Still No Difference

Historically, wetlands in the Prairie Pothole Region (PPR), located in southern Canada and the upper Midwest of the United States, were restored through drainage removal, which reestablished the natural hydrology of the site. By restoring hydrology, wetland vegetation would recolonize on its own via long-lived seedbanks and propagules dispersed by waterfowl (Galatowitsch and van der Valk 1996). However, reinstatement of natural hydrology alone is often not enough to restore wetlands to reference conditions (Aronson and Galatowitsch 2008, Salaria et al. 2019). These restorations are limited in their success due to depleted seedbanks, isolation from other restored or natural wetlands, and the prolific establishment of invasive species post-restoration (Mulhouse and Galatowitsch 2003, Aronson and Galatowitsch 2008).

Plant community composition can be improved by using more comprehensive restoration techniques including removing sediment, using diverse seed mixes, and/or transplanting live plant or soil material into the restoration site (Lindig-Cisneros and Zedler 2002, Beas et al. 2013, Strehlow et al. 2017, Larson et al. 2020). Excavation works to remove exotic species from the seedbank, allowing for long-buried native seed to emerge post-restoration (Beas et al. 2013, Larson et al. 2020). However, there may not be enough viable native species in the seed bank to produce a diverse wetland site, creating a need to introduce native species by seed (Strehlow et al. 2017, van der Valk and Baalman 2018). Another way to improve plant community composition is by transplanting live material from natural sites to the restoration sites. Transplanting soil from natural wetlands improves species richness and vegetation cover (McKinstry and Anderson 2005), and transplanting live plant material can aid in establishing species that are historically absent in wetland restorations (Fraser and Kindscher 2005).

Although applying multiple restoration methods has helped improve plant community composition, applying a diverse seed mix may be all that is needed to improve the plant community. Two years following a wetland restoration, Strehlow et al. (2017) determined that plant community composition did not differ across three separate restoration strategies: 1) native seed only; 2) native seed and native hay mulch; and 3) native seed, native hay mulch, and transplanted vegetation plugs. To determine whether these results would differ over time, this study looks at how plant community composition differs across these same three treatments 10 years following the restoration of this wetland.

Our study was conducted on a previously cropped, approximately 19-ha wetland at the Albert K. Ekre Grass-land Preserve in Richland County, ND, which is managed by North Dakota State University. Prior to restoration, the site was dominated by introduced species including Typha × glauca (hybrid cattail) and Phalaris arundinacea (reed canarygrass). To prepare the site for restoration, the site was burned, planted with Roundup Ready soybeans, and treated five times with glyphosate (Roundup PowerMax, Monsanto, St. Louis, MO) between May 2013 and July 2014 (Strehlow 2015). To restore the site, a randomized complete block design was then created with approximately 11 ha of the study site split into nine equal plots organized into three rows of three. In late July and early August of 2014, three different vegetative restoration treatments were repeated three times and replicated once per block: 1) native seed only, 2) native seed + native hay mulch, and 3) native seed + native hay mulch + transplanted vegetation plugs (Strehlow 2015).

Following the restoration, vegetation surveys have been conducted every year from 2015 to 2021 (Strehlow 2015, Durant 2020, Polgar 2022). In all instances, there was no difference among treatments for mean relative cover of native species or introduced species, species richness, or species diversity (Durant 2020, Polgar 2022). To determine whether there are more nuanced differences in the plant community composition, we collected vegetative data differently, using Modified-Whitaker plots (Stohlgren et al. 1995) in the summer of 2024. The Modified-Whitaker design detects more rare or patchy plant species, providing a more accurate representation of a site’s species richness (Barnett and Stohlgren 2003). One 20 m × 50 m (1000 m²) Modified-Whitaker plot was randomly assigned within each treatment plot and located using GPS. Nested within each Modified-Whitaker plot, we recorded percent cover of each species in 10 1-m² nonoverlapping subplots, and we recorded species presence in 10 10-m2 nonoverlapping sub-plots, one 100 m² subplot, and in the entire area (1000 m²) of the plot. Our study only looked at 2024 collected data to determine plant community differences among treatments.

We performed all statistical analyses in R version 4.5.1 (R Core Team 2025). To get an in-depth analysis of the plant community composition, we compared total species richness, native species richness, and introduced species richness across all treatments. To understand the effects of the treatments on introduced species, we compared the relative cover of P. arundinacea, Poa pratensis (Kentucky bluegrass), Bromus inermis (smooth brome), and Agrostis gigantea (redtop) across all treatments as these were the most abundant introduced species found within the study site.

Our analyses showed similar results to the prior studies done at this site. We found no differences in total species richness (F_2,6 = 1.45^, p = 0.31), native species richness (F_2,6 = 0.16, p = 0.33), or introduced species richness (F_2,6 = 2.27, p = 0.18) among the three treatments (Figure 1). We also found no differences in the relative cover of P. arundinacea (F_2,6 = 1.75, p = 0.25)^, P. pratensis (F_2,6 = 0.33, p = 0.73), B. inermis (F_2,6 = 1.54^, p = 0.29), or A. gigantea (F_2,6 = 1.10, p = 0.39) among the three treatments (Figure 2).

This study aimed to determine how three different inten-sive restoration techniques affected the plant community 10 years after restoration. Our results showed no difference in species richness or relative cover across the three restoration methods used in this study, which are similar to prior findings at this site. To best understand the lack of significance found in this study, it is important to note the land use history of our study site.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Average species richness (±SE) in a 19-ha wet-land restoration at the Albert K. Ekre Grassland Preserve near Kindred, ND across treatments: native seed only (seed), native seed + native hay mulch (seed + mulch), and native seed + native hay mulch + transplanted vegetation plugs (seed + mulch + plugs). A) total spe-cies richness, B) native species richness, C) introduced species richness. Different lower-case letters indicate significant differences between treatments.

Prior to restoration, our study site was cropped and left fallow in years when crop production was not viable, leading to an increase in introduced and weedy species. Although the site was cropped and herbicide was used to pretreat the site for restoration, introduced species were still prominent after restoration for all three treatments (Durant 2020, Polgar 2022). This remains an issue 10 years after treatment. In all our vegetative restoration treatments, introduced species make up approximately half of the relative cover (Figure 2). This can largely be attributed to P. arundinacea and P. pratensis (Figure 2). Once P. arundinacea is present in a restoration site, it spreads prolifically, taking over the site and outcompeting native sedge species (Budelsky and Galatowitsch 2000, Green and Galatowitsch 2001). Poa pratensis also outcompetes native species, often hindering restoration success (Toledo et al. 2014, DeKeyser et al. 2015).

It often takes years after restoration to show improve-ments in the plant community (Aronson and Galatowitsch 2008). However, in areas with high levels of introduced species, it is more difficult for native species and plant guilds to colonize effectively over time. The high level of invasion by P. pratensis and P. arundinacea in this wetland may be contributing to the slow expansion of native species, limiting the space and nutrients they need to prosper (Mulhouse and Galatowitsch 2003). Once established, these introduced species create a positive feedback loop in favor of P. arundinacea and other introduced species, continually increasing introduced species colonization while limiting the establishment of native species (Farrer and Goldberg 2009, Galatowitsch and Bohnen 2021).

This site has been actively managed since its initial restoration in 2014 with grazing and prescribed fire to control introduced species. Adaptive management and control of invasives species early in the restoration process is critical to reduce the establishment of fast-spreading weedy species such as P. arundinacea (Green and Galatowitsch 2001, Galatowitsch and Bohnen 2021). With introduced species occupying much of the restoration site, a more adaptive management plan may need to be implemented to achieve more native species colonization (Galatowitsch and Bohnen 2021). Grazing and prescribed fire did not seem to enhance any vegetative restoration method used in this study since we found no differences between treatments 10 years after restoration. However, although not significant, P. arundinacea relative cover was highest in the seed only treatment, indicating added mulch or vegetation plugs during restoration may inhibit the spread of P. arundinacea when the site is actively managed post-restoration. Three months before the data were collected for this study, the site was burned with prescribed fire in the spring of 2024. Prescribed fires, when used in conjunction with grazing, can improve and maintain habitat heterogeneity (Fuhlendorf and Engle 2004). While this site was still invaded by introduced species, native species richness was high across all treatments (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Average relative cover (±SE) of introduced species in a 19-ha wetland restoration at the Albert K. Ekre Grassland Preserve near Kindred, ND across treatments: native seed only (seed), native seed + native hay mulch (seed + mulch), and native seed + native hay mulch + transplanted vegetation plugs (seed + mulch + plugs). A) Phalaris arundinacea cover, B) Poa pratensis cover, C) Bromus inermis cover, D) Agrostis gigantea cover. Different lower-case letters indicate significant differences between treatments.

Because we found no differences among treatments, the most cost-effective vegetative restoration treatment was to use native seed only (Strehlow 2017). Typically, researchers have concluded utilizing additional advanced restoration techniques improves plant community composition and overall restoration success throughout the PPR (Galatowitsch 2006, van der Valk and Baalman 2018). However, as we found no differences among restoration treatments, our results suggest that certain depleted wetlands may be restored to a more diverse, natural state without spending excessive time or money on restoration treatments beyond seeding with natives.