Identifying and mapping invasive alien plant individuals   and stands from aerial photography and satellite images in   the central Hawequa conservation area   Aurelia Therese Forsyth A thesis submitted in partial fulfilment of the requirements for the degree of Magister Scientiae, in the Department of Biodiversity and Conservation Biology, University of the Western Cape Supervisor: Dr. Richard S. Knight Co-Supervisor: Dr. Helen M. de Klerk November 2012 i Identifying and mapping invasive alien plant individuals   and stands from aerial photography and satellite images in   the central Hawequa conservation area   Aurelia Therese Forsyth KEYWORDS Acacia species Fynbos High-resolution imagery Image classification Invasive alien plants Pinus species Plant densities Prioritising clearing Remote sensing WorldView-2 satellite imagery ii ABSTRACT   Identifying and mapping invasive  alien plant individuals and stands from aerial photography and satellite images in the central Hawequa   conservation area A.T. Forsyth MSc Thesis, Department of Biodiversity and Conservation Biology, University of the Western Cape The Cape Floristic Region, situated at the southern tip of Africa, is one of the world’s most botanically diverse regions. The biodiversity of this region faces various types of threats, which can be divided into three main categories, namely increasing urbanisation, agriculture expansion, and the spread of invasive alien vegetation. It has been shown that botanically diverse areas are more prone to invasion by invasive alien plant (IAP) species. The Hawequa conservation area, in the south-western Cape, is particularly botanically diverse, such that it is very prone to aggressive invasion by IAP species. Therefore, conservation management of the Hawequa conservation area urgently need to map, prioritise and clear IAP species. Due to the topographical complexity of this mountainous area, it is not possible to map the distribution of IAP species throughout the protected area by conventional field methods. Remote sensing may be able to provide a suitable alternative for mapping. The aim of this research was to assess various image classification methods, using two types of high-resolution imagery (colour aerial photography and WorldView-2 satellite imagery), in order to map the distribution of IAP species, including small stands and individuals. Specifically, the study will focus on mapping Pinus and Acacia spp. in a study site of approximately 9 225ha in the Hawequa conservation area. iii Supervised classification was performed using two different protocols, namely   per-pixel and per-field. For the per-pixel classification Iterative Self-Organising Data Analyses Technique (ISODATA) w as used, a method supported by ERDAS Imagine. The per-field (object-based) classification was done using fractal net   evolution approach (FNEA), a method supported by eCognition. The per-pixel classification mapped the extent of Pinus and Acacia spp. in the study area as 1 205.8 ha (13%) and 80.1 ha (0.9%) respectively, and the per- field classification as 1 120.9 ha (12.1%) and 96.8 ha (1.1%) respectively. Accuracy assessments performed on the resulting thematic maps generated from these two classification methods had a kappa coefficient of 0.700 for the per-pixel classification and 0.408 for the per-field classification. Even though the overall extent of IAP species for each of these methods is similar, the reliability of the actual thematic maps is vastly different. These findings suggest that mapping IAP species (especially Pinus spp.) stands and individuals in highly diverse natural veld, the traditional per-pixel classification still proves to be the best method when using high-resolution images. In the case of Acacia spp., which often occurs along rivers, it is more difficult to distinguish them from the natural riverine vegetation. Using WorldView-2 satellite images for large areas can be very expensive (approximately R120 per km2 in 2011), but in comparison with the cost of mapping and the subsequent clearing, especially in inaccessible areas, it might be a worthwhile investment. Alternative image sources such as the high- resolution digital colour infrared aerial photography must be considered as a good source for mapping IAP species in high altitude areas. November 2012 iv DECLARATION   I declare that Identifying and mapping invasive alien plant individuals and stands   from aerial photography and satellite images in the central Hawequa conservation area is my own work, that it has not been submitted before for any degree or examination in any other university, and that all the sources I have used or quoted have been indicated and acknowledged as complete reference. Aurelia Therese Forsyth November 2012 Signed: …………………………………………………………. v ACKNOWLEDGEMENTS   I would like to thank the following people for their assistance with this research:   My thanks go to supervisor Dr. Helen de Klerk for her guidance, support and advice. Thank you to her for also facilitating the purchase of the WorldView-2 satellite images used in this project through the University of Stellenbosch. Many thanks to Dr. Richard Knight for his supervision. A special thanks to the following colleagues and experts for their availability to assist with advice on various aspects such as sampling, image processing, and image classification: Dr. Wesley Roberts (CSIR), Kevin Shaw (CapeNature), Dr. Andrew Turner (CapeNature), Mark Thompson (GeoTerraImage), Fanie Ferreira (GeoTerraImage), and Wolfgang Lück. In addition I would also like to thank Arthur and Stephen Chapman for assisting with high-altitude fieldwork and sampling. Then, most of all I would like to thank Greg Forsyth (my husband) for his support with botanical advice, and also high-altitude fieldwork and sampling. vi CONTENTS   TITLE PAGE   KEYWORDS ii ABSTRACT   iii DECLARATION v ACKNOWLEDGEMENTS vi CONTENTS vii LIST OF TABLES ix LIST OF FIGURES xiii LIST OF ACRONYMS xvi GLOSSARY I – TERMS xiii GLOSSARY II – FILE EXTENSIONS xv Chapter 1: GENERAL INTRODUCTION 1 1.1. Introduction....................................................................................................... 1 1.2. IAP species ...................................................................................................... 2 1.3. Selection of IAP species to map ....................................................................... 4 1.4. Remote Sensing and GIS in IAP species mapping ........................................... 5 1.5. Aim of study ...................................................................................................... 7 1.6. Objectives ........................................................................................................ 7 1.7. Fynbos biome ................................................................................................... 8 Chapter 2: LITERATURE REVIEW 11 2.1. Introduction..................................................................................................... 11 2.2. IAP species: What it is .................................................................................... 13 2.3. IAP species: Impact on ecosystems ............................................................... 14 2.4. Mapping of IAP species using remote sensing ............................................... 16 2.5. Choosing appropriate remotely sensed imagery for the research ................... 19 2.6. Remote sensing: Classification techniques ..................................................... 25 2.7. Remote sensing: Protocols and algorithms ..................................................... 29 2.8. Conclusion...................................................................................................... 35 vii Chapter 3: RESEARCH METHODS 38   3.1. Introduction..................................................................................................... 38 3.2. Data acquisition ............................. ................................................................. 38 3.3. Study area selection and description .............................................................. 40   3.4. Selection of IAP species for mapping ............................................................. 42 3.5. Pre-processing of satellite images .................................................................. 42 3.6. Selection of vegetation information classes .................................................... 45 3.7. Survey design for the selection of training and reference sites ....................... 48 3.8. Classification Protocol .................................................................................... 53 3.9. Accuracy assessment ..................................................................................... 64 Chapter 4: RESULTS: PRESENTATION AND DISCUSSION 67 4.1. Introduction..................................................................................................... 67 4.2. Summary of results per method ...................................................................... 68 4.3. Accuracy of results per method ...................................................................... 73 4.4. General discussion ......................................................................................... 86 4.5. Efficiency per method ..................................................................................... 91 4.6. Conclusion...................................................................................................... 92 Chapter 5: CONCLUSIONS AND RECOMMENDATIONS 94 5.1. Introduction..................................................................................................... 94 5.2. Application and limitations of remote sensing in mapping IAP species, using high resolution imagery ................................................................................. 94 5.3. Recommendations ........................................................................................ 102 REFERENCES 103 APPENDICES 116 viii LIST OF TABLES Table 1. A list of multispectral satellite i mage products available in South Africa. The resolution reflected in the table below is the finest resolution available within   all the bands provided with the image. The following acronyms for the bands were used: NIR = near infrared, MIR = mid infrared, TIR = thermal infrared. 20 Table 2. Details of two adjacent WorldView-2 images received from the Satellite Application Centre (SAC). 39 Table 3. Number of sample sites (reflected as actual numbers) calculated per vegetation information classes (rows) stratified across the different landform categories (columns) were randomly selected for use during the image classification. 50 Table 4. Percentage calculation of where the training sites plot within the north and south aspects. Number of training sites randomly selected per vegetation information classes (rows) overlayed with the aspect shapefile (columns). 52 Table 5. The different settings used for the spectral Euclidean distance (recorded in digital numbers) and geographical constraint (number of pixels) during the capturing of the areas of interest (AOI) per vegetation information class. 56 Table 6. Summed areas classified per vegetation information class extracted from the per-pixel classified thematic map. The percentages calculated represent the actual invasive alien plants (IAP) species cover. 69 Table 7. The summarised areas classified per invasive alien plants (IAP) species, Afrotemperate forest, and ‘other’ extracted from the per-pixel classified thematic map. The Pinus spp. vegetation information classes were combined to give one area. 70 Table 8. The summarised areas classified per vegetation information class extracted from the per-field classified thematic map. The actual areas for invasive alien plants (IAP) species were calculated using average percentages per pre- defined density categories. The remainder areas not covered by IAP species were added as ‘other’. 71 ix Table 9. Summed areas, classified per invasive alien plants (IAP) species,   Afrotemperate forest, and ‘other’ extracted from the per-field classified thematic map. The Pinus spp. vegetation informa tion classes were combined to give one area. 72   Table 10. Comparison of the results of the summed areas, classified per invasive alien plants (IAP) species, Afrotemperate forest, and ‘other’ extracted from the per-pixel and per-field classified thematic map. 73 Table 11. Definition and mathematical calculation summary of the five calculations performed within each confusion matrix (Campbell 1996). 74 Table 12. Confusion matrix to assess the accuracy of the per-pixel classified thematic map. This matrix was done using all the vegetation information classes. The following acronyms were used in the matrix; Acacia = Acacia stand dense, Afro.forest = Afrotemperate forest, P. indiv. = Pinus individual, P. sparse = Pinus stand sparse, P. scattered = Pinus stand scattered, P. dense = Pinus stand dense, Omission = Omission error (%), Commission = Commission error (%), Prod. acc = Producer’s accuracy (%), and Cons. acc. = Consumer’s accuracy. The diagonal values represent accurately classified pixels (match between classes assigned to pixels by the classification and reference sites). 74 Table 13. Confusion matrix to assess the accuracy of the per-pixel classified thematic map. For this matrix all the Pinus spp. vegetation information classes were combined. The following acronyms were used in the matrix; Acacia = Acacia stand dense, Afro.forest = Afrotemperate forest, Omission = Omission error (%), Commission = Commission error (%), Prod. acc = Producer’s accuracy (%), and Cons. acc. = Consumer’s accuracy. The diagonal values represent accurately classified pixels (match between classes assigned to pixels by the classification and reference sites). 76 Table 14. Confusion matrix to assess the accuracy of the original level one per- field classified thematic map. This matrix was done using all the vegetation information classes, excluding ‘Pinus individual’. The following acronyms were used in the matrix; Acacia = Acacia stand dense, Afro.forest = Afrotemperate forest, P. sparse = Pinus stand sparse, P. scattered = Pinus stand scattered, P. dense = Pinus stand dense, Omission = Omission error (%), Commission = x