DNA barcoding: error rates based on comprehensive sampling

Abstract

DNA barcoding has attracted attention with promises to aid in species identification and discovery; however, few well-sampled datasets are available to test its performance. We provide the first examination of barcoding performance in a comprehensively sampled, diverse group (cypraeid marine gastropods, or cowries). We utilize previous methods for testing performance and employ a novel phylogenetic approach to calculate intraspecific variation and interspecific divergence. Error rates are estimated for (1) identifying samples against a well-characterized phylogeny, and (2) assisting in species discovery for partially known groups. We find that the lowest overall error for species identification is 4%. In contrast, barcoding performs poorly in incompletely sampled groups. Here, species delineation relies on the use of thresholds, set to differentiate between intraspecific variation and interspecific divergence. Whereas proponents envision a "barcoding gap" between the two, we find substantial overlap, leading to minimal error rates of approximately 17% in cowries. Moreover, error rates double if only traditionally recognized species are analyzed. Thus, DNA barcoding holds promise for identification in taxonomically well-understood and thoroughly sampled clades. However, the use of thresholds does not bode well for delineating closely related species in taxonomically understudied groups. The promise of barcoding will be realized only if based on solid taxonomic foundations.

Figure 1. Phylogenetic Relationships and Terminology

(A) Reciprocal monophyly. Members of each species share a unique common ancestor. For each species, the white star represents the coalescent, the point at which all extant haplotypes share a common ancestry. (B) Paraphyly. One species (Y), is monophyletic, but nests within another recognized species (X). Thus, the coalescent of species Y (small star) is contained within the coalescent of species X (large star). (C) Polyphyly. Neither species X or Y are monophyletic, and both coalesce to the white star.

Figure 2. Schematic of the Inferred Barcoding Gap

The distribution of intraspecific variation is shown in red, and interspecific divergence in yellow. (A) Ideal world for barcoding, with discrete distributions and no overlap. (B) An alternative version of the world with significant overlap and no gap.

Figure 3. Intraspecific and Interspecific Estimations

A subclade of five cowrie ESUs shows how both coalescent and divergence depths are generated. The two most disparate individuals are culled from within each ESU (left—red) and used in a constrained phylogeny with a molecular clock enforced (right) to recover both the maximum coalescent depth (red) and the divergence depths between sisters (yellow). Two young ESUs (stars) would be missed (false negatives) if a 3% threshold cutoff (shown) was employed. Note that Palmadusta artuffeli, a Japanese endemic species, is nested among monophyletic subspecies of the paraphyletic species P. clandestina. The black circle indicates the coalescent for the species P. clandestina, and the black star indicates the interspecific divergence for species-level analyses.

Figure 4. Sample Size Effect on Intraspecific Variation

A. Coalescent depth vs. sample size. B. Histograms for coalescent depths of various sample size classes. Mean coalescent depth increases with increased sample size, from 0.0049 for n ≥ 2, to 0.0057 for n ≥ 5, to 0.0068 to 0.0070 for n ≥ 10.

Figure 5. Alternative Metrics for Intraspecific Variation

(A) Distribution of all intraspecific pairwise K2P distances for cowrie ESUs with n ≥ 10, turbinids and limpets. Left y-axis for cowries; right y-axis for others. (B) Comparison between estimated theta versus estimated maximum coalescent for each cowrie ESU with n ≥ 10; r² = 0.837. (C) Distribution of theta values for cowrie (n ≥ 10), turbinid and limpet ESUs.

Figure 6. Interspecific Variation

Distribution of divergence depths between terminal ESUs and their sister ESU(s) in cowries, turbinids and limpets.

Figure 7. Barcoding Overlap: Cowrie ESUs

(A) Relative distributions of intraspecific variability (coalescent depth—red) and interspecific divergence between ESUs (yellow), demonstrating significant overlap and the lack of a barcoding gap. Note that the x-axis scale shifts to progressively greater increments above 0.02. (B) Cumulative error based on false positives plus false negatives for each threshold value. The optimum threshold value is 0.013 (2.6%), where error is minimized at 17%.

Figure 8. Intraspecific Variation Based on Recognized Species

(A) The distribution of all intraspecific pairwise distances for traditionally recognized cowrie species with n ≥ 10. The white bars represent intraspecific distances where the two specimens compared fall into separate ESUs. (B) Theta values for traditionally recognized cowrie species. Black bars are species that correspond to an ESU; white bars are species that include multiple ESUs.

Figure 9. Barcoding Overlap: Cowrie Species

Data are presented as in Figure 7; however, estimates of intraspecific variation and interspecific divergence are based on traditionally recognized cowrie species. (A) Relative distributions of intraspecific variability (coalescent depth—red) and interspecific divergence between species (yellow), demonstrating a more pronounced overlap than when utilizing ESUs. Note that the x-axis scale shifts to progressively greater increments above 0.02. (B) Cumulative error based on false positives plus false negatives for each threshold value. The optimum threshold value is 0.025 (5%), where error is minimized at 33%.