Fig. 1. Study sites and catchments in the European Alps with indicated glacier cover (right; CORINE land coverage 2018).

At each site, river water temperature was recorded continuously (hourly measurements) with digital loggers (Onset Tidbit2, MA, U.S.A.) and water was filtered through pre-combusted Whatman^TM GF/F filters (0.7 um) to quantify sediment concentrations (3x 0.5 L). Water discharge was estimated using depth/velocity transects and the Ott discharge probe (Ott MF Pro, Kempten, Germany), and water samples were taken to quantify nitrate (NH₃-N), dissolved nitrogen, and dissolved and total phosphorous concentrations. The concentration of suspended sediment (mg L^-1) was assessed by combusting filters at 450°C for 4 hours. We modelled the sediment load (mg s^-1) by combining discharge (L s^-1) and sediment concentrations (mg L^-1). Each catchment was delineated using QGIS (QGIS Development Team, 2009) using highly resolved digital elevation data retrieved from the Austrian data repository (www.data.gv.at). The ratios of glacial land-cover were calculated based on the freely available CORINE Land Cover data (https://land.copernicus.eu/pan-european/corine-land-cover) from 2018.

We used the BenthoTorch² (bbe Moldaenke, Germany) to differentiate among the main periphyton groups that estimates the composition of the following larger periphyton groups based on their in vivofluorescence excitation spectra: cyanobacteria (pigment-group A), diatoms and chrysophytes (pigment-group B), and green algae (pigment-group III). The fast non-invasive measurements allowed spatially highly resolved assessments of total chlorophyll aconcentrations and relative contributions of the different periphyton groups across a river. We analyzed 10-15 cobble surfaces (diameter >10 cm) in each of the 3 transects across each study stream on three occasions during summer 2022 (July, August, September), resulting in up to 45 sampled cobble surfaces per study river and month.

Periphyton was sampled over the same three transects across each study reach with up to 7 sampled cobble surfaces per transect. All scraped surfaces per transect were pooled and divided into 50 mL fractions (3): two were transferred onto GF/C filters (1.2 µm; Whatman^TM) and one was kept frozen until fatty acid analyses. Ash-free dry mass (AFDM, expressed in mg cm^-2) was quantified by drying filters for 24 hours at 60°C followed by ashing at 450°C for 4 hours (loss on ignition (Heiri, Lotter & Lemcke, 2001)). The chlorophyll a content (µg cm^-2) was quantified using a U-3900H spectrophotometer (Hitachi, Tokyo, Japan) following the extraction protocol as described elsewhere (Lorenzen, 1967).

Lipids and their FA were analyzed as described by (Guo et al. , 2015). Briefly, total lipids from freeze-dried (i.e., all lipids and FA were reported as dry weight; DW) periphyton retained on filters (~5 mg) were dissolved in ice-cold chloroform (2 mL) and stored under N₂ atmosphere over night at −80°C to improve lipid extraction efficiency. Samples were then further extracted in chloroform-methanol (2:1) and NaCl (0.8 mL; salt wash), vortexed and sonicated, and subsequently analyzed gravimetrically in pre-weighed tin capsules (total lipid content determination). Fatty acids from total lipid extracts were derivatized to fatty acid methyl esters (FAME) in a H₂SO₄ methanol solution for 16 hours at 50°C. All FAME were stored at -80°C until being separated using gas chromatography (THERMO^TM Trace GC) and detected using flame ionization detection (FID). FAME were separated by a Supelco™ SP-2560 column (100 m, 25 mm i.d., 0.2 µm film thickness), identified by comparison to the retention times of known standards (37-component FAME Mix, Supelco 47885-U; Bacterial Acid Methyl Ester Mix, Supelco 47080-U). The FAME concentrations were quantified using calibration curves based on known standard concentrations. All FAME analyses were replicated within the study design (e.g., 3 periphyton samples per sampling at each transect).

Habitat conditions crucial for primary producers in alpine rivers (water temperature indices, sediment concentration and sediment load, concentrations of nutrients such as nitrate and dissolved phosphorous) were linked to glacier cover in the catchment (GCC, %) using general linear models (GLMs) or generalized additive models (GAMs), depending on the type of the relationship. The percentage of explained deviance (R²) was used to evaluate model performance. As different water temperature indices we calculated the average of daily means, daily ranges, and daily maximum and minimum for two weeks after sampling (July 2022). For comparisons of sediment and nutrient concentrations, we also considered data from the previous year, 2021. This extended dataset allowed us to assess the variability for these parameters.

The relationships between glacier cover (and associated habitat conditions) and the biomass of periphyton groups were tested using GLMs/GAMs. Data manipulation and testing was done using R v.4.2.2 (R Core Team, 2021), GAMs were built using the mgcv package (Wood, 2011), graphics were produced using ggplot2 (Wickham, 2016, 2) and ggtern (Hamilton & Ferry, 2018). Biotic data were assessed in the summer months (July, August, and September) in the year 2022 only.