Friday, September 4, 2009

Thoughts on ESEB and the Morphometrics Workshop

I returned from the ESEB meetings, plus a follow-up workshop on Geometric Morphometrics, last night. Boy am I tired!! I'm on my 3rd cup of coffee this morning and still feel sleepy; more than one good night's sleep may be needed for a full recovery.

I thought the ESEB meeting was outstanding!!! There were lots of great talks, and I got to meet several evolutionary heros of mine, such as Russell Lande and Steve Arnold. Both were present at my talk, and when Lande asked me a question I nearly shat my britches in front of everyone. I don't think anyone noticed. One honest Swede (who will remain anonymous here) told me that my talk "was good, but not as good as your others," an assessment that I can live with. It was my first talk on entirely new material, so I forgive myself; I look forward to a 2nd opportunity! =-) Erik gave a great, wide-ranging talk on fitness trade-offs between fitness related traits. Some of his slides didn't work (the entire meeting suffered from technical misfires), but he was able to recover and present his messages clearly. I particularly appreciated his incorporation of an explicit adaptive landscapes perspective. The talk was also an interesting contrast with Roff's talk (which I also really enjoyed), which was right before Erik's and was a much more traditional quantitative genetic perspective on the same topic (not so wide ranging as Erik's approach).

I am biased I suppose, but I also LOVED LOVED LOVED the talks by Ryan Calsbeek and Robert Cox -- both on intralocus sexual conflict. A year ago I barely knew that this even was; now I want to study it (by the way, Erik, Barry Sinervo, and Andrew McAdam have a super paper in the "Early View" page in Evolution right now on this very topic). Ryan is my next postdoctoral advisor, and Bob is a postdoc in their lab. They both reported on their studies of Anolis lizards. In a nutshell, there is intralocus sexual conflict on body size -- that is, females and males have different optimal body sizes. How is it resolved (or is it?)? In part, it seems that females use cryptic female choice, and sort sperm from large males to make sons, and sperm from small males to make daughters. This maximizes the fitness of offspring. What made the talks great? First, they were intellectually rigorous (I learned a lot). This is the most important part. Second, the slides were perfect. This matters more than some people realize, and is hard to do, too. Every slide was readable (no axes labeled with tiny text), not overproduced, and informative. Third, they are both great speakers, with confident, engaging styles. I'd like to respectfully request that Bob not apply for any jobs that I want. =-)

Finally, Ian Dworkin gave a very cool talk on the last day (I stole a picture from his web page for this blog post). He is basically doing the same project I am (looking at selection on wing shape by predators), but using lab experiments with Drosophila and mantids. He has an amazing study system going here, I think, and can do things Erik and I can't do, such as concordant artificial selection experiments and studies of the genetics of development. I am certain to be citing his papers a lot. I got to speak with Ian, too, and he's a cool dude.

After ESEB I attended a workshop on geometric morphometrics by Chris Klingenberg and Andrea Cardini. It was a great experience for me. Chris and Andrea are both in the "cool dude" category, too. We learned how to use the program MorphoJ, which Chris created. It's an amazing, integrative program that does some analyses no other program does (I have lots of new ideas now!). Both Chris and Andrea are real experts on morphometrics, and it was a great opportunity to have a chance to interact with them and learn from them. That is one of my principles: never pass on the chance to learn from some of the best. We had some great discussion about semilandmarks, which both Chris and Andrea view with skepticism (semilandmarks are placed on edges and outlines and can mathematically slide along along the edge; they do not identify homologous points). 13 of my 16 landmarks are semilandmarks. I believe I convinced them that they work very well for my project, but Chris and Andrea have an excellent point that we must all remember: just because you can mathematically accomplish a task doesn't mean that it has biological relevance. Care is needed. In the case of semilandmarks, they become meaningless when structures twist (for example, during ontogeny) or when one has, for example, pumps 0n the outline with no linking homology. I have no such issues, but these problems are very common. I like my use of semilandmarks because it allows me to look at total shape. Several studies of insect wings have many good landmarks but few (or none) of them are on the wing outline, so that overall wing shape is not quantified, only patterns of homology within the wing. These could in principle change without a change in overall shape.

So, that's that! Two very good meetings, one very full brain.


  1. I agree - Calsbeek's & Cox's talk were GREAT!

  2. Thx for a superinteresting bloggpost with personal reflections that largely coincide with my impressions from this great meeting!