1
|
Resheff YS, Bensch HM, Zöttl M, Harel R, Matsumoto-Oda A, Crofoot MC, Gomez S, Börger L, Rotics S. How to treat mixed behavior segments in supervised machine learning of behavioural modes from inertial measurement data. MOVEMENT ECOLOGY 2024; 12:44. [PMID: 38858733 PMCID: PMC11165886 DOI: 10.1186/s40462-024-00485-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 02/21/2024] [Accepted: 06/06/2024] [Indexed: 06/12/2024]
Abstract
The application of supervised machine learning methods to identify behavioural modes from inertial measurements of bio-loggers has become a standard tool in behavioural ecology. Several design choices can affect the accuracy of identifying the behavioural modes. One such choice is the inclusion or exclusion of segments consisting of more than a single behaviour (mixed segments) in the machine learning model training data. Currently, the common practice is to ignore such segments during model training. In this paper we tested the hypothesis that including mixed segments in model training will improve accuracy, as the model would perform better in identifying them in the test data. We test this hypothesis using a series of data simulations on four datasets of accelerometer data coupled with behaviour observations, obtained from four study species (Damaraland mole-rats, meerkats, olive baboons, polar bears). Results show that when a substantial proportion of the test data are mixed behaviour segments (above ~ 10%), including mixed segments in machine learning model training improves the accuracy of classification. These results were consistent across the four study species, and robust to changes in segment length, sample size, and degree of mixture within the mixed segments. However, we also find that in some cases (particularly in baboons) models trained with mixed segments show reduced accuracy in classifying test data containing only single behaviour (pure) segments, compared to models trained without mixed segments. Based on these results, we recommend that when the classification model is expected to deal with a substantial proportion of mixed behaviour segments (> 10%), it is beneficial to include them in model training, otherwise, it is unnecessary but also not harmful. The exception is when there is a basis to assume that the training data contains a higher rate of mixed segments than the actual (unobserved) data to be classified-such a situation may occur particularly when training data are collected in captivity and used to classify data from the wild. In this case, excess inclusion of mixed segments in training data should probably be avoided.
Collapse
Affiliation(s)
- Yehezkel S Resheff
- Hebrew University Business School, The Hebrew University of Jerusalem, Jerusalem, Israel.
| | - Hanna M Bensch
- Department of Biology and Environmental Science, Centre for Ecology and Evolution in Microbial Model Systems (EEMIS), Linnaeus University, 391 82, Kalmar, Sweden
- Kalahari Research Centre, Kuruman River Reserve, Van Zylsrus, South Africa
| | - Markus Zöttl
- Department of Biology and Environmental Science, Centre for Ecology and Evolution in Microbial Model Systems (EEMIS), Linnaeus University, 391 82, Kalmar, Sweden
- Kalahari Research Centre, Kuruman River Reserve, Van Zylsrus, South Africa
| | - Roi Harel
- Department for the Ecology of Animal Societies, Max Planck Institute of Animal Behavior, Constance, Germany
- Department of Biology, University of Konstanz, Constance, Germany
- Centre for the Advanced Study of Collective Behaviour, University of Konstanz, Constance, Germany
- Mpala Research Centre, Nanyuki, Kenya
| | - Akiko Matsumoto-Oda
- Graduate School of Tourism Sciences, University of the Ryukyus, Nakagami, Okinawa, Japan
| | - Margaret C Crofoot
- Department for the Ecology of Animal Societies, Max Planck Institute of Animal Behavior, Constance, Germany
- Department of Biology, University of Konstanz, Constance, Germany
- Centre for the Advanced Study of Collective Behaviour, University of Konstanz, Constance, Germany
- Mpala Research Centre, Nanyuki, Kenya
| | - Sara Gomez
- Department of Biosciences, Swansea University, Swansea, Wales, UK
| | - Luca Börger
- Department of Biosciences, Swansea University, Swansea, Wales, UK
| | - Shay Rotics
- School of Zoology, Faculty of Life Sciences, and the Steinhardt Museum of Natural History, Tel Aviv University, Tel Aviv, Israel
- Kuruman River Reserve, Kalahari Research Centre, Van Zylsrus, South Africa
| |
Collapse
|
2
|
Berry P, Dammhahn M, Hauptfleisch M, Hering R, Jansen J, Kraus A, Blaum N. African dryland antelope trade-off behaviours in response to heat extremes. Ecol Evol 2024; 14:e11455. [PMID: 38855312 PMCID: PMC11157150 DOI: 10.1002/ece3.11455] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/08/2024] [Revised: 04/30/2024] [Accepted: 05/07/2024] [Indexed: 06/11/2024] Open
Abstract
Climate change is predicted to narrow the prescriptive zone of dryland species, potentially leading to behavioural modifications with fitness consequences. This study explores the behavioural responses of three widespread African antelope species-springbok, kudu and eland-to extreme heat in a dryland savanna. We classified the behaviour of 29 individuals during the hot, dry season on the basis of accelerometer data using supervised machine learning and analysed the impact of afternoon heat on behaviour-specific time allocation and overall dynamic body acceleration (ODBA), a proxy for energy expenditure, along with compensatory changes over the 24-hour cycle. Extreme afternoon heat reduced feeding time in all three antelope species, increased ruminating and resting time, while only minimally affecting walking time. With rising heat, all three species reduced ODBA on feeding, while eland reduced and kudu increased ODBA on walking. Diel responses in behaviour differed between species, but were generally characterised by daytime reductions in feeding and increases in ruminating or resting on hot days compared to cool days. While antelope compensated for heat-driven behavioural change over the 24-hour cycle in some cases, significant differences persisted in others, including reduced feeding and increased rumination and resting. The impact of heat on antelope behaviour reveals trade-offs between feeding and thermoregulation, as well as between feeding and rumination, the latter suggesting a strategy to enhance nutrient uptake through increased digestive efficiency, while the walking response suggests narrow constraints between cost and necessity. Our findings suggest that heat influences both behaviour-specific time allocation and energy expenditure. Altered diel behaviour patterns and incomplete compensation over the 24-hour cycle point to fitness consequences. The need to prioritise thermoregulation over feeding is likely to narrow the prescriptive zone of these dryland antelope.
Collapse
Affiliation(s)
- Paul Berry
- Plant Ecology and Nature Conservation, Institute of Biochemistry and BiologyUniversity of PotsdamPotsdamGermany
| | - Melanie Dammhahn
- Behavioural Biology, Institute for Neuro‐ and Behavioural Biology (INVB)University of MünsterMünsterGermany
| | - Morgan Hauptfleisch
- Research DirectorateNamibia Nature FoundationWindhoekNamibia
- Unit for Environmental Sciences and ManagementNorth West UniversityPotchefstroomNort West ProvinceSouth Africa
- Biodiversity Research CentreNamibia University of Science and TechnologyWindhoekNamibia
| | - Robert Hering
- Ecology/Macroecology, Institute of Biochemsitry and BiologyUniversity of PotsdamPotsdamGermany
| | - Jakob Jansen
- Plant Ecology and Nature Conservation, Institute of Biochemistry and BiologyUniversity of PotsdamPotsdamGermany
| | - Anna Kraus
- Plant Ecology and Nature Conservation, Institute of Biochemistry and BiologyUniversity of PotsdamPotsdamGermany
| | - Niels Blaum
- Plant Ecology and Nature Conservation, Institute of Biochemistry and BiologyUniversity of PotsdamPotsdamGermany
| |
Collapse
|
3
|
Pohle J, Signer J, Eccard JA, Dammhahn M, Schlägel UE. How to account for behavioral states in step-selection analysis: a model comparison. PeerJ 2024; 12:e16509. [PMID: 38426131 PMCID: PMC10903358 DOI: 10.7717/peerj.16509] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/19/2023] [Accepted: 11/01/2023] [Indexed: 03/02/2024] Open
Abstract
Step-selection models are widely used to study animals' fine-scale habitat selection based on movement data. Resource preferences and movement patterns, however, often depend on the animal's unobserved behavioral states, such as resting or foraging. As this is ignored in standard (integrated) step-selection analyses (SSA, iSSA), different approaches have emerged to account for such states in the analysis. The performance of these approaches and the consequences of ignoring the states in step-selection analysis, however, have rarely been quantified. We evaluate the recent idea of combining iSSAs with hidden Markov models (HMMs), which allows for a joint estimation of the unobserved behavioral states and the associated state-dependent habitat selection. Besides theoretical considerations, we use an extensive simulation study and a case study on fine-scale interactions of simultaneously tracked bank voles (Myodes glareolus) to compare this HMM-iSSA empirically to both the standard and a widely used classification-based iSSA (i.e., a two-step approach based on a separate prior state classification). Moreover, to facilitate its use, we implemented the basic HMM-iSSA approach in the R package HMMiSSA available on GitHub.
Collapse
Affiliation(s)
- Jennifer Pohle
- Institute of Biochemistry and Biology, University of Potsdam, Potsdam, Germany
| | - Johannes Signer
- Wildlife Sciences, Faculty of Forest Sciences and Forest Ecology, University of Goettingen, Göttingen, Germany
| | - Jana A. Eccard
- Animal Ecology, Institute of Biochemistry and Biology, University of Potsdam, Potsdam, Germany
| | - Melanie Dammhahn
- Department of Behavioural Biology, University of Münster, Münster, Germany
| | - Ulrike E. Schlägel
- Institute of Biochemistry and Biology, University of Potsdam, Potsdam, Germany
| |
Collapse
|
4
|
Sur M, Hall JC, Brandt J, Astell M, Poessel SA, Katzner TE. Supervised versus unsupervised approaches to classification of accelerometry data. Ecol Evol 2023; 13:e10035. [PMID: 37206689 PMCID: PMC10191777 DOI: 10.1002/ece3.10035] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/04/2022] [Revised: 03/31/2023] [Accepted: 04/05/2023] [Indexed: 05/21/2023] Open
Abstract
Sophisticated animal-borne sensor systems are increasingly providing novel insight into how animals behave and move. Despite their widespread use in ecology, the diversity and expanding quality and quantity of data they produce have created a need for robust analytical methods for biological interpretation. Machine learning tools are often used to meet this need. However, their relative effectiveness is not well known and, in the case of unsupervised tools, given that they do not use validation data, their accuracy can be difficult to assess. We evaluated the effectiveness of supervised (n = 6), semi-supervised (n = 1), and unsupervised (n = 2) approaches to analyzing accelerometry data collected from critically endangered California condors (Gymnogyps californianus). Unsupervised K-means and EM (expectation-maximization) clustering approaches performed poorly, with adequate classification accuracies of <0.8 but very low values for kappa statistics (range: -0.02 to 0.06). The semi-supervised nearest mean classifier was moderately effective at classification, with an overall classification accuracy of 0.61 but effective classification only of two of the four behavioral classes. Supervised random forest (RF) and k-nearest neighbor (kNN) machine learning models were most effective at classification across all behavior types, with overall accuracies >0.81. Kappa statistics were also highest for RF and kNN, in most cases substantially greater than for other modeling approaches. Unsupervised modeling, which is commonly used for the classification of a priori-defined behaviors in telemetry data, can provide useful information but likely is instead better suited to post hoc definition of generalized behavioral states. This work also shows the potential for substantial variation in classification accuracy among different machine learning approaches and among different metrics of accuracy. As such, when analyzing biotelemetry data, best practices appear to call for the evaluation of several machine learning techniques and several measures of accuracy for each dataset under consideration.
Collapse
Affiliation(s)
- Maitreyi Sur
- Conservation Science Global, Inc.West Cape MayNew JerseyUSA
- Present address:
Radboud Institute for Biological and Environmental Sciences (RIBES)Radboud UniversityNijmegenThe Netherlands
| | - Jonathan C. Hall
- Department of BiologyEastern Michigan UniversityYpsilantiMichiganUSA
| | - Joseph Brandt
- U.S. Fish and Wildlife Service, Hopper Mountain National Wildlife Refuge ComplexVenturaCaliforniaUSA
| | - Molly Astell
- U.S. Fish and Wildlife Service, Hopper Mountain National Wildlife Refuge ComplexVenturaCaliforniaUSA
- Department of BiologyBoise State UniversityBoiseIdahoUSA
| | - Sharon A. Poessel
- U.S. Geological Survey, Forest and Rangeland Ecosystem Science CenterBoiseIdahoUSA
| | - Todd E. Katzner
- U.S. Geological Survey, Forest and Rangeland Ecosystem Science CenterBoiseIdahoUSA
| |
Collapse
|
5
|
Brennan JR, Menendez HM, Ehlert K, Tedeschi LO. ASAS-NANP symposium: mathematical modeling in animal nutrition-Making sense of big data and machine learning: how open-source code can advance training of animal scientists. J Anim Sci 2023; 101:skad317. [PMID: 37997926 PMCID: PMC10664406 DOI: 10.1093/jas/skad317] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/09/2023] [Accepted: 09/21/2023] [Indexed: 11/25/2023] Open
Abstract
Advancements in precision livestock technology have resulted in an unprecedented amount of data being collected on individual animals. Throughout the data analysis chain, many bottlenecks occur, including processing raw sensor data, integrating multiple streams of information, incorporating data into animal growth and nutrition models, developing decision support tools for producers, and training animal science students as data scientists. To realize the promise of precision livestock management technologies, open-source tools and tutorials must be developed to reduce these bottlenecks, which are a direct result of the tremendous time and effort required to create data pipelines from scratch. Open-source programming languages (e.g., R or Python) can provide users with tools to automate many data processing steps for cleaning, aggregating, and integrating data. However, the steps from data collection to training artificial intelligence models and integrating predictions into mathematical models can be tedious for those new to statistical programming, with few examples pertaining to animal science. To address this issue, we outline how open-source code can help overcome many of the bottlenecks that occur in the era of big data and precision livestock technology, with an emphasis on how routine use and publication of open-source code can help facilitate training the next generation of animal scientists. In addition, two case studies are presented with publicly available data and code to demonstrate how open-source tutorials can be utilized to streamline data processing, train machine learning models, integrate with animal nutrition models, and facilitate learning. The National Animal Nutrition Program focuses on providing research-based data on animal performance and feeding strategies. Open-source data and code repositories with examples specific to animal science can help create a reinforcing mechanism aimed at advancing animal science research.
Collapse
Affiliation(s)
- Jameson R Brennan
- Department of Animal Science, South Dakota State University West River Research and Extension Center, Rapid City, SD 57701, USA
| | - Hector M Menendez
- Department of Animal Science, South Dakota State University West River Research and Extension Center, Rapid City, SD 57701, USA
| | - Krista Ehlert
- Department of Natural Resource Management, South Dakota State University West River Research and Extension Center, Rapid City, SD 57701, USA
| | - Luis O Tedeschi
- Department of Animal Science, Texas A&M University, College Station, TX 77843-2471, USA
| |
Collapse
|
7
|
Gunner RM, Holton MD, Scantlebury DM, Hopkins P, Shepard ELC, Fell AJ, Garde B, Quintana F, Gómez-Laich A, Yoda K, Yamamoto T, English H, Ferreira S, Govender D, Viljoen P, Bruns A, van Schalkwyk OL, Cole NC, Tatayah V, Börger L, Redcliffe J, Bell SH, Marks NJ, Bennett NC, Tonini MH, Williams HJ, Duarte CM, van Rooyen MC, Bertelsen MF, Tambling CJ, Wilson RP. How often should dead-reckoned animal movement paths be corrected for drift? ANIMAL BIOTELEMETRY 2021; 9:43. [PMID: 34900262 PMCID: PMC7612089 DOI: 10.1186/s40317-021-00265-9] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 06/02/2021] [Accepted: 09/25/2021] [Indexed: 05/19/2023]
Abstract
BACKGROUND Understanding what animals do in time and space is important for a range of ecological questions, however accurate estimates of how animals use space is challenging. Within the use of animal-attached tags, radio telemetry (including the Global Positioning System, 'GPS') is typically used to verify an animal's location periodically. Straight lines are typically drawn between these 'Verified Positions' ('VPs') so the interpolation of space-use is limited by the temporal and spatial resolution of the system's measurement. As such, parameters such as route-taken and distance travelled can be poorly represented when using VP systems alone. Dead-reckoning has been suggested as a technique to improve the accuracy and resolution of reconstructed movement paths, whilst maximising battery life of VP systems. This typically involves deriving travel vectors from motion sensor systems and periodically correcting path dimensions for drift with simultaneously deployed VP systems. How often paths should be corrected for drift, however, has remained unclear. METHODS AND RESULTS Here, we review the utility of dead-reckoning across four contrasting model species using different forms of locomotion (the African lion Panthera leo, the red-tailed tropicbird Phaethon rubricauda, the Magellanic penguin Spheniscus magellanicus, and the imperial cormorant Leucocarbo atriceps). Simulations were performed to examine the extent of dead-reckoning error, relative to VPs, as a function of Verified Position correction (VP correction) rate and the effect of this on estimates of distance moved. Dead-reckoning error was greatest for animals travelling within air and water. We demonstrate how sources of measurement error can arise within VP-corrected dead-reckoned tracks and propose advancements to this procedure to maximise dead-reckoning accuracy. CONCLUSIONS We review the utility of VP-corrected dead-reckoning according to movement type and consider a range of ecological questions that would benefit from dead-reckoning, primarily concerning animal-barrier interactions and foraging strategies.
Collapse
Affiliation(s)
- Richard M. Gunner
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| | - Mark D. Holton
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| | - David M. Scantlebury
- School of Biological Sciences, Queen’s University Belfast, Belfast, 19 Chlorine Gardens, Belfast BT9 5DL, Northern Ireland, UK
| | - Phil Hopkins
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| | - Emily L. C. Shepard
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| | - Adam J. Fell
- Biological and Environmental Sciences, University of Stirling, Stirling FK9 4LA, Scotland, UK
| | - Baptiste Garde
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| | - Flavio Quintana
- Instituto de Biología de Organismos Marinos (IBIOMAR), CONICET. Boulevard Brown, 2915, U9120ACD Puerto Madryn, Chubut, Argentina
| | - Agustina Gómez-Laich
- Departamento de Ecología, Genética y Evolución & Instituto de Ecología, Genética Y Evolución de Buenos Aires (IEGEBA), CONICET, Pabellón II Ciudad Universitaria, C1428EGA Buenos Aires, Argentina
| | - Ken Yoda
- Graduate School of Environmental Studies, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan
| | - Takashi Yamamoto
- Organization for the Strategic Coordination of Research and Intellectual Properties, Meiji University, Nakano, Tokyo, Japan
| | - Holly English
- School of Biology and Environmental Science, University College Dublin, Belfield, Dublin, Ireland
| | - Sam Ferreira
- Savanna and Grassland Research Unit, Scientific Services Skukuza, South African National Parks, Kruger National Park, Skukuza 1350, South Africa
| | - Danny Govender
- Savanna and Grassland Research Unit, Scientific Services Skukuza, South African National Parks, Kruger National Park, Skukuza 1350, South Africa
| | - Pauli Viljoen
- Savanna and Grassland Research Unit, Scientific Services Skukuza, South African National Parks, Kruger National Park, Skukuza 1350, South Africa
| | - Angela Bruns
- Veterinary Wildlife Services, South African National Parks, 97 Memorial Road, Old Testing Grounds, Kimberley 8301, South Africa
| | - O. Louis van Schalkwyk
- Department of Agriculture, Government of South Africa, Land Reform and Rural Development, Pretoria 001, South Africa
- Department of Migration, Max Planck Institute of Animal Behavior, 78315 Radolfzell, Germany
- Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria, Onderstepoort 0110, South Africa
| | - Nik C. Cole
- Durrell Wildlife Conservation Trust, Les Augrès Manor, Channel Islands, Trinity JE3 5BP, Jersey, UK
- Mauritian Wildlife Foundation, Grannum Road, Indian Ocean, Vacoas, Mauritius
| | - Vikash Tatayah
- Mauritian Wildlife Foundation, Grannum Road, Indian Ocean, Vacoas, Mauritius
| | - Luca Börger
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
- Centre for Biomathematics, Swansea University, Swansea SA2 8PP, UK
| | - James Redcliffe
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| | - Stephen H. Bell
- School of Biological Sciences, Queen’s University Belfast, Belfast, 19 Chlorine Gardens, Belfast BT9 5DL, Northern Ireland, UK
| | - Nikki J. Marks
- School of Biological Sciences, Queen’s University Belfast, Belfast, 19 Chlorine Gardens, Belfast BT9 5DL, Northern Ireland, UK
| | - Nigel C. Bennett
- Mammal Research Institute. Department of Zoology and Entomology, University of Pretoria, Pretoria 002., South Africa
| | - Mariano H. Tonini
- Instituto Andino Patagónico de Tecnologías Biológicas y Geoambientales, Grupo GEA, IPATEC-UNCO-CONICET, San Carlos de Bariloche, Río Negro, Argentina
| | - Hannah J. Williams
- Department of Migration, Max Planck Institute of Animal Behavior, 78315 Radolfzell, Germany
| | - Carlos M. Duarte
- Red Sea Research Centre, King Abdullah University of Science and Technology, Thuwal 23955, Saudi Arabia
| | - Martin C. van Rooyen
- Mammal Research Institute. Department of Zoology and Entomology, University of Pretoria, Pretoria 002., South Africa
| | - Mads F. Bertelsen
- Center for Zoo and Wild Animal Health, Copenhagen Zoo, Roskildevej 38, DK-2000 Frederiksberg, Denmark
| | - Craig J. Tambling
- Department of Zoology and Entomology, University of Fort Hare, Alice Campus, Ring Road, Alice 5700, South Africa
| | - Rory P. Wilson
- Swansea Lab for Animal Movement, Department of Biosciences, Swansea University, Singleton Park, Swansea SA2 8PP, Wales, UK
| |
Collapse
|