Investigating attributes for selecting nurse sows in swine herds of Minnesota, USA, using Best-Worst Scaling analysis

Nurse sow selection is aimed at ensuring ideal sows are chosen to manage surplus piglets in hyper prolific herds. This study investigated attributes farm managers preferred when choosing nurse sows. We evaluated seven attributes: Body Condition Score (BCS), Sows' Current Litter Size (SCLS), Sows' Current Litter Health (SCLH), Sow Behaviour (SB), Sow Teat Number (STN), Lactation Stage (LS), and Parity (P). Choice decisions of 51 managers were stratified into five subgroups: gender, age, pig-farming experience, education level, and herd size managed. We used Best Worst Scaling analysis to evaluate relative importance of attributes, while Conditional Logit Model assessed their utility importance, setting P as the baseline. Results showed that SCLH was the most important while P the least important. All six attributes were significantly preferred over the baseline. Female managers showed significant preferences for all attributes than male managers. Younger managers (18-40 years old) placed significant preference on sow physical qualities compared to older managers (> 40 years old). Managers with ≤10 years' experience demonstrated wider significant preference profiles than those of >10 years. Managers with college degree preferred half of the attributes while those with professional degrees (BSc/MSc) gave significant relevance to almost all attributes. Managers managing herds of up to 5000 sows viewed all attributes significant while those in charge of extra-large herds (>5000 sows) gave no preference to SB and LS. This research provides new empirical evidence and nuanced analysis of nurse sow selection, placing SCLH as a priority proxy for selecting a nurse sow.