Andrew Kaul is a new Restoration Ecology Post-doc in the Center for Conservation and Sustainable Development at the Missouri Botanical Garden. Here he describes some projects from his dissertation work conducted with Brian Wilsey at Iowa State University.
Tallgrass prairies once covered most of the central United States, but much of this historic ecosystem was lost to agriculture during the 19th and early 20th Centuries. Iowa sits at the heart of the tallgrass prairie range and lost more of its prairie than any other state except Illinois.
Throughout the Midwest, prairie restoration efforts have become increasingly common, and after several decades of research, the practice of prairie restoration has become increasingly complicated. We now know that restoration outcomes can be highly variable, and it is difficult to predict the outcome of any given restoration because there are so many factors that have been documented to influence restoration success, in terms of species diversity and establishment of target species from the seed mix.
For my PhD at Iowa State I studied 93 grassland restoration projects across Iowa. Previous work on grassland restoration had included careful experiments and thorough investigations of novel restoration techniques. What hadn’t been done before was to treat existing restorations each as their own little experiment and to sample broadly across the wide diversity of restorations in the real world. This approach allowed us to describe general patterns across many sites and to investigate which of the many potentially important processes tended to drive restoration outcomes.
With this project, my advisor Brian Wilsey and I sought to test which factors are the best predictors of restoration success in terms of species diversity, degree of invasion, and establishment of sown species. We considered factors including management history, the diversity of the seed mix, land use history of the site, site size and shape, soil characteristics, and weather during the first couple years of vegetation development. We took a retrospective approach to answer these questions, using existing restorations, which were highly variable in their age and how they were undertaken. We sampled vegetation at 93 restoration sites across Iowa over two summers and interviewed the managers of each of those sites afterwards to get information on when the restoration was started, what seed was used, and how it had been managed. We also sampled 5 prairie remnants, as a reference.
We found that by far the strongest predictor of plant diversity and recruitment of species from the seed mix was the degree of invasion by exotic species, where the more heavily invaded a site was, the lower the plant diversity and recruitment of target species. The influence of exotic species was more important than soil type, site management, restoration age, or any other aspects of the restoration, indicating control of exotic species is key to restoring prairies, and other temperate grasslands. The degree of invasion was higher in more linear shaped sites, sites with higher soil organic matter, and sites with fewer species in the seed mix, so we found that these variables were negatively related to our restoration success measures because of their indirect effects through exotic species. More linear habitats tend to have more “edge effects” where there are more colonization opportunities for exotic species. The higher invasion rates we found in sites with greater soil organic matter indicate the exotic species are better able to take advantage of nutrient availability. The lower invasion rates in sites seeded with more prairie species indicate that these mixes contain species that together, occupy more niche space and leave less open niches for exotic species to colonize. We also found that sites mowed during the first two years of establishment had higher diversity and establishment of sown species. This practice is supposed to suppress the annual weeds, which start growing before the seeded prairie species can establish.
Another goal of this project was to examine the ecology of milkweeds in prairie habitats. Milkweeds are obligate host plants for larvae of the monarch butterfly (Danaus plexippus), and in recent years, conserving and restoring milkweed populations in service of monarchs has become a major conservation priority in North America, especially in the Midwest, where many of the migratory monarchs breed. We counted milkweed stems within a meter of our sample quadrats at each prairie, and used these count data to examine what prairie habitats have the highest milkweed abundances, and what features of a prairie habitat best predict stem density. Specifically, we tested whether stem densities were different between remnant prairies, roadside restorations, and the non-roadside “conservation” restorations, most of which are managed by the Iowa Department of Natural Resources.
Milkweeds were far more abundant in remnants than restorations. Among restorations, roadsides had higher milkweed densities. Remnant prairies also had a higher diversity of milkweeds, so they are clearly an important habitat for this forb assemblage. Most of the milkweeds we sampled in restorations were common milkweed, even though it is rarely planted. On the other hand, Swamp milkweed (Asclepias incarnata) and butterfly milkweed (Asclepias tuberosa) are often included in restorations seed mixes, but were not nearly as abundant as volunteer common milkweeds.
Across all the restorations, we tested whether milkweed stem density was related to management (burning and/or mowing) or environmental variables including soil characteristics, plant diversity, degree of invasion, and site shape (linearity). We found that milkweeds were more abundant in more linear and invaded sites, and sites with lower soil density, and higher soil pH. These factors indicate that milkweeds are more abundant in areas with more soil disturbance. This is not surprising, considering the “weedy” ruderal nature of many milkweeds, especially common milkweed. The relationship with pH was a novel discovery, and future work will be needed to experimentally test whether milkweed germination or growth is higher in more basic soils, which is what our study indicates.
I am continuing my research on tallgrass prairie restoration with new projects examining plant functional traits to help understand why certain species are under- or over-represented in restorations. We have collected data on plant and leaf functional traits for over a hundred prairie plants and will test how the mean traits of plant communities differ between seed mixes, restorations, and remnants. Additionally, I am working with the Wilsey Lab on a related project examining phenological differences between plant communities in remnant and restored prairies.
To learn more, follow me on Twitter @andrew_kaul and check out our milkweed paper in Restoration Ecology. The prairie restoration study was recently accepted in Ecological Applications, under the title, “Exotic species drive patterns of plant species diversity in 93 restored tallgrass prairies.” Look for it to come out soon!