Understanding the Variation in Vegetation Composition of Prairie Restorations within Crop Fields

Abstract:

<p>Restoring native habitat within agricultural fields offers a way to balance the provisioning of ecosystem services like soil retention with the conservation of native habitat in extremely simplified landscapes, while simultaneously promoting sustainable crop production. As a farming conservation practice, these in-field restorations should be able to reliably achieve target goals across a wide variety of landscapes and contexts. For the vegetation community these goals could include species diversity and consistent native cover, which may facilitate ecosystem service provisioning and invasion resistance in high distrubed areas.</p>

<p>For this thesis I performed a comprehensive sampling of farms that have implemented in-field tallgrass prairie restorations (hereinafter "prairie strips") to understand which factors contribute to variation in the restoration outcomes of diversity and target species cover. I focused my analysis on the seed mix, since it is the largest economic investment a landowner will make in this type of restoration, but controlled for other differences between sites such as age, planting size, and planting season. I found that the seed mix richness was a strong predictor of diversity and target species richness across sites. I also found that the season in which a site was planted (spring, summer, fall) affected the establishment of target species, especially forbs, but that this effect was only detected in a subset of farms that had utilize the same seed mix. I found few predictors were associated with the abundance of non-target, or weedy, species across sites, and instead found that site differences explained the majority of variation in these outcomes. Non-target weedy richness and prairie species cover were negatively associated with one another, futher underscoring the importance of controlling weedy vegetation for successful prairie establishment.</p>

<p>I explored which species were consistently detected across sites in Appendix C. I found that nine species were reliably detected across sites and seeded at least five times. I only found that two species were never detected across sites but seeded a number of times (n=4), indicating they may be unreliable for this type of restoration practice.</p>

<p>In Appendix D I examined the results of a before-after survey that I distributed to my cooperating landowners and farmers to assess their confidence and interested in indentifying plants. I found that the majority of my cooperators were interested in learning to identify more species and would find it most effective for their learning if another "knowledgeable" person taught them. I also found that the majority of my cooperators percieved their botanical literacy to have increased during the two years that this work took place.</p>

<p>In sum, this work demonstrates that certain restoration outcomes can be predictably achieved across a variety of farms through management decisions like the seed mix and that some species can reliably establish across these in-field restorations. More work will be needed to address the factors that consistently suppress non-target vegetation, as it contibues to be an unpredictable and problematic aspect of prairie restoration and management.</p>