Discriminating tropical grass ( Cenchrus ciliaris) canopies grown under different nitrogen treatments using spectroradiometry

Abstract

Techniques for estimating and mapping pasture quality are critical for a better understanding of wildlife and livestock grazing patterns. Nitrogen is one of the most important elements that determine quality in plants. We assessed the potential to discriminate differences in nitrogen concentration using high-resolution reflectance by growing Cenchrus ciliaris grass with different fertilization treatments in a greenhouse. Canopy spectral measurements from each treatment were taken under controlled laboratory conditions within a period of 4 weeks using a GER 3700 spectroradiometer. Results show that there were statistically significant differences in spectral reflectance between treatments within certain wavelength regions—an encouraging result for classifying and mapping grasslands with different levels of nutrients using hyperspectral remote sensing. We further investigated the effect of varying nitrogen supply to a specific absorption feature in the visible between 550 and 750 nm (R 550– 750) using continuum-removed spectra. Results show that the high nitrogen treatment had deeper and wider absorption pits as compared to the low nitrogen treatment as well as the control (no nitrogen), which is important for the prediction of nitrogen in grass canopies. This is a promising result for the remote sensing of canopy chemistry since emphasis can be shifted from the mid-infrared region (which is highly masked by water absorption) to the visible region. Overall, the results provide the possibility to map variation in pasture quality using hyperspectral remote sensing.