Landscape drivers and spatial scale predict epiphyte biodiversity across human-modified landscapes

Abstract

ContextHuman-driven landscape processes such as habitat loss and fragmentation act on biodiversity but their effects are mediated by spatial scale. Quantification of the effects of the spatial scale of landscape processes on biodiversity moves beyond a purely ecological concern, towards a pragmatic tool for conservation aiming to designing biodiversity-rich landscapes.ObjectivesWe conducted a multiscale study to quantify the direction of effect (positive, negative, or neutral) and spatial scale (scale of effect, SoE) that three landscape processes—habitat loss (forest cover), fragmentation (number of patches), and edge effects (edge density)—exert independently on epiphyte biodiversity. MethodsWe focused on hyperdiverse vascular epiphytes and orchid communities, sampled across 23 human-modified landscapes in the mountainous region of the eastern Colombian Andes. We consider 12 spatial scales ranging from 100-2400 m radii and use metrics of Hill numbers to quantify biodiversity responses. ResultsHabitat loss and edge effects act at fine spatial scales (SoE= 200 m radius), predicting low species richness and abundance distributions across epiphytes and orchid communities. Fragmentation negatively impacts orchid communities at coarse scales (SoE= 2400 m radius). However, the biodiversity-landscape relationship might be undetected if assessed at the incorrect spatial scales (Forest cover 800-1500 m; Fragmentation ≥ 300 m; and edge density ≤800 radii).ConclusionsWe showed that habitat loss, fragmentation, and edge effects all play a negative role on biodiversity, and their detectability and impact is scale dependant. To design landscapes that are beneficial for mountainous plant biodiversity, it is important to improve forest cover as well as to reduce fragmentation and edge effects.