Movement predictability of individual barn owls facilitates home range size and survival estimation

doi:10.21203/rs.3.rs-1778136/v1

Download PDF

Research Article

Movement predictability of individual barn owls facilitates home range size and survival estimation

https://doi.org/10.21203/rs.3.rs-1778136/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 07 Feb, 2023

Read the published version in Movement Ecology →

You are reading this latest preprint version

Background.

There is growing attention to individuality in movement, its causes and consequences. Accumulating evidence demonstrates that conspecifics differ repeatably in their spatial behaviors, forming behavioral types (“spatial-BTs”), similarly to other well-established personality traits (e.g., boldness or sociability). Spatial-BTs are typically described as the difference in the mean level among individuals, while the intra-individual variation (IIV, i.e., predictability) is only rarely considered. Studies identifying the factors that determine predictability or its ecological consequences for broader space use patterns are even more sparse, in part because it was mostly tested in captivity (e.g., with repeated boldness assays). Here we test if [i] individuals differ in their movement and specifically in their predictability. We then investigate [ii] the consequences of this variation for home range size and survival estimates, and [iii] the factors that affect individual predictability.

Methods.

We tracked 92 barn owls (Tyto alba) at very high resolution (every few seconds) using an ATLAS system, and monitored their survival. We calculated movement and space-use indices (e.g., max-displacement and home-range size, respectively) from the trajectories (115.2 ± 112.1 nights; X̅±SD) and then used double-hierarchical and general mix models to assess spatial-BTs, individual predictability in nightly max-displacement, and its consistency across time. Finally, we explored if predictability levels were associated with home-range estimates and survival, as well as the seasonal and demographic factors (age, sex) that affect it.

Results.

Our dataset revealed clear patterns of individualism in owls’ movement. Individuals differed consistently both in their mean movement (e.g., max-displacement) and their IIV around it (i.e., predictability). Further, more predictable individuals had smaller home-ranges and lower survival rates, on top and beyond the expected effects of their spatial-BT (max-displacement), sex, and age. Juveniles were less predictable than adults, but sexes did not differ in their predictability.

Conclusion.

These results demonstrate that individual predictability may act as an overlooked axis of spatial-BT with potential implications for relevant ecological processes at the population level and individual fitness. Considering how individuals differ in their IIV of movement beyond the mean-effect can facilitate understanding the intraspecific diversity, predicting their responses to changing ecological conditions and their population management.

Animal personality

behavioral plasticity

behavioral syndromes

bio-telemetry

double-hierarchical general linear model (DHGLM)

high-throughput tracking

movement ecology

Animal movement reflects the interaction between an individual's needs and its changing environment, with direct impacts on its fitness and central ecological processes, such as foraging and dispersal [1, 2]. Movement patterns show remarkable diversity across different scales, varying among species, populations, individuals, and also within individuals across time [3–5]. Differences in movement among conspecifics are often related to environmental heterogeneity (e.g., in resource distribution), demographic factors (e.g., sex and age), or short-term behavioral states [e.g., hunger level, 6–9]. In addition to these well-established factors, there is a growing attention to individuality in movement and its influence on ecological patterns and population response to changing conditions. Yet, both the patterns of individuality and their consequences remain understudied [10–12].

Studies in behavioral ecology have long-established the framework and tools to quantify consistency in individual behaviors, and the existence of animal personalities (aka temperament or behavioral types; hereafter BTs) in traits such as boldness, exploration and sociability [13, 14]. More recently, accumulating evidence demonstrates that individuals are also repeatable in their movement and spatial behaviors, forming “spatial-BTs” [10, 15]. Examples include repeatable home range behaviors [16, 17], habitat use [18, 19], or finer scale movement indices like maximal daily displacement [20, 21]. The growing interest in spatial-BTs is driven in part by the ever-improving ability to track more individuals at better resolution and accuracy, providing enough data on sufficient sample size to dissect the contribution of among-individual variation from other factors like background environments [22]. These rich datasets, in turn, can reveal the impact of fine-scale movements on social interactions, disease transmission and other behaviors that may influence fitness and population dynamics and ultimately the ability of species to adapt to environmental changes [23–25].

Despite the immense progress in describing spatial-BTs, this is mostly done as the difference in the mean level among individuals [e.g., 15,17,21]. Differences among individuals in their intra-individual variation (IIV) have been thoroughly discussed for other behaviors [e.g., boldness, 26] but are rarely considered in movement studies. IIV is often termed as individual unpredictability [26–28]. It is defined as the residuals in a measured behavior after accounting for the among-individual differences in their mean response, and for the plasticity of their response to environmental gradients [which may also differ among individuals, 26,28–30]. For instance, Stamps et al. [26] repeatedly assayed the boldness (latency to emerge after disturbance) of hermit crabs (Pagurus bernhardus). They found that individual crabs differed in their mean level and had a plastic response of diminishing trend (habituation). Yet individuals also differed in their predictability around the individual-specific expected level: some were predictably closer to the expected trend while others showed high IIV around the trend-line. Broadly speaking, predictability usually increases (IIV decreases) with age or experience [sometimes referred as canalization, 31]. Predictability may also vary between sexes, depending on the species’ natural history. Both predictability and plasticity can be regarded as axes of BT if individuals are stable over time. Quantifying predictability (or IIV) requires repeated assessments of the behavior in mind. Hence, most studies on this topic were done in captivity [e.g., 31,or 32]. Despite the methodological advantages of captive-assays, this approach is limiting the ability to link observed differences in predictability with their in-situ ecological consequences for individual fitness, space use, and ultimately population dynamics [33]. Exploring predictability in movement behavior can overcome these cons and offer a suitable alternative approach.

Biologging datasets may provide the required large samples of repeated measures per individual [e.g., daily routines, 22,34] with a balanced design and large sample sizes across individuals. Studies by Cleasby et al. [35] and Hertel and Niemela [36] also provide a methodological route to test predictability in movement by double-hierarchical general linear models (DHGLMs) that simultaneously capture individual variation in intercept (behavioral type), slope (behavioral plasticity), and residual variance (predictability). Recent work by Hertel and colleagues [36, 37] suggests that highly predictable individuals coexist along with unpredictable individuals within the same population, and that predictability may co-vary with movement rate creating a movement syndrome [sensu 21] that may have potential ecological impacts.

Ideally, quantifying predictability in movement indices requires large datasets with multiple individuals simultaneously tracked at high resolution over long periods (thus experiencing relatively similar external conditions) [30]. Indeed, empirical in situ examples are still particularly rare, at least in part due to these data requirements. Thus, it is still unknown whether spatial-BTs differ also in their predictability (or just in the mean trait), and if the predictability is potentially another relevant spatial BT axis. Further, to the best of our knowledge, none of the previous examples tested if predictability in movement are associated with other ecological properties, such as broad-scale space-use patterns or fitness proxies [as was done for other behaviors, 33]. Similarly, it remains unknown if age and other plausible predictors (like sex) influence predictability in movement, as was shown for other behaviors [31 and reference therein]. This sets the ground for broadening the axes of individual variation in exploring also consistent differences among individuals in their movement predictability.

To address these gaps, we focused on a population of barn owls (Tyto alba). This nocturnal predator plays a key role as a pest control agent in agricultural areas. Accordingly, barn owls’ natural history has been relatively well described [38], and their high reliance on nest boxes [aimed to enhance carrying capacity, 39] along with ongoing monitoring facilitates captures and tracking. High-resolution tracking data allows us to quantify their predictability and its ecological consequences. Specifically, we ask three broad questions (followed by corresponding hypotheses). First, what are the patterns of among-individual variation in movement? We hypothesize that [H1.1] individuals differ consistently in their movement patterns (i.e., showing a spatial-BTs in their mean levels); that [H1.2] they also differ in their predictability (IIV around the mean); and that [H1.3] these varying levels of predictability are repeatable across time [i.e., an individual that was relatively predictable in period A will be relatively predictable in period B; e.g., 33]. Second, we ask what are the ecological consequences of predictability? Since BTs are known to affect broader space use pattern and fitness [10, 40], we hypothesize that predictability may be associated with both [H2.1] home range size and [H2.2] individual survival. Lastly, we ask what factors affect the predictability of individuals? We hypothesize that [H3.1] age and sex will co-vary with this index, as was shown in several previous examples [limited mostly to captive behavioral assays, 31,36,41].

2.1 Study species

The barn owl (Tyto alba) is a medium-size owl (length 33–39 cm; wingspan 80–95 cm) from the Tytonidae family, with a cosmopolitan distribution [38]. It is known as a generalist top predator, which feeds almost exclusively on small mammals located in agricultural fields and open grasslands. At the same time, as secondary cavity breeders, barn owls are limited by availability of suitable sites in these habitats. As a result, they tend to breed in man-made structures (e.g., barns, farmhouses, and ruins) - as their name indicates.

Barn owls have a remarkable breeding capacity. They lay large clutches ranging from 2–11 [e.g., ~ 4.6 chicks in Israel, 42]. The laying interval is about two to three days, with asynchronous hatching (i.e., the female incubates from the first egg laid) resulting in substantial age differences within the brood. Eggs require, on average, 32 days (range 27–36 days) of incubation. Hence, the overall incubation period varies from 34 for up to 59 days for clutches of 11 chicks (32 + 27). During incubation the female usually remains inside the nest cavity, while the male provides her food requirements. When all eggs are hatched (usually about two weeks after the first one is hatched), the female starts leaving the nest more regularly. Fledglings leave the nest for the first time at approximately the age of 55 days [43], but the parental care extends for ~ 30 more days [44]. In Israel, eggs are laid around March [extending from January to June, 42], and pairs can sometimes raise up to two broods annually.

2.2 Study system

The study was conducted at the Harod valley (32°30'N 35°29'E), north-eastern Israel. This valley is dominated by intensive agricultural landscapes (field crops, orchards, banana fields and fishponds), and holds a few rural settlements. Over the last three decades, dozens of nesting boxes were deployed throughout the area for barn owls` usage, enhancing their local resident population [39]. Every breeding season, nest boxes are being monitored during the day (inactivity period for the owls) as part of the ongoing monitoring project. These efforts include deployment of metal rings for individual identification, standard measurements (wing length and body mass), and aging. Sex was determined with genetic kits on feathers samples (Karnieli-Vet Ltd., Kiryat Tiv’on, Israel). In addition to the ongoing monitoring, from December 2019 to February 2022, we fitted tracking devices (hereafter tags) to adults and fledglings (wing > 280 mm) in good body conditions (a total of 94 owls). Tags were attached with a Teflon harness in a backpack or leg-loop configuration (total weight of the device and harness was 13 ± 1 g; <4.3% of the body mass in all cases; Fig. 1). Trapping and tagging procedures were authorized by permits 2019/42155 and 2020/42502 from Israel Nature and Parks Authority.

2.3 The ATLAS tracking system

We have used the ATLAS wildlife tracking system to address our hypotheses and track barn owls. In addition to the tags, the ATLAS system includes a set of ground stations with tower-mounted antennas and central data-processing and storage servers [45–47]. The system uses a reverse-GPS approach (i.e., localizations estimated from data captured by the stations whose coordinates are known and not by the tags themselves). Tags transmit radio signals with unique tag-IDs at 1/4 or 1/8 Hz (depending on settings) and these are received by ground stations. If three or more ground stations successfully receive a given signal, then tags’ location is computed at high accuracy [ideally about ± 5m, 47,48], from the times of arrival of the signals. The use of ground stations limits the coverage to a particular region (whereas GPS offers global coverage), but this is acceptable for studies of barn owls that largely remain local. Such an approach allows tags to be inexpensive, lightweight, and battery efficient, compared to other alternatives (e.g., GPS where the tag localizes itself), providing many more data points per gram. These features make the system particularly effective in collecting high resolution and accurate data for many individuals over extended periods (months-years) and hence appropriate for investigating individual variation in movement patterns and predictability.

2.4 Data analysis and statistical methods

Filtering and processing. All data analyses were done in the R environment [49] with Rstudio [50]. To ensure data quality we applied several data filtering and segmentation steps. First, following a previously described pipeline [51], we filtered out points with low system-accuracy estimate (STD > 50m) or with non-realistic speed-line > 15m sec^− 1. Second, to reflect the owl's nocturnal activity, the data was segmented into nights, starting at 5 p.m. and ending at 6:00 a.m. Daytime data was excluded, and nights were used as the base unit for all further analyzes. Third, we computed the nightly movement indices (see below) only if tracking provided sufficient data points (more than 1000 points per night). Fourth, our data [female static behavior, chick ages using wing length of the oldest fledgling in the nest, 52] allows us to back-calculate the start and end dates of the incubation period for each female and hatching dates for the fledglings. We excluded incubation periods from female movement analysis and considered transition from fledgling to adult at the age of one year. Fifth, because barn owl behavior and movement patterns change drastically along the year, we divided the data into three trimesters, representing different stages in the owls’ reproductive cycle and seasonality (1st-trimester: Feb-May, incubating/nesting period; 2nd-trimester: Jun-Sep, rearing/post-breeding period; and 3rd-trimester: Oct-Jan, fall-winter time). We then calculated movement indices for each trimester excluding owls with less than 25 nights in a given trimester.

Several movement indices are conceivable for quantifying individual consistency, such as the nightly total-distance (i.e., the sum of all movements/flight segments in the night), trajectory openness and others [see 34, this volume]. To minimize the influence of location errors during stops, we segmented the trajectories by their activity mode, move (i.e., fly) or stop segments, and estimated total distance for moving segments only. These indices were computed using the AdpFixedPoint function from R-package toolsForAtlas. For simplicity, we focus here on results from the commonly used nightly max-displacement as the main index of movement, defined as the distance between the first point of the night (typically the nest box) to the most distant point in the nightly trajectory.

Estimating individual repeatability. To test our initial hypothesis regarding consistent individual variation in movement [H1.1], we calculated repeatability (Rp) using Nakagawa and Schielzeth [53], as the proportion of the total variance accounted for by differences among individuals. In addition to repeatability, we also report the coefficient of variation for among-individual variance (CVi), calculated as the among-individual variance standardized by the trait mean. Both repeatability and CVi are population-level estimates of the degree of individual variation, with CVi suggested as a more robust estimate [36].

Individuals may differ in their average nightly max-displacement (flying near or far) and also in their variability around their mean. Some individuals are unpredictable and are producing a broad range of nightly movement ranges, whereas more predictable individuals are narrowly centered around their own average (Fig. 1). We measured individual predictability [H1.2] by estimating variation in residual intra-individual variation (rIIV), i.e., the spread of residuals around an individual reaction norm following the protocol of Cleasby et al. [35] and Hertel et al. [36]. We used the R-package brms [54] to fit a double-hierarchical general linear model (DHGLM) to our datasets with nightly max-displacement as a response variable. Individuals with a high residual variance in the DHGLMs are accordingly more unpredictable than individuals with a low residual variance. When necessary for biological interpretation of rIIV values we back-transformed them to the original scale (km) by taking the exponent of its logged values from the DHGLM outputs.

Finally, to determine whether individual predictability is consistent across time [H1.3; e.g., 33], we calculated trimester-specific predictability values and established individual consistency also in this aspect, similarly to the calculation of repeatability and CVi of max-displacement. Then, we established the ecological relevance of individual predictability for broad topics in movement ecology by exploring the predictive power of this index on home-range (HR) estimates [H2.1] and survival [H2.2], on top and beyond commonly investigated indices of age, sex, and mean max-displacement.

Home range analysis. We estimated the owls’ home range for each trimester, using the autocorrelated kernel density estimate [akde, 55] from the R-package ctmm [56]. Due to the function's long running time over our huge dataset (with millions of points), for this analysis we have subsampled the data at a 10 min interval. Further, for representing its HR as locations where the owl actually chooses to stand on (e.g., perch or rest) and not places where it flies through, we have used only the segments of activity mode identified as stops (and not flight), as described above. An individual with unknown sex and an individual whose HR model did not converge were excluded from this analysis.

To explore the factors related with the HR size, and in particular the effect of predictability [H2.1] while accounting for other factors, we compared a set of linear mixed effect models [lmer function from R-package lme4, 57]. All models included Log(HR) as the dependent variable and individual’s ID as a random effect. Fixed effects varied among models and included three categorical ones: age (adult vs. juvenile), sex and trimester. Continuous fixed effects (standardized before inclusion) included the number of tracking nights in the trimester, mean max-displacement, mean total-distance and mean rIIV (index of predictability of max-displacement). We checked for collinearity among fixed effects and considered models with most of their possible combinations, excluding a few models with singularity issues. Models were ranked with the corrected Akaike information criterion (AICc) using R-package AICcmodavg; [58] and prediction plots from the top models were generated with the R-package effects [59].

Survival analysis. For testing our hypothesis that predictability affects survival [H2.1], beyond the effects of age, sex, and max-displacement we modeled survival using Cox hazard regression [60, 61]. Out of our focal sample (N = 74) 22 individuals died during the study, many of them through collision with cars, a well-documented mortality factor of barn owls around the world [38, 62, 63]. The regression was calculated using coxph [R-package survival, 64,65], and a Forest plot for the model was drawn using ggforest [R-package survminer, 66]. Tracking duration (days until last known status), and final status were recorded for each individual, where status can be “censored” (live or unknown) or dead. Age was classified as juvenile or adult for each individual. Five individuals that transitioned between age groups during the tracking period were considered separately for each group (i.e., surviving for one year as juvenile, and then according to their adult fate). Since our regression model does not account for the repeated measures of these five individuals, we repeated the analysis using mixed effects Cox hazard regression model [coxme, 67]. This version (with ring-ID used as a random factor) resulted in very similar results to the simple regression described above (see Tables A1, A2), and we kept the simple model.

Factors affecting predictability. Finally, for testing our third hypothesis regarding the factors affecting individual predictability, we have modelled rIIV (un-predictability) as a function of age group, sex, year, and trimester. We constructed General linear mixed-models similarly to the ones described above. For this model set, we considered all possible combinations of the above-mentioned fixed effects using the function dredge [R-package MuMIn, 68]. We ranked the models according to their AICc values and investigated the effect size of the relevant models. As yet another validation of the factors affecting rIIV, we also fitted a DHGLM with nightly max-displacement as a response variable, individual identity as a random effect and sex and age as fixed effects both in the mean model and in the model of the residuals. These models showed the same qualitative results as the simpler models described above.

We obtained data for 92 different individuals during 2019–2021, with tracking duration of 115.2 ± 112.1 (X̅±SD; range: 1-545) nights per individual, and a total of over 10,600 nightly trajectories. After data filtering and excluding birds with insufficient data per trimester, our dataset includes 128 trimesters-bird combinations with 74 individuals and an average of 1.7 ± 1.0 (range: 1–5) trimesters per individual. These birds include 45 females and 29 males from which 34 are adults and 45 are juveniles (five individuals transitioned from juvenile to adult during the tracking period). See Table A3 for full details.

Barn owls typically start their nocturnal activity at their central roosting site (often a next box) and use several perching points during the night before returning to their day roost. They pass a nightly distance (i.e., a ‘total distance’) of 11.82 ± 7.73 km (range: 2.26 to 51.89), with maximal displacement of 2.45 ± 0.93 km (range: 0.56 to 5.41 km). Season has a strong effect on barn owls’ movement. During the breeding season, owls act as central place foragers, frequently returning to the nest to provide for the incubating female (the males) and later for the chicks (both parents, males often more intensively that females). After fledging and a short flight-learning period juveniles had longer distances compared to adults, and males show longer distances compared to females (one-way ANOVA, F_df=3=8.97, P < 0.001; see Fig. A1).

3.1 Individual variation in barn owl movement

In agreement with our hypothesis [H1.1] barn owls show a significant individualism in their movement, with repeatable movement indices. For max-displacement Rp = 0.228 [0.158, 0.297; confidence interval] and coefficient of variation for among-individual variance (CVi) = 0.362 [0.303, 0.426]. Individuals also differ in their overall predictability estimates (Fig. 3). Trimester-specific predictability values (for birds tracked in more than one trimester) reveal that also this trait is strongly repeatable for each individual, with repeatability estimated as Rp = 0.424 [0.185, 0.638]. These results were not limited to the index of max-displacement, we found quantitatively similar patterns also for the index of nightly ‘total distances’, with repeatable values and varying predictability among individuals (Table A4). Overall, these findings support our second [H1.2] and third [H1.3] hypotheses regarding IIV values being consistent individual traits, and below we turn to explore the ecological consequences of this variation and the factors that affect it.

3.2 The ecological consequences of predictability in movement

Individual predictability is related to its HR size. Ranking models for HR size show that the only two models reliably explain the data, while all other considered models have weak predictive power (Delta_AICc > 4). The top model and second-best model (Delta_AICc = 2.24) include almost the same fixed effects (Table 1, see model ranks for all top-ten models in Table A5). Most importantly, our index of predictability (rIIV) was included in both top models, together with mean max-displacement as well as age, trimester, and number of tracking nights within the trimester. Sex was included in the second-best model.

Table 1

Model ranking for models of HR size using modified Akaike Information Criterion (AICc). The top two models are shown, all other models have Delta_AICc > 4 (Table A5). Both leading models include the following fixed effect **Age**; **Trimester**, **TrackingNights** (the number of tracking night within each trimester), **MaxDisp** (mean maximal displacement) and most importantly for our hypothesis - also the **rIIV** (the index of predictability) for max-displacement. The second model also included **Sex**.
Fixed effect included in the model	K	AICc	Delta_AICc	AICcWt	Cum.Wt
Age, Trimester, TrackingNights; MaxDisp; rIIV	9	345.64	0	0.54	0.54
Age, Trimester, TrackingNights; MaxDisp; rIIV, Sex	10	347.88	2.24	0.18	0.72
All other models	7–10	> 349.94	≥ 4.3	< 0.06

Focusing on predicted effect sizes (Table A6, Fig. 4) from the leading model (which were almost identical in the second model) demonstrates that - in agreement with our hypothesis [H2.1] - the predictability per se has a major effect on HR size, independent of all other considered factors. Trivially, max-displacement (Fig. 4A) has the strongest effect on HR, with individuals moving further at the nightly scale having also large HR at the trimester scale. More surprising is the strong effect of predictability (Fig. 4B), which equals to about a half of the former effect of the max-displacement. The positive effect indicates that individuals with higher rIIV (less predictable) have larger HRs. Age has an expected effect with juvenile barn owls having larger HR than adults (Fig. 4C). Trimesters differed as well (Fig. 4D), with HR size in the 1st trimester (Feb-May; breeding season) being lower compared to 2nd (Jun-Sep, rearing/post-breeding period) and 3rd (Oct-Jan, fall-winter time) trimesters (with no clear difference among the latter two). Tracking nights (both models; Fig. 4E) and sex (second model only, not shown) had very weak effects. While intuitively longer tracking should yield larger HR (at least over short periods), we find a weak but negative effect of tracking nights. This likely reflects our conservative minimal threshold (at least 25 nights), ensuring this initial growth is already included. The negative trend may indicate growing representation of intensively used sites resulting in slightly smaller kernels. Notably, although total distance was included as a possible fixed effect in various models within the set, it was not highly ranked, and is absent from all leading models.

Individual predictability is related to its survival. In agreement with the hypothesis [H2.2] that survival will be related with predictability, we find that rIIV of max-displacement negatively and significantly associate with risk level in the Cox hazard regression. This negative effect means that individuals with higher rIIV value (i.e., less predictable) seem to survive longer, while more predictable individuals (lower rIIV values) are more prone to mortality (Fig. 5; Table A1). None of the other predictors significantly affected regression results. Neither mean max-displacement nor did male differ from the females (reference category) survival.

While age-dependent diminishing mortality risk is commonly observed in many biological systems - barn owls included [69] - this general pattern is only weakly reflected in our data. Juveniles showed a (non-significant) trend toward higher hazard compared to adults (the reference category) and this modeling result is also reflected in actual mortality rates (10/33 adults vs. 12/46 juveniles died during the study). Presumably, two factors may contribute to this relatively high adult mortality: first, dispersing juveniles readily leave the region (resulting in ‘censored’ status). Indeed, tracking duration for juveniles was marginally shorter than adults (127.9 ± 92.7 vs. 172 ± 112.3, respectively; T_df=60.6=1.84 P = 0.06). Second, anthropogenic factors (mostly collision with cars) that are disproportionately hitting adults (66% of car hits were of adults although they were only 42% or tracked individuals), presumably since they engage in more intensive foraging than juveniles.

3.3 Factors affecting individual predictability

Our last hypothesis [H3.1] that predictability depends on age and sex was only partially supported. The data show that rIIV depends on max-displacement, and the age group but was independent of sex (Fig. 6), with barn owls becoming predictable with age (aka ‘behavioral canalization’). First, a direct comparison reveals that juveniles are less predictable than adults with an average value of rIIV of 2.25 ± 0.83 km vs. 1.23 ± 0.59 km, respectively. In contrast to our expectation, sex did not affect rIIV and females and males have average rIIV values of 1.74 ± 0.81 km and 1.98 ± 1.02 km, respectively (two-way ANOVA for sex and age F_df=1,1=44.05, P < 0.001). This basic result is also supported by model comparison: only a few models in our compared set of 16 models (with combinations of age, sex, trimester, and year as fixed effect) gained sufficient support (Delta_AICc < 4). The top two models (with a cumulative weight of 85%) both included age, which was also the only predictor included in all other subsequent relevant models (Tables 2, A7). Age shows a significant positive effect on rIIV in these models (Table A8), implying that barn owls become predictable as they mature. Year (second-best model) had intermediate effects size indicating some variation in this category, with slightly higher predictability (low rIIV) in 2021 in comparison to 2020.

Table 2

Top two models estimating rIIV. The leading model includes only age as a fixed effect and the second model fixed effects are age and year.
Fixed effect included in the model	K	logLik	AICc	Delta_AICc	AICcWt	Cum.Wt
Age	4	-145.97	300.26	0	0.49	0.49
Age, Year	5	-145.20	300.88	0.62	0.36	0.85
All other models	3–8	<-146.87	> 304.22	≥ 3.96	< 0.07

In this study, we have used the cutting-edge ATLAS system to track barn owl movement at high resolution (fixes every few seconds). This rich dataset, including 74 individuals tracked over long periods of several months each, revealed clear patterns of individualism in movement. Individuals differed consistently from each other in their mean movement indices (max-displacement, total distance) and in their residual intra-individual variation (rIIV), a measure of (un)predictability. Some individuals were highly predictable while other were more variable around their mean nightly movement, and this trait was repeatable across trimesters. These two axes of spatial BT (mean movement and predictability level) are not only distinguishing different conspecifics, but they are also associated with apparent ecological consequences: more predictable individuals (low rIIV) have smaller HRs and lower survival rate. The effects on HR and survival are above-and-beyond the commonly studied effects of an individual's max-displacement (positively affecting HR size), age (juveniles have larger HR but similar survival in our dataset), and sex (no strong effect on HR or survival). Age explained some of the differences among individuals’ predictability, with juveniles being largely less predictable than adults, while sex had no effect. To the best of our knowledge, these links are largely under-explored in the literature of movement ecology. Indeed, there is a growing recognition in the importance of spatial-BTs for various ecological properties [10, 15, 21], yet only a few studies have quantified the predictability of individuals in their movement patterns over time in natural settings [e.g., 36]. Further, almost none has shown predictability can actually relate (or even influence) broader ecological patterns nor explored what determines individual predictability in movement. Below we discuss the ecological importance of predictability, compare our findings to similar systems and point out some possible future directions.

4.1 Spatial behavioral types (BTs) and movement predictability

Consistent among-individual behavioral variation (aka personality, or behavioral types; BTs) has become a mainstream in behavioral ecology [70]. More recently, approaches developed in behavioral ecology has been adopted to study individuality in movement and spatial-BTs [10, 12]. Indeed, movement and space use behaviors show relatively high repeatability [~ 0.67, 15] compared to the mean level of other behaviors [Rp = 0.37, 70]. These higher values reflect the fact that individuals are repeatedly “assayed” in different environments. The observed differences among them reflect, at least to some extent, the environmental heterogeneity among their home ranges (HR) [71, 72]. Our repeatability estimates for the barn owl movements fall on the lower side of this range (R < 0.4). This lower value might reflect a considerable HR overlap among many of our individuals, tracked within a relatively small area. Thus, owls might be experiencing relatively similar conditions that reduce the contribution of spatial heterogeneity to apparent repeatability.

Studies investigating predictability in movement are particularly rare, even though movement data is very suitable for such investigation by offering numerous repeated “assays” (e.g., days) for each tracked individual [36, 37]. Here we have used merely two indices of movement (max displacement and total distances), but numerous other indices can be relevant. For instance, Michelangeli et al. [21] have used these two indices, and several others in order to demonstrate that sleepy lizards (Tiliqua rugosa) have distinct spatial BTs, and that some indices co-vary (i.e., movement syndromes). Because some of these commonly used movement indices (e.g., total tracking distance) are sensitive to sampling rate [73], it is essential to unify rates across individuals (or to subsample some for unifying it). Future studies may sub-sample larger datasets like the one included here to directly test sensitivity of predictability estimates of different movement indices, or of their combined scores (e.g., via principal component analysis), to the properties of tracking datasets (e.g., resolution, duration, number of individuals).

Broadly speaking, quantifying predictability from movement tracks should be more accurate when using high resolution data, simply because it improves the estimation of the track properties. Our data is of exceptionally high temporal resolution (a fix every ~ 8 sec), but we suggest that slightly lower resolutions (e.g., fixes every few minutes) should still be suitable for quantifying predictability as they are quite reliably daily movement patterns. Further, predictability estimates for other temporal scales (e.g., weekly, seasonally, yearly) might be highly relevant for other systems, depending on the duration and resolution of the tracking data at hand. Even coarse resolution data might still be suitable for space-use patterns [e.g., seasonal HR, 16] but probably less for precise description of the daily movements as done here or for brown bears (Ursus arctos) [37]. Given that high-resolution datasets with tracking of numerous individuals over longer periods are becoming more readily available thanks to improving technologies [23, 45], we suggest that quantifying spatial-BTs and specifically movement predictability will be very feasible for many more study systems. These detailed tracking datasets, along with complementary sensor data (e.g., accelerometers) will allow more scholars to address individuality of movement, and investigate the factors that shape it, and its consequences for individuals and species across space and time [22].

4.2 The ecological importance of movement predictability

An association with home range size. The ecological or evolutionary consequences of BTs in populations is well acknowledged for various systems and contexts [74, 75]. Less is known about the implications of spatial BTs, and the association of their fine-scale movements (e.g., daily indices) with broader space use patterns at the HR scale. Accumulating examples demonstrate that variation in HR size is repeatable, and that it can be linked to independently measured BTs [e.g., 16,76,77], or to spatial-BTs assess from the movement data itself [17, 21]. Arguably, for indices like the ones used here (namely max-displacement) local and global scales are expected to co-vary, with individuals that move more at the daily scale also covering larger HRs. Indeed, we also find this positive effect size of max displacement on HR (Fig. 4A).

In addition to this somehow trivial effect, we considered sex as a predictor of HR. In contrast to previous studies of barn owls that found larger male HR sizes during the breeding season [38], we did not find such effect. This result might reflect the exclusion of incubation periods from the females data (something that was not possible in previous studies); the longer scope of our tracking where seasonal changes across the annual cycle mask the differences during breeding season; or due to differences in home-range estimators [78]. Interestingly, even while accounting for other confounding factors (including max displacement), we discover that predictability per se was a strong predictor of HR size, with an effect size of almost a half of the max-displacement (Fig. 4B).

We find that unpredictable individuals (high rIIV) have larger HRs (even within a cohort and sex group). This finding might reflect more consistent and repeatable use of the same locations and is in agreement with previously described syndromes in movement patterns [15, 20, 21]. Future studies may explore if predictable individuals (in our single index of max displacement) tended to also do their longer movement in the same locations, and also in similar directions resulting in smaller HRs. Un-predictable individuals, in contrast, might also show lower predictability in where, when and in what direction they do their long movement, revealing further structures in barn owl movements. More generally, if the strong association between individual predictability and its HR are generalizable beyond our focal population (regardless the specific mechanism proposed here) this could provide an insightful currency that does not require additional data collection, but merely to include this aspect in relevant models of the existing tracking data. Given the high relevance of accurate HR estimates for diverse questions in ecology and conservations [55, 79], adding this predictive power can facilitate ecological application and improve prediction accuracy of future models. For instance, HR overlap among neighbors is often an important for disease and parasite transmission [80, 81].

Consequences for survival. Studying behavioral predictability in the wild might reveal life history trade-offs which are not present in laboratory studies [82]. Only a handful of studies quantified IIV in situ permitting investigation of the ecological significance of predictability. Examples include IIV in aggressiveness of male deer affecting their mating success [33]; predictability affecting foraging site fidelity of kittiwakes [83], and brown bear movement predictability potentially affecting their ability to cope with changes in the environment [37]. Such studies demonstrate various potential ecological implications of IIV, but to the best of our knowledge none of the previous examples demonstrated that predictability affected survival, which is arguably a corner stone in individual fitness.

Our long tracking and large sample size allowed us to identify the fate of many individuals, with considerable mortality rates during the study. We find that predictable individuals survive less. This might appear as counter intuitive since predictability can be preserved as a risk averse behavior. Yet, this interpretation for our context (simply higher rIIV in max displacement) is not accurate. Un-predictable movement can in fact reflect the adaptive response to changing ecological conditions that were unmeasured (hence adaptive plasticity). It can also reflect exploration bouts with long flights (as evident also in Fig. A2) providing information and access to spatially different resources [84]. Further, although individuals expressing greater “risky” behaviors should in-theory suffer from higher mortality, bolder BTs were sometimes found to live significantly longer in the wild (in contrast to the laboratory), suggesting that apparently risky behavior in one context can be beneficial in others [82]. Such finding agrees with our result that unpredictable barn owl have higher survival rate.

4.3 Factors affecting predictability

Identifying what determines individual BT, and here specifically predictability in movement is essential for a mechanistic understating of this phenomena and its ecological consequences as well as generalizing across systems. Variation in predictability can reflect both proximate and ultimate reasons. At the proximate level, we find that older individuals are more predictable than young ones, as was demonstrated before for the predictability of other behaviors [31, 36, 41]. Common explanations for such ‘canalization’ include the idea that with increased exposure to predictable cues, individuals may become more certain in their assessment of the environment allowing traits to become fixated. Other suggested explanations have to do with traits being linked to physiological factors such as growth rate and hormonal profiles, and behaviors change concurrently with these physiological reorganizations.

Observed consistent behavioral differences among individuals recorded under natural conditions may not reflect only inherent differences in spatial BTs, but also arise from (at least partially) differences in the geographical features or resource distribution [15, 72]. Here we do not find support for this general pattern in explaining variation in predictability. In our study system there is an elevation gradient (from + 100m ASL at the west, down to -250m BSL in the east), affecting local temperature, soil type and crop type, which in turn determine food availability (abundance and species composition of the rodent community). Yet, variation in predictability was not associated with the gradient (Fig. A2), and more/less predictable individuals where apparently randomly distributed along the region, supporting the argument that our measured behavior results from a genuine inherent behavioral tendency and preference. In the future, experimental approaches (e.g., relocation) can reveal the distinct contributions of background heterogeneity and the individual tendency to spatial-BTs [72].

On a more ultimate level, differences in spatial-BTs can arise from individual diet specialization - variations in preferences for certain habitat features, resources (a particular type of food), and/or social interactions [85]. While some of these are likely consistent at the individual level [86] others will depend on the environment. In the future, studies can explore if differences in predictability are associate with diet preferences, and also what are the mechanisms that generate apparent differences in predictability. One may ask whether the shape of the distribution of the index in mind (here max displacement) changes? Namely, this asks if unpredictable individuals just having larger variation, or do they have a fat tailed distribution? Indeed, here we find support for the latter, with high rIIV values appearing through some night of very long-range displacement (Fig A3), and not though a homogenous larger variation. Whether these long range reflect a distinct behavioral mode expressed only in certain individuals (and why) remains unknown. Such modes can be inferred from modeling step size and turning angle distributions, allowing identification of different subgroups in the population [87, 88]. Applying such an approach to spatial-BTs of different predictability may offer a strong integration of existing tools from movement ecology with new concepts from behavioral ecology.

4.4 Concluding remarks

In this initial study we only explored spatial BT in two traits (travel distance and max-displacement) and predictability of the latter, but future studies can expand to various other classical indices of movement [e.g., tortuosity, 20,21] and space use [15, 89], that affect Spatial-BTs and spatial syndromes, as well as newer indices such as movement openness [34]. Similarly, other factors affecting rIIV beyond those explored here (sex, age, year and season) and other ecological and evolutionary consequences (beyond HR size and survival) can be investigated, such as spatial-BT implications for parasite prevalence [90], or breeding success [33]. Finally, our tracking is biased to owls staying within the ATLAS region. If individuals with remarkably different spatial-BTs choose to leave the region completely [91] then our sample might be underestimating the variation in the population. Despite these limitations, we were able to describe the prevalence of spatial-BTs and consistent variation in individual predictability within our population. We also validated the ecological relevance of predictability to their fitness and broader space use. Unpredictable individuals may cope better under changing environmental conditions [37]. Human-induced environmental changes act as a non-random filter selecting for species capable of coping with such changes. Likewise, they can also affect intra specific composition, with predictability as a key axes that is still poorly explored [26, 30]. Understanding the scope and role of intraspecific variation in movement and other spatial behaviors will help us better conserve species and their habitats in the face of unprecedented environmental change.

Ethical Approval and Consent to participate

Trapping and tagging procedures were authorized by permits 2019/42155 and 2020/42502 from Israel Nature and Parks Authority.

Consent for publication

All authors have agreed to publish the manuscript’ content.

Availability of supporting data

The raw data will be published in Movebank

Competing interests

The authors declare that they have no competing interests.

Funding

This work was funded by grants from the Israeli Science Foundation (ISF396/20) and ISF1919/19, the Data Science Center at Tel Aviv University (TAD), and the Koret-UC Berkeley-Tel Aviv University Initiative in Computational Biology and Bioinformatics. SC was also supported by the Hoopoe Foundation, Society for the Protection of Nature in Israel, Ministry of Agriculture and Rural Development, Ministry of Regional Cooperation, Larry Kornhauser, and Peter and Naomi Neustadter.

Authors’ contributions

SC and OS conceived the study. YL and OS raised the funds for the study. OS, ST, EA, and SC established and maintaind the core system (i.e., the ATLAS). SC led the fieldwork and data collection, with assistance from OS. SC and TS analyzed the data and created the figures with the involvement of OS and EA. SC and TS and OS wrote the manuscript with critical inputs and feedback from all other co-authors, who also confirmed final version.

Acknowledgments

Assaf Uzan, Michal Handel, Nadav Amir, Shani Cain and Yohay Wasserlauf helped with fieldwork and other components of data collection. We are also grateful to many people and institutes who permit us to use their towers and power supply for our base stations. These include Eli Chupa and Pablo Uner (Harod Valley Water Supply Co-op LTD), Shmuel Wirzberger (Kibbutz Beit Hashita), Eytan Ofer (Kibbutz Sde Eliyahu), Yitzhak Rivlin, Yado Shalev and Elon Shaked (Kibbutz Ein HaNetziv). Shaldag Refrigeration Storeroom LTD, Avihu Levanon (Partner PHI LTD), Michael Zelmanov (Pelephone Communications Ltd), Asher Nahmani and Rafi Levi (Mekorot), and Eli Cohen and Eli Mimran (Bezeq Inc.), and The Israeli Telecommunication Corp Ltd.

Holyoak M, Casagrandi R, Nathan R, Revilla E, Spiegel O. Trends and missing parts in the study of movement ecology. Proc Natl Acad Sci. 2008;105:19060–5.
Nathan R, Getz WM, Revilla E, Holyoak M, Kadmon R, Saltz D, et al. A movement ecology paradigm for unifying organismal movement research. Proc Natl Acad Sci. 2008;105:19052–9.
Abrahms B, Seidel DP, Dougherty E, Hazen EL, Bograd SJ, Wilson AM, et al. Suite of simple metrics reveals common movement syndromes across vertebrate taxa. Mov Ecol. 2017;5:1–11.
Getz WM, Saltz D. A framework for generating and analyzing movement paths on ecological landscapes. Proc Natl Acad Sci. 2008;105:19066–71.
Joo R, Picardi S, Boone ME, Clay TA, Patrick SC, Romero-Romero VS, et al. Recent trends in movement ecology of animals and human mobility. Mov Ecol. 2022;10:1–20.
Spiegel O, Harel R, Getz WM, Nathan R. Mixed strategies of griffon vultures’ (Gyps fulvus) response to food deprivation lead to a hump-shaped movement pattern. Mov Ecol. 2013;1:1–12.
McIntyre NE, Wiens JA. Interactions between landscape structure and animal behavior: the roles of heterogeneously distributed resources and food deprivation on movement patterns. Landsc Ecol. 1999;14:437–47.
Cartar R V., Real LA. Habitat structure and animal movement: the behaviour of bumble bees in uniform and random spatial resource distributions. Oecologia. 1997;112:430–4.
Oudman T, Piersma T, Ahmedou Salem M V., Feis ME, Dekinga A, Holthuijsen S, et al. Resource landscapes explain contrasting patterns of aggregation and site fidelity by red knots at two wintering sites. Mov Ecol. 2018;6:1–12.
Spiegel O, Leu ST, Bull CM, Sih A. What’s your move? Movement as a link between personality and spatial dynamics in animal populations. Ecol Lett. 2017;20:3–18.
Nilsson J-Å, Brönmark C, Hansson L-A, Chapman BB. Individuality in movement: the role of animal personality. In: Hansson L-A, Åkesson S, editors. Animal movement across scales. Oxford, UK: Oxford University Press; 2014. p. 90–190.
Shaw AK. Causes and consequences of individual variation in animal movement. Mov Ecol. 2020;8:1–12.
Sih A, Bell A, Johnson JC. Behavioral syndromes: an ecological and evolutionary overview. Trends Ecol Evol. 2004;19:372–8.
Réale D, Reader SM, Sol D, McDougall PT, Dingemanse NJ. Integrating animal temperament within ecology and evolution. Biol Rev. 2007;82:291–318.
Stuber EF, Carlson BS, Jesmer BR. Spatial personalities: A meta-analysis of consistent individual differences in spatial behavior. Behav Ecol. 2022;33:477–86.
Payne E, Spiegel O, Sinn DL, Leu ST, Gardner MG, Godfrey SS, et al. Intrinsic traits, social context, and local environment shape home range size and fidelity of sleepy lizards. Ecol Monogr. 2022;e1519.
Harrison PM, Gutowsky LFG, Martins EG, Patterson DA, Cooke SJ, Power M. Personality-dependent spatial ecology occurs independently from dispersal in wild burbot (Lota lota). Behav Ecol. 2015;26:483–92.
Picardi S, Ranc N, Smith BJ, Coates PS, Mathews SR, Dahlgren DK. Individual variation in temporal dynamics of post-release habitat selection. Front Conserv Sci. 2021;2.
Trevail AM, Green JA, Bolton M, Daunt F, Harris SM, Miller PI, et al. Environmental heterogeneity promotes individual specialisation in habitat selection in a widely distributed seabird. J Anim Ecol 2021;90:2875–87.
Hertel AG, Leclerc M, Warren D, Pelletier F, Zedrosser A, Mueller T. Don’t poke the bear: using tracking data to quantify behavioural syndromes in elusive wildlife. Anim Behav. 2019;147:91–104.
Michelangeli M, Payne E, Spiegel O, Sinn DL, Leu ST, Gardner MG, et al. Personality, spatiotemporal ecological variation and resident/explorer movement syndromes in the sleepy lizard. J Anim Ecol 2022;91:210–23.
Nathan R, Monk CT, Arlinghaus R, Adam T, Alós J, Assaf M, et al. Big-data approaches lead to an increased understanding of the ecology of animal movement. Science. 2022;375:eabg1780.
Milles A, Dammhahn M, Grimm V. Intraspecific trait variation in personality-related movement behavior promotes coexistence. Oikos. 2020;129:1441–54.
Merrick MJ, Koprowski JL. Should we consider individual behavior differences in applied wildlife conservation studies? Biol Conserv. 2017;209:34–44.
Dougherty ER, Seidel DP, Carlson CJ, Spiegel O, Getz WM. Going through the motions: incorporating movement analyses into disease research. Ecol Lett 2018;21:588–604.
Stamps JA, Briffa M, Biro PA. Unpredictable animals: individual differences in intraindividual variability (IIV). Anim Behav. 2012;83:1325–34.
O’Dea RE, Noble DWA, Nakagawa S. Unifying individual differences in personality, predictability and plasticity: a practical guide. Methods Ecol Evol. 2022;13:278–93.
Biro PA, Adriaenssens B. Predictability as a personality trait: consistent differences in intraindividual behavioral variation. Am Nat. 2013;182:621–9.
Westneat DF, Wright J, Dingemanse NJ. The biology hidden inside residual within-individual phenotypic variation. Biol Rev. 2015;90:729–43.
Mitchell DJ, Fanson BG, Beckmann C, Biro PA. Towards powerful experimental and statistical approaches to study intraindividual variability in labile traits. R Soc Open Sci. 2016;3:160352.
Kok EM., Burant JB, Dekinga A, Manche P, Saintonge D, Piersma T, et al. Within-individual canalization contributes to age-related increases in trait repeatability: a longitudinal experiment in red knots. Am Nat. 2019;194:455–69.
Sinn DL, Gosling SD, Moltschaniwskyj NA. Development of shy/bold behaviour in squid: context-specific phenotypes associated with developmental plasticity. Anim Behav. 2008;75:433–42.
Jennings DJ, Hayden TJ, Gammell MP. Personality and predictability in fallow deer fighting behaviour: the relationship with mating success. Anim Behav. 2013;86:1041–7.
Luisa Vissat L, Cain S, Nathan R, Toledo S, Spiegel O, Getz WM. Categorizing animal diel movement patterns with examples from high-resolution barn owl tracking. 2022. submitted to this issue.
Cleasby IR, Nakagawa S, Schielzeth H. Quantifying the predictability of behaviour: statistical approaches for the study of between-individual variation in the within-individual variance. Methods Ecol Evol. 2015;6:27–37.
Hertel AG, Niemelä PT, Dingemanse NJ, Mueller T. A guide for studying among-individual behavioral variation from movement data in the wild. Mov Ecol. 2020;8:1–18.
Hertel AG, Royauté R, Zedrosser A, Mueller T. Biologging reveals individual variation in behavioural predictability in the wild. J Anim Ecol. 2021;90:723–37.
Roulin A. Barn owls: Evolution and ecology. Cambridge University Press; 2020.
Meyrom K, Motro Y, Leshem Y, Aviel S, Izhaki I, Argyle F, et al. Nest-Box use by the Barn Owl Tyto alba in a Biological Pest Control Program in the Beit She’an Valley, Israel. Ardea. 2009;97:463–7.
Patrick SC, Weimerskirch H. Personality, foraging and fitness consequences in a long lived seabird. PLoS One. 2014;9:e87269.
Favati A, Zidar J, Thorpe H, Jensen P, Løvlie H. The ontogeny of personality traits in the red junglefowl, Gallus gallus. Behav Ecol. 2016;27:484–93.
Charter M, Izhaki I, Leshem Y, Meyrom K, Roulin A. Relationship between diet and reproductive success in the Israeli barn owl. J Arid Environ. 2015;122:59–63.
Pickwell G. Barn owl growth and behaviorisms. Auk. 1948;65:359–73.
Almasi B, Massa C, Jenni L, Roulin A. Exogenous corticosterone and melanin-based coloration explain variation in juvenile dispersal behaviour in the barn owl (Tyto alba). PLoS One. 2021;16:e0256038.
Toledo S, Shohami D, Schiffner I, Lourie E, Orchan Y, Bartan Y, et al. Cognitive map-based navigation in wild bats revealed by a new high-throughput tracking system. Science. 2020;369:188–93.
Toledo S, Mendel S, Levi A, Vortman Y, Ullmann W, Scherer L-R, et al. Vildehaye: a family of versatile, widely-applicable, and field-proven lightweight wildlife tracking and sensing tags. Proc ACM/IEEE Int Conf Inf Process Sens Networks. 2022.
Weller-Weiser A, Orchan Y, Nathan R, Charter M, Weiss AJ, Toledo S. Characterizing the accuracy of a self-synchronized reverse-GPS wildlife localization system. 15th ACM/IEEE Int Conf Inf Process Sens Networks,. 2016. p. 1–12.
Beardsworth CE, Gobbens E, Maarseveen F van, Denissen B, Dekinga A, Nathan R, et al. Validating a high-throughput tracking system: ATLAS as a regional-scale alternative to GPS. bioRxiv. 2021.
R Core Team. R: a language and environment for statistical computing. R Found Stat Comput Vienna, Austria. 2013.
RStudio Team. RStudio: Integrated Development Environment for R. Boston, RStudio Inc.; 2021.
Gupte PR, Beardsworth CE, Spiegel O, Lourie E, Toledo S, Nathan R, et al. A guide to pre-processing high-throughput animal tracking data. J Anim Ecol. 2022;91:287–307.
Roulin A. The function of food stores in bird nests: observations and experiments in the barn owl Tyto alba. Ardea. 2004;92:69–78.
Nakagawa S, Schielzeth H. Repeatability for Gaussian and non-Gaussian data: a practical guide for biologists. Biol Rev. 2010;85:935–56.
Bürkner PC. brms: an R package for Bayesian multilevel models using stan. J Stat Softw 2017;80:1–28.
Fleming CH, Fagan WF, Mueller T, Olson KA, Leimgruber P, Calabrese JM, et al. Rigorous home range estimation with movement data: a new autocorrelated kernel density estimator. Ecology. 2015;96:1182–8.
Calabrese JM, Fleming CH, Gurarie E. ctmm: an r package for analyzing animal relocation data as a continuous-time stochastic process. Methods Ecol Evol. 2016;7:1124–32.
Bates D, Mächler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1–48.
Mazerolle MJ. AICcmodavg: Model selection and multimodel inference based on (Q)AIC(c). R package version 2.3-1. 2020.
Fox J, Weisberg S. An R companion to applied regression, 3rd edition. Thousand Oaks CA: Sage; 2019.
Cox DR. Regression Models and Life-Tables. J R Stat Soc Ser B. 1972;34:187–202.
Fox J, Weisberg S. Cox proportional-hazards regression for survival data. An R and S-Plus companion to applied regression. In: Fox J, editor. An R S-PLUS companion to applied regression. Thousand Oaks, CA: Sage Publications Inc.; 2002.
Borda-de-Água L, Grilo C, Pereira HM. Modeling the impact of road mortality on barn owl (Tyto alba) populations using age-structured models. Ecol Modell. 2014;276:29–37.
Boves TJ, Belthoff JR. Roadway mortality of barn owls in Idaho, USA. J Wildl Manage. 2012;76:1381–92.
Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. Springer, New York; 2000.
Therneau TM. A package for survival analysis in R. R package version 3.3-1. 2022.
Kassambara A, Kosinski M, Biecek P, Fabian S. Package ‘survminer’. 2017.
Therneau TM. Package ‘coxme’. Mixed effects cox models. R package version 2.2–10. 2018.
Barton K. Package ‘MuMIn’. Model selection and model averaging based on information criteria. R package version 1.15.11. 2015.
Altwegg R, Schaub M, Roulin A. Age-specific fitness components and their temporal variation in the barn owl. Am Nat. 2007;169:47–61.
Bell AM, Hankison SJ, Laskowski KL. The repeatability of behaviour: a meta-analysis. Anim Behav. 2009;77:771–83.
Dingemanse NJ, Kazem AJN, Réale D, Wright J. Behavioural reaction norms: animal personality meets individual plasticity. Trends Ecol Evol. 2010;25:81–9.
Spiegel O, Pinter-Wollman N. Dissecting how behavior and environment shape spatial personalities: a comment on Stuber et al. Behav Ecol. 2022;33:489–91.
Noonan MJ, Fleming CH, Akre TS, Drescher-Lehman J, Gurarie E, Harrison AL, et al. Scale-insensitive estimation of speed and distance traveled from animal tracking data. Mov Ecol. 2019;7:1–15.
Sih A, Cote J, Evans M, Fogarty S, Pruitt J. Ecological implications of behavioural syndromes. Ecol Lett. 2012;15:278–89.
Dall SRX, Bell AM, Bolnick DI, Ratnieks FLW. An evolutionary ecology of individual differences. Ecol Lett. 2012;15:1189–98.
Minderman J, Reid JM, Hughes M, Denny MJH, Hogg S, Evans PGH, et al. Novel environment exploration and home range size in starlings Sturnus vulgaris. Behav Ecol. 2010;21:1321–9.
Spiegel O, Leu ST, Sih A, Godfrey SS, Bull CM. When the going gets tough: Behavioural type-dependent space use in the sleepy lizard changes as the season dries. Proc R Soc B Biol Sci. 2015;282.
Fieberg J, Börger L. Could you please phrase “home range” as a question? J Mammal. 2012;93:890–902.
Börger L, Dalziel BD, Fryxell JM. Are there general mechanisms of animal home range behaviour? A review and prospects for future research. Ecol Lett. 2018;11:637–50.
Nunn CL, Thrall PH, Leendertz FH, Boesch C. The spread of fecally transmitted parasites in socially-structured populations. PLoS One. 2011;6:e21677.
Wilson-Aggarwal JK, Goodwin CED, Moundai T, Sidouin MK, Swan GJF, Monique Léchenne, et al. Spatial and temporal dynamics of space use by free-ranging domestic dogs Canis familiaris in rural Africa. Ecol Appl. 2021;31:e02328.
Moiron M, Laskowski KL, Niemelä PT. Individual differences in behaviour explain variation in survival: a meta-analysis. Ecol Lett. 2020;23:399–408.
Harris SM, Descamps S, Sneddon LU, Bertrand P, Chastel O, Patrick SC. Personality predicts foraging site fidelity and trip repeatability in a marine predator. J Anim Ecol. 2020;89:68–79.
Spiegel O, Crofoot MC. The feedback between where we go and what we know — information shapes movement, but movement also impacts information acquisition. Curr Opin Behav Sci. 2016;12:90–6.
Toscano BJ, Gownaris NJ, Heerhartz SM, Monaco CJ. Personality, foraging behavior and specialization: integrating behavioral and food web ecology at the individual level. Oecologia. 2016;182:55–69.
Webber QMR, Laforge MP, Bonar M, Robitaille AL, Hart C, Zabihi-Seissan S, et al. The ecology of individual differences empirically applied to space-use and movement tactics. Am Nat. 2020;196:E1–15.
Fryxell JM, Hazell M, Börger L, Dalziel BD, Haydon DT, Morales JM, et al. Multiple movement modes by large herbivores at multiple spatiotemporal scales. Proc Natl Acad Sci. 2008;105:19114–9.
Spiegel O, Harel R, Centeno-Cuadros A, Hatzofe O, Getz WM, Nathan R. Moving beyond curve fitting: Using complementary data to assess alternative explanations for long movements of three vulture species. Am Nat. 2015;185:E44–54.
Leclerc M, Vander Wal E, Zedrosser A, Swenson JE, Kindberg J, Pelletier F. Quantifying consistent individual differences in habitat selection. Oecologia. 2016;180:697–705.
Payne E, Sinn DL, Spiegel O, Leu ST, Wohlfeil C, Godfrey SS, et al. Consistent individual differences in ecto-parasitism of a long-lived lizard host. Oikos. 2020;129:1061–71.
Cote J, Clobert J, Brodin T, Fogarty S, Sih A. Personality-dependent dispersal: Characterization, ontogeny and consequences for spatially structured populations. Philos Trans R Soc B Biol Sci. 2010;365:4065–76.

No competing interests reported.

Appendix.docx

Download PDF

Journal Publication

published 07 Feb, 2023

Read the published version in Movement Ecology →

Editorial decision: Major revision
09 Aug, 2022
Reviews received at journal
19 Jul, 2022
Reviewers agreed at journal
07 Jul, 2022
Reviewers agreed at journal
06 Jul, 2022
Reviewers invited by journal
05 Jul, 2022
Editor assigned by journal
26 Jun, 2022
Submission checks completed at journal
26 Jun, 2022
First submitted to journal
20 Jun, 2022

You are reading this latest preprint version

Movement predictability of individual barn owls facilitates home range size and survival estimation

Status:

Journal Publication

Version 1

Abstract

Background.

Methods.

Results.

Conclusion.

Figures

1. Introduction

2. Methods

2.1 Study species

2.2 Study system

2.3 The ATLAS tracking system

2.4 Data analysis and statistical methods

3. Results

3.1 Individual variation in barn owl movement

3.2 The ecological consequences of predictability in movement

3.3 Factors affecting individual predictability

4. Discussion

4.1 Spatial behavioral types (BTs) and movement predictability

4.2 The ecological importance of movement predictability

4.3 Factors affecting predictability

4.4 Concluding remarks

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1