Estimating seed production quantification using drone imagery in the Walloon region (Belgium)

Walloon forests are dominated (65%) by three main tree species: oak, beech, and spruce, often occurring in monospecific stands. This homogeneity makes forest habitats particularly vulnerable to climate change, as illustrated by bark beetle outbreaks and beech decline. Functional diversification of forest stands therefore appears to be a priority pathway for restoring more resilient forests by mobilizing tree species already present within the landscape.

This study focuses on the ecological trajectory of clear-cut areas by examining propagule production capacity at the landscape scale, with particular attention to two species that are essential for the recolonization of clear-cuts in the Belgian Ardennes: sessile oak (Quercus petraea) and silver birch (Betula pendula). The objective is to estimate propagule production of these two species based on their multispectral emissions during the seed production period.

Ten sites per species were selected and surveyed every two weeks using a DJI Mavic 3M drone equipped with a multispectral camera. In addition, approximately twenty individuals per site were selected, and their fruit production was estimated through ground-based observations. Finally, phenological monitoring images were taken for each studied individual at a height of two meters above the canopy.

The variability in seed production among individuals provided a sufficiently wide gradient to highlight a relationship between seed production and individual spectral signals. However, a second measurement campaign is required, with an extended monitoring period, to strengthen these results.

Estimating annual seed production makes it possible to assess the potential for dispersal and establishment within a given clear-cut site targeted for restoration. This information will contribute to better adaptation of silvicultural management practices at the stand level by integrating landscape-scale propagule availability.