Thursday, March 19, 2015

Adopting Orphaned Taxa (TAD2015).

(For Taxonomist Appreciation Day 2015)

I know it’s been a while since my last blog post. I’m deep into research right now and about to defend my thesis proposal. I’m also working on a publication. Which means that my writing time is going elsewhere and not here.

My thesis research overall concerns a large and varied tribe of tachinid flies called Blondeliini (Blond-el-ee-ai-nai) or the blondeliines. The core of the work is the Blondelia group of genera, called such because it includes the type genus of the tribe, Blondelia. Females of the Blondelia group have a boat-like keel on the abdomen and a sharp piercing hook for poking holes in things, usually caterpillars.

The piercer isn’t an “ovipositor” in the homologous sense, because it doesn’t contain the tube that carries the egg into the host. Instead, the egg tube (the mostly membranous segments 8-10 of the abdomen) travels down the posterior groove in the piercer and into the hole the piercer has made in the host.


A female 'sword fly' of the genus Eucelatoria with hind legs removed. (1992: Mexico, Portillo de Reon.)

In one genus (Eucelatoria), the piercer can be half the body length! I’m not sure why these species have a piercer that long, but there’s some evidence they parasitize caterpillars hidden in rolls of leaves. Some Blondelia group species have spines on the ventral keel, and others have only bristles. Some males of the Blondelia group have hairy patches on their abdomens, and other closely related species are clean shaven. Host use varies; Costa Rican species of the Eucelatoria armigera complex are particular to one or a few species of noctuid moth caterpillar, while the polyphagous species Compsilura concinnata feeds on over 200 species of Lepidoptera.

With interesting oviposition behavior, morphology, and a large number of species (>135; not including all the many undescribed Neotropical species) the Blondelia group is an enticing project for a young taxonomist. Do not fall for this trap!

Orphaned taxa are those genera or families that are without a current expert or active worker. The Blondelia group, and Blondeliini in general, are a particularly frightening example. Abandonment can be for a number of reasons. In some cases the group isn’t charismatic enough, or is of minimal economic importance. In other cases high diversity and difficult diagnosis are deterrents. A history of poor descriptions and over-splitting genera may be to blame; for this final reason orphaned taxa often have a taxonomic impediment. The longer the period between experts, the greater the impediment to future research becomes.

In the case of the Blondelia group, our good friend Dr. Townsend is mostly to blame. He is responsible for naming nearly half of all valid blondeliine genera, and most of these with one species apiece. Add to this his notorious over-splitting, his mediocre descriptions, and his terrible, no good, very bad Manual of Myiology genus key, and very few people are courageous enough to venture forth.

However, not all blame can be placed on Townsend. Disregarding the history, blondeliines are a difficult group with many examples of morphological convergence. They are small, usually dark colored, and told apart mostly by arrangements of bristles. Color patterns often fool me.


Left: Sword fly male. Right: NOT sword fly male. Really similar, but really different. Can you see the difference? (Click for embiggen)


I thank Monty Wood for his great work on Blondeliini (1981), without which I would be lost. But this is a preliminary work of broad scope. Efforts focused on a single or a few genera have revealed the scale of the mess yet to be resolved.

Diego Inclan (a graduated MS from my lab) and Dr. John Stireman (my PI and advisor) provide a vexing example of this mess in their recent Zookeys paper. In his masters thesis, Diego found that some species considered part of the Neotropical blondeliine genus Erythromelana were clearly not. This lead to a convoluted taxonomic investigation. Below is a paragraph from the Zookeys paper as illustration.


“An example of the taxonomic instability of Neotropical tachinid genera is witnessed in the species Euptilodegeeria obumbrata (Wulp). This species was first classified in the former tachinid genus Hypostena by Wulp (1890; along with many other blondeliines), based on specimens collected in Guerrero, (southwest) Mexico. […] The species was moved by Townsend (1931) to the new genus Euptilodegeeria, moved again to the genus Erythromelana Townsend by Wood (1985) and recently excluded from Erythromelana and resurrected to its previous genus (Euptilodegeeria) by Inclán and Stireman (2013). Although the taxonomy of Tachinidae, particularly of the Blondeliini, is challenging due to the scarcity of clear synapomorphies, the confusion in the generic assignment of E. obumbrata was also due to the limited number of specimens evaluated, the lack of examination of male terminalia and the use of only males for the descriptions. In the present study, we use additional information from male and female terminalia to demonstrate that these “obumbrata” specimens, previously assigned to Hypostena, Euptilodegeeria and Erythromelana, actually belong to the genus Eucelatoria Townsend (1909), in which females possess a sharp piercer for internal oviposition in the host. We also argue that the former species Machairomasicera carinata described from a single female by Townsend (1919) in the monotypic genus Machairomasicera, and later synonymized with Eucelatoria by Wood (1985), belongs to this same species group of Eucelatoria, which we here define and characterize. In the end, taxa that were assigned to four different genera in fact belong to one species group of Eucelatoria, providing an example of the taxonomic confusion that plagues many groups of Neotropical tachinids.” [Emphasis mine]

Many similar issues remain in the genus Eucelatoria. The group may not even be monophyletic. I am not revising all the species in the Blondelia group for my dissertation—or even all the species in Eucelatoria—but the challenge feels insurmountable.

There are two ways to publish natural history research. One is to be cautious, to wait until all possible evidence is covered and carefully recorded, all the museums have been visited, and every last lead has been pursued. “I only have one more type to look at, and it’s been missing for 40 years. But I can’t publish until I find it.” The other is to rush wildly into publication with any new finding, getting the information out as quickly as possible. “Never mind the types in that European museum, I have the specimens here and there’s nothing (well) written in the literature to say I’m wrong!” 

Taxonomists, myself included, fall more on the cautious side. Townsend was an exception. We want all the bits of evidence before we publish our scientific opinions, whether that be new species, synonyms, homonyms, or redescriptions. Caution is great when you start with a clean slate. But in the face of a huge mess caution is paralyzing. How do I start? I’m looking at a great wreck of a building. Do I take the debris out piece by piece and slowly repair? Or do I knock it down, bulldozer the area, and pour a new foundation?

I have sat and looked and sat and looked and wondered at my specimens guessing and second guessing myself if what I am seeing is really separate species, or if they have been previously described. This back and forth mental motion is useless. I fear too much of wreaking havoc. But plenty havoc has already been wrought.

I think there is a middle ground. Stride boldly, but document everything. Don’t worry too much about creating new species synonyms or mis-associating males and females. Those issues can easily be fixed later. Otherwise you’ll spend the rest of your life waiting for that visit to that one university in Chile (when the type was long since moved to a museum elsewhere). At the same time, document everything and carefully record your findings. If you provide photographs, written description, genitalia drawings, and adequate references to collections and literature in your publications, someone else can build upon this firm basis and correct your mistakes.

An excellent example of walking this line is Dr. Lee Herman’s 2013 revision of the New World species of Oedichirus, a genus of rove beetles (Staphylinidae). Dr. Herman, a Curator Emeritus at the American Museum of Natural History, received the “J. O. Westwood Medal and Award for Insect Taxonomy” [PDF] for this publication. Rove beetles have a taxonomic history as equally tortured as tachinid flies. As in tachinids, associating males and females of the same species is difficult. At times only male or female specimens are available, and species were described sometimes based on the male and sometime based on the female. Furthermore, the majority of specimens available (including the types) are too old for modern techinques like DNA Barcoding. Herman could have waited until new specimens were available, but instead he pushes forward. In the methods section he writes, “hypotheses of male/female association proposed herein for the other species can be corroborated or refuted by DNA barcoding techniques using newly collected specimens.”

Remember that every “opinion” of natural history is a hypothesis subject to further testing. When our hypotheses are presented as expert opinion but rest on shoddy work they are an obstacle. When we refuse to present hypotheses for fear of being wrong they are also an obstacle. But when our hypotheses are presented boldly and rest on good work, even our mistakes are outweighed by the scientific contribution. I have discovered that it does no good to worry. Any well documented progress is good progress. Anything mopped up is better than the mess we have now.

3 comments:

Dave said...

I like it

Gunnar said...

Working on a taxon with the taxonomic history dominated by over- or undersplitters, I agree very much with that middle ground!
A less scientific fact I find notable, is that the name Blondelia makes the mind drift towards certain people based on very superficial criteria. I wonder if these criteria affect patronymy desicions? In Psychodidae there is another genus with similar connotations, namely Brunettia...

Kai said...

As far as I know, there are no patronyms in the genus Blondelia.