Search for top-down and bottom-up drivers of latitudinal trends in insect herbivory in oak trees in Europe
Aim: The strength of species interactions is traditionally expected to increase toward the Equator. However, recent studies have reported opposite or inconsistent latitudinal trends in the bottom-up (plant quality) and top-down (natural enemies) forces driving herbivory. In addition, these forces have rarely been studied together thus limiting previous attempts to understand the effect of large-scale climatic gradients on herbivory.
Location: Europe.
Time period: 2018-2019. Major taxa studied: Quercus robur.
Methods: We simultaneously tested for latitudinal variation in plant-herbivore-natu-ral enemy interactions. We further investigated the underlying climatic factors associated with variation in herbivory, leaf chemistry and attack rates in Quercus robur across its complete latitudinal range in Europe. We quantified insect leaf damage and the incidence of specialist herbivores as well as leaf chemistry and bird attack rates on dummy caterpillars on 261 oak trees.
Results: Climatic factors rather than latitude per se were the best predictors of the large-scale (geographical) variation in the incidence of gall-inducers and leaf-miners as well as in leaf nutritional content. However, leaf damage, plant chemical defences (leaf phenolics) and bird attack rates were not influenced by climatic factors or latitude. The incidence of leaf-miners increased with increasing concentrations of hydrolysable tannins, whereas the incidence of gall-inducers increased with increasing leaf soluble sugar concentration and decreased with increasing leaf C:N ratios and lignins. However, leaf traits and bird attack rates did not vary with leaf damage.
Main conclusions: These findings help to refine our understanding of the bottom-up and top-down mechanisms driving geographical variation in plant-herbivore interactions, and indicate the need for further examination of the drivers of herbivory on trees.
Valdés‐Correcher, E., Moreira, X., Augusto, L., Barbaro, L., Bouget, C., Bouriaud, O., ... & Castagneyrol, B. (2021). Search for top‐down and bottom‐up drivers of latitudinal trends in insect herbivory in oak trees in Europe. Global Ecology and Biogeography, 30(3), 651-665.
The distribution range of Quercus robur L. (shaded in yellow) and locations of trees sampled by professional scientists (orange symbols) and schoolchildren (blue symbols) in 2018 (circles and 2019 (squares). Additional maps showing oak trees used for estimating leaf herbivory, attack rates on dummy caterpillars and trait analyses are provided in Supporting Information Figure S1.1
Effects of latitude, mean spring temperature and mean spring precipitation on bird attack rates (a, b, c), gall-inducer incidence (d, e, f), leaf-miner incidence (g, h, i), leaf damage (j, k, l) and total defences (m, n, o). Dots represent raw data averaged at the tree level. Solid and dashed lines respectively represent model predictions and corresponding standard error calculated after other significant variables (see Supporting Information Table S2.1) were set to their mean value. Only statistically significant relationships are shown. Regression line equations are as follows: (e) y = −3.32 + 0.44 · x; (f) y = −3.32 + 0.32 · x − 0.51 · x2; (h) y = −1.98 + 0.37 · x - 0.36 · x2; (i) y = −1.98 - 0.44· x
Effect of mean spring precipitation and latitude on soluble sugar (a) and C:N ratio (b) on leaves, respectively. Dots represent raw data averaged at the tree level. Solid and dashed lines respectively represent model predictions and corresponding standard error for temperature and latitude calculated after other significant variables (see Supporting Information Table S2.2) were set to their mean value. Only significant relationships are shown. Regression line equations are as follows: (a) y = 1.53 - 0.10· x; (b) y = 17.9 - 0.86 · x + 0.70 · x2
Effects of latitude, mean spring temperature, mean spring precipitation and leaf chemistry on gall-inducer (left) and leaf-miner (right) incidences. Circles and error bars represent standardized parameter estimates and corresponding 95% confidence intervals (CI), respectively. The vertical dashed line centred on zero represents the null hypothesis. Black and grey circles represent significant and non-significant effect sizes, respectively