Assessing trade-offs in avian behaviour using remotely collected data from a webcam.

Individual animals engage in many behaviours which are mutually exclusive, and so where individuals increase the duration of time spent on one type of behavioural activity, this must be offset by a corresponding decrease in at least one other type of behaviour. To understand the variation observed in animal behaviour, researchers need to know how individuals trade-off these mutually-exclusive behaviours within their time-activity budget. In this study, we used remotely collected behavioural observations made from a live-streaming webcam to investigate trade-offs in the behaviour of two bird species, the mute swan (Cygnus olor) and whooper swan (Cygnus cygnus). For both species, we tested for correlations in the duration of time spent on key mutually exclusive behaviours: aggression, foraging, maintenance, and resting. We detected a negative association between aggression and resting behaviours in both species, indicating that increased aggression is achieved at the expense of resting behaviour. In contrast, there was no apparent trade-off between aggression and foraging, aggression and maintenance, or maintenance and resting. Foraging and resting behaviours were negatively correlated in both species, highlighting a trade-off between these distinct modes of behaviour. A trade-off between foraging and maintenance behaviours was detected for the sedentary mute swans, but not the migratory whooper swans. Our findings show how birds can trade-off their time investments in mutually exclusive behaviours within their time-activity budgets. Moreover, our study demonstrates how remotely-collected data can be used to investigate fundamental questions in behavioural research.

Introduction

Many of the behaviours exhibited routinely by animals are mutually exclusive [1]. For example, an individual animal cannot forage and rest at the same time. Individuals therefore face key decisions about which behaviours to engage in at any given time. Such decision-making is expected to lead to trade-offs in the types of behaviour exhibited by individuals [1, 2]. Where individuals increase the duration of time spent on one type of behavioural activity, this must be offset by a decrease in one or more other types of behaviour [2-4]. As such, the behaviours that an individual exhibits represent a zero-sum game.

Behavioural researchers routinely use time-activity budgets to describe the relative amounts of time that animals spend engaged in different behaviours [5-7]. Time-activity budgets can therefore be a powerful tool for assessing differences in the behaviour of different animal species, as well as differences in behaviour within animal species [8]. The literature on time-activity budgets recorded for birds show that the time investment made on different behaviours can be highly variable, even for concurrent observations of different birds within the same population in a shared habitat [8, 9]. At least some of this variation within a population may reflect trade-offs in behavioural activities, as individuals prioritise certain behaviours over others, depending on their state and environmental conditions [10]. To improve our understanding of the variation observed in animal behaviour, researchers require knowledge of how animals may trade-off mutually-exclusive behaviours within their time-activity budget. In particular, information is needed on the degree to which the different behaviours displayed by a species are traded off against each other.

In this study, we used remote behavioural observations made via a live-streaming webcam to investigate the trade-offs in avian behaviour. Webcams have become a valuable tool in the study of animal behaviour, as they allow behavioural observations to be made without disturbance to focal individuals and without the need for researchers to be physically present at study sites [11-15]. We selected two common waterbirds as our focal species, the mute swan (Cygnus olor) and whooper swan (Cygnus cygnus). Their large body size, distinctive white plumage, and use of open-water habitats have allowed researchers to make detailed behavioural observations [9, 13, 16–20].

Where the two species co-exist, mute and whooper swans show similar, but not identical, patterns of behaviour [9]. Among swans, foraging, resting, and maintenance behaviours such as preening are the most commonly observed activities [21], although the amount of time spent on these behaviours shows marked variation between studies [9, 18–20]. Other key behaviours include aggression towards conspecifics and heterospecifics, although previous studies have found high levels of variation in how much time is spent on aggression between individual swans [9, 13] and between swan populations [31]. Swans cannot be engaged in aggressive interactions such as threat displays, pecking, or striking with their wings, at the same time as behaviours such as foraging, maintenance or resting [16–20, 22]. The prominence of foraging, resting, and maintenance behaviours within the time-activity budgets of swans [21] suggests that as additional behaviours such as aggression are undertaken, these more common behaviours might be involved in any trade-offs. However, the extent to which swans trade-off different behavioural activities within their time-activity budgets, has received little attention from researchers.

As aggression is a means of acquiring and maintaining access to food resources [23-25], we expected that individuals who were motivated to forage would engage in aggressive interactions with other individuals. However, because aggression and foraging are mutually exclusive types of behaviour, over short time durations we would not expect a positive association between these two behaviours either. Our first prediction was therefore that we would not observe a correlation between the time spent on aggressive and foraging behaviours. Instead, we expected birds to gain the time spent on aggression from other behaviours that can be deferred, such as maintenance and resting behaviours. Our second and third predictions were therefore that higher durations spent on aggression would be associated with lower durations spent on resting and maintenance behaviours, respectively. Finally, our fourth prediction was that we would observe trade-offs (i.e. a negative statistical association) between all non-aggressive behaviours (e.g. foraging, maintenance, and resting), as these activities represent distinct modes of behaviour [1].

Methods

Study system

The Wildfowl & Wetland Trust (WWT) Caerlaverock reserve (54°59ʹ2.4ʺ N, 3°30ʹ0ʺ W) in southwest Scotland is an important site for wintering waterbirds, including both mute and whooper swans [13, 16, 26]. The 587 ha site comprises a patchwork of aquatic and terrestrial habitats, including small lakes used by waterbirds for feeding and roosting. The non-migratory mute swans are resident at Caerlaverock throughout the year, whilst whooper swans are winter visitors present at Caerlaverock between October and March [16]. Both swan species are known to use the habitats within the reserve for all behaviours within their ethograms, including foraging, roosting and maintenance behaviours, as well as aggressive interactions with conspecifics and heterospecifics [9, 16]. Tracking studies of GPS-tagged individuals indicated that swans fed and roosted on the reserve, as many remained there for extended periods of time [27]. Nationally, the mute swan population has been stable since c.2000, whilst whooper swan numbers have increased steadily in recent decades [28, 29]. Within the Solway Estuary (which includes the WWT Caerlaverock reserve) the mean of the peak winter counts of individual swans between 2015/2016 and 2019/2020 was 76 mute swans (range 73–83) and 303 whooper (range 160–487) [26]. Previous research has found a balanced sex ratio among swans overwintering at our study site [16]. Whooper swan family groups arrive together and typically remain together during winter [30], although this behaviour is somewhat less common among mute swans, for which family groups may break up sooner [31].

Data collection

To quantify swan behaviour, we used a focal sampling approach, sensu [32], in which individual swans were observed for 15 minute periods. An observation period of 15 minutes was therefore selected as a duration for which it was practical to follow a single individual, and made our approach comparable with those of earlier studies of swan behaviour which used similar durations of observations [9, 17, 20, 33, 34]. During each observation, the number of seconds spent by the focal swan on each of four mutually exclusive behaviours was recorded: aggression, foraging, maintenance, and resting. Aggressive behaviours include threat displays, as well as striking at opponents with wings or beak, and are commonly observed among birds [35-37]. Foraging referred to all behaviours related to seeking, acquiring, and consuming food, maintenance behaviours included all preening, stretching and comfort activities, whilst resting represented periods of apparent inactivity such as sleeping [9]. Any time for which the swan was not visible during the observation period, for example, if it was obscured by other individuals, was not included in the calculations of the time spent on behaviours. The time spent on each behaviour was expressed as a proportion of the total time for which the swan was visible during the observation period. The species identity of the focal individual was recorded during each observation; the two species could be distinguished readily during observations due to interspecific differences in the dominant colouration of the bill (orange and yellow for mute and whooper swans, respectively) and their body posture (mute swans typically hold their neck in a characteristic s-shaped curve, whereas whooper swans typically hold their neck straight) [30, 31].

In order to select focal individuals at random, and thus avoid any observer bias, we superimposed a numbered 18 x 10 grid over the webcam image; each grid cell was assigned an unique number, and we used a random number generator to select a grid cell. Of the swans within that cell, the individual closest to the centre of the grid cell was selected as the focal individual for that 15 minute observation period. If the selected grid cell did not contain any swans, the process was repeated until an occupied cell was selected. This allowed us to randomise the selection of individuals, as well as the swan species that was selected. As there were markedly fewer juveniles at the study sites, our observations were made of adult birds. As the two swan populations were observed to use all grid cells, as well as all parts of the study lake, we see no reason why any individual would have been sampled more frequently than others.

A total of 119 observations were carried out between mid-November 2020 and mid-March 2021, inclusive (November n = 9, December n = 22, January n = 41, February n = 32, March n = 15). This sampling pattern was organised to obtain data across the months during which both species of swans were present at the site [13, 16, 34]. To obtain data from across the diurnal active period of the swans, observations were made only at 09:30, 12:00, 12:30, and 14:30. Overall, behavioural observations were made for a total of 119 swans (81 mute swans and 38 whooper swans). These observations comprising 1,448 minutes of observations (mute swans = 983 minutes, whooper swans = 465 minutes).

All behavioural observations were made remotely via a live-streaming webcam (AXIS Q6035-E PTZ Dome Network Camera), which was fixed in place facing directly outwards over the main roosting and feeding lake used by the swans. The angle and magnification were not adjusted during the study so that the field-of-view was standardised across all observation periods. Further information on the webcam set up is available in earlier published work [13].

Ethics

This observational study was carried out with the prior approval of the ethics committee of the College of Life and Environmental Sciences of the University of Exeter (eCLESPsy002195). Our observational study of non-human animals did not feature human participants and so participant consent was not required. As data collection was conducted virtually via a publically-accessible live-streaming webcam, no physical visits to the study site were undertaken, and hence no study site permits were required.

Statistical analyses

All statistical analyses were carried out in R version 3.6.3 [38]. Prior to assessing correlations between behaviours, we first tested whether either swan species showed marked differences in behaviour across the four times of day at which we made our observations. We therefore used Kruskal-Wallis tests, implemented using the kruskal.test function in R, to determine whether there were statistically significant differences in the time spent by each swan species on each behaviour (aggression, maintenance, foraging, resting) between each of the four time periods at which we made our behavioural observations. As we were interested in differences between any time period, we modelled the time periods as factors rather than as a continuous variable. We selected the non-parametric Kruskal-Wallis tests as the presence of zeros in our data set (e.g. where behaviours were not observed for some focal individuals) meant that the assumption of parametric tests such as Analysis of Variance (ANOVA) could not be met [39]. The P values were adjusted using Holm-Bonferroni corrections, implemented via the p.adjust function in R, to account for multiple comparisons [40].

We subsequently ran pairwise Kendall’s τ correlations [41] using the cor.test function in R to test the strength and direction of the association between the proportions of time spent on each behaviour: aggression vs foraging, aggression vs maintenance, aggression vs resting, foraging vs maintenance, foraging vs resting, and maintenance vs resting. Again, we selected the non-parametric Kendall’s τ correlations as the presence of zeros in our data set meant that the assumption of parametric correlation analyses such as Pearson’s r could not be met [39]. Kendall’s τ values range from -1.0 (a perfect negative association), to +1.0 (a perfect positive association), with a value of 0.0 indicating no association between two variables [41]. As we knew a priori that the two swan species exhibit differences in behaviours [9, 13], we ran the correlation analyses separately for each species. Furthermore, as previous research on the swans at WWT Caerlaverock had shown that behaviour did not vary markedly with time of day [13], data from all observation periods were combined for the analyses; the validity of this assumption was tested using the Kruskal-Wallis tests previously described. Statistically significant positive or negative associations between the proportions of time spent on behaviours were attributed where P < 0.05, after P values had been adjusted using Holm-Bonferroni corrections, implemented via the p.adjust function in R, to account for multiple comparisons [40].

Results

Mute swans spent the highest proportion of their time engaged in foraging behaviour (mean ± 95% CI = 0.421 ± 0.043), and the lowest proportion of their observed time on aggressive interactions (0.141 ± 0.027; Fig 1). Mute swans spent intermediate proportions of their time-activity budget on resting (0.184 ± 0.036) maintenance behaviours (0.254 ± 0.030; Fig 1).

Fig 1

Time-activity budgets.

The mean (±95% CI) proportion of observed time that mute swans (black bars; n = 81) and whooper swans (grey bars; n = 38) spent on each behaviour.

In contrast, whooper swans spent more time resting (0.334 ± 0.067) than on any other behaviour (Fig 1). Similar to mute swans, the whooper swans spent the lowest proportion of their observed time on aggressive interactions (0.126 ± 0.047; Fig 1). Foraging and resting accounted for similar proportions of their time-activity budget, 0.323 ± 0.060 and 0.334 ± 0.063, respectively (Fig 1).

Our analyses detected no statistically significant differences in the time spent on each behaviour by mute swans between the four time periods at which we made our observations: aggression (χ = 1.81, d.f. = 3, P = 1.000; Fig 2a); maintenance (χ = 5.14, d.f. = 3, P = 0.973; Fig 2b); foraging (χ = 4.12, d.f. = 3, P = 0.996; Fig 2c); resting (χ = 2.48, d.f. = 3, P = 1.000; Fig 2d).

Fig 2

Observed mute swan behaviour at different times of day.

The bars represent the mean ± 95% CI time spent on each behaviour, whilst the filled circles represent the individual data points.

We found no statistically significant differences in the time spent on each behaviour by whooper swans between the four time periods at which we made our observations: aggression (χ = 2.33, d.f. = 3, P = 1.000; Fig 3a); maintenance (χ = 7.63, d.f. = 3, P = 0.435; Fig 3b); foraging (χ = 7.28, d.f. = 3, P = 0.444; Fig 3c); resting (χ = 4.90, d.f. = 3, P = 0.973; Fig 3d).

Fig 3

Observed whooper swan behaviour at different times of day.

The bars represent the mean ± 95% CI time spent on each behaviour, whilst the filled circles represent the individual data points.

Among mute swans, the proportion of time spent on aggression was negatively correlated with the proportion of time spent resting (τ = -0.225, P = 0.036; Fig 4). Similarly, we also detected negative correlations between the proportion of time engaged in foraging and both maintenance (τ = -0.359, P < 0.001; Fig 4) and resting behaviours (τ = -0.385, P < 0.001; Fig 4).

Fig 4

Mute swan behavioural trade-offs.

Kendall’s tau correlations between each of the recorded mute swan behaviours.

No statistically significant correlations were detected, however, for mute swans between the proportion of time spent on aggression and foraging (τ = -0.185, P = 0.122; Fig 4), aggression and maintenance (τ = -0.029, P = 1.000; Fig 4), or maintenance and resting (τ = -0.016, P = 1.000; Fig 4).

For our observations of whooper swans, we found that the proportion of time spent on aggressive interactions was negatively correlated with the proportion of time spent resting (τ = -0.337, P = 0.036; Fig 5), whilst the proportion of time spent on foraging behaviour was negatively correlated with the proportion of time spent on resting behaviour (τ = -0.415, P = 0.003; Fig 5).

Fig 5

Whooper swan behavioural trade-offs.

Kendall’s tau correlations between each of the recorded whooper swan behaviours.

We detected no statistically significant correlations between the proportions of time that whooper swans spent engaged in aggression and foraging (τ = -0.149, P = 0.820; Fig 5), aggression and maintenance (τ = 0.078, P = 1.000; Fig 5), foraging and maintenance (τ = -0.149, P = 0.685; Fig 5), or between maintenance and resting (τ = -0.190, P = 0.593; Fig 5).

Discussion

In this study, we used remotely collected behavioural observations made from a live-streaming webcam to investigate trade-offs in the behaviour of two bird species, the mute swan and whooper swan. These remotely-collected data allowed us to test four predictions regarding swan behaviour, as set out in our introduction: (i) no correlation between the time spent on aggressive and foraging behaviours; (ii) a correlation between aggression and resting; (iii) a correlation between aggression and maintenance behaviours; (iv) negative correlations between all non-aggressive behaviours. The findings that we report here illustrate how birds can trade-off their time investments in mutually exclusive behaviours within their time-activity budgets.

The time spent on foraging by mute swans at our study site (mean ± 95% CI = 42.1 ± 4.3%) was similar to the 43.3 ± 2.6% reported by another recent study of mute swan behaviour in eastern England [9]. Indeed, our observed value was also similar to other studies of foraging effort during winter, which reported 48% at a rural site in Poland [42], 41% in Denmark [21], and 36% in Ireland [33]. In contrast, an earlier study of mute swan behaviour at another wintering site in Poland found that foraging accounted for c.20% of time [43], whilst mute swans wintering in Scotland spent 58% of time foraging [44]. Our finding that our focal whooper swans spent 32.3 ± 6.0% of their time-activity budget on foraging was similar to the 35% reported from a study in China [45] and the c.30% reported by an earlier study conducted at our study site [16]. Research from Ireland and eastern England reported slightly higher foraging efforts of 40% [46] and 43% [9], respectively, whilst observations of whooper swans wintering in Turkey reported only 12% of time spent on foraging [22]. Some caution is needed in comparisons of foraging effort between different studies, as the time spent on foraging depends on multiple factors including daily energy requirement, food biomass, food energetic content, as well as the energetic cost of foraging behaviour [47-50]. However, because our observed time investments in foraging were well within the range reported previously for both species by studies conducted during winter, our data suggest that swans not only roosting on our study lake and feeding elsewhere, as is sometimes observed for overwintering swan populations [51]. Similarly, an earlier tracking study of GPS-tagged whooper swans at our study site indicated that individuals fed and roosted on the reserve, as many remained there for extended periods of time [27]. Although the study observed some localised (<10 km) movements from the reserve to adjacent areas, it did not find the systematic daily movements that would be expected if swans were only roosting on the reserve and were feeding elsewhere [27].

Maintenance behaviours such as preening were important activities for both swan species, accounting for 25% and 22% of the time activity budgets of our focal mute swans and whooper swans, respectively. These values are intermediate in comparison to previously published observations of wintering swans, which have ranged from 5% for mute swans wintering in Scotland [44] up to 44% for whooper swans wintering in Ireland [46]. Resting behaviour was also a major component of the time-activity budget of both focal species, accounting for 18% and 33% of time among mute swans and whooper swans, respectively. Again, these values appear intermediate when compared with previously published observations of wintering swans. As examples, whooper swans wintering in China spent 14% of time resting [45], whilst overwintering mute and whooper swans in eastern England spent 7% and 18% on resting, respectively [9]. In contrast, mute swans wintering in Denmark and Poland rested for 35% [21] and 46% [43] of their total time-activity budgets, respectively, whilst whooper swans wintering in Turkey spent 55% of their time resting [22].

Our observations that mute swans and whooper swans spent 14% and 13% of their time on aggressive encounters with other birds were higher than in many previous studies. As examples, whooper swans wintering in China spent 0.3% of time on aggression [45], mute swans in the USA spent 2% of time on aggression [20], whilst overwintering mute and whooper swans in eastern England spent 2% and 1%, respectively, of their time on aggression [9]. Our observed values were similar to the 16% recorded for mute swans during the breeding season in Poland [52]. However, studies of ecologically-similar large-bodied herbivorous avian species have recorded higher levels of aggression; for example, among barnacle geese (Branta leucopsis) up to 35% of time is spent engaged in aggressive interactions [53]. Indeed, a recent meta-analysis of aggression among waterbird species found that the time spent on aggression ranged from 0–35% [36], and so our observed values were within the range reported previously for waterbirds. Among swans, the time spent engaged in aggressive interactions has been shown to rise with increasing swan densities [9, 13]; therefore future work at our study site could assess whether the relatively high levels of aggression that were observed are due to higher densities of swans using the study lake.

Our observations of swan behaviour provided support for our first prediction, that there would be no trade-off between aggression and foraging. Aggressive behaviours represent a means of acquiring and maintaining access to food resources at the expense of competitors [24, 54]. Clustered food resources and the benefits of collective anti-predator behaviours mean that bird species often aggregate into flocks at feeding sites [55]. However, in such situations where multiple individuals seek to exploit shared food resources, those individuals that can dominate access to those resources will gain a competitive advantage. Previous research [56] has showed that as swan densities increased, competitively dominant individuals incurred smaller penalties to food intake rates compared with subdominant individuals. Aggressive behaviours are therefore frequently observed among birds such as swans at their feeding sites [36]. Indeed, among swans most aggression is directed towards conspecifics rather than heterospecifics, as the former represent the greatest competition for limiting resources such as food [9, 13, 20]. Individual swans which are actively foraging would therefore be expected to be the most likely to engage in aggression, whilst those not foraging would move away from feeding areas to minimise aggression [16].

Our findings also offered support for our second prediction, that there would be a trade-off between aggression and resting behaviours. In contrast, neither swan species showed any association between aggression and maintenance behaviours, thereby providing no support for our third prediction, that there would be a trade-off between these behaviours. In our two focal species, increased aggression therefore appears to come at the expense of resting behaviour. Resting can be considered to represent a pool of time that can, at least to an extent, be redistributed to other activities as the individual requires. It is unclear how much time devoted to resting could be redistributed to activities such as aggression before there could be negative consequences on individuals; thus it is unclear what the ultimate limits of this behavioural trade-off might be. Aside from reducing the risk of predation [57, 58], periods of rest are known to be important in birds for health and cognition [59, 60]. To our knowledge there has been no research to date which has assessed the daily amount of rest that swans require to meet these needs. At our study site, mute and whooper swans spent on average 18.4% and 33.4% of their time engaged in resting behaviours. Given that previous studies have reported that swans can spend as little as 0.5% of their time resting [17], the redistribution of time from rest to other activities such as aggression is unlikely to be having marked impacts on our focal swans. Previous research has shown that male swans typically spend more time engaged in aggressive interactions compared with females [36]; however, sex cannot be readily determined visually for individual swans within a flock, and so the sex of our focal individuals could not be accounted for in our study. Previous research has found a balanced sex ratio among swans overwintering at our study site [16], and so any such sex effects were unlikely to lead to marked biases in our population-level data. Future research based on individuals of known sex could expand upon our findings by testing for trade-offs within each sex separately.

We found mixed support for our fourth prediction, that there would be a trade-off between all non-aggressive behaviours. As mutually exclusive activities, foraging and resting represent distinct modes of behaviour. Foraging is an activity that allows individuals to acquire the energy and nutrients needed to survive and reproduce, but exposes them to mortality risks such as predation, whereas resting allows individuals to minimise energy expenditure and predation risk [57]. Individual birds regulate their foraging effort to trade-off starvation and predation risk [58]. For animals such as swans that may fly between feeding sites, the mortality risk associated with foraging extends beyond predation to include flying accidents such as collisions with natural or man-made objects [61, 62]. Given the need for animals to achieve optimal trade-offs between foraging and resting, it is perhaps unsurprising that the strongest negative associations for both swan species were detected between foraging and resting.

Whilst we found no association between the durations of time spent on maintenance and resting, a trade-off between foraging and maintenance was detected only among mute swans. The biological significance of this disparity is unclear, especially as it is difficult to infer mechanistic explanations for behaviour from short observations. We quantified the behaviours of our focal swans over relatively short durations of 15 minutes, in common with most other studies of swan activities [9, 17, 20, 33, 34]. Whilst longer observation periods would have increased the information gained from each individual, they would also have increased the chance that the individual would have moved out of view during the observation. As a consequence, we were only able to assess for potential trade-offs over these short time periods. An unresolved question is therefore whether such trade-offs may persist over longer durations, or whether individuals alter their behaviour to undertake behaviour that was previously avoided. For example, where swans redistribute potential resting time to engage instead in aggressive interactions, do they later undertake additional resting behaviour to account for the earlier trade-off? Or is that potential resting time lost permanently? Such questions can likely only be addressed by quantifying swan behaviour over much longer time periods. The current widely used visual-based methods of quantifying swan time-activity budgets are likely to be unsuitable for longer term studies, as swans frequently move out-of-view. Alternative methods, such as the construction of time-activity budgets from data from individuals fitted with accelerometers and other devices that allow behavioural activities to be inferred [50], may be needed to address these further questions.

Our work demonstrates how remotely-collected data can be used to investigate fundamental questions in behavioural research. Such remote data collection provides a number of advantages to behavioural scientists, including reduced impacts of disturbance on focal animals, reduced carbon footprint associated with repeated visits to observation sites, and greater accessibility for scientists who cannot physically travel to study sites [11-15]. Remote methods can also offer a means to collect data during the Covid-19 pandemic, which has curtailed the ability of researchers to visit field sites to undertake traditional methods of in-person data collection [63]. Given these advantages, we expect that remote methods of data collection will become an increasingly valued tool for behavioural research.

Observational studies have confirmed that animals are flexible in the amount of time that they spend on different behaviours, and as a consequence populations can exhibit high variation in the amount of time spent on different activities [64]. Such variation may reflect trade-offs in behavioural activities, whereby animals prioritise certain behaviours, depending on their current state and environmental conditions [10]. Our observations of two overwintering swan species are consistent with the idea that birds can trade-off their time investments in mutually exclusive behaviours within their time-activity budgets, at least over short time periods. For both of our study species, negative associations between foraging and resting, and between resting and aggression, suggest that swans can trade-off time investment in these behaviours. However, we also recognise that it is difficult to draw broader conclusions about the implications of such patterns of behaviour, including fitness impacts, from short-term observations conducted over periods of 15 minutes. Future research that undertook behavioural observations of known individuals over longer time periods, would further improve our understanding of the capacity of birds to trade-off behavioural activities within their time-activity budgets. The patterns of behaviour that we have documented here can inform the development of hypotheses regarding behavioural trade-offs in such studies.

30 May 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Claudia Mettke-Hofmann, PhD

Section Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. In your Methods section, please provide additional information regarding the permits you obtained for the work. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why."

KLC (Straive) 15 Feb 22: Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

4. Please note that in order to use the direct billing option the corresponding author must be affiliated with the chosen institute. Please either amend your manuscript to change the affiliation or corresponding author, or email us at plosone@plos.org with a request to remove this option.

5. Please amend your list of authors on the manuscript to ensure that each author is linked to an affiliation. Authors’ affiliations should reflect the institution where the work was done (if authors moved subsequently, you can also list the new affiliation stating “current affiliation:….” as necessary).

Additional Editor Comments:

Comments by the Academic Editor:

Abstract:

Lines 32-34: Aggression and foraging are not negatively correlated. But are they positively correlated justifying the interpretation that aggression serves to secure access to food?

Lines 36-40: There is no trade-off between maintenance behaviour and resting for migratory whooper swan as it may be important to maintain good feather quality. Were there other trade-offs e.g. maintenance vs aggression? Is there more maintenance behaviour in whooper swans than mute swans justifying the interpretation?

Introduction:

Lines 54pp: This is interesting but deviates from your study as this links to individual differences, whereas your study is about differences between species. Please provide background information about differences in time budgets between species, ideally closely related species and also in relation to migration and residency to lead to your aim and selection of behaviours studied. The background should also include information about which behaviours are often traded-off against each other (e.g. vigilance-foraging, aggression-foraging, resting-foraging etc). Again, this then justifies the selection of behaviours and predictions for your study. You should also provide information about any known differences in the selected behaviours between the two focal species.

Methods:

Lines 128pp: Swans were randomly selected using a grid and random number generator. Were all grids selected at a comparable rate or was there e.g. a clumping in the middle? I ask as individuals may be quite stationary and any random clumping could lead to sampling a particular individual more often than others. Also, were only adult swans sampled or also juveniles?

Did you observe both swan species on the same day and time slot or how did you distribute observations?

Results:

Please provide all results even when they are not significant.

Discussion:

Lines 212pp: It would be good to repeat the predictions to remind the reader.

Lines 246pp: Potential sex effects. I am sure you know the sex ratio on the reserve for both species. I assume that usually entire families overwinter together, but it would be nice to have this mentioned somewhere.

Lines 278pp: You may also discuss that the birds may perform some particular behaviours on the lake you monitored (e.g. resting), whereas they may move somewhere else to perform other behaviours (e.g. foraging in the estuary). To which extend they move away for foraging may also differ between the two species. This should be discussed.

Lines 306pp: You mention individual differences here and that individuals trade-off behaviours. However, as far as I understand you were not able to identify individuals and your results reflect a population average rather than differences between individuals.

Fig. 1: Summing up the different behaviours seems to result in a value over 1 for both species. Wouldn’t you expect that the proportion of behaviours within each species adds up to 1 or below when you consider that there may be other behaviours and out-of-site? It definitely should not be above 1.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript PONE-D-22-04625 is fairly well written but would be stronger if the authors (1) included more background information on the use of the reserve wetlands of the two focal species (2) conducted focal observations during a broader range across daylight hours of the day (3) kept conclusions regarding trade-offs in behaviour to the scope of the data as 15 minute intervals are ok to show that differences exist, but the implications of those differences likely have little biological relevance at that timescale for an individual. The authors did do a nice job of pointing out some of the benefits of the web cam approach for observations but missed an opportunity to take advantage of a constant ability to collect data across daily time periods via this type of data stream. A major revision to the paper as written to address these primary topics would be warranted prior to publication. Please see review file for specific notes on the manuscript.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Review of PONE-D-22-04625.docx

Click here for additional data file.

16 Jun 2022

Dear Professor Chenette,

My co-authors and I have submitted a revision to our manuscript, “Assessing trade-offs in avian behaviour using remotely collected data from a webcam” (PONE-D-22-04625), as requested. We would like to thank the section editor Dr Claudia Mettke-Hofmann, as well as the anonymous reviewer, for their positive responses and constructive comments on the earlier version of our manuscript. We have revised our manuscript to address each of these comments and feel that our study has been greatly improved. In this letter of response, we present our detailed responses to each of the comments and discuss the changes we have made. Furthermore, we have now uploaded our data, as well as the code used in our analysis, to a figshare repository which is publically accessible. We would be grateful if the Data Availability statement could be updated to reflect this. Both the data files and the analytical code used in our study can be accessed via the following DOI: https://doi.org/10.6084/m9.figshare.20063225

Should you have any queries please do not hesitate to contact me.

Yours faithfully,

Dr Paul Rose

For clarity, the comments from the reviewers are presented in bold and our responses are in italics. We have also indicated in our responses the line number(s) in the revised version of the manuscript where the relevant changes have been made.

Section Editor’s comments:

1. The manuscript is well written and makes interesting use of webcam data. Both, the reviewer and I would like to see the following comments addressed before the manuscript can be considered for publication in PLOS ONE. A) Provide more background information about the species and similar studies that lead to the aims and predictions. Predictions may need to be reformulated.

– Our response: We are grateful for this positive and constructive feedback; revising our manuscript to address these comments has helped us to improve our work. In this letter we have responded to each of the comments in turn, detailing the changes that we have made. On the specific point raised in this comment, we have amended our introduction to provide greater background information on the behaviour of our two focal species (lines 77-92). This information is supported by citations of relevant previous studies of the behaviour of our two focal swan species (e.g. Holm 2002; Rees et al. 2005; Tatu et al. 2007; Włodarczyk 2017; Nergiz 2019; all of which are now cited in our revised manuscript). This new information has been integrated with our study predictions, so that the predictions presented in our revised manuscript are justified by the information that we provide (lines 77-106). We also more clearly draw the reader’s attention to key knowledge gaps that we intended our study to address (e.g. lines 90-92). Our revised manuscript therefore now features a much stronger link between our study predictions and the background information on the behaviour of both swan species.

2. B) Consider including more data covering the entire day to control for any species-specific use of the reserve across the day. The way of data collection (webcam) would support this wider approach.

– Our response: Please see our response to Reviewer 1 Comment 7. In our revised manuscript we now present new analyses of each behaviour across all of the different times of day at which observations were conducted. With these data and additional analyses, we now show that there are no temporal patterns in the behaviour of either swan species that could have confounded our original analysis (in particular, the pooling of data across time-periods).

3. C) Provide a more detailed discussion better fitting to the scope of the data.

– Our response: We have amended the discussion section in several ways to improve the text. New text has been added to discuss the behavioural findings for both species (lines 300-357). We have now included new text which provides better ecological context for our findings by comparing our findings with comparable previous studies of swan behaviour (lines 300-357). We now also provide the reader with information from previous studies on known factors which influence the amount of time spent on key behaviours, such as foraging (lines 312-316) and aggression (lines 353-357). Furthermore, we have added text to discuss evidence from tracking research that swans use the reserve for extended periods of time, and hence do not systematically move out of the reserve to feed (lines 320-325). Please also see our response to Reviewer 1 Comment 11 for details of additional revisions that we have made to our discussion.

4. Lines 32-34: Aggression and foraging are not negatively correlated. But are they positively correlated justifying the interpretation that aggression serves to secure access to food?

– Our response: On reflection we accept that the abstract focused too much on interpretations and not enough on reporting our findings. We have therefore revised the abstract to report all of the pairwise correlations between behaviours, even those that were non-significant (lines 32-39). Where correlations were detected, we have now stated whether these were positive or negative. We have also removed the interpretative text regarding access to food from the abstract, as we accept that these were not well-supported.

5. Lines 36-40: There is no trade-off between maintenance behaviour and resting for migratory whooper swan as it may be important to maintain good feather quality. Were there other trade-offs e.g. maintenance vs aggression? Is there more maintenance behaviour in whooper swans than mute swans justifying the interpretation?

– Our response: On reflection we accept that the abstract focused too much on interpretations and not enough on reporting our findings. We have therefore revised the abstract to report all of the pairwise correlations between behaviours, even those that were non-significant (lines 32-39). We have also removed the interpretative text from the abstract, as we accept that these were not well-supported.

6. Lines 54pp: This is interesting but deviates from your study as this links to individual differences, whereas your study is about differences between species. Please provide background information about differences in time budgets between species, ideally closely related species and also in relation to migration and residency to lead to your aim and selection of behaviours studied. The background should also include information about which behaviours are often traded-off against each other (e.g. vigilance-foraging, aggression-foraging, resting-foraging etc). Again, this then justifies the selection of behaviours and predictions for your study. You should also provide information about any known differences in the selected behaviours between the two focal species.

– Our response: As suggested, we have amended the text here to include mention of differences between species and between populations, rather than focusing on individuals. We now state that time-activity budgets can be a powerful tool for assessing differences in the behaviour of different animal species (lines 55-58). Our revised introduction includes much more background information on the types of behaviour shown by our two focal species (lines 77-92). We now mention that, where our two focal species co-exist, they show similar, but not identical, patterns of behaviour (lines 77-78); we provide additional ecological context for these two species (including their migratory/residency status) in the section on our study system (lines 114-122). We also present information on which behaviours may be traded off against others (lines 87-92), although we highlight that there has been little research on this topic, and so our study is a timely attempt to address this knowledge gap.

7. Lines 128pp: Swans were randomly selected using a grid and random number generator. Were all grids selected at a comparable rate or was there e.g. a clumping in the middle? I ask as individuals may be quite stationary and any random clumping could lead to sampling a particular individual more often than others. Also, were only adult swans sampled or also juveniles?

– Our response: We have added text to the methods section to confirm that, because swans were observed to use all grid cells (as well as all parts of the study lake), we see no reason why any individual grid cell would have been sampled more frequently than others (lines 163-166). Furthermore, we have now added text to explain that, because there were markedly fewer juveniles at the study sites, our observations were made of adult birds (lines 162-163). Our revised manuscript therefore now includes these important pieces of information to help the reader better understand our sampling approach.

8. Did you observe both swan species on the same day and time slot or how did you distribute observations?

– Our response: We have amended the text in our methods section to clarify that by randomly selecting the individual swan that was closest to the centre of randomly-generated grid co-ordinates, we effectively randomised the selection of swan species as well as the selection of an individual, because the individual that was closest to the centre of that grid could have been either a mute swan or a whooper swan (lines 161-162). This approach allowed us to gain behavioural data on both swan species without clumping of data on one species on certain days or time-periods.

9. Please provide all results even when they are not significant.

– Our response: As suggested, we have now included all behavioural results within our results section, including all non-significant findings, for both focal species (lines 268-271 and 281-285). Furthermore, we have also added text to report the non-significant correlations in the abstract (lines 32-39).

10. Lines 212pp: It would be good to repeat the predictions to remind the reader.

– Our response: We have followed the recommendation and have repeated the four predictions at the start of the introduction (lines 293-298). We also point the reader to the full rationale for these predictions, which can be found in the introduction.

11. Lines 246pp: Potential sex effects. I am sure you know the sex ratio on the reserve for both species. I assume that usually entire families overwinter together, but it would be nice to have this mentioned somewhere.

– Our response: We agree that it is useful to clarify these points. We have therefore amended our text to mention, with an appropriate supporting reference, that previous research has found a balanced sex ratio among swans overwintering at our study site (lines 126-127, 396-397). We also discuss how future research could investigate possible sex effects on behaviour, building upon our study (lines 399-400). Furthermore, we have also added text to confirm that whooper swan family groups typically remain together during winter, although this is somewhat less common among mute swans, for which family groups may break up sooner (lines 127-130).

12. Lines 278pp: You may also discuss that the birds may perform some particular behaviours on the lake you monitored (e.g. resting), whereas they may move somewhere else to perform other behaviours (e.g. foraging in the estuary). To which extend they move away for foraging may also differ between the two species. This should be discussed.

– Our response: We agree that this is an important point that deserved greater attention in our manuscript. Therefore, we have added additional text to the discussion and methods to address this point. In our revised discussion we now compare the percentage of time spent on each behaviour with those reported in previous studies (lines 300-357). If our focal individuals had, for example, only been roosting on our study lake and had been undertaking daily foraging flights to feed elsewhere, then our data would show much higher resting and much lower foraging than in comparable studies. However, for all behaviours the time spent by our focal swans is well within the range reported in previous studies, and indicates that our focal swans were exhibiting their range of behaviours. For example, we now report that the 42% of time spent on foraging by mute swans at our study site was similar to the 43% reported for mute swans in eastern England [Wood et al. 2021], 48% at a rural site in Poland [Józkowicz & Gorska-Klek 1996], 41% in Denmark [Holm 2002], and 36% in Ireland [Keane & O'Halloran 1992]. In addition, it is also clear from previous research (Griffin et al. 2010; now cited in our study) that tracked individual swans at the study site that individuals fed and roosted on the reserve, as many remained there for extended periods of time. Although they observed some localised (<10 km) movements from the reserve to adjacent areas, they did not observe the systematic daily movements that would be expected if swans were only roosting on the reserve and were feeding elsewhere ; we now mention this study in our discussion (lines 320-325), as well as in our methods section in the text on the study system (lines 119-121). Overall, the evidence from our own work, and previous studies, suggests it is unlikely that the swans were performing only certain behaviours on our study lake and other behaviours unobserved away from our web cam; as described above, our revised manuscript now discusses this evidence.

13. Lines 306pp: You mention individual differences here and that individuals trade-off behaviours. However, as far as I understand you were not able to identify individuals and your results reflect a population average rather than differences between individuals.

– Our response: We accept the reviewer’s point and so we have removed the focus on individual animals from this paragraph, instead referring to populations (lines 446-463).

14. Fig. 1: Summing up the different behaviours seems to result in a value over 1 for both species. Wouldn’t you expect that the proportion of behaviours within each species adds up to 1 or below when you consider that there may be other behaviours and out-of-site? It definitely should not be above 1.

– Our response: We have checked our data thoroughly and we confirm that the time values displayed in Fig 1 for Mute Swans (aggression = 0.141, foraging = 0.421, maintenance = 0.254, resting = 0.184) and Whooper Swans (aggression = 0.126, foraging = 0.323, maintenance = 0.217, resting = 0.334) each sum to exactly 1.0, as would be expected. To ensure that readers can accurately evaluate the data presented in Fig 1, we have amended Fig 1 to increase the number of y-axis labels on the graph; the value of the y-axis is now indicated every 0.05% rather than every 0.10%. We have also rewritten our results section to report the precise values for the time-activity budgets of each species (lines 228-237).

Reviewer 1’s comments:

1. The manuscript PONE-D-22-04625 is fairly well written but would be stronger if the authors (1) included more background information on the use of the reserve wetlands of the two focal species.

– Our response: We are grateful to the reviewer for their positive response and for the constructive comments, which have helped us to improve our manuscript. We have responded to each of the reviewer’s comments in turn, detailing the changes that we have made. On this particular point regarding our focal species, we have added additional information on the two swan species at our study site, including their use of the reserve wetland for all behaviours (lines 116-119), as well as information on their behaviours (lines 77-92). Please see also our response to Section Editor’s Comment #12, which made a similar point.

2. conducted focal observations during a broader range across daylight hours of the day

– Our response: Please see our response to Reviewer 1 Comment 7, where we present new analyses of each behaviour across all of the different times of day at which observations were conducted, which addresses the point raised here.

3. kept conclusions regarding trade-offs in behaviour to the scope of the data as 15 minute intervals are ok to show that differences exist, but the implications of those differences likely have little biological relevance at that timescale for an individual.

– Our response: On reflection we acknowledge the reviewer’s point and so we have revised our text in order to more firmly ground our discussion and conclusions in our data. We accept that the paragraph on the possible biological significance of the species disparity in whether maintenance and foraging showed a negative correlation, was rather speculative given our short observation durations. We have therefore removed this paragraph and instead we acknowledge that the biological significance of this disparity is unclear, and further acknowledge that it is difficult to infer mechanistic explanations for behaviour from short observations (lines 414-424). We have also softened the tone in the final paragraph to better reflect the tentative nature of conclusions drawn from short-term observations (lines 446-463). We have included within our discussion a paragraph which highlights the short duration of our observation periods, and proposes how future research using longer observations could shed further light on the questions that we address in our study (lines xx-xx); please see our response to Reviewer 1 Comment #10.

4. Line 75-83: Following your line of reasoning here that aggression is a means of maintaining access to food resources, I would have hypothesized that birds that were more aggressive had better access to high quality food resources and therefore would need to spend less time feeding as a result.

– Our response: We certainly see the reviewer’s point and agree that, over longer time periods, swans that maintain access to food resources through aggression might spend less time feeding, as they could gain their required energy more efficiently from a high quality food resource. However, over the short time durations (such as the 15 minute observation periods used in our study), any time that is spent on aggression activities is time that cannot be spent on foraging. This would offset any time gained from access to food resources. We have amended the text justifying our prediction to clarify this point about time scales (lines 95-97), so that our revised manuscript sets out the rationale for our predictions more clearly.

5. Line 85-88: These predictions contradict my expectations for why/how aggressive behavior would be rewarded. If a bird that spends a lot of time being aggressive to other birds is rewarded by the richest feeding areas and can quickly obtain the resources it needs then its time spent feeding is less and it receives other benefits like increased survival as feeding can be a risky behavior. Not sure that it is that critical a point as just testing these hypotheses is interesting but the assumption of no correlation between aggression and feeding time stated on Ln 80 is not intuitive.

– Our response: Please see our response to the previous comment (Reviewer 1 Comment #4), in which we explain how we have amended the text of our manuscript to improve the justification associated with our study prediction. Moreover, we also agree with the reviewer’s point that the direction of our prediction would not have affected the result found in our study, as our analyses tested for all types of association between the two behaviours (both positive and negative correlations, as well as no correlations).

6. Line 96-98: Curious if the migratory whooper swans use the reserve for all of their feeding and roosting needs while in the area or if they predominantly feed or roost somewhere outside the reserve?

– Our response: Please see our response to Section Editor’s Comment #12, which made a similar point. In our revised discussion we now compare the percentage of time spent on each behaviour with those reported in previous studies (lines 300-357). If our focal individuals had, for example, only been roosting on our study lake and had been undertaking daily foraging flights to feed elsewhere, then our data would show much higher resting and much lower foraging than in comparable studies. However, for all behaviours the time spent by our focal swans is well within the range reported in previous studies, and indicates that our focal swans were exhibiting their range of behaviours. In addition, it is also clear from previous research (Griffin et al. 2010; now cited in our study) that tracked individual swans at the study site that individuals fed and roosted on the reserve, as many remained there for extended periods of time. Although they observed some localised (<10 km) movements from the reserve to adjacent areas, they did not observe the systematic daily movements that would be expected if swans were only roosting on the reserve and were feeding elsewhere ; we now mention this study in our discussion (lines 320-325), as well as in our methods section in the text on the study system (lines 191-121). Overall, the evidence from our own work, and previous studies, suggests it is unlikely that the swans were performing only certain behaviours on our study lake and other behaviours unobserved away from our web cam; as described above, our revised manuscript now presents and discusses this evidence.

7. Line 168-169: While behaviour may not have varied significantly, I would argue that there was a suggestion of differences between them during different times of day and month – see figure from (13). I would really like to see a breakdown of behaviour by hour from sunrise to sunset to test this.

– Our response: We understand the reviewer’s concerns, and agree that it is important to demonstrate that the time spent on each behaviour did not vary markedly across our observation periods, as such temporal variability within the diurnal period would have meant that it would be inappropriate to pool the observations from different times of day for our main analysis. We have therefore followed the reviewer’s suggestion of assessing whether the time spent on each behaviour by each of the two swan species did vary significantly between the different time periods for which we have data. We used Kruskal-Wallis tests to determine whether there were statistically significant differences in the time spent by each swan species on each behaviour (aggression, maintenance, foraging, resting) between each of the four time periods at which we made our behavioural observations. As we were interested in differences between any time period, we modelled the time periods as factors rather than as a continuous variable. Our new analysis found no statistically-significant differences in any behaviour between the four times-of-day at which sampling occurred, for either swan species. We have updated our manuscript to include this new analysis, in particular with new text in the methods (lines 193-205) and results (lines 243-247,253-257). We have also added two new figures to illustrate the results (new Fig 2 and Fig 3). Whilst we did not have data for every hour of the day, because day-length was short during our winter study period, our observations did span from early morning to dusk. Our earliest observations were from 09:30 and our latest observations were from 14:30; because day-length was short during our winter observations, it would not have been possible to collect data consistently at times that were outside of this sampling window. Crucially, we have now shown that we did not confound our analysis by pooling our data across all four times-of-day. The new analysis that we have added will reassure readers that our decision to pool data did not influence our results or conclusions.

8. Line 176-178: Was it assumed that both species of swans were present on the reserve for the entire day? If not, could the whooper swans be feeding offsite and roosting on the reserve?

– Our response: Please see our responses to Section Editor’s Comment #12 and Reviewer 1 Comment #6, which also address this question.

9. Lines 267-268: The disparity could also be due to differences in behavior by time of day between the species. If the more mobile whooper swans are flying out to other areas for various resources, the behaviour observed in the reserve may be biased. If there is movement data associated with the flocks or individuals it would be helpful to present for a clearer picture of their daily behaviour.

– Our response: The section referred to here by the reviewer has been deleted from the manuscript as part of our revisions in response to another comment (Reviewer 1 Comment #3). However, we agree that the point raised here regarding the potential movements of the birds, and how this could have impacted on our observational data, is an important one that we needed to address as part of our revisions. Therefore, we have added additional information from tracking studies and behavioural studies to provide the reader with information on this point. We now state that previous research (Griffin et al. 2010; now cited in our study) that tracked individual swans at the study site reported that individuals fed and roosted on the reserve, as many remained there for extended periods of time. Although they observed some localised (<10 km) movements from the reserve to adjacent areas, they did not observe the systematic daily movements that would be expected if swans were only roosting on the reserve and were feeding elsewhere ; we now mention this study in our discussion (lines 320-325), as well as in our methods section in the text on the study system (lines 119-121). In our revised discussion we now compare the percentage of time spent on each behaviour with those reported in previous studies (lines 300-357). If our focal individuals had, for example, only been roosting on our study lake and had been undertaking daily foraging flights to feed elsewhere, then our data would show much higher resting and much lower foraging than in comparable studies. However, for all behaviours the time spent by our focal swans is well within the range reported in previous studies, and indicates that our focal swans were exhibiting their range of behaviours. For example, we now report that the 42% of time spent on foraging by mute swans at our study site was similar to the 43% reported for mute swans in eastern England [Wood et al. 2021], 48% at a rural site in Poland [Józkowicz & Gorska-Klek 1996], 41% in Denmark [Holm 2002], and 36% in Ireland [Keane & O'Halloran 1992]. Overall, the evidence from our own work, and previous studies, suggests it is unlikely that the swans were performing only certain behaviours on our study lake and other behaviours unobserved away from our web cam; as described above, our revised manuscript now presents and discusses this evidence. We are grateful to the reviewer for prompting us to include this new information, which we believe has strengthened our manuscript.

10. Lines 278-295 This is a very nice summary of the limitations of the short duration observations and potential alternative approaches in the future.

– Our response: We thank the reviewer for their positive response to this section. We agree that it is important to discuss the limitations of the study, and so we have kept this section in our revised manuscript (lines 417-435).

11. Lines 313-317: Given data is collected in 15-minute blocks and the total budget for an individual for a given day would be a much better measure of trade-offs I would caution not to overstate the findings. Does a trade-off during a 15 minute bout have biological (impacts to fitness) for any individual or should those measurements be compared across a greater timescale before we draw conclusions regarding plasticity in behaviour.

– Our response: We agree that undertaking observations over longer time periods, such as determining time budgets for individuals for a given day, would provide more robust estimates of potential trade-offs between different behaviours; however, this was not an option for our study as we could not identify individuals from web cam footage and individuals typically do not remain visible ‘on screen’ for long enough to permit that approach. We have included text to explain these points (lines 419-423). We have also softened the tone in the final paragraph to better reflect the tentative nature of conclusions drawn from short-term observations (lines 446-463), and we have removed the mention of plasticity in behaviour. In particular, we have been careful not to attempt to draw any conclusions regarding fitness consequences. Instead, our revised conclusion acknowledges the interesting patterns that we have found were derived from short time periods of 15 minutes (lines 416-423, 456-458). We hope that the findings from our short-term study will inform future longer-tem studies, from which more robust conclusions regarding trade-offs can be drawn. Indeed, our study provides advice on how such studies could be carried out (lines 432-435, 458-461). We argue that the patterns of behaviour that we have documented in our study can inform the development of hypotheses regarding behavioural trade-offs in future longer-term studies.

Journal requirements:

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

– Our response: We have amended our manuscript to conform to PLOS ONE’s style requirements. On the title page, we have removed the postal code information from the affiliations listed for each author. The corresponding author’s initials are now given in parentheses after the email address. Within the manuscript, figures are now referred to using the abbreviation “Fig” and the corresponding number. Each figure now has a short title in bold before the main legend text. We have also renamed our files as recommended (for example, the file for figure 1 is now named “Fig1.tiff”.

2. In your Methods section, please provide additional information regarding the permits you obtained for the work. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why."

– Our response: We have amended the methods section to include a specific ethics subsection to provide this information (lines 183-190). This subsection states that our study was carried out with the prior approval of the ethics committee of the College of Life and Environmental Sciences of the University of Exeter (eCLESPsy002195). Moreover, we state that as our data collection was conducted virtually via a publically-accessible live-streaming webcam, no physical visits to the study site were undertaken, and hence no study site permits were required.

3. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

– Our response: We confirm that our study did not feature human participants and so participant consent was not required. Our manuscript reports an observational study of non-human animals. We have amended the manuscript to include an ethics section within the methods (lines 183-190), in order to explain this point.

4. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

– Our response: We have now uploaded our data, as well as the code used in our analysis, to a figshare repository, which is publically available. We would be grateful if the Data Availability statement associated with our manuscript could be updated to reflect that the data and code can be accessed via this DOI: https://doi.org/10.6084/m9.figshare.20063225

5. Please note that in order to use the direct billing option the corresponding author must be affiliated with the chosen institute. Please either amend your manuscript to change the affiliation or corresponding author, or email us at plosone@plos.org with a request to remove this option.

– Our response: The corresponding author, Dr Paul Rose, is affiliated with the University of Exeter, which is the institution selected for billing. We confirm that this affiliation is listed in our manuscript.

6. Please amend your list of authors on the manuscript to ensure that each author is linked to an affiliation. Authors’ affiliations should reflect the institution where the work was done (if authors moved subsequently, you can also list the new affiliation stating “current affiliation:….” as necessary).

– Our response: We confirm that the affiliations for all three authors are listed on the title page. For each author, the relevant affiliation is indicated with a superscript number, as per PLOS ONE’s guidelines.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

28 Jun 2022

Assessing trade-offs in avian behaviour using remotely collected data from a webcam

PONE-D-22-04625R1

Dear Dr. Rose,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Claudia Mettke-Hofmann, PhD

Section Editor

PLOS ONE

Additional Editor Comments (optional):

All issues have been addressed.

Reviewers' comments:

30 Jun 2022

PONE-D-22-04625R1

Assessing trade-offs in avian behaviour using remotely collected data from a webcam

Dear Dr. Rose:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Claudia Mettke-Hofmann

Section Editor

PLOS ONE