The Fellows and I traveled to southwest Minnesota (Windom, MN) this week for the 2023 Grassland Restoration Network’s annual workshop. This version was hosted by the Minnesota Department of Natural Resources and spearheaded by Jeff Zajac. We looked at five prairie restoration sites, each of which stimulated conversation and provided helpful lessons. I’ll try to share some of my bigger takeaways in this post.

If you’re interested, please check out Bill Kleiman’s summary of the meeting at the GRN blog site. Following that blog is also a great way to find out about next year’s meeting, as well as to get a wide range of valuable and practical information on prairie restoration work.

The point of these workshops is to bring together people from around the country to compare notes and learn from each other. To that end, we didn’t just want to see examples of where everything had gone right. How much can you learn from that, after all? Fortunately, Jeff and his colleagues were willing to share the full range of their experiences, including some gorgeous prairies and some that hadn’t yet lived up to expectations.

As an example, Jeff took us to a site where he’d been planning to use herbicide to kill smooth brome before re-seeding an old Conservation Reserve Program (CRP) planting with a diverse prairie mix. Because of various factors, including abundant gopher mounds, discing issues, soil erosion potential, unrelenting wet weather, and worries about seed viability over time, he wasn’t able to get the spraying done. Faced with difficult choices, he went ahead and planted the seed. You can probably guess what kinds of issues he faces now.

While it might sound silly to plant expensive seed into a site you don’t feel is ready, I think Jeff made a very sensible decision, given the difficult circumstances he faced. Most of us who have been restoring prairies for a long time have faced similar challenges. Plus, the site isn’t a total disaster. There are parts that look like the above photo, where plant diversity is relatively low and smooth brome and other invasives are abundant. But there are other parts of the site, especially on higher, drier areas, that are much more diverse and look promising.

Jeff’s plan is to try to incrementally increase the plant diversity and habitat quality of the site over time. He figures it’s already better than it was, since a major project goal was to create grassland bird and insect habitat. The alternate choice would be to start over. That’s always a difficult choice, and rarely an easy one. Do you push forward with a site that will need frequent and intensive management to deal with invasives and/or increase plant diversity? Or do you cut your losses and hope that by starting over you can spend a bunch of time and money briefly, in return for much less expensive and time consuming stewardship needs afterward? This site provided an excellent platform to contemplate and discuss those kind of choices.

Another frequent topic of discussion during the two-day workshop had to do with how much, and what species of grass should be in seed mixes. This has been an intriguing topic that has evolved considerably over the last 21 years of this workshop series. There’s a decided trend toward using lighter seeding rates of grasses. Data and lots of experience show that if you put too much grass seed in a mix, it doesn’t much matter how much forb (wildflower) seed you use because the grasses will end up dominating anyway.

What kind of grass to use, though, is also a big topic. Traditionally, big bluestem, Indiangrass, and switchgrass were major components of seed mixes. There are various reasons for that, including the price and availability of seed and the quick establishment and space-filling ability of those grasses. However, many restoration projects have more recently chosen to greatly reduce, or even eliminate seed from those big, strongly-rhizomatous grasses in planting efforts. Instead, they focus on shorter species and bunchgrasses, including prairie dropseed, little bluestem, sideoats grama, and others.

Even when using these other species, grass seeding rates are still quite a bit lower in many of today’s plantings than they were in the past. On the flip side, forb seeding rates are often higher. Using lots of forb seed can be helpful, not only in ensuring a nice (and quick-to-establish) diverse plant community, but also to help outcompete weedy plants during the early years of a project. The weed competition aspect seems to be especially important in more productive soils in the more eastern parts of the tallgrass prairie region.

Speaking of controlling weeds during the early years of plantings, we had some good discussion about the necessity of mowing. In Nebraska, most of our high-diversity prairie restoration plantings use relatively low seeding rates and we don’t typically mow weeds. Further east, mowing weeds has been seen as more of a necessity, mostly because soils there support stronger, more dense growth that can suppress new prairie seedlings.

It sounded like mowing has been a frequent part of the Minnesota DNR’s restoration process, but a recent study might change that. Dr. Brian Wilsey from Iowa State University was on hand to talk about an experiment he helped the DNR staff with. One of the results was that weed mowing seemed not to have any significant impact on prairie establishment.

Does that mean no one in southwest Minnesota should mow weeds anymore during prairie plantings? Probably not, but it’s always helpful to check to see if common practice is actually the best practice. In this case, think of all the hours local DNR staff can save by not having to drive tractors/mowers across each site multiple times!

The topic of mowing and weeds was an excellent example of the kind of practices that vary in necessity and effectiveness from place to place. There are numerous examples of restoration and stewardship techniques that work well in some locations but not others. Sometimes, that variance is explained by precipitation patterns and other factors tied to geography. Other times, two sites close to each other might face very different weed problems just because of local differences in soils and the kinds of weed legacy found at each site.

Not only do the species of weeds and their abundance differ between places, the way stewards need to deal with them can too. Canada thistle is a good example of this. The DNR site managers leading this workshop said Canada thistle usually doesn’t need much control in the restorations. It is often abundant soon after prairies are planted, but it tends to diminish over time with or without mowing.

I’ve seen other sites in places like Illinois and Wisconsin where the Canada thistle story is similar, though sometimes they feel like mowing during the first couple years is helpful. In the sites I’m familiar with in Nebraska, Canada thistle expands dramatically, even in remnant prairie, without aggressive suppression, usually with herbicide. It’s flabbergasting to me that it’s so tame further east. Good for them, I guess.

On the other hand, I don’t worry about sweet clover or Canada goldenrod in our prairies – restored or otherwise. Sweet clover comes and goes without changing plant diversity and our cattle grazing helps suppress it during its flowering years. Canada goldenrod can be abundant in young restorations, but tends to become less common over time. That’s very different from how those species seem to act in some other locations. Good for us!

We talked for a while about Canada goldenrod because it is seen as a real threat in many locations (most sites in Illinois, for example). There were very different predications among the group about how the current amount of goldenrod in the DNR plantings we saw would change over time. Most of the locals, who probably know best, weren’t concerned. They were more focused on other invasives like crown vetch, birdsfoot trefoil, and wild parsnip.

All of this reinforces a major theme of these GRN workshops over the years – every site is different. There are certainly some commonalities, however. For example, years ago, there was unanimous agreement among workshop participants that a dormant season broadcast seeding on a recently-harvested Roundup-ready soybean field is a consistently successful approach. That doesn’t mean other approaches can’t also be successful, but it was striking to find so much agreement on that particular one. Also, I wonder how many hyphens that previous sentence should actually have (dormant season, broadcast seeding, recently harvested, Roundup ready, consistently successful). The world will never know.

The variation between sites, though, even within a fairly small geographic area, is really important. To me, it highlights the value of starting small on any new project, and of trying a variety of approaches during that early experimental period. Lessons learned from trial plots can inform larger efforts that follow and it’s relatively painless to start over when something goes catastrophically wrong on a small project. We don’t always have the option of building restoration projects incrementally and evolving our tactics as we go. It sure seems like a smart approach, though, when possible.

I’m very grateful to Jeff and all of the Minnesota DNR staff who helped put on a terrific workshop this year. I also appreciate the disparate ideas and experiences shared by the participants, all of whom made our conversations thought-provoking and constructive. I’ve included only a few highlights from those conversations. I’ll surely weave other bits into future posts.

Maybe I’ll see you at a future GRN workshop. We’re looking at Missouri, Kansas, Illinois, and Indiana sites as potential hosts over the next few years. Regardless, stay tuned to this platform and the wonderful GRN blog to keep the discussion going.