FREE SHIPPING over $40
OVER 10000 FIVE-STAR REVIEWS
5% OFF ALL SUBSCRIPTIONS
100% HAPPINESS GUARANTEE
Analyze Diet
0
Environmental monitoring and assessment2025; 197(8); 922; doi: 10.1007/s10661-025-14363-1

Artificial intelligence tools to assess different levels of activity performed by semi-wild horses in grassland ecosystems.

Abstract: In order to understand the role of horses in ecosystems and to effectively use their grazing in the protection of grasslands, it is important to assess where they primarily stay, followed by whether these habitats are used for grazing or resting. The main goal of the study was the model development based on artificial intelligence tools which allow to distinguish the basic levels of activity performed by horses using data from an accelerometer mounted in a collar worn by animals. The model calibration was based on direct observations of five randomly selected Polish primitive horse mares. In order to create a model that allows for classification into three groups of behaviours: grazing, resting, and moving, an approach based on machine learning, one of the basic technologies of artificial intelligence, was used. The carried out analyses allowed for the determination of the most important features, among the fourteen determined from raw X, Y, and Z axis acceleration values across 5-s measurements. The recommended method for the classification of behaviours of primitive Konik horses based on the selection of variables observed from the accelerometer is the CART method, whereas the most accurate tool for its construction is learning neural networks. Our research indicates the usefulness of the accelerometer and proposed artificial intelligence methods in distinguishing the main activities performed by horses.
© 2025. The Author(s).
Get Full Text
Publication Date: 2025-07-16 PubMed ID: 40665113PubMed Central: PMC12263774DOI: 10.1007/s10661-025-14363-1Google Scholar: Lookup
The Equine Research Bank provides access to a large database of publicly available scientific literature. Inclusion in the Research Bank does not imply endorsement of study methods or findings by Mad Barn.
LinkedInCopy
  • Journal Article

Summary

This research summary has been generated with artificial intelligence and may contain errors and omissions. Refer to the original study to confirm details provided. Submit correction.

This research developed an artificial intelligence (AI) model to assess the levels of activity performed by semi-wild horses, enabling a better understanding of their role and behavior in grassland ecosystems, which facilitates more effective grazing management strategies.

Research Objective and Methodology

  • The main objective of the study was to develop an AI-based model that accurately categorizes the horses’ behaviors into primary activities: grazing, resting, and moving. To achieve this, the researchers collected data using accelerometers attached to the collars worn by the semi-wild Polish primitive horse mares.
  • The model development process involved calibrating the accelerometer data against direct observations of the horses’ activities. This approach allowed for machine learning techniques to accurately classify activities based on the accelerometer’s raw X, Y, and Z axis acceleration values from 5-second measurements.

Key Findings and Recommendations

  • The analysis identified the most crucial features out of fourteen derived from raw acceleration values that helped distinguishing the horses’ main activities. These features were used to teach the AI model to classify the horses’ behaviors accurately.
  • The study recommends using the Classification And Regression Tree (CART) method for classifying the behaviors of primitive Konik horses based on accelerometer data. Notably, learning neural networks was found to be the most effective tool for building the model.
  • The research underscores the potential of accelerometers and AI methods in identifying the main activities of semi-wild horses, suggesting their valuable application in monitoring and managing these animals in grassland ecosystems.

Implications of the Research

  • The findings of this study could significantly improve the understanding of semi-wild horses’ behavior in their natural habitats. The application of AI tools could make it easier for ecologists and other stakeholders to monitor and manage these horses, thus promoting the preservation of grassland ecosystems.
  • The developed AI model also demonstrates the potential for similar applications in other wildlife monitoring efforts, thereby expanding the utility of these technologies in conservation and ecosystem management.

Cite This Article

APA
Chodkiewicz A, Prończuk M, Studnicki M. (2025). Artificial intelligence tools to assess different levels of activity performed by semi-wild horses in grassland ecosystems. Environ Monit Assess, 197(8), 922. https://doi.org/10.1007/s10661-025-14363-1

Publication

Environmental monitoring and assessment
ISSN: 1573-2959
NlmUniqueID: 8508350
Country: Netherlands
Language: English
Volume: 197
Issue: 8
Pages: 922
PII: 922

Researcher Affiliations

Chodkiewicz, Anna
  • Department of Agronomy, Institute of Agriculture, Warsaw University of Life Sciences, 159 Nowoursynowska Str., 02-776, Warsaw, Poland. anna_chodkiewicz@sggw.edu.pl.
Prończuk, Martyna
  • , Warsaw, Poland.
Studnicki, Marcin
  • Department of Biometry, Institute of Agriculture, Warsaw University of Life Sciences, 159 Nowoursynowska Str., 02-776, Warsaw, Poland.

MeSH Terms

  • Animals
  • Horses / physiology
  • Artificial Intelligence
  • Grassland
  • Environmental Monitoring / methods
  • Behavior, Animal
  • Ecosystem

Conflict of Interest Statement

Declarations. Ethics approval: All authors have read, understood, and have complied as applicable with the statement on “Ethical responsibilities of Authors” as found in the Instructions for Authors. A statement on the welfare of animals if the research involved animals: No approval of research ethics committees was required to accomplish the goals of this study because experimental work did not involve capturing and chemical immobilisation of the horses. Competing interests: The authors declare no competing interests. Clinical trial number: Not applicable.

References

This article includes 40 references
  1. Altenburg TM, Wang X, van Ekris E, Andersen LB, Møller NC, Wedderkopp N, Chinapaw MJM. The consequences of using different epoch lengths on the classification of accelerometer based sedentary behaviour and physical activity. (7), Article Article e0254721.
    pmc: PMC8282067pubmed: 34265011doi: 10.1371/journal.pone.0254721google scholar: lookup
  2. Alvarenga FAP, Borges I, Palkovič L, Rodina J, Oddy VH, Dobos RC. Using a three-axis accelerometer to identify and classify sheep behaviour at pasture. , 91–99.
    doi: 10.1016/j.applanim.2016.05.026google scholar: lookup
  3. Auer U, Kelemen Z, Engl V, Jenner F. Activity time budgets – A potential tool to monitor equine welfare?. (3), Article 850.
    pmc: PMC8002676pubmed: 33802908doi: 10.3390/ani11030850google scholar: lookup
  4. Barwick J, Lamb DW, Dobos R, Welch M, Trotter M. Categorising sheep activity using a tri-axial accelerometer. , 289–297.
  5. Barwick J, Lamb DW, Dobos R, Welch M, Schneider D, Trotter M. Identifying sheep activity from tri-axial acceleration signals using a moving window classification model. (4), Article 646.
    doi: 10.3390/rs12040646google scholar: lookup
  6. Berger A, Scheibe K-M, Eichhorn K, Scheibe A, Streich J. Diurnal and ultradian rhythms of behaviour in a mare group of Przewalski horse (), measured through one year under semi-reserve conditions. , 1–17.
    doi: 10.1016/s0168-1591(99)00026-xgoogle scholar: lookup
  7. Błażek, E., Stopa-Boryczka, M., Boryczka, J., Wawer, J., Dobrowolska, M., Grabowska, K., Grzęda, M., Osowiec, M., & Skrzypczuk, J. (2013). Wydział Geografii i Studiów Regionalnych Uniwersytetu Warszawskiego.
  8. Campbell HA, Gao L, Bidder OR, Hunter J, Franklin CE. Creating a behavioural classification module for acceleration data: Using a captive surrogate for difficult to observe species. (24), 4501–4506.
    pubmed: 24031056
  9. Chodkiewicz A. Advantages and disadvantages of Polish primitive horse grazing on valuable nature areas – A review. Article e00879.
    doi: 10.1016/j.gecco.2019.e00879google scholar: lookup
  10. Collins GH, Petersen SL, Carr CA, Pielstick L. Testing VHF/GPS collar design and safety in the study of free-roaming horses. (9), Article e103189.
    pmc: PMC4157739pubmed: 25198704doi: 10.1371/journal.pone.0103189google scholar: lookup
  11. Cosyns E, Degezelle T, Demeulenaere E, Hoffman M. Feeding ecology of Konik horses and donkeys in Belgian coastal dunes and its implication for nature management. (Suppl. 2), 111–118.
  12. Duncan P. Time-budget of Carmague horses. , 188–208.
  13. Dunford CE, Marks NJ, Wilson RP, Scantlebury DM. Identifying animal behaviours from accelerometers: Improving predictive accuracy of machine learning by refining the variables selected, data frequency, and sample duration. Article e11380.
    pmc: PMC11097004pubmed: 38756684doi: 10.1002/ece3.11380google scholar: lookup
  14. Hampson BA, Morton JM, Mills PC, Trotter MG, Lamb DW, Pollit CC. Monitoring distances travelled by horses using GPS tracking collars. (5), 176–181.
    pubmed: 20529024doi: 10.1111/j.1751-0813.2010.00564.xgoogle scholar: lookup
  15. Hansen BD, Lascelles DX, Keene BW, Adams AK, Thomson AE. Evaluation of an accelerometer for at-home monitoring of spontaneous activity in dogs. (5), 468–475.
    pubmed: 17472445doi: 10.2460/ajvr.68.5.468google scholar: lookup
  16. Hennig JD, Scasta JD, Beck JL, Schoenecker KA, King SRB. Systematic review of equids and telemetry collars: Implications for deployment and reporting. , 361–371.
    doi: 10.1071/wr19229google scholar: lookup
  17. Howley T, Madden MG, O’Connell ML, Ryder AG. The effect of principal component analysis on machine learning accuracy with high dimensional spectral data. (1st ed., pp. 209–222).
    doi: 10.1007/1-84628-224-1_16google scholar: lookup
  18. Kaczensky P, Ganbaatar O, von Wehrden H, Walzer C. Resource selection by sympatric wild equids in the Mongolian Gobi. (6), 1762–1769.
    doi: 10.1111/j.1365-2664.2008.01565.xgoogle scholar: lookup
  19. King SRB. Home range and habitat use of free-ranging Przewalski horses at Hustai National Park, Mongolia. , 101–113.
    doi: 10.1016/s0168-1591(02)00087-4google scholar: lookup
  20. Kleanthous N. Machine learning techniques for classification of livestock behavior. (1st ed., pp. 304–315).
    doi: 10.1007/978-3-030-04212-7_26google scholar: lookup
  21. Köhler M, Hiller G, Tischew S. Year-round horse grazing supports typical vascular plant species, orchids and rare bird communities in a dry calcareous grassland. , 48–57.
    doi: 10.1016/j.agee.2016.03.020google scholar: lookup
  22. Krop-Benesch A, Berger A, Hofer H, Heurich M. Long-term measurement of roe deer () (Mammalia: ) activity using two-axis accelerometers in GPS-collars. (1), 69–81.
    doi: 10.1080/11250003.2012.725777google scholar: lookup
  23. Krużel J, Ziernicka-Wojtaszek A, Borek Ł, Ostrowski K. Zmiany czasu trwania meteorologicznego okresu wegetacyjnego w Polsce w latach 1971–2000 oraz 1981–2010. , 47–52.
  24. Menard C, Duncan P, Fleurance G, Georges J, Lila M. Comparative foraging and nutrition of horses and cattle in European wetlands. , 120–133.
    doi: 10.1046/j.1365-2664.2002.00693.xgoogle scholar: lookup
  25. Metera E, Sakowski T, Słoniewski K, Romanowicz B. Grazing as a tool to maintain biodiversity. (4), 315–334.
  26. Moindardeau C, Mesléard F, Ramone H, Dutoit T. Grazing in temporary paddocks with hardy breed horses (konik polski) improved species-rich grasslands restoration in artificial embankments of the Rhône river (Southern France). Article e01874.
    doi: 10.1016/j.gecco.2021.e01874google scholar: lookup
  27. Morrison R, Sutton DGM, Ramsoy C, Hunter-Blair N, Carnwath J, Horsfield E, Yam PS. Validity and practical utility of accelerometry for the measurement of in-hand physical activity in horses. , 233.
    pmc: PMC4566433pubmed: 26362544doi: 10.1186/s12917-015-0550-2google scholar: lookup
  28. Pasicka E. Polish Konik horse – Characteristics and historical background of native descendants of tarpan. (2–4), 25–38.
  29. Polish Horse Breeders Association. (2020). Program ochrony rasy konik polski. Retrieved February 7, 2025, from https://www.pzhk.pl/wp-content/uploads/pr-hodow-kn-2020-04-01.pdf
  30. Prache S, Gordon IJ, Rook AJ. Foraging behaviour and diet selection in domestic herbivores. , 335–345.
  31. R Core Team. R: a language and environment for statistical computing. .
  32. Reķe A, Zariņa A, Vinogradovs I. Management of semi-wild large herbivores’ grazing sites in Latvia. , 241–244.
  33. Riaboff L, Shalloo L, Smeaton AF, Couvreur S, Madouasse A, Keane MT. Predicting livestock behaviour using accelerometers: A systematic review of processing techniques for ruminant behaviour prediction from raw accelerometer data. Article 106610.
    doi: 10.1016/j.compag.2021.106610google scholar: lookup
  34. Sakai K, Oishi K, Miwa M, Kumagai H, Hirooka H. Behavior classification of goats using 9-axis multi sensors: The effect of imbalanced datasets on classification performance. Article 105027.
    doi: 10.1016/j.compag.2019.105027google scholar: lookup
  35. Schaefer JA, Morellet N, Pépin D, Verheyden H. The spatial scale of habitat selection by red deer. , 1337–1345.
    doi: 10.1139/z08-122google scholar: lookup
  36. Simanungkalit G, Barwick J, Cowley F, Dobos R, Hegarty R. A pilot study using accelerometers to characterise the licking behaviour of penned cattle at a mineral block supplement. (4), Article 1153.
    pmc: PMC8073741pubmed: 33920600doi: 10.3390/ani11041153google scholar: lookup
  37. Stroh PA, Mountford JO, Hughes FMR. The potential for endozoochorous dispersal of temperate fen plant species by free-roaming horses. , 359–368.
    doi: 10.1111/j.1654-109x.2011.01172.xgoogle scholar: lookup
  38. Vázquez Diosdado JA, Barker ZE, Hodges HR, Amory JR, Croft DP, Bell NJ, Codling EA. Classification of behaviour in housed dairy cows using an accelerometer-based activity monitoring system. , 15.
    doi: 10.1186/s40317-015-0045-8google scholar: lookup
  39. Vulink, J.T. (2001). . Rijkswaterstaat.
  40. Zhang M, Sawchuk A. A feature selection-based framework for human activity recognition using wearable multimodal sensors. (pp. 92–98).

Citations

This article has been cited 0 times.
Subscribe to our newsletter
Join 99,850 horse owners like you who receive our email updates on horse health and nutrition.