Exploring effective detection and spatial pattern of Prickly Pear Cactus (Opuntia genus) from airborne imagery before and after prescribed fires in the Edwards Plateau

Authors: JAIME, XAVIER - Texas A&M University; Angerer, Jay; Yang, Chenghai; WALKER, JOHN - Texas A&M Agrilife; MATA, JOSE - Texas A&M University; TOLLESON, DOUG - Texas A&M Agrilife; WU, X. BEN - Texas A&M University

Submitted to: Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/12/2023
Publication Date: 8/15/2023
Citation: Jaime, X., Angerer, J.P., Yang, C., Walker, J., Mata, J., Tolleson, D., Wu, X. 2023. Exploring effective detection and spatial pattern of Prickly Pear Cactus (Opuntia genus) from airborne imagery before and after prescribed fires in the Edwards Plateau. Remote Sensing. 15(16). Article 4033. https://doi.org/10.3390/rs15164033.
DOI: https://doi.org/10.3390/rs15164033

Interpretive Summary: During the last 100 years, prickly pear cactus has been increasing on rangelands in the US and elsewhere. Detecting and mapping prickly pear has been challenging because of the mixture and diversity of plants across rangeland landscapes. A series of image classification methods were evaluated to determine which method would work best for mapping prickly pear abundance and pattern on the landscape. The image classification methods included object-based feature extraction, random forest machine learning, and spectral endmember classification. Field data on prickly pear were collected to identify prickly pear locations and sizes, determine plant cover, and detect spectral signatures. The field data were then used to train and test the image classification methods, and statistics from the testing allowed the identification of the best method. The spectral endmember classification method was found to be the most accurate, with an overall accuracy of 96%. It was then used to map prickly pear distribution and to quantify abundance before and after prescribed fire. A spatial comparison of the images before and after the fire revealed that fire reduced prickly pear abundance by 46.5%. However, reductions varied by soil type, with deeper soils having the highest reductions in prickly pear, likely because of higher fine fuel around prickly pear on these soils. Patterns of prickly pear on the landscape were also altered by fire and soil type. Results from this study indicate that mapping prickly pear on rangelands is feasible and can be used to aid in the design and effectiveness evaluation of rangeland management strategies such as fire.

Technical Abstract: Over the past century, prickly pear (PP) cactus (e.g., genus Opuntia; subgenus Platyopuntia) has increased on semi-arid rangelands. Effective detection of cacti abundance and spatial pattern is challenging due to the inherent heterogeneity of rangeland landscapes. In this study, high-resolution multispectral imageries (0.21-m) were used to test object-based (OB) feature extraction, random forest (RF) machine learning, and spectral endmember (n-D) classification methods to map PP and evaluate its spatial pattern. We trained and tested classification methods using field-collected GPS location, plant cover, and spectrometry from 288 2-m radius polygons before a prescribed burn and 480 samples after the burn within a 69.2-ha burn unit. The most accurate classification method was then used to map PP distribution and to quantify abundance before and after fire. As a case study, we assessed the spatial pattern of mapped PP cover considering topoedaphic setting and burn conditions. The results showed that the endmember classification method, spectral angle mapper (SAM), outperformed the RF and OB classification with higher kappa coefficients (KC) (0.93 vs. 0.82 and 0.23 respectively) and overall accuracies (OA) (0.96 vs. 0.91 and 0.49) from pre-fire imagery. KC and OA metrics of post-fire imagery were lower, but rankings among classification methods were similar. SAM classifications revealed that fire reduced PP abundance by 46.5%, but reductions varied by soil type, with deeper soils having greater decreases (61%). Kolmogorov-Smirnov tests indicated significant changes before and after fire in the frequency distribution of PP cover within deeper soils (D = 0.64, p=0.02). Two-way ANOVA revealed that the interaction of season (pre vs. post-fire) and soils significantly (p<0.00001) influenced the spatial pattern of PP patches. Fire also reduced the size and shape of PP patches depending on the topoedaphic settings. These findings provide opportunities for accurate mapping of PP to aid in the design and implementation of spatially-explicit rangeland management strategies, such as fire, that can help reduce and mitigate the ecological and economic impacts of prickly pear expansion.