How canopy shadow affects invasive plant species classification in high spatial resolution remote sensing

Plant invasions can result in serious threats for biodiversity and ecosystem functioning. Reliable maps at very-high spatial resolution are needed to assess invasions dynamics. Field sampling approaches could be replaced by unmanned aerial vehicles (UAVs) to derive such maps. However, pixel-based species classification at high spatial resolution is highly affected by within-canopy variation caused by shadows. Here, we studied the effect of shadows on mapping the occurrence of invasive species using UAV-based data. MaxEnt one-class classifications were applied to map Acacia dealbata, Ulex europaeus and Pinus radiata in central-south Chile using combinations of UAV-based spectral (RGB and hyperspectral), 2D textural and 3D structural variables including and excluding shaded canopy pixels during model calibration. The model accuracies in terms of area under the curve (AUC), Cohen’s Kappa, sensitivity (true positive rate) and specificity (true negative rate) were examined in sunlit and shaded canopies separately. Bootstrapping was used for validation and to assess statistical differences between models. Our results show that shadows significantly affect the accuracies obtained with all types of variables. The predictions in shaded areas were generally inaccurate, leading to misclassification rates between 65% and 100% even when shadows were included during model calibration. The exclusion of shaded areas from model calibrations increased the predictive accuracies (especially in terms of sensitivity), decreasing false positives. Spectral and 2D textural information showed generally higher performances and improvements when excluding shadows from the analysis. Shadows significantly affected the model results obtained with any of the variables used, hence the exclusion of shadows is recommended prior to model calibration. This relatively easy preprocessing step enhances models for classifying species occurrences using highresolution spectral imagery and derived products. Finally, a shadow simulation showed differences in the ideal acquisition window for each species, which is important to plan revisit campaigns.

[1]  Thomas A. Groen,et al.  Transferability of species distribution models: The case of Phytophthora cinnamomi in Southwest Spain and Southwest Australia , 2016 .

[2]  E. Marchante,et al.  Mapping the Flowering of an Invasive Plant Using Unmanned Aerial Vehicles: Is There Potential for Biocontrol Monitoring? , 2018, Front. Plant Sci..

[3]  Dimitri Lague,et al.  3D Terrestrial LiDAR data classification of complex natural scenes using a multi-scale dimensionality criterion: applications in geomorphology , 2011, ArXiv.

[4]  María Flor Álvarez-Taboada,et al.  Mapping of the Invasive Species Hakea sericea Using Unmanned Aerial Vehicle (UAV) and WorldView-2 Imagery and an Object-Oriented Approach , 2017, Remote. Sens..

[5]  Steven E. Franklin,et al.  Northern Conifer Forest Species Classification Using Multispectral Data Acquired from an Unmanned Aerial Vehicle , 2017 .

[6]  Matthew J. Smith,et al.  Protected areas network is not adequate to protect a critically endangered East Africa Chelonian: Modelling distribution of pancake tortoise, Malacochersus tornieri under current and future climates , 2013, bioRxiv.

[7]  P. Binggeli.  A taxonomic, biogeographical and ecological overview of invasive woody plants , 1996 .

[8]  K. Moffett,et al.  Remote Sens , 2015 .

[9]  Antonio Robles-Kelly,et al.  Shadow modelling based upon Rayleigh scattering and Mie theory , 2014, Pattern Recognit. Lett..

[10]  Roger Alex Clapp,et al.  The Unnatural History of the Monterey Pine , 1995 .

[11]  S. Schmidtlein,et al.  Identification of high nature value grassland with remote sensing and minimal field data , 2017 .

[12]  Daniel Spring,et al.  Evaluating the Feasibility of Eradicating an Invasion , 2006, Biological Invasions.

[13]  W. V. Reid,et al.  Biodiversity hotspots. , 1998, Trends in ecology & evolution.

[14]  F. Fassnacht,et al.  Monitoring Andean high altitude wetlands in central Chile with seasonal optical data: A comparison between Worldview-2 and Sentinel-2 imagery , 2018, ISPRS Journal of Photogrammetry and Remote Sensing.

[15]  Ribana Roscher,et al.  Mapping raised bogs with an iterative one-class classification approach , 2016 .

[16]  Gregory Asner,et al.  Applications of Remote Sensing to Alien Invasive Plant Studies , 2009, Sensors.

[17]  Fabian Ewald Fassnacht,et al.  Differentiating plant functional types using reflectance: which traits make the difference? , 2018, Remote Sensing in Ecology and Conservation.

[18]  Henrik Skov Midtiby,et al.  Plant species classification using deep convolutional neural network , 2016 .

[19]  Karen Anderson,et al.  Lightweight unmanned aerial vehicles will revolutionize spatial ecology , 2013 .

[20]  Anthony Ricciardi,et al.  Should Biological Invasions be Managed as Natural Disasters? , 2011 .

[21]  Cang Hui,et al.  Tree invasions: patterns, processes, challenges and opportunities , 2013, Biological Invasions.

[22]  S. Schmidtlein,et al.  Mapping plant species in mixed grassland communities using close range imaging spectroscopy , 2017 .

[23]  Hongming Zhang,et al.  An Analysis of Shadow Effects on Spectral Vegetation Indexes Using a Ground-Based Imaging Spectrometer , 2015, IEEE Geoscience and Remote Sensing Letters.

[24]  N. R. Spencer,et al.  The biocontrol of gorse, Ulex europaeus, in Chile: a progress report. , 2000 .

[25]  G. Hoarau,et al.  The fate of the Arctic seaweed Fucus distichus under climate change: an ecological niche modeling approach , 2016, Ecology and evolution.

[26]  M. Vilà,et al.  Ecological impacts of invasive alien plants: a meta-analysis of their effects on species, communities and ecosystems. , 2011, Ecology letters.

[27]  H. Nagendra.  Using remote sensing to assess biodiversity , 2001 .

[28]  S. Schoenholtz,et al.  Buffer effects of streamside native forests on water provision in watersheds dominated by exotic forest plantations , 2015 .

[29]  Ron Kohavi,et al.  A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection , 1995, IJCAI.

[30]  Adrien Michez,et al.  Mapping of riparian invasive species with supervised classification of Unmanned Aerial System (UAS) imagery , 2016, Int. J. Appl. Earth Obs. Geoinformation.

[31]  Juha Hyyppä,et al.  Individual Tree Detection and Classification with UAV-Based Photogrammetric Point Clouds and Hyperspectral Imaging , 2017, Remote. Sens..

[32]  Matthew O. Anderson,et al.  Radiometric and Geometric Analysis of Hyperspectral Imagery Acquired from an Unmanned Aerial Vehicle , 2012, Remote. Sens..

[33]  Lin Liu,et al.  Object-Based Mangrove Species Classification Using Unmanned Aerial Vehicle Hyperspectral Images and Digital Surface Models , 2018, Remote. Sens..

[34]  S. Schmidtlein,et al.  Mapping an invasive bryophyte species using hyperspectral remote sensing data , 2016, Biological Invasions.

[35]  Robert P. Anderson,et al.  Maximum entropy modeling of species geographic distributions , 2006 .

[36]  Wen Liu,et al.  Object-Based Shadow Extraction and Correction of High-Resolution Optical Satellite Images , 2012, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[37]  C. Peota.  Novel approach. , 2011, Minnesota medicine.

[38]  Aníbal Pauchard,et al.  Survival and growth of Acacia dealbata vs. native trees across an invasion front in south-central Chile , 2011 .

[39]  Oliver Q. Whaley,et al.  Identifying species from the air: UAVs and the very high resolution challenge for plant conservation , 2017, PloS one.

[40]  M. Westoby,et al.  ‘Structure-from-Motion’ photogrammetry: A low-cost, effective tool for geoscience applications , 2012 .

[41]  P. Gong,et al.  Object-based Detailed Vegetation Classification with Airborne High Spatial Resolution Remote Sensing Imagery , 2006 .

[42]  P. Pyšek,et al.  Timing Is Important: Unmanned Aircraft vs. Satellite Imagery in Plant Invasion Monitoring , 2017, Front. Plant Sci..

[43]  Feng Liu,et al.  Addendum: Using Satellite Data for the Characterization of Local Animal Reservoir Populations of Hantaan Virus on the Weihe Plain, China. Remote Sens. 2017, 9, 1076 , 2018, Remote. Sens..

[44]  X. Briottet,et al.  Shadow detection in very high spatial resolution aerial images: A comparative study , 2013 .

[45]  M. Holmgren,et al.  Why have European herbs so successfully invaded the Chilean matorral? Effects of herbivory, soil nutrients, and fire , 2000 .

[46]  Manoj K. Arora,et al.  Land Cover Classification Using IRS LISS III Image and DEM in a Rugged Terrain: A Case Study in Himalayas , 2005 .

[47]  R. Mittermeier,et al.  Biodiversity hotspots for conservation priorities , 2000, Nature.

[48]  Vassilia Karathanassi,et al.  De-shadowing of airborne imagery using at-sensor downwelling irradiance data , 2014 .

[49]  Adriaan van Niekerk,et al.  Incorporating risk mapping at multiple spatial scales into eradication management plans , 2014, Biological Invasions.

[50]  B. Koch,et al.  TREESVIS-A SOFTWARE SYSTEM FOR SIMULTANEOUS 3 D-REAL-TIME VISUALISATION OF DTM , DSM , LASER RAW DATA , MULTISPECTRAL DATA , SIMPLE TREE AND BUILDING MODELS , 2004 .

[51]  H. Ishii,et al.  The role of crown architecture, leaf phenology and photosynthetic activity in promoting complementary use of light among coexisting species in temperate forests , 2010, Ecological Research.

[52]  Anita Simic Milas,et al.  Different colours of shadows: classification of UAV images , 2017 .

[53]  S. Marhan,et al.  Forest Soil Phosphorus Resources and Fertilization Affect Ectomycorrhizal Community Composition, Beech P Uptake Efficiency, and Photosynthesis , 2018, Front. Plant Sci..

[54]  Junichi Kurihara,et al.  A novel approach for vegetation classification using UAV-based hyperspectral imaging , 2018, Comput. Electron. Agric..

[55]  Klara Dolos,et al.  Comparing Generalized Linear Models and random forest to model vascular plant species richness using LiDAR data in a natural forest in central Chile , 2016 .

[56]  M. J. Nigam,et al.  Shadow Detection and Removal from Remote Sensing Images using NDI and Morphological Operators , 2012 .

[57]  Mauricio Galleguillos,et al.  Predicting Vascular Plant Diversity in Anthropogenic Peatlands: Comparison of Modeling Methods with Free Satellite Data , 2017, Remote. Sens..

[58]  M. Neteler,et al.  Potential of remote sensing to predict species invasions , 2015 .

[59]  G. Kattenborn,et al.  PILOT STUDY ON THE RETRIEVAL OF DBH AND DIAMETER DISTRIBUTION OF DECIDUOUS FOREST STANDS USING CAST SHADOWS IN UAV-BASED ORTHOMOSAICS , 2018, ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences.

[60]  Barbara Koch,et al.  Automatic Single Tree Detection in Plantations using UAV-based Photogrammetric Point clouds , 2014 .

[61]  Madodomzi Mafanya,et al.  Evaluating pixel and object based image classification techniques for mapping plant invasions from UAV derived aerial imagery: Harrisia pomanensis as a case study , 2017 .