Extending the Functionality of Behavioural Change-Point Analysis with k-Means Clustering: A Case Study with the Little Penguin (Eudyptula minor)

Classifying behavior from short-interval biologging data: An example with GPS tracking of birds

Recent advances in digital data collection have spurred accumulation of immense quantities of data that have potential to lead to remarkable ecological insight, but that also present analytic challenges. In the case of biologging data from birds, common analytical approaches to classifying movement behaviors are largely inappropriate for these massive data sets.We apply a framework for using K-means clustering to classify bird behavior using points from short time interval GPS tracks. K-means clustering is a well-known and computationally efficient statistical tool that has been used in animal movement studies primarily for clustering segments of consecutive points. To illustrate the utility of our approach, we apply K-means clustering to six focal variables derived from GPS data collected at 1-11 s intervals from free-flying bald eagles (Haliaeetus leucocephalus) throughout the state of Iowa, USA. We illustrate how these data can be used to identify behaviors and life-stage- and age-related variation in behavior.After filtering for data quality, the K-means algorithm identified four clusters in >2 million GPS telemetry data points. These four clusters corresponded to three movement states: ascending, flapping, and gliding flight; and one non-moving state: perching. Mapping these states illustrated how they corresponded tightly to expectations derived from natural history observations; for example, long periods of ascending flight were often followed by long gliding descents, birds alternated between flapping and gliding flight.The K-means clustering approach we applied is both an efficient and effective mechanism to classify and interpret short-interval biologging data to understand movement behaviors. Furthermore, because it can apply to an abundance of very short, irregular, and high-dimensional movement data, it provides insight into small-scale variation in behavior that would not be possible with many other analytical approaches.

Sequential movement pattern-mining (SMP) in field-based team-sport: A framework for quantifying spatiotemporal data and improve training specificity?

Athlete external load is typically quantified as volumes or discretised threshold values using distance, speed and time. A framework accounting for the movement sequences of athletes has previously been proposed using radio frequency data. This study developed a framework to identify sequential movement sequences using GPS-derived spatiotemporal data in team-sports and establish its stability. Thirteen rugby league players during one match were analysed to demonstrate the application of the framework. The framework (Sequential Movement Pattern-mining [SMP]) applies techniques to analyse i) geospatial data (i.e., decimal degree latitude and longitude), ii) determine players turning angles, iii) improve movement descriptor assignment, thus improving movement unit formation and iv) improve the classification and identification of players' frequent SMP. The SMP framework allows for sub-sequences of movement units to be condensed, removing repeated elements, which offers a novel technique for the quantification of similarities or dis-similarities between players and playing positions. The SMP framework provides a robust and stable method that allows, for the first time the analysis of GPS-derived data and identifies the frequent SMP of field-based team-sport athletes. The application of the SMP framework in practice could optimise the outcomes of training of field-based team-sport athletes by improving training specificity.

Fine-scale movements and behaviors of coyotes (Canis latrans) during their reproductive period

In canids, resident breeders hold territories but require different resources than transient individuals (i.e., dispersers), which may result in differential use of space, land cover, and food by residents and transients. In the southeastern United States, coyote (Canis latrans) reproduction occurs during spring and is energetically demanding for residents, but transients do not reproduce and therefore can exhibit feeding behaviors with lower energetic rewards. Hence, how coyotes behave in their environment likely differs between resident and transient coyotes. We captured and monitored 36 coyotes in Georgia during 2018-2019 and used data from 11 resident breeders, 12 predispersing residents (i.e., offspring of resident breeders), and 11 transients to determine space use, movements, and relationships between these behaviors and landcover characteristics. Average home range size for resident breeders and predispersing offspring was 20.7 ± 2.5 km² and 50.7 ± 10.0 km², respectively. Average size of transient ranges was 241.4 ± 114.5 km². Daily distance moved was 6.3 ± 3.0 km for resident males, 5.5 ± 2.7 km for resident females, and 6.9 ± 4.2 km for transients. We estimated first-passage time values to assess the scale at which coyotes respond to their environment, and used behavioral change-point analysis to determine that coyotes exhibited three behavioral states. We found notable differences between resident and transient coyotes in regard to how landcover characteristics influenced their behavioral states. Resident coyotes tended to select for areas with denser vegetation while resting and foraging, but for areas with less dense vegetation and canopy cover when walking. Transient coyotes selected areas closer to roads and with lower canopy cover while resting, but for areas farther from roads when foraging and walking. Our findings suggest that behaviors of both resident and transient coyotes are influenced by varying landcover characteristics, which could have implications for prey.

Linking behavioral states to landscape features for improved conservation management

A central theme for conservation is understanding how animals differentially use, and are affected by change in, the landscapes they inhabit. However, it has been challenging to develop conservation schemes for habitat-specific behaviors.Here we use behavioral change point analysis to identify behavioral states of golden eagles (Aquila chrysaetos) in the Sonoran and Mojave Deserts of the southwestern United States, and we identify, for each behavioral state, conservation-relevant habitat associations.We modeled behavior using 186,859 GPS points from 48 eagles and identified 2,851 distinct segments comprising four behavioral states. Altitude above ground level (AGL) best differentiated behavioral states, with two clusters of short-distance movement behaviors characterized by low AGL (state 1 AGL = 14 m (median); state 2 AGL = 11 m) and two associated with longer-distance movement behaviors and characterized by higher AGL (state 3 AGL = 108 m; state 4 AGL = 450 m).Behaviors such as perching and low-altitude hunting were associated with short-distance movements in updraft-poor environments, at higher elevations, and over steeper and more north-facing terrain. In contrast, medium-distance movements such as hunting and transiting were over gentle and south-facing slopes. Long-distance transiting occurred over the desert habitats that generate the best updraft.This information can guide management of this species, and our approach provides a template for behavior-specific habitat associations for other species of management concern.

Measuring ontogenetic shifts in central-place foragers: A case study with honeybees

Measuring time-activity budgets over the complete individual life span is now possible for many animals with the recent advances of life-long individual monitoring devices. Although analyses of changes in the patterns of time-activity budgets have revealed ontogenetic shifts in birds or mammals, no such technique has been applied to date on insects. We tested an automated breakpoint-based procedure to detect, assess and quantify shifts in the temporal pattern of the flight activities in honeybees. We assumed that the learning and foraging stages of honeybees will differ in several respects, to detect the age at onset of foraging (AOF). Using an extensive dataset covering the life-long monitoring of 1,167 individuals, we compared the AOF outputs with the more conventional approaches based on arbitrary thresholds. We further evaluated the robustness of the different methods comparing the foraging time-activity budget allocations between the presumed foragers and confirmed foragers. We revealed a clear-cut learning-foraging ontogenetic shift that differs in duration, frequency and time of occurrence of flights. Although AOF appeared to be highly plastic among bees, the breakpoint-based procedure seems better capable to detect it than arbitrary threshold-based methods that are unable to deal with inter-individual variation. We developed the aof r-package including a broad range of examples with both simulated and empirical datasets to illustrate the simplicity of use of the procedure. This simple procedure is generic enough to be derived from any individual life-long monitoring devices recording the time-activity budgets, and could propose new ecological applications of bio-logging to detect ontogenetic shifts in the behaviour of central-place foragers.

Integrating direct observation and GPS tracking to monitor animal behavior for resource management

Monitoring the behavior of pack animals in protected areas informs management about use patterns and the potential associated negative impacts. However, systematic assessments of behavior are uncommon due to methodological and logistical constraints. This study integrated behavior mapping with GPS tracking, and applied behavior change point analysis, as an approach to monitor the behaviors of pack animals during overnight periods. The integrated approach identified multiple grazing patterns (i.e., locally intense grazing, ambulatory grazing) not feasible through a single methodology alone. Monitoring behavior and corresponding environmental conditions aid managers in implementing strategies designed to mitigate impacts associated with pack animals in natural areas. Results also contrast the influence of temporal scale on behavior segmentation to inform decisions for further monitoring and management of domestic animal use and impacts in natural areas. This integrated approach reduced time and logistical constraints of each method individually to promote ongoing monitoring and highlight how multiple management tactics could reduce impacts to sensitive habitats.

Search and foraging behaviors from movement data: A comparison of methods

Search behavior is often used as a proxy for foraging effort within studies of animal movement, despite it being only one part of the foraging process, which also includes prey capture. While methods for validating prey capture exist, many studies rely solely on behavioral annotation of animal movement data to identify search and infer prey capture attempts. However, the degree to which search correlates with prey capture is largely untested. This study applied seven behavioral annotation methods to identify search behavior from GPS tracks of northern gannets (Morus bassanus), and compared outputs to the occurrence of dives recorded by simultaneously deployed time-depth recorders. We tested how behavioral annotation methods vary in their ability to identify search behavior leading to dive events. There was considerable variation in the number of dives occurring within search areas across methods. Hidden Markov models proved to be the most successful, with 81% of all dives occurring within areas identified as search. k-Means clustering and first passage time had the highest rates of dives occurring outside identified search behavior. First passage time and hidden Markov models had the lowest rates of false positives, identifying fewer search areas with no dives. All behavioral annotation methods had advantages and drawbacks in terms of the complexity of analysis and ability to reflect prey capture events while minimizing the number of false positives and false negatives. We used these results, with consideration of analytical difficulty, to provide advice on the most appropriate methods for use where prey capture behavior is not available. This study highlights a need to critically assess and carefully choose a behavioral annotation method suitable for the research question being addressed, or resulting species management frameworks established.

Elastic K-means using posterior probability

The widely used K-means clustering is a hard clustering algorithm. Here we propose a Elastic K-means clustering model (EKM) using posterior probability with soft capability where each data point can belong to multiple clusters fractionally and show the benefit of proposed Elastic K-means. Furthermore, in many applications, besides vector attributes information, pairwise relations (graph information) are also available. Thus we integrate EKM with Normalized Cut graph clustering into a single clustering formulation. Finally, we provide several useful matrix inequalities which are useful for matrix formulations of learning models. Based on these results, we prove the correctness and the convergence of EKM algorithms. Experimental results on six benchmark datasets demonstrate the effectiveness of proposed EKM and its integrated model.

Path segmentation for beginners: an overview of current methods for detecting changes in animal movement patterns

Increased availability of high-resolution movement data has led to the development of numerous methods for studying changes in animal movement behavior. Path segmentation methods provide basics for detecting movement changes and the behavioral mechanisms driving them. However, available path segmentation methods differ vastly with respect to underlying statistical assumptions and output produced. Consequently, it is currently difficult for researchers new to path segmentation to gain an overview of the different methods, and choose one that is appropriate for their data and research questions. Here, we provide an overview of different methods for segmenting movement paths according to potential changes in underlying behavior. To structure our overview, we outline three broad types of research questions that are commonly addressed through path segmentation: 1) the quantitative description of movement patterns, 2) the detection of significant change-points, and 3) the identification of underlying processes or 'hidden states'. We discuss advantages and limitations of different approaches for addressing these research questions using path-level movement data, and present general guidelines for choosing methods based on data characteristics and questions. Our overview illustrates the large diversity of available path segmentation approaches, highlights the need for studies that compare the utility of different methods, and identifies opportunities for future developments in path-level data analysis.