Semlitsch Talks about His Study on Salamander Abundance
Nov. 13, 2014
The abundance and biomass of salamanders in the Missouri Ozark forest ecosystem is the focus of a new study from the Semlitsch lab. Below is an extended interview with Dr. Semlitsch about the study, which was published in the 20 November issue of the Canadian Journal of Zoology.
What motivated you to do this study?
Back in 1975, Thomas Burton and Gene Likens published two papers on the abundance of salamanders in the Hubbard Brook Experimental Forest in New Hampshire. One was in Ecology and one was in Copeia. These papers basically report the biomass of salamanders at Hubbard Brook and say that, of all the vertebrates, salamanders have more biomass — actual tissue — in these forest landscapes than any other vertebrate. These two papers are the top citations talking about the importance of amphibians in terms of biomass and ecosystem function.
I read these papers when I was a graduate student in 1975. But, recently I went back to reread these papers, and the abundance estimates that these papers are based on are very poor by modern standards, by standards in 2014. In 1975, that is all they had, and they were good. But I knew the work that my graduate student, Katie O’Donnell, was doing down in the Ozarks was, in my view, state-of-the-art in terms of abundance sampling and abundance modeling of salamanders. I talked to her, and I said, “Let’s use your abundance estimates to calculate biomass of Southern Redback Salamanders in the Ozarks and see how it compares to Burton and Likens’ estimates.” And so we did. We totaled up all the numbers, and then we sent the animals off to get nutrient analyses done. And, low and behold, our numbers were 2-4 times greater than Burton and Likens’ estimates. I sort of knew that, but this study confirmed it.
Then, I thought, let’s go in the literature and just see if they’re just super abundant in the Ozarks and nowhere else on the planet. So I went in the literature and looked at abundance estimates all over eastern North America where they occur and even some in the west. When you tally up all those estimates of abundance, the average is 10 times higher than Burton and Likens’ estimates. So that suggests to me that Burton and Likens’ estimates are either poor and low or New Hampshire is just the pauperite of salamanders. But the point is that the big deal they made out of salamanders being important in New Hampshire, in my view, is now accentuated 10-fold.
The ultimate reason for doing this study is that we have long asked these three related questions: Why are amphibians important? Why do we care if they survive? And why are we concerned about amphibian conservation at all? Well I would contend that if you have that much biomass and if that biomass is important in forest ecosystems and if you lose it, then something is going to go wrong. It goes back to this idea of the rivet theory in airplanes: if you start popping rivets out of the wings of an airplane, at some point the wings are going to fall off. Well, I would contend that salamanders represent a whole lot of rivets. If you lose all these salamanders in the Ozarks or at Hubbard Brook or in the Appalachians or in the Shenandoah Mountains, then something is going to happen. There’s no question in my mind, now, that salamanders are really important.
How is your method of estimating abundance better than Burton and Liken’s method?
There are basically two methods for estimating abundance. One is what would now be considered a fairly naive approach: you simply go out in the forest and you count salamanders that you encounter. You do that on a designated area, a plot. You dig through the leaf litter, you flip things over, and you count them, and then you say you have X number of salamanders per meter squared. That is what is called a naive count, and that is what Burton and Likens did.
The problem is that we know from behavioral studies and life history studies that only a certain percent — and that may be only 10% or 20% or 30% — of all the salamanders that actually live in forests are on the surface at any one time. The vast majority of them are underground. Given that, what you really have to do is one of two other things to get a more contemporary or sophisticated estimate of abundance. You have to go in and capture those animals, mark them, and then go out and do multiple recaptures. So, over time, you get animals that are marked and unmarked, and by using that in a population modeling formula, you can then determine the population that is above ground plus the population that is below ground. That is using mark-recapture to estimate abundance.
We did something that was even more contemporary, something called unmarked hierarchical modeling. So our animals were not marked, but what we did is intensive repeated surveys over time. By doing that you get an estimate of variation of the salamanders that are present on any one day compared to all other days that you census. Those numbers are put into a population model that takes into account how often you detect animals on those plots and how often you don’t. That’s called hierarchical modeling. The model takes into account detection and non-detection of animals. It also gives you a population estimate that takes into account surface-active animals and belowground-active animals without marking them.
Our population abundance models also took into account factors of the environment. Factors such as day of collection, time since last rainfall, month, year, slope, aspect, canopy cover, etc., were plugged into the model to then correct all those different aspects of what was out there on the day you collected those salamanders. Again, that’s what the model does and that’s what makes this so powerful. Now we can get estimates that are precise for a particular condition at a particular location in time. We could never do that before.
Moreover, because it’s predictive, you can then do the same thing for other areas around it that have not been directly measured for salamanders. Again, by knowing the slope, by knowing the canopy, by knowing other characteristics, you can plug those factors into this formula and estimate salamander abundance. We did that for a larger landscape to just get an estimate for a bigger landscape, just to show people how many salamanders were actually present.
And how many salamanders are present?
Based on our abundance model, we estimate that there are 1.88 – 2.65 billion salamanders, representing 1,485.2-2,093.5 metric tons of wet biomass, distributed across the Salem-Postosi District of the Mark Twain National Forest.
To put that in some context, I calculated that the biomass of white tailed deer in the Ozarks is about 3.95 kg/hectare. Our study shows that salamanders range from 5.77 –10.20 kg/ hectare!!!!
What is a good working definition of biomass?
Basically, meat on the hoof. So when you talk about biomass, you’re basically talking about food. If you go out and catch a fish, you catch a five pounds bass in your farm pond, that’s five pounds of meat that can be consumed by a predator or a decomposer.
Why did you chose to do this study in the Ozarks as opposed to, say, the Appalachian Mountains where there is the greatest diversity of salamander species?
The Appalachians do have the highest diversity in terms of species, but the Ozarks have a fairly high species diversity for the Midwest because it has a lot of unique habitat and it’s well protected and its unique. The Redbacked Salamander is the single dominant species in the Ozark forests. Sinkin is embedded in the Mark Twain National Forest. It’s representative of the Ozarks. That’s what makes it such a powerful site to work. This is an Ozark system and this is an Ozark result. I think we should be really proud of it. The Ozarks are a tremendous resource and salamanders are part of that system.
What role do salamanders play in forested ecosystems?
We know from studying salamanders in these forests that the salamanders live in the leaf litter and they live underground and they go back and forth. In the leaf litter, they eat all the little invertebrates that eat leaf litter. There are some studies that suggest that salamanders actually control populations of the invertebrates that consume leaf litter. Salamanders are part of that food chain. That’s what connects them to the forest. Then we know from a whole other group of studies that salamanders are then eaten by lots of other animals, from garter snakes, to ringneck snakes, to other salamanders, to blue jays, to raccoons, to wild turkeys. You name it — they eat the salamanders. Again, it links those predators to the salamanders to the forest litter invertebrates to the leaf litter, and that’s what contributes to this ecosystem function. That is the ecosystem function. And the nutrients that are transferred and the energy and the protein, all that, salamanders play an integral role in. We still don’t understand the full extent of that role, but the first thing we need to understand is how many are out there and what the biomass is and what they contain.
Are there any management/conservation implications of your results?
One of the things we found when we modeled salamander abundance and put in all the different habitat variables was that northeast facing slopes — which are the shadiest and hence the coolest and the moistest slopes — had more salamanders than south facing slopes and west facing slopes, which are hotter and drier. That’s just a matter of exposure to the sun. If you were out collecting salamanders, a good field herpetologist would know that you go to the moistest places, the shadiest places to find more salamanders. This absolutely confirms that it’s where they occur on the landscape.
For conservation and management, if we want to protect the greatest number of salamanders, the greatest biomass of salamanders, the greatest ecosystem function, then we want to conserve these slopes. If you wanted to protect an area and keep it as a “seed source” that could be used to repopulate areas after you harvest timber or you do some other kind of land management, then that would be the best area to protect. If you protect the other slopes, the south facing slopes, then you’re not going to have many animals and they’re going to be hot and dry. Then, when you cut the timber completely around them, they might dry out even more and you might lose everything.
What is the message you hope the average person walks away with from this study?
I think some people are going to be amazed at the numbers and the biomass. We’ve talked about hidden biodiversity of amphibians in general. This is an example of hidden biomass. There’s all this “meat” out there that’s used in this ecosystem function that we don’t see. Salamanders are seldom seen by the average person for a few reasons: they are nocturnal; they’re very small; and the vast majority live underground and when they are on the surface they’re usually in the leaf litter. They’re hidden. I think what we’re trying to hammer home is that just because we don’t see them doesn’t mean they’re not important. In fact, I think, based on the science, they may be more important than anything else out there, at least in terms of vertebrates. Again, that’s what Burton and Liken concluded as well in 1975.
Who were your coauthors?
Katie O’Donnell is a Ph.D. student in my lab. She’s finishing this semester and will be starting a postdoc at the USGS Center in Gainesville, Fl, where she’ll be working on Flatwoods salamanders at St. Marks National Wildlife Refuge.
Frank Thompson is the Director of the Forest Service Experimental Station here on campus. He funded the project and was part of the design of the experiment. The Sinkin Experimental Forest is where he does a lot of his work on oak regeneration in the Ozarks.
Interviewed and written up by: Melody Kroll
News by research strength
- Cell Biology
- Genetics & Genomics
- Molecular Biology
- Plant Biology
- Quantitative & Computational Biology