A decade of movement ecology
Rocío Joo, Simona Picardi, Matthew E. Boone, Thomas A. Clay, Samantha C. Patrick, Vilma S. Romero-Romero, Mathieu Basille
AA decade of movement ecology
Roc´ıo Joo , Simona Picardi , Matthew E. Boone , Thomas A. Clay , Samantha C.Patrick , Vilma S. Romero-Romero , and Mathieu Basille Department of Wildlife Ecology and Conservation, Fort Lauderdale Research and EducationCenter, University of Florida, Fort Lauderdale, FL, USA Department of Wildland Resources, Jack H. Berryman Institute, Utah State University, LoganUt 84322, USA School of Environmental Sciences, University of Liverpool, Liverpool, L69 3GP, UK Universidad de Lima, Peru * Corresponding author: Rocio Joo, rocio.joo@ufl.edu
Abstract
Movement is fundamental to life, shaping population dynamics, biodiver-sity patterns, and ecosystem structure. Recent advances in tracking technol-ogy have enabled fundamental questions about movement to be tackled, lead-ing to the development of the movement ecology framework (MEF), considereda milestone in the field [33]. The MEF introduced an integrative theory oforganismal movement, linking internal state, motion capacity and navigationcapacity to external factors. Here, a decade later, we investigated the currentstate of research in the field. Using a text mining approach on > Keywords: movement ecology paradigm, technology, text mining, biologging, in-terdisciplinarity 1 a r X i v : . [ q - b i o . P E ] M a y ntroduction: the rise of a field called movementecology Movement, defined as a change in position of an individual in time, has been stud-ied for millennia, from philosophical (Aristotle’s
De motu animalium
De motu musculorum > Snapshot: research topics in movement ecology
We screened all abstracts from 2009 until 2018 to identify 15 broad topics fromthe words used (via Latent Dirichlet Allocation; see Material and Methods). Wechose 15 topics as a reasonable number that would not be too large to prevent theinterpretation of all topics, or too small that the topics would be too general (seediscussion in the online Appendix section 3.1.2). These topics were, in descendingorder of prevalence:1)
Social interactions and dispersal , a broad topic encompassing interactionswith conspecifics or the environment, as well as group movement, changes inpopulation density and dynamics.2)
Movement models , encompassing any type of model (e.g. generalized linearmodel, model selection criterion, or even schematical models) that could beused to study dynamics, patterns, and populations.3)
Habitat selection , which encompasses choices in space use, influenced byresource availability or risks (e.g. natural predators or human disturbance),mainly in mammalian systems (Fig. 2).3)
Detection and data , focused on the collection of movement information andthe required technological devices. This topic is also mostly related to mammalstudies.5)
Home ranges , mostly focused on the identification of areas where animalslive and develop their activities, and the geographical extent of this area.6)
Aquatic systems , involving the study of aquatic species, particularly fish,their migration, reproductive behavior and habitat, mostly for managementpurposes.7)
Foraging in marine megafauna , consisting of foraging strategies and be-havior of marine top predators, mostly birds and mammals (Fig. 2).8)
Biomechanics , focused on body motion, swimming or flight power, and kine-matics.9)
Acoustic telemetry , used to monitor animal movement (mostly fish), or insome cases, effects of anthropogenic noise on animal behavior.10)
Experimental designs , which involve analyzing behavioral and movementresponses based on multiple stimuli, mostly on cattle and domestic animals.11)
Activity budgets , investigating—mostly using telemetry data—the effect ofenvironmental conditions on the time allocated to different activities.12)
Avian migration , encompassing migration routes, orientation and flightstrategies.13)
Sports , consisting of motion analysis of sports players for better performance.14)
Human activity patterns , mostly related to health and physical activity inchildren and adults, often sampled with accelerometers.15)
Breeding ecology , involving space use and movement corridors during breed-ing seasons; mostly, but not exclusively on turtles and whales.While Sports and Human activity patterns are not strictly ecological topics, theyare an integral part of the literature studying movement phenomena, and benefitfrom advances in movement ecology. For instance, works on Sports have tried to un-derstand why and how certain individual and collective behaviors emerge in games,using principles from ecological psychology (e.g. [39]), focused on the interdepen-dencies of humans and their environments [5]. Moreover, some studies related toHuman activity patterns were also inspired from animal studies (e.g. [42]).Social interactions and dispersal, Movement models, and Habitat selection re-mained the most relevant topics throughout the decade (Fig. 3). The prevalenceof Home ranges studies decreased over the years. In contrast, Sports has becomea more recurrent topic in the literature. The prevalence of the other topics hasremained relatively stable in time. The division into research topics has revealedsome distinction between marine and terrestrial realms, as four topics pertainedspecifically to breeding or foraging ecology in marine species.4 n = 345)(n = 233)(n = 290)(n = 122) (n = 200)(n = 384)(n = 227)(n = 106) (n = 535)(n = 323)(n = 191)(n = 159) (n = 134)(n = 268)(n = 146)
Sports Human activity patterns Breeding ecologyAcoustic telemetry Experimental designs Activity budgets Avian migrationHome ranges Aquatic systems Foraging in marine megafauna BiomechanicsSocial interactions and dispersal Movement models Habitat selection Detection and data0.000.250.500.750.000.250.500.750.000.250.500.750.000.250.500.75 T o t a l p r opo r t i on o f pape r s w i t h i n a t op i c TaxonMammalsFishBirdsHumansInsectsReptilesothersCrustaceansAmphibiansMollusks
Figure 2: For each topic, relative frequencies of papers studying each taxonomicalgroup. Only papers with more than 50% of association to each topic ( γ , see Materialsand Methods) are used for this graph.Figure 3: Time series of the relative prevalence of each topic every year. To improvereadability, the topics with the most pronounced increases and decreases have beenhighlighted by continuous and dashed lines, respectively.5 avigation(9.0%)Internalstate(49.0%)Externalfactors(77.3%) Movementpath l o c a t i o n ( x , y , z , t ) Motion(26.2%)
Figure 4: Representation of the components of the movement ecology frameworkand how much they were studied in the last decade: external factors, internal state,motion and navigation capacities, whose interactions result in the observed move-ment path. The size of each component box is proportional to the percentage ofpapers (in parentheses) tackling them irrespectful of whether they are only aboutthis component or in combination with another one. The latter is specified throughthe segments that join the components to the observed movement path. One fillcolor corresponds to papers that only studied one component, while two or morecolors correspond to papers that tackled two or three components, respectively (theones from those colors). The width of the segment is proportional to the percentageof papers that studied that combination (or single component). Only combinationscorresponding to >
5% of papers are shown; e.g. combinations involving navigationand papers studying navigation on its own had <
5% of papers each therefore theyare not shown in the graph.
The movement ecology framework
The MEF introduced by [33] consisted of four components: external factors (i.e. theset of environmental conditions that affect movement), internal state (i.e. the intrin-sic factors affecting motivation and readiness to move), navigation capacity (i.e. theset of traits enabling the individual to orient), and motion capacity (i.e. the setof traits enabling the individual to execute movement). The outcome of the in-teractions between these four components would be the observed movement path(plus observation errors). We found that, in the last decade, most studies tackledmovement in relation to external factors (77%), while a minority of them studiedthe three other components (49%, 26%, and 9%, for internal factors, motion, andnavigation capacity, respectively). For the most part, studies did not look into inter-actions between these components, except for external factors with internal states(25% of the studies; Fig. 4). Quite strikingly, this is the same overall pattern as inthe decade before (1999-2008; Appendix section 3.3.1).6igure 5: Proportion of papers in each year using each type of device. To im-prove readability, the devices with the most pronounced increases or decreases werehighlighted by continuous and dashed lines, respectively.
Tools for movement ecology
Technology has been a major driver of trends in data collection and scientific pub-lications in movement ecology. Past reviews have highlighted an increase in theamount and variety of tracking devices, which are becoming more affordable, withmore efficient battery capabilities, and reductions in size (see [38], [25], [21], andsurvey to movement ecologists in Appendix section 4). Here, we categorized track-ing device observations as accelerometer, acoustic telemetry, body condition mea-surements, encounter observations, GPS, light loggers, pressure data, radar, radiotelemetry, satellite, and video/image (details of these categories and the analysis arein Appendix section 3.4).Throughout the last decade, GPS has not only remained the most popular devicein movement studies, but its popularity in relation to other methods has increased(Fig. 5). This is likely due to the development of cheaper, smaller and more effi-cient devices, which make them a feasible option for small and medium-sized animals[25]. While in 2009 radio telemetry was as popular as GPS, later in the decade GPSseems to have increasingly replaced radio telemetry [1], which has been experiencinga substantial decrease in parallel. The use of accelerometers and video is becomingmore popular; the former allows for finer spatio-temporal resolution movement data(Fig. 5), opening avenues to exploring physiological aspects of movement like energyexpenditure [45], while the latter gives us an animal’s-eye view of its local environ-ment, providing information on visual cues used, foraging behavior and movementstrategies [37, 24].The increasing volume and diversity of movement data obtained through these7igure 6: Proportion of papers of each year using each software package. Shownare the five most mentioned software packages. The ones with the most pronouncedincreases or decreases were highlighted by continuous and dashed lines, respectively.tracking devices require appropriate software tools for data management, processing,and analysis [40, 23]. We evaluated the use of 33 software packages (see section 3.5of the Appendix for a full list), and the five most popular through the decade wereR, ArcGIS, Matlab, SPSS and SAS, in that order. Among those, R experienced aconstant and strong growth in the last 10 years, while usage of all others substantiallydecreased, making R an undisputed preference in the field (Fig. 6).In another study of ecology in general, the same pattern in reported R usage wasobserved [26]. According to both [26] and this study, the popularity of R ten yearsago was low (used by >
10% of the papers), while the majority of articles publishednowadays have reported its use, indicating a homogenization of not only movementecology but ecology in general towards R. This success is likely due to the fact that Roffers a free software environment to program and create new methods, share them,and improve them, facilitating transparency, collaboration, and reproducibility [28],and at the same time it can be extended with more than 50 specialized packages toprocess and analyze movement [23]. R also leverages other programming languages(e.g. C, python, Fortran, etc.) by allowing internal access to their use within an Rworkflow (and R syntax).In parallel with the development and improvement of tracking devices and soft-ware, there has been substantial progress in the number and sophistication of quan-titative methods for the study of movement (e.g. [3, 35, 2]). We investigated the useof statistical methods in the movement ecology literature (see Material and Meth-ods, and Appendix section 3.6, for more details). Most studies (68%) used, at theleast, generic statistical methods (i.e. with no explicit spatial, temporal or social in-teraction component in its definition) such as regression models. A smaller number857%) used at least one or more specialized methods, i.e. movement, non-movementspatiotemporal (e.g. spatiotemporal geostatistics), spatial (e.g. point process), timeseries, and social analysis methods (e.g. social networks). Our analysis revealsthat researchers are not necessarily using movement-specific techniques to analyzemovement (only 33% of the studies), and, in some cases (42%), not using spatial,temporal, or social analyses either.While the availability of movement data and associated software tools and meth-ods are increasing (see a summary list in [10]), the proportion of papers usingmovement-specific analytical methods does not show the same pattern (Fig. 7).Actually, the proportion of usage of generic methods is increasing. In addition, andbased on a trigram analysis, we found that the most common methods were linearmixed models (Table in section 3.6.1 of the Appendix).This raises the question: why were the majority of papers not using movement-specific methods? Certainly, not all studies require movement-specific methods; thechoice of method should depend on the research question, assumptions, and data.Another reason for the use of non-movement methods could be related to manyscientists coming-of-age in a time when autocorrelation in movement was consid-ered a nuisance, or they do not possess the quantitative skills necessary to use thesemethods. Movement is a complex process, and in most cases, statistically noisy,nonlinear, and spatially and temporally correlated (28). Interdisciplinary work be-tween ecologists and statisticians to “decomplexify” movement models (either mak-ing them more simple or usable for different datasets and situations) may still be achallenge to overcome [44]. Moreover, as we intensify data collection and processing,the use of movement models – for a statistical representation of movement and forpopulation-level inference – can be expected to increase.
Open questions for the future of movement ecology
Technology has undeniably been driving research in movement ecology in the pastdecade. With access to numerous and diverse tracking data [4], and tools for dataprocessing and analysis [10, 23], researchers have been able to sample spatiotemporalbehavior and changes in physiology, to investigate subjects like social interactions,habitat selection, foraging behavior, physiological performance, and migration; top-ics that were revealed by our text mining analysis. However, the technological ad-vances have not structurally changed the field of movement ecology: none of thoseresearch topics are new, and, we have not moved towards the integrative study ofmovement advised by the MEF.The movement ecology framework was a revolutionary idea: an integrating vi-sion of the study of movement, represented by the interaction between the fourcomponents of the framework. As argued by [33], it is only through combining dif-ferent components of the framework that we can gain a mechanistic understandingof movement, from the neurological and physiological drivers to the life-history andevolutionary consequences. While it has been recognized as a seminal, if not themost influential, publication in the field (with > Materials and methods
We selected scientific peer-reviewed papers in English that studied the voluntarymovement of one or more living individuals. We used Web of Science (WoS) asa search engine. Since very few papers mention “movement ecology” in their ab-stracts, titles and keywords, we did not use “movement ecology” as a search phrase.After a detailed testing phase, we came up with search terms within four groups ofwords: behavior, movement (e.g. motion, moving), biologging (e.g. telemetry, gps)and individuals (e.g. animal, human; we focused our efforts on Animalia). To bequalified, papers needed to use words from at least 3 of these groups in their ab-stract, keywords or title. Also, papers studying movement of objects other thanwhole organisms (e.g. cell, neuron), were filtered out. See more details on the searchterms and their quality control in section 2 of the Appendix. More than 8 thousand(8,007) papers met our criteria. Results from the WoS (title, keywords, abstracts,authors, DOI, etc.) on these 8,007 papers were downloaded and used for most ofthe analyses. In addition, we used the fulltext package [11] in [34], using Elsevier,Springer, Scopus, Wiley, BMC, and PLOS one API keys to download full texts of4,037 papers. Finally, using an automatic detection algorithm (see section 2.3 of theAppendix for a description), 3,674 “Materials and Methods” sections were extractedfrom this set of papers.
Topic analysis
Topics were not defined a priori . Instead, we fitted Latent Dirichlet Allocation(LDA) models to the abstracts [6]. LDAs are basically three-level hierarchicalBayesian models for documents (in our case, abstracts). Here we assumed thatthere was a fixed number of latent or hidden topics behind the abstracts, and thatthe choice of words in the abstracts were related to the topics the authors wereaddressing. Thus, an abstract would have been composed of one or more topics,and a topic would have been composed of a mixture of words. The probability of11 word appearing in an abstract depended on the topic the abstract was adressing.Here we used the LDA model with variational EM estimation [41] implemented inthe topicmodels package [17]. More details about the practical aspects of LDAmodeling and a short discussion on the number of topics can be found in section 3.1of the Appendix.From the fitted LDA model, we obtained 1) for each topic, the probability ofthe topic being referred to in each document (denoted by γ ), and 2) for each word,the probability of appearing in a document given the presence of a certain topic(denoted by β ). The β values were thus a proxy of the importance of a word in atopic. They were used to interpret and label each topic, and to create wordcloudsfor each topic, shown in section 3.1 of the Appendix. The sum of γ values for eachtopic served as proxies of the “prevalence” of the topic relative to all other topicsand were used to rank them. Taxonomical identification
To identify the taxonomy of the individuals studied in the papers, the ITIS (Inte-grated Taxonomic Information System) database ( ) was used to detect names of any animal species (kingdom Animalia) that werementioned in the abstracts, titles and keywords. We screened these sections for latinand common (i.e., vernacular) names of species (both singular and plural), as wellas common names of higher taxonomic levels such as orders and families. Afterhaving identified any taxon mentioned in a paper, we summarized taxa at the Classlevel (except for superclasses Osteichthyes and Chondrichthyes which we mergedinto a single group labeled Fish, and for classes within the phylum Mollusca andthe subphylum Crustacea which we considered collectively). Thus, each paper wasclassified as focusing on one or more class-like groups; for example - mammals, birds,insects, etc. For the purpose of our analysis, we kept humans as a separate categoryand did not count them within Class Mammalia. See section 3.2 of the Appendixfor more details.
Framework and tools
To assess the study of the different components of the movement ecology framework,we built what we call here a “dictionary”. A dictionary is composed of words andtheir meanings. Here, the words of interest were the components of the framework(i.e. internal state, external factor, motion, and navigation), and their meanings werethe terms potentially used in the abstracts to refer to the study of each component.For example, terms like “memory”, “sensory information”, “path integration”, or“orientation” were used to identify the study of navigation. Similarly, the devices,software, and statistical methods used were also assessed through dictionary ap-proaches. More details, including quality control of the dictionaries, in sections 3.3to 3.6 of the Appendix.
Access to data and codes
We provide details on all data processing and analyses at https://rociojoo.github.io/mov-eco-review/ , from descriptions of word search on Web of Science12nd scripts to download the papers, up to the codes to reproduce every single plotin this manuscript. The website, hosted in the https://github.com/rociojoo/mov-eco-review repository, works as an online Appendix to this manuscript. Theauthors can be directly contacted for further development and questions about thedataset, which has not been released to respect Text and Data Mining rights of thepublishers.
Acknowledgments
The authors would like to thank Susana Clusella-Trullas for fruitful exchanges aboutinternal states and physiology. Trey Shelton, from UF library, gave advice and guid-ance about TDM rights and obtaining APIs from publishers, which was very usefulat early stages of this study. We are also grateful to Luis Cajachahua Espinoza,for his help to explore scrapping possibilities at the very beginning of this work.RJ, MEB, TAC, SCP and MB were funded by a Human Frontier Science ProgramYoung Investigator Grant (SeabirdSound; RGY0072/2017).13 eferences [1] B. M. Allan, D. G. Nimmo, D. Ierodiaconou, J. VanDerWal, L. P. Koh, andE. G. Ritchie. Futurecasting ecological research: the rise of technoecology.
Ecosphere , 9(5):e02163, May 2018.[2] M. Auger-Mth, K. Newman, D. Cole, F. Empacher, R. Gryba, A. A. King,V. Leos-Barajas, J. M. Flemming, A. Nielsen, G. Petris, and L. Thomas. Anintroduction to state-space modeling of ecological time series. arXiv:2002.02001[q-bio, stat] , Feb. 2020. arXiv: 2002.02001.[3] T. Avgar, J. R. Potts, M. A. Lewis, and M. S. Boyce. Integrated step selectionanalysis: bridging the gap between resource selection and animal movement.
Methods in Ecology and Evolution , 7(5):619–630, 2016.[4] M. Baratchi, N. Meratnia, P. J. M. Havinga, A. K. Skidmore, and B. A. G.Toxopeus. Sensing Solutions for Collecting Spatio-Temporal Data for WildlifeMonitoring Applications: A Review.
Sensors (Basel, Switzerland) , 13(5):6054–6088, May 2013.[5] R. G. Barker et al. Ecological psychology: Concepts and methods for studyingthe environment of human behavior. 1968.[6] D. M. Blei, A. Y. Ng, and M. I. Jordan. Latent Dirichlet Allocation.
Journalof Machine Learning Research , 3(Jan):993–1022, 2003.[7] L. J. Bracken and E. A. Oughton. What do you mean? The importance oflanguage in developing interdisciplinary research.
Transactions of the Instituteof British Geographers , 31(3):371–382, Sept. 2006.[8] G. D. Brewer. The Challenges of Interdisciplinarity.
Policy Sciences , 32(4):327–337, 1999.[9] L. Bromham, R. Dinnage, and X. Hua. Interdisciplinary research has consis-tently lower funding success.
Nature , 534(7609):684–687, June 2016.[10] L. Brger. EDITORIAL: Stuck in motion? Reconnecting questions and tools inmovement ecology.
Journal of Animal Ecology , 85(1):5–10, 2016.[11] S. Chamberlain. fulltext: Full Text of ’Scholarly’ Articles Across Many DataSources , 2019. R package version 1.3.0.[12] S. J. Cooke, S. G. Hinch, M. Wikelski, R. D. Andrews, L. J. Kuchel, T. G.Wolcott, and P. J. Butler. Biotelemetry: a mechanistic approach to ecology.
Trends in ecology & evolution , 19(6):334–43, June 2004.[13] F. J. Dyson. Is Science Mostly Driven by Ideas or by Tools?
Science ,338(6113):1426–1427, Dec. 2012.[14] R. Frodeman, J. T. Klein, and R. C. D. S. Pacheco.
The Oxford Handbookof Interdisciplinarity . Oxford University Press, Jan. 2017. Google-Books-ID:d1XjDQAAQBAJ. 1415] P. Gloaguen, S. Mah´evas, E. Rivot, M. Woillez, J. Guitton, Y. Vermard, andM. P. Etienne. An autoregressive model to describe fishing vessel movementand activity.
Environmetrics , 26(1):17–28, nov 2015.[16] S. Goossens, N. Wybouw, T. Van Leeuwen, and D. Bonte. The physiology ofmovement.
Movement Ecology , 8(1):5, Feb. 2020.[17] B. Gr¨un and K. Hornik. topicmodels: An R package for fitting topic models.
Journal of Statistical Software , 40(13):1–30, 2011.[18] E. Gurarie, C. Bracis, M. Delgado, T. D. Meckley, I. Kojola, and C. M. Wagner.What is the animal doing? Tools for exploring behavioural structure in animalmovements.
Journal of Animal Ecology , 85(1):69–84, 2016.[19] E. Haeckel.
Generelle Morphologie der Organismen. Allgemeine Grundz¨ugeder organischen Formen-Wissenschaft, mechanisch begr¨undet durch die von C.Darwin reformirte Descendenz-Theorie, etc , volume 2. 1866.[20] P. H¨anggi and F. Marchesoni. Introduction: 100 years of Brownian motion.
Chaos , 15(2):26101, June 2005.[21] R. Harcourt, A. M. Martins Sequeira, X. Zhang, F. Rouquet, K. Komatsu,M. Heupel, C. R. McMahon, F. G. Whoriskey, M. Meekan, G. Carroll, S. Brodie,C. Simpfendorfer, M. A. Hindell, I. D. Jonsen, D. P. Costa, B. A. Block, M. M.Muelbert, B. E. Woodward, M. J. Weises, K. Aarestrup, M. Biuw, L. Boehme,S. J. Bograd, D. Cazau, J.-B. Charrassin, S. Cooke, P. D. Cowley, P. J. NicoDe Bruyn, T. J. Du Dot, C. M. Duarte, V. M. Eguluz, L. Ferreira, J. Fernndez-Garcia, K. Goetz, Y. Goto, C. Guinet, M. Hammill, G. C. Hays, E. Hazen,L. Huckstadt, C. Huveneers, S. Iverson, S. A. Jaaman, K. Kittiwattanawong,K. Kovacs, C. Lydersen, T. Moltmann, M. Naruoka, L. Phillips, B. Picard,N. Queiroz, G. Reverdin, K. Sato, D. W. Sims, E. B. Thorstad, M. Thums,A. M. Treasure, A. Trites, G. D. Williams, Y. Yonehara, and M. A. Fedak.Animal-borne telemetry : an integral component of the ocean observing toolkit.
Frontiers in Marine Science , 6:326, June 2019.[22] M. Holyoak, R. Casagrandi, R. Nathan, E. Revilla, and O. Spiegel. Trends andmissing parts in the study of movement ecology.
Proceedings of the NationalAcademy of Sciences of the United States of America , 105(49):19060–5, Dec.2008.[23] R. Joo, M. E. Boone, T. A. Clay, S. C. Patrick, S. Clusella-Trullas, andM. Basille. Navigating through the r packages for movement.
Journal of AnimalEcology , 89:248–267, 2020.[24] S. A. Kane, A. H. Fulton, and L. J. Rosenthal. When hawks attack: animal-borne video studies of goshawk pursuit and prey-evasion strategies.
Journal ofExperimental Biology , 218(2):212–222, Jan. 2015.[25] R. Kays, M. C. Crofoot, W. Jetz, and M. Wikelski. Terrestrial animal trackingas an eye on life and planet.
Science , 348(6240):aaa2478, 2015.1526] J. Lai, C. J. Lortie, R. A. Muenchen, J. Yang, and K. Ma. Evaluating thepopularity of R in ecology.
Ecosphere , 10(1):e02567, 2019.[27] S. K. Lowerre-Barbieri, R. Kays, J. T. Thorson, and M. Wikelski. The oceansmovescape: fisheries management in the bio-logging decade (20182028).
ICESJournal of Marine Science , 76(2):477–488, Apr. 2019.[28] J. S. S. Lowndes, B. D. Best, C. Scarborough, J. C. Afflerbach, M. R. Frazier,C. C. O’Hara, N. Jiang, and B. S. Halpern. Our path to better science in lesstime using open data science tools.
Nature Ecology & Evolution , 1(6):0160,2017.[29] M. Marzano, D. N. Carss, and S. Bell. Working to Make InterdisciplinarityWork: Investing in Communication and Interpersonal Relationships.
Journalof Agricultural Economics , 57(2):185–197, 2006.[30] K. McMahon, K.-j. van Dijk, L. Ruiz-Montoya, G. A. Kendrick, S. L. Krauss,M. Waycott, J. Verduin, R. Lowe, J. Statton, E. Brown, and C. Duarte. Themovement ecology of seagrasses.
Proceedings of the Royal Society B: BiologicalSciences , 281(1795):20140878, Nov. 2014.[31] H. J. Miller, S. Dodge, J. Miller, and G. Bohrer. Towards an integrated scienceof movement: converging research on animal movement ecology and humanmobility science.
International Journal of Geographical Information Science ,33(5):855–876, May 2019.[32] K. Morelle, F. Lehaire, and P. Lejeune. Is wild boar heading towards movementecology? A review of trends and gaps.
Wildlife Biology , 20(4):196–205, Aug.2014.[33] R. Nathan, W. M. Getz, E. Revilla, M. Holyoak, R. Kadmon, D. Saltz, andP. E. Smouse. A movement ecology paradigm for unifying organismal movementresearch.
PNAS , 105(49):19052–19059, 2008.[34] R Core Team.
R: A Language and Environment for Statistical Computing . RFoundation for Statistical Computing, Vienna, Austria, 2018.[35] D. P. Seidel, E. Dougherty, C. Carlson, and W. M. Getz. Ecological metricsand methods for GPS movement data.
International Journal of GeographicalInformation Science , 32(11):2272–2293, Nov. 2018.[36] R. K. Swihart and N. A. Slade. Testing For Independence of Observations inAnimal Movements.
Ecology , 66(4):1176–1184, 1985.[37] A. Thiebault, R. Mullers, P. Pistorius, M. A. Meza-Torres, L. Dubroca,D. Green, and Y. Tremblay. From colony to first patch: Processes of preysearching and social information in Cape Gannets.
The Auk: OrnithologicalAdvances , 131(4):595–609, Aug. 2014.[38] M. Thums, J. Fernndez-Gracia, A. M. M. Sequeira, V. M. Eguluz, C. M. Duarte,and M. G. Meekan. How Big Data Fast Tracked Human Mobility Research andthe Lessons for Animal Movement Ecology.
Frontiers in Marine Science , 5:21,2018. 1639] B. Travassos, K. Davids, D. Araujo, and P. T. Esteves. Performance analysis inteam sports: Advances from an Ecological Dynamics approach.
InternationalJournal of Performance Analysis in Sport , 13(1):83–95, 2013.[40] F. Urbano, F. Cagnacci, C. Calenge, H. Dettki, A. Cameron, and M. Neteler.Wildlife tracking data management: a new vision.
Philosophical Transactionsof the Royal Society of London, Series B: Biological Sciences , 365:2177–2185,2010.[41] M. J. Wainwright and M. I. Jordan. Graphical Models, Exponential Fami-lies, and Variational Inference.
Foundations and Trends in Machine Learning ,1(12):1–305, Nov. 2008.[42] N. S. Wasenius, K. P. Grattan, A. L. J. Harvey, N. Barrowman, G. S. Goldfield,and K. B. Adamo. Maternal gestational weight gain and objectively measuredphysical activity among offspring.
PLOS ONE , 12(6):e0180249, June 2017.[43] P. Weingart. Interdisciplinarity: The paradoxical discourse. In P. Weingartand N. Stehr, editors,
Practising interdisciplinarity , chapter 2, pages 25–41.University of Toronto Press Toronto, Toronto Buffalo London, 2000.[44] H. J. Williams, L. A. Taylor, S. Benhamou, A. I. Bijleveld, T. A. Clay, S. d.Grissac, U. Demar, H. M. English, N. Franconi, A. GmezLaich, R. C. Griffiths,W. P. Kay, J. M. Morales, J. R. Potts, K. F. Rogerson, C. Rutz, A. Spelt,A. M. Trevail, R. P. Wilson, and L. Brger. Optimizing the use of biologgers formovement ecology research.
Journal of Animal Ecology , 89(1):186–206, 2020.[45] R. P. Wilson, L. Brger, M. D. Holton, D. M. Scantlebury, A. GmezLaich,F. Quintana, F. Rosell, P. M. Graf, H. Williams, R. Gunner, L. Hopkins,N. Marks, N. R. Geraldi, C. M. Duarte, R. Scott, M. S. Strano, H. Robotka,C. Eizaguirre, A. Fahlman, and E. L. C. Shepard. Estimates for energy expen-diture in free-living animals using acceleration proxies: A reappraisal.
Journalof Animal Ecology , 89(1):161–172, 2020.[46] Y. U. Zheng. Trajectory Data Mining : An Overview.