Animal Trajectories Research Articles

Imputation Approaches for Animal Movement Modeling

The analysis of telemetry data is common in animal ecological studies. While the collection of telemetry data for individual animals has improved dramatically, the methods to properly account for inherent uncertainties (e.g., measurement error, dependence, barriers to movement) have lagged behind. Still, many new statistical approaches have been developed to infer unknown quantities affecting animal movement or predict movement based on telemetry data. Hierarchical statistical models are useful to account for some of the aforementioned uncertainties, as well as provide population-level inference, but they often come with an increased computational burden. For certain types of statistical models, it is straightforward to provide inference if the latent true animal trajectory is known, but challenging otherwise. In these cases, approaches related to multiple imputation have been employed to account for the uncertainty associated with our knowledge of the latent trajectory. Despite the increasing use of imputation approaches for modeling animal movement, the general sensitivity and accuracy of these methods have not been explored in detail. We provide an introduction to animal movement modeling and describe how imputation approaches may be helpful for certain types of models. We also assess the performance of imputation approaches in two simulation studies. Our simulation studies suggests that inference for model parameters directly related to the location of an individual may be more accurate than inference for parameters associated with higher-order processes such as velocity or acceleration. Finally, we apply these methods to analyze a telemetry data set involving northern fur seals (Callorhinus ursinus) in the Bering Sea. Supplementary materials accompanying this paper appear online.

Measuring site fidelity and spatial segregation within animal societies.

Summary Animals often display a marked tendency to return to previously visited locations that contain important resources, such as water, food, or developing brood that must be provisioned. A considerable body of work has demonstrated that this tendency is strongly expressed in ants, which exhibit fidelity to particular sites both inside and outside the nest. However, thus far many studies of this phenomena have taken the approach of reducing an animal's trajectory to a summary statistic, such as the area it covers.Using both simulations of biased random walks, and empirical trajectories from individual rock ants, Temnothorax albipennis, we demonstrate that this reductive approach suffers from an unacceptably high rate of false negatives.To overcome this, we describe a site‐centric approach which, in combination with a spatially‐explicit null model, allows the identification of the important sites towards which individuals exhibit statistically significant biases.Using the ant trajectories, we illustrate how the site‐centric approach can be combined with social network analysis tools to detect groups of individuals whose members display similar space‐use patterns.We also address the mechanistic origin of individual site fidelity; by examining the sequence of visits to each site, we detect a statistical signature associated with a self‐attracting walk – a non‐Markovian movement model that has been suggested as a possible mechanism for generating individual site fidelity.

Zebrafish tracking using convolutional neural networks

Keeping identity for a long term after occlusion is still an open problem in the video tracking of zebrafish-like model animals, and accurate animal trajectories are the foundation of behaviour analysis. We utilize the highly accurate object recognition capability of a convolutional neural network (CNN) to distinguish fish of the same congener, even though these animals are indistinguishable to the human eye. We used data augmentation and an iterative CNN training method to optimize the accuracy for our classification task, achieving surprisingly accurate trajectories of zebrafish of different size and age zebrafish groups over different time spans. This work will make further behaviour analysis more reliable.

Size Matters: Observed and Modeled Camouflage Response of European Cuttlefish (Sepia officinalis) to Different Substrate Patch Sizes during Movement.

Camouflage is common throughout the phylogenetic tree and is largely used to minimize detection by predator or prey. Cephalopods, and in particular Sepia officinalis cuttlefish, are common models for camouflage studies. Predator avoidance behavior is particularly important in this group of soft-bodied animals that lack significant physical defenses. While previous studies have suggested that immobile cephalopods selectively camouflage to objects in their immediate surroundings, the camouflage characteristics of cuttlefish during movement are largely unknown. In a heterogenic environment, the visual background and substrate feature changes quickly as the animal swim across it, wherein substrate patch is a distinctive and high contrast patch of substrate in the animal's trajectory. In the current study, we examine the effect of substrate patch size on cuttlefish camouflage, and specifically the minimal size of an object for eliciting intensity matching response while moving. Our results indicated that substrate patch size has a positive effect on animal's reflectance change, and that the threshold patch size resulting in camouflage response falls between 10 and 19 cm (width). These observations suggest that the animal's length (7.2–12.3 cm mantle length in our case) serves as a possible threshold filter below which objects are considered irrelevant for camouflage, reducing the frequency of reflectance changes—which may lead to detection. Accordingly, we have constructed a computational model capturing the main features of the observed camouflaging behavior, provided for cephalopod camouflage during movement.

Visual animal biometrics: survey

Visual animal biometrics is an emerging research discipline in computer vision, pattern recognition and cognitive science. It is a promising research field that encourages new development of quantified algorithms and methodologies for representing, detection of visible features, phenotypic appearances of species, individuals and recognition of morphological and animal biometric characteristics. Furthermore, it also assists the study of animal trajectory and behaviours analysis of species. Currently, real-world applications of visual animal biometric systems are gaining more proliferation due to a variety of applications and use, enhancement of quantity and quality of the collection of extensive ecological data and processing. However, to advance visual animal biometrics will require integration of methodologies among the scientific disciplines involved. Such valuable efforts will be worthwhile due to the enormous perspective of this approach rests with the formal abstraction of phenomics, to build well-developed interfaces between different organisational levels of life. This study provides a comprehensive survey of visual animal biometric systems and recognition approaches for various species and individual animal based on their morphological image pattern and biometric characteristics. This comprehensive review paper encourages the multidisciplinary researchers, scientists, biologists and different research communities to design the better platforms for the development of efficient algorithms and learning models to solve the massive data processing, classification and identification of different species related problems.

Temporal segmentation of animal trajectories informed by habitat use

AbstractMost animals live in seasonal environments and experience very different conditions throughout the year. Behavioral strategies like migration, hibernation, and a life cycle adapted to the local seasonality help to cope with fluctuations in environmental conditions. Thus, how an individual utilizes the environment depends both on the current availability of habitat and the behavioral prerequisites of the individual at that time. While the increasing availability and richness of animal movement data has facilitated the development of algorithms that classify behavior by movement geometry, changes in the environmental correlates of animal movement have so far not been exploited for a behavioral annotation. Here, we suggest a method that uses these changes in individual–environment associations to divide animal location data into segments of higher ecological coherence, which we term niche segmentation. We use time series of random forest models to evaluate the transferability of habitat use over time to cluster observational data accordingly. We show that our method is able to identify relevant changes in habitat use corresponding to both changes in the availability of habitat and how it was used using simulated data, and apply our method to a tracking data set of common teal (Anas crecca). The niche segmentation proved to be robust, and segmented habitat suitability outperformed models neglecting the temporal dynamics of habitat use. Overall, we show that it is possible to classify animal trajectories based on changes of habitat use similar to geometric segmentation algorithms. We conclude that such an environmentally informed classification of animal trajectories can provide new insights into an individuals' behavior and enables us to make sensible predictions of how suitable areas might be connected by movement in space and time.

Effects of Land Cover on the Movement of Frugivorous Birds in a Heterogeneous Landscape.

Movement is a key spatiotemporal process that enables interactions between animals and other elements of nature. The understanding of animal trajectories and the mechanisms that influence them at the landscape level can yield insight into ecological processes and potential solutions to specific ecological problems. Based upon optimal foraging models and empirical evidence, we hypothesized that movement by thrushes is highly tortuous (low average movement speeds and homogeneous distribution of turning angles) inside forests, moderately tortuous in urban areas, which present intermediary levels of resources, and minimally tortuous (high movement speeds and turning angles next to 0 radians) in open matrix types (e.g., crops and pasture). We used data on the trajectories of two common thrush species (Turdus rufiventris and Turdus leucomelas) collected by radio telemetry in a fragmented region in Brazil. Using a maximum likelihood model selection approach we fit four probability distribution models to average speed data, considering short-tailed, long-tailed, and scale-free distributions (to represent different regimes of movement variation), and one distribution to relative angle data. Models included land cover type and distance from forest-matrix edges as explanatory variables. Speed was greater farther away from forest edges and increased faster inside forest habitat compared to urban and open matrices. However, turning angle was not influenced by land cover. Thrushes presented a very tortuous trajectory, with many displacements followed by turns near 180 degrees. Thrush trajectories resembled habitat and edge dependent, tortuous random walks, with a well-defined movement scale inside each land cover type. Although thrushes are habitat generalists, they showed a greater preference for forest edges, and thus may be considered edge specialists. Our results reinforce the importance of studying animal movement patterns in order to understand ecological processes such as seed dispersal in fragmented areas, where the percentage of remaining habitat is dwindling.

Estimating where and how animals travel: an optimal framework for path reconstruction from autocorrelated tracking data

AbstractAn animal's trajectory is a fundamental object of interest in movement ecology, as it directly informs a range of topics from resource selection to energy expenditure and behavioral states. Optimally inferring the mostly unobserved movement path and its dynamics from a limited sample of telemetry observations is a key unsolved problem, however. The field of geostatistics has focused significant attention on a mathematically analogous problem that has a statistically optimal solution coined after its inventor,Krige. Kriging revolutionized geostatistics and is now the gold standard for interpolating between a limited number of autocorrelated spatial point observations. Here we translate Kriging for use with animal movement data. Our Kriging formalism encompasses previous methods to estimate animal's trajectories—the Brownian bridge and continuous‐time correlated random walk library—as special cases, informs users as to when these previous methods are appropriate, and provides a more general method when they are not. We demonstrate the capabilities of Kriging on a case study with Mongolian gazelles where, compared to the Brownian bridge, Kriging with a more optimal model was 10% more precise in interpolating locations and 500% more precise in estimating occurrence areas.

Quantifying Animal Trajectories Using Spatial Aggregation and Sequence Analysis: A Case Study of Differentiating Trajectories of Multiple Species

The increasing availability of telemetry data with high spatial and temporal resolution promises to greatly advance scientific understandings of the movement patterns of individual organisms across space and time. The amount of data provided by such methods, however, can be challenging to analyze and interpret. In this study, we present a new approach for analyzing animal movements that aggregates telemetry locations into spatial clusters and extracts the information from sequences formed by individuals passing through these spatial clusters. We applied this integrated approach of spatial aggregation and sequence analysis to quantify and compare trajectories of cattle (Bos taurus), mule deer (Odocoileus hemionus), and elk (Cervus elaphus) tracked by automated telemetry at the Starkey Experimental Forest and Range in northeastern Oregon, USA. Our approach effectively differentiated movement patterns of the three species. It provides a useful mean of quantifying movement patterns of species in a landscape.

Ecological Trade-offs between Migration and Reproduction Are Mediated by the Nutrition-Sensitive Insulin-Signaling Pathway.

Crowding and changes in food availability are two critical environmental conditions that impact an animal's trajectory toward either migration or reproduction. Many insects facing this challenge have evolved wing polyphenisms. When conditions favor reproduction, wing polyphenic species produce adults that either have no wings or short, non-functional wings. Facultative wing growth reflects a physiological and evolutionary trade-off between migration and reproduction, triggered by environmental conditions. How environmental cues are transduced to produce these alternative forms, and their associated ecological shift from migration to reproduction, remains an important unsolved problem in evolutionary ecology. The brown planthopper, a wing polymorphic insect exhibiting strong trade-offs in investment between migration and reproduction, is one of the most serious rice pests in Asia. In this study, we investigated the function of four genes in the insulin-signaling pathway known to couple nutrition with growth, PI3 Kinase (PI3K), PDK1, Akt (Protein Kinase B), and the forkhead gene FOXO. Using a combination of RNA interference and pharmacological inhibitor treatment, we show that all four genes contribute to tissue level regulation of wing polymorphic development in this insect. As predicted, silencing of the NlPI3K, NlAkt and NlPDK1 through dsRNA and with the pharmacological inhibitor Perifosine resulted in short-winged brown planthoppers, whereas knockdown of NlFOXO resulted in long-winged planthoppers. Morphometric analyses confirm that phenotypes from our manipulations mimic what would be found in nature, i.e., major parameters such as bristle number, wing area and body weight are not significantly different from non-experimental animals. Taken together, these data implicate the insulin-signaling pathway in the transduction of environmental factors into condition-dependent patterns of wing growth in insects.

Estimating where and how animals travel: an optimal framework for path reconstruction from autocorrelated tracking data.

An animal's trajectory is a fundamental object of interest in movement ecology, as it directly informs a range of topics from resource selection to energy expenditure and behavioral states. Optimally inferring the mostly unobserved movement path and its dynamics from a limited sample of telemetry observations is a key unsolved problem, however. The field of geostatistics has focused significant attention on a mathematically analogous problem that has a statistically optimal solution coined after its inventor, Krige. Kriging revolutionized geostatistics and is now the gold standard for interpolating between a limited number of autocorrelated spatial point observations. Here we translate Kriging for use with animal movement data. Our Kriging formalism encompasses previous methods to estimate animal's trajectories--the Brownian bridge and continuous-time correlated random walk library--as special cases, informs users as to when these previous methods are appropriate, and provides a more general method when they are not. We demonstrate the capabilities of Kriging on a case study with Mongolian gazelles where, compared to the Brownian bridge, Kriging with a more optimal model was 10% more precise in interpolating locations and 500% more precise in estimating occurrence areas.

Machine vision application in animal trajectory tracking

This article was motivated by the doctors’ demand to make a technical support in pathologies of gastrointestinal tract research [10], which would be based on machine vision tools. Proposed solution should be less expensive alternative to already existing RF (radio frequency) methods. The objective of whole experiment was to evaluate the amount of animal motion dependent on degree of pathology (gastric ulcer). In the theoretical part of the article, several methods of animal trajectory tracking are presented: two differential methods based on background subtraction, the thresholding methods based on global and local threshold and the last method used for animal tracking was the color matching with a chosen template containing a searched spectrum of colors. The methods were tested offline on five video samples. Each sample contained situation with moving guinea pig locked in a cage under various lighting conditions.

Prying into the intimate secrets of animal lives; software beyond hardware for comprehensive annotation in 'Daily Diary' tags.

BackgroundSmart tags attached to freely-roaming animals recording multiple parameters at infra-second rates are becoming commonplace, and are transforming our understanding of the way wild animals behave. Interpretation of such data is complex and currently limits the ability of biologists to realise the value of their recorded information.DescriptionThis work presents Framework4, an all-encompassing software suite which operates on smart sensor data to determine the 4 key elements considered pivotal for movement analysis from such tags (Endangered Species Res 4: 123-37, 2008). These are; animal trajectory, behaviour, energy expenditure and quantification of the environment in which the animal moves. The program transforms smart sensor data into dead-reckoned movements, template-matched behaviours, dynamic body acceleration-derived energetics and position-linked environmental data before outputting it all into a single file. Biologists are thus left with a single data set where animal actions and environmental conditions can be linked across time and space.ConclusionsFramework4 is a user-friendly software that assists biologists in elucidating 4 key aspects of wild animal ecology using data derived from tags with multiple sensors recording at high rates. Its use should enhance the ability of biologists to derive meaningful data rapidly from complex data.

Extracting spatio‐temporal patterns in animal trajectories: an ecological application of sequence analysis methods

Summary Digital tracking technologies have considerably increased the amount and quality of animal trajectories, enabling the study of habitat use and habitat selection at a fine spatial and temporal scale. However, current approaches do not yet explicitly account for a key aspect of habitat use, namely the sequential variation in the use of different habitat features. To overcome this limitation, we propose a tree‐based approach that makes use of sequence analysis methods, derived from molecular biology, to explore and identify ecologically relevant sequential patterns in habitat use by animals. We applied this approach to ecological data consisting of simulated and real trajectories from a roe deer population (Capreolus capreolus), expressed as ordered sequences of habitat use. We show that our approach effectively captured spatio‐temporal patterns of sequential habitat use by roe deer. In our case study, individual sequences were clustered according to the sequential use of the elevation gradient (first order) and of open/closed habitats (second order). We provided evidence for several behavioural processes, such as migration and daily alternating habitat use. Some unexpected patterns, such as homogeneous sequences of use of open habitat, could also be identified. Our findings advocate the importance of dealing with the sequential nature of movement data. Approaches based on sequence analysis methods are particularly useful and effective since they allow exploring temporal patterns of habitat use in a synthetic and visually captive manner. The proposed approach represents a useful and effective way to classify individual movement behaviour across populations and species. Ultimately, this method can be applied to explore the temporal scale of ecological processes based on movement.

MODELLING FREE-RANGE CATTLE MOVEMENTS IN FORESTS USING MULTISTATE RANDOM WALKS

In this study multi-state random walk models were tested as general approach to analyse movement states in animal relocation data. For this purpose, GPS collars were fit to free-range cattle inhabiting Andean forests in Patagonia, Argentina. The results indicate that animal trajectories have re-visitation patterns in restricted areas of few kilometres and that a large part of behaviors occur at a time scale lower than six hours. Simple multi-state random walk models proved to be practical in modelling characteristics such as step and turning angles distribution in addition to emergent properties of animal trajectories such as temporal autocorrelation. Re-visitation patterns and the emergence of home-range, among other memory processes could be marked as possible limitations of these models. This work is the first description of free-range cattle movement inhabiting the andean-Patagonian forests in semi-feral conditions.

Automated high-throughput measurement of body movements and cardiac activity of Xenopus tropicalis tadpoles

Xenopus tadpoles are an emerging model for developmental, genetic and behavioral studies. A small size, optical accessibility of most of their organs, together with a close genetic and structural relationship to humans make them a convenient experimental model. However, there is only a limited toolset available to measure behavior and organ function of these animals at medium or high-throughput. Herein, we describe an imaging-based platform to quantify body and autonomic movements of Xenopus tropicalis tadpoles of advanced developmental stages. Animals alternate periods of quiescence and locomotor movements and display buccal pumping for oxygen uptake from water and rhythmic cardiac movements. We imaged up to 24 animals in parallel and automatically tracked and quantified their movements by using image analysis software. Animal trajectories, moved distances, activity time, buccal pumping rates and heart beat rates were calculated and used to characterize the effects of test compounds. We evaluated the effects of propranolol and atropine, observing a dose-dependent bradycardia and tachycardia, respectively. This imaging and analysis platform is a simple, cost-effective high-throughput in vivo assay system for genetic, toxicological or pharmacological characterizations.

Bivariate Gaussian bridges: directional factorization of diffusion in Brownian bridge models

BackgroundIn recent years high resolution animal tracking data has become the standard in movement ecology. The Brownian Bridge Movement Model (BBMM) is a widely adopted approach to describe animal space use from such high resolution tracks. One of the underlying assumptions of the BBMM is isotropic diffusive motion between consecutive locations, i.e. invariant with respect to the direction.Here we propose to relax this often unrealistic assumption by separating the Brownian motion variance into two directional components, one parallel and one orthogonal to the direction of the motion.ResultsOur new model, the Bivariate Gaussian bridge (BGB), tracks movement heterogeneity across time. Using the BGB and identifying directed and non-directed movement within a trajectory resulted in more accurate utilisation distributions compared to dynamic Brownian bridges, especially for trajectories with a non-isotropic diffusion, such as directed movement or Lévy like movements. We evaluated our model with simulated trajectories and observed tracks, demonstrating that the improvement of our model scales with the directional correlation of a correlated random walk.ConclusionWe find that many of the animal trajectories do not adhere to the assumptions of the BBMM. The proposed model improves accuracy when describing the space use both in simulated correlated random walks as well as observed animal tracks. Our novel approach is implemented and available within the “move” package for R.Electronic supplementary materialThe online version of this article (doi:10.1186/2051-3933-2-5) contains supplementary material, which is available to authorized users.

Measuring effects of linear obstacles on wildlife movements: accounting for the relationship between step length and crossing probability

Animal movements in the landscape are influenced by linear features such as rivers, roads and power lines. Prior studies have investigated how linear features, particularly roads, affect movement rates by comparing animal's movement rate measured as step lengths (i.e., the distance between consecutive observations such as GPS locations) before, during and after crossing of a linear feature. The null hypothesis has been that the length of crossing steps should not differ from other steps, and a deviation from this, mainly that steps are longer during crossing, has been taken as support for a disturbance effect of the linear feature. However, based on the simple relationship between the length of a step and its probability to cross a linear feature, we claim that this assumption is inappropriate to test for behavioural responses to linear features. The probability is related to the proportion of the total length of the trajectory (i.e., the path of movement) a step constitutes. Consequently, care should be taken when formulating hypotheses about how animal moves in relation to linear features in the landscape. Statistical tests should be set up with respect to the expected length based on the distribution of step lengths in the trajectory. We propose two methods that accounts for the bias in crossing frequency that is caused by step lengths, and illustrates their applications by using simulated animal trajectories as well as empirical data on reindeer in an area with a power line.

Splitting animal trajectories into fine-scale behaviorally consistent movement units: breaking points relate to external stimuli in a foraging seabird

Animal movements are widely studied in ecology, and the analysis of tracking data usually gains from splitting the time-series into different parts before interpreting movement strategies. The recent increase in data accuracy and resolution allows for the study of fine-scale movements where each behavioral change is recorded. We propose a simple method to identify the elementary units of movement in a trajectory, resulting in breaks in the track corresponding to the animals' decisions to change its movement. We quantify the movement between successive steps with a vector of speed and direction and represent a movement path in a trigonometric circle space in order to visualize behavioral changes instead of spatial changes. In this space, the distance between successive points informs about their similarity in both speed and direction. We quantify the temporal changes in these distances with a cumulative sum and use a line simplification algorithm to identify breaks in the slope that correspond to breaks in the consistency of successive distance values. We test the algorithm on simulated trajectories and show that the expected number of segments is accurately identified. Moreover, we relate the resulting segmentation from recorded trajectories to events observed using animal-borne video footage and show that the presence of stimuli in the surroundings of the animal is associated with a higher frequency of changes in movement. As an applied example, we propose a descriptive analysis of the segments and show that segments of particular characteristics are not distributed equally along the trajectory, highlighting larger-scale behavioral strategies.

Medial Entorhinal Grid Cells and Head Direction Cells Rotate with a T-Maze More Often During Less Recently Experienced Rotations

Prior studies of head direction (HD) cells indicate strong landmark control over the preferred firing direction of these cells, with few studies exhibiting shifts away from local reference frames over time. We recorded spiking activity of grid and HD cells in the medial entorhinal cortex of rats, testing correlations of local environmental cues with the spatial tuning curves of these cells' firing fields as animals performed continuous spatial alternation on a T-maze that shared the boundaries of an open-field arena. The environment was rotated into configurations the animal had either seen or not seen in the past recording week. Tuning curves of both cell types demonstrated commensurate shifts of tuning with T-maze rotations during less recent rotations, more so than recent rotations. This strongly suggests that animals are shifting their reference frame away from the local environmental cues over time, learning to use a different reference frame more likely reliant on distal or idiothetic cues. In addition, grid fields demonstrated varying levels of "fragmentation" on the T-maze. The propensity for fragmentation does not depend on grid spacing and grid score, nor animal trajectory, indicating the cognitive treatment of environmental subcompartments is likely driven by task demands.