Window installations were monitored and re-secured monthly until sample collection in June 2019. At that time, one UV-filtering window had been lost to weather and that pair was excluded from downstream analyses. Although it is not possible to determine when the mosses were last naturally hydrated in their habitat, it had rained 1.6 cm 1 month prior to collection, at which time samples were not observed to be hydrated beneath the windows. Cushions were collected dry using 9-cm diameter culture dishes from each of the remaining 19 pairs along with an additional third, unmanipulated site reference (within 1 m of the window pair). This latter “site reference” was used to test for effects of a window treatment, per se. Specimens were stored dry (Fig. 1A) and in the dark until analyses.