Collaborating through the years to understand Grinnell flora

Johanna Swanson, Nehir Ergun, Nell Badgley 

As students in the course Evolution of the Iowa Flora (BIO305), our guiding question has often been, “Why is the Iowa flora the way it is?” To better understand the present Iowa flora that we have studied in the first part of our class, we have to understand the past.

The curiosity of a Grinnell biology professor over 60 years ago gave us a chance to reconstruct the flora of Grinnell College about 27,000 years ago. In 1962, Professor Ben Graham took advantage of construction excavations at what would later become Bucksbaum Center for the Arts, on the Grinnell College campus. He sampled some exposed peat, about 5-6 m below the surface, and he and his students performed initial analyses on pollen and wood fragments (Graham, 1962).

In his initial paper, Graham focused mostly on the geology of Grinnell, did not depict any photos of pollen or wood anatomy, and promised a “more complete analysis” that never came (Graham, 1962). Hence, students of BIO305 have taken up this analysis. 

A cardboard box containing remnants of Graham’s peat, rediscovered in the Noyce basement in 2018, has since been passed down between BIO305 classes who have continued trying to identify the wood, macrofossils, and pollen preserved in the peat. Together, we have collaborated across time to build a story of the Grinnell flora while learning about paleoecological research methods. This year, it was our turn to work with the peat. Using our own analyses and Iowa Flora classes that took place in 2018 and 2022, we hoped to gain a more comprehensive understanding of the vegetation present in this historic sample to expand upon and perhaps confirm the analyses composed in Graham (1962).

As a meager but mighty cohort of three students, we weren’t able to divide up into teams like previous classes. We decided to focus on wood samples. Like our BIO305 predecessors, we began our work with the box of peat– parsing through one manageable chunk of dirt to try to find preserved elements of interest. Here’s what we found. 

Results

1. Larix (larch) wood 

Peat sample W8.2(3) 200x (left); Modern Larix tracheids 200x (right);

This peat sample (right) is from a wood chip sample extracted from the peat by students in 2018, but digested and plated on a microscope slide this year. Of note in this image is both the crosshatching on several of the fibres and the pattern on the pitting of the tracheids. Crosshatching often indicates mechanical stress in a tree while the pattern, shape, and size of tracheids can be used for identification. In this case the tracheids’ pattern of paired, bordered pits with a width of 0.7 micrometers (μm), as well as larger pits with a width of 2.0 μm, is indicative of species in the genus Larix. We suspect that this could be Larix larcina, a species commonly found in bogs.

Our reference sample from a living larch tree on campus had a paired pit width of 1.0 μm and a larger circle width of 2.3 μm (right). 

Pollen 200x (from 2022 class)

While we focused on wood samples in the lab, Professor Eckhart was photographing pollen samples from the same peat and discovered a pollen grain that could be attributed to the Larix genus as well. We students hypothesized that due to the similar size of our sample to the reference Larix coupled with a pollen sample that looks remarkably similar to modern Larix pollen, the Larix genus was likely present over 26,000 years ago. Professor Eckhart, who analyzed the pollen during our course, indicated that Picea (spruce) pollen is the most common in our peat sample. 

Should our hypothesis be valid, this would be the first time a BIO305 class has identified larch wood in the peat, but not the first time it was ever identified. In 1962, our old friend Ben Graham elicited the help of fellow Grinnell biology professor (and plant anatomist) Waldo S. Walker, who identified large wood fragments in the peat mass as Larix laricina (Graham, 1962).

2. Definitive angiosperm wood 

P3(2) 200x. Wood vessel elements extracted, stained, and imaged in 2025. Left photo shows distinct scalariform perforation plates (circled) and dense side-wall pitting (boxed). Right photo shows a single vessel element.

Perhaps our most important find was definitive angiosperm wood. One of the clearest indicators of its character are the scalariform perforation plates and parts of a broken vessel (Right image, circled). While some other years there had been students who argued for finding magnolia pollen, their evidence was weak. We, however, have much stronger evidence for angiosperm wood. The vessel width is about 30 µm, the scalariform plates have about 12 bars and have a height of 0.7 µm, and the pits are around 0.32 µm wide. A major resource for identification came from the Inside Wood identification system created by North Carolina State University Libraries and a paper pulp identification manual created by Russell A. Parham and Richard L. Gray. With these two tools, our group and Professor Eckhart came up with two tentative hypotheses about what our mystery tree could be.

The first one is that our wood is a member of the Betulaceae family, which includes alders and birches. Some members of this family are native to cold bogs. All are wind pollinated. 

Radial section of Alnus incana, which is native to colder climates of North America. Image from Inside Wood and contributed by Elisabeth Wheeler.

In this image, there are 23 bars in the scalariform plates at the end of each vessel element, with a maximum width of 60 µm, and each plate bar is at least 6 µm wide. This is fairly different to what was found in our wood sample, as the width of the scaliform plate section was similar to the width of a vessel with no plating. Similar problems arise with looking at plants in the closely related genus, Betula. Another flaw in this hypothesis is that we have found no pollen from anything in the Betulaceae family, even though they are wind pollinated. As we have seen pollen from numerous wind pollinated species, it seems bizarre to find wood of a Betulaceae tree but find no pollen, although it’s possible no one has been able to isolate pollen from the peat thus far.

The other potential family is Ericaceae, more specifically the genus Kalmia. Kalmia includes species such as the famous mountain laurel and other species adapted to cold bogs. 

Radial sections of Kalmia latiflora, which is native to the Eastern United States. Image from Inside Wood (left: USw748; right: Keating 48784) and contributed by Elisabeth Wheeler.

 

With the right image as a reference for measurements, we believe our wood looks remarkably similar to the second image. In the first image, there are over 30 bars and the scaliform plates have a height of roughly 3.7 µm. While the heights of each scalaform plate are closer in Kalmia than it was for Alnus, there are still too many bars in a single section and each plate is much smaller in our sample.

Graham (1962) pointed out a number of potential species in the peat sample, among them Larix (larch), Picea (spruce), Abies (fir), Pinus, Alnus (alder) and Acer (maple). In our analysis, we were able to illuminate evidence of larch, spruce, and alder wood to support Graham’s analysis, as well as posit some hypotheses of our own.

Methodology: past and present 

Another goal of our class this year was to refine the sample processing and plating methodology in hopes of getting clearer images to more specifically identify the flora of the past.

Both this year and in classes past, the process began by selecting toothpick-sized wood fragments with tweezers and submerging them in a maceration fluid (equal parts glacial acetic acid 3% hydrogen peroxide) before placing these samples on a heating plate.The maceration fluid helped to separate individual cells (such as fibers, vessels, and tracheids) which would be crucial in identifying our samples. 

The old method, used by both the 2018 and 2022 classes, stained the samples with safranin dye and immediately plated them onto glass slides. This year, we incorporated an ethanol washing series into our methodology and experimentally applied different concentrations of safranin stain (0.1%, 1%, and 10%) to the wood samples to see if this affected sample clarity. We also compared two different mounting mediums: glycerin and tacky glue.

While we did not find much of a difference between samples that were and weren’t washed with ethanol, we did note that tacky glue was a much more efficient mounting agent than glycerin when it came to larger wood samples. The tacky glue enabled us to “mash up” larger wood pieces more efficiently so that we could view them under the microscope. 

What we found was most important for clarity of samples was how macerated (or “broken up”) samples were before analysis. What was difficult was that even if samples were well-macerated, sometimes the state of the original sample was too degraded to identify distinct species structures in the cells or to parse individual cells out at all. Furthermore, if samples were too degraded or if not enough care was taken during the ethanol washing process, sometimes samples were lost altogether.

Over the course of our few weeks working with this peat, we were able to build on the discoveries of Grinnell students and professors past and contribute to understanding the history of our campus flora. Perhaps thousands upon thousands of years in the past where Bucksbaum now stands was a shrubby bog with some pines nearby. Little did those plants know they’d be buried deep below the soil one day and picked apart by some curious college biologists far, far in the future. 

References

Graham Jr, B. F. (1962). A post-Kansan peat at Grinnell, Iowa: a preliminary report. In Proceedings of the Iowa Academy of Science 69(1), 39-44.