Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Australie → Prediction and assessment of forage production and ecological function in wooded rangelands using remote sensing

University of New South Wales (1998)

Prediction and assessment of forage production and ecological function in wooded rangelands using remote sensing

Knight, Arthur William

Titre : Prediction and assessment of forage production and ecological function in wooded rangelands using remote sensing

Auteur : Knight, Arthur William

Etablissement de soutenance : University of New South Wales (Australia)

Grade : Doctor of Philosophy (PhD) 1998

Résumé
A new, spatially explicit, remote sensing method using five Landsat Thematic Mapper (TM) images to predict forage production was developed. The method also assessed ecological functions of forage change in wooded semi-arid rangelands of Eastern Australia. The aim was to define characteristics of cover change to produce a method to interpret rangeland function, its potential for production, and capacity to sustain functions at their scale of occurrence. Abstract : A new, spatially explicit, remote sensing method using five Landsat Thematic Mapper (TM) images to predict forage production was developed. The method also assessed ecological functions of forage change in wooded semi-arid rangelands of Eastern Australia. The aim was to define characteristics of cover change to produce a method to interpret rangeland function, its potential for production, and capacity to sustain functions at their scale of occurrence. Four Classes (mosaics) of cover change are defined by change in cycles of cover change (in patch points) to define patch-mosaic cover change stability. Cover change cycles are driven by pluriannual cycles (large seasonal rainfall, drought and grazing effects). Within each Class, indices of cover increase (mean cover increase and change in cover increase) are derived using TM band 3 in good habitats of perennial forage production over two periods of cover increase. The index of mean cover increase is combined with field data to predict palatable perennial grass production four years and eight months after image acquisition, after five years of El Nino driven climate change (drought). A compilation of simple associations between forage and soil patch attributes shows that unique functions of stability (constancy) and resilience (tolerance to change) occur in each Class. These functions sustain different forage types and have different capacities to improve function quality and size of production as soil conditions improve (which indicates potential rehabilitation benefit). The remote sensing indices of cover increase and change in cycles of cover change reveal a history of Class associations that can be used to interpret rangeland function and function change using remote sensing methods alone. The primary product is a successful model of production assessment. The model accounts for the ecological functions of cover change at their spatial and temporal scales of occurrence. An appropriately scaled and calibrated model for cover change assessment could provide information to manage forage dynamics, the value of production, and potential production following habitat improvement (e.g., rehabilitation), even when rangeland functions (e.g., production) are infrequent and heterogeneously mixed. A Class of pre-eminent productive potential, stability and resilience (especially for palatable perennial grass) was found, which also had the greatest capacity to improve habitat functions and productive potential (by up to 30%), through simple improvements to soil condition and function

Présentation (ALWELAIE)

Page publiée le 2 février 2018