CURM Twenty-fourth Meeting, 3/5/2009
"The world is full of mysteries, and I love mysteries." -- Freeman J. Dyson
"I never know, when that phone rings, what my next mystery might be." -- Medical Examiner Daniel Spitz.
- KYMAA Meeting, March 27 - 28, 2009 at Kentucky State University in Frankfort, Kentucky.
- You've been offered the opportunity to present by Will Faris, who runs the show. The question is "One talk, or two?"
- The answer would appear to be two talks, early (8:20) on Saturday morning.
- In new developments,
(generated by simulate_st_johns_pop.lsp).
- Grayson is at work on slide one, giving an overview of the project, the dogmas, the biological facts, etc. A picture of "what's to come". He's going to add a "title slide" to have up at the beginning of the show, including author names (and mine should appear to, upon consideration, as "Advisor: Andy Long"). Things we discussed today:
- Keep the "horror" in at the outset; make sure that we know that it's the females who are the killers.
- Make sure we have a picture of the male and female wasps from Florida.
- We want to recast that second mystery, as how to get a mass-provisioning model that will allow three different sized cicada populations to produce a single male wasp size.
- Katie is at work on slide two, which shoots down the dogma of opportunistic hunting.
- We need to have all the plots on the same platform -- the data won't change over the course of the images, so changing the format would simply distract. I suggest R.
- Keep the technical vocabulary down, emphasizing the concepts in simple words.
- Grayson is working on slide three, the Florida slide, which is supposed to be a cliff-hanger: it will leave us at the half-way mark with a lead-in to the fundamental mystery (and principal stake in the heart of the one-if-by-male dogma): how is it that with three populations of cicadas of radically different sizes, we don't see three radically different sized male wasps? Mass provisioning could explain the maintenance of two distinct sizes of wasps in Florida, if the wasp female population is ever truncated by some event to eliminate the largest females.
- Katie is at work on slide four, which addresses issues of allometry: as we prepare to create a model of mass-provisioning (rather than the one-if-by-male, two-if-by-female sex-allocation theory), we need to move between RWL and mass for cicadas, and between mass and RWL for wasps. In between those two processes, we need to transform the cicada mass into wasp mass, which we believe occurs as 1/4 of the mass of the cicada-meal (laboratory results).
- Grayson is working on slide 5, the "mass provisioning solutions" slide. In conjunction with this, I show the following results of a simulation which illustrates that if a population of wasps is truncated so that they can't access the (generated by simulate_pop.lsp).
In this case the populations I used had 100 wasp females, a heft of 2.2, and a cicada population of mean 1000 in mass, and 200 in SD. One thing to notice is that the steady state mass of the wasps is below half the mass of the cicadas....This shows that, if the Newberry, Florida wasps had been truncated so that the TB populations were inaccessible, then the population of wasps might have fallen back to smaller sizes related the TG population. (which would explain the mystery of the two-size-wasp-populations)
When Grayson and I discussed this yesterday, we had the idea of focusing on the directed graph showing female wasps "recreating themselves (and their spouses)", and the idea of "sweet spots" in the stair-step hunting curve (we need names for all our recycled curves, and stair-step seems like a good one for that one)..
- Katie is working on the summary slide. No doubt she is waiting for us to decide what it is that we want to summarize!;)
- Grayson and I got onto the idea of looking at wasp masses, and looking at the cicada distributions, to see if a 1-if-by strategy was even possible. Then we talked about the relative masses of male and female wasps. I ran a little regression program which produced estimates of the best multiplier (e.g. factor of 2?) which would transform the male sample cdf into the female cdf. The results were as follows: