Six herbarium samples did not establish a common temperature slope difference between anther groups in 1941-2017

We tested a heat-protection explanation for flower UV pigmentation. We used 1,224 complete herbarium specimens from a public archive. Six tests covered Australia, Eurasia and North America in 1941-1985 and 1986-2017. Models compared temperature slopes for protected and exposed anthers. They controlled species-specific collection months and time trends, ozone and location. The protected-minus-exposed slopes were -1.71 and +5.39 percentage points per C in Australia, -4.55 and +10.96 in Eurasia, and -3.53 and -2.65 in North America. All six corrected intervals included zero. Equal-species weights and a transformed response kept all six signs. Removing month controls changed the sign in later Eurasia. Two protected-anther taxa in each northern region gave weak support for group inference. The results do not establish a common heat-protection response or its absence. Nonrandom collection, few species, source exposure errors and one archive limit the test. The package includes code, a specimen panel, source hashes and numerical checks. We tested no surface air pollutant or weather extreme.

floral pigmentation · temperature · anther architecture · herbarium · observation bias · regional replication · null result