Towards a taxonomy of stereotypic behaviours in the American mink (Neovison vison), a model Carnivore: Homogeneous or heterogeneous?

Stereotypic behaviours (SBs) are sometimes assumed homogeneous, despite their diverse morphologies, as if sharing a common aetiology. However, if different SB forms are instead heterogeneous, they may have different causes, potentially impacting how best to prevent or manage them. We therefore conducted two exploratory studies with mink (Neovison vison) to investigate whether 'loco' SBs (locomotor forms like pacing, and 'stationary' forms like head-bobbing) are homogeneous or heterogeneous, and to corroborate previous findings that 'scrabbling' (stereotypic scratching at cage walls) is distinct from loco SBs. We assessed correlations between their time-budgets (Studies 1 and 2), followed young mink as they aged (Study 2), and used chi-square goodness-of-fit tests to explore whether some SB forms co-occur within bouts more or less frequently than expected by chance (Study 2). Co-occurrence in bouts and positive correlations between different SB forms would indicate homogeneity and perhaps shared causal factors. Conversely, SB forms appearing on their own in bouts and/or showing absent or negative correlations with other forms would indicate heterogeneity. Study 1 demonstrated negative correlations between scrabbling and most loco SB forms (p < 0.05), but positive correlations between most loco forms (p < 0.05). Study 2 improved upon Study 1's methods with a new sample of mink, a refined ethogram, and the collection of longitudinal data and data on multiform sequences during SB bouts. Scrabbling, again, negatively correlated with several loco SB forms (p < 0.05). It was also likely to occur in bouts without any other SB form. Moreover, within the loco group, the two subtypes of route-tracing ('nestbox pacing' and pacing) positively correlated (p < 0.05), and 'nestbox pacing' often occurred in bouts with pacing. Stationary SBs, however, appeared heterogeneous: head-twirling and head-bobbing tended to positively correlate with each other (p < 0.10), but not other forms, and to appear in individual bouts; while whole-body bobbing and head-weaving positively correlated with route-tracing (p < 0.10), and they also co-occurred. Overall, results confirmed that scrabbling is distinct from other SBs. Loco SBs, however, appeared either homogeneous (Study 1) or heterogeneous (Study 2), depending on the ethogram. We conclude that a SB ethogram precisely based on morphology (e.g., whole-body versus head-based SBs) is most likely to reveal underlying heterogeneities; that future researchers should validate our new proposed mink SB subcategories, 'whole-body' and 'head-based'; and that our methods could usefully be applied to other species to categorise their SBs.