Abstract:
In tropical regions insect herbivores are considered one of the key players in the ecosystem function by effecting the nutrient cycle as well as the carbon sequestration of these systems. However, their feeding is strongly influenced by abiotic factors, like temperature, leading to
a lower performance and abundance of herbivores under lower temperature, as well as biotic factors, like plant defence mechanisms (e.g. tougher leaves or a higher concentration of secondary metabolites). But plants are facing a constant trade-off between growth and defence, which is stronger under harsher environmental conditions. The resulting differences in defence further lead to a big variability in the rate of herbivory between the plants. Although several studies investigated the interaction between herbivory, biotic and abiotic factors,
most of them concentrate on single species and systems. However, since herbivory varies between species and has high impacts on the whole ecosystem by influencing plants viability and the carbon sequestration, it is crucial to investigate the interaction on community level.
We therefore assessed herbivory across an elevational gradient of 1000 m to 3000 m a.s.l. on a community level, i.e. by measuring 313 representative tree species. To determine the biotic interaction we measured leaf traits using both conventional and remote sensing methods, to look for the applicability of remote sensing for extrapolation of community patterns. Our study shows, that herbivory decreased with increasing elevation. Furthermore, sulphur showed a positive association with herbivory and SLA, nitrogen and tannin showed a negative
association. The influence of carotenoid on herbivory changed from a positive to a negative one with elevation and anthocyanin showed the opposite pattern. Phylogeny thereby had the same effect on both conventional and remote sensing measured leaf traits. Our results showed,
that the influence of the abiotic factors especially at higher elevations were strong, leading to the conclusion, that community level herbivory is more effected by the abiotic interaction along the elevational gradient than the biotic interaction. Moreover, the models of conventionally and remote sensing determined leaf traits had a strong congruence regarding the influence of abiotic and biotic factors on herbivory, although they comprise different leaf traits. Consequently we can state, that remote sensing can be used to predict herbivory on a community level in a tropical mountain forest.