Leaf damage by insect herbivores can severely affect the growth and fitness of plants and is considered to play an important role in the maintenance of high tree diversity in (sub)tropical forests. The trees of the main experimental plantations of BEF China are strongly affected by herbivory and understanding patterns in plant growth, trait expression and other ecosystem processes thus requires incorporation of potential herbivory effects and knowledge of how herbivory may change across the tree diversity gradient. This dataset quantifies leaf damage by insect herbivores for the trees of the main experimental study plots. - Only living trees are included in the dataset. Dead trees, trees with no leaves and missing trees were deleted.

Herbivory was measured on those trees which are also analysed by subproject (SP) 2, i.e., in the enlarged core area of each plot. In monocultures (16 large 4-mu plots and 34 small 1-mu plots per site) and two species-mixtures (8 large and 36 small plots per site), all trees in the enlarged core area of 6 x 6 individuals (= 36 trees) were measured. In species-mixtures with 4 species (4 large and 24 small plots per site), the core area to be measured comprised 9 x 9 individuals (= 81 trees) and in 8, 16 and 24 species-mixtures (2, 1 and 1 large and 18, 15 and 1 small plots, respectively, per site) the core area was 12 x 12 individuals (= 144 trees). On each large 4-mu plot, measurements were conducted in the core area of one of the four 1-mu subplots (i.e. in the VIP subplots). The degree of foliar damage by insects, defined as the combined removal of photosynthetic tissue by leaf-chewing, mining and galling (and, if visible, sucking) insects, was estimated using percentage classes of herbivore damage. Damage classes were defined beforehand and appropriateness of the estimates was checked by analysing samples of randomly collected leaves with optical scanning methods. Categories were then expressed as mean percentage of leaf tissue removal based on the scanned leaves. Herbivory was assessed for 21 fresh leaves (i.e. expanded in the current gowing season) on three randomly selected branches (7 leaves x 3 branches) of each tree individual to estimate overall damage levels for each individual. As additional data, the total number of leaves was estimated and the main damage type was recorded.

Main Experiment Site A 29°08'-29°11'N 117°90'-117°93'E

Summer 2011 (May-June)