Upscaling hypotheses on herbivore damage in plants facing environmental stress: Variation among scales and plant enemies in a relict tree

The few existing analyses of broad-scale variation in plant-herbivore interactions are largely exploratory and not based on hypotheses on the effect of varying environmental conditions on the patterns of herbivore damage. However, there are two established hypotheses that relate herbivore damage and environmental stress mostly tested at the within-population level: the Plant Stress Hypothesis (PSH) and the Plant Vigor Hypothesis (PVH). To examine if support for the PSH holds true at broader scales, we tested it at the micro- (among individuals), meso- (between habitats) and macroscale (between geographical regions, Iberia and Macaronesia). We measured plant stress and leaf damage by different natural enemies (chewers, miners and pathogens) in 12 populations across the range of the relict tree Prunus lusitanica. Results regarding the proportion of leaves attacked by any enemy showed no support for the PSH at either micro or mesoscale, observing instead higher damage in less stressed individuals within populations and in high-quality habitats. However, the latter was true exclusively within the Macaronesian region, highlighting the important effect of the type of abiotic stress on plant susceptibility to herbivory. At the macroscale, alternatively, results showed support for the PSH: we observed a higher proportion of leaves attacked in Macaronesia than in Iberia, where plants unexpectedly exhibited greater stress symptoms. Among-scale inconsistencies may respond to the effect of distinct climatic factors governing both plant and herbivore dynamics at large scales, particularly in species with distributional ranges across contrasting climatic regions such as P. lusitanica. Finally, different plant enemies showed different patterns of damage, corroborating the need to study them separately to achieve consistent generalizations of these hypotheses.