Analysis of horse movements from non-commercial horse properties in New Zealand

Abstract

AIMS: To investigate property-level factors associated with the movement of horses from non-commercial horse properties, including the size and location of the property, number and reason for keeping horses.

METHODS: Using a cross-sectional survey 2,912 questionnaires were posted to randomly selected non-commercial horse properties listed in a rural property database. The survey collected information about the number of horses, and reasons for keeping horses on the property, and any movement of horses in the previous 12 months. Three property-level outcomes were investigated; the movement status of the property, the frequency of movement events, and the median distance travelled from a property. Associations were examined using logistic regression and Kruskal-Wallis analysis of variance.

RESULTS: In total 62.0% (488/791) of respondents reported at least one movement event in the year prior to the survey, for a total of 22,050 movement events. The number of movement events from a property varied significantly by the number of horses on the property (p<0.02), while the median distance travelled per property varied significantly by both region (p<0.03) and property size (p<0.01). Region, property size, the number of horses kept, and keeping horses for competition, recreation, racing or as pets were all significantly associated with movement status in the multivariable analyses (p<0.001).

CONCLUSION AND CLINICAL RELEVANCE: This study showed that there are characteristics of non-commercial horse properties that influence movement behaviour. During an exotic disease outbreak the ability to identify properties with these characteristics for targeted control will enhance the effectiveness of control measures.

Keywords:

• Equine
• epidemiology
• logistic regression
• movement
• infectious disease control
• equine influenza

Acknowledgements

This work was funded by the Ministry of Agriculture and Forestry (MAF) New Zealand. AsureQuality provided the AgriBase TM dataset, with particular thanks to Robert Sanson. The authors wish to thank the participants, without whom this research would not have been possible. Appreciation is extended to Sarah Moore and Juan Sanchez for their help with address checking and Simon Verschaf felt for his IT support (Massey University).