A multi-proxy framework to detect insect defoliations in tree rings: a case study on pine processionary

Assessing and reconstructing the impacts of defoliation caused by insect herbivores on tree growth, carbon budget and water use, and differentiating these impacts from other stresses and disturbances such as droughts requires multi-proxy approaches. Here we present a methodological framework to pinpoint the impacts of pine processionary moth (Thaumetopoea pityocampa), a major winter-feeding defoliator, on tree cover (remote-sensing indices), radial growth and wood features (anatomy, density, lignin/carbohydrate ratio of cell walls, δ13C and δ18O of wood cellulose) of drought-prone pine (Pinus nigra) forests in north-eastern Spain. We compared host defoliated (D) and coexisting non-defoliated (ND) pines along with non-host oaks (Quercus faginea) following a strong insect outbreak occurring in 2016 at two climatically contrasting sites (cool-wet Huesca and warm-dry Teruel). Changes in tree-ring width and wood density were analyzed and their responses to climate variables (including a drought index) were compared between D and ND trees. The Normalized Difference Infrared Index showed reductions due to the outbreak of –47.3% and –55.6% in Huesca and Teruel, respectively. The D pines showed: a strong drop in growth (–96.3% on average), a reduction in tracheid lumen diameter (–35.0%) and lower lignin/carbohydrate ratios of tracheid cell-walls. Both pines and oaks showed synchronous growth reductions during dry years. In the wet Huesca site, lower wood δ13C values and a stronger coupling between δ13C and δ18O were observed in D as compared with ND pines. In the dry Teruel site, the minimum wood density of ND pines responded more negatively to spring drought than that of D pines. We argue that multi-proxy assessments that combine several variables have the potential to improve our ability to pinpoint and reconstruct insect outbreaks using tree-ring data.