Studies on the Ecology of Roadside Turfgrass Mixtures in Minnesota

Abstract

Establishing persistent vegetation along roadsides is challenging in cold climates. These areas are subject to snowplow damage, winter freezing and ice encasement, excessive heat and drought, application of deicing salt, and poor maintenance; they also often contain poor soils. The result is low vegetation cover which leads to more soil erosion and greater vulnerability to weed invasion. We wanted to identify adapted turfgrass mixtures and characteristics or disturbances of a site that leads to poor coverage over time. All experiments were conducted at 14 roadside research sites in Minnesota. Each of these sites was located along a two to four lane road with different traffic volumes. We collected soil for a seed bank analysis that took place in the greenhouse using the soil emergent method. Then, at these same roadsides a turfgrass mixture experiment was seeded that consisted of 40 treatments composed of monocultures, two-way mixtures, some three-way mixtures, and a single six-way mixture, in addition to four currently recommended Department of Transportation seed mixtures. Each site contained three blocks in a randomized complete block design; seven of the sites were seeded in the fall of 2018 and seven in the fall of 2019. The total turfgrass coverage was assessed at each site twice per year using the quadrat-grid intersection method. In the first experiment, we wanted to characterize the seed bank at these different sites and understand if it affects the weed coverage over time in the field plots. We found that there were differences in the seed bank at many sites. A range of more than 9 times was found in seedling density between sites (23-209 seedlings L-1). Differences were also found in observed species density (8-16 species L-1) and Chao estimated species density (9-32 species L-1) for each site. Despite the significant differences in the type and density of the seed banks, its impact was relatively low on weed coverage over time. Weed coverage was found to be lower when turfgrass coverage was maintained over time. The next experiment sought to identify the effect of including greater turfgrass species richness in a seed mixture on the coverage over time. We found a significant positive interaction with turfgrass coverage as a function of the number of species and time (Est=0.08, S.E.=0.02, p<0.001). This suggests that turfgrass coverage is increasing through time when more species are included in a mixture. This finding shows that roadsides maintained without regular fertilizer applications and no supplemental irrigation after establishment would benefit from greater species richness in a seed mixture. Finally, we wanted to identify different seeding clusters for the state of Minnesota, because the Minnesota Department of Transportation is currently recommending statewide turfgrass mixtures. If specific clusters are identified it could improve the applicability of turfgrass recommendations, and likely result in more turfgrass coverage over time. We collected soil and weather variables from each site and performed an agglomerative hierarchical cluster analysis. We validated the results of the clustering by comparing the species composition at sites. Our results suggested an optimal clustering would consist of two geographical seeding clusters in Minnesota (north and central/south) and one non-geographical cluster for sites that contain poor soil quality. A poor soil quality site generally contained more sand, greater bulk density, a higher saturated paste extract electrical conductivity, and lower organic matter. We recommend more soil testing procedures for practitioners before seeding a site. Seeding mixtures designed for these clusters in Minnesota will result in improved coverage over time allowing roadside vegetation to fulfill its intended functions.