LAUSR

Simple Summary Assessing the biodiversity of a site is challenging, as a comprehensive survey of all plant and animal species is usually intractable due to limited resources. There is great interest in finding groups of organisms that may serve as proxies for the total biodiversity of a site. Numerous taxa have been proposed and utilized for rapid biodiversity assessments that allow for such a survey in a matter of weeks or less. Herein, we test the idea that nocturnal moths have many of the characteristics that make them ideal for such surveys. We demonstrate that even in a few hours of sampling during single night surveys, moth communities are predictive of regional forest types at sampling sites in New Jersey. An ordination method, known as non-metric multidimensional scaling (NMDS) analyses, revealed that moth communities differ significantly across the four forest types that we assessed. We used Analysis of Similarity (ANOSIM) R tests to quantify the degree of differentiation among moth communities, and found that the moth families Tortricidae and Geometridae predict forest communities nearly as well as the total moth diversity. Uncommon species were better predictors than common species. Host plant generalists were better predictors than specialists, which was a surprising find. Abstract Biodiversity conservation decisions are typically based on limited data and resources. For this reason, there is great interest in surveying taxa that may allow for a rapid assessment of the biodiversity at a site. Numerous taxa have been proposed and utilized for rapid assessments that allow for such a survey in a matter of weeks or less. Herein, we test the idea that nocturnal moths have many of the characteristics that make them ideal for such surveys, such as relative ease of identification, strong ecological association with specific plant species and habitats, high alpha diversity, extended seasonal activity, and ease of trapping. We demonstrate that even in a few hours of sampling during single night surveys, moth communities are predictive of regional forest types at sampling sites in New Jersey. We sampled moths in five different forest habitats in New Jersey, USA: Pine Barrens, Upland Deciduous Forest, Palustrine Deciduous Forest, Maritime Forest, and Ruderal/Disturbed Forests, at four sites per forest type. Non-metric multidimensional scaling (NMDS) analyses revealed that moth communities differ significantly across these four forest types (p < 0.01). We used Analysis of Similarity (ANOSIM) R tests to quantify the degree of differentiation among moth communities, and found that Tortricidae (R = 0.657) and Geometridae (R = 0.637) predict forest communities nearly as well as the total moth diversity (R = 0.668). Uncommon species (R = 0.665) were better predictors than common species (R = 0.500). Host plant generalists (R = 0.654) were better predictors than specialists (0.538), which was a surprising find.