Improving plant diversity in non-diverse grasslands on a small family farm in southwest North Dakota

Results

Our study was conducted in 2019 and 2020. Overall, 2019 was a much wetter and cooler summer than 2020. April through September precipitation totaled 18.5 inches in 2019 and declined sharply to 6.7 inches in 2020. 

Our fields prior to research activities were similar and not species rich with only 7 species occurring on-site. Crested wheatgrass was the dominant plant accounting for nearly 40% of canopy cover, followed closely by smooth brome at 30%. Alfalfa was the dominant forb making up 11% of the overall canopy cover. Western wheatgrass was the most prevalent native species, but still accounted for less than one percent of the cover. Inland salt grass, dandelion, yellow sweet clover and Kentucky bluegrass were also found on-site in small percentages. Roughly 5% of fields prior to treatment establishment were considered bare ground while litter covered roughly 18% of the area.

Section 1. 2020 represented the second growing season since plots were seeded. Species richness was 14 in the native seed treatment and 11 in the non-native/native treatment. Richness was similar in 2019 on the native treatment at 14, but down some in the non-native/native treatment where 18 species were recorded in 2019. Planted cover for seeded species was ~ 43% in the native treatments and ~ 55% in the non-native/native treatment. Alfalfa was included in the non-native/native treatment and accounted for 38% of the total planted cover. In contrast, alfalfa was not included in the native seed treatment and did not contribute to the total planted cover. Nonetheless, alfalfa was common in the native treatments occupying roughly 27% of the canopy cover. In the native treatment, planted grasses made up 17% of the canopy cover while planted forbs made up 26% (Table 2). Side-oats grama had the greatest average cover for planted grasses at ~ 4% while Maximillian sunflower had the greatest cover of planted forbs at 15%. Planted forbs accounted for the majority of the planted canopy cover in the non-native/native treatment with 38% being alfalfa. Maximillian sunflower was the second most prevalent forb and accounted for 6% of the canopy cover (Table 2). Canopy cover of planted grasses in the non-native/native treatment was low at ~ 5%. Remnant grasses remained from the pre-restoration plant community with smooth bromegrass comprising between 3 and 10% of the canopy across treatments.

Section 2. German millet which was included in the 2019 cover crop, was dominant throughout section 2 making up 38% of the cover in the native treatment and 44% in the non-native/native treatment (Figure 2; Table 3). Species richness was 18 in the native treatment and 12 in the non-native/native treatment. Cover of planted grasses and forbs were similar between treatments at 6% and 8%, respectively. Crested wheatgrass and smooth brome were found in section 2 treatments occupying ≤1% of the cover.

Section 3. Species richness ranged between 7 and 9 in treatments regardless if oats were included or not (Table 4). Cover of planted grasses and forbs was 4% in the native treatment when planted with oats and 41% when planted without. Similarly, cover of planted grasses and forbs was greater in the non-native/native treatment when oats weren’t included (32%) versus when they were (25%). Oats occupied between 42 and 57% of the canopy in treatments that included oats. Big bluestem, side-oats grama and Maximillian sunflowers were the most common species found in section 3. Smooth bromegrass and crested wheatgrass were not observed in section 3.

Floral Resources

Section 1. Floral resources peaked in July and August when alfalfa and Maximillian sunflower, the two dominant flowering species were in bloom (Figure 3, Table 5). Five species were flowering in both the native and non-native treatments in June. Native treatments had 9 species flowering in July and 6 in August while 8 species were flowering in non-native/native treatments in July and 6 in August. In general, across these plots, there were more flowers blooming in August of 2019 relative to 2020 (Table 5).

Section 2. No plants along transects were observed flowering outside of Maximillian sunflower and plains coreopsis in this section. Maximillian sunflower averaged 8 flowering stems per transect for both treatments while plains coreopsis had 14 flowers per transect in the native treatment and 15 in the non-native/native treatment.

Section 3. Plants began to flower in late-July regardless if planted with oats or not (Table 6). Plains coreopsis and Maximillian sunflower were the two most common flowering plants across all treatment (Table 6). Though we did not assess floral resources in September, 2020, visual observations revealed heavy flowering of Maximillian sunflower and other plants throughout the beginning of September (Figure 4).

Butterfly Surveys

Instead of breaking down butterfly surveys by planting date and treatment, we choose to report the monthly average number of butterflies observed along each transect across the entire restoration (Table 7). Butterfly numbers varied across months with a low of 55 counted during June surveys and a high of 119 observed in July. The most frequently observed species also varied with Melissa blue and cabbage white common during June surveys and clouded sulfurs most common in July (Table 7).

Discussion

Many restored grasslands found throughout the North-Central Region lack diversity and struggle to provide ecological services to society (forage for livestock and pollinators). As grasslands continue to be converted to other uses, future restoration efforts should work to promote resilient grasslands that increase diversity and maximize ecological services. Here, we evaluated two seed mixes; the first a native seed mix and the second a mix that consisted of both native and non-native plants, seeded under a variety of situations across two years. We developed our seed mixes to provide forage for livestock and floral resources for pollinators. Because grasses tend to become more common over time at the expense of forbs, our seed mixes included a greater proportion of forbs than is typically used in grassland restorations with the expectation of maintaining a strong forb community into the future. Forb seed accounted for 52% of seed weight in the non-native/native mix and 56% in the native mix. Following two growing seasons, plants in section 1 appear to be establishing across both treatments as canopy cover of planted species are increasing. Both mixes two years post-seeding have resulted in a more species rich grassland than previously existed prior to restoration efforts. Slender wheatgrass and side-oats grama had the greatest cover of native grasses in the native planting while despite being in very low percentages, side-oats grama and green needle grass were the most common grasses in the non-native/native mix.

            Grassland restorations, especially those associated with native plants can take years before the final outcome can be predicted. The outcome can be heavily influenced by precipitation amounts and timing. We had above average rain in 2019, but below average precipitation in 2020. This may have given plants, like alfalfa that establish early and can tap into moisture reserves an edge over those that require more time before emergence. Alfalfa was common in both treatments despite only being included in the non-native/native treatment. Section 1 received only a single application of glyphosate prior to planting and this was likely not sufficient to eradicate alfalfa from the original grassland. Therefore, alfalfa has been able to survive and appears to be on the increase. Our concern with the planting in section 1 is that the one application of glyphosate may not have killed off all existing plants prior to restoration and over time, the alfalfa, smooth brome and crested wheatgrass will outcompete the other grasses and forbs. Time will tell. Despite the ecological services provided by alfalfa, we will target management actions to prevent it from becoming dominant.

            Restoration plantings completed in 2020 received little moisture until late-June and July, likely suppressing seed germination for some species. In section 2, the area planted to a cover crop in 2019, volunteer millet germinated prior to some seeded plants and may have further restricted establishment beyond the general lack of rain. Landowners and managers interested in planting a cover crop prior to grassland restoration should consider terminating the cover crop prior to flowering as seed could impact the long-term outcome of a grassland restoration. However, flowers associated with cover crops may benefit pollinators. Despite the lack of moisture and establishment of volunteer millet, species richness of planted cover was well above that found in the original planting. Percent cover of planted species was low in both treatments which is expected in first year plantings.

            Nurse crops are sometimes used during grassland restorations. We found that at our farm, seeding oats with perennial grasses and forbs reduced the percent canopy cover of perennial plants. Native grasses and forbs that can take longer to establish than non-native plants, accounted for 32% of the cover when seeded without oats versus 4% when oats were included. Similarly, planted perennial cover was 41% in the non-native/native treatment when oats were not included and 25% when they were. However, our results represent only one growing season and we expect changes in the plant community to occur over time, some of which may show benefits associated with the nurse crop. One benefit of planting oats with perennials is the wildlife appreciate the winter food source. Having oats interspersed with bare ground, perennial plants and annual weeds makes fantastic winter cover for upland birds.

            The seeds associated with our study mixes were expensive. Elevated costs were associated with many of the native forbs and grasses. It is difficult to accurately assess with any degree of certainty which mix was more cost affective relative to the other given the newness of restoration efforts. Nonetheless, there are some commonalities in species between mixes that have established and should be strongly considered in future plantings. Maximillian sunflower and plains coreopsis are two forbs that did well regardless of mix while side-oats grama was a common grass found throughout restored grasslands.

            Restoration efforts at our farm have resulted in greater floral diversity than previously existed with fields supporting flowers with an array of shapes, sizes, colors, and blooming periods. Assessing butterflies in small plots proved difficult as they freely fly from one treatment to the next. Nonetheless, we found that our fields provided for numerous different species of butterfly from June through August. Furthermore, our results show that providing flowering resources to pollinators throughout the growing season is important as the species we recorded using our fields shifted over time.

Management Implications

            Our grassland restorations are in their infancy and we expect changes in the plant community to occur over time. However, based on what we’ve observed we offer a couple of suggestions that may be of use going forward. First, one application of glyphosate is likely not sufficient to kill perennial cover prior to seeding. Additional applications of glyphosate prior to seeding, though an added expense is likely worth it. Second, the use of cover crops prior to planting perennial grasses and forbs may have benefits, but could also have the potential to introduce a huge amount of seed to the seedbed if not managed properly. We let our cover crop reach maturity which allowed pollinators to use flowers and provided food for wildlife, but also introduced competition for seedlings in 2020 when millet seed germinated. Instead of using chemical to terminate the cover crop prior to seed set, an alternative option may be to allow livestock to graze the cover crop while it is blooming which in return may allow flowering but prevent some seed set. Third, although very premature, it may be important to include many species of forbs during seeding if providing a diverse suite of floristic resources is important as many of the different forb species we included have not been found in the fields as of yet.

Moving Forward

            We intend to leave the restored area idle in 2021. However, managing these fields to maintain floristic resources will be a priority moving forward and we intend to conduct research to find the best way to maintain flowers in our fields. We also intend to fence the restored area and begin grazing livestock in spring 2022. Finally, we intend to use the information gleaned during this project to restore the remaining grasslands on our farm.

Table 2. Canopy cover of planted species two years post-seeding in section 1. Section 1 was treated once with glyphosate prior to seeding in May, 2019.

Table 3. Canopy cover of planted species during the first growing season post-seeding in section 2. Glyphosate was applied once in 2019, planted to a cover crop in 2019, treated with glyphosate again in 2020, and seeded in May, 2020.

Table 4. Canopy cover of planted species during the first growing season post-seeding in section 3. Glyphosate was applied three times in 2019, treated with glyphosate again in 2020, and seeded in May, 2020.

Table 5. Average number of flowering stems per 25 m transect in two different plantings established in section 1 in 2019 in Hettinger, ND.

Table 6. Average number of flowering stems per 25 m transect in two different seed mixtures seeded with and without oats as a cover crop in May 2020 in Hettinger, ND.

Table 7. Average number of butterflies observed along 25 m transects in a restored grassland in Hettinger, ND.