Bird and mammal sampling strategies: NEON's contribution to the continental-scale ecology of vertebrates

Background/Question/Methods

A key component of the National Ecological Observatory Network (NEON) design is terrestrial organismal sampling, intended to increase our understanding of (1) the regulation of biodiversity and ecosystem function, (2) climatic forcing of phenology and demography, and (3) the ecological and evolutionary aspects of infectious diseases at the continental-scale. This component involves, in part, evaluating the multi-decadal community and phenological dynamics of small mammals and breeding birds throughout sites that can vary in area from 10km² up to 500km².  Standardized sampling at these spatial and temporal scales for these groups is unprecedented, with the exception of the North American Breeding Bird Survey (BBS). Thus, in developing the design, we face significant challenges such as:  How do we standardize sampling across different biomes to get interpretable and comparable data? How do we best concentrate certain sampling to maximize efficiency and use of resources? Here we outline the current design strategies for the small mammal and breeding bird sampling protocols, which involve analyses of both prototype sampling efforts and existing large-scale or long-term data, and present preliminary results of these analyses.

Results/Conclusions

Species accumulation curves were used to evaluate the impacts of sampling frequency on estimates of small mammal and breeding bird richness. The long-term dataset from the Portal LTREB site in Arizona, USA (Ernest et al. 2009), revealed that sampling twice a year captured only 17 of the 21 species documented via monthly sampling from 1977-2010. However, sampling 6, rather than 12, times a year did successfully capture the higher richness. Similarly, density estimates based on only 2 samples per year were significantly different than those based on either 6 or 12 samples annually. For the breeding birds, a season of intense point count sampling at a site in northeastern Colorado revealed that average species richness increased dramatically with sampling, from approximately 17 for one grid (consisting of 16 points, 250m apart) to 29 species for 5 grids, and continued to increase, albeit more gradually, with each additional grid (up to 35 species across 9 grids). Additional analyses for informing sampling effort will be presented, and community input solicited, to help NEON achieve the most effective and efficient designs possible for the lifetime of the observatory.