The shrinkage effect
Imagine an experiment. Take an iceberg lettuce, Lactuca sativa, and leave it in a warm oven for a few days. What will happen? The outcome is not surprising. It shrinks by a very large amount. Who cares? I’ve been thinking about this ‘shrinkage effect’ for some time now, and it turns out that it matters in some important ways for ecology and climate research. In this post I want to take you through the story of how an chance observation can become a scientific publication.
The story begins with some fieldwork in the mountains of Colorado, measuring the size of leaves. We noticed that they get a lot smaller when they dried out in the air, which was a problem for our work. Below you can see the leaves of Senecio soldanella, an alpine herb.
We informally called this a shrinkage problem. But beyond the fun name (check urbandictionary if you need a definition) – it seemed as though this effect could be important. Leaf area is used for a range of different analyses in ecology, including models of plant resource use or ecosystem carbon fluxes, and also in reconstructions of past climate based on fossils. If measurements of leaf area were always biased, because of this shrinkage, then so would be the conclusions of many studies. The reason is that often studies rely on dried leaf specimens stored in museum collections. Below you can see a collection of a Dubautia species from Hawai’i.
So we decided to see how big the shrinkage effect was, all around the world. We measured leaves from lowland tropics through to the alpine temperate zone. This wasn’t so difficult, because we already had expeditions planned to all these places, and measuring shrinkage is easy (all you need is a digital scanner). We found that leaves shrink in area by about twenty percent, and up to eighty percent. This is a very large bias – much larger than we ever expected!
So we wanted to understand what makes leaves shrink, to be able to correct for this bias. After some adventures with statistics, we found that a range of things predict how much leaves shrink – for example, whether or not a species is woody, whether or not its leaves are succulent.
But we also wanted to know if fossils would also suffer from shrinkage, or if we could actually ‘unshrink’ leaves instead of relying on statistical corrections. This required some experimentation – and I was fortunate enough to know just the right group of investigators.
Last year I was based in a middle-school science classroom in Tucson, Arizona. We were studying ecology and considering a large inquiry-based project for all the classes – so this was a perfect opportunity to do some real science with over 100 new scientists. We worked through a series of experiments that simulated fossilization or tried to unshrink a leaf. Three weeks of classroom time including practice sessions and lab-report writing resulted in some surprising results. We all found that fossil leaves may not shrink, and that any shrinkage can be reversed by rehydrating a leaf. A very helpful conclusion for scientists worldwide, thanks to our middle-school collaborators.
In the end we wrote a peer-reviewed scientific article on this study – and over forty middle school students who did good jobs with their lab reports became published authors before they ever made it to high school!
I’m very proud of this project – it’s a great example of when a moment of idle curiosity can lead to new science, and when the joy of discovery can be shared with a new generation of scientists!