Managing Sedge for Better Northern Hardwood Regeneration (Stockbridge-Munsee Reservation, WI)

State or Province
Nearest city or town
Stockbridge Munsee Reservation
Describe the location
Stockbridge Munsee Reservation
Stockbridge Munsee Reservation
Cover type
Forest health threats
Silviculture system
Estimated year of stand origin
Site index
75 feet
for species
Acer saccharum
Brief silvicultural objective
Better understand the relationship between Carex pennsylvanica and Carex pedunculata, deer browse, and regeneration of quality northern hardwood species in a way to create future management options to promote seedling success.
Site preparation method
Soil texture
Soil details
Fine Sandy Loams to Loamy Fine Sands over glacial outwash.
Stand area
18 acres
Treatment area
5 acres

44.869536, -88.904822


This study is centered in a stand of rich mesic northern hardwoods indicative of the local landscape. The basal area is about 63ft2/acre (14.5m2/ha) with an average DBH of 12 inches (30.5cm). The stand is 90-110 years old and has the same cutover history of the rest of the local lands, and most of Wisconsin for that matter. It is speculated that the area may have been pastured for a period of time due to the evidence of some branching that appears more common on openly grown trees, but this was long ago. The soils are mostly excessively well drained Rosholt Fine Sandy Loams which support the rich mesic (ATM grading towards AH) plant community. To the south the soils transitions to Shawano Loamy Fine Sand but the difference is very minimal.

Management was absent until the Tribe entered the stand in 1999 and conducted a commercial thinning with the objective to begin converting the even-aged stand to an all-aged stand. Years after the harvest, little to no regeneration had developed however a sedge component was evident covering 75% to 100% of the forest floor. The problematic sedge component consisted mostly of Penn sedge (Carex pennsylvanica) with a component of longstalk sedge (Carex pedunculata). A few scattered sugar maple (Acer saccharum), white ash (Fraxinus americana), and black cherry (Prunus serotina) seedlings were present but were so deformed from years of deer browse it was a surprise they were even alive.

Though not excessively widespread, more and more areas are being noticed with this sedge carpet and little to no tree regeneration. Often these stands are associated with areas of high deer numbers, possible disturbance (i.e. pasturing), and along the edges of the forest. However, as the Tribe’s land base expands and the sedge inevitably continues its encroachment into the forest, the problem is sure to grow.

A few researchers have begun looking at the interactions of sedges and regeneration, often including the influence of deer. Randall and Walters (2005 Michigan SAF) studied the survival rates of planted sugar maple seedlings with and without sedge and with and without deer. Though the sedge limited seedling success, the effects were overshadowed by those of the deer. Similarly, Matula with the WDNR (Un-published as of 2014) showed the benefits of sedge control but still found that deer browse overshadowed the sedges. Interestingly, the discussions highlight the interactions between the sedges, the selectivity of deer browsing, and even earthworms all exacerbating the issue. Further, if climate change predictions are included, warmer drier summers can be expected in the future which seems to favor sedges. Warmer and drier summers, warmer and drier soils due to worms, mineral seedbed due to worms, potential for frost heaving, the impenetrable mat created by the sedges, and the un-palatable nature of sedge to deer all promote sedge over nearly all other species. How are managers supposed to combat such an issue with so much weighted against success? Though both great studies, the Tribal lands are unique with their extremely high nutrient levels, lower deer numbers, limited access, and greater freedom of management. The Tribe felt that the results of these studies could be used to develop a complementary study for the reservation and look at the effects of treatments within this different environment.

Silviculture Objective(s)

The study hopes to evaluate the effectiveness of scarification and an herbicide treatment to control the sedge component and promote successful hardwood regeneration (4,500 stems/acre or 11,000 /ha) after 5 years. Deer exclosures will also help to show best case scenarios and help evaluate the success of the treatments.

Figure 1. The study area stand is infested with a dense carpet of sedge that covers nearly 100% of the forest floor. Little natural regeneration is occurring on the site even though the stand was recently harvested leaving about 80 ft2/ac of residual basal area.

Pre-treatment stand description and condition

Landowner objectives/situation:

The landowner’s overall goal is to promote a healthy forest and one component of this is healthy and successful regeneration. The limitations to adequate regeneration are:

  1. Management Techniques – There is lots of research on hardwood densities, gaps, species composition, seed bed preparations, etc. Though important, it is not a focus of this study as the same management strategies failing in stands with sedges are successful in areas absent of the sedges; notably these successful areas also have low deer numbers.
  2. Climate Change – Not a whole lot to work with here, other than beginning assisted migration and promoting more southern species. Surely a contributing factor but not a fiscally feasible treatment option.
  3. Earthworms – They are here. They remove the duff layer, mix the soil, allow for excessive drainage, reduce soil moisture, alter nutrient fluxes, warm the soil in spring (no duff layer), contribute to potential frost heaving, etc… Again, surely a contributing factor but no fiscally feasible treatment option available at this time.
  4. Deer – They are also here. Over the years, study after study shows how excess deer densities hinder the ability of successful forest management as well as exacerbate the majority of other forest health issues. However, with the political and social stigma surrounding deer their control as a treatment is essentially a moot point. Though moot, creating small deer exclosures are cheap and easy and may shed some light on how other treatments are working.

Sedges – The presence of dense mats of sedges seems to cause a problem though the interactions of other factors may be compounding these effects. Treating sedge is relatively cheap and easy with either herbicides or ground disturbing activities when compared to the other factors. However, will treating just the sedges be enough to overcome the effects of the other factors and promote successful regeneration?

Figure 2. The test plot in the foreground was treated with herbicide which seems to have controlled the sedge mat over the past growing season. 

Silviculture Prescription

The prescription was to create 32 plots (3m x 3m or 9.84ft x 9.84ft) to explore the effects of herbicide, scarification, deer exclosures and their interactions on hardwood regeneration success. The plots were to be chosen in an area to keep as many natural variables as constant as possible, for instance: soil, light, moisture, tree density, sedge density, etc. Seedling and herbaceous plant cover surveys were scheduled for years 1, 2, 3, 5, and 10 years using a plot centered 2m x 2m (6.56ft x 6.56ft) grid divided into 16 quarter meter squares (2.69ft2). The outside area was left un-surveyed as a buffer to any effects the edge of the plots (specifically exclosures) may have had on the future regeneration.  Surveying the amount of sunlight reaching each plot was also planned to determine how homogeneous each plot was.

Figure 3. Deer exclosures were erected to investigate the cumulative effects of sedge and deer. This test plot shows only the deer exclosure treatment where previously browsed seedlings are responding well to the absence of browse pressure.

What actually happened during the treatment

During 2013, plots were chosen and randomly treated with herbicide, scarification, deer exclosures and all combinations of such. Four (4) replication of each were created to make the study more robust. The scarification was completed May 29th using a rototiller set to 3 inches (7.6cm), making 4 passes in each needed plot. Originally, Oust XP at a rate of 2oz/acre (140g/ha) was recommended and applied on June 3rd. This failed to be effective and in the meantime it was discovered that Randall and Walters had good success with Glyphosate and fall treatments. November 15th, all herbicide treatments were treated again with 2.7% active ingredient. The fall treatment was intended to be following leaf fall but before freeze up as the forbs and trees were dormant however the sedges were still actively growing and easily killed. The exclosures were erected on June 25th using 8 foot (2.4m) fence posts and a 6 foot (1.8m) wire fence. It was felt, due to the small size of the exclosures, a 6 foot (1.8m) high fence was adequate.

On August 21st a windstorm damaged some of the trees in the study area. Only one exclosure was damaged but repaired. Large branches that fell into plots were gently removed.

During June of 2014, sampling began measuring percent coverage of plants in the plots, recorded by species. Seedling counts were also conducted. Unfortunately, time constraints prevented crews from complete the sampling for this year. It was disappointing to miss this sampling period but also the crew felt that this data was not completely necessary to answer the ‘big’ question. 

Figure 4. The test plots are also showing signs of other herbaceous plant species colonizing in the absence of the dense sedge mat. In other plots, especially ones that were scarified, non‐native invasive species are also being introduced which may be of concern.

Post-treatment assessment

Though very early in the study and with only observations being made; the herbicide seemed to do a good job controlling the sedge component the first year and allowed for some germination of hardwood species. Sugar maple seedlings were very prevalent and often numbered 100+ seedlings per plot. Other tree species were observed: basswood, red maple, black cherry and white pine. Spring ephemeral forbs were also observed frequently in these plots. The scarification plots also showed good control of the sedge component but looked as if it was rebounding quicker than in the herbicide treated plots. Notably, the scarified plots were more populated with ‘weed’ forbs and exotic invasives plants were noticed. The exclosures showed similar results as the other plots, but this is expected as browse is a more long term effect.

As planned from the beginning, long term monitoring is needed to really tease out the effects of the treatments and answer the overlying question of regeneration.

Plans for future treatments

No plans for future treatments are expected other than the obvious continued monitoring. As more distinguishable results are discovered future treatments may be formulated to develop more detail but that is likely years down the road.

Costs and economic considerations

The project was funded by the BIA and all needed supplies cost about $2,000 ($1,600 for exclosures, and the remainder for a lux meter, soil probe, flagging, and sampling odds and ends). The project took about 64 man hours to set up and complete all treatments. Also, the crew was finding that sampling will take about 2 man hours per plot per sampling period or about 64 man hours; this includes data entry but not analysis.

Extrapolating the costs out for actual stand treatments is a bit harder at this point and probably easier to find somewhere else from other studies. However herbicide was applied at about 4.4 gallons (16.7L) of product (GlyStar Plus 41% ai) per acre at a cost of about $46/2.5gal jug ($4.86/L). Obviously a tank sprayer would be needed to cover stand sized acreages and this would influence the cost greatly (tractor? ATV?)… Similarly, the rototilling may or may not be feasible at a stand scale as rocks, roots, topography, and time may limit its usefulness. The crew chose to use the rototiller in this study because the Tribe owned the tiller, the plots were small where anchor chain or larger more common scarification equipment would not have been appropriate, and the results were felt to be the best case scenario (if we are going to see results at all this would provide the best possible situation).  

Other notes

Randall, J.A., and Walters M.B. (2005) Deer and sedge: bottlenecks to seedling regeneration in northern hardwood forests: Potential restoration techniques aimed at reversing the effects. Michigan Society of American Foresters. Forests and Whitetails – striving for balance. St. Ignace, MI, USA. June 9th – 10th.

Randall and Walters have continued their research and may have more to offer.

Colleen Matula of the WiDNR has been conducting research in this area but has not published anything yet. She was gracious enough to share some of the preliminary data but was not able to provide any document for distribution at this time.

Summary / lessons learned / additional thoughts

This project is still in its infancy but already the crew is seeing some interesting interactions between removing the sedges, seedling germination, and forb recovery. The fall herbicide treatment is showing some potential for control while still promoting the seedlings and native forbs, which is pretty exciting. It will take more time to develop any definitive answers from these treatments.

There are already things the crew would like to do differently if we could, first would be better setup. The crew would have liked to have tested plots for consistent light conditions before setup so that this variable could have been more consistent. Second, the crew would have liked pre-treatment data; did plots differ greatly before the treatments already and was this a product of chance or some microsite conditions? Could it have been corrected for? Finally, and the biggest short coming, consistent and regular sampling is needed; the crew just hopes to find the time.