The frequency of sex: population genomics reveals differences in recombination and population structure of the aflatoxin-producing fungus Aspergillus flavus
Figure S1. Sampling locations of two north-south transects in the United States. The eastern transect fields were in Pennsylvania (PA), North Carolina (NC), and Florida (FL). The central transect fields were in Iowa (IA), Indiana (IN), Oklahoma (OK), and Texas (TX). Three fields were sampled in each state; when only one or two stars are present in a state on the map, then more than one field fell within the area covered by the star. No two fields were within 1 km of each other. The underlying map is modified from a public domain image (https://commons.wikimedia.org/wiki/File:Blank_US_map_borders.svg). Reprinted from Drott, M.T., Fessler, L.M. and Milgroom, M.G., 2019. Population subdivision and the frequency of aflatoxigenic isolates in Aspergillus flavus in the United States. Phytopathology, 109(5), pp.878-886.
Figure S2. Estimates of the linkage disequilibrium decay (r2) from rarefaction analysis (detailed in Supplemental methods) using 60 random samples with replacement of 50,000 SNPs and four, six, or eight clone-corrected isolates (sample sizes are indicated on right side of graph) sampled from three populations of Aspergillus flavus in the United States.
Figure S3. Maximum likelihood tree of concatenated nucleotide sequences of CaM, BenA, RPB2, and the ITS rDNA region from 94 Aspergillus flavusisolates sampled across the United States. Data from other species in section Flavi of Aspergillus were obtained from Frisvad et al. (2019). A. flavusisolates from this study have been color-coded to reflect populations A (green), B (red), C (purple) determined from whole-genome sequences. Isolates 21, 55 and 83 are S-strain isolates and were not included in population-level analyses. Isolate 83 was determined in this study to be A. texensis, a recently discovered species (Singh et al. 2018). Branch tip labels refer to isolate numbers that are defined in Table S1.
Figure S4. The mean number (A) and size of sclerotia (B) in isolates of Aspergillus flavus sampled from three L-strain populations of Aspergillus flavus in the United States. From each population we randomly sampled 11 isolates and assessed sclerotial phenotypes after seven days of growth on both Czapek-dox agar (CZA) and potato dextrose agar (PDA). Additionally, two S-strain isolates were assessed. Size measurements were performed on a random sample of 30 sclerotia (when available) from each isolate. Many isolates did not produce sclerotia on one or either medium. From comparison of sclerotial sizes from isolates that produced on both media, we did not find an effect of media type on sclerotial size (data not shown). We thus averaged sclerotial sizes across media types for comparison between populations in panel B. Sample sizes indicated in panel B reflect the number of independent isolates. Error bars are SE.
Figure S5. Plot of Bayesian information criterion (BIC) versus the number of possible clusters representing SNP data from 94 isolates of Aspergillus flavus from the United States. BIC is a goodness of fit measure, whereas clusters represent potential populations. The lowest point indicates the optimal number of populations. We inferred three populations of L-type A. flavus, with a fourth population comprising two S-type isolates, which was not well supported. S-type isolates were not included in subsequent population-level analyses.
Figure S6. Neighbor-net network generated using single nucleotide polymorphisms from whole-genome sequences of 94 Aspergillus flavus isolates collected in the United States and five Aspergillus oryzae isolates previously sequenced by Gibbons et al. (2012) (network A) or one Aspergillus texensis isolate (network B). The overall population of A. flavusL-type isolates is subdivided into three populations: A (green), B (red), and C (purple). Branch tip labels refer to isolate numbers that are defined in Table S1. S-type isolates (21 and 55 at top center of both networks) and A. texensis (83 at top center of network B) A. oryzae isolates (bottom right of network A; names start with “RIB”) were not colored-coded as they were not included in population-level analyses. Branch tip labels refer to isolate numbers that are defined in Table S1.
Figure S7. Maximum likelihood tree generated from concatenating nucleotide sequences of amdS12, omt12, and trpC13 genes in Aspergillus flavus isolates from this study and those from Geiser et al. (2000). The overall population of L-type isolates from this study is subdivided into three populations: A (green), B (red), and C (purple); S-type isolates were not colored as they were not included in population-level analyses. Several isolates from population B are missing the omt12 gene from the aflatoxin gene cluster and were thus not included in this analysis. Annotations on the tree refer to clade names from Geiser et al. (2000). Branch tip labels refer to isolate numbers that are defined in Table S1.
Figure S8. The number of effective codons (Nc) in (A) all genes and (B) the top 100 highly expressed genes (HEGs) across three populations of L-type Aspergillus flavus in the United States.
The file "aflavus.arrow.pilon.gff3" contains the annotation of the high-quality NRRL3357 genome that was used as a reference. The corresponding version of the genome has been uploaded as "aflavus_scaffolds.fasta.gz". While this annotation and genome version were used for this study, newer versions are available on the JGI portal.
