NestWatch first-egg dates of 22 North American bird species (1997–2025) were 3 days later in the first year of a nest site, and this caused the apparent extra advance within sites

Question: in NestWatch nest records, does the spatial temperature slope of first-egg date agree with the year-to-year slope at a site, and did dates within sites advance more than local warming predicts? Data: 300,895 first nesting attempts of 22 species (NestWatch Open Dataset v6, USA and Canada, 1997–2025) and NOAA CPC daily 0.5° temperature. Method: for each species we selected a temperature window on within-site fit only, with cross-fitting over random halves of sites. We estimated spatial (cell climatology, year fixed effects), equal-latitude spatial, within-site temporal and anomaly slopes, with two-way clustered (space block x year) joint covariance, and pooled species with random effects. Results: the spatial slope was −2.11 days/°C (95% CI −2.34 to −1.88) and the within-site slope −1.40 (−1.68 to −1.11). The difference was −0.70 (−1.09 to −0.31), and the CI excluded zero under all eight sample definitions (−0.65 to −1.08). With latitude in the spatial model, the difference was −0.18 (−0.70 to 0.34). This independently reproduces SCI 0bcf42b4 with a different design. New result: at constant window temperature, within-site dates advanced by −2.47 days/decade (−3.54 to −1.40). But the first recorded year of a site was 3.08 days (2.33 to 3.84) later than later years. The point estimate was positive in all 22 species. There was no second-year effect (0.00 days), and the first year of an observer was not late (−0.78 days). With a first-year term, the residual trend was +0.15 days/decade (−0.73 to 1.04), and the temperature slope did not change. Thus the apparent faster-than-temperature advance within sites is a site-tenure artifact, and lay dates followed the temporal temperature slope. Limitations: participatory data with changing methods, 0.5° temperature, one calendar window per species, and earliest attempt per site as a proxy for first brood. NestWatch records (CC BY-NC 3.0) are not included; code and aggregate results are.

breeding phenology · lay date · NestWatch · space-for-time substitution · phenotypic plasticity · citizen science bias · nest boxes · climate change · temperature sensitivity · North American birds · reproduction study