Phylogeny of Ephemeroptera (mayflies) based on molecular evidence
Introduction
Ephemeroptera (mayflies) is a monophyletic group of semi-aquatic pterygote insects, comprising 3083 species, 376 genera, and 37 described families (Brittain and Sartori, 2003). They are present on all continents, excluding Antarctica, and are associated with freshwater and brackish water habitats. Nymphs have much longer antennae, functioning mandibulate mouthparts, and are aquatic, in contrast to the adults which lack mouthparts and do not feed, relying on the nutritional buildup from the immature stages. Mayfly diversity is greatest in lotic habitats in the temperate and tropic regions, where they are an important part of the food chain, consuming primary producers such as algae and plants, and as a food source for vertebrate predators like fish. Additionally, these insects are used as bioindicators of pollution and water quality. The adults are soft-bodied insects possessing short antennae, vestigial mouthparts, two long cerci, and typically possess a medial caudal filament originating from the last abdominal segment. Adult mayflies typically have two pairs of wings, however, the second pair is considerably smaller than the first and in some species is absent altogether. Ephemeroptera is unique among pterygote insects in possessing functional wings at the penultimate molt (subimago stage), prior to the full development of genitalia; in all other insects the presence of functional wings occurs only after the final molt (Brittain, 1982, Brittain and Sartori, 2003, Edmunds, 1996).
Ephemeroptera has been considered by many to be sister to Odonata + Neoptera (Fürst von Lieven, 2000, Kristensen, 1991, Staniczek, 2000, Wheeler et al., 2001, Whiting et al., 1997) although alternate hypotheses have been suggested (Boudreaux, 1979, Brodsky, 1994, Hennig, 1981, Hovmöller et al., 2002, Kukalova-Peck, 1991, Kukalova-Peck, 1997, Martynov, 1924, Matsuda, 1970, Riek and Kukalova-Peck, 1984). Recently, it was shown that, while data from three molecular loci ambiguously resolved basal pterygote relationships, total evidence analysis (combined molecular and morphological data) strongly supports the position of mayflies as sister to all other extant pterygotes (Ogden and Whiting, 2003).
Since, the current suite of evidence supports mayflies as sister group to all other winged insect orders, a robust phylogeny for Ephemeroptera should shed light on proposed hypotheses concerning the evolution of wings in insects. Specifically, the proposed hypothesis of pleural origins for wings from gills could be examined (Brodsky, 1994, Kukalova-Peck, 1978, Kukalova-Peck, 1983, Kukalova-Peck, 1991). One of the underlying assumptions of this hypothesis is that articulated pleural extensions, initially used as gills for respiration, served as a morphological transition to wings from the immature to the mature life stages. This particular assumption could be examined in a phylogenetic framework. For example, if the basal lineages of mayflies do not have highly innervated, movable, paddle-like gills, with well-developed associated musculature, then the proposed “mayfly gills to wing origin” hypothesis loses strength. Additionally, the pattern of loss of the imago of certain mayfly lineages could elucidate hypotheses concerning the homologies between mayfly life stages (subimago, imago) and life stages of other winged insects (imago). Moreover, these patterns could support or reject the notion that flight evolved out of the water as opposed to a terrestrial origin. There are also some interesting evolutionary trends within the mayflies that could be examined given a robust topology, such as the burrowing lifestyle and associated morphological features (i.e., mandibular tusks, gills, etc.), the fishlike body form and swimming behavior, and the presence of a carapace (fused pronotum, mesonotum, and wing buds), among others. Therefore, a robust phylogeny for Ephemeroptera should assist further investigation of important evolutionary trends, not only within the mayflies, but in the winged insect groups as well.
Section snippets
Review of mayfly classification and phylogeny
After the earliest taxonomic treatments (Leach, 1815, Linnaeus, 1758, Pictet, 1843–1845), more comprehensive works began to subdivide mayflies into more taxa based on diagnosed adult characters, with some larval structures depicted in the descriptions (Eaton, 1871, Eaton, 1883–1888, Eaton, 1968). Later classifications began to use more larval characters, due to their apparent usefulness in determining natural groupings (Ulmer, 1920) and this practice for proposing classifications continued up
Taxon sampling
Taxonomic sampling consisted of exemplars representing 94 spp. of Ephemeroptera, 9 spp. of Odonata, and 5 spp. of non-pterygote insects for a total of 108 taxa (Table 2). All direct optimization analyses were rooted to the Collembola (Hypogastruridae). Within Ephemeroptera, 89 genera, from all four suborders, and from 31 families, representing ∼24% of the genera and 84% of families were included. Numerous genera from large, diverse families were included to better represent the major lineages
Results
All of the amplified H3 sequences exhibited a conserved reading frame. A total of 365 bases from this gene were used in phylogenetic reconstruction and were treated as pre-aligned data. The longest complete mayfly sequences and average lengths (respectively) for the remaining genes were: 385 (Epeorus sp.) and ∼380 bp for 12S; 593 (Tricorythodes sp. and Neoephemera youngi) and ∼570 bp for 16S; 1932 (Yarina mota) and ∼1850 bp for 18S; and 3223 (Hexagenia sp.) and ∼3100 bp for 28S.
Direct optimization
Higher level
The direct optimization analyses of the data strongly support a monophyletic Ephemeroptera with a Bremer support value (Bs) of 84, and bootstrap value (bt) of 100. Nodal support for the placement of Baetidae as sister to all remaining clades is strong (Bs = 40 and bt = 100), however, the node is not stable in the parameter landscape, being present in only one other parameter set (2:2:1). The character state reconstruction in MacClade for gill movabililty is equivocal when mapped on the parsimony
Conclusion
This analysis represents the first formal analysis across almost every major lineage of mayflies and is the first molecular phylogeny for the Order Ephemeroptera. The analyses included 31 of the 37 families, representing ∼24% of the genera. 11 families were supported as being monophyletic, although four others (Behningiidae, Caenidae, Ephemerellidae, and Leptohyphidae) were recovered in a large portion of the parameter landscapes supporting their monophyly as well; five families were supported
Acknowledgments
We thank M. Sartori, J-L Gattolliat, Sartori Lab, P. McCafferty, L. Jacobus, P. Randolph, McCafferty Lab, A. Haybach, T. Hitchings, H. James, M. Pescador, FAMU group, E. Dominguez, J. Peters, B. Richard, A. Huryn, K. Tojo, K. Finley, J. Skivington, D. Lemkuhl, N. Kluge, and J. Webb, for providing specimens and other advise on the project; Whiting Lab for valuable discussion and assistance. Analyses were performed in the Fulton Supercomputer Center, Brigham Young University, and parallel
References (76)
Exploring the behavior of POY, a program for direct optimization of molecular data
Cladistics
(2001)- et al.
Data exploration in phylogenetic inference: scientific, heuristic, or neither
Cladistics
(2003) - et al.
The Paleoptera problem: basal pterygote phylogeny inferred from 18S and 28S rDNA sequences
Cladistics
(2002) - et al.
The problem with “the Paleoptera Problem”: sense and sensitivity
Cladistics
(2003) - et al.
Multiple sequence alignment in phylogenetic analysis
Mol. Phylogenet. Evol.
(2000) - et al.
Probabilism and phylogenetic inference
Cladistics
(1997) - et al.
The phylogeny of the extant hexapod orders
Cladistics
(2001) - et al.
Phylogenetic systematics of the Potamanthidae (Ephemeroptera)
Transactions of the American Entomological Society
(1991) Arthropod Phylogeny With Special Reference to Insects
(1979)The limits of amino acid sequence data in angiosperm phylogenetic reconstruction
Evolution
(1988)