NAAMP visits with six or more cars had 0.10 fewer call groups after site and survey control in 2001–2015

We tested whether passing cars were linked to shorter frog and toad call lists. We used open USGS North American Amphibian Monitoring Program data. The sample had 126,114 visits at 8,750 sites on 877 routes in 23 US states from 2001 to 2015. We joined overlapping species labels into call groups. A linear model held site and survey window effects and each route survey fixed. It also used air temperature terms. Visits with six or more cars had 0.0975 fewer recorded groups than visits with no cars. The nominal 95% state cluster interval was minus 0.1327 to minus 0.0623 groups. The raw mean gap was minus 0.5430 groups. All planned checks kept a negative association. A fresh offline run matched all 15 result and figure files. The result supports the use of visit conditions in roadside survey analysis. It does not estimate species loss or true detection probability. Noise can mask calls or change calling. Local change and recording error can also affect the result. Missing car records limit the sample. The package has unchanged source files, code, provenance, figures, results and reproduction steps.

amphibians · NAAMP · traffic · calling surveys · recorded richness · fixed effects · open data · survey conditions