Effects of Nocturnal Grazing Behavior and Stock Density on Cattle Dung in Sloped Rotational Grazing

Abstract

In this study, we investigated the effects of nocturnal variation in grazing behavior and stock density on the distribution of cattle dung in a sloped rotational grazing system in Japan. The specific objectives were to determine the influence of watering sites on dung distribution at night and the relationship between stock density and dung distribution. The study was conducted on a 2.56-ha steeply sloped pasture that was divided into four sub-paddocks for rotational grazing. Drone imagery and the YOLO deep learning algorithm were used to automate the mapping of more than 400 cattle dung pats per grazing session, with a mean average precision of 0.75. The YOLO algorithm demonstrated high accuracy in identifying dung that was up to 4 days old. Our analyses revealed that the distance to a watering site influenced fecal distribution during the day, but had a weaker and insignificant effect at night. With low-density grazing, dung pats showed a clumped distribution across the entire pasture, whereas at higher stock density the distribution approached randomness. The formation of four dung hotspots was observed in the pasture, although these spots were not consistently used throughout subsequent grazing periods. The cattle tended to migrate from more elevated foraging areas to more level grazing areas for rest, resulting in a loose connection between foraging and resting sites, and this behavior may have a bearing on the lack of successive utilization of these fecal hotspots. Subsequent research must examine whether these patterns persist across seasonal variations and different grazing lands. Despite this need for validation, our findings can provide insights for guiding sustainable nutrient redistribution and grassland management on sloped terrain through the management of stock density and grazing schedules.