Determinants of taxonomic, functional and phylogenetic diversity that explain the distribution of woody plants in tropical Andean montane forests along altitudinal gradients

Abstract

Understanding the evolutionary and/or ecological mechanisms (including biotic, abiotic and stochastic factors) determining community assembly –i.e., the composition, structure and distribution of biological communities on an ecosystem– remains a central theme in ecology. Beyond the classical taxonomic scope, functional and phylogenetic approaches have certainly improved our knowledge about the mechanisms generating biodiversity patterns within the last decades. However, many discrepancies still exist about which is the relative extent of those mechanisms and under which circumstances they exert their influence. In this sense, contradictory conclusions about what processes drive community assembly seem to be caused by different spatial scale considerations, hampering and embroiling the understanding of this topic. To incorporate this perspective, a hierarchical community assembly model has been proposed, which postulates that different community assembly mechanisms operate sequentially, at increasingly finer spatial scales, to render the observed biodiversity distribution patterns. This model comprises from evolutionary processes such as historical patterns of speciation, extinction or migration, that mostly act at large spatial scales, to abiotic and biotic processes, like environmental filtering or competitive exclusion, mainly performing at smaller scales. We here propose a multi- spatial scale approach aiming to validate some of the hierarchical model postulates and to clarify which assembly mechanisms drive species occurrence in biological communities. To achieve this goal, we use the tropical montane forests (TMFs) in the Andes as model system. Mountain ranges provide an excellent natural laboratory to investigate how ecological and evolutionary mechanisms influence community assembly along their slopes. Besides, Andean TMFs foster an enormous biodiversity and their large geographical extent enables the replication of large altitudinal gradients over different latitudes, in areas with different climates, topographies and biogeographic and evolutionary histories. The replication of gradients is particularly important for the sake of the generalisation of conclusions beyond the singular contingencies of each studied site. Hence, in this doctoral thesis I study woody plant biodiversity patterns in Andean TMFs along altitudinal gradients to accomplish the following goals: 1) To investigate taxonomic, functional and phylogenetic diversity patterns along Andean TMFs altitudinal gradients and explore some of the mechanisms responsible for community assembly (Chapter I). This is a fundamental step for a later sound exploration of such mechanisms. 2) To test the postulate that ecological mechanisms responsible for community assembly –namely environmental filtering and competitive exclusion– operate at different spatial scales –small and large, respectively– (Chapter II). 3) To address how different environmental filters affect spatial variation of a specific functional trait –namely compound leaves– by quantifying changes in this trait patterns along altitudinal gradients and linking them to underlying potential ecological drivers –drought, light availability and wind– (Chapter III). To accomplish these goals, we considered 114 (Chapter I), 60 (Chapter II) and 390 plots (Chapter III) scattered along 4, 2 and 11 altitudinal gradients in Andean TMFs, respectively. Plots’ bioclimatic features were retrieved from world climatic databases. In all cases, all occurring woody individuals were surveyed. Individuals were identified taxonomically and the resulting taxa characterized functionally. In addition, we built a phylogenetic tree (Chapters I and III). That information was summarized using different metrics and analysed using statistical techniques, such as null models or generalized linear mixed effects models (GLMMs). We found in Chapter I that taxonomic, functional and phylogenetic diversity monotonically decrease along altitude. This decrease is positively correlated with a decrease in temperature and gives clues about the importance of ecological processes shaping community assembly. The functional and phylogenetic diversity decrease suggests that the increasingly colder temperatures, as we move upslope, operate as an environmental filter (abiotic mechanism) that selects only for those species capable to adapt and survive. Those species, which historically migrated towards highlands and succeeded belong to few closely related clades originated in tropical warm lowlands (evolutive mechanism), in agreement with the Tropical Niche Conservatism hypothesis. The importance of environmental filtering for community assembly and the spatial extent of its effect was evidenced in Chapter II by showing that a functional traits clustering pattern only appears at large spatial scales (between plots at different altitudinal belts) that embrace stark environmental differences. Moreover, the effect of environmental filtering was confirmed by demonstrating that its fingerprint goes beyond the observed pattern, since along the altitudinal gradient plant species exhibit strong climatic preferences and their abundances dramatically shift. However, competitive exclusion was not found to act in this ecosystem, nor even at small spatial scales (within subplots) where plant species effectively compete for limited resources, at least by mediating a functional divergence among co-occurring species. Finally, in Chapter III, we reported a decrease on species bearing compound leaves –richness and abundance– from lower to upper altitudes, consistent across the Andean mountain range, but we found no evidence that this trend was caused by the environmental conditions hypothesized to favor this type of leaf in lower TMFs. Finally, we explored whether compound leaves altitudinal pattern could be the potential result of TMFs evolutionary history, but we found either support for this potential explanation. In summary, by addressing functional and phylogenetic diversity in addition to the taxonomic one, we can obtain a much more complete understanding of the mechanisms responsible of community assembly that ultimately lead to the observed biodiversity patterns. Phylogenetic diversity offers us a window to the past that permits to infer the evolutionary mechanisms and events that explain the historical plant colonization of the Andean slopes. Our results suggest that few closely related lineages migrated from tropical warm lowlands towards the highlands. Functional diversity reveals how the species cope with their surrounding environment and shows that environmental filter is the overriding ecological mechanism determining the differential distribution of species along altitude. However, environmental filters seem not to shape the altitudinal trends of compound leaves, and thus, considering also that the evidences in this sense found in other studies are weak, we cannot advocate for the usage of compound leaf as an easily measurable, qualitative functional trait. The fact that our results are consistent among different study sites highlights the robustness of our conclusions and permits to generalize them in the Andean TMFs and beyond, in other tropical mountain ranges.