Search

As human settlement expands farther into previously uninhabited areas, interactions with wild animals are likely to increase. The nature of these interactions can be detrimental to humans and animals alike. We focus on the relationship between urban areas and bears, and the consequences of a bear's dietary choices. Using an agent-based model we investigated the effects of educating humans about waste management and bear deterrence methods on the number of bears that enter urban areas repeatedly. Variables tested included the percentage of the landscape that is urban, probability of deterrence, percentage of the human population educated about bear safe behaviours, types of bear management strategies (BMSs) implemented in educated urban areas, and the bear management spatial configurations (BMCs). The results indicate that all education methods reduce the number of human–bear conflicts. For each percent of the population that is taught, there is a 5% decrease in the probability that a bear becomes a conflict bear. We also found that the existing residential spatial configuration and the bear management strategies to be implemented are important considerations when creating an education program. Our results suggest that agent-based models can be used to identify viable management strategies and to determine the most effective human education program (BMS and BMC) when trying to reduce the number of conflict bears.

12 a priori models used to explain moose grouping behaviour in Sweden from aerial survey data collected in 2006, 2009, 2011. Models are shown in order of decreasing rank with model log-likelihood (-logLik), Akaike’s information criterion (AIC), AIC differences (∆i) and AIC weights (ωi) and deviance. (DOCX 14 kb)

Background Predation risk is a primary motivator for prey to congregate in larger groups. A large group can be beneficial to detect predators, share predation risk among individuals and cause confusion for an attacking predator. However, forming large groups also has disadvantages like higher detection and attack rates of predators or interspecific competition. With the current recolonization of wolves (Canis lupus) in Scandinavia, we studied whether moose (Alces alces) respond by changing grouping behaviour as an anti-predatory strategy and that this change should be related to the duration of wolf presence within the local moose population. In particular, as females with calves are most vulnerable to predation risk, they should be more likely to alter behaviour. Methods To study grouping behaviour, we used aerial observations of moose (n = 1335, where each observation included one or several moose) inside and outside wolf territories. Results Moose mostly stayed solitary or in small groups (82% of the observations consisted of less than three adult moose), and this behavior was independent of wolf presence. The results did not provide unequivocal support for our main hypothesis of an overall change in grouping behaviour in the moose population in response to wolf presence. Other variables such as moose density, snow depth and adult sex ratio of the group were overall more influential on grouping behaviour. However, the results showed a sex specific difference in social grouping in relation to wolf presence where males tended to form larger groups inside as compared to outside wolf territories. For male moose, population- and environmentally related variables were also important for the pattern of grouping. Conclusions The results did not give support for that wolf recolonization has resulted in an overall change in moose grouping behaviour. If indeed wolf-induced effects do exist, they may be difficult to discern because the effects from moose population and environmental factors may be stronger than any change in anti-predator behaviour. Our results thereby suggest that caution should be taken as to generalize about the effects of returning predators on the grouping behaviour of their prey. Electronic supplementary material The online version of this article (doi:10.1186/s12983-017-0195-z) contains supplementary material, which is available to authorized users.

Conditional logistic regression (CLR) is widely used to analyze habitat selection and movement of animals when resource availability changes over space and time. Observations used for these analyses are typically autocorrelated, which biases model-based variance estimation of CLR parameters. This bias can be corrected using generalized estimating equations (GEE), an approach that requires partitioning the data into independent clusters. Here we establish the link between clustering rules in GEE and their effectiveness to remove statistical biases in variance estimation of CLR parameters. The current lack of guidelines is such that broad variation in clustering rules can be found among studies (e.g., 14-450 clusters) with unknown consequences on the robustness of statistical inference. We simulated datasets reflecting conditions typical of field studies. Longitudinal data were generated based on several parameters of habitat selection with varying strength of autocorrelation and some individuals having more observations than others. We then evaluated how changing the number of clusters impacted the effectiveness of variance estimators. Simulations revealed that 30 clusters were sufficient to get unbiased and relatively precise estimates of variance of parameter estimates. The use of destructive sampling to increase the number of independent clusters was successful at removing statistical bias, but only when observations were temporally autocorrelated and the strength of inter-individual heterogeneity was weak. GEE also provided robust estimates of variance for different magnitudes of unbalanced datasets. Our simulations demonstrate that GEE should be estimated by assigning each individual to a cluster when at least 30 animals are followed, or by using destructive sampling for studies with fewer individuals having intermediate level of behavioural plasticity in selection and temporally autocorrelated observations. The simulations provide valuable information to build reliable habitat selection and movement models that allow for robustness of statistical inference without removing excessive amounts of ecological information.