Date of Award

Fall 2023

Project Type


Program or Major

Natural Resources

Degree Name

Master of Science

First Advisor

David A Mortensen

Second Advisor

Rebecca G Sideman

Third Advisor

Jessica G Ernakovich


Most arable weeds arise from a resident soil seedbank and are typically controlled with tillage and herbicides, each of which pose sustainability challenges to growers and consumers. However, agronomic management practices that reduce weed seed input to the seedbank and that accelerate mortality of seeds already in the soil could reduce the need for more conventional weed control. Previous research and farmer practice has demonstrated that rotation of annual crops with several years of perennial forage can reduce weed abundance and the need for chemical weed control. In perennial forage systems, crop species and harvest intensity may have important effects on seedbank composition due to the multiple stress and mortality factors these systems impose on weeds, including facilitation of strong crop-weed competitive interactions, periodic defoliation of crop (and weed) canopies, and potentially crop species-specific associations with soil faunal and microbial communities that cause seed damage and mortality. Unfortunately, we know little about the influence of agronomic practices such as forage crop species selection or harvest frequency and harvest height on weed population dynamics or the specific mechanisms that regulate weed seed survival in the soil in perennial forage systems. We used a three-year factorial field experiment established in 2018 that included four mixtures of perennial forage legumes, each grown with orchardgrass, Dactylis glomerata L., two harvest height treatments (5 and 10 cm residual forage height), and two harvest frequency treatments (three and five harvests per year) to investigate how perennial forage composition and management act as community assembly filters on the composition and abundance of the weed seedbank. The composition and abundance of the weed seedbank was quantified at the end of the third year of the study using the direct germination method in a heated greenhouse. While overall seedbank density did not differ across treatments, seedbank community composition was influenced by the interactive effects of both harvest height and harvest frequency. More intensely harvested plots (shorter stubble remaining) favored weed seed bank communities more heavily dominated by weed species with specific functional traits: mat-forming weeds that set seed close to the soil surface were favored in short stubble harvest regimes, possibly due to their ability to grow and reproduce while evading defoliation. Conversely, the harvest regime in less intensely harvested plots enabled a more generalist weed community. In order to determine whether weed seed decay mechanisms might vary across the treatments, we conducted a weed seed burial experiment with seeds of a common weed, velvetleaf (Abutilon theophrasti Medik.). Seeds were buried in each experimental unit to simulate seed shed and incorporation in the soil and were extracted in the following spring. Seed death over the burial period was not different in more intensely harvested plots compared to less intensely harvested plots. This suggests that when considering how forage canopy management practices mediate seed mortality, seed decay may be of lesser importance compared to other mechanisms of seedbank decline such as fatal germination and seed predation. An indicator plant material was buried to gain a secondary measurement of cellulose decay in the system without the complex effects of seed dormancy and viability. In red clover and white clover biculture plots, cellulose decay was greater in more frequent harvest treatments when compared to less frequent. Lastly, both the decay of velvetleaf seeds and of indicator cellulose were significant predictors of seedbank community composition, demonstrating that microbial activity acts as a biological filter on weed seedbank community assembly.