Kamran Safi

Fuel economy matters to both animals and machines, as the cost of transport can determine where, when, and how far locomotion is possible. However, established estimates of these values for animal flight come from laboratories and treat the atmosphere as constant, even though free-flying animals routinely extract energy from atmospheric flows. Using high-resolution movement data, we estimate the cost of transport for 50 bird and bat species flying in the wild. Our results revise the classic ranking of locomotion costs, suggesting that flying can be cheaper than swimming though almost as costly as running, depending on interactions between physical environment, morphology, and behavior. Linking transport costs to atmospheric conditions provides a mechanistic basis for predicting how changing windscapes will reshape the energetics, movements, and distributions of flying animals.

Many animals must gain independence from parental care by acquiring skills necessary for survival. Thus, the rate of skill acquisition can predict when parental care ends and animals emigrate. However, young animals sometimes acquire adult-like skills long before they emigrate. We asked why the timing of emigration decouples from the rate of skill acquisition and proposed that skill acquisition may instead link to the timing of excursions from natal territories before emigration. We used global positioning system tracking of 14–64 juvenile golden eagles (Aquila chrysaetos), which learn to fly and hunt under parental care. We expected flight performance to predict the timing of excursions better than the timing of emigration. Using time-to-event models, we predicted excursion and emigration rates based on flight performance and patterns of daily movement. Contrary to our prediction, flight performance failed to predict either event, whereas daily movement predicted both. Daily movement may capture motivation and weather together with flight skills, and those factors probably drive emigration. Juveniles may delay emigration not because they lack skill, but because the costs of independence outweigh its benefits. This decoupling of ability and choice may also obscure the link between skill acquisition and the end of parental care in other systems.

Understanding how animals respond to their physical environment requires environmental observations at the scale at which behavioural decisions are made. For soaring birds, the coarse resolution of weather products has long hindered the analysis of their behavioural response to fine-scale atmospheric dynamics, forcing uplift sources to be inferred largely from behaviour itself. Here, we combined high-resolution movement data from 24 golden eagles with the kilometre-scale COSMO weather model. We first classified thermal, orographic, and gravity-wave uplifts using independent atmospheric predictors and then quantified the birds' use of each uplift type and their fine-scale behavioural responses. Eagles relied predominantly on thermals, but opportunistically adjusted their use of uplift sources seasonally. The birds' flight behaviour could not reliably indicate which uplift type was primarily used, and thus suggests that both atmospheric processes and behavioural responses are better described as continua than discrete categories. Finally, we compared vertical wind velocities derived from eagles soaring behaviour with those modelled by the COSMO weather model, showing that most of the thermals exploited by eagles remain unresolved at kilometre-scale model resolution. Our results demonstrate how high-resolution weather models provide new insights into bird movement decisions, while also highlighting the potential of soaring birds as biologically embedded atmospheric sensors that could help closing the resolution gap in atmospheric models.

Individual behavioural variation is a driver of population dynamics and key for the adaptation and persistence of species. In colonial species, individuals share the same ecological niche and aggregate in space, but they might use different foraging strategies and incur differential costs. Using location and accelerometer data from 21 adult oilbirds ( Steatornis caripensis ), which are colonial frugivores, we characterized nocturnal movement patterns and measured total nightly flight time, as a proxy of movement effort, for different movement strategies during their non-breeding period. We collected location and accelerometer data using 22 g loggers (e-obs GmbH; Munich, Germany), between August 22nd, 2008 and February 14th, 2009. The tracked oilbirds displayed three distinct movement strategies for the entire sampling period: Intra-tropical migrants (ITM), which performed roost switches that lasted between 17 and 73 continuous days, consistent with partial nomadic migrations rather than seasonal movements; venturers, which performed long-distance excursions but returned to the main cave within 3 days, and roost-faithful individuals that never performed such excursions and never switched roosts. Among the three strategies, we estimated that ITM showed a lower movement effort, based on total nightly flight time, throughout the sampling period. Despite travelling long distances, ITM birds reduced overall flight effort by day-roosting near foraging sites, minimizing the need for long commuting flights. The oilbirds showed site fidelity to multiple foraging areas, without among-individual spatial overlap. These findings give insights into how behavioural flexibility could facilitate resource partitioning and coexistence in this colonial species, as well as the consequences of inter-individual variation for adapting to habitat change. Such information is relevant to evaluate the role of oilbirds as ecosystem service providers and habitat connectivity.

Erroneous locations are ubiquitous in animal tracking, propagating into esti- mates of movement rate, home range, habitat selection and behavioural state. Existing filters often need a threshold chosen per species and schedule, assume a movement process heterogeneous data rarely satisfy, or stop at a plot for manual filtering. A block of consecutive wrong positions, caused for example by GNSS spoofing, whose interior looks ordinary step by step, is caught only by one supervised filter. I introduce mt_clean_track(), an unsupervised cleaning cascade in the R package move2utils. Four detectors score every location, each for an error class the others cannot: a Brownian-bridge geometric residual, a movement- probability score, a path-versus-displacement detour ratio, and a speed cap. A fifth stage, not a detector, identifies a block of false positions at a track’s start or end, where no per-location test can tell which side is the artefact. Every threshold bar the detour ratio’s is read from the track’s own scores, and a location is flagged only on corroborated evidence, so the composition controls the false-positive rate. I evaluate the cascade on synthetic trajectories with known truth, spanning spikes, spoofing blocks, colony halos and multi-state anomalies, against four alternatives (a naïve speed cap, atlastools, trip::sda, SDLfilter). Each alternative is given the same speed ceiling, calibrated on the cohort’s uncon- taminated track; the cascade, nothing but the data. Even so, it recovered all 141 injected outliers on every contaminated track, whereas they left 27 to 33 behind at recalls of 0.766–0.809. No alternative scored better on recall and precision together; where it ranked below one, the cost was -precision, never a retained outlier. On a coherent spoofing block it recovered 30 of 30 injected locations against their 0 to 3; on a clean track with 30 m positional error it flagged one location across 100 realisations. Unsupervised operation, data-driven thresholds and a read-out of what trig- gered each flag make outlier screening routine and reproducible instead of be- spoke. The cascade screens rather than models, and the functions beneath it stay open to further detectors and to new consensus-finding rules.

Studying animal movement at the population scale requires a stable, modern software substrate. Within R, the legacy move package supplied that substrate for over a decade, but its sp / rgeos backbone has been retired. The successor package move2 deliberately confined its scope to the data class and core movebank API functions. The analytical machinery of move , namely dynamic Brownian-bridge utilisation distributions, the directional bivariate-Gaussian variant, corridor segmentation, and along-track thinning, was left to port to the modern sf / terra stack. We present move2utils , an R package that completes and complements that transition. move2utils provides move2 -native ports of the move analytical functions, preserves the original C kernels where they exist, and replaces the deprecated spatial scaffolding around them. It additionally ports some of the legacy R code to faster C kernels. move2utils also exposes novel outlier-detection methodology described in detail in a companion paper. The package is open-source (GPL ≥ 3), is developed on the MPCDF GitLab and mirrored on GitHub for public installation, and ships with vignettes and a CI-tested check suite. We illustrate it with a worked example on real tracking data and synthetic datasets.

Early-life conditions are often associated with fitness of animals later in life. Although recent studies show that individual characteristics during development affect the formation of early behavioural phenotypes, we have a limited understanding of the cascade of life-history transitions from early phenotypes to fitness-relevant behaviours in later life stages. Here, we used GPS and body-acceleration data of 35 juvenile golden eagles (Aquila chrysaetos) to investigate factors associated with nestling activity levels (overall dynamic body acceleration; ODBA), the relationship between nestling activity and the timing of fledging and post-fledging activity, as well as the effects of post-fledging activity on post-fledging movements and the timing of dispersal. We found that brood size, but no other intrinsic or extrinsic factors, was associated with activity levels. Active nestlings fledged earlier and remained more active after fledging than passive nestlings. Increased post-fledging activity levels were related to an increased number of pre-dispersal forays and an early timing of dispersal. Even though some effects showed reduced certainty, our results provide evidence for downstream effects of behavioural differences expressed in the nest. They suggest that early activity phenotypes of nestlings drive the timing of subsequent life-history transitions, thereby representing a key mechanism that links behavioural phenotypes in early life with future performance.

Scavengers generally rely on patchily distributed, unpredictable carrion. A long-standing hypothesis suggests scavenging ravens reliably locate such food by directly following large carnivores to their kills. However, by satellite tracking 69 ravens, 20 wolves, and 11 cougars in Yellowstone National Park, we found that following of predators over large distances rarely occurred. Instead, ravens routinely revisited sites where wolf kills were common—returning from distances of up to 155 kilometers to find carrion. Much like navigating to permanent anthropogenic subsidies, ravens appear to remember potential sources of carrion shaped by previous encounters with wolves or their kills. These findings suggest that spatial memory and navigation play a considerably greater role than previously assumed among scavengers, and possibly other wide-ranging species, in search of ephemeral resources.

The expanding human footprint drives profound transformations in species, including not only morphological trait loss but also behavioural trait extinction. In contrast to behavioural trait extinction, which entails a reduction in genetic variability, behavioural plasticity maintains behavioural trait diversity and the associated genetic variability required for evolutionary adaptation. Yet, while behavioural plasticity shaping non-human capacity to coexist with humans has evolved under past, human-modified, environmental conditions, few studies have examined whether species retain a level of behavioural plasticity after past interactions with humans. Using data from 64 juveniles golden eagles ( Aquila chrysaetos ) from a population demographically recovering after centuries of persecution, yet still restricted to less human-dominated alpine massifs, we tested whether human avoidance is innate or learned through experience during dispersal or parental influences during the pre-dispersal period. We performed a Principal Component Analysis (PCA) on landscape features used during the pre-dispersal and the first fifteen weeks of the dispersal phase, and used Mahalanobis distance as an indicator for quantifying changes in individual spatial behaviour over time. The lower frequency of use of human-dominated landscape features, compared with the more sparsely populated mid- and high-elevation landscape features, resulted from individuals’ consistent use of similar landscape features across the pre-dispersal and dispersal periods. Avoidance of humans was therefore better explained by a learning or imprinting process prior to dispersal, which predisposed individuals to use landscape features similar to those experienced during pre-dispersal, rather than by learning during dispersal, or by an innate avoidance. Imprinting on natal habitat features corroborates the observed limited range expansion of the alpine population beyond the massif borders. Differences among European populations in range recolonisation suggest contrasting histories of persecution, which may have left varying degrees of behavioural plasticity across populations. As a guide for future conservation priorities, these findings underscore the need to reconstruct past conditions and/or to document contemporary animal behaviour in order to build a library of behaviours that can be used in the future to track behavioural changes.