Estimating grass nutrients and biomass as an indicator of rangeland (forage) quality and quantity using remote sensing in Savanna ecosystems

Abstract

Grass quality and quantity information plays a crucial role in understanding the distribution, densities and population dynamics of herbivores (i.e. livestock and wildlife). Leaf nitrogen (N) and biomass (g/ m2) are indicators of grass quality and grass quantity, respectively. The objective of the study is to estimate and map leaf N and biomass as an indicator of rangeland quality and quantity using vegetation indices derived from one RapidEye image taken at peak productivity. The study was undertaken in the north-eastern part of South Africa, in a transect extending from protected areas such as Kruger National Park and a privately owned game reserve to the communal areas of Bushbuckridge. Field work was undertaken to collect data on biomass and grass samples for retrieving leaf N, in April 2010, same time with image acquisition. RapidEye image was atmospherically corrected using atmospheric correction software for flat surfaces (ATCOR 2). Environmental or ancillary data sets were also collected from various sources as to develop an integrated modeling approach with the remotely-sensed indices. Commonly used vegetation index such as simple ratio was used exploiting a new red-edge band embedded in the RapidEye sensor. Leaf N regression models were developed using simple regression. Biomass (g/m2) prediction models were developed by applying bootstrapped stepwise regression using a combination of vegetation index and environmental or ancillary variables. Simple ratio (SR54) based on red-edge band produced higher grass N estimation accuracy. For the biomass estimation, vegetation indices produced poor results explaining less than 15% of variation. Biomass estimation was significantly improved to 27% of explained biomass variation by integrating vegetation index (SR54) and ancillary data. The latter approach is crucial because biomass is influenced by various environmental variables, which therefore play a crucial role in model development. The study demonstrated a potential of forage quantity and quality estimation using new high spatial remote sensing data with the red edge band. Integrating vegetation indices and ancillary data provides an opportunity to map grass biomass during peak productivity. Forage quality and quantity information is crucial for planning and management of grazing resources.