Multiple processes act simultaneously to shape diversity patterns across the globe. The species composition of a community is likely the outcome of a hierarchical interplay of stochastic and deterministic processes. Processes occurring at a regional scale (e.g., dispersal and temperature) can determine which species can colonise and establish in a location, while those acting at the local scale (e.g., productivity, biotic interactions) further constrain local community composition. Nevertheless, different mechanisms of community assembly in each biological group can influence the responses of organisms to their surrounding environment, as a result of their ability to cope with it or to disperse to other locations.
Accurate quantification of biodiversity is essential to assess the diversity and changes in community and food-web structure across scales (temporal and spatial). Environmental (e)DNA metabarcoding allows for biodiversity assessments through non-invasive, cost-efficient, and rapid surveys. However, the approach struggles to outperform conventional morphological approaches in providing reliable quantitative estimates for surveyed species (e.g., abundance and biomass).
This work advocates for the use of both eDNA- and morphological-based approaches as
complementary methods. Building up on the strengths of these approaches, a framework was developed to pair datasets originated with eDNA metabarcoding and conventional methods, finetuning high-resolution biodiversity assessments, improving taxonomic resolution, and assigning abundances and traits to taxa in the hybrid dataset.
Despite eDNA being used to reconstruct co-occurrence networks, co-occurrences alone are not evidence of species interactions. eDNA datasets would further benefit from background knowledge on species traits and realised interactions, not only for assessments of communities across scales but also in the reconstruction of food webs. In this thesis it is shown how this could be achieved by developing sampling strategies (random or stratified) pairing eDNA- (high taxonomic resolution and sampling replication) and morphological-based (quantification of species abundance and measurement of functional traits) approaches, accounting for scale and/or relevant environmental gradient likely to driving the responses of interest. To further account for challenges and limitations identified in a review done regarding studies reconstructing trophic interactions with either DNA metabarcoding or trait-matching (or both), an eDNA-trait matching framework is presented. In this work it is showcased how pairing eDNA and trait data, obtained through morphological-based approaches, improves our ability to characterise greater numbers of food webs across scales.
Throughout the present work, it is demonstrated that biodiversity assessments with highresolution (temporal, spatial, and trophic coverage) are needed. It is also shown how it can be accomplished by integrating different types of data (e.g., observations, species interactions and traits), obtained with already available methods (e.g., eDNA metabarcoding, morphological, and network approaches), future-proofing and leverage eDNA datasets to a higher, and more functional level. Additionally, to describe diversity patterns and changes in communities and food-webs structures, multi-scale and multi-trophic assessments are essential, conveying their complexity, particularly in response to the ongoing environmental and anthropogenic changes.