It is essential to have an accurate picture of the spatial distribution of equines to be able to monitor equine health events effectively. In France, this information is only available for certain categories of live equines kept in professional structures and for dead equines removed by renderers. This limits the surveillance, prevention and control methods able to be used to prevent the spread of equine diseases. Our study aimed to provide a realistic estimate of the spatial distribution of the French equine population at the detailed scale of the French commune (France’s smallest administrative unit). For this purpose, we adapted the Bayesian method used by Lo Lacono et al., based on the distance between the owner’s location and the location of his/her equines, and on the percentage of urban coverage. To assess whether the location of dead equines could be representative of the location of live equines, the distribution of distances between equines and owners was calculated from a sample of live equines on the one hand, and a sample of dead equines on the other, both accurately located. We also tested two different assignment methods for locating equines: Method 1 assigned to each owner a single holding commune, while Method 2 allowed more variability in holding communes for owners associated with multiple equines. A marked difference was observed between Methods 1 and 2 regardless of the sample used, with only 2.4% and 4.3% respectively of the communes having the same number of equines. Conversely, little difference was observed in the results whether the live or dead equine sample was used, with approximately 45% of the communes having the same number of equines. Regarding differences in spatial distribution, Method 1 based on the live equine sample estimated higher local densities of equines without considering urban areas. In contrast, Method 2 provided more dispersed maps, with low densities in the densest urban areas. In conclusion, dead equines appeared to be representative of live equines and some of our estimates are consistent with the information collected by the French horse and riding institute (IFCE). These results now have to be compared with field data to test their relevance.
Read full abstract