Conceived and designed the experiments: JMW LMJ DHF XZ BK CJG RED DJB BR IP PDNH. Performed the experiments: JMW LMJ DHF XZ BK CJG RED DJB BR IP. Analyzed the data: JMW LMJ DHF XZ BK CJG RED DJB BR IP. Contributed reagents/materials/analysis tools: JMW LMJ DHF XZ BK CJG RED DJB BR IP PDNH. Wrote the paper: JMW PDNH LMJ DHF.
The authors have declared that no competing interests exist.
DNA barcoding of aquatic macroinvertebrates holds much promise as a tool for taxonomic research and for providing the reliable identifications needed for water quality assessment programs. A prerequisite for identification using barcodes is a reliable reference library. We gathered 4165 sequences from the barcode region of the mitochondrial cytochrome
DNA barcoding
The application of DNA barcoding to freshwater biomonitoring has recently generated much interest for several reasons
An essential requirement of biomonitoring, conservation biology, and ecology is a sound taxonomy of the study organisms. Although taxa could be based purely on operational taxonomic units (OTUs) defined by sequence data, information on key functional niche traits (e.g.
Mayflies are abundant in most aquatic habits, and show varied tolerance to differing disturbance regimes, making them ideal candidates for monitoring water quality. Together with the caddisflies (Trichoptera) and stoneflies (Plecoptera), they comprise one of the most commonly used biomonitoring metrics, EPT richness
The development of a barcode reference library for North American mayflies was initiated by sequencing 150 specimens from 80 North American species
The mayfly fauna of North America includes 651species (Mayfly Central,
Although most specimens derived from east of the Rocky Mountains, collections were made across North America (
Family | # known NAspecies andsubspecies | # barcodedspecies | # barcodedprovisionalspecies | # species withMXID ≥2.2% | Mean MXID | Mean MNID |
Acanthametropodidae | 2 | 0 | 0 | 0 | – | – |
Ameletidae | 35 | 20 | 1 | 4 | 1.2 | 13.6 |
Ametropodidae | 2 | 1 | 0 | 0 | 0.2 | – |
Baetidae | 146 | 58 | 37 | 32 | 6 | 15.3 |
Baetiscidae | 11 | 4 | 0 | 2 | 2.8 | 9.3 |
Behningiidae | 1 | 1 | 0 | 1 | 7.9 | – |
Caenidae | 36 | 12 | 1 | 6 | 8 | 11.6 |
Ephemerellidae | 72 | 48 | 13 | 21 | 4.7 | 11.7 |
Ephemeridae | 13 | 9 | 0 | 3 | 3.3 | 8.2 |
Euthyplociidae | 1 | 0 | 0 | – | – | – |
Heptageniidae | 130 | 56 | 21 | 19 | 2.8 | 10.2 |
Isonychiidae | 15 | 7 | 4 | 2 | 1.5 | 13.3 |
Leptohyphidae | 35 | 6 | 8 | 2 | 2.4 | 14.1 |
Leptophlebiidae | 88 | 24 | 10 | 9 | 3.4 | 14.1 |
Metretopodidae | 9 | 2 | 0 | 1 | 5.1 | 14.3 |
Neoephemeridae | 4 | 1 | 0 | 0 | 0.8 | – |
Oligoneuriidae | 8 | 0 | 0 | – | – | – |
Palingeniidae | 1 | 1 | 0 | 0 | 1.6 | – |
Polymitarcyidae | 7 | 2 | 0 | 1 | 4.8 | 6.3 |
Potamanthidae | 5 | 1 | 0 | 0 | 0.2 | – |
Siphlonuridae | 24 | 11 | 5 | 1 | 1.8 | 9.4 |
All distances are % K2P; MNID = minimum interspecific K2P distance, MXID = maximum intraspecific K2P distance.
No species shared haplotypes but previously published sequences for nine currently valid species clustered with specimens assigned to a different taxon. Whenever we were able to reexamine specimens, the original morphological identification proved incorrect.
The average maximum intraspecific divergence was 3.9% (max = 26.7%) and the average intraspecific divergence was 1.97%. Minimum interspecific distances ranged from 0.3–24.7% (mean: 12.5%). One hundred five species (29.7%) had maximum intraspecific divergences greater than 2.2%, a level of divergence found to delimit species across diverse groups of insects
Increasing the geographic range between samples did not always lead to large increases in intraspecific divergence. For example,
Among the morphologically distinct species
Species | Diagnostic Nucleotides (position:nucleotide) |
|
364:T, 502:C |
|
208:A, 232:C, 574:A |
79:A, 172:C, 397:T, 433:A, 541:A, 553:A |
All but 9 of the 44 species with a maximum intraspecific divergence greater than the minimum interspecific divergence were polyphyletic or paraphyletic.
New distribution records were discovered for
This paper presents the largest barcoding dataset of any order of aquatic insects, providing records for more than 4000 individuals from over 350 species. The analysis of both larvae and adults from several localities and some topotypes enabled identification with less ambiguity than in the past. This allowed the correction of previously misidentified species and strengthened knowledge of the levels of COI variation both within and between species. Our results confirm that levels of sequence divergence among closely allied species are generally high. For example, congeneric species of North American Ephemeroptera showed a mean barcode divergence of 13.9%, a much higher value than the 7–8% divergence reported for congeneric species of Lepidoptera in both North America
Our results indicate that much more taxonomic work is required on North American Ephemeroptera as many currently recognized species include several highly divergent, often polyphyletic, haplotypes, usually correlated with morphological differentiation among lineages. Most of these species have complex histories of synonymy, reflecting a 60 year trend in North American mayfly systematics towards inclusive species concepts. This trend was driven by the observation of individuals morphologically intermediate between named species, or by observations suggesting that members of one species fell within ‘the expected range of variation’ of another. Nearly all of these decisions of synonymy were based only on limited morphological studies, without consideration of biogeographic, ecological, behavioral or molecular data. While this fusion of species has simplified morphology-based identifications, it now seems likely that this trend often ignored biological reality. For example, the four species in the
Our results reveal that many species currently treated as synonyms may well represent valid species. A critical step in the resolution of this uncertainty lies in the acquisition of DNA barcodes from the holotype of each taxon as it represents the only unambiguously identified specimen. Because type specimens of North American Ephemeroptera have a mean age of 75 years (range 0–250 years), DNA degradation will be significant. However, prior work on Lepidoptera has shown that DNA barcode sequences can regularly be recovered from specimens that are less than a century old by assembling short amplicons
While obtaining barcodes from type specimens will aid in the application of names to barcode clusters, further taxonomic work integrating multiple lines of evidence may be required to test current species hypotheses. For example, our results show that additional examination of the North American Baetidae and Ephemerellidae is required as they both contain large numbers of species with multiple barcode clusters, high intraspecific morphological variation, and many interim species identifications. DNA barcoding results cannot only serve as a guide of where to focus these efforts, but also be used as part of an iterative revisionary process
This study has increased barcode coverage for North American Ephemeroptera from 121 to 354 species, or over 50% of the known fauna. Because the present library includes records for nearly all common species, most Ephemeroptera taken in routine biomonitoring samples may now be rapidly and accurately identified through DNA barcoding, albeit with the proviso that further taxonomic work is needed to clarify species boundaries. Our progress in constructing a barcode library for North American Ephemeroptera further indicates the feasibility of gaining global coverage both for this order and for other key groups of aquatic insects.
Detailed collection data are included in
Topotypes, specimens collected from the type locality and which have a high certainty of representing the type concept of a species, were sequenced for the following species:
Each barcode sequence (>400 bp) for North American Ephemeroptera from previous publications (150 -
Sequences were obtained from the following Barcode of Life Data systems (BOLD systems:
COI sequences were aligned in MEGA 5
Neighbour Joining tree using Kimura-2-Parameter distance for COI DNA sequences from 4065 individuals of North American Ephemeroptera. Specimens and species with topotypes are indicated with ‘*’; specimens and species from the same general area as the type locality are indicated with ‘#’.
(PDF)
Species-level summary of K2P distance, sample distribution, and sample size for North American Ephemeroptera. All distributional records use standard 2-letter (or 3-letter, for Mexico) postal abbreviations. MNID = minimum interspecific K2P distance, MXID = maximum intraspecific K2P distance, species with maximum intraspecific sequence divergence ≥5.0% are indicated with ‘*’.
(DOC)
Updated identifications for Ephemeroptera specimens with published barcode records.
(DOC)
Collection data for all barcoded specimens of Ephemeroptera.
(XLS)
Taxonomic notes on selected North American Ephemeroptera species with barcode records.
(DOC)
Special thanks are given to Parks Canada for help with access to field sites in Cape Breton Highlands National Park, and to the Stroud Water Research Center, Pennsylvania, the Southern California Coastal Water Research Project, Costa Mesa, California, and the Aquatic Bioassessment Lab, California State University, Chico for providing California specimens and valuable discussion. Additional specimens were provided by the Highlands Biological Station, North Carolina. S. Bateson, Biodiversity Institute of Ontario, provided assistance in production of trees and maps.