Piloting a criterion-based and in-cage behavioural monitoring approach in assessing the sociality of shelter cats towards humans.

The suitability of rehoming sheltered cats hinges on their sociality towards humans. However, the socialisation history of these cats is often unknown. While a structured and validated assessment is available, it requires considerable time. Here we pilot a criterion-based assessment as a quicker alternative while exploring the basis for an automated assessment via monitoring in-cage behaviours, in the absence of an observer. Supposing our criterion works in evaluating the sociality of sheltered cats, the behavioural progression exhibited during the assessment will differ between cats that reached the criterion and those that did not. Similarly, should in-cage behaviour be able to differentiate between more and less socialised cats, we would expect in-cage behavioural transitions to differ between cats that reached the criterion and individuals that did not. In our investigations, we presented two modified assays for five days to 16 newly sheltered cats, noting if individuals reach the criterion while recording their behaviours. To avoid the carryover effect, in-cage behaviours were recorded before each assay. Next, we identified the significant behavioural transitions using sequential analysis for responses elicited during the assays and when cats were in their respective cages. Subsequently, we used PERMANOVA to compare the behavioural transitions between cats that reached the criterion and those that did not. Corresponding to the comparison between more and less socialised cats, cats that reached the criterion displayed more behaviours and these were distinct from individuals that failed (P value = 0.019). Likewise, we found in-cage behaviours to be distinct between cats that reached the criterion and those that failed (P value = 0.001). Despite the need for further validation and consideration of the likely effect of shelter stress, our findings seemingly point to the criterion-based assessment as a quicker alternative. Furthermore, our results propound the possibility of automating sociality assessment through the monitoring of in-cage behaviours using artificial intelligence. • Rehoming cats that are socialised towards humans is crucial for their welfare. • The current evaluation of shelter cats' sociality towards humans is time-consuming. • We proposed a faster criterion-based alternative while exploring a passive approach. • Cats that reached the criterion behaved congruously to human-socialised individuals. • In-cage behaviours differ between cats that reached the criteria and those not. [ABSTRACT FROM AUTHOR]