The relative influence of ecological assembly processes such as environmental filtering, competition, and dispersal are known to vary across spatial scales, but detecting these drivers’ relative influence has proven difficult. In order for ecologists and land managers to make informed, data-driven conservation decisions across scales, straightforward and flexible sampling schemes are needed to quantify how environmental factors, like terrain, weather, and land-use, affect and shape community composition. Classically, sampling locations are selected at random, across regularly-spaced grids, and transects to assess biodiversity. But such designs often require a detailed understanding of the processes underlying a system before their application, as well as intensive effort to capture diversity across scales. Here we describe a new fractal-based sampling scheme, and test its efficacy to sample an area’s diversity through both simulations and empirical data collected at the Right Hand Fork of the Logan River in the Bear River Mountains of northeastern Utah.
Results/Conclusions
Our simulations show increased statistical power to detect ecological patterns when a fractal sampling design is used. We argue that fractals distribute the number of pairwise comparisons across spatial distance classes and therefore allow us to investigate differences between patterns at different scales. Preliminary analysis of both phylogenetic and functional trait diversity indices suggests ecologically relevant shifts in composition across gradients in elevation and aspect. We found statistically significant increases in the community weighted means of maximum height and seed mass in our data, suggesting that larger species are found on north-facing slopes. Phylogenetic diversity, represented by the standard effect size of the mean nearest taxon distance, also shifted from strongly negative (indicating close-relatives were co-occurring) to strongly positive (indicating distant-relatives were co-occurring) across the south-north aspect gradient. Together, this theoretical and empirical evidence supports the flexibility and efficiency of this method to simultaneously sample environmental and community heterogeneity at fine and broad scales.