Measurement of spatial distribution of cattle dung under high and low stocking densities using remote sensing : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Agricultural Science at Massey University, Manawatū, New Zealand

Abstract

Regenerative pasture management is increasingly being practiced in New Zealand and encompassing a range of principles which generally focus on using diverse plant species to maximise photosynthesis and strategically use grazing livestock, with an overall aim of improving soil health. An example of regenerative pasture management is undertaken on Mangarara Farm, a non-irrigated sheep and beef farm located in Elsthorpe in central Hawkes Bay. Mangarara farm focuses on having higher pre and post-grazing pasture biomass than more conventional practices and grazes these pastures at higher stocking density. In theory, this grazing practice leads to more pasture trampling into the ground and more even dung distribution than conventional grazing practices. A new measurement method was developed in this study to test whether there is a change in the spatial distribution of cattle dung under regenerative management compared to conventional management. This method used a drone fitted with a red, green and blue (RGB) camera to identify and spatially map dung patches following grazing in a defined area (cells) on Mangarara farm. The research trial compared conventional and regenerative management using low and high stocking density. The control grazing (conventional) had a low stocking density of 6 Angus heifers moved every four days. In comparison, regenerative grazing had a high stocking density of 57 Angus heifers moved multiple times daily. The novel drone method was validated against a systematic measurement approach to assess the accuracy of the drone in detecting dung patches, compared to the systematic manual marking of dung patches using a survey-grade Trimble GPS to manually mark every dung patch within the cell. The results showed that the drone detected for all cells a mean of 57% of the dung patches within the cells. Data analysis revealed that multiple key factors affected the drone accuracy, including trees, pasture height and the amount of bare soil and it is recommended that lower pasture height and less bare soil present will minimise variation in future measurements. The same drone method was then used to compare the spatial distribution of dung under regenerative and conventional management. The results showed a significant difference between the median number of dung patches/ha for regenerative and conventional management. The analysis showed that the dung was not randomly distributed throughout the cell and that the regenerative management had slightly less clustering than the conventional management, indicating that the dung was more evenly distributed through the cell under regenerative management. The results from this study have shown that a drone fitted with an RGB camera successfully detected the spatial distribution of dung. However, some key limitations were identified, including wet soil conditions, bare soil and pasture height, which made it difficult to identify the dung due to a lack of colour contrast with the pasture and/or muddy soils. Dung with a higher liquid content was also difficult to delineate as one or several dung patches. Despite these limitations, this novel drone fitted with an RGB camera method offers a cheaper alternative to the traditional labour-intensive method of measuring dung distribution via the grid method and provides scope to measure dung distribution under a range of topographies such as hill country. This new method provides an opportunity for more research on the distribution of dung under different grazing management conditions and offers the potential to improve our understanding of soil nutrient distribution and nutrient loss risk.