Data and code for manuscript titled "Contrast in mycorrhizal associations leads to divergent rhizosphere metabolomes and plant-soil feedback among grassland species"

Dataset to test how variation in mycorrhizal associations and associated rhizosphere metabolomes among co-existing temperate grassland species leads to species-specific plant-soil feedback. Ten common temperate grassland species were examined, with seeds and soil collected from two sites: Briza media L. (graminoid), Carex flacca L. (graminoid),Leontodon hispidus L. (forb), Prunella vulgaris L. (forb) and Succisa pratensis Moench (forb) from Laelatu wooded meadow (58°35’06”N 23°34’09”E), and Carlina vulgaris L. (forb), Festuca rubra L. (graminoid), Galium verum L. (forb), Pimpinella saxifraga L. (forb) and Silene vulgaris (Moench) Garcke (forb) from Uisu alvar grassland (58°38’31”N 23°30’51”E). The experiment consisted of two phases.
In the conditioning phase, sand and soil was sterilised using gamma irradiation (25 Gy) and a sterile background soil mixture was prepared by mixing 1 part of soil with 2 parts of sand. Pots were filled with 1.4 kg of sterilized soil mixture. To create soil treatments with and without AM fungal presence, 300g of sterilized or live soil inoculum (prepared from soil originating from same site as the seeds) were added to pots. To homogenise other microbiota across all pots, a liquid microbial wash was prepared by mixing 5 litres of water with 1 kg of live, which was filtered through a sieve with 32 µm aperture to exclude AM fungal spores, and was added to all pots at the rate of 60ml per pot. Plants were grown in a greenhouse for two months (April-June). After six weeks of growth, the pots were placed over a wide plastic container and 250ml of tap water was added to each pot, resulting in ca 75 ml of leachate collected per pot. The leachates were filtered with syringe filters (pore size 0.2 μm, Minisart, Sartorius Stedim biotech, Göttingen, Germany) and were immediately frozen at -20°C. Leachates were collected four times during a two-week period. Leachates were analysed for total organic C (TOC) using a 5000A TOC-L analyser (Shimadzu, Japan) and ammonium, nitrate, phosphate and total organic nitrogen (TON) were measured in an Auto Analyser AA3 (Seal Analytical, UK). Subsamples for metabolomics were freeze-dried, and 10mg (dry weight) were analysed using UPLC-QTOF-MS (LCMS). Plants were harvested after eight weeks of growth.

In the feedback phase, the conditioned soil and roots from each pot were mixed with an equal weight of sterlised background soil mixture and used to fill five new pots of 0.5 L volume. Individual seedlings were planted into conspecific soil and soils conditioned by four other species from the same site. Soils originated from both soil inoculation treatments (whole vs sieved inoculum), and each seedling-soil combination was replicated five times, resulting in 50 pots per species altogether. Due to insufficient seed germination in S. pratensis and P. saxifraga, these species were excluded from the experiment, resulting in a total of 400 pots distributed among eight focal species. Plants were grown under the same conditions as in the conditioning phase. Plants were harvested after eight weeks of growth.