Ecologists have widely observed a transition in forest structure and tree community composition at the elevation of cloud formation on tropical mountains, often referred to as the tropical montane forest (TMF) syndrome. However, difficulty in systematically quantifying variation in fog and low cloud moisture inputs among sites has prevented ecologists from directly assessing the role fog and low cloud moisture inputs likely play in the formation of these ecosystems. Due to advancements in remote sensing and modeling efforts, global hydrological models now provide reliable estimates of fog inputs at the one-hectare scale. Our objective was to evaluate if model-derived fog metrics predict differences among forests sites in species composition and forest structure aside from what can be predicted by elevation alone. We compiled a database of forest attributes (basal area, stem density, and tree diversity) and woody plant species inventories from over 25 tropical mountain ranges provided by contributors to the CloudNET research network. For each site, we used the WaterWorld global hydrological model to extract estimates of mean annual air temperature, mean annual precipitation, and percent rainfall from fog inputs for each site.
Results/Conclusions
We found substantial variation among sites in fog and total rainfall inputs that could not be explained by temperature or elevation. In multiple regression models evaluating the influence of percent fog, rainfall, and temperature on forest attributes, stem density was significantly related to temperature, basal area was significantly related to fog inputs, and woody plant diversity significantly declined with both decreasing temperature and increasing fog inputs, indicating that changes in fog and temperature over elevation gradients independently contribute to the TMF syndrome. Multivariate analysis of tree genera distributions indicated that while elevation and temperature explained the majority of variation among abiotic variables in the presence of montane forest species, fog inputs loaded strongly on the second most important axis. Based on this analysis, we present a list of “fog-associated” genera that that may be particularly sensitive to changes in cloud climatology on tropical mountains. This study is among the first to quantitatively demonstrate the importance of fog inputs in shaping montane forest communities across worldwide sites. We recommend that fog niches and inputs should be considered as a key variable in modeling tree species distributions and ecosystems function of tropical montane forests under future climate scenarios.