The Georgia Southern University-Savannah Science Museum Herpetology Collection is something of a liminal space. The gently buzzing lights, the rows and rows of dead snakes and frogs preserved in alcohol and formalin, the unadorned concrete walls and floors, and the utter lack of any other living things in a hall full of bodies make it an oddly calming place to sit and study. When running through flashcards or gazing through a magnifying glass looking for scars in the scales of a snake that died in 1972, the collection room lends everything a feeling of gravity and peace. I worked in this collection for three years as an undergrad at Georgia Southern, changing fluids, preparing new specimens, and fighting with our database. In that time I also used some of the collection’s residents to write my Honors thesis, which eventually became the paper published in the Journal of Zoology.
The GSU collection holds over 35,000 individual reptiles and amphibians, almost 400 of which are rough earth snakes (Haldea striatula), unassuming little non-venomous snakes that due to their small size and nearly uniform brown coloration look like what might result if a child’s clay model of a snake came to life (that will become important later). In life, they’re secretive and fossorial – so much so that despite ample evidence that their range includes the areas I’ve lived in for twenty three years, I’ve never seen one in the wild – but what ultimately became of interest to us was the predation risk faced by these small, largely defenseless snakes.
When I first decided to work with rough earth snakes for my Honor’s thesis under the mentorship of Dr. Christian Cox, an assistant professor at Georgia Southern University, our project had nothing to do with predation risk. Our initial idea was to use them to study the reproductive strategies of small, live-bearing snakes. Which would be sacrificed in lean years, the number of babies or their size? Both? To what degree?
The first step in answering these questions was to retreat into the solitude of the herpetology collection, where we examined all 398 rough earth snakes that had been preserved over the years, noting over a dozen morphological traits. The plan was to focus on the pregnant females – by measuring their size and weight and comparing those to the number of embryos they carried, we could establish a foothold on the topic and move forward into experimentation.
Unfortunately the collection had other ideas. Less than ten of the snakes were carrying young, not nearly enough to draw any conclusions, even highly speculative ones. But I did notice a different pattern. A number of our snakes were carrying completely different but equally interesting set of data in their scales and tails that spun us off into a new line of inquiry. Our snakes had scars.
Many had old wounds evident in their scales, and more had tails blunted by some kind of attack. Only a handful of our snakes were pregnant, but enough of them had evidence of attack by some sort of predator that we quickly refocused to determine whether there was any significant pattern to the scarring. There was. Larger snakes were statistically more likely to have scars or blunted tails than smaller snakes. This was interesting. Received wisdom is that the smaller an animal is, the more vulnerable it is to predators, and the more likely it is to have wounds, and this pattern appeared to run counter to that wisdom.
Of course, there were other potential explanations. Perhaps the scars were obtained in the snakes’ youth, with adults simply accumulating them over time. Perhaps the survivorship bias inherent in a herpetology collection (to make it to me, a snake has to not be eaten by the predator that attacks it) was simply excluding the small snakes. To sort out what was driving this pattern in the collection, we had to move into the field, by which I mean a wooded area of South Georgia.
This is where rough earth snakes resembling some sort of platonic ideal of “snake” came in handy. In order to standardize exposure time to predators, I dug deep into my kindergarten craft skills and set to work making scores of little brown snakes out of clay, a process that would have been much more grueling if we were working with a more colorful species. In the end we had three sizes of clay snakes that roughly equated to a newborn, juvenile, and adult in length and girth. Each model spent five days in the woods, subject to attack by any predators that might be present. Based on the bite marks on the recovered models, those predators were mostly rodents and small birds, although one model might have been bitten by an armadillo, and a fair few were so heavily damaged by the attack that we weren’t able to make heads or tails of it (in large part because the attacker had removed the model’s head and tail).
Once we’d analyzed the data from the models, the pattern was clear. Even over the same time-frame, in the same environment, larger models were attacked with a significantly greater frequency than smaller models. Our initial, exciting hypothesis was supported. Larger rough earth snakes seem to actually draw more attacks than their young, running counter to the assumption that “bigger is better” when it comes to not being eaten. This result has a number of interesting implications, the most obvious of which is that predators may play an important role in the evolution of maximum body size in small-bodied snakes (an upper limit often attributed to food availability).
While it’s still unclear if this pattern is simply due to smaller snakes being harder to spot, larger snakes presenting more energetically tempting targets to the most common predators, or some as yet unknown mechanism, the fact remains that we ultimately did gain some insights into the evolution of this little snake’s life history traits, just not the reproductive ones we set out to find.
So go to a collection and study dead snakes, as they might actually hold some answers. And say “Hi” to the undergrad working the collection. We get a little lonely in there.