26 December 2011

Sand crab shrinkage?

I spent the day in my office working on two talks I have to give next month. One is for the Subtropical Biology meeting we are hosting (13 January), and one is a public talk at the South Padre Island Birding Center (28 January).

It was lovely to just have the day to think about how to do these talks. Nobody else around. No interruptions. Just a chance to think about how to explain the science in a (hopefully) engaging way.

And along the way, I solved a puzzle that lets me fix a 25 year old error in the scientific literature. Even more satisfying.

ResearchBlogging.orgThe talks I am giving are both about one of the local species of sand crabs, Lepidopa benedicti. There’s not a lot of papers on any sand crab, including this species. But there is a nice paper that describes how the little ones grow and develop, which is relevant to my talks.

Sand crabs, like many arthropods, have a larval stage. The adult sand crabs are dedicated diggers, and are rarely seen above sand, but the babies (zoea) are adrift in the waves. They spend the first few weeks of their life as plankton, up in the water. This might give them a chance to drift away, so that populations can colonize brave new beaches, where no sand crab has dug before. Because they are tiny and tasty, many crustacean larvae have spines of one sort or another to try to keep themselves out of the mouths of predators.

Then, the larvae undergo metamorphosis into a stage the looks much more like the adult (the megalopa) and they settle out into the sand.

I wanted to put all the different larval stages together on a single slide, all to the same scale, so that people could see the growth. But when I grabbed the pictures and rescaled them all to a single size, I got this:

I was confused. Why was the second stage smaller than the first one? And the third stage was still about the same size as the first. Typically, animals, you know, grow. Not shrink.

My first thought was I’d made a mistake. Rescaling things can be tricky, with lots of multiplications and divisions, especially since these all had different size scale bars (one was a bar for 1 mm, another was for 1.5 mm, and so on). (This led to some slightly goofy efforts of me holding up rulers to my computer screen to try to measure the sizes of the pictures ad scale bars from the PDF.) But no, it didn’t seem to be a math mistake on my part.

I went back to the text, and found there was something weird going on. Stuck and Truesdale had measurements of the size listed in the text, and they weren’t in line with the pictures. Even accounting for the fact that they probably didn’t draw an “average” larvae, there was no way they could be right: the most of the stages were about half the size in the pictures that the text said they should be!

Based on the measurements in the text, the sizes should be more like this:

Ahhhhh. That makes more sense.

Besides fixing this particular mistake, there are bigger lessons for making graphics. This might have been caught if:

  • Images weren’t spread over several pages.
  • A single scale bar wasn’t doing triple or quadruple duty on a single figure.
  • All figures were using the same scale (like 1 mm), instead of switching from figure to figure (1.5, mm, 2.4 mm, 1 mm).

Now I have my slide, and I have, in a very small way, perhaps helped there be just a little less error in the world. Even though it may not be a big deal, it’s still satisfying. Like pushing that jutting book back into place on the shelf.


Stuck KC, Truesdale FM. 1986. Larval and early postlarval development of Lepidopa benedicti Schmitt, 1935 (Anomura: Albuneidae) reared in the laboratory. Journal of Crustacean Biology 6(1): 89-110. DOI: 10.2307/1547933

No comments: