The school year has ended, but with the ongoing pandemic pushing most usual summertime activities off the table, my thoughts have been turning to how I spent my last summer: organizing fishing trips to Nevada. These excursions weren’t exactly spent relaxing on a boat in Tahoe reeling in trout. Rather, my companions and I were kneeling in a defunct quarry under the desert sun, prying apart layers of blindingly white rock in search of the fossils of small fish that once lived in a long-disappeared lake. The goal of my sweat-inducing, but ultimately enjoyable fieldwork season was to investigate an ecological replacement event that occurred 10 million years ago in the Miocene period. Thanks to funding from the UCMP and the immense help of friends and colleagues (most of whom came from the museum community) that volunteered their help in the field, I was able to collect over 300 fossils of sticklebacks and killifish, and present my preliminary results at the Annual Meeting of the Society of Vertebrate Paleontology in Brisbane, Australia in October 2019.
I am interested in studying how the interactions between species in an ecosystem change over time, particularly over periods longer than can be studied by observing living systems. The fossil record is a unique window into the deep past, and has offered plenty of insight into how ecosystems have changed over Earth’s history. But because the ages of the rock layers that contain fossils usually can only be constrained within several hundred thousand years (and often worse than that, even leaving aside the issue of varyingly-aged fossils within a bed), it has been difficult to connect events on paleontological timescales with processes ecologists observe in today’s world. I was contemplating this challenge last year when I heard about Michael Bell’s work on fossil sticklebacks, a kind of anchovy-sized fish still around today that has been the focus of many evolutionary and ecological studies. The fossils came from lake-deposited diatomite—a rock composed of the countless glass skeletons of single-celled algae called diatoms. Most of the diatoms would have settled onto the lakebed during their summer blooms, whereas in the winter only a relatively thin layer of mud would have been deposited. The seasonal cycle of diatoms and mud accumulating on the bottom remarkably went undisturbed, and can be seen today in the rock as white layers of diatomite divided by thin darker bands of mud. Each layer, known as a varve, represents a single year’s deposition in much the same way as rings in a tree trunk represent a year’s growth. With varves, the fossil record theoretically can provide a snapshot of every year—higher temporal resolution than the U.S. census!
Dr. Bell had used his site’s exceptional, essentially year-by-year record to study the fine-scale evolution of sticklebacks over several millennia. I wondered about the possibility of studying how the ecology of the lake changed over the same time. I spoke with Mike, who was joining the UCMP community as a research associate, and had donated his considerable stickleback collection to the museum. He told me how sticklebacks comprise virtually all of the animal fossils found in the quarry below a certain part of the section—and correspondingly, until a certain period in the history of the lake—but are replaced by an unrelated, though superficially similar species of fish in the killifish family. Mike had been focused on the sticklebacks themselves, and so nobody had studied this period of transition. This left an opportunity for a project—to chart the course of a species replacement event that occurred 10 million years ago at a nearly annual resolution.
My basic goal for fieldwork was to intensively sample fossils from just above and below the stickleback-killifish transition, with the aim of documenting the pace and tempo of relative population change between the species. Ultimately, I want to compare the results to data from modern ecological invasions in lakes, such as by lampreys and carp in the Great Lakes. I initially felt a bit guilty about going out to bring back more fish, what with the UCMP still in the process of housing the thousands of specimens collected by Mike, but there was no other way to get data from the right time and with the necessary stratigraphic precision.
Before beginning my sampling in earnest, I scouted out the site in the spring with Mike and Yoel Stuart, an evolutionary biologist at Loyola University Chicago interested in studying the tempo of evolution in the sticklebacks. The site, near Fernley NV, was once an active quarry (diatomite is powdered to make diatomaceous earth, used in pest control, filtration, etc.), and is still owned by the mining company. Mike introduced us to a company geologist named Michel Houseman, who is an expert in the geologic history of the area. Mr. Houseman had previously published in the academic literature on the species of diatoms found in the rocks and their significance for the history of the lakes, and explained to us the complex tectonics behind the evolving system of basins in the region. He posited that killifish may have entered the local lake at a time when the various bodies of freshwater in the region became interconnected. On the trip, Mike also introduced us to the art of collecting fossil fish: how to use putty knives to pry apart the layered diatomite in wide sheets that maximized the chance of finding whole specimens, and to measure specimens’ stratigraphic heights from marker beds. Our reference beds were often layers of volcanic ash which were primarily deposited from past eruptions of the Yellowstone supervolcano.
After the scouting trip, I came up with a plan for the summer and developed a simple method for estimating the relative population densities of stickleback versus killifish for a stratigraphic interval (i.e., a series of layers representing a chunk of time) by taking into account the sampled surface area and average splitting thickness of the rock for the interval. The money I received from the UCMP helped fund my five subsequent trips, allowing me to purchase tools for working through the diatomite, fuel to make the drive from Berkeley, and of key importance, a shade tarp that I obtained only after suffering through a weekend baked by 100°F temperatures. I was using my personal vehicle, whose low clearance and lack of 4-wheel drive were rather unsuited to the rock-strewn, very much unpaved road leading to the quarry, so the funding was also useful in repairing a couple incidents of tire and undercarriage damage. While the working conditions in the quarry might have been considered harsh between the heat, dust, and frequent dust devils that tossed around our things, overall the experience was a lot of fun. Rarely does fieldwork in vertebrate paleontology involve finding as many beautifully preserved fossils as we were, and we had the pleasure of seeing lots of hopping kangaroo rats, a panoply of desert lizards, and awe-inspiring night skies. There were also the refreshing dips in Donner Lake punctuating the drive home. Above all, the most enjoyable part of fieldwork was sharing these experiences with the people who accompanied me. Especially in this current time of social distancing, it’s nice to think back to the memories I made fossil fishing last summer.
In the end, we collected hundreds of sticklebacks and killifish from the quarry. Despite the sizeable sample, we found only a single killifish fossil below the youngest stickleback, suggesting that the transition was very rapid—the overlap covers a period of less than 50 years (one exception was a killifish I found on our scouting trip in a layer thousands of years older than the transition, indicating that killifish were entering the lake infrequently or at low levels before the shift). This is a pretty interesting result, given that invasions in the fossil record tend to be uneventful affairs with respect to the rest of the ecological community, with species popping in and out like tenants in an apartment complex. Finding a replacement event in the Miocene that appears to be comparably dramatic and fast to modern, human-caused invasions supports the idea that the general lack of such events in the fossil record is at least in part due to our coarse record of geologic time. Major disturbances to an ecosystem that last several decades, or even centuries or millennia, before giving way to a new stable state are likely to be entirely invisible to paleoecologists millions of years later—as long as the post-recovery community regains a broadly similar structure to its former state, and no global extinctions occur from the disturbances. That said, the disparity between past and present ecosystem responses to invasion may well reflect real differences in today’s world, like introduction of species co-occurring with other human impacts on nature, and the ability of humans to transport species in ways unlikely or impossible to occur naturally. Knowing the extent to which modern invasions are distinct in the history of life is important for predicting what the ultimate consequences of ongoing species invasions will be. My study certainly can’t answer this complex question by itself, but it does offer a rare window into what one invasion process looked like long before humans were around.
Thanks in large part to further UCMP funding, I was able to present a poster of this result to the annual conference of the Society of Vertebrate Paleontology in Brisbane this October. In addition to learning a lot and getting to see Australia with several fellow UCMP’ers, I received valuable feedback and ideas for my project. Given the short overlap between sticklebacks and killifish and consequently few data points within the transition, it hasn’t been as simple as I hoped to answer questions about the pattern of killifish invasion. For example, did sticklebacks only decline once the killifish arrived, or was their population dropping earlier? Did the replacement event happen immediately after the first killifish appeared, or was there some sort of lag time or fluctuation between the species populations? To address these and related questions as best I can, I plan to return to the quarry when possible and sample more intensely the few centimeters above and below the now well-defined transition. I have also begun looking into the diatoms that make up the rock itself, with the guidance of UCMP’s Assistant Director Dr. Lisa White. Diatoms are at the base of the food web and the presence of certain species tends to reflect environmental conditions, so a change in diatom composition correlated with the transition in fish could suggest potential causes (or effects) of the replacement. Lastly, investigating how the size and/or morphology of the fish varies around the transition could help reveal whether the species were really competing.
This project has been my first experience doing independent research as a graduate student. I have gained invaluable experience organizing fieldwork and practicing curatorial skills thanks to the UCMP’s funding and community support—including that of my companions in the field—as well as the invaluable advice and collaboration of Mike Bell. I look forward to sharing the ultimate results of my study.