We had our first week in the Discovering Insect Species course (see this post for an introduction to the course), and it was a good start.
Tuesday
Tuesday’s session was a gentle introduction species delimitation. I wanted to get the students thinking about the sorts of data that are used to delimit species, and to try it out in some examples. This is a “full immersion” class, and so a lot of the context and background will come in the midst of our experiences, and not explained up front. For example, we didn’t at all talk about what we mean by “species”, and instead just dove right in to experiencing properties of species. (We will have an explicit discussion of the meaning of “species” later.)
After we took care of the administrative details, we gave half the students 14 laminated, numbered cards, each of which had a picture of a tamarin, and other half 13 cards with different tamarin pictures. Here’s what the full set looked like:
Each group’s task was to group the specimens depicted in the photographs into species. Here’s one group discussing the visual evidence provided by the photographs, evidence that might allow them to group the specimens into species.
Once each group was done, the two groups got together, and discussed how they might merge their proposed species. Once they had come up with proposed groupings, we talked about the evidence they used, and what evidence they might ideally have to make the decisions. I then revealed how the specimens they examined are viewed by biologists. The class’s results exactly matched those of biologists.
I then talked a little bit about the taxonomic history of the tamarin species they sorted. The earliest species to be described was Sanguinus midas, which was originally described by Linnaeus (1758) as Simia midas. I showed the class the original description:
I told them how Linnaeus’s 1758 work was the starting point of zoological nomenclature, and the species names proposed there represent the first animal species to be described using modern names (with the exception of the spider names in Clerck 1757). I showed them the very first species to be described in Linnaeus 1758, with the simple description “Know Thyself”:
We then switched over to carabids. After I introduced the class to carabids in general and Bembidion in particular, we did a similar card-based exercise with Bembidion subgenus Liocosmius. This is a subgenus that I revised recently with Ken Cooper. The cards for these specimens each had a photograph of the beetle, as well as a picture of the genitalia (all photographed specimens were males), and the voucher number:
The class attempted to reached a consensus about the species represented among the cards, but the species boundaries weren’t nearly as obvious as it was for the tamarins. I asked if they would like any more data for the specimens, and it was eventually decided that clearer views of the genitalic structures would help. So I then gave them accessory cards with better views of the internal structures:
This definitely helped, and they were able to make more progress.
I then gave them more data in the form of images of gene trees of four genes. Here’s what the tree of COI looked like:
They matched the voucher numbers one the cards with those in the trees, and looked at the correlations between the tree shape and the morphological characters. We talked for quite a while about the meaning of the tree diagram and the nature of gene trees.
After they had looked at all four gene trees, and the cards again, I showed them the four gene trees again, this time labelled to reveal the conclusions I came to in my 2014 study of the group.
We ended this session by looking a specimens of Bembidion (Trepanedoris) under the microscope. We also looked at other species of Bembidion that we would likely see on our field trip on Thursday, in particular Bembidion (Eupetedromus) incrematum and B. (Furcacampa) timidum.
The session included a lot of concepts, but I think it was a good start to this immersive experience.
Thursday
The class Thursday afternoon was all about seeing the beetles we will be researching, Bembidion subgenus Trepanedoris, in the field. We went to some wilder lands near the Willamette River in Corvallis, and managed to find quite a few Trepanedoris. We will find out next Tuesday exactly what we found, but at first glance it looks as if we found three species of Trepanedoris.
We had an excellent time, and all managed to do a great job sucking up the little beetles. Here we are (except for my grad student John, who is the photographer) proudly holding up our catches.
David Maddison
The Subulate Palpomere 
As soon as you folks get images of the habiti and median lobes made of the Trepanedoris species made I would love to see them. As you know, there are a number of species in the Washington+British Columbia region and both Hatch (1953) and Lindroth (1961-69) are incomplete.
I have been going through all my Trepanedoris specimens from the San Juan (WA) and Gulf (BC) islands and adjacent regions. Bembidion fortestriatum are common, but siticum, canadianum, connivens and elizabethae appear to be in some of the many samples too but at much lower numbers. Some of these species are obviously very good fliers, perhaps all of them.
One interesting characteristic about B. acutifrons that I have noticed after looking through thousands of pointed Trepanedoris specimens over the last week is that the males for some reason often conveniently project their median lobe during the process of being pickled in ethanol. I have noticed that Pterostichus (Pseudoferonina) do the same thing. I wonder why this is? Many carabid species that I regularly collect never seem to do it.
James Bergdahl
Conservation Biology Center
Spokane, WA, USA
James,
We will be making the images public in the next few weeks.
David
I forgot to mention that B. acutifrons is also fairly widespread in my San Juan & Gulf island samples, and it is found on a wide variety of stillwater wetland types, including brackish seashore marshes with some classic “salty” carabid species (e.g. Bembidion vile, B. indistinctum and Dyschirius integer). It is also regularly encountered on the adjacent mainland. My gesalt, based on habitus and experience with B. acutifrons in the field, suggests your tree based on DNA analysis will place this species as an outlier from the main group of species. Among what are now considered the Trepanedoris (Bousguet 2012) it seems to me this species is the easiest to identitfy, at least in my region of collecting.
I regularly pray to GOD that state of the art technology for determining phylogenies based on DNA will somehow be reflected in morphological attributes that are reasonably easy to observe on whole specimens with a decent dissecting scope. It is always important to keep in mind that whatever phylotree one builds, no matter what materials and techniques are employed, in the end they are all “just” hypotheses. Thomas Casey provides an good example. He is somewhat infamous among carabidologists for naming new species based on small morphological variation, and presumably a healthy dose of providence. (I keep a file on all the derogatory comments about Casey I come across in the literature.) Casey described many of the “valid” Trepanedoris species recognized today in North America (Bousquet 2012). I predict that the more sophisticated phylotrees based on DNA (barcodes) become, the more species and subspecies Casey described will be considered valid. If this proves to be the case, some very famous and respectable carabidologists, many of which are now dead, will have to “eat their words”. Will DNA tools for building phylogenies eventually re-erected Casey as a visionary/genius, not a crazy splitter? Casey is of course not the only “productive” taxonomist with such great skill for seeing new species and subspecies.
Note: I am not a “splitter”, and therefore have some reservation about using gene data to do so.
James Bergdahl
Conservation Biology Center
Spokane, WA, USA
James,
Regarding ” It is always important to keep in mind that whatever phylotree one builds, no matter what materials and techniques are employed, in the end they are all “just” hypotheses.”: I wouldn’t have had a successful career if I were naive about this.
Regarding Casey: I have noted in the past that Casey was correct more often than we had thought. That said, I don’t think he will ever be viewed as a visionary/genius. Even when all is said and done, there will be many of his names that are synonyms. Yes, he did detect valid differences that others failed to detect, but he also oversplit too much too. I think there was a mix there of insight and being correct just by the shotgun approach – if he described enough he would be bound to hit something.
Using DNA data is independent of being a “splitter” or a “lumper”: DNA data is simply data that provides evidence closer to what we need for species inference. But at least for those of us who are classically trained morphological systematists, we also thoroughly consider morphological data in any of these decisions about species boundaries. As with any scientific study, naive or dogmatic interpretation of limited data can lead to poor inferences; thorough, careful, and skeptical interpretation of more data can lead to better inferences.
Pingback: Discovering Insect Species: hands-on with Trepanedoris | The Subulate Palpomere
Pingback: Discovering Insect Species: Overview in Rearview | The Subulate Palpomere