Ecological Indicators (Dec 2021)
Fine-grained topographic diversity data improve site prioritization outcomes for bees
Abstract
Biodiversity hotspots are often found in areas of high environmental diversity. This may be explained by an abundance of varying niche space promoting coexistence, buffering effects allowing for persistence during periods of climate change, and by a diversity of selective pressures driving speciation. Thus, environmentally diverse areas are attractive targets for conservation. Few aspects of the environment are unchanging, however, and strategies targeting comparatively durable surrogates for biodiversity are a focus of applied conservation research. For example, topographic diversity (topodiversity), or the variation of landforms present in an area, has proven useful for site prioritization among some taxa. The suitability of this approach for conservation of ecologically important taxa, such as bee pollinators, has not been studied exhaustively, however. Furthermore, the importance of topographic features smaller than 30 m is generally unknown, due to the limited availability of fine-grained elevation data, though could prove relevant for this group and others. We gathered sub-meter elevation data with structure-from-motion photogrammetry and surveyed for bees and plants in the Sapphire Mountains of western Montana, USA to explore if fine-grained topodiversity data are useful for site prioritization. We calculated topodiversity metrics for a range of spatial grain-sizes (0.25–250 m) and neighborhoods (1–500 m) and conducted sensitivity analyses studying correlations with biodiversity patterns. We prioritized sites according to rarity-weighted richness (RWR), which indicates complementarity, or the degree to which a site can contribute unique species to a preserve network. Bee RWR was most strongly correlated with 2 m grain topodiversity data (r 0.73), while plant RWR was most strongly correlated with 235 m data (r 0.44). Though 2 m data were optimal in bees, larger resolutions (35 m) also were effective for site prioritization and use of these meso-grained surrogates only increased site prioritization error by one rank position. We explored multiple regression models predicting RWR with topodiversity metrics of various grains and neighborhood but found no improvement over linear models incorporating only the single most correlated grain-size and neighborhood metric. The best model for plants performed relatively poorly for the task of bee site prioritization. Modeling complementarity with pooled biodiversity data aggregated from both groups slightly improved site rankings for bees and slightly decreased ranking performance in plants. While the costs of fine-grained elevational remote sensing technologies may be high at present, national efforts to gather these data are on the horizon and may facilitate conservation of bees and other groups at the local scale.
