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

Accueil du site → Master → Afrique du Sud → 2019 → Multi-date remote sensing assessment of invasive Lantana camara species distribution in semi-arid savanna rangelands of South Africa.

University of KwaZulu-Natal (2019)

Multi-date remote sensing assessment of invasive Lantana camara species distribution in semi-arid savanna rangelands of South Africa.

Madileng, Ngwanamapotu Paschaline.

Titre : Multi-date remote sensing assessment of invasive Lantana camara species distribution in semi-arid savanna rangelands of South Africa.

Auteur : Madileng, Ngwanamapotu Paschaline.

Université de soutenance : University of KwaZulu-Natal

Grade : Master of Science in Environmental Sciences 2019

Résumé
Invasive alien species pose a massive threat to biodiversity globally. A comprehensive analysis of the spatial distribution of invasive species, like Lantana camara, is essential for providing appropriate management strategies on both a local and regional scale. The main aim of the study was to map and assess the spatial distribution of Lantana camara invasion in savanna rangeland ecosystems over time in Agincourt, South Africa, using high-resolution SPOT 6 data. The first objective of the study is to focus on reviewing the progress in the remote sensing of Lantana camara. A review of the literature shows that previous studies on mapping and monitoring invasive Lantana camara have relied on traditional methods, such as visual interpretations and field surveys, which have been insufficient, particularly for large-scale monitoring. The use of commercial satellite data with a high-resolution has demonstrated the potential for providing fine spectral and spatial resolution capabilities that are essential for offering precise and reliable data on the spatial distribution of invasive species. The challenges encountered in remote sensing of Lantana camara include the problem of similarity in the spectral signatures of the weed and other vegetation species, which leads to a low classification accuracy. The second objective was to map the spatial distribution and rate of change of Lantana camara in savanna ecosystems over time and space. To achieve this objective, a supervised maximum likelihood classification was used for the SPOT 6 satellite images acquired over a period of three years (2014, 2016 and 2018). The results showed that, Lantana camara was distributed over almost the whole study area for all the three years, yet it decreased with time, due to the clearing and disaster programs. Furthermore, the weed then increased and re-established itself between 2016 and 2018, due to failure to do follow-up control after the initial attempts to eradicate it and the failure to remove rootstock. It was again observed that Lantana camara species can be accurately detected and mapped with an overall classification accuracy of >80% for all the three years. However, research with the enhanced spatial and spectral capabilities such as SPOT 6, has shown the importance of remotely sensed data in predicting lantana camara distribution

Présentation

Version intégrale (1,6 Mb)

Page publiée le 22 janvier 2021