The Berkeley Science Review: Articles

Getting Back To Nature
Revisiting the 1914 survey of California wildlife (view PDF)
by Erica Spotswood

In the Museum of Vertebrate Zoology (MVZ), director Craig Moritz walks to a row of cabinets and pulls out a shelf. Inside lie rows of chipmunks, carefully stuffed and labeled, with a tiny skull sealed in a glass jar next to each. To the untrained observer, they look like replicas of the same species. To Moritz, they tell a story that crosses the boundaries of both space and time. These specimens are part of a unique biological survey project launched by Joseph Grinnell, the museum’s first director, in 1908. The Grinnell survey, which lasted over 30 years, covered over 700 locations spanning the state of California. The resulting database, encompassing over 20,000 specimens, 13,000 pages of field notes, and 2,000 photographs, represents one of the most comprehensive collections of its kind in the world.

Moritz must have known he was stepping onto the shoulders of giants when he began his position as director in April 2001. Looking for background information on the history of the museum, he was given Grinnell’s Philosophy of Nature, a compilation of writings published by his predecessor in the late 1940s. In the book, Grinnell predicts that the real value of his field work “will not be realized until the lapse of many years, possibly a century.” Excited by the idea of using the museum centennial to complete Grinnell’s prophecy, Moritz began to think about returning to the original sites to see how the ecological communities had changed over the years. What followed was the development of the Grinnell resurvey project, begun in 2002 in Yosemite.

After three summers of intensive fieldwork and collaboration between the National Park Service, the MVZ, and the U.S. Geological Survey, the resurvey team has revisited all of the original 42 sites. Armed with the detailed information from the past provided by the original survey and the newly collected data from the resurvey, a diverse group of contemporary Berkeley scientists is using the Grinnell collection to study a series of interrelated issues—from the changing distribution of vertebrates to the impacts of climate change to the developing patterns of genetic diversity. Because the original database is so complete, it is providing a rare opportunity for modern researchers to get a glimpse into the past, to examine the present, and to predict the future.

Remembrance of Things Past
On October 28, 1907, benefactor and avid naturalist Annie Montague Alexander wrote a letter to the UC Berkeley president proposing $7,000 towards the running of a museum dedicated exclusively to the mammals, birds, and reptiles of the west coast. All the University had to do was come up with the means to construct a building complete with electric light and heat. And so the Museum of Vertebrate Zoology was born.

Joseph Grinnell, who was its first director from 1908 until his death in 1939, was not simply concerned about collecting specimens for the museum. His goal was to understand how species and communities were distributed across space and across ecological gradients within the state. According to Jim Patton, Professor Emeritus and curator of mammals, “He was looking at geographic variation and change of characters in space and time. He wanted to understand the kinds of factors that might influence local adaptation and … variation among individuals and within populations.” These ideas were unique at the time because they called into question the accepted notion that species are static and unchanging.

Grinnell’s ideas were more contemporary with those of the biologists of the 1940s, who developed the notion that differences between species are driven by ecological and geographical barriers. The result of this philosophy was a one-of-a-kind collection. “There are lots of specimen collections in the world, but what is missing from them is the Grinnell philosophy and the methods he used,” Patton explains. “He went out and looked at organisms in a controlled way rather than haphazardly saying ‘we don’t have any specimens from location X so let’s go out and get some.’ There is an ecological and conceptual framework that underlies all of the localities that were visited.” Grinnell also developed a method for recording information (the Grinnell field note system) that is still used around the world to this day. The resurvey team is attempting to adhere as closely as possible to Grinnell’s original methodology.

First, they must find the exact location where a given survey was conducted. In some cases, this is easy. A description of the site plus a point on a topographic map was sufficient for Jim Patton to find the exact slope in Lyell canyon where Grinnell set his traps. Where precise information is missing, or where changes in land use have rendered a resurvey at a location irrelevant, things are more complicated. For example, the original trapping location at one site now sits in the parking lot of a Wal-Mart. Instead of trapping next to the dumpster, a comparable site nearby with similar vegetation in a similar habitat was chosen in its place.

Once the location is determined, a camp site is chosen close by and traps are set out for four days. Here, too, it is impossible to mimic precisely the methods Grinnell used. For one thing, Grinnell’s team shot animals—something that’s impossible inside the park, and impractical, at best, outside of it. As a result, the resurvey team does not survey for carnivores (which are usually larger, rarer, and more difficult to trap without shooting them), though they have made use of data collected by the park to inform them about current distributions.

The traps they use are different as well. Live traps are used in the resurvey, whereas a small lethal trap called the “museum special” was used for most of the small mammals in the original survey. Named for its niche market, the trap protected the skull by breaking an animal’s neck instead of hitting it on the head. Valuable to taxonomists, the skull is used in identifying closely related species. Current bird survey methods have also been modified slightly. The surveyors still walk along a path, as Grinnell did, but now birds are surveyed only at specific points along the way. At these locations, called point counts, all birds heard or seen over a seven minute period are recorded.

Survey Says…
Equipped with volumes of data from the two surveys of Yosemite, a small army of people associated with the MVZ is now working to analyze and catalog the differences in vertebrate communities between the two time periods. Documenting and verifying these changes is no small feat though. In order to prove beyond all reasonable doubt that a species is present where it did not exist before (or vice versa), one must be able to show that the absence of that species was because it was not there and not simply because it was not found. The presences are more straightforward. If a species is found and you have a good taxonomist to identify it (and a specimen to prove it), you know it was there. But how do you prove something is really not there if you can’t find it?

Moritz is working with population biologist Steve Beissinger from the Department of Environmental Science, Policy, and Management to build models of how “trappable” each species is by looking at the total number of sites and the animals observed at each site. Mammal curator Chris Conroy explains that by using this method, “If an animal was always trapped on every trap line, every night, you get an idea that it is an easily trapped animal. If you then go to a place and don’t trap it, you can be more confident that it truly isn’t there and that you didn’t just miss it.”

Determining what was trapped and when during the original survey has also proved more difficult than expected. Roughly three times more information exists in the field notes than in the specimen collection. The field notes are scanned and available for anyone to view online via the MVZ webpage, but there is currently no easy way to search this database, other than, of course, by looking through each entry. The museum wants to make this simpler, and they are working to develop software to recognize key words in the field notes or to convert them all into text. For now though, each question asked can only be answered by hiring someone to pore through all of the field notes.

In some cases, this has been worth the effort. Juan Para, a PhD student in integrative biology, spent a year sifting through 13,000 pages of field notes recording every mammal caught on every trap line between 1910 and 1925. The database shows that small mammals have been moving around in some surprising ways. Several species have shown a shift in their altitudinal ranges of up to 2,000 meters. Four species not originally found in the park have expanded their ranges upward in elevation into the park. Four species of small mammals which were formerly common have contracted their ranges. One, the shadow chipmunk (Tamias senex), has gone from being very common to virtually non-existent. In some cases, the reasons for these changes in species distribution are related to fire. Since the mid 20th century, the National Park Service has aggressively suppressed fires inside the park.

Comparing current photographs with those taken during the original Grinnell survey show marked increases in tree density, as well as some encroachment of trees into what were once meadows. Corresponding decreases in the abundance of small mammals that prefer forest floors that are open with dappled sunlight, such as the Golden Mantled Ground Squirrel, are easy to explain when one considers the increase in forest canopy density. But there are other species in which no such explanation can be found. Why, for example, has the piñon mouse expanded its range into the park? Now found 2,000 meters higher in elevation at locations as high as 10,200 feet, the mouse has been trapped miles from the its nearest preferred habitat of piñon pines and junipers. Likewise, the alpine chipmunk and the American pika were formerly common at elevations as low as 7,800 feet. Far less numerous today, neither has been found below 9,500 feet.

Movin’ On Up
Moritz, Patton, and their crew of researchers fear that these changes in elevation could be linked to global climate change. There is another line of evidence that supports this idea. Contrary to what the mammal researchers have been finding, bird diversity appears to be increasing inside the park. Birds such as the blue winged teal are now found breeding in the high lakes of Yosemite. These birds look for lakes that are free of ice to land in. Earlier ice-out dates associated with global warming could be the explanation. Additionally, several high elevation species are declining in numbers. Thus similar evidence across the very different bird and mammal taxa suggest that climate change may be an important factor influencing the declines in abundance of high elevation species.

To explore further the impacts of climate change on the survey species, researchers have been using climate data from the early 1900s to develop species distribution models. Most climatologists do not have access to species distribution data from multiple time periods and therefore cannot directly test how species have moved as the climate has warmed. To get around this, models must look at changes across many locations during the same period of time. The assumption is that the locations differ in climate, and nothing else. In practice, nature is never so simple.

The Grinnell project offers a rare opportunity to do the opposite—look at the effects of climate change over time instead of across space. PhD candidates Bill Monahan, Juan Parra, and Morgan Tingley have been taking the opportunity to use climate models created from the Grinnellera to predict species distribution in the present. Then, the current survey will show if their predictions match up with what the survey team actually finds. Likewise, current climate models can be used to predict past species distribution based on past climate, which can then be compared to the original Grinnell survey findings.

As Monahan explains, the project provides an opportunity to train the models and increase their accuracy, which can then be used to more precisely predict how species will change in the future. What they have done so far is preliminary. “For some species, the model did really well while for others, the model did a horrible job,” he adds. But when Jim Patton looked at the predictions for the alpine chipmunk, what he saw was accurate. “If you model its distribution based on Grinnell climate and distribution and then predict its distribution now, you actually see this altitudinal shift. It is impressively clear.”

The high elevation species are of particular interest for several reasons. First, high elevation areas are those in California that are most likely to have experienced the least amount of land use change in the past 100 years. The effects of climate change can therefore be isolated and investigated alone. Second, the high elevation areas are predicted both to experience more warming and to contain species that are more vulnerable to climate change. Restricted to high elevations to begin with, as the climate warms, the habitats of these species are predicted to shrink, eventually leading to their extinction. These patterns should be visible much sooner in animals than in plants because they move around so much faster.

Other explanations for the altitudinal shifts in mammal distributions do exist, and more work needs to be done before the Grinnell team will be able to say with certainty if climate change is to blame. One hurdle in this research is the lack of a good control—a place where the climate is known not to change—since climate change is a worldwide phenomenon. It’s also possible that competition between species for similar resources like food could be the cause of the shifts in population—a good hypothesis, but one that is difficult to measure. The Grinnell resurvey team has not looked closely enough at the behavior of the study species to rule competition out as an explanation. Only further research and the completion of the current resurvey project can hope to shed more light on the potential causes of these observed trends.

What the Future Holds
With the resurvey of Yosemite largely completed, Lassen National Park is next on the team’s list. Work will begin this spring on this part of the project, which extends from Red Bluff in northern California, east to the Nevada border. Post-doctoral fellow John Perrine has been working for the last five months in the planning stages. Finding the sites where Grinnell originally surveyed has proved much more difficult than it was in Yosemite and has taken a great deal of historical sleuthing.

Digging through old maps, tax records, and historical land tenure documents, Perrine has managed to locate many of the sites, though he has had to contend with quite a few obstacles: towns that have disappeared, names that have changed, railroads that have been built and then abandoned, ferries that used to transport people across rivers that now have bridges, and giant cattle ranches owned through land grants by Spanish rancheros that no longer exist and whose precise locations were never defined. If his work is any indication of how the rest of the project will go, the team will learn a lot about history in the process. More importantly, they hope to build on what they learned in Yosemite, verifying or disproving the patterns they have begun to see emerging. It’s a big world out there, and with the Grinnell data and the resurvey team’s effort, we’ll be able to sneak a peek into how human activities are changing, and will continue to change, that world in the future.

Erica Spotswood is a graduate student in environmental science,
policy, and management.

Want to know more?
Check out:
http://mvz.berkeley.edu/Grinnell/index.html

Genes from Drawers:
In addition to shifts in distribution, the genetic diversity of mammal populations may be changing as well. Access to museum specimens from 100 years ago with precise information on the locality where they were collected provides a rare opportunity to study how changes in distribution have influenced the genetics of modern populations. Emily Rubidge, PhD student in the Department of Environmental Science, Policy, and Management, is using new techniques for extracting DNA from old museum skins to compare them to the resurveyed collection. Looking at a set of genetic markers to determine variability, she will be able to determine if there has been an overall change in the total genetic diversity between the two time periods.

The way in which a species has declined is expected to be reflected in the present gene pool. If an entire population of alpine chipmunks moved up in elevation, one might expect that they would have maintained the same degree of genetic diversity within the contracted range. Alternatively, if the range contracted when the lower elevation population went extinct, one would expect the current population to be less genetically diverse than the original. Rubidge’s preliminary results suggest that the alpine chipmunk has lost genetic diversity, suggesting the latter hypothesis. As she explains, “One of the big problems conservation biologists face is that we don’t know what things were like before. Although the environment obviously wasn’t unaltered in the 1900s, it is a baseline that we can use to look at changes. It is exciting to be able to ask population genetic questions about populations 100 years ago.”

Comments on this article? Drop us a line at with 'letter to the editor' in the subject!