LiDAR and multispectral imagery predict the occurrence of tree cavities suitable for a cavity‐nesting duck

• DOI: 10.1002/rse2.236
• Publication Date: 2021-08-20T10:27:41Z
• Authors: Edmund J. Zlonis; Ram Deo; James B. Berdeen
• Keywords: Tree (set theory); Aerial imagery

Abstract Tree cavities are an essential habitat component for wildlife species across diverse taxa, from insects to large mammals. Many of these imperiled by loss cavities. Further, conservation action is hindered limited information on the spatial distribution cavities, largely due difficulties in developing useful models their presence or abundance. Accurately predicting fine‐scale, landscape‐wide, important features would greatly benefit measures. In this study, we evaluated efficacy using remotely sensed data, including LiDAR, multispectral imagery, and SAR, predict locations suitable nesting Wood Ducks Aix sponsa at fine scales (≤40 × 40 m) a broad landscape (~254 000 ha). We used Random Forest classify presence–absence four prediction (5, 10, 20, 40‐m pixels) as well three groupings predictor variables (LiDAR‐derived metrics, forest‐inventory [derived via LiDAR imagery], ancillary data SAR imagery]), then compared accuracy between models. The 20‐m response‐scale had highest accuracy. Variables each were relied imagery data. Our final model predictive map 84% overall Out Bag and, when tested with independent cavity dataset, correctly identified 80% trees presences. can be researchers land‐managers determine how past management actions have affected availability complexes potentially, other secondary‐cavity‐nesting wildlife. addition, our analysis serve methodological case‐study cavity‐nesting regions, use like high‐density increases.