Remote Sensing of Invasive Lantana camara (Verbenaceae) in Semiarid Savanna Rangeland Ecosystems of South Africa Timothy Dube a, * , Cletah Shoko b , Mbulisi Sibanda c , Paschaline Madileng d , Xivutiso G. Maluleke d , Velma R. Mokwatedi d , Lorencia Tibane d , Tebogo Tshebesebe d a Institute for Water Studies, Department of Earth Sciences, University of the Western Cape, Bellville, 7535, South Africa b Division of Geography, School of Geography, Archaeology and Environmental Studies, University of Witwatersrand, 2050, Johannesburg, South Africa c Discipline of Geography, School of Agriculture Earth & Environmental Sciences, University of KwaZulu Natal, Scottsville, 3209, Pietermaritzburg, South Africa d Department of Geography & Environmental Studies, University of Limpopo, Sovenga, 0727, South Africa article info Article history: Received 3 March 2019 Received in revised form 21 June 2019 Accepted 8 January 2020 Key Words: agroecosystems climate change environmental impacts invasive species livelihoods native species rangelands abstract Invasive Lantana camara (L. camara) is one of the key drivers of social-ecological and environmental change. Understanding its distribution is critical in determining its impact on the environment and livelihoods and in developing sustainable remediation and rehabilitation strategies. In this study we demonstrate the first comparative assessment of Landsat 8 Operational Land Imager (OLI) and Sentinel-2 Multispectral Instrument (MSI) satellite data in detecting and mapping of invasive L. camara from other land cover types (i.e., built up, fields/bare patches, grassland, and shrub) in semiarid rangeland eco- systems of South Africa. Discriminant analysis (DA) classification technique was used to detect and characterize the spatial distribution of L. camara using Landsat 8 OLI and Sentinel-2 derivatives (i.e., spectral bands, indices, and combined variables). Comparatively, the results show that Sentinel-2 data were able to detect and map L. camara with a high overall accuracy (78.4%) than Landsat 8 OLI, which yielded an accuracy of 65.5%. Further, Student’s t-test statistical analysis results showed that Sentinel-2 outperformed Landsat 8 (P < 0.05, Student’s t < 0.233) in mapping L. camara from other land cover types. High performance from Sentinel-2 data indicates the relevance and potential of characterizing and profiling invasive species with the new-generation sensors, a previously daunting task, with broadband multispectral sensors. © 2020 The Society for Range Management. Published by Elsevier Inc. All rights reserved. Introduction Savanna rangeland invaders, mainly Alien Invasive Plants (AIPs), cause far-reaching socioeconomic, environmental, and agro- ecosystem disturbances that can result in the loss of native species (Vil a et al. 2011). These plants effectively outcompete native species in the utilization of natural resources like nitrogen, water, and space and can swiftly replicate, which gives these species an advantage to spread across rangelands (Seabloom et al. 2003). Previous studies show that plant invaders like L. camara have the capability to maximize their invasion through the modification of the environments (i.e., nutrient and water cycling among others), creating unfavorable conditions for native species (Kruse et al. 2000; Bouda et al. 2001; Kohli et al. 2006; Ghayal and Dhumal 2011). In addition, invasive species like L. camara can harm liveli- hoods and increase vulnerability through encroaching on produc- tive land and reducing access or mobility. More importantly, they can reduce the supply of natural capital or resources critical for rural livelihoods and degrade rangelands, which can result in in- come losses and increased vulnerability. L. camara has potential negative impacts on human or wildlife/livestock health, besides its potential to reduce the landscape’s cultural value. If uncontrolled, the impacts of this species can minimize community resilience or adaptive capacity to further changes or invasion. As such, local and national governments in affected countries have constantly embarked on physical eradication and management efforts toward controlling these plants, but still their spread is on the rise. This is largely because some of the affected areas are unknown to au- thorities. For effective control mechanisms to be put in place, there is, therefore, a need for up-to-date, timely, and accurate informa- tion on the spatial distribution and configuration of these plants (Matongera et al. 2016). Accurate digital data sets and automated * Correspondence: Timothy Dube, Institute for Water Studies, Dept of Earth Sciences, University of the Western Cape, Private Bag X17, Bellville, 7535, South Africa, þ27735470645 E-mail address: tidube@uwc.ac.za (T. Dube). Contents lists available at ScienceDirect Rangeland Ecology & Management journal homepage: http://www.elsevier.com/locate/rama https://doi.org/10.1016/j.rama.2020.01.003 1550-7424/© 2020 The Society for Range Management. Published by Elsevier Inc. All rights reserved. Rangeland Ecology & Management xxx (xxxx) xxx Please cite this article as: Dube, T et al., Remote Sensing of Invasive Lantana camara (Verbenaceae) in Semiarid Savanna Rangeland Ecosystems of South Africa, Rangeland Ecology & Management, https://doi.org/10.1016/j.rama.2020.01.003