Herbarium sampling-
We used nodules collected from herbarium specimens of Medicago lupulina housed at the Marion Ownbey Herbarium at Washington State University. Marion Ownbey houses 400,000 vascular and non-vascular plant specimens, including 122 specimens of Medicago lupulina collected primarily from North America (109 specimens). For our experiments we used nodules from historic and contemporary specimens collected around Washington State from Skagit County (1 nodule), Klickitat County (2 nodules), Grant County (1 nodule), Yakima County (1 nodule), Ferry County (1 nodule) and Whitman County (2 nodules) from 1925-2019 (Table 1). For each specimen we use a clean, autoclaved razor blade to excise nodules and a similarly sized piece of leaf material to be used as a control for environmental contamination (Heberling and Burke 2019). We used gloves and a clean surface (ethanol sterilized, autoclaved trays) to collect material. Nodules and leaf material were stored in sterile Eppendorf tubes at room temperature until use in the culturing experiment. We used contemporary nodules collected from the Washington State University campus in Pullman, Washington. Contemporary nodules included one air dried nodule, stored at room temp in an Eppendorf tube for one year (2018), a fresh collected nodule (2019), and a lab strain of Ensifer medicae (WSM-419).
Culture collection -
Seven different media types were prepared: Luria-Bertani (LB), Tryptone-Yeast (TY), Yeast Mannitol (YM), diluted YM, YM plus calcium, YM plus sodium chloride (NaCl), and YM plus pyruvate. The diluted YM was prepared as a ½X dilution of YM media for all ingredients except agar. The YM plus calcium was prepared according to the YM recipe with the addition of 2 grams calcium carbonate (CaCO3) per liter, instead of 1 gram per liter. Similarly, the YM plus NaCl was prepared with 1 gram of NaCl per liter, which is 10 times the amount in the protocol. The YM plus pyruvate was prepared as indicated in the protocol, with the addition of 600ul of filter-sterilized (Acrodisc syringe filter, 0.2µM, VWR, Atlanta, GA, USA) 2M sodium pyruvate to the media post-autoclaving to produce a final concentration of 2mM sodium pyruvate. Media recipes can be found in Table S1. Luria-Bertani (LB), Tryptone-Yeast (TY), Yeast Mannitol (YM) were all chosen for this study because they are commonly used to isolate rhizobia and non-rhizobia species in plants. We used a dilute YM in the hopes that by diluting nutrient rich media, we might discourage the growth of fast-growing species such as Bacillus , allowing slower-growing species to grow, as well to lower the presence of free-radicals that can inhibit cell growth (Oliver 2005). Salt was added to retain osmotic balance and regulate membrane potential (Cinátl 1969). Calcium was added because it has been demonstrated to increase viability of rhizobia, by strengthening the cell wall (Vincent 1962). Finally, pyruvate was added to prevent oxidative damage (Giandomenico et al. 1997) by eliminating metabolic by-products and free-radicals (Stevenson et al. 2004).
Nodules and leaves were rehydrated in filter-sterilized (Acrodisc syringe filter, 0.2µM, VWR, Atlanta, GA, USA) 10% sucrose solution for 48 hours. After rehydration, nodules were rinsed in autoclaved, filtered water Milli-Q® seven times before transfer to 200ul ½X phosphate buffered saline (PBS) for an additional 24 hours. Nodules (and leaves) were crushed then streaked onto ten plates of each media type for the 1950 specimen, but three plates of each media type for years 1974-2019. After completing the nodule and leaf control for the 1950 sample it seemed the amount of liquid was not enough to adequately streak 70 plates. An autoclaved, filtered Milli-Q®water blank and an autoclaved PBS blank were used as negative controls and were streaked onto five plates of each media type. Plates were sealed with parafilm then incubated at 30°C for 12 hours before being observed for growth.
Plates were observed for growth every 24 hours. Plates with growth after five days were re-streaked for isolation onto the same media type using the quadrant streak method. Each sample underwent at least two isolation re-streaks onto the same media type. If a plate had a single isolated strain after two sets of re-streaks, cells were harvested for glycerol stocks and pellets for DNA extractions. If the plate was not a single isolated strain (>1 colony morphology on the plate), isolation restreaks continued for each morphologically unique colony on the plate until they were isolated. This took up to four sets of re-streaks for some isolates. We described the isolates using criteria from the American Society for Microbiology (Breakwell et al. 2007), describing the pigmentation, form, margin, elevation, and texture (Table S2). Glycerol stocks were made by flooding the plates with 3ml of sterile ½X PBS. Cells were lifted from the surface of media and mixed into the buffer using sterile plastic pestles. We added 750µl of cell/buffer slurry to each 2 ml stock tube each with 750µl of 80% glycerol. Stock tubes were vortexed then stored in the -80 °C freezer. Remaining cell/buffer slurry was added to a 2ml Eppendorf tube and centrifuged at 10,000rpm for 3 minutes or until a pellet formed. Once pelleted, all supernatant was removed. Pellets were stored at -80 °C prior to use for DNA extractions.