kunz, T.H. (1974). Feeding ecology of a temperate insectivorous bat (Myotis velifer). — Ecology 55: 693-711.
Langwig, K.E., Frick, W.F., Bried, J.T., Hicks, A.C., Kunz, T.H. & Kilpatrick, A.M. (2012). Sociality, density-dependence and microclimates determine the persistence of populations suffering from a novel fungal disease, white-nose syndrome. — Ecol. Lett. 15: 1050-1057.
Langwig, K.E., Frick, W.F., Reynolds, R., Parise, K.L., Drees, K.P., Hoyt, J.R., Cheng, T.L., Kunz, T.H., Foster, J.T. & Kilpatrick, A.M. (2015). Host and pathogen ecology drive the seasonal dynamics of a fungal disease, white-nose syndrome. — Proc. Roy. Soc. Lond. B: Biol. Sci. 282: 20142335.
Le Coeur, C., Thibault, M., Pisanu, B., Thibault, S., Chapuis, J.-L. & Baudry, E. (2015). Temporally fluctuating selection on a personality trait in a wild rodent population. — Behav. Ecol. 26: 1285-1291.
Martin, J.G.A. & Réale, D. (2008). Temperament, risk assessment and habituation to novelty in eastern chipmunks, Tamias striatus. — Anim. Behav. 75: 309-318.
Maslo, B. & Fefferman, N.H. (2015). A case study of bats and white-nose syndrome demonstrating how to model population viability with evolutionary effects. — Conserv. Biol. 29: 1176-1185.
Menzies, A.K., Timonin, M.E., McGuire, L.P. & Willis, C.K.R. (2013). Personality variation in little brown bats. — PLoS ONE 8: e80230.
Metheny, J.D., Kalcounis-Rueppell, M.C., Willis, C.K.R., Kolar, A.A. & Brigham, R.M. (2008). Genetic relationships between roost-mates in a fission-fusion society of tree-roosting big brown bats (Eptesicus fuscus). — Behav. Ecol. Sociobiol. 62: 1043-1051.
Montiglio, P.-O., Ferrari, C. & Réale, D. (2013). Social niche specialization under constraints: personality, social interactions and environmental heterogeneity. — Philos. Trans. Roy. Soc. B: Biol. Sci. 368: 20120343.
Morton, F.B. & Altschul, D. (2019). Data reduction analyses of animal behaviour: avoiding Kaiser’s criterion and adopting more robust automated methods. — Anim. Behav. 149: 89-95.
Nicolaus, M., Tinbergen, J.M., Ubels, R., Both, C. & Dingemanse, N.J. (2016). Density fluctuations represent a key process maintaining personality variation in a wild passerine bird. — Ecol. Lett. 19: 478-486.
Olson, C.R. & Barclay, R.M.R. (2013). Concurrent changes in group size and roost use by reproductive female little brown bats (Myotis lucifugus). — Can. J. Zool. 155: 149-155.
Patriquin, K.J., Leonard, M.L., Broders, H.G. & Garroway, C.J. (2010). Do social networks of female northern long-eared bats vary with reproductive period and age?. — Behav. Ecol. Sociobiol. 64: 899-913.
Patriquin, K.J., Palstra, F., Leonard, M.L. & Broders, H.G. (2013). Female northern myotis (Myotis septentrionalis) that roost together are related. — Behav. Ecol. 24: 949-954.
Patriquin, K.J., Leonard, M.L., Broders, H.G., Ford, W.M., Britzke, E.R. & Silvis, A. (2016). Weather as a proximate explanation for fission–fusion dynamics in female northern long-eared bats. — Anim. Behav. 122: 47-57.
Pike, T.W., Madhumita, S., Lindström, J. & Royle, N.J. (2008). Behavioural phenotype affects social interactions in an animal network. — Proc. Roy. Soc. Lond. B: Biol. Sci. 275: 2515-2520.
R Core Team (2019). R: a language and environment for statistical computing. — R Foundation for Statistical Computing, Vienna.
Rapin, N., Johns, K., Martin, L., Warnecke, L., Turner, J.M., Bollinger, T.K., Willis, C.K.R., Voyles, J. & Misra, V. (2014). Activation of innate immune-response genes in little brown bats (Myotis lucifugus) infected with the fungus Pseudogymnoascus destructans. — PLoS ONE 9: e112285.
Réale, D., Reader, S.M., Sol, D., McDougall, P.T. & Dingemanse, N.J. (2007). Integrating animal temperament within ecology and evolution. — Biol. Rev. 82: 291-318.
Reckardt, K. & Kerth, G. (2006). The reproductive success of the parasitic bat fly Basilia nana (Diptera: Nycteribiidae) is affected by the low roost fidelity of its host, the Bechstein’s bat (Myotis bechsteinii). — Parasitol. Res. 98: 237-243.
Reckardt, K. & Kerth, G. (2007). Roost selection and roost switching of female Bechstein’s bats (Myotis bechsteinii) as a strategy of parasite avoidance. — Oecologia 154: 581-588.
Roberts, G. & Sherratt, T.N. (2002). Behaviorial evolution: does similarity breed cooperation? — Nature 418: 499-500.
Rydell, J., Entwistle, A. & Racey, P.A. (1996). Timing of foraging flights of three species of bats in relation to insect activity and predation risk. — Oikos 76: 243-252.
Shea, T.J.O. & Vaughan, T.A. (1977). Nocturnal and seasonal activities of the pallid bat, Antrozous pallidus. — J. Mammal. 58: 269-284.
Sih, A., Bell, A.M., Johnson, J.C. & Ziemba, A.R.E. (2004). Behavioural syndromes: an integrative overview. — Q. Rev. Biol. 51: 211-244.
Sih, A., Cote, J., Evans, M., Fogarty, S. & Pruitt, J. (2012). Ecological implications of behavioural syndromes. — Ecol. Lett. 15: 278-289.
Silvis, A., Kniowski, A.B., Gehrt, S.D. & Mark Ford, W. (2014). Roosting and foraging social structure of the endangered Indiana bat (Myotis sodalis). — PLoS ONE 9: e96937.
Smith, B.R. & Blumstein, D.T. (2008). Fitness consequences of personality: a meta-analysis. — Behav. Ecol. 19: 448-455.
Sueur, C., Petit, O., De Marco, A., Jacobs, A.T., Watanabe, K. & Thierry, B. (2011). A comparative network analysis of social style in macaques. — Anim. Behav. 82: 845-852.
Sutherland, W.J. (1998). The importance of behavioural studies in conservation biology. — Anim. Behav. 56: 801-809.
Turner, J.M., Warnecke, L., Wilcox, A., Baloun, D., Bollinger, T.K., Misra, V. & Willis, C.K.R. (2015). Conspecific disturbance contributes to altered hibernation patterns in bats with white-nose syndrome. — Physiol. Behav. 140: 71-78.
United States Fish & Wildlife Service (2016). National white-nose syndrome decontamination protocol. Available online at https://www.whitenosesyndrome.org/sites/default/files/resource/national_wns_decon_protocol_04.12.2016.pdf.
VanderWal, E., Festa-Bianchet, M., Réale, D., Coltman, D.W. & Pelletier, F. (2015). Sex-based differences in the adaptive value of social behavior contrasted against morphology and environment. — Ecology 96: 631-641.
Verant, M.L., Boyles, J.G., Waldrep, W., Wibbelt, G. & Blehert, D.S. (2012). Temperature-dependent growth of Geomyces destructans, the fungus that causes bat white-nose syndrome. — PLoS ONE 7: e46280.
von Merten, S., Zwolak, R. & Rychlik, L. (2017). Social personality: a more social shrew species exhibits stronger differences in personality types. — Anim. Behav. 127: 125-134.
Webber, Q.M.R., Mcguire, L.P., Smith, S.B. & Willis, C.K.R. (2015). Host behaviour, age and sex correlate with ectoparasite prevalence and intensity in a colonial mammal, the little brown bat. — Behaviour 152: 83-105.
Webber, Q.M.R., Brigham, R.M., Park, A.D., Gillam, E.H., O’Shea, T.J. & Willis, C.K.R. (2016). Social network characteristics and predicted pathogen transmission in summer colonies of female big brown bats (Eptesicus fuscus). — Behav. Ecol. Sociobiol. 70: 701-712.
Webber, Q.M.R. & Willis, C.K.R. (2018). An experimental test of effects of ambient temperature and roost quality on aggregation by little brown bats (Myotis lucifugus). — J. Therm. Biol. 74: 174-180.
Whitehead, H. (2008). Analyzing animal societies: quantitative methods for vertebrate social analysis. — University of Chicago Press, Chicago, IL.
Wilcox, A., Warnecke, L., Turner, J.M., McGuire, L.P., Jameson, J.W., Misra, V., Bollinger, T.K. & Willis, C.K.R. (2014). Behaviour of hibernating little brown bats experimentally inoculated with the pathogen that causes white-nose syndrome. — Anim. Behav. 88: 157-164.
Willis, C.K.R. (2015). Conservation physiology and conservation pathogens: white-nose syndrome and integrative biology for host-pathogen systems. — Integr. Comp. Biol. 55: 631-641.
Willis, C.K.R. & Brigham, R.M. (2004). Roost switching, roost sharing and social cohesion: forest-dwelling big brown bats, Eptesicus fuscus, conform to the fission-fusion model. — Anim. Behav. 68: 495-505.
Willis, C.K.R. & Brigham, R.M. (2007). Social thermoregulation exerts more influence than microclimate on forest roost preferences by a cavity-dwelling bat. — Behav. Ecol. Sociobiol. 62: 97-108.
Willis, C.K.R., Lane, J.E., Liknes, E.T., Swanson, D.L. & Brigham, R.M. (2005). Thermal energetics of female big brown bats (Eptesicus fuscus). — Can. J. Zool. 879: 871-879.
Willis, C.K.R., Voss, C.M. & Brigham, R.M. (2006). Roost selection by forest-living female big brown bats (Eptesicus fuscus). — J. Mammal. 87: 345-350.
Wilson, A.J., Réale, D., Clements, M.N., Morrissey, M.M., Postma, E., Walling, C.A., Kruuk, L.E.B. & Nussey, D.H. (2010). An ecologist’s guide to the animal model. — J. Anim. Ecol. 79: 13-26.
Summary of results for principal component analysis (PCA) of behavioural responses of little brown bats in the first trial for a modified hole-board test (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Summary of results for principal component analysis (PCA) of behavioural responses of little brown bats in the second trial for a modified hole-board test (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Summary of results for principal component analysis (PCA) of behavioural responses of little brown bats in the first trial for a modified Y-maze test (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Summary of results for principal component analysis (PCA) of behavioural responses of little brown bats in the second trial for a modified Y-maze test (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Summary table of repeatability for five personality traits (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Summary table of repeatability for four roosting behaviours (
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Summary timeline of bat captivity and behavioural trials. Note, orange brackets and text represent periods of time where different activities happened, while horizontal bars and light-blue text represent occasions where we released bats due to inability to feed independently (day 6) and at the end of the experiment (day 18).
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Screen capture of modified hole-board test from a video recording of a behavioural trial.
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Comparison of observed correlation coefficients for eigenvector centrality and eight behavioural traits compared to a distribution of correlation coefficients generated based on 1000 iterations of randomized centrality values correlated with each behaviour. Vertical red lines represent the observed correlation coefficient and vertical dashed lines represent the 95% quantiles from the distribution of randomly generated correlation coefficients. Note: only group size and roost fidelity were significant in the original correlation matrix and our randomization procedure corroborates this finding, indicating this correlation was not random.
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585
Comparison of observed correlation coefficients for group size and eight behavioural traits compared to a distribution of correlation coefficients generated based on 1000 iterations of randomized centrality values correlated with each behaviour. Vertical red lines represent the observed correlation coefficient and vertical dashed lines represent the 95% quantiles from the distribution of randomly generated correlation coefficients. Note: centrality and roost fidelity were significant correlations in the original correlation matrix and our randomization procedure corroborates this finding, indicating these correlations are not random.
Citation: Behaviour 157, 2 (2020) ; 10.1163/1568539X-00003585