Brush Puzzles in Somme Prairie Nature Preserve

Draft - August 2020

Accentuate the Positive

The concerns raised in this report should not detract from appreciation for the good and valuable work done by the contractors, who reduced the dominance of brush over large areas without damaging the prairie intermixed with the brush in any observable way in most areas. 

Eliminate the Negative

But where apparent problems emerge, especially in regard to high-quality, remnant natural areas, we stewards strive to understand the causes and develop improved techniques and safeguards as best we can.  

 

Background

Trees and brush in Somme Prairie have been fought mostly by volunteers for four decades. In recent years, Forest Preserve staff has hired contractors to speed up the process; the last of the large brush and trees should all be removed before the end of September. 

But the small high-quality areas have had their edges nipped at by short, dense brush, mostly gray dogwood, still, after all these years. The dogwood spreads in clones, with stems connected by roots.  

Foreground (above) shows dogwood with prairie species, stunted but surviving.

Top of photo is Grade A prairie.

For decades the dogwood areas have languished. They have not, as hoped, responded to management by recovering the health and diversity that's found in those very-high-quality areas nearby. Instead, portions of them burn, recover somewhat, degrade again as the brush increases, on and on. Yet these repeatedly injured areas have the best potential in the short run to increase the size of the adjacent small high-quality prairie that represents the heart of the biodiversity of this site. The Grade A diverse populations of plants,  animals, and soil biota need room to spread. 

In 2019, the Somme Seminar group was offered special grant funding by three foundations, initially to support our work in Somme Woods. But Somme Prairie Nature Preserve seemed to be a very high priority. The Seminar group approached the Forest Preserve staff and stewards Laurel Ross and Lisa Musgrave, offering to use some of that funding to hire contractors to do jobs that professionals could perhaps do better than we volunteers. Friends of the Forest Preserves served as fiscal agent. We contracted with two restoration companies, both of which came highly recommended for their willingness to use great care in and around high-quality areas. They each proposed a different strategy. That seemed good to us, as we could compare them.

For Area A – contractor #1 chose to apply 20% Garlon 4 to stems (“basal barking”) in winter. 

For Area B – contractor #2 chose to cut stems and apply 20% Garlon 4 to the cut stumps in winter.  

Observations on the work in summer 2020:

1. Work seemed most successfully done in the highest quality areas where dogwood was small and sparse, indicated this summer by occasional dead woody stems. This work seemed to have been effective without damage to nearby vegetation. That seemed also true in lesser quality areas nearby, wherever brush was sparse. Most of the work falls in this "successful" category.

2. Other work was done where brush was denser. In some of these areas, the brush had already shaded out the prairie vegetation, so now in early summer, dead woody stems stood over mostly bare ground and moss. Some of these areas were substantial; one was six feet wide and forty feet long. 

3. Typically in dense dogwood, the treatment was effective at killing most of the taller stems, but in every square meter one or a few stems survived, some large and some small. New shoots emerged in June. Red dye was visible on some of the surviving stems, indicating that the problem was not that they’d been missed. 

4. In other areas, quality vegetation had survived under the relatively dense brush up until the herbicide treatment. But here, in many cases, the herb-layer species seemed to be dead or much reduced. Similar untreated areas nearby showed considerable undergrowth of sedges, grasses, and forbs. Small areas that had been missed (no red dye on stems) also have a surviving understory despite dogwood stems as tall and dense as in the deadzone areas.  In sum, this evidence points to translocated herbicide (rather than shade) killing off the prairie herbs in these areas. 

5. Some quality species survived in most of the treated areas, including those now looking rather bare. These included Baptisia leucophaea, Platanthera leucophaea, and Lithospermum canescens. Other species surviving there (and now likely to act aggressively in this weakened community) included Helianthus grosseserratus, Rosa sp., Rubus sp., and Solidago altissima.

  

In the photo above, most dogwood stems are dead. Bare ground and moss suggest that the prairie vegetation had been shaded out here. Most vegetation appearing in June this year consists of young weedy species. A few mature dogwood stems survive along with small new dogwood shoots. 

In other nearby areas (below) the presence of last year’s thatch suggests that the prairie survived the dogwood, but not  the herbicide. Parts of Area A now consist of dead dogwood stems standing over dead herb vegetation. 

For another example, the photo below shows such an area on June 8. Most dogwood stems are standing and dead. Stems that survived the cutting along with new shoots that emerged this spring have been cut and their stems treated with glyphosate. More than half of the herb vegetation in this areas appears dead.   It would have been good if the approach taken in Area B (by contractor #2) was better, but unfortunately it was not. In many of the areas where the dogwood was cut and the stumps painted, there were also areas of mostly dead vegetation around the stumps (see below).

 

Original prairie is precious beyond measure. Rhizomatous prairie species may reinvade the dead areas, and other prairie species seeded in fall 2020 may help recovery starting in 2021. But the adjacent untreated dogwood stems may also reinvade - or re-invigorate weakened roots. We do not want to kill prairie plants while killing the dogwood. But dense brush also kills prairie, so leaving it is also not an option.

  

Next steps recommended by the two contractors

 
When asked for recommendations, the contractors did not initially offer insights on the dead zones, but they did propose the following:
 
Contractor for Area A: Stay out of these sensitive areas during the growing season. Repeat basal bark treatments on living stems next winter. 
 
Contractor for Area B:  Follow up in August, cut living stems and herbicide before the plants have time to translocate most resources to the roots. 
 
Both of these proposals seemed worth considering. 

 

Next steps taken in July 2020

 
The stewards decided to ignore the surviving brush in the high-quality areas until late fall or winter, as growing-season trampling in these areas seemed not worth it. 
 
For dead zones (where there was less concern about trampling, as the ground was now mostly bare), there was concern about the dogwood root systems recovering, if untreated stems and this year’s re-sprouting stems (as in photos above) were left to photosynthesize over the summer. We decided to treat the re-sprouts and missed stems in and near the dead zones. That is, we cut stems in the dead zones and others near enough that they could be cut while we stood in the dead zones, leaving the buckthorn, ash, and other species until next fall or winter. The dogwood recovery threat seemed to be especially serious as few other plants would be competing for water or nutrients. 

When we cut the live stems, we treated them with 55% AquaNeat, applied with absorbent-cloth-tipped wands. AquaNeat (glyphosate) is said to have little or no impact through soil spread. As some aggressive species seemed poised to “take over” these empty areas, we also cut and painted some of them (Solidago altissima, Helianthus grosseseratus, and Rubus flagellaris). The plan is to seed these areas with locally gathered seed in fall 2020.
 
Belatedly, as we continued to search for answers, we decided to compare our problem areas with an area nearby cut by the volunteers last winter. To our surprise, it looked similar, with dead vegetation from the previous year often surrounding the cut stumps (see photo, below).  When cutting brush among high-quality prairie, is this a more widespread problem, and we just hadn’t been watching closely enough? 

Suggestions to consider in winter 2021

We are reaching out for comments and suggestions. For a start, some experienced prairie stewards and managers offer suggestions below. They agreed that one important principle is to treat the whole dogwood clone at once if possible (rather than treating scattered parts of it).

Dave Bart of Stantec studied the results and was puzzled. In his experience, it should have been possible to control the dogwood without substantial harm to the nearby vegetation. He initially suggested that basal barking with 20% Garlon would be the best approach. But he was concerned about the very large numbers of very small stems in some areas and later suggested that to cut and paint, meticulously, might be best in those areas. He judged that three years of dormant season treatment would be sufficient to reduce the dogwood to the point that fire alone would control it adequately.

After additional research, Dave made a summary recommendation:

First of all, thanks again for reaching out to me on this. Not only was it important that this was noticed, but that you are working with others to try to figure out why. From what you said, it sounds like both contractors approached the work with sensitivity and applied the herbicide in a careful manner. You raised a good question, is this happening elsewhere, have we not paid close enough attention? What is most alarming to me is that these effects were noticed in the middle of summer several months after a dormant season application. I suspect that the conditions observed at the site are more related to residual herbicide rather than application technique. A lot of factors here but a few things to consider:

Herbicides - some (likely including yourself) are using Glyphosate or Triclopyr amine (3A) mixed with water to conduct cut stump treatment - above freezing - these products are known to break down quickly in soil. Triclopyr ester (4E) products are not all created equal. Petroleum based vs. plant based carriers, etc. - but it's difficult to figure out what the inert ingredients are because companies in the US do not have to disclose it. 

In this case, it appears that the herbicide has not broken down, but why? Could concentrations be high enough in the roots, and when adjacent plants with fibrous roots mature and grow through these root channels, could they be impacted? - Sounds like a long shot?
Are particular soils and favorable drainage conditions required to break down the herbicide? We know that substances can remain in root pores longer in particular soils, like clay.

Again, I believe this is somehow related to herbicide in soil/roots and the lack of breakdown before the growing season. I would propose the following:

Brushcut plants to a height of 2" in late fall (late October - November) when most natives are dormant, utilize a sponge applicator to carefully apply Glyphosate at 50% mixed with water.
Complete work with the following conditions:
Temperatures above 35F, soil temperatures above 40F, no snow cover, no precipitation in forecast following application
To be clear I believe there will be a decrease in the percentage of Grey Dogwood mortality using Glyphosate compared to Triclopyr 4. This methodology was chosen to reduce the impact to high quality native vegetation in grade A and B prairie. The timing and frequency of fire following brush clearing will also be important to achieve long-term control.  

This screams out for an experiment, I'm sure you are already planning to gather data and watch this closely. Let me know if you have any questions or would like to discuss further. 

Jenny Flexman of Poplar Creek Prairie Stewards wrote: I'd be interested in knowing how they applied the herbicide.  We apply with paint brushes.  Some people apply with squirt bottles or sprayers, which is not as exact and perhaps not a good idea in high quality areas, especially sprayers.

We found in Shoe Factory Nature Preserve that repeated cutting, cut-stump herbiciding (even though we often cut without herbiciding,or missed a lot while herbiciding because we had more cutters than herbiciders) and burning eventually got us to the tipping point where burning alone controls the dogwood: no more clones. We did not see herbicide damage to remnant species. Outside the preserve we did the same thing on a slope that was massively covered by dogwood, but we also seeded, and there was so much dogwood that we often cut far more than we herbicided.  I guess what that all means is that if herbiciding is causing damage to remant plants, cutting and burning alone might eventually be successful, since the cutting gets sun to the ground layer for better growth, and thus better burns, which further knock back the dogwood.  (One word of caution: our effort was accomplished by burning more frequently than once every three years.) 

Rob Sulski of Footstone also recommended repeated dormant season treatment for gray dogwood, although he suggested careful cut-and-paint for all stems. Cut the stems 2 inches high, flat so the cut area will hold maximum Garlon 4 (20%) without running off, but also paint the top half-inch of the bark. He reported treating dogwood at varied times during the winter and found March to be the most effective time. Regular burning is critical; but burn no sooner than a month after herbicide treatment to give the herbicide enough time to act. Do this in three consecutive winters to reduce dogwood sufficiently that fire will control it from then on. Interestingly, Rob is both a volunteer steward and a contractor, and he said that for the level of care needed in the highest quality areas, he thought few contractors be as painstaking as required, but that volunteers could. 

Jim Vanderpoel, Citizens for Conservation, board: Gray dogwood is easily managed by prescribed burns in the open prairie reconstructions of Grigsby and Flint Creek Savannah. Both preserves went right from row crop, hay meadow and overgrazed pasture to restoration--they skipped that old field stage that is so favorable to gray dogwood.

A much closer example to Somme Prairie Grove is Barrington's Baker's Lake preserve. It had a large population of gray dogwood that we have cut down time and again and it keeps coming back. In one area east of the trail the dogwood is kept at bay by frequent burns. That is where the richest forb, sedge and grass community is. That rich community must burn hot enough to kill the gray dogwood--the vicious circle is that other parts of the preserve are too brushy to support the hot grass fires that can kill the brush!

I guess I can just repeat the recommendation that the team persevere at Somme Prairie.

Debbie Antlitz, ecologist, Cook County Forest Preserves wrote:

One factor of a true remnant prairie community is that many of the mychorhizal networks, soil biota, structural aggregates, capillary action, cation exchange capacity etc are still intact, and more intricate and extensive compared to restoration sites. The myco networks in particular can connect and transport nutrients across species, and through this method selective application of herbicide to one species of plant could eventually impact another species, as appears to have happened in your photos. This may also apply to elemental components of the overall herbicide as they are metabolized by the fungi, such as ammonia, ethyl groups, nitrogen, etc, which depending on concentration could work as a fertilizer or a poison. 

I have from time to time seen this phenomenon on localized scales, in remnant prairies of varying quality. Usually within a year or two the patch will have recovered to the point where the dead zone is no longer evident. 

My recommendation for remnant areas with this phenomenon would be to consider treatment with herbicides with gentler base components, shorter half-life, and perhaps less systemic (depending on effectiveness if the invasive species being treated is not well established with deep roots).  

Kevin Scheiwiller, Citizens for Conservation, staff: We do not have much dogwood in our prairie reconstructions, so the following may not be comparing apples to apples.

At Baker's Lake savanna, in March 2019, we took on a couple of large dogwood clones in an attempt to free up room for some teenage bur oaks. We decided to cut and treat stumps with 20% Garlon 4 immediately after the workday, very meticulously. I would say only about 65-70% of the clone was killed. The first picture shows shows what about 30% of the clone looks like, a dead stump with new sprouts growing immediately adjacent. Similar to your experience, you can still see the oil on the stump from a year and a half ago, and there are no new lateral shoots, which leads me to believe this stump was not missed with herbicide, and yet there is new dogwood growth.

Dave Bart's comments make a lot of sense to me, as it probably takes multiple years of attacking these clones to finally defeat them. Fire obviously has helped keep some areas relatively clean, but the second photo shows the results after we got a nice hot fire in Baker's Lake this March. The dogwood was top killed but has already grown back to 70% of its original size in just 5 months!

Your fears that aggressive natives will take over these areas are well founded as tall goldenrod and woodland sunflower are quickly moving into the areas we cleared. Might be a good place to sow some dodder and seed with our savanna mix.

The collateral damage to the prairie plants is a bit of head scratcher. Garlon 4 in the dormant season never seems to have much lasting effect. I use 50% Glyphosate in any wet areas where we cut brush in the winter and have noticed some bizarre residual effects on Glossy Buckthorn, but not on any herbaceous veg. Might be worth trying Glyphosate out in an area to see if that keeps the off-target kill down.
 
CFC would be happy to run parallel or complementary experiments on tackling dogwood (i.e. basal barking vs. cut stump vs. glyphosate treatments). Let me know if I can provide anything else for this project!

 

(For additional good contributions, see "Comments" - below.)

 

---------------

 

Stewards have prepared this report to reach out to others for additional possible insights and suggestions. 

When all the evidence and analysis are in, we'll do a new post on lessons and next steps. 

 

Somme Prairie Site Stewards: Laurel Ross and Lisa Musgrave

 

Somme Prairie Seminar Group stewards: Matt Evans, Katie Kucera, Christos Economou, Emma Leavens, Eriko Kojima, Sai Ramakrishna, and Stephen Packard. 

 

A note on the Seminars: Starting in 2017, the "learning community" at Somme Woods began a process where stewards would draft plans for a part of the preserve, share them in writing, graphically, and orally during a field seminar. All stewards would contribute ideas, and the plans and work would benefit. Since that time, a variety of different seminar steward sub-groups have expanded this work-and-learning approach to other sites, as in this case at Somme Prairie. This draft post is part of that process.