Spring wetness predicted frog and toad calling occupancy on 1,036 NAAMP routes in the eastern and central USA (2001–2015), but wetness in the three previous summers did not

Question: Does drought in past breeding seasons decrease later adult calling of frogs and toads, as recruitment failure predicts? Approach: We used the unified-protocol data (2001–2015) of the North American Amphibian Monitoring Program (NAAMP), a roadside calling survey in 25 US states. For 18 taxa, the response was stop occupancy: the fraction of listening stops on a route with the taxon heard in at least one of three annual runs. The data had 4,858 route-years on 1,036 routes. Predictors were the Palmer Drought Severity Index (PDSI) of the NOAA climate division for the current spring (March–May) and for the April–August period of each of the three previous years. Models had fixed effects for route × observer and for year. The pre-specified primary statistic was the legacy effect L, the sum of the three past-summer coefficients. The null distribution came from 92 time-shifted PDSI series of the same divisions (10–100 years earlier, or 10 years later). Results: The pooled L was −0.02 percentage points (pp) of occupancy per PDSI unit (95% null-spread interval −0.56 to +0.52; one-sided p = 0.52). No taxon had a clear legacy effect. Current-spring wetness increased occupancy by +0.53 pp per PDSI unit (more extreme than all 92 null values; 14 of 18 taxa positive). The lag with the largest coefficient was not related to age at maturity (Spearman rho = −0.04, 14 taxa). Seven sensitivity analyses gave the same pattern: PHDI, random routes only, a calling-index response, route-observer trends without year effects, single-observer route-years, survey air temperature, and a binomial GLM. Limits: Stop occupancy is a coarse index of abundance. Climate divisions cannot resolve drought at single ponds. The panels are short (median 4 years per route). The data do not exclude a legacy smaller than about 0.5 pp per PDSI unit. The analysis plan, code, derived tables and null distributions are included.

amphibians · anurans · NAAMP · calling survey · drought · Palmer Drought Severity Index · lagged effects · occupancy · climate variability · time-shift null · null result · North America