In GBIF records from 11 European countries, 1970–2025, autumn mushroom dates moved later by 1.5 days per decade, but not more than dates of perennial fungi, lichens and recording visits

Kauserud et al. (2012) reported that in 1970-2007 the autumn fruiting of European mushrooms moved later by 1.2-3.4 days per decade. Critics said that a change in recorder behaviour can cause the same pattern. We tested this with 34.9 million GBIF fungus records (snapshot 2026-10-01) from 11 countries, in the window 1 July to 31 December. We compared the species-centred record dates of 2,381 autumn-fruiting species with three controls that do not have an autumn fruiting peak: 22 perennial wood-decay fungi, 1,528 lichens, and the dates of all fungus-recording visits. Uncertainty comes from a bootstrap of 50 km blocks and years, and we pooled countries with random effects. The raw record dates of autumn fungi moved later by 1.41 days per decade (95% CI 0.71-2.11) in 1970-2007, in agreement with the earlier study. But in the same period, perennial fungi moved later by 1.80, lichens by 2.89 and visits by 1.35 days per decade. For 1970-2025 the values were 1.47, 2.92, 3.41 and 2.36. The autumn-minus-perennial contrast was -0.43 (-1.60 to 0.75) in 1970-2007 and -1.40 (-2.25 to -0.56) in 1970-2025. The autumn-minus-lichen contrast was -1.05 and -1.88. With only the national databases that Kauserud et al. used, the 1970-2007 autumn-minus-perennial contrast was -0.13 +/- 1.13 days per decade. The end of the recorded season (90th percentile) moved later by 2.1 days per decade for autumn fungi, and by 8.6-8.9 for the controls. A per-visit correction did not remove the trends from the controls. The contrast can detect a phenology signal: from year to year it responded to summer rain (-4.9 days per log unit). Weather trends predicted a contrast change of +0.2 days per decade, but the observed change was -1.4. Thus, these records show no delay in autumn fruiting beyond the change in the recording calendar. Main limitation: the controls are imperfect, and recorders partly follow fruiting (slope about 0.6 for perennial fungi). A recorder change that moves the controls more than the autumn fungi can hide a real delay, so the result does not exclude a phenological change.

fungal phenology · mushroom fruiting · GBIF · recorder effort · observation bias · citizen science · negative control · climate change · Europe · lichens · polypores · opportunistic records