top of page

Experimental Design

3 treatments
6 replicate plots of each type

​

Addition of the following:
1) Prairie grasses  (G)
2) Prairie grasses + 1x forbs (GF1)
3) Prairie grasses + 2x forbs (GF2)

 

A, B, C, D refer to 1m x1m permanent vegetation sampling plots

It was essential to establish replicates of each of these sowing treatments.  When scientists ask whether their treatments (in our case, the G, GF1, and GF2 sowing treatments) lead to different biological results, they are in effect asking:  Can we see bigger differences in whatever we are measuring “among the treatments” than “among each of the replicate plots of the same treatment?"

Full details of the experimental design and establishment of the site are in the first published research paper and supplemental materials based on data from the Free State Prairie site.  Click the buttons below to read these documents.

​

In addition to the main experiment, Peggy Schultz and colleagues also established an “experiment within the experiment.”  Peggy Schultz is a scientist at the University of Kansas in the Kansas Biological Survey and Environmental Studies program.  She wanted to explore whether addition of small amounts of soil from remnant unplowed prairies would promote prairie plant establishment.  The logic was that such soil would contain prairie microbes, including fungal mutualists called mycorrhizae.  As shown on the left of the experimental design figure, her experiment was established within the twelve GF1 and GF2 plots.  Click the button below for full details of the design and results of this study.

As opposed to simply planting a field of prairie plants, Free State Prairie was set up as a “restoration experiment.”   Why?  The answer is simple: restoration ecology is an active scientific discipline, and much can be learned from experimentation and research.  Importantly, we wanted students to have first-hand exposure to scientific inquiry.

IMG_3733.jpg
IMG_4546.JPG

But what kind of experiment?  We had two criteria.  First and most importantly, we sought an experimental design that allowed us to explore testable hypotheses relevant to current ecological questions.  In contrast to the traditional, perfunctory laboratory in a classroom, the “answers” to authentic research questions are not known a priori.  Second, we wanted experimental treatments that would create large areas of different vegetation types.  Such an ecologically heterogeneous environment would attract diverse fauna and provide contrasting habitats for sampling exercises by high school classes. More generally, the creation of this site would add an interesting natural environment to the campus.   

After many, many conversations and consultations we decided on the details of our restoration experiment.  Bryan Foster, Helen Alexander, Julie Schwarting, Kelly Kindscher, Jim Bever, Peggy Schultz, Ben Sikes, Dan Hirmas, university and high school students and many community experts helped in the original research design. 

Screen Shot 2020-12-16 at 9.32.27 PM.png

For our overarching research topic, we focused on a well-known challenge in prairie restoration: restored prairies do not have the high diversity of forbs (non-grass flowering plants, “wildflowers”) present in remnant prairies.  Numerous factors may influence forb diversity, including the particular mix of species being sown, the sowing rate, effects of weedy plants (non-sown species), and the nature of the soil microbe community.   Several of these factors could be addressed using our experimental design.  

IMG_0163.HEIC

In brief, we decided to plant prairie grass seed over the entire site, and then divide the area into 18 plots, with six replicates of each of three treatments.  One treatment (G) was the grass-only treatment (only addition of prairie grasses).  The other two treatments (GF1, GF2) had not only prairie grasses but prairie forbs at two different sowing densities (GF2 has twice the density of forbs than GF1).  This simple design manipulated both species diversity (G vs. GF1/GF2) and sowing density (GF1 vs. GF2) and lent itself to diverse questions.  For example, we can ask: (1) does sowing twice as many forb seeds result in twice the establishment of such species?  (2) does the establishment of weedy, non-sown species depend on sown species diversity or sowing rate?  Finally, by including plots without forb seeds (G plots), our work can contribute to current discussions about forb dispersal and colonization.  For instance, we explored (3):  how quickly do forbs colonize the grass-only treatment and are there species-specific dispersal patterns?

prairie plots design.png
early aerial .jpg

Aerial view of Free State Prairie in 2015

bottom of page