Does One Size Fit All? Variations in the DNA Barcode Gaps of Macrofungal Genera

Abstract

The nuclear ribosomal internal transcribed spacer (nrITS) region has been widely used in fungal diversity studies. Environmental metabarcoding has increased the importance of the fungal DNA barcode in documenting fungal diversity and distribution. The DNA barcode gap is seen as the difference between intra- and inter-specific pairwise distances in a DNA barcode. The current understanding of the barcode gap in macrofungi is limited, inhibiting the development of best practices in applying the nrITS region toward research on fungal diversity. This study examined the barcode gap using 5146 sequences representing 717 species of macrofungi from eleven genera, eight orders and two phyla in datasets assembled by taxonomic experts. Intra- and inter-specific pairwise distances were measured from sequence and phylogenetic data. The results demonstrate that barcode gaps are influenced by differences in intra- and inter-specific variance in pairwise distances. In terms of DNA barcode behavior, variance is greater in the ITS1 than ITS2, and variance is greater in both relative to the combined nrITS region. Due to the difference in variance, the barcode gaps in the ITS2 region are greater than in the ITS1. Additionally, the taxonomic approach of "splitting" taxa into numerous taxonomic units produces greater barcode gaps when compared to "lumping". The results show variability in the barcode gaps between fungal taxa, demonstrating a need to understand the accuracy of DNA barcoding in quantifying species richness. For taxonomic studies, variability in nrITS sequence data supports the application of multiple molecular markers to corroborate the taxonomic and systematic delineation of species.

Keywords: DNA barcode; ITS1; ITS2; nrITS; pairwise distances.

Figure 1

Distributions of intra- and inter-specific pairwise distances around the barcode gap (a). When means are closer and/or variances around the means are greater, distributions can overlap, eliminating the barcode gap (b). An examination of distribution quantiles can be used to understand the effectiveness of molecular barcode sequence data and the barcode gap for a taxon.

Figure 2

The following macrofungal genera were studied: (a) Hebeloma celatum (H. J. Beker), (b) Laccaria amethystina group, (c) Marasmius sp., (d) Suillus tomentosus (N. Nguyen), (e) Russula magnarosea clade (C. R. Noffsinger), (f) Stereum hirsutum (M. G. Wood, Mycoweb.org), (g) Trametes versicolor (M. G. Wood), (h) Sarcodon imbricatus, (i) Hydnum sp. (R. A. Swenie), (j) Morchella elata group and (k) Elaphomyces muricatus (M. G. Wood). All pictures by A. W. Wilson unless otherwise indicated.

Figure 3

Pairwise distances of DNA barcode sequence data using a K80 substitution model. Horizontal lines are the inter-specific means for the nrITS, ITS1 and ITS2 pairwise distance distributions from 5146 sequences. Stereum is evaluated under two partitions. The “lumping” partition has 20 species and is represented by open boxplots. The “splitting” partition has 30 species and is represented in the filled boxplots.

Figure 4

Variances in pairwise distance distributions, showing ranked nrITS sequence data for intra-specific (a) and inter-specific (b) distances.

Figure 5

How “lumping” and “splitting” approaches to species delineation affect pairwise distance distributions. Intra-specific variation is broader when using the “lumping” approach (a, red distribution) than when using the “splitting” approach (b). For inter-specific variation, this relationship is inverted such that “lumping” (c, blue distribution) is narrower than “splitting” (d).