From communities to individuals: Using remote sensing to inform and monitor woodland restoration

Abstract

Individual-tree data from sensors mounted on drones are uncovering genetic-based differences in crown architecture traits. Understanding how these differences interplay with animal habitat will inform provenance choices in climate-ready restoration.

Four reasons why functional traits are not being used in restoration practice

“…Despite important challenges and limitations (e.g. "big data" storage and processing, exclusion of belowground and seed traits), the ability of remote sensing technology to track an array of aboveground traits through time and across management actions will facilitate both planning and monitoring of ecological restoration, and guide adaptive management (Harrison et al 2021) (Fig. 1).…”

Existing and emerging uses of drones in restoration ecology

“…This planning phase typically involves baseline surveys (incl. Species/ecosystem/vegetation surveys, abiotic condition surveys, characterisation of types and degrees of ecosystem degradation), reference ecosystem surveys, goal development (including scenario planning), restoration intervention option analysis, and mapping of associated logistical issues (Gann et al, 2019; Harrison, Camarretta, et al, 2021; Harrison, Davidson, et al, 2021). Drones can have a demonstrable role in each step, which we outline below.…”

“…Nonetheless, it has been estimated that using drones to survey 100 ha of land is ~10 times more efficient than surveys based on traditional field surveys (Filipovs et al, 2021). In addition, a recent study linked drone photogrammetry derived vegetation structural traits with animal tracking to understand the structural components required to elucidate animal behaviour (Harrison, Camarretta, et al, 2021; Harrison, Davidson, et al, 2021). Lee et al (2021) pointed out that other approaches may be more efficient because automated detection systems require significant amounts of time to generate training data and train detection models.…”

“…Available sensors include active sensors, such as light detection and ranging (LiDAR) modules, which use laser‐based technology to generate very high‐density point clouds (e.g., >1,000 points/m 2 ) to detect objects at fine scales such as individual trees to model provenance performance differences (Camarretta et al, 2020). Passive sensors are also available, such as optical sensors, including high‐definition cameras that can be used in combination with object recognition algorithms to monitor species and ecological communities (Dalla Corte et al, 2020; Hamylton et al, 2020; Harrison, Camarretta, et al, 2021; Harrison, Davidson, et al, 2021; Lu et al, 2020). In addition, drones can carry other payloads that are not sensors.…”