2020•PLoS ONEOpen access
Integrating high resolution drone imagery and forest inventory to distinguish canopy and understory trees and quantify their contributions to forest structure and dynamics
Raquel Fernandes Araujo, Jeffrey Q. Chambers, Carlos Henrique Souza Celes, Helene C. Muller‐Landau, Ana Paula Ferreira dos Santos, Fabiano Emmert, Gabriel H. P. M. Ribeiro, Bruno Gimenez, Adriano José Nogueira Lima, Moacir Alberto Assis Campos, Níro Higuchi
Open full text 33 citations
Abstract
Tree growth and survival differ strongly between canopy trees (those directly exposed to overhead light), and understory trees. However, the structural complexity of many tropical forests makes it difficult to determine canopy positions. The integration of remote sensing and ground-based data enables this determination and measurements of how canopy and understory trees differ in structure and dynamics. Here we analyzed 2 cm resolution RGB imagery collected by a Remotely Piloted Aircraft System (RPAS), also known as drone, together with two decades of bi-annual tree censuses for 2 ha of old growth forest in the Central Amazon. We delineated all crowns visible in the imagery and linked each crown to a tagged stem through field work. Canopy trees constituted 40% of the 1244 inventoried trees with diameter at breast height (DBH) > 10 cm, and accounted for ~70% of aboveground carbon stocks and wood productivity. The probability of being in the canopy increased logistically with tree diameter, passing through 50% at 23.5 cm DBH. Diameter growth was on average twice as large in canopy trees as in understory trees. Growth rates were unrelated to diameter in canopy trees and positively related to diameter in understory trees, consistent with the idea that light availability increases with diameter in the understory but not the canopy. The whole stand size distribution was best fit by a Weibull distribution, whereas the separate size distributions of understory trees or canopy trees > 25 cm DBH were equally well fit by exponential and Weibull distributions, consistent with mechanistic forest models. The identification and field mapping of crowns seen in a high resolution orthomosaic revealed new patterns in the structure and dynamics of trees of canopy vs. understory at this site, demonstrating the value of traditional tree censuses with drone remote sensing.
Open-access reader