Data from the National Ecological Observatory Network (NEON) is intended to be the foundation for a new generation of models that predict biodiversity and ecosystem function in a changing world. Conclusions drawn from any ecological dataset must be understood in the context of its spatial and temporal extent. Without such context, it is difficult to generalize results into the past or future, or to draw conclusions about other locations and systems. The ARGON project (Augmenting Research Grounded On NEON) focuses on synthesizing ecological trait and community data relevant to contextualize NEON data. In this presentation, we address two questions: (1) To what extent are NEON sites representative of surrounding sites – their ecological context – as described within the ARGON dataset? (2) How are the species of NEON sites distributed within the tree of life – their evolutionary context – and how does this affect the functional composition of NEON sites? We also describe how ARGON differs from most data compilation attempts by not to compiling data, but rather programming code to compile data. This has the advantage of ensuring that all data collectors receive full credit for their efforts in every analysis making use of their data.
Results/Conclusions
ARGON database in terms of taxonomic, functional, and phylogenetic diversity. Despite the opportunistic sampling of sites within ARGON, we present attempts to measure the spatial decay of similarity in the data. We argue that NEON is capturing the regional ecological context in which each site is located. (2) As with most ecological communities, the plant and bird species of the NEON sites are not randomly sampled from the tree of life. We do find, however, that the phylogenetic signal of the functional traits of the species within the sites is approximately the same as across the ARGON dataset as a whole. We speculate that, while NEON may not perfectly capture its evolutionary context, this invariance of phylogenetic signal suggests ecological processes can readily be studied in NEON data. ARGON is an incomplete dataset and so, at best, these results remain preliminary. The underlying R software is reasonably complete, however, and so we conclude by providing a basic set of instructions as to how scientists can become members of the ARGON project by contributing data.