[Odonata-l] Odonata phylogeny

[creagrus]

 In today's postings, Roy Beckemeyer helpfully provided a web link to a 
paper by Hasegawa & Kasuya (2006) that discusses odonate phylogeny. As 
it happens, I had read this paper last week, along with others on this 
general topic, and found it interesting to observe the progress underway 
in this area.

My background is in field ornithology and world birding. I have a 
website that features Bird Families of the World 
[http://montereybay.com/creagrus/list.html  ] and I have been an avid 
reader of taxonomic papers, as well as a correspondent with many of the 
professional ornithologists working in that field.  In ornithology, for 
example, a set of new papers signal a major advance in understanding 
bird phylogeny as it relates to the traditional "Old World Warblers," an 
assemblage of ~700 species that now proves to be paraphyletic. Indeed, 
the "Sylvidae" is actually composed of 8-10 separate lineages. I have a 
3-web page set on the topic of the "Breakup of the Old World Warblers" 
that summarizes much of the new work, with a host of citations, 
beginning at http://montereybay.com/creagrus/sylvid-intro.html

To comment briefly on odonate phylogeny, let me digress slightly to 
offer a very simplified overview of progress in the avian world. For 
many years, work on avian phylogeny was based on studies of morphology, 
internal structures, behavior, voice, parasites, and numerous other 
lines of evidence. In the last couple of decades of the 20th century, 
the field of modern cladistic studies, based on this type of evidence, 
dominated the literature. In the 1980s, Charles Sibley and colleagues 
used an indirect method for recovering molecular evidence [measuring 
'melting curves' on lab-created DNA-DNA hybridization strands] that led 
to influential papers in 1990 that upended many classical traditions 
about avian evolution (Sibley &Ahlquist 1990, Sibley & Monroe 1990). 
Many of their findings have proved prescient but others have not been 
supported by later studies. In upsetting the apple cart, they got only 
part of it right. By the late 1990s, studies based on mitochondrial DNA 
became the rage, but these, too, had limitations. While very useful to 
test hypotheses of relationships between closely-related taxa, they 
proved less helpful in looking at larger issues at the family or order 
level, with statistical problems like 'long branch attraction' giving 
misleading 'positives' at unexpected places. Recently, however, the use 
of the sequences of nuclear DNA has proven much more reliable. A growing 
consensus is that nuclear DNA may hold the real key to direct evidence 
of avian relationships. Progress has been unusually rapid in the last 
couple of years, and many suspect that within, say, 10 years, all the 
major issues related to phylogeny may be sorted out. My web pages have 
tried to stay abreast of this progress -- some of which is quite 
surprising, yet quite satisfying. The most satisfying of all the studies 
have been those that use a variety of evidence -- multiple nuclear DNA 
genes, mitochondrial studies, cladistic work on morphology, etc -- that 
support the same phylogeny at a high degree of confidence.

I have no background in the Odonata literature, so these comments may 
reflect some ignorance. As I understand it, it has been generally 
thought that there are three major suborders of the order Odonata: the 
Zygoptera [damselflies], the Anisoptera [dragonflies], and the 
Anisozygoptera [two species of Epiophlebia in Japan and the eastern 
Himalayas]. It has generally been thought that the Anisozygoptera is a 
basal lineage, the two species are often referred to as "living 
fossils"; e.g., Silsby (2001). The recent genetic analysis by Hasegawa & 
Kasuya (2006), based on nuclear DNA, was intended to consider the 
placement of the Anisozygoptera. Their evidence placed that suborder 
between the other two suborders, consistent with physical characters 
that are "half-damselfly, half-dragonfly." [Much of the paper discusses 
the comparative value of two specific genes for recovering odonata 
phylogeny, rather than the phylogeny itself.]

What I found particularly interesting in Hasegawa & Kasuya (2006) is 
that a Japanese species of spreadwing (Lestes japonicus) also falls 
intermediate between damselflies and dragonflies, and rather close to 
Epiophlebia. The authors do not discuss this finding in any detail, and 
do not express an opinion as to which suborder the spreadwings might 
belong.  This finding, however, is similar with phylogenetic analysis of 
Saux et al. (2003), who looked at mitochondrial DNA evidence. They found 
that spreadwings [family Lestidae] was more closely related to 
dragonflies than to damselflies. The damselflies [suborder Zygoptera] 
are a paraphyletic group until the Lestidae are removed; at that point 
the molecular evidence finds them to be a monophyletic clade. The 
dragonflies are a monophyletic clade, whether of not Lestidae is 
included within the Anisoptera. Saux et al. (2003) did not analyze 
Epiophlebia, so again we do not know whether the spreadwings are (a) 
part of the dragonflies, (b) part of the Anisozygoptera that is 
otherwise restricted to Asia, or (c) should be placed in their own 
suborder. What they found, however, is that spreadwings are not typical 
damselflies.

In avian phylogeny, nuclear DNA evidence is often better to determine 
family or order level relationships than is mitochondrial DNA evidence. 
Therefore, the results of Hasegawa & Kasuya (2006) are an unexpected and 
quite interesting confirmation of what mitochondrial DNA evidence had 
suggested: that spreadwings are not damselflies. This is contrary to 
cladistic studies of morphology, wing venation, and larvae (e.g., the 
extensive work of Rehn 2003). In the avian world, it appears to me that 
properly constructed nuclear DNA studies have overwhelmed even the best 
cladistic work based solely on
morphology. Of course, studies that use both lines of evidence produce 
the most satisfying results. Quite by accident, today I came across 
reference to Bybee, Ogden, Rehn & Whiting, in prep., "Phylogeny of 
Odonata (Dragonflies and Damselflies) based on molecular evidence." 
Perhaps they are working on such a broad-scale multi-dimensional approach?

The two published papers using mitochondrial and nuclear DNA would not 
seem to be enough to change the classical phylogeny yet, but perhaps the 
handwriting is on the wall. Perhaps it is time for us to consider the 
real possibility that spreadwings are not damsels, and (possibly) adjust 
our checklists accordingly?

Beyond the Lestidae, there is the question of whether the same results 
would apply to all the lestoids in the world. Other questions are 
apparently not yet settled. Are gomphids or aeshnids more basal? Do 
cordulegastrids belong with libellulids, as the Saux paper concluded? 
[The Hasegawa paper hints otherwise.] Dennis Paulson, in litt., points 
out that all the studies so far are holarctic-centric, while the bulk of 
Odonata is tropical. So there is a lot yet to be learned. Indeed, we may 
be comparatively early in the analysis.

In looking through the archives and in general chit-chat, I haven't seen 
much talk on this stuff. I find it fascinating and await the newest 
developments. Does anyone else? How much of our local checklists should 
we be ready to revise? In the bird world, we are just now getting used 
to putting ducks & geese first, then qrouse & quail, before the 
traditional stuff like loons, grebes, and tubenoses.  In the ode world, 
should we be preparing to think of spreadwings -- and perhaps others -- 
in a different way?

Literature cited:
Hasegawa, E., and E. Kasuya. 2006. Phylogenetic analysis of the insect 
order Odonata using 28S and 16S rDNA sequences: a comparison between 
data sets with different evolutionary rates. Entomological Science 9: 
55-66.
Rehn, A.C. 2003. Phylogenetic analysis of higher-level relationships of 
Odonata. Systematic Entomology 28: 181-239.
Saux, C., C. Simon, and G.S. Spicer. 2003. Phylogeny of the dragonfly 
and damselfly Order Odonata as inferred by mitochondria 12S ribosomal 
RNA sequences. Ann. Entomol. Soc. Am. 96: 693-699.
Silsby, J. 2001. Dragonflies of the World. Smithsonian Instit. Press, 
Washington, D.C.