Where is the unseen fungal diversity hidden

Estimates of global fungal species richness have increased almost 3—5 fold in the past 20 years, from 1. In order to obtain more accurate estimates of true fungal diversity, increased sampling using high throughput DNA sequencing of many different types of environments is needed, and DNA banks may significantly contribute to filling this knowledge gap.

However, we know very little about Hawaiian fungi, their potential rates of endemism, and patterns of biodiversity. Similar rates of endemism are found in the Hawaiian flora. Similar to other DNA banks across the world, collections for common species contain multiple individuals with their own accession number.

All naturally occurring plant tissues harbor fungi as both endophytes, living in between plant cells Rodriguez et al.

Utilizing several plant bank samples from the HPDL we investigated the diversity of unintentionally co-sampled fungi found within banked plant samples. Thus, rather than superficially sequencing the fungi from all available replicate DNA extracts from a single plant species, we chose to deeply sequence ten samples of a common endemic Hawaiian plant genus, Clermontia Campanulaceae , with species found across the Hawaiian Islands Givnish et al.

This deep sequencing was done in hopes that the vast majority of fungi from our samples would be recovered from each sample. In this study, we utilized historical DNA bank samples to validate the use of plant bank samples as a resource for elucidating phyllosphere fungal biodiversity, while subsequently examining plant-associated fungal diversity across space. Our two main questions were do DNA bank samples store microbial diversity?

And can these previously collected samples be used to uncover ecological patterns, such as changes in microbial community similarity over space? Leaves were not disturbed by rinsing prior to DNA extraction. Approximately 1. De-multiplexed fastq files were obtained from the sequencing facility from the ten Clermontia plant bank samples.

These paired-end reads were merged with the Illumina Paired-End reAd mergeR PEAR , keeping reads with a minimum assembly length of bp, average quality threshold of 15 and above, and discarding all reads with any uncalled bases Zhang et al.

Potential chimeras were removed in vsearch Rognes et al. The most abundant sequence for each OTU was chosen as a representative sequence. All statistical analyses were conducted in R version 3.

Samples were rarefied to 16, reads, the minimum sample depth. Rarefaction, species accumulation curves were generated using the vegan package for: all samples, individual samples, and samples pooled by island Oksanen et al. Because observed species richness often under estimates true species richness Hughes et al.

Hill numbers and extrapolations were generated based on individual samples and individual species. A Venn diagram was generated to visualize overlapping taxa between islands using the VennDiagram package Chen, To investigate ecological patterns, we accounted for variables that may be influencing the fungal communities found in these banked samples.

These factors were temporal and physical distances between sample collections, as well as fungal community dissimilarity. Pairwise distance matrices were calculated for physical distance in kilometers using the geosphere package Hijmans, , time between sample collections in days, and Bray-Curtis community dissimilarity using the vegan package Oksanen et al.

Separate Mantel tests for each combination of the following pairwise distance matrices: time between sample collections days , as well as Euclidean physical distance between samples km , and community dissimilarity Bray-Curtis , were run for 10, permutations.

To investigate the effects of these variables a final partial Mantel test for physical distance and community dissimilarity, while controlling for time, was run for 10, permutations Oksanen et al. A total of 4,, sequence reads were obtained from the plant DNA library samples. Of these, 3,, paired-end reads Taxonomic assignment yielded a total of 1,, fungal reads that were binned into 2, fungal OTUs for use in in downstream analyses.

After removing OTUs with less than ten reads and rarefying to the sample with the minimum number of reads, we removed 1, OTUs While the observed OTU accumulation curve for all ten samples combined did not reach an asymptote Fig.

S1 , OTU accumulation curves by sample and by island except in the case of C. S2 and S3. We investigated patterns of fungal diversity at the phyla and ordinal levels. Overall, the majority of fungi in the subkingdom Dikarya dominated all of the phylloplane samples, with phylum Ascomycota being most abundant followed by Basidiomycota Fig.

Fungi belonging to the phylum Chytridiomycota and Zygomycota were also present in lower abundances. The top ten most abundant orders were Capnodiales, Chaetothyriales, Exobasidiales, Peltigerales, Pertusariales, Pleosporales, Tremellales, Ustilaginales, and two unknown orders Fig. Based on our observed data, total average fungal OTU richness by island was Overall about twenty OTUs were found on all of the five islands Fig.

Over this spatial range, while taking into account time number of days between sample collections, the fungal phylloplane communities exhibit a significant decrease in community similarity across increasing geographic distance Fig. Time was also significantly correlated with physical distance between sample sites Fig.

Time between sampling and community dissimilarity was marginally significantly correlated Fig. Despite the significant relationships with sampling time the Partial Mantel between distance and community dissimilarity while accounting for time, was significant Fig. In this study we investigated the diversity of phylloplane fungi associated with Clermontia spp.

We found that these specimens harbored a considerable diversity of fungi. After quality control, we found 1, fungal OTUs from just ten samples, representing 20 Clermontia individuals and eight species. Fungal richness ranged from to OTUs per plant sample. Incredibly, this diversity was recovered from a total of just 20 grams of leaf tissue from which DNA was extracted and preserved. Despite our high sequencing depth, the observed species accumulation curves for all samples and islands did not saturate, indicating our sequencing efforts likely underestimated true Clermontia phylloplane fungal diversity.

However, this novel use of DNA bank samples revealed substantial undiscovered fungal biodiversity stored in plant samples. This study highlights a new and underutilized function of biological collections, as well as gives insights into regional fungal diversity patterns.

Previous estimates of total regional fungal richness have been based off of plant to fungi ratios ranging from Hawksworth, to Taylor et al.

Our data supplement these studies using environmental NGS data. S6 that the entire Hawaiian flora c. This results in an approximate plant to fungi species ratio. In addition to the study of microbial diversity, questions regarding microbial biogeography, host specificity, and the effects of global change on microbial communities could be addressed with DNA banks.

For example, we were able to confirm the distance decay of microbial community similarity from DNA bank samples collected across the Hawaiian Islands. This finding is similar to other microbial systems where significant distance decay patterns were found in foliar endophytic Vaz et al. You can revoke your consent at any time. Where is the unseen fungal diversity hidden? A study of Mortierella reveals a large contribution of reference collections to the identification of fungal environmental sequences.

New Phytol 3 , Identifier doi: PMID: Privacy settings. Storage duration: This cookie remains only for the current browser session. Vimeo Display of videos. Youtube Display of videos.