Determining yield of forage crops using the Canopeo mobile phone app

Abstract

One of the biggest challenges in pastoral systems is accurately determining forage availability. Destructive sampling is the most accurate method to use but it is very time-consuming, so non-destructive methods are usually preferred. This study compared data on the percentage of green canopy cover (%GCC) obtained using a recently developed image-analysis mobile phone app (Canopeo®) with crop biomass, light interception and NDVI measurements. Data were obtained for a range of crop species from two dryland experiments of lucerne (E1) and winter forage crops (E2). Linear and quadratic regression models were constructed to evaluate the data obtained using Canopeo® to estimate biomass accumulation, light interception and NDVI. For lucerne, positive linear relationships were found between %GCC and biomass (0.77 and 0.79 for Spring + Summer and Autumn + Winter, respectively; P<0.05), and between %GCC and light interception (R2=0.83; P<0.05). For winter forage crops, there was a linear relationship between biomass and %GCC (R2=0.81; P<0.05), and a linear relationship between %GCC and light interception (R2=83; P<0.05). NDVI data obtained using the GreenSeeker® crop sensing system was linearly related to %GCC data from Canopeo® in both experiments (R2=0.84 in E1 and R2=0.93 in E2; P<0.05). The Canopeo® app proved to be a fast and reliable method to estimate biomass accumulation and light interception of forage crops in this study and has potential wider applications. However, it does not distinguish between crops and weeds.