Methodology:

In February 2007, a field warming experiment was set up in 'the biogully' on Rothera Point. Ten plots were established, five of which were covered with open top chambers (OTCs) to warm mats of Cephaloziella varians. Samples of C. varians were collected from control and OTC plots on 16 February 2007, 7 March 2012, 27 January 2014, 10 February 2014, 28 February 2015 and 21 March 2017 and were stained with aniline blue. Using epifluorescent and light microscopy, plants were examined for fungal structures, viz., percentage stem length colonised (SLC) by dark septate hyphae, hyaline septate hyphae, fine hyphal coils and coarse hyphal coils. When at least 30 rhizoids were found in each sample, the percentage that were colonised by hyphae was also recorded, and the number of rhizoids per mm of stem was also calculated. Random numbers were assigned to each sample so that the author was blinded to the plots from which the plants had been sampled during microscopic analyses.

Mat temperatures were measured every 3 hours between 17 February 2010 and 23 February 2011 by inserting button loggers (SL51 Smart Logger, Status Instruments, Tewkesbury, UK) at depths of c. 10 mm into C. varians mats in four control plots and four OTCs.

Mat moisture concentrations were measured on 11 January, 31 January, 14 February and 28 February 2014 by sampling areas of mat from five control plots and five OTCs, drying them at 80 deg C for 48 h and then reweighing them.

Data collection:

Microscopy: Olympus BX51 compound microscope (Olympus Life Science, Tokyo, Japan) fitted with a UV fluorescence filter cube (U-MWU2, consisting of a BP 330-385 excitation filter, a DM 400 dichromatic mirror and an LP 420 emission filter).

Mat temperatures: SL51 Smart Logger (Status Instruments, Tewkesbury, UK).

Data quality:

Raw data are shown. The data have not been cleaned. Missing data are indicated by NA. Three rows of data are missing from the colonisation data owing to a lack of material (plot 6, sampled on 7 March 2012, and plot 9, sampled on 28 February 2015) and an OTC having become damaged (plot 8 on 21 March 2017). Sixteen values are missing from the "rhizoids_colonised_percentage" column in the colonisation data owing to there being less than 30 rhizoids in each of these samples. One value is missing from the mat moisture concentration data (plot 2, sampled on 11 January 2014) owing to a data processing error.