Evaluating an unmanned aerial vehicle‐based approach for assessing habitat extent and condition in fine‐scale early successional mountain mosaics

Abstract

Nomenclature Castroviejo (1986Castroviejo ( -2013 for published volumes, Franco (1984) and Franco & Afonso (1998) for other groups, except Agrostis (Romero Garc ıa et al. 1988) Abbreviations DSM = digital surface model; EU = European Union; GCP = ground control points; GPS = differential global positioning system; UAV = unmanned aerial vehicle; RF = random forest Abstract Question: Can very high-resolution colour orthophotography and digital surface models (DSMs) from an unmanned aerial vehicle (UAV) be effectively used for assessment of habitat extent and condition in fine-scale disturbancedependent mosaics? Location: Serra de Arga mountain range, a Natura 2000 protected site in the NW region of Portugal where drastic changes in pastoral activities have occurred over recent decades.Methods: An UAV platform was used to collect very high-resolution (6 cm) images and to produce a DSM (10 cm). From these data, several features were extracted related to colour, band ratios, as well as texture features calculated from colour imagery and surface elevation. Based on a systematic sampling design, field data were collected for both training and validation of a supervised classifier. Extracted features and ground truth training data were combined to calibrate a pixel-based Random forest classifier, with the purpose of devising a habitat map for the entire study area. Map validation was performed to assess classification accuracy, and feature importance metrics were calculated.Results: Validation results revealed good mean overall accuracy (0.89), with some performance decrease in situations of high interspersion of habitat types. The priority habitat type 6230* (Nardus grasslands), defining the vegetation matrix of the test site, obtained 0.96 and 0.91, considering, respectively, producer and user accuracy. In turn, priority habitat type 4020* (Atlantic wet heathlands) recorded 0.68 and 0.77. The obtained habitat map allowed measurement of the extent, description of the spatial arrangement and provided an indication of the conservation condition of target habitat types. Test results regarding the discrimination ability of different features highlighted the importance of surface elevation textures derived from the DSM, followed by band ratios textures and other more complex texture features calculated from colour imagery.Conclusions: Overall, the developed methodology showed promising results for assessing the extent and condition of habitats of high conservation priority in fine-scale, dynamic vegetation mosaics. Future advances in the use of UAV platforms may play an important role in monitoring protected sites and fulfil legal reporting obligations of EU member states, while reducing the costs associated with intensive in-field assessments.

Estimating Invasion Success by Non-Native Trees in a National Park Combining WorldView-2 Very High Resolution Satellite Data and Species Distribution Models

“…In contrast to previous studies, the classification was performed using spectral-derived (e.g., NDVI) and contextual information (e.g., slope). The application of ancillary layers during OBIA classification is increasingly used [12,17,62], since image band derivatives and ancillary data sources provide useful information to help distinguish between spectrally inseparable vegetation classes and enhance classification [62]. Große-Stoltenberg [63] demonstrated that vegetation indices estimated with hyperspectral data successfully contributed to distinguish Acacia longifolia and the RF classifier presented the best classification performance at the canopy level.…”

“…In this front, Object-Based Image Analysis (OBIA) techniques have facilitated the use of very high resolution (VHR) optical imagery (e.g., WorldView-2) [13][14][15], namely with sub-metric spatial resolution of image pixels. The merging of terrain and spectral data into image classifications [16,17] and the use of texture measures as contextual descriptors [18] has also enhanced the classification ability of machine-learning algorithms including random forests (RF) [19,20] and support vector machines (SVMs) [21].…”

Comparing ultra‐high spatial resolution remote‐sensing methods in mapping peatland vegetation

“…Spatial patterns of peatlands can be tracked with remotely sensed data, but it has been argued that the spatial resolution in common mapping approaches is too coarse (Palace et al, 2018). Ultra-high spatial resolution (UHSR) remote sensing, which provides data with cm-level pixel size, can reveal such patterns in vegetation composition that are lost in coarser resolution (Díaz-Varela, Calvo Iglesias, Cillero Castro, & Díaz Varela, 2018;Gonçalves et al, 2016;Lehmann et al, 2016;Mora, Vieira, Pina, Lousada, & Christiansen, 2015). In particular, the benefits of UHSR are evident in fragmented landscapes such as peatlands (Arroyo- Mora, Kalacska, Lucanus, Soffer, & Leblanc, 2017;Lehmann et al, 2016;Lovitt et al, 2017;Palace et al, 2018).…”

Adapting a Natura 2000 field guideline for a remote sensing-based assessment of heathland conservation status

“…The authors conclude that UAV missions could help to support field surveys and could play a major role in future monitoring tasks. In further related UAV applications, Dufour et al (2013) tested RGB-mosaics and digital surface models (DSMs) derived from UAV flights for mapping riparian vegetation and Gonçalves et al (2016) used high resolution color orthophotography and DSMs derived from UAV for assessing habitat extent and condition by applying a Random Forest classifier in a heathland ecosystem. Both studies confirmed the potential of UAV systems to supplement field-work by providing spatially continuous data.…”