Supplementary Table 1.Sample sites, collection information, burn status, microsite, and isolation frequency data for each tree. Individuals sampled more than once are listed separately by sampling event. JD: J. deppeana, QH: Q. hypoleucoides, QT: Q. turbinella.
Supplementary Table 2.Top BLAST hits of 1298 sequences.
Supplementary Table 3. Ca, K, Mg, and N content in mature, symptomless leaves of J. deppeana and Q. hypoleucoides. Values indicate percentage of leaf dry weight. JD: J. deppeana, QH: Q. hypoleucoides.
Supplementary Table 4.Occurrence of 95 OTU as a function of host taxon and burn status, and class-level taxonomic placement of each OTU. Data represent sequenced isolates including both ITS1 and ITS2 sequence (see methods).
Supplementary Table 5.PERMANOVA shows community structure significantly differed as a function of host taxon and fire age class but did not differ as function of host taxon X fire age class. * indicates significant results
Supplementary figure legends
Supplementary Fig. 1 Regression analyses with least-squares contrast for endohpyte isolation frequency and diversity in the Santa Rita Mountains reveal differences between fire age classes defined by≤6 y and 7–18 y, providing a basis for delineating very recent (≤6 y since fire) and recent (7–18 y since fire) age classes. a log-transformed isolation frequency, with season and host species set as random factors and fire age class as the explanatory variable (F1, 19.82=10.96, P=0.0035). b log-transformed Fisher's alpha, with host species set as a random factor and fire age class as the explanatory variable (F1, 11.04=5.60, P=0.0374). Different letters represent significant differences between fire age classes. Season was not included in the latter analysis because it was not associated with differences in diversity (see methods)
Supplementary Fig. 2 Herbivory and pathogen damage scores for Quercus spp., and their relationship to fire age class, endophyte isolation frequency, and diversity. a Herbivory damage was significantly lower (*) in≤6 y fire age class (ANOVA, F 2, 148=8.4947, P=0.0003) relative to the other fire age classes. Pathogen damage did not differ as a function of fire age class (ANOVA, F 2, 148=0.5559, P=0.5747). b-e Herbivory and pathogen damage scores were not related to isolation frequency and diversity.
Supplementary Fig. 3In vitro fungal growth on leaf extracts prepared from healthy foliage of J. deppeanaor Q. hypoleucoides. Colony radius was compared for four endophyte isolates grown on extracts from the same host species from which they were isolated (same host), vs. the other (different host). Different letters indicate significant differences. Data show the colony radius of YLH0024 after 18 d (t 22=−1.21, P=0.2389), YLH0063 after 10 d (t 21=2.53, P=0.0193), YLH0084 after 8 d (t 22=−0.97, P=0.3419), and YLH0112 after 20 days (t 22=−3.07, P=0.0057)
Supplementary Fig. 4 Linear regression of a isolation frequency and b diversity vs. foliar Ca, K, Mg, and N content in J. deppeana
Supplementary Fig. 5Linear regression of aisolation frequency and b diversity vs. foliar Ca, K, Mg, and N content in Q. hypoleucoides
Supplementary Fig. 6Leaf Ca, K, Mg and N content did not differ significantly among fire age classes in a J. deppeana (Ca: F 2, 30=1.38, P=0.2679; K: F 2, 30=1.50, P=0.2401; Mg: F 2, 30=0.82, P=0.4486; N: F 2, 30=2.47, P=0.1019) or bQ. hypoleucoides (Ca: F 2, 20=1.49, P=0.2496; K: F 2, 20=0.60, P=0.5591; Mg: F 2, 20=0.31, P=0.7387; N: F 2, 20=0.02, P=0.9851)