Submitted to: The Rangeland Journal
Publication Type: Peer reviewed journal
Publication Acceptance Date: 12/1/1999
Publication Date: N/A
Citation: NORTHUP, B.K., BROWN, J.R., DIAS, C.D., SKELLY, W.C., RADFORD, B. A. 1999. TECHNIQUE FOR NEAR-GROUND REMOTE SENSING OF HERBACEOUS VEGETATION IN TROPICAL WOODLANDS. THE RANGELAND JOURNAL. v. 21(2). p. 229-243. Interpretive Summary: A study was undertaken to develop a flexible and adaptable near-ground remote sensing system that allowed description of the herbaceous layer of the dry tropical woodlands in northern Australia. The objective was to develop a system that could describe spatially referenced estimates of projected herbaceous cover, bare ground and standing crop with a level of accuracy similar to standard field techniques currently used in Australia. The system consisted of a gantry frame and telescoping boom arm mounted on a quad bike, digital camera (or analog camera using slide film) to capture plot images, Leica TC1010 station (electronic theodolite with data capture card) to measure distance and bearings to centers of imaged plots, and a suite of readily available software packages to process the images. Information quality and time required for the developed system and the BOTANAL field estimation technique to describe herbaceous canopy cover, bare ground and standing crop were undertaken. The remote sensing procedures were as accurate as BOTANAL techniques at estimating canopy cover or bare ground, with small increases in time required, but less accurate at describing standing crop. Also, the remote sensing system produced measures of cover with lower levels of variation (and hence more accurate) than BOTANAL techniques. Since the collected data was also spatially referenced (not currently used with BOTANAL procedures), accurate maps of cover or bare ground for managed pastures could be developed to better describe landscape function. Strengths and weaknesses of both systems were also identified and discussed.
Technical Abstract: Cost-effective, accurate, repeatable methods for assaying vegetative characteristics are critical to develop management strategies for grasslands in tropical Australia. We outline a set of flexible and adaptable techniques that allow the description of plant canopy cover, bare ground, and standing crop through near-ground remote sensing. The system consisted of a gantry frame and telescoping boom arm mounted on a quad bike, digital camera to capture plot images, Leica TC1010 total station (electronic theodolite with data capture card to measure distance and bearings to centers of imaged plots) and a suite of readily available software packages to process the images. During 1997, data were collected from three experimental paddocks in different stages of degradation within a dry tropical woodland located near Charters Towers, Queensland to compare data quality and time requirements for digital and standard field estimation (BOTANAL) techniques. The near-ground remote sensing system wa as accurate as BOTANAL at estimating herbaceous canopy cover or bare ground, with only minimal increases in time requirements. However, the system was less accurate at defining standing crop, due to the 3- dimensional nature of biomass distribution within tussock canopies. The remote sensing system also had better control of within and between plot variance, and was capable of producing more accurate mean estimates of cover. Since the collected data was spatially referenced (not currently used with BOTANAL), accurate distribution maps of cover could be interpolated to define landscape organization and response to management. Strengths and weaknesses of the 2 systems were also discussed.