Czech breeding bird community temperature index rose at 29% of full tracking in 1982–2024, but not more than with shuffled species values and not more where sites warmed more

Question: did Czech breeding bird communities change towards warm-climate species in 1982–2024, was the change specific to species thermal preference, and was it larger where sites warmed more? We repeated the analysis on the UK Breeding Bird Survey (1994–2024) as a replication. Data: Czech JPSP transects (351 sites, 3,398 site-years, 154 species) and UK BBS squares (7,070 squares, 95,212 square-years, 113 species); May–July temperature from CRU TS 4.10, checked with E-OBS 32.0e; species temperature indices (STI) from GBIF May–July records in European 0.5-degree cells, independent of the survey counts. Method: log-weighted community temperature index (CTI), trends with site x observer fixed effects, spatial-block bootstrap, joint site and species bootstrap, 1,000 shuffles of STI among species, exact species contributions, and regression of site CTI trends on site warming trends. Results: in Czechia the CTI increased by 0.017 per decade (0.011 to 0.021), which is 29% (19% to 41%) of full tracking of the 0.46 °C per decade warming. With species resampling the interval included zero (tracking ratio −0.09 to 0.69). Shuffled STI gave an equal or larger trend in 12% of cases. Site CTI trends did not increase with site warming (slope −0.024 with CRU and −0.020 with E-OBS; full tracking predicts 0.126). Year-to-year CTI changes did not follow temperature changes. Decreases of cold-climate species (Goldcrest, Tree Pipit) added most to the increase; decreases of warm species (Turtle Dove, Serin) removed much of it. In the UK the CTI increased by 0.010 per decade (11% of full tracking; shuffle p = 0.33), and European Goldfinch supplied 73% of the net increase. This agrees with an earlier SCI analysis of the UK data that used different choices. Results were stable in 7–8 variants per scheme (weights, land birds, STI thresholds and season, survey-based STI, period, block size). Conclusion: both communities moved slowly towards warm species, much slower than warming, not more where warming was larger, and not more than with random species values. Limitations: observational design; GBIF sampling bias in the STI; the shuffle test detects only tracking above about 60–66% of full tracking; site warming trends differ between temperature products; 88 Czech sites qualify for the site test.

community temperature index · climatic debt · breeding birds · climate change · Czech Republic · United Kingdom · Breeding Bird Survey · species temperature index · permutation test · GBIF · CRU TS · E-OBS · replication