Restoring Cottonwoods to the Missouri River (Part 2)

Back in February I wrote a short post about our cottonwood tree planting efforts on the Missouri River.  Here’s a more detailed description of that project, including some early (and interesting) results.

Tyler Janke, our Missouri River restoration ecologist, holds a young cottonwood seedling from our demonstration project’s first season.

The project started as some informal discussions with Scott Josiah, the state forester for the Nebraska Forest Service.   I mentioned to Scott my concerns that most of our cottonwood woodlands on Nebraska’s big rivers consist primarily of mature trees, and that we aren’t seeing many new cottonwood stands becoming established.  As existing cottonwood trees die of old age, they’ll be replaced by the trees that are now filling the understory of the woodlands, including species such as ash, mulberry, hackberry, and eastern redcedar.  Those are all fine trees, but provide a very different kind of habitat than a cottonwood woodland.  Scott was interested in the same issue and, a few months later, approached me with an opportunity to join them on a grant application to fund a collaborative project.

We successfully obtained a grant from the U.S. Forest Service for a three year demonstration project, and the National Arbor Day Foundation jumped in with additional funding and support.  Tyler Janke, our restoration ecologist along the Missouri River is spearheading the project, and will be restoring approximately 300 acres of cottonwood woodland over the next three years.  Along the way, we’ll be trying and comparing a number of techniques to figure out what works best, including:

1. Planting cuttings and rooted seedlings.  Directly planting young trees into the ground is the surest way to establish a woodland, but is also expensive, time-consuming, and difficult to do at large scales.  We expect that cuttings (the harvested tips of young branches with buds on them) may survive less well than actual rooted seedlings, but they’re cheaper and easier to deal with.  We’ll compare the two methods over three years and multiple sites.

2. Disking ground near existing mature cottonwoods.  We hope to take advantage of the abundant seed that falls from existing trees by preparing good seed beds for them to land on and germinate in.  Cottonwoods establish in moist bare soil.  We can create the bare soil with disking, and then hope that the moisture will come from a combination of high groundwater (in some sites) and rainfall. 

3. Irrigation.  Rainfall may or may not come at the right times or amounts to stimulate seed germination in disked sites or to ensure survival of our tree plantings.  Irrigation will allow us to narrow down the causes for good or poor tree establishment by controlling the amount of moisture available for young trees – at least in small areas.  In addition, if irrigation turns out to make a really big difference in survival and establishment, there are significant areas of land along the Missouri River that have center pivot irrigation systems in place, and if those areas are converted to natural habitats in the future, those irrigation systems could be put to use. 

4. Herbaceous seedings.  Starting next year, we’ll be planting seed of grasses, sedges, and forbs within some of our cottonwood plantings.  We want to see whether or not we can establish a sedge meadow-type community in the understory of developing cottonwood woodlands.  If so, we might be able to avoid the more typical understory community of reed canarygrass and a few other invasive or monoculture-forming plant species.  This is kind of a wild card portion of the project, but I think it’s an important one.  Most woodland restoration projects focus solely on getting trees started, and I think there’s a danger that we might be missing our best (only?) opportunity to create a quality herbaceous vegetation layer beneath those trees by waiting until the trees are relatively mature before starting to worry about what’s growing beneath them.

Our demonstration project is taking place on private lands that are enrolled in the Wetland Reserve Program (WRP) but where converting cropland to grassland hasn’t worked very well for various reasons.  All of them were underwater last year during the big flood, and there are a number of sites that have been flooded at least a couple times since being planted to prairie/wetland vegetation.  It might turn out that cottonwood establishment is the best fit for land where flooding is too frequent or severe to establish and maintain grassland habitat.  If so, we hope we can help that effort by developing and testing techniques that can make that conversion as efficiently and effectively as possible.

This spring, we harvested and planted about 30,000 cuttings (see the February post) in addition to about 1,000 rooted seedlings.  Most of those were planted in early May with a mechanical tree planter, but some were hand-planted by volunteers from the National Arbor Day Foundation.  At about the same time, Tyler had a contractor disk up about 100 acres across four different properties.  In mid May, Tyler started irrigating 2 acres of disked land at one site with a big “traveling gun” irrigation system (used commonly on athletic fields – it has a sprinkler head on a long hose and a reel that winds the hose in as it waters). 

In this photo, Tyler is using an ATV to pull the sprinkler head and hose out from the reel. Once it’s stretched out, the pump is turned on and the sprinkler head is slowly pulled back to the reel as it waters. This allows a long area to be watered.


It’s still pretty early to start analyzing the results of the project, and we’ve got two more seasons to go.  However, we have seen some interesting things already, and the very warm and dry weather this year has provided us with an interesting opportunity to see what happens on the extreme end (we hope) of the moisture continuum.  Tyler has been watching the sites closely all season, and collecting a little data.  More extensive data will be collected in the coming weeks to help us analzye our first year results.  However, I had a chance to visit the sites with Tyler a few weeks ago, and can talk about what we saw at that point.

First, in this dry year, irrigation made a huge difference in the establishment and survival of trees.  Between May 15 and August 1, Tyler figures he put a little more than 18 inches of “rainfall” on the two acres he irrigated.  The rest of the sites only got about three and a half inches of rain during that same time period – most of that coming in a single rainfall event.  Within the disked areas that were irrigated, germination from seed falling from nearby trees was very successful.  Tyler’s winding down the irrigation now, so we’ll see how those young seedlings survive through the winter and into next year.

Irrigation in progress.


Flags mark some of the seedlings that established on their own in a disked area under irrigation.


A young seedling that established and grew under irrigation.

Second, we are seeing surprisingly good establishment of cottonwood seedlings in some portions of the disked/non-irrigated areas.  Weed growth is very strong in most of the sites, so it’s hard to see the ground very well in many places.  However, even accounting for that, it appears that the best seedling establishment is happening in places with sparse weed growth.  What’s interesting about that is that those areas of sparse weeds appear to be in soils with the least organic matter – and most prone to drought stress.  The fact that the cottonwood seedlings seem to be thriving where most annual weeds can’t is really intriguing.  Tyler’s going to do some more thorough searching to confirm (or deny!) what we were seeing a few weeks ago, and will also have a soil scientist help him evaluate the soil properties where cottonwoods did and didn’t establish well.

Third, because of the tall weed growth in all four sites, Tyler mowed some plots in both the irrigated and non-irrigated portion of one site in early July.  So far, it’s too soon to see whether or not that made a difference, but it gives us something else to look at.  Tyler thinks there maybe a difference in seedling height between the mowed and unmowed portion of the disked/irrigated site, but whether or not that will affect long-term survival is unknown.  Dealing with weeds may turn out to be a major challenge going forward.  Tyler also sprayed some of the disked areas with Roundup – as well as the tree planting sites – early this spring, and that did seem to keep the weed growth suppressed somewhat, but we don’t know yet whether it was significant or not.

Some of the sites have impressive stands of weeds – especially considering the low rainfall amounts for this year.

Finally, we saw a big difference between the survival of rooted seedlings and cuttings.  Outside the irrigated area, most of the rooted seedlings are alive and growing, but very few of the cuttings are.  Within the irrigated area, Tyler planted two small groups of cuttings – a group of 10 larger diameter cuttings and a group of 10 smaller diameter cuttings.  So far, four out of the 10 big cuttings are alive and two out of the 10 small cuttings are alive.  The vast majority of the rooted seedlings under irrigation have survived.

A rooted seedling in an non-irrigated portion of a site. Most of these seedlings appeared to be surviving at this point.


A cutting that sprouted and survived at least the first half of the hot dry summer.

Very likely, the low survival of cuttings is due to the drought, but Tyler has a few other ideas to pursue before next year’s plantings.  He’s wondering whether the warm dry winter affected the bud development on the branches we harvested for cuttings, and if so, whether we should have harvested the cuttings earlier in the season than we did. 

Tyler also has some ideas about how to harvest higher quality cuttings next year – or at least to ensure more uniform quality.  This year, we just harvested cuttings from the branches of small trees, but next year Tyler wants to harvest cuttings from “stool beds” created by cutting those small trees down this year and letting them create clusters of new stems from the ground.  Harvesting those new stems will help ensure that we’re getting only one-year-old stems.  He also wants to be more uniform about the length and diameter of cuttings we harvest, because that may be making a difference as well.  Hopefully we can improve upon this first year’s survival rates by improving and standardizing cutting quality.  A little rain would help too.


Overall, it was a great first year.  The nice thing about a project like this is that because it’s an experiment/demonstration project, failure is still success.  As the saying goes, we’ll learn more from our failures from our successes – and this first year provided us with an excellent opportunity to see failures!  Of course, we also had some important successes and have the trees to show for it.  In addition, the drought provided us an with an important opportunity to see and analyze where cottonwood seeds established on their own in our disked areas – under very stressful conditions.  We hope to learn from that and incorporate those lessons into what we try in years two and three. 

Tyler Janke celebrating our first year successes.

As Tyler continues to evaluate this first year, I’ll let you know if we find anything else interesting.  Otherwise, stay tuned for year two!

20 thoughts on “Restoring Cottonwoods to the Missouri River (Part 2)

  1. Chris, I’m curious, how far apart were the cuttings planted? I’ve heard that sometimes cuttings are dipped in a rooting compound to promote growth. Is that true, and was it done on any of these cutting?

  2. It seems odd to me that you are intentionally trying to establish cottonwoods. They appeared everywhere early on in local prairie restorations. This was actually quite a concern until the first prescribed burn seemed to have taken care of the problem. They are also a constant nuisance in pots and trays I use to grow native plants.

    Cottonwoods seem to be a disturbance adapted species. Will not some epic flood detroy the entire mature cottonwood woodland along with all invading trees only for the process to start all over again? If not, are your efforts to restore cottonwood woodlands appropriate to the site? Or are your efforts more of an attempt to remediate conditons caused by changing hydrology (dams, lower water tables, etc.)? It seems you are substituting a man-made disturbance (old cultivated fields) for a natural disturbance (floods). If the floods never return because the hydrology has been changed, will the cottonwoods in your restoration area meet the same fate as those in the cottonwood woodlands?


    • James, it’s a question we’ve talked a lot about. On the surface, it does seem ironic that we’re spending time and money to establish a species that often shows up where it’s not wanted. I’ve spent a lot of time removing them from wetland restorations on the Platte River. Unfortunately, the species’ habit of episodic establishment means that’s it’s difficult to get it to show up in abundance where we’d like to have it (and where we think future conditions will favor it). Once the trees get some size to them (5-7 years?), they’ve got an excellent chance of surviving most flood events that occur where we’re trying to establish them. Tyler and others have done excellent modeling of floodplain conditions, including soils, groundwater, and flood frequency/intensity, and we’re choosing sites that we feel pretty confident will allow the cottonwoods to survive.

      To your last two questions: Yes, we’re substituting man-made disturbance for natural disturbance. It doesn’t look like the Missouri will be returning to 1700’s condition, so we need to make some decisions about how to manage the floodplain under what appears to be its new regime. The decisions aren’t easy. And to your last question, if the floods don’t return, these cottonwoods we’re establishing should survive just fine. They won’t reproduce successfully on their own very well, but if we can figure out how to guide that process along, we should be able to create new woodlands in a staggered way through time and maintain a good diversity of stand ages across the floodplain. (And if we can do it simply by disking, the cost will be pretty low) Lots to learn and do over the next decades…

      • For me, the idea of trying to establish cottonwoods is a hard one to love. I think it might be easier if I was familar with a high quality cottonwood woodland. I’ve always admired the resiliency of the cottonwoods in the dunes around lake Michigan. If I had pictures of a high quality cottonwood woodland in Nebraska, along with information about the reason this habitat is a conservation priority, and other expected benefits, then maybe I could see through your eyes.


    • James,

      You are right, Cottonwoods are a disturbance adapted species. They essentially require some type of disturbance event to facilitate large scale establishement. The former crop fields where prairie restorations take place provide favorable conditions for cottonwood germination early on (lots of bare soil and sunlight), but, as the sites establish, cottonwood germination suitability sharply declines.

      In planning this project, we were very careful to select sites where cottonwood woodland/forest is the most suitable community. We researched old photos and maps, as well as evaluated each sites modeled flood frequency, flood duration, soils and landscape position. Additionally, most of these sites were initially planted with a grassland seed mix 8 to 10 years ago, but repeated flooding either killed any established grassland, or didn’t permit it to establish. They were essentially non-functioning habitats.

      We are substituting a man-made disturbance for a natural disturbance because the natural disturbance regime has been modified by humans. In Southeast Nebraska, the Missouri river still floods regularly, but the timing of these floods has changed and is not always in sync with cottonwood seedfall. Furthermore, erosion and deposition of sediment by floods has been nearly eliminated by channelization of the river, severly limiting the extent and spatial distribution of “natural” germination sites.

  3. Great summary – thanks so much for sharing your efforts. I appreciated your comment above regarding decision making under new regimes. It is a difficult task to balance historic landscapes and current realities!

  4. Are not cottonwoods a larger water user than prairie plants and hence will be a contributor to dewatering these sites even further? Stressing wetland dependent species even further for some eastern birds and bugs that are restricted to current riparian woodlands in NE but common a hundred miles east.

    • Gus – it’s an interesting point. I don’t know for sure what the relative water use is – and it would obviously vary by the density and age of the trees, etc. Dewatering wetlands is an important issue, of course, but I also think there’s value in cottonwood woodland habitat. Balancing those and all the other factors is the challenge, right? I also don’t know the Missouri system well enough to tell you how important the dewatering issue might be. It’s pretty different from the Platte wetlands I work with and know better. Maybe Tyler can chime in here and add his perspective.

    • Gus,

      Although these project sites are taking place on “Wetland Reserve Program” easements, no substantial marsh, wet meadow, or other typical wetland communities exist on these sites. In some cases, attempts to create traditional wetland habitat has been made by making shallow excavations, but nearly all of them have failed. It’s entirely likely that cottonwoods use more water than other species since they typically grow in places where flooding and high water tables are common. However, groundwater monitoring efforts along the Missouri consistantly report that water table depth and fluctuation are highly correlated to river stage, which is governed by runoff, tributary flows, and of course, reservoir operations. I’m not aware of any groundwater monitoring efforts that have compared forested and non-forested areas. It’s an interesting question that would be worth pursuing.

  5. I hope I’m wrong, but it seems I’m detecting some antipathy toward this fully native and quite naturally occurring tree, which we know has been along the Platte and all other major prairie streams for as long as anyone can imagine. Aesthetically, I love a grand old cottonwood as much as an old bur oak , but I must admit I have almost no knowledge of why cottonwoods are important ecologically, or what biological diversity they support.
    We started to talk about this just a little bit when I visited up there, but were interrupted and didn’t go far with it. Chris, do you have any good links to help dilute my ignorance about cottonwood ecology?

    • James,

      Here a paper on the ecology of Cottonwoods.

      Click to access Life_History_Ecology_and_Conservation_of_Riparian_Cottonwoods_in_North_America.pdf

      Probably the best quote I saw from this paper follows.

      Conservation and Restoration of Riparian Cottonwood Forests
      -For example, although riparian vegetation occurs on less than 1% of the western North American landscape, it provides habitat for more bird species than all other habitat combined.

      It’s not that I don’t like cottonwoods. I just really think the loss of the loess hill ecosytem around the Missouri River is a greater priority. When I revisited Nebraska I could not believe how much the loess hills have changed in just over a decades. Where a National Treasure used to exist, now there is only a continuous scrub forest of Eastern Red Cedar trees. :( I wonder if the people living in the area have even noticed the change. It’s like the old tail of the gold fish in water that slowly boils. It was very apparent to me because I left then came back for a visit years later.


      • James — I hear you, man! It’s the same sad story in the “glades” of the Ozarks and other southern hills; all going to cedars where not being managed. And to hear Chris tell it, same sort of thing out there around Niobrara, too. But of course, Chris works along the Platte, not in the Loess Hills, nor normally out at Niobrara, so it makes sense to me for there to be some cottonwood component to his restoration activities. – The other James

        • James, Chris works along the Platte. However, according to the post this project is occurring along the Missouri.

          Speaking of conserving trees around the Platte River … I nominate the bluffs above Fishery Road in Gretna, NE. There are some nice Oak trees on those bluffs. The land has either been abused by over grazing and invaded by cedars. However, it still has the bones for a great Savannah restoration. It would be a great place to demonstrate the benefit of management. I think it is a plus that it is right along the busiest travel corridor in the state. It is highly visible from the road and people would stop just to take a break from driving. It is also midway between the two most populous cities in the state. Restoring that location would be a great ambassador for the benefits of ecological management.

          James M.

  6. “What’s interesting about that is that those areas of sparse weeds appear to be in soils with the least organic matter – and most prone to drought stress. The fact that the cottonwood seedlings seem to be thriving where most annual weeds can’t is really intriguing.”

    Since it’s a player in early dune succession I don’t see that as surprising at all!

  7. Many years ago, I contacted the federal caretakers of land along the Missouri Breaks in Montana. I had read articles about the woodhawks such as Jeremiah Johnson that had cut down vast areas of cottonwoods to supply the steamboats coming up the Missouri River. When we canoed this area, there were almost no large cottonwoods except on the islands in the river. I offered to organize a trip to plant cottonwoods along the river banks since they had obviously never returned. They said the reason the trees had never returned was that cattle ate the seedlings. Although the banks were protected by the Wild and Scenic River designation, the area was certainly overgrazed. Was I given good information? Cattle obviously don’t relish the cedar but the buffalo herds on the Nature Conservancy Preserve seemed to do a fine job of keeping them off sections of the preserve where they were allowed to graze.

    • Duke – I don’t know the ecology of the upper Missouri enough to confirm or deny your information. Cattle certainly do feed on young cottonwoods, though in my area they eat them only now and then – when they’re looking for some particular addition to their diet, perhaps. We can get cattle to eat cottonwoods more dependably if we put in a high enough stocking rate. If the area was being overgrazed, cottonwoods probably wouldn’t have had much of a chance to grow. Cedars are a more interesting case. I’ve never seen cattle eating cedar trees, but it does sometimes seem that there are fewer in grazed areas than ungrazed (though the opposite can also be true). Not sure if cattle alter competition in a way that works against cedars? I would have thought the opposite, but I’ve never figured it out…

  8. This is an interesting project and since I am unfamiliar with this locale, I can only compare this to the region in which I work and am most knowledgeable. In the Texas Plains, Plains cottonwood was the primary native tree creating ribbons of green winding in riparian areas between expanses of short and midgrass prairie. These historic cottonwoods provide cover for many animals and are especailly important as turkey roosts. With heavy grazing pressure we see the same situation as described, old giant cottonwoods with little regeneration or understory. Some of the incentive programs provided cost share to landowners to fence off riparian areas and deferments to reduce the grazing pressure allowing natural regeneration. In some cases, watering facilities were installed to distribute cattle over areas that may not have been utilized in grazing rotations. Cottonwood pole plantings are another method that speeds up the process by quite a few years. If you have a good source available, you can jump start the cottonwood restoration by cutting 7-8 foot “broom stick” diameter poles during dormancy and planting those using an auger to dig very deep holes close to the stream side water level. Success rate depends upon many factors such as quality of the pole stock, length of time between cutting and planting, water availablility, etc. High survival rates (60-70%) are possible if all conditions are met and nature cooperates. Again, the need for maintaining roosting grounds for native birds was the primary goal and removing water sapping invasive species such as salt cedar and Russian olive was necessary to maintain stream flow.
    Good work Chris, I enjoy your site.

    • Vicki,

      I have been thinking about planting some cottonwood poles. I would love to speak to you to learn more about your methods and success rates.


  9. Pingback: Best of Prairie Ecologist Photos – 2012 | The Prairie Ecologist

Leave a Reply to James McGee Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.