A new reversal learning test for horses

Mira Joanna Hämäläinen, Iina Liisa Brotherus, Tuire Kaimio, Laura Hänninen, Heli Maria Suomala, Anna-Mari Olbricht, Anna Mykkänen

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Abstract

Learning difficulties may be linked to sleep disorders in horses, but so far, no practical test suitable for stable conditions has been available. We tested a novel reversal learning test (RLT) we have previously developed, on 16 riding school horses. Before the RLT, horses were conditioned to touch an object presented individually from both sides and at different distances for median (CL95%) duration of 16.0 (16.8–19.6) min. During RLT, horses were required to reach a rate of 7 touches per minute, after which the target side was reversed until the test was over. The end criteria were that the treats were finished, the horse stopped working or 30 min had elapsed. Two days after RLT, we recorded the duration of REM-like-sleep (REMLS, resting neck relaxed) the horses performed in their stalls from 16 to 08. We scored it further as lasting either for less or at least 20 min (short or long REMLS). We tested with a log-rank test the possible differences in learning curves between short or long REMLS. Overall, 15/16 horses completed the test, with a median (CL95%) number of 6.0 (6.1–7.4) successful turns in 14.0 (12.4–14.6) min. However, the median number of successful turns was at 5 (3.9–6.1) for short REMSL and at 6 (4.3–7.7) for long REMSL (χ2= 6.4, p = 0.01). We concluded, that RLT can be performed under stable conditions. More studies are needed, but differences in test success may be explained by differences in sleep.
Original languageEnglish
Title of host publicationProceedings of the 5th congress of ECAWBM
Publication date2023
Publication statusPublished - 2023
MoE publication typeA4 Article in conference proceedings

Cite this