Who's afraid of the big bad glove? Testing for fear and its correlates in mink

Fear in farm animals is a welfare and economic concern. For Scandinavian mink, the stick test is common for assessing fearfulness: a spatula is inserted into the cage and minks' immediate responses are noted. However, on Ontario farms, fearfulness in the stick test was very rare and aggressive responses were prevalent, rendering this test poor for testing hypotheses related to fear and welfare. We therefore developed a modified version, the glove test, where the finger of a handling glove is inserted into the cage. This proved more sensitive than the stick test for detecting fearfulness in Ontario mink (20% vs. 2.5%, P < 0.0001); and successfully reduced aggressive responding (22% vs. 41%, P < 0.0001). When test-retest reliability was assessed conventionally, it was only moderate (e.g., 37.5% mink behaved the same in three tests). However, it is biologically realistic to expect habituation over repeated trials (e.g., reduced fearfulness), and treating such changes as acceptable, results were reliable for 76% of mink over three tests. Reliability could be further improved by taking location into account, since some mink were unresponsive only if in the nest box, half-asleep (Experiment 3: kappa of 0.38 if never in nest box vs. kappa of 0.002 if were). Glove tests had construct validity: mink immediately classified as fearful spent more time exhibiting other fear-related behaviours (ambivalence: mean 30 s vs. 4 s, P = 0.009), while mink immediately classified as curious then spent more time investigating the glove (mean 103 s vs. 57 s, P < 0.0001). Glove tests also revealed expected temperament differences between Black and Pastel colour-types, with Pastels being less fearful (P = 0.001). Finally, we tested whether fearfulness in the glove test is associated with decreased reproductive success. Pastel mink that were fearful during the presumed gestation period were less likely to reproduce (P = 0.006). Like stick tests, glove tests are thus practical, valid and reliable for assessing fearfulness in farmed mink, but they are better for detecting fearfulness in low fear populations. (C) 2011 Elsevier B.V. All rights reserved.