Identifying drivers of fox and cat faecal deposits in kitchen gardens in order to evaluate measures for reducing contamination of fresh fruit and vegetables

Abstract

Preventing foodborne pathogen contamination of raw fruit and vegetables in the field is critically important for public health. Specifically, it involves preventing faecal deposit by wildlife or domestic animals in fields of crops and kitchen gardens. The present study aims to identify the drivers of fox, dog and cat faecal deposits in kitchen gardens in order to mitigate the risk of contamination of raw produce with parasites shed in carnivore faeces. The focus was on Echinococcus multilocularis, ranked highest in the importance of foodborne parasites in Europe, but attention was also paid to other parasites of major concern - Toxoplasma gondii and Toxocara spp. During the winters of 2014 to 2016, faecal samples were collected from 192 kitchen gardens located in north-eastern France. From these samples, 77% contained scat of carnivores. Molecular analyses revealed that 59% of the 1016 faeces collected were from cats, 31% from foxes, and 10% from dogs. The ease of accessibility to kitchen gardens, the presence of food in the vicinity, and the composition of the surrounding vegetation were used to explain the distribution of fox and cat faeces. Generalized Linear Mixed Effects modelling showed that: i) fencing was not efficient in reducing cat faecal deposits, but drastically decreases those of foxes; ii) the abundance of Microtus sp. indicates a reason for the presence of both fox and cat faecal deposits, iii) the abundance of Arvicola terrestris, the proximity of fruit trees or farms and the predominance of forest and grassland around the village are all drivers of fox faecal deposits. These results point to the importance of fencing around kitchen gardens located in E. multilocularis endemic areas, particularly those surrounded by forest and grassland or close to fruit trees or farms.

Keywords: Echinococcus multilocularis; Environmental contamination; Foodborne parasites; Toxocara sp.; Toxoplasma gondii.

Fig. 1

Localization of the two study areas (Ardennes and Moselle) in France (a), the prospected villages in Ardennes and Moselle regions (b), and the kitchen gardens sampled in one village (c).

Fig. 2

Outputs from the fox faeces deposit model. (a) Variable effects (odds ratio and their 97.5% confidence intervals) for the full model. The more fare away a variable is from the intercept, the highest is its effect on faeces deposit. (b). Expected number of fox faeces per kitchen garden predicted by the parsimonious model according to the different levels of each influential variables.

Fig. 3

Outputs from the cat faeces deposit model. (a) Variable effects (odds ratio and their 97.5% confidence intervals) for the full model. The more fare away a variable is from the intercept, the highest is its effect on faeces deposit. (b). Expected number of cat faeces per kitchen garden predicted by the parsimonious model according to the different levels of each influential variables.