Demographic studies measure drivers of plant fecundity such as seed production and survival, but few address environmental drivers of seed viability, such as germination and dormancy. Variation in climate and seed type may both directly and indirectly alter seed germination via altered fungal pathogen abundance.

We examined seed germination and microbial communities of seeds of Danthonia californica, which are either chasmogamous (external, wind-pollinated) or cleistogamous (internal, self-fertilized) and Festuca roemeri, which are solely chasmogamous. Seed populations were sourced across environmental gradients. We tested germination and characterized seed fungal community structure, using high-throughput sequencing.

For F. roemeri, maternal plant significantly influenced germination as did climate and pathogens; germination increased in wetter, cooler sites. For D. californica, the main drivers of germination were maternal plant, seed type and pathogens; on average, more chasmogamous seeds germinated. Fungal composition depended on seed type, with fewer fungi associated with cleistogamous seeds. Seed fungal composition varied with climate, plant density and mean proportion of seeds germinated.

Putative pathogens that were negatively correlated with germination were more abundant for both Danthonia and Festuca chasmogamous seeds than Danthonia cleistogamous seeds. In D. californica, cleistogamous and chasmogamous seeds contain vastly different fungal communities.

Germination: Sterile Petri dishes (90 mm) were lined with sterile filter paper and up to 25 seeds per dish (range 12-25) were arranged to reduce contact, with D. californica seeds separated into different dishes according to their seed type (up to 25 seeds/seed type/maternal family) . Hands, forceps and countertop were cleaned with alcohol between seed families to reduce carryover of seed fungi. All petri dishes were watered to dampness with distilled water, placed immediately into cold storage at 4°C for four weeks to break seed dormancy, and watered when necessary to keep damp but not wet. After cold stratification the seeds were kept at room temperature. Germinated seeds were recorded and removed every two days and petri dishes re-watered as needed.

Seed epiphytic fungi: When the number of germinating seeds approached zero (approximately 6 weeks after removal from 4°C), all ungerminated seeds (N=5,521), were examined with a dissecting microscope. Microbes on the seeds were categorized as fungi, bacteria, or unknown. When possible, fungi were keyed to genus. A small subset of seeds that germinated (N=27) were also examined for microbes.

To identify the fungi growing on ungerminated seeds to species and to verify the accuracy of morphology-based genus-level identifications, we isolated the fungi on culture media before extracting and sequencing DNA. The seeds were cultured on diameter 90 mm plates of 2% water agar medium supplemented with 20 mL/L of an antibiotic solution (5 g/L of penicillin, 5 g/L streptomycin, and 1.5 g/L chloramphenicol) to inhibit bacterial growth. Fungi growing from the cultured seeds were isolated and maintained on 2% malt agar plates. A representative sample of each morphotype was subcultured in 2% malt liquid media and incubated at approximately 25 C for 1-2 weeks to ensure sufficient mycelial growth for DNA extraction. DNA extraction and Sanger sequencing protocols followed Thomas et al. (2016). Sequences were then aligned against the UNITE fungal database using the BLAST algorithm (Altschul et al., 1990; Abarenkov et al., 2010) to obtain species hypotheses (Koljalg et al., 2013).

Please see the associated ReadMe files.