Highlights d Ungulates moved to track forage in landscapes with wavelike spring green-up d Patterns of green-up explained where migratory behavior occurred in many ecosystems d At the species level, migrants and residents received equivalent foraging benefits d Movement tactics represent behavioral adaptations to specific landscapes

The most common framework under which ungulate migration is studied predicts that it is driven by spatio–temporal variation in plant phenology, yet other hypotheses may explain differences within and between species. To disentangle more complex patterns than those based on single species/ single populations, we quantified migration variability using two sympatric ungulate species differing in their foraging strategy, mating system and physiological constraints due to body size. We related observed variation to a set of hypotheses. We used GPS‐collar data from 537 individuals in 10 roe Capreolus capreolus and 12 red deer Cervus elaphus populations spanning environmental gradients across Europe to assess variation in migration propensity, distance and timing. Using time‐to‐event models, we explored how the probability of migration varied in relation to sex, landscape (e.g. topography, forest cover) and temporally‐varying environmental factors (e.g. plant green‐up, snow cover). Migration propensity varied across study areas. Red deer were, on average, three times more migratory than roe deer (56% versus 18%). This relationship was mainly driven by red deer males which were twice as migratory as females (82% versus 38%). The probability of roe deer migration was similar between sexes. Roe deer (both sexes) migrated earliest in spring. While territorial male roe deer migrated last in autumn, male and female red deer migrated around the same time in autumn, likely due to their polygynous mating system. Plant productivity determined the onset of spring migration in both species, but if plant productivity on winter ranges was sufficiently high, roe deer were less likely to leave. In autumn, migration coincided with reduced plant productivity for both species. This relationship was stronger for red deer. Our results confirm that ungulate migration is influenced by plant phenology, but in a novel way, that these effects appear to be modulated by species‐specific traits, especially mating strategies.

Humans, as super predators, can have strong effects on wildlife behaviour, including profound modifications of diel activity patterns. Subsequent to the return of large carnivores to human‐modified ecosystems, many prey species have adjusted their spatial behaviour to the contrasting landscapes of fear generated by both their natural predators and anthropogenic pressures. The effects of predation risk on temporal shifts in diel activity of prey, however, remain largely unexplored in human‐dominated landscapes. We investigated the influence of the density of lynx Lynx lynx, a nocturnal predator, on the diel activity patterns of their main prey, the roe deer Capreolus capreolus, across a gradient of human disturbance and hunting at the European scale. Based on 11 million activity records from 431 individually GPS‐monitored roe deer in 12 populations within the EURODEER network (http://eurodeer.org), we investigated how lynx predation risk in combination with both lethal and non‐lethal human activities affected the diurnality of deer. We demonstrated marked plasticity in roe deer diel activity patterns in response to spatio‐temporal variations in risk, mostly due to human activities. In particular, roe deer decreased their level of diurnality by a factor of 1.37 when the background level of general human disturbance was high. Hunting exacerbated this effect, as during the hunting season deer switched most of their activity to night‐time and, to a lesser extent, to dawn, although this pattern varied noticeably in relation to lynx density. Indeed, in the presence of lynx, their main natural predator, roe deer were relatively more diurnal. Overall, our results revealed a strong influence of human activities and the presence of lynx on diel shifts in roe deer activity. In the context of the recovery of large carnivores across Europe, we provide important insights about the effects of predators on the behavioural responses of their prey in human‐dominated ecosystems. Modifications in the temporal partitioning of ungulate activity as a response to human activities may facilitate human–wildlife coexistence, but likely also have knock‐on effects for predator–prey interactions, with cascading effects on ecosystem functioning.

Background: Biotelemetry offers an increasing set of tools to monitor animals. Acceleration sensors in particular can provide remote observations of animal behavior at high temporal resolution. While recent studies have demonstrated the capability of this technique for a wide range of species and behaviors, a coherent methodology is still missing (1) for behavior monitoring of large herbivores that are usually tagged with neck collars and frequently switch between diverse behaviors and (2) for monitoring of vigilance behavior. Here, we present an approach that aims at remotely monitoring different types of large herbivore behavior including vigilance with acceleration data. Methods:We pioneered this approach with field observations of eight collared roe deer (Capreolus capreolus). First, we trained a classification model for distinguishing seven structural behavior categories: lying, standing, browsing, walking, trotting, galloping and 'others' . Second, we developed a model that predicted the internal states, active and resting, based on the predicted sequence of structural behaviors and expert-based rules. Further, we applied both models to automatically monitor vigilance behavior and compared model predictions with expert judgment of vigilance behavior. To exemplify the practical application of this approach, we predicted behavior, internal state and vigilance continuously for a collared roe deer. Results:The structural behaviors were predicted with high accuracy (overall cross-validated accuracy 71%). Only behaviors that are similar in terms of posture and dynamic body movements were prone to misclassification. Active and resting states showed clear distinction and could be utilized as behavioral context for the detection of vigilance behavior. Here, model predictions were characterized by excellent consistency with expert judgment of vigilance behavior (mean accuracy 96%). Conclusion:In this study, we demonstrated the strong potential and practical applicability of acceleration data for continuous, high-resolution behavior monitoring of large herbivores and showed that vigilance behavior is well detectable. In particular, when combined with spatial data, automated behavior recognition will enrich many fields in behavioral ecology by providing extensive access to behaviors of animals in the wild.

Background Our understanding of movement patterns and behaviours of wildlife has advanced greatly through the use of improved tracking technologies, including application of accelerometry (ACC) across a wide range of taxa. However, most ACC studies either use intermittent sampling that hinders continuity or continuous data logging relying on tracker retrieval for data downloading which is not applicable for long term study. To allow long-term, fine-scale behavioural research, we evaluated a range of machine learning methods for their suitability for continuous on-board classification of ACC data into behaviour categories prior to data transmission. Methods We tested six supervised machine learning methods, including linear discriminant analysis (LDA), decision tree (DT), support vector machine (SVM), artificial neural network (ANN), random forest (RF) and extreme gradient boosting (XGBoost) to classify behaviour using ACC data from three bird species (white stork Ciconia ciconia, griffon vulture Gyps fulvus and common crane Grus grus) and two mammals (dairy cow Bos taurus and roe deer Capreolus capreolus). Results Using a range of quality criteria, SVM, ANN, RF and XGBoost performed well in determining behaviour from ACC data and their good performance appeared little affected when greatly reducing the number of input features for model training. On-board runtime and storage-requirement tests showed that notably ANN, RF and XGBoost would make suitable on-board classifiers. Conclusions Our identification of using feature reduction in combination with ANN, RF and XGBoost as suitable methods for on-board behavioural classification of continuous ACC data has considerable potential to benefit movement ecology and behavioural research, wildlife conservation and livestock husbandry.