Natural selection works on Ecologists, too

My study site in Washington...not too shabby

Lots of folks don’t realize that, in many cases, big scientific discoveries emerge from lucky situations. However, for most ecologists, scientific discoveries are hard fought. In ecology and evolutionary biology, the lab and field work required to advance our scientific understanding can be wrought with unexpected and unavoidable obstacles. Through experience and seemingly countless failures, these scientists must learn how to preempt, circumvent and mitigate the unforeseen that could ruin an experiment. If they don’t learn this skill, natural selection could weed them out of the field of successful scientists, preventing them from obtaining the coveted faculty position. Ok, yes, I’m being dramatic, but, seriously, the ability to simultaneously prepare for the worst and hope for the best is perhaps one of the most important qualities of a successful field ecologist.

Last summer, from mid-May to early August, I conducted lab and field studies in Wenatchee, WA, a small town on the eastern slopes of the Cascades that is actually known as the Apple Capital of the World. When I learned of this accolade, I was pretty shocked, because in the summer Wenatchee is bone dry and blazing hot thanks to the rain shadow effect. Starting in June, the daily high rarely dips below 90 degrees Fahrenheit, and after a few months there, it was hard to remember what rain sounded like. Needless to say, irrigation is critical to the agricultural industry in Wenatchee.

Douglas-fir tussock moths (DFTMs) chowin' down on their favorite tree

I was based out of a federal research lab in Wenatchee and conducted some field experiments on Forest Service land outside of Chelan, about an hour and half drive from the lab. My collaborators and I are studying a microbial control agent of the Douglas-fir tussock moth. Yes, that’s hard to digest, so let me break it down a bit. The Douglas-fir tussock moth (DFTM) is a native moth species of western North America, with a pretty large range that extends from southern British Columbia down to New Mexico, with an eastern range limit near the Rockies of Colorado. At the end of summer, adult moths emerge from cocoons, and they have about 48 hours to find a partner to mate, before they perish (talk about dramatic!). Adult females are wingless and sit atop their recently spent cocoons, releasing pheromones that will hopefully lure a male to fly down. After mating, the female wraps her eggs in that same cocoon and then dies. Those eggs spend the winter in diapause - think insect hibernation - and emerge as tiny larvae the following spring. Now, in years when there are a lot of these moth larvae around, as the larvae feed and grow, they can completely strip a Douglas fir tree of its needles. This defoliation can drastically reduce tree growth rates or even kill the tree. And, in really bad years, hundreds of thousands of acres of Douglas fir can be defoliated and killed by these ravenous little larvae. This has huge ecological and economic impacts, especially considering the value of Douglas fir timber. Check out this media coverage of a recent, smaller outbreak of the moth near Colorado Springs.


A branch that has been entirely defoliated by the DFTM caterpillars


In order to combat large DFTM outbreaks, the US Forest Service has taken advantage of one of the moth’s natural enemies, a virus (a nucleopolyhedrovirus for the savvy). This virus is species-specific, meaning that it can only infect DFTM and not any related moth species, and it can drastically reduce the number of larvae over the course of the summer. Larvae of the DFTM become infected with the virus when they consume contaminated needles on the tree. The virus incubates in the larval body for about 7 days, at which point the virus particles release an enzyme in mass quantities that causes the insect’s tissues to dissolve. Basically, the virus liquefies its host in a pretty dramatic fashion. This causes billions of virus particles to spill out over the tree’s needles, which can then be consumed by other uninfected larvae, and the process of transmission continues. After decades of research on these moth-virus interactions, conducted mainly in the Pacific Northwest from about 1960 - 1980, the Forest Service came up with Tussock Moth Biocontrol-1 (TMB1). This is a powerderized, water soluble form of the virus isolated from infected larvae that can be sprayed from helicopters. This is not some genetically engineered virus; Forest Service personnel reared thousands and thousands of DFTM larvae, infected them with a naturally occurring virus isolate, incubated the virus in the hosts, collected the virus, and powderized it. TMB1 became the first EPA-approved, viral biocontrol agent for use against a forest pest. Pretty cool! Nowadays, the US Forest Service uses pheromone traps to keep tabs on the abundance of moths in western forests. If densities are high enough to cause serious ecological or economic losses, the Forest Service can aerially spray TMB1 after larvae hatch in the spring, which can reduce the overall defoliation of the forest.

What’s interesting - and problematic - about this microbial control program, is that strains of this virus (i.e. strains other than the sprayed TMB1 type) circulate naturally in DFTM populations. And it can be difficult to predict when these naturally circulating virus strains might cause an epidemic that would obviate the need to spray TMB1. There’s a similar problem in predicting which strains of influenza virus will be dominant in the next year, and therefore, which strains should be used to make vaccines. In our case, knowing whether a naturally occurring strain will cause a large epidemic is important, because the spray program can be quite costly. For example, in 2000, the Forest Service sprayed about 40,000 acres in Oregon at a cost of approximately $2.6 million, roughly $67/acre. That’s not cheap.

Our research goal is to conduct experiments that will allow us to create mathematical models that will predict the conditions under which it would be most economically efficient to spray TMB1. This is some complicated stuff, so, I’ll move on to what’s relevant here. In the field, we create mini-epidemics using DFTM larvae that we have reared in the lab. We place these larvae into mesh bags that we secure on real Douglas-fir branches and add virus. This way, we can measure how the virus spreads, and we can use these measurements to improve our models.

We wanted to conduct experiments on a few different virus isolates with many different experimental treatments, tweaking things here and there to see how the virus behaves. This meant we needed to create hundreds of these mini-epidemics in the field. Our field site was an hour and a half’s drive from our lab up into the mountains, outside of the town of Chelan. Because of this, my field technician, Jeff, and I camped near the field site each night. We collected samples from the experiments during the morning, drove the samples down to the lab in Wenatchee to sort during the day with the help of more technicians, and then we would drive back up to the field site in the evening to camp and repeat for a few weeks. Remember though, that this part of Washington is hot and dry, and because we were up in the mountains, we had no cellphone reception.

On the evening of 28 June, Jeff and I were heading north out of Wenatchee to our field site. Just as we were at the edge of town, we saw a small funnel of smoke billowing from behind the foothills to the west; we couldn’t make out the source. It was a pretty small fire burning, and it wasn’t very windy, so we weren’t worried at all. Small fires - along with some very large ones - had been cropping up all around Washington at the time, so this wasn’t something unexpected. There was a fire ban all over the state, so we were also used to eating out of cans at the campsite most nights. The next morning we collected a lot of samples - this was one of the busiest days of the experiment, and one of the most critical. We packed them tightly in the truck bed, and headed down from the mountains back to Wenatchee.

About 45 minutes into the drive, with Jeff at the wheel, we hit the point of regaining cellphone reception, and my phone went berserk. A stream of text messages and voicemails came barreling in. At the time, my postdoctoral advisor, Greg, was in town with this wife, Alison, to help out with the experiment; it takes a team of 5 or 6 to complete the lab processing in one day. It would take me too long to read and listen to all of those messages, so I decided to just call Greg to see what was happening. It turns out that the small fire, which started just outside the town of Cashmere, had grown enormous over night and had swept across the mountains, down the foothills towards Wenatchee. At the same time, an ember from this fire had swooped into an agricultural factory in the middle of Wenatchee and had lit a stack of wooden pallets on fire. This grew to explosive levels due to adjacent, volatile agro-chemicals. The middle of town was on fire, and a forest fire was encroaching down the mountains just to the west of town, all throughout the night and into the morning. The federal lab, where we usually sorted our samples, was shut down, because the fire had stopped only 200 yards from the building. Fortunately, the fire was under control by this time, but windy conditions threatened its stability.

So here I am, in the passenger seat of our research vehicle, with hundreds of perishable samples in the truck bed, and no place to process them. The experiment will be ruined if I don’t get these things sorted out! When I received this news, we hadn’t even gotten to the point where we could actually see the town of Wenatchee. Jeff and I had no idea what to expect. As we pulled into town, we could see the scorched hillsides, and we could feel the smoke in our lungs from the factory fire. But science could not rest!

I got in touch with our collaborator at the Forest Service, Karl, and he graciously offered to let us use his tiny garage to process samples. Unfortunately, there was no air conditioning in this garage, and it was a sizzling 102 degrees outside. At this point we had no other options. Greg, Karl, Allison, Jeff and I packed into the garage with makeshift tables loaned to us from neighbors, and we sorted away for hours, sweating. We were also battling time, because in such heat, our samples - which were actually live moth larvae - might die from the heat inside the mesh bags, which was climbing inside the garage.

You have to wear a mask when working with caterpillars. They have "urticating" (bristly, irritating) hairs they release as a defense mechanism. Over time they can cause a severe allergic reaction and affect your breathing or cause a bad rash.

Our makeshift workshop/ sauna

Fortunately, with the help of the team, we salvaged the majority of the samples. Also, thanks to the efforts of the fire department and a fair bit of luck, the town of Wenatchee survived with relatively minor damage and no casualties. Many other towns and cities in Washington were not so fortunate that summer as fire swept through the state’s public and private lands.

Cute, but troublesome little buggers

Although my field site was spared from the fires last summer, the record heat waves still took their toll, and I lost 25% of my experiment. Many of the DFTM larvae inside the experimental bags dessicated, which is a natural source of mortality for these populations, but one I was hoping to avoid in my experiments. However, from years of experience with experiments going wrong, I had prepared for the worst and increased my sample sizes beyond what was statistically recommended. Even though this led to many many more hours of labor in the lab and in the field, it turned out to be necessary to salvage my experiment from the heat and fires.

There are so many ways an ecological field experiment can go wrong. As scientists, we need to prepare for the worst and hope for the best. But in the end – and I think most field ecologists will agree – the extra work is totally worth it.