Towards a more practical attention bias test to assess affective state in sheep

Abstract

Tests for attention bias potentially offer more rapid assessment of affective state in animals than existing cognitive methods. An attention bias test has previously been developed for sheep and validated as a measure of anxious states. The 3 minute test assessed behavioural responses of sheep in an enclosed arena after brief exposure to the threat of a dog. Experiment 1 of the current study aimed to refine the previously developed method, removing the need for a habituation period and shortening the test duration. Sheep were given either an anxiolytic drug, an anxiogenic drug or a control treatment prior to testing to induce contrasting affective states. Differences in behaviour were found between the treatment groups within the first 45s of the test, indicating the original test duration could be shortened from 180 s. During testing, 36 of 40 animals in the control and anxiolytic groups ate the novel feed offered in the test, indicating it is not necessary to habituate animals to a feed container. Experiment 2 aimed to confirm the responses measured in the test were primarily towards the dog rather than other aspects of the test environment. Sheep exposed to an empty window at the beginning of the test behaved differently to those which were exposed to a dog, indicating sheep behaviour in the test is at least partially a response to the dog. A third group of sheep were also tested with the dog immediately after having small data loggers attached to their necks. Behaviour of these sheep did not differ from the sheep tested without loggers, indicating data logger attachment did not impact their behaviour in the test. In both experiments, treatments did not appear to modify activity (zones crossed), which we propose indicates the test was primarily detecting valence of the affective state rather than arousal.

Peer Review #2 of "The influence of pharmacologically-induced affective states on attention bias in sheep (v0.1)"

“…All latency data were analyzed with Cox's proportional hazards model using survival analysis (Therneau & Grambsch, 2000;Therneau, 2015), as described by Monk et al (2018b). These data included latencies to sniff the photo, sniff the environment and become non-vigilant.…”

“…Tamir & Robinson, 2007;Grafton, Ang & MacLeod, 2012;Sanchez & Vazquez, 2014;Caudek, Ceccarini & Sica, 2017), although these results have not been consistent in all tested populations (Isaacowitz et al, 2008). In sheep, Lee et al (2016) and Monk et al (2018b) demonstrated that the anxiolytic drug diazepam induced an attention bias away from a threatening stimulus and towards a food reward compared to saline treated control animals. This suggests that the attention bias test may be used to assess positive, or at least less negative, affective states in sheep.…”

“…A calm-like state was induced in the Calm group using the anxiolytic drug diazepam (Troy Laboratories, Sydney, Australia). The diazepam was administered at a dose rate of 0.1mg/kg as per previous studies (Lee et al, 2016;Monk et al, 2018b). The drug was administered i.m.…”

“…For example, humans in anxious states pay more attention towards threatening information than non-anxious individuals (Bradley et al, 1995;Bradley, Mogg & Lee, 1997;Bar-Haim et al, 2007). Based on this principle, studies were conducted to determine whether an attention bias test developed for sheep was sensitive to changes in anxiety-like states (Lee et al, 2016;Monk et al, 2018b). The test measured time allocation of attention between a dog (threat) and a food reward (positive).…”

“…Few studies have used pharmacological treatments in sheep to induce and assess positive affective states. The anxiolytic drug diazepam was used to reduce anxiety-like behaviors during development of the original attention bias test method (Lee et al, 2016;Monk et al, 2018b). Further, it has been used to reduce anxiety-like behaviors in a range of other contexts, such as during isolation and suddenness tests (e.g.…”