Search

Animal movement arises from a complex interaction between an animal and its heterogeneous environment. The large amount of variables involved in that interaction complicates the ecological inference thereof. In order to better understand movement, the myriad of involved variables can be grouped into behavioural, spatial, and temporal components of the movement process (related to the questions ofwhy, whereandwhento move), which should be addressed simultaneously.We studied the influence of spatial environmental variation on fine-scale movement trajectories of eland antelope, blue wildebeest and plains zebra in a South-African game reserve, in simultaneous consideration with temporal rhythms and behavioural movement modes.We described the animals’ movement process as multi-mode random walks using hidden Markov models (HMMs), where both the probabilities of switching between modes as well as the mode-dependent speed and turning angle distributions were a function of time of day, tree cover and terrain slope.The fitted HMMs clustered the trajectories of all three species into a fast and directed “transit” mode and a slow and tortuous “encamped” mode. All movement properties displayed a strong diel pattern, with the effects of tree cover and slope on speed being largest in the evening when animals generally moved faster. Wildebeest and zebra in their transit mode moved faster with increasing tree cover whereas eland moved slower, but only so on steep terrain. In the encamped mode, tree cover had almost no effect on speed, but a strong effect on tortuosity. The animals also generally moved more tortuously with higher tree cover, even more so on sloping terrain.Due to our integrated approach, we showed that the influence of spatial environmental heterogeneity on animal movement varies greatly in diel cycles and between movement modes, reflecting variation in the animals’ habitat selection and decision making throughout the day. Thus, focusing on the interactions betweenwhy, whenandwhereto move is needed to more truthfully describe and understand animal movement.

Motivation Certain early-succession habitats may emerge only at restricted locations following disturbance. Therefore, whether disturbances tend to occur at certain sites or not can significantly affect habitat availability and metapopulation persistence of early-successional habitat specialists. Available models that combine metapopulation and landscape processes do not address how to model mobile, spatially shifting disturbance intensities independent of factors of site suitability. We present a model that allows the study on how a mobile disturbance pattern, of either natural or anthropogenic origin, affects patch network and metapopulation dynamics in realistic, heterogeneous landscapes. Methods We simulate metapopulation dynamics using a realistic landscape and varying patch destruction (and turnover) rates. We model the local patch emergence rate as the function of site suitability to patch emergence –a permanent factor–and local disturbance intensity, which we first estimate from empirical data and then simulate using annually updating spatial random fields. Using this model, we test whether and how a mobile disturbance pattern affects metapopulation persistence of the false heath fritillary butterfly (Melitaea diamina). Results In our case study, a mobile disturbance pattern caused new patches to emerge further away from occupied patches over time. This decreased the probability of new patches becoming colonized and thus impaired metapopulation persistence even when the median distance between patches appeared unchanging. However, if disturbances moved to areas that were highly suitable to patch emergence, increased habitat availability could compensate the otherwise detrimental effects of a mobile disturbance pattern. Disturbances that had a moderate degree of mobility had the most uncertain effects to metapopulation persistence. Conclusions Our modelling approach distinguishes between two processes behind the spatio-temporal pattern and rates of patch emergence–disturbance dynamics and varying site suitability. It enables the use of social and environmental data for forecasting habitat availability for early-succession habitat specialists under alternative future scenarios. It can be applied and developed further to suit multiple study systems. Our case study suggests that for species conservation, it is either beneficial to organize recurring management activities to take place at constant locations, or to gradually shift them towards areas that are highly suitable to patch emergence.

Wildlife crime is one of the most profitable illegal industries worldwide. Current actions to reduce it are far from effective and fail to prevent population declines of many endangered species, pressing the need for innovative anti-poaching solutions. Here, we propose and test a poacher early warning system that is based on the movement responses of non-targeted sentinel animals, which naturally respond to threats by fleeing and changing herd topology. We analyzed human-evasive movement patterns of 135 mammalian savanna herbivores of four different species, using an internet-of-things architecture with wearable sensors, wireless data transmission and machine learning algorithms. We show that the presence of human intruders can be accurately detected (86.1% accuracy) and localized (less than 500 m error in 54.2% of the experimentally staged intrusions) by algorithmically identifying characteristic changes in sentinel movement. These behavioral signatures include, among others, an increase in movement speed, energy expenditure, body acceleration, directional persistence and herd coherence, and a decrease in suitability of selected habitat. The key to successful identification of these signatures lies in identifying systematic deviations from normal behavior under similar conditions, such as season, time of day and habitat. We also show that the indirect costs of predation are not limited to vigilance, but also include (1) long, high-speed flights; (2) energetically costly flight paths; and (3) suboptimal habitat selection during flights. The combination of wireless biologging, predictive analytics and sentinel animal behavior can benefit wildlife conservation via early poacher detection, but also solve challenges related to surveillance, safety and health.

Background: Animals respond to environmental variation by changing their movement in a multifaceted way. Recent advancements in biologging increasingly allow for detailed measurements of the multifaceted nature of movement, from descriptors of animal movement trajectories (e.g., using GPS) to descriptors of body part movements (e.g., using tri-axial accelerometers). Because this multivariate richness of movement data complicates inference on the environmental influence on animal movement, studies generally use simplified movement descriptors in statistical analyses. However, doing so limits the inference on the environmental influence on movement, as this requires that the multivariate richness of movement data can be fully considered in an analysis.Methods: We propose a data-driven analytic framework, based on existing methods, to quantify the environmental influence on animal movement that can accommodate the multifaceted nature of animal movement. Instead of fitting a simplified movement descriptor to a suite of environmental variables, our proposed framework centres on predicting an environmental variable from the full set of multivariate movement data. The measure of fit of this prediction is taken to be the metric that quantifies how much of the environmental variation relates to the multivariate variation in animal movement. We demonstrate the usefulness of this framework through a case study about the influence of grass availability and time since milking on cow movements using machine learning algorithms.Results: We show that on a one-hour timescale 37% of the variation in grass availability and 33% of time since milking influenced cow movements. Grass availability mostly influenced the cows’ neck movement during grazing, while time since milking mostly influenced the movement through the landscape and the shared variation of accelerometer and GPS data (e.g., activity patterns). Furthermore, this framework proved to be insensitive to spurious correlations between environmental variables in quantifying the influence on animal movement.Conclusions: Not only is our proposed framework well-suited to study the environmental influence on animal movement; we argue that it can also be applied in any field that uses multivariate biologging data, e.g., animal physiology, to study the relationships between animals and their environment.

Animal population sizes are often estimated using aerial sample counts by human observers, both for wildlife and livestock. The associated methods of counting remained more or less the same since the 1970s, but suffer from low precision and low accuracy of population estimates. Aerial counts using cost-efficient Unmanned Aerial Vehicles or microlight aircrafts with cameras and an automated animal detection algorithm can potentially improve this precision and accuracy. Therefore, we evaluated the performance of the multi-class convolutional neural network RetinaNet in detecting elephants, giraffes and zebras in aerial images from two Kenyan animal counts. The algorithm detected 95% of the number of elephants, 91% of giraffes and 90% of zebras that were found by four layers of human annotation, of which it correctly detected an extra 2.8% of elephants, 3.8% giraffes and 4.0% zebras that were missed by all humans, while detecting only 1.6 to 5.0 false positives per true positive. Furthermore, the animal detections by the algorithm were less sensitive to the sighting distance than humans were. With such a high recall and precision, we posit it is feasible to replace manual aerial animal count methods (from images and/or directly) by only the manual identification of image bounding boxes selected by the algorithm and then use a correction factor equal to the inverse of the undercounting bias in the calculation of the population estimates. This correction factor causes the standard error of the population estimate to increase slightly compared to a manual method, but this increase can be compensated for when the sampling effort would increase by 23%. However, an increase in sampling effort of 160% to 1,050% can be attained with the same expenses for equipment and personnel using our proposed semi-automatic method compared to a manual method. Therefore, we conclude that our proposed aerial count method will improve the accuracy of population estimates and will decrease the standard error of population estimates by 31% to 67%. Most importantly, this animal detection algorithm has the potential to outperform humans in detecting animals from the air when supplied with images taken at a fixed rate.

Background: Animals respond to environmental variation by changing their movement in a multifaceted way. Recent advancements in biologging increasingly allow for detailed measurements of the multifaceted nature of movement, from descriptors of animal movement trajectories (e.g., using GPS) to descriptors of body part movements (e.g., using tri-axial accelerometers). Because this multivariate richness of movement data complicates inference on the environmental contribution to animal movement, studies generally use simplified movement descriptors in statistical analyses. However, doing so limits the inference on the environmental contribution to movement, as this requires that the multivariate richness of movement data can be fully considered in an analysis. Methods: We propose a data-driven analytic framework to quantify the environmental contribution to animal movement that can accommodate the multifaceted nature of animal movement. Instead of fitting the response of a simplified movement descriptor to a suite of environmental variables, our proposed framework centres on predicting an environmental variable from the full set of multivariate movement data, i.e., the reverse of the route of causal inference. The measure of fit of this prediction is taken to be the metric that quantifies how much of the environmental variation relates to the multivariate variation in animal movement. We demonstrate the usefulness of this framework through a case study about the contribution of grass availability and time since milking to cow movements using machine learning algorithms. Results: We show that on a one-hour timescale 37% of the variation in grass availability and 33% of time since milking contributed to cow movements. Grass availability contributed mostly to the cows’ neck movement during grazing, while time since milking contributed mostly to the movement through the landscape and the shared variation of accelerometer and GPS data (e.g., activity patterns). Furthermore, this framework proved to be insensitive to spurious correlations between environmental variables in quantifying the contribution to animal movement. Conclusions: Not only is our proposed framework well-suited to study the environmental contribution to animal movement; we argue that it can also be applied in any field that uses multivariate biologging data, e.g., animal physiology, to study the relationships between animals and their environment.

Migration can influence dynamics of pathogen-host interactions. However, it is not clearly known how migration pattern, in terms of the configuration of the migration network and the synchrony of migration, affects infection prevalence. We therefore applied a discrete-time SIR model, integrating environmental transmission and migration, to various migration networks, including networks with serial, parallel, or both serial and parallel stopover sites, and with various levels of migration synchrony. We applied the model to the infection of avian influenza virus in a migratory geese population. In a network with only serial stopover sites, increasing the number of stopover sites reduced infection prevalence, because with every new stopover site, the amount of virus in the environment was lower than that in the previous stopover site, thereby reducing the exposure of the migratory population. In a network with parallel stopover sites, both increasing the number and earlier appearance of the stopover sites led to an earlier peak of infection prevalence in the migratory population, because the migratory population is exposed to larger total amount of virus in the environment, speeding-up the infection accumulation. Furthermore, higher migration synchrony reduced the average number of cumulative infection, because the majority of the population can fly to a new stopover site where the amount of virus is still relatively low and has not been increased due to virus shedding of infected birds. Our simulations indicate that a migration pattern with multiple serial stopover sites and with highly synchronized migration reduces the infection prevalence.

Large areas of agricultural land have been abandoned in European Russia since 1991, triggering succession toward more wooded landscapes, especially in northern regions where conditions for agriculture are more challenging. We hypothesize that this process has contributed to a southward shift by migratory Atlantic Greater White‐fronted geese, as stopover sites in northern Russia became progressively less suitable. To test this hypothesis, we located stopover sites from information contained in 2976 ring recoveries and sightings of neck‐collared geese. These records were divided into three time periods, chosen to reflect major changes in the economy and land use of European Russia: 1960–1990, 1991–2000, and 2001–2013. We used a kernel density estimator grid to delineate areas surrounding 300 putative stopover sites, and statistically evaluated the effects of latitude, distance to nearest waterbody, settlement, and period on stopover site usage by geese. Our results show that over the three periods, usage of the stopover sites has shifted southward, indicating that Greater White‐fronted geese have shifted their migration pathway, with the greatest shift in the most recent period. This shift was confirmed by a highly significant squared latitude term and significant interaction term between periods. The nearest settlements showed no significant effect on stopover site usage while the nearest waterbody term was negative, suggesting higher waterbody densities contributed to higher densities of stopover sites. We attribute the shift to the successional reforestation of the Russian landscape that has followed widespread land abandonment, especially that following the break‐up of the former USSR.

Elephant and impala as intermediate feeders, having a mixed diet of grass and browse, respond to seasonal fluctuations of forage quality by changing their diet composition. We tested the hypotheses that (1) the decrease in forage quality is accompanied by a change in diet from more monocots in the wet season to more dicots in the dry season and that that change is more pronounced and faster in impala than in elephant; (2) mopane (Colophospermum mopane), the most abundant dicot species, is the most important species in the elephant diet in mopane woodland, whereas impala feed relatively less on mopane due to the high condensed tannin concentration; and (3) impala on nutrient-rich soils have a diet consisting of more grass and change later to diet of more browse than impala on nutrient-poor soils. The phosphorus content and in vitro digestibility of monocots decreased and the NDF content increased significantly towards the end of the wet season, whereas in dicots no significant trend could be detected. We argue that this decreasing monocot quality caused elephant and impala to consume more dicots in the dry season. Elephant changed their diet gradually over a 16-week period from 70% to 25% monocots, whereas impala changed diets rapidly (2-4 weeks) from 95% to 70% monocots. For both elephants and impala, there was a positive correlation between percentage of monocots and dicots in the diet and the in vitro digestibility of these forage items. Mopane was the most important dicot species in the elephant diet and its contribution to the diet increased significantly in the dry season, whereas impala selected other dicot species. On nutrient-rich gabbroic soils, impala ate significantly more monocots than impala from nutrient-poor granitic soils, which was related to the higher in vitro digestibility of the monocots on gabbroic soil. Digestibility of food items appears to be an important determinant of diet change from the wet to the dry season in impala and elephants.

1. Scaling theory predicts that organisms respond to different scales of resource patchiness in relation to their own size. We tested the hypothesis that the scale of nutrient patchiness mediates resource partitioning between large trees and small grasses in a semi-arid savanna. 2. In a factorial field experiment, Colophospermum mopane trees and associated grasses were fertilized at either a fine or coarse scale of patchiness with nitrogen (N), phosphorus (P) or N + P. The growth of marked tree shoots, herbaceous biomass and leafNand P concentrations were monitored for 2 years following fertilization. 3. Responses of trees were partly scale dependent. Tree leaf N concentration and shoot length relatively increased with fertilization at a coarse scale. Tree leaf mass decreased when P was supplied at a fine scale of patchiness, suggesting intensified grass competition. 4. Phosphorus fertilization increased leaf P concentrations more in grasses than trees, whereas N fertilization increased leaf N concentration moderately in both trees and grasses. Herbaceous above-ground biomass around focal trees was negatively correlated with tree size when fertilized with N, suggesting intensified tree competition. 5. Synthesis. Our results support the hypothesis that trees benefit more from nutrients supplied at a relatively coarse scale of patchiness. No direct responses of grasses to scale were detected. In trees, the scale effect was surpassed by the effect of sample year, when rainfall varied

1. Elucidation of the mechanism determining the spatial scale of patch selection by herbivores has been complicated by the way in which resource availability at a specific scale is measured and by vigilance behaviour of the herbivores themselves. To reduce these complications, we studied patch selection by an animal with negligible predation risk, the African elephant. 2. We introduce the concept of nutrient load as the product of patch size, number of patches and local patch nutrient concentration. Nutrient load provides a novel spatially explicit expression of the total available nutrients a herbivore can select from. 3. We hypothesized that elephant would select nutrient-rich patches, based on the nutrient load per 2500 m(2) down to the individual plant scale, and that this selection will depend on the nitrogen and phosphorous contents of plants. 4. We predicted that elephant would cause more adverse impact to trees of lower value to them in order to reach plant parts with higher nutrient concentrations such as bark and root. However, elephant should maintain nutrient-rich trees by inducing coppicing of trees through re-utilization of leaves. 5. Elephant patch selection was measured in a homogenous tree species stand by manipulating the spatial distribution of soil nutrients in a large field experiment using NPK fertilizer. 6. Elephant were able to select nutrient-rich patches and utilized Colophospermum mopane trees inside these patches more than outside, at scales ranging from 2500 down to 100 m(2) . 7. Although both nitrogen and phosphorus contents of leaves from C. mopane trees were higher in fertilized and selected patches, patch choice correlated most strongly with nitrogen content. As predicted, stripping of leaves occurred more in nutrient-rich patches, while adverse impact such as uprooting of trees occurred more in nutrient-poor areas. 8. Our results emphasize the necessity of including scale-dependent selectivity in foraging studies and how elephant foraging behaviour can be used as indicators of change in the availability of nutrients.

In savannas, the tree-grass balance is governed by water, nutrients, fire and herbivory, and their interactions. We studied the hypothesis that herbivores indirectly affect vegetation structure by changing the availability of soil nutrients, which, in turn, alters the competition between trees and grasses. Nine abandoned livestock holding-pen areas (kraals), enriched by dung and urine, were contrasted with nearby control sites in a semi-arid savanna. About 40 years after abandonment, kraal sites still showed high soil concentrations of inorganic N, extractable P, K, Ca and Mg compared to controls. Kraals also had a high plant production potential and offered high quality forage. The intense grazing and high herbivore dung and urine deposition rates in kraals fit the accelerated nutrient cycling model described for fertile systems elsewhere. Data of a concurrent experiment also showed that bush-cleared patches resulted in an increase in impala dung deposition, probably because impala preferred open sites to avoid predation. Kraal sites had very low tree densities compared to control sites, thus the high impala dung deposition rates here may be in part driven by the open structure of kraal sites, which may explain the persistence of nutrients in kraals. Experiments indicated that tree seedlings were increasingly constrained when competing with grasses under fertile conditions, which might explain the low tree recruitment observed in kraals. In conclusion, large herbivores may indirectly keep existing nutrient hotspots such as abandoned kraals structurally open by maintaining a high local soil fertility, which, in turn, constrains woody recruitment in a negative feedback loop. The maintenance of nutrient hotspots such as abandoned kraals by herbivores contributes to the structural heterogeneity of nutrient-poor savanna vegetation.

1. Understanding and accurately predicting the spatial patterns of habitat use by organisms is important for ecological research, biodiversity conservation and ecosystem management. However, this understanding is complicated by the effects of spatial scale, because the scale of analysis affects the quantification of species-environment relationships. 2. We therefore assessed the influence of environmental context (i.e. the characteristics of the landscape surrounding a site), varied over a large range of scales (i.e. ambit radii around focal sites), on the analysis and prediction of habitat selection by African elephants in Kruger National Park, South Africa. 3. We focused on the spatial scaling of the elephants' response to their main resources, forage and water, and found that the quantification of habitat selection strongly depended on the scales at which environmental context was considered. Moreover, the inclusion of environmental context at characteristic scales (i.e. those at which habitat selectivity was maximized) increased the predictive capacity of habitat suitability models. 4. The elephants responded to their environment in a scale-dependent and perhaps hierarchical manner, with forage characteristics driving habitat selection at coarse spatial scales, and surface water at fine spatial scales. 5. Furthermore, the elephants exhibited sexual habitat segregation, mainly in relation to vegetation characteristics. Male elephants preferred areas with high tree cover and low herbaceous biomass, whereas this pattern was reversed for female elephants. 6. We show that the spatial distribution of elephants can be better understood and predicted when scale-dependent species-environment relationships are explicitly considered. This demonstrates the importance of considering the influence of spatial scale on the analysis of spatial patterning in ecological phenomena.

The concentrations of resources in forage are not perfectly balanced to the needs of an animal, and food species differ in these concentrations. Under many circumstances, animals should thus forage on multiple food species to attain the maximum and most balanced intake of several resources. In this article we present a model to extend optimal foraging theory to incorporate concurrent foraging for multiple resources from several food species. A balancing of resources is achieved by representing the amount of a resource as the time during which it is used. Optimization is considered at two hierarchical levels: the time spent in a patch and the proportion of patches of each food species included in the foraging path. Our model results show that the balancing of resource intake can be achieved at the level of the foraging path, while the maximization of intake can be realized at the nested patch level. The choice for a food species should be dependent on the differences in intake and resource ratios between the food species. Under free choice of food species, the optimal patch residence time is subject not to differences between patches but to the local intake rate.

Background Urbanization causes modification, fragmentation and loss of native habitats. Such landscape changes threaten many arboreal and gliding mammals by limiting their movements through treeless parts of a landscape and by making the landscape surrounding suitable habitat patches more inhospitable. Here, we investigate the effects of landscape structure and habitat availability on the home-range use and movement patterns of the Siberian flying squirrel (Pteromys volans) at different spatial and temporal scales. We followed radio-tagged individuals in a partly urbanized study area in Eastern Finland, and analysed how landscape composition and connectivity affected the length and speed of movement bursts, distances moved during one night, and habitat and nest-site use. Results The presence of urban habitat on movement paths increased both movement lengths and speed whereas nightly distances travelled by males decreased with increasing amount of urban habitat within the home range. The probability of switching from the present nest site to another nest site decreased with increasing distance among the nest sites, but whether the nest sites were connected or unconnected by forests did not have a clear effect on nest switching. Flying squirrels preferred to use mature forests for their movements at night. Conclusions Our results suggest that the proximity to urban habitats modifies animal movements, possibly because animals try to avoid such habitats by moving faster through them. Urbanization at the scale of an entire home range can restrict their movements. Thus, maintaining a large enough amount of mature forests around inhabited landscape fragments will help protect forest specialists in urban landscapes. The effect of forested connections remains unclear, highlighting the difficulty of measuring and preserving connectivity in a species-specific way. Electronic supplementary material The online version of this article (doi:10.1186/s40462-016-0071-z) contains supplementary material, which is available to authorized users.

Urbanization causes modification, fragmentation and loss of native habitats. Such landscape changes threaten many arboreal and gliding mammals by limiting their movements through treeless parts of a landscape and by making the landscape surrounding suitable habitat patches more inhospitable. Here, we investigate the effects of landscape structure and habitat availability on the home-range use and movement patterns of the Siberian flying squirrel (Pteromys volans) at different spatial and temporal scales. We followed radio-tagged individuals in a partly urbanized study area in Eastern Finland, and analysed how landscape composition and connectivity affected the length and speed of movement bursts, distances moved during one night, and habitat and nest-site use.The presence of urban habitat on movement paths increased both movement lengths and speed whereas nightly distances travelled by males decreased with increasing amount of urban habitat within the home range. The probability of switching from the present nest site to another nest site decreased with increasing distance among the nest sites, but whether the nest sites were connected or unconnected by forests did not have a clear effect on nest switching. Flying squirrels preferred to use mature forests for their movements at night.Our results suggest that the proximity to urban habitats modifies animal movements, possibly because animals try to avoid such habitats by moving faster through them. Urbanization at the scale of an entire home range can restrict their movements. Thus, maintaining a large enough amount of mature forests around inhabited landscape fragments will help protect forest specialists in urban landscapes. The effect of forested connections remains unclear, highlighting the difficulty of measuring and preserving connectivity in a species-specific way.

Wildlife management to reduce the impact of wildlife on their habitat can be done in several ways, among which removing animals (by either culling or translocation) is most often used. There are however alternative ways to control wildlife densities, such as opening or closing water points. The effects of these alternatives are poorly studied. In this paper, we focus on manipulating large herbivores through the closure of water points (WPs). Removal of artificial WPs has been suggested to change the distribution of African elephants, which occur in high densities in national parks in Southern Africa and are thought to have a destructive effect on the vegetation. Here, we modeled the long-term effects of different scenarios of WP closure on the spatial distribution of elephants, and consequential effects on the vegetation and other herbivores in Kruger National Park, South Africa. Using a dynamic ecosystem model, SAVANNA, scenarios were evaluated that varied in (1) availability of artificial WPs, (2) levels of natural water, and (3) elephant densities. Our modeling results showed that elephants can indirectly negatively affect the distributions of mesomixed feeders, mesobrowsers and some mesograzers under wet conditions. The closure of artificial WPs hardly had any effect during these natural wet conditions. Only under dry conditions the spatial distribution of both elephant bulls and cows changed when the availability of artificial water was severely reduced in the model. These changes in spatial distribution triggered changes in the spatial availability of woody biomass over the simulation period of 80 years and this led to changes in the rest of the herbivore community, resulting in increased densities of all herbivores, except for giraffe and steenbok, in areas close to rivers. The spatial distributions of elephant bulls and cows showed to be less affected by the closure of WPs than most of the other herbivore species. Our study contributes to ecologically informed decisions in wildlife management. The results from this modeling exercise imply that long-term effects of this intervention strategy should always be investigated at an ecosystem scale.

Surface water is one of the constraining resources for herbivore populations in semi-arid regions. Artificial waterpoints are constructed by wildlife managers to supplement natural water supplies, to support herbivore populations. The aim of this paper is to analyse how a landowner may realize his ecological and economic goals by manipulating waterpoints for the management of an elephant population, a water-dependent species in the presence of water-independent species. We develop a theoretical bio-economic framework to analyse the optimization of wildlife management objectives (in this case revenue generation from both consumptive and non-consumptive use and biodiversity conservation), using waterpoint construction as a control variable. The model provides a bio-economic framework for analysing optimization problems where a control has direct effects on one herbivore species but indirect effects on the other. A landowner may be interested only in maximization of profits either from elephant offtake and/or tourism revenue, ignoring the negative effects that could be brought about by elephants to biodiversity. If the landowner does not take the indirect effects of waterpoints into consideration, then the game reserve management, as the authority entrusted with the sustainable management of the game reserve, might use economic instruments such as subsidies or taxes to the landowners to enforce sound waterpoint management.

1. Understanding and accurately predicting the spatial patterns of habitat use by organisms is important for ecological research, biodiversity conservation and ecosystem management. However, this understanding is complicated by the effects of spatial scale, because the scale of analysis affects the quantification of species-environment relationships. 2. We therefore assessed the influence of environmental context (i.e. the characteristics of the landscape surrounding a site), varied over a large range of scales (i.e. ambit radii around focal sites), on the analysis and prediction of habitat selection by African elephants in Kruger National Park, South Africa. 3. We focused on the spatial scaling of the elephants' response to their main resources, forage and water, and found that the quantification of habitat selection strongly depended on the scales at which environmental context was considered. Moreover, the inclusion of environmental context at characteristic scales (i.e. those at which habitat selectivity was maximized) increased the predictive capacity of habitat suitability models. 4. The elephants responded to their environment in a scale-dependent and perhaps hierarchical manner, with forage characteristics driving habitat selection at coarse spatial scales, and surface water at fine spatial scales. 5. Furthermore, the elephants exhibited sexual habitat segregation, mainly in relation to vegetation characteristics. Male elephants preferred areas with high tree cover and low herbaceous biomass, whereas this pattern was reversed for female elephants. 6. We show that the spatial distribution of elephants can be better understood and predicted when scale-dependent species-environment relationships are explicitly considered. This demonstrates the importance of considering the influence of spatial scale on the analysis of spatial patterning in ecological phenomena.

Predation risk from lions (Panthera leo) has been linked to habitat characteristics and availability and traits of prey. We separated the effects of vegetation density and the presence of drinking water by analyzing locations of lion kills in relation to rivers with dense vegetation, which offer good lion stalking opportunities, and artificial water points with low vegetation density. The spatial distribution of lion kills was studied at the Klaserie Private Nature Reserve, South Africa. The distance between 215 lion kills and the nearest water source was analyzed using generalized linear models. Lions selected medium-sized prey species. Lion kills were closer to rivers and to artificial water points than expected by random distribution of the kills. Water that attracted prey, and not the vegetation density in riverine areas, increased predation risk, with kills of buffalo (Syncerus caffer), kudu (Tragelaphus strepsiceros), and wildebeest (Connochaetes taurinus) as water-dependent prey species. Traits of prey species, including feeding type (food habits), digestion type (ruminant or nonruminant), or body size, did not explain locations of lion kills, and no seasonal patterns in lion kills were apparent. We argue that the cascading impact of lions on local mammal assemblages is spatially heterogeneous.

1. Globally, both climatic patterns and nitrogen deposition rates show directional changes over time. It is uncertain how woody seedlings, which coexist with herbaceous plants in savannas, respond to concurrent changes in water and nutrient availability. 2. We investigated competition effects between herbaceous vegetation and tree seedlings (Colophospermum mopane) under changed water and nutrient (fertilized) conditions in a garden experiment situated in a semi-arid savanna. 3. Herbaceous competition significantly suppressed woody seedling growth. The effect of herbaceous competition on woody seedling growth remained constant with both increasing water and nutrient availability. However, during a wet-season drought, herbaceous competition apparently caused premature leaf senescence in non-irrigated treatments. Fertilization exacerbated negative competition effects during the drought, while irrigation prevented leaf loss of tree seedlings in spite of herbaceous competition and fertilization. 4. Based on a conceptual model, we propose that the vigorous response of herbaceous plants to increased nutrient availability leads to faster depletion of soil water, which increasingly causes water stress in woody seedlings if the interval between watering events is prolonged, e.g. during wet-season droughts. 5. Synthesis. Our data support the notion that changes in drought frequency are of greater importance to woody recruitment success than changes in annual rainfall amount. Based on the water and nutrient interactions observed in our experiment, we suggest that the effect of increased nitrogen deposition on woody seedling recruitment is contingent on water availability.

In this paper, we address the question whether and through which mechanisms herbivores can induce spatial patterning in savanna vegetation, and how the role of herbivory as a determinant of vegetation patterning changes with herbivore density and the pre-existing pattern of vegetation. We thereto developed a spatially explicit simulation model, including growth of grasses and trees, vertical zonation of browseable biomass, and spatially explicit foraging by grazers and browsers. We show that herbivores can induce vegetation patterning when two key assumptions are fulfilled. First, herbivores have to increase the attractiveness of a site while foraging so that they will revisit this site, e.g. through an increased availability or quality of forage. Second, foraging should be spatially explicit, e.g. when foraging at a site influences vegetation at larger spatial scales or when vegetation at larger spatial scales influences the selection and utilisation of a site. The interaction between these two assumptions proved to be crucial for herbivores to produce spatial vegetation patterns, but then only at low to intermediate herbivore densities. High herbivore densities result in homogenisation of vegetation. Furthermore, our model shows that the pre-existing spatial pattern in vegetation influences the process of vegetation patterning through herbivory. However, this influence decreases when the heterogeneity and dominant scale of the initial vegetation decreases. Hence, the level of adherence of the herbivores to forage in pre-existing patches increases when these pre-existing patches increase in size and when the level of vegetation heterogeneity increases. The findings presented in this paper, and critical experimentation of their ecological validity, will increase our understanding of vegetation patterning in savanna ecosystems, and the role of plant¿herbivore interactions therein.