Hardwood Reforestation in a Creek Valley Dominated by Reed Canary Grass

by Tim and Susan Gossman

"Our farm is located in the karst region of Southeast Minnesota.  This area includes karst features such as caves, sinkholes and springs.  Lost Creek flows through our farm in a narrow valley surrounded by bluffs, steep forested land and former pasture.  Thirty years ago the 20 acres of creek bottom land on our farm was dominated by a floodplain forest comprised mostly of American Elm.  As these trees were killed by Dutch elm disease, and the shade disappeared, Reed Canary Grass (RCG) (Phalaris arundinacea) began to move into the area.  RCG is an aggressive perennial grass that threatens wetland and riparian areas where it forms a monoculture, eventually smothering the native grasses and forbs and preventing any regeneration of trees or shrubs.  It now dominates most of the 20 acres except for pockets of natural stands of native hardwoods and tees that were planted before it moved in to the area.  RCG provides almost no wildlife benefits, makes poor pasture or forage and provides no economic gain.  Returning this area to forest will provide food and habitat for birds and wildlife and provide enhanced hunting opportunities.  The trees will also offer short term economic returns from nut and acorn production as the trees begin to bear seed in about 20 years and long term economic benefits from the sale of timber.  The trees will shade Lost Creek, a designated trout stream, providing better trout habitat.  We have planted Spruce, Pine and Fir trees on our farm and have been selling Christmas trees for over 10 years.  We have also planted hardwood trees in appropriate areas of our farm and restored native grasses and wildflowers in other areas.  These have increased the wildlife benefits as well as current and future income on our farm. This project is a continuation of that process. 

Due to the persistence of RCG and its resistance to control by non-chemical practices, we were faced with an environmental decision:  whether it was better to leave the creek valley and RCG untreated and allow the RCG to dominate and spread but not expose the area to herbicides or to explore several alternatives including treating an area with chemical herbicides for several years in an attempt to reforest the area.  After much research and deliberation, we decided that including a plan to determine the effectiveness of the application of herbicides with a track record of low environmental impact at rates no higher than would be used in a field of soybeans for a period of only 3 or 4 years to reestablish a forest that should remain for over 100 years is the more sustainable, environmental benefits to our farm and to the Lost Creek and Root River Watersheds.  When using herbicides, we followed label instructions for rates and application techniques, including any special instructions for spraying near water.  We plan to reach the goal of reforestation by the use of alternative plans using different techniques of suppressing the RCG and growing trees.

It has been taken over by RCG.  We realize that this a long term project and plan to complete the project over 7 to 10 years.  This long term plan exposes no more than 2 acres of tilled soil to erosion in any one year.  Over the past 20 years we have planted tree seedlings and tree seeds such as walnuts and acorns in the creek valley.  We have had fair survivability in the ears not yet overtaken by RCG, but near 100% failure when planting in areas dominated by RCG.

The four strategies that we are using to control the RCG and eventually return the area to a mix of bottomland forest with a healthy understory and open areas of sedges, reeds and native forbs are listed below.  All four methods utilize the fact that RCG does not reproduce or survive in heavy shade.  We will repeat the four plans over the 3 years of the Sustainable Agriculture Demonstration Grant to test the procedures in different weather conditions.

Plan A:  Controlling the RCG with a combination of prescribed burning, herbicide application, mowing and tillage, followed by a direct seeding using nuts, acorns and seeds of a diverse mix of bottomland hardwoods and shrubs, supplemented by planting tree seedlings are the techniques we are using in this research area.   Plan Details     Photo Gallery

Plan B:  Planting fence post sized poles of willow and cottonwood in areas that are not accessible by machinery to shade out the RCG if the method used in this plan.  When the RCG is eventually controlled, the site will be planted to additional tree species for more diversity.   Plan Details     Photo Gallery

Plan C:  Planting a direct tree seeding of a diverse mix of bottomland hardwoods and shrubs in areas where the shade of Box Elders has already controlled the RCG, followed by cutting down the Box Elders is the procedure we are using in this plan.   Plan Details     Photo Gallery    Collecting your own seeds photos     Seed Storage photos      Seed Photos

Plant D:  One year of herbicide treatment and tillage adjacent to stands of Box Elder trees to allow natural seeding by the Box Elders and encourage new stands of Box Elder to shade out the RCG is the experiment we are doing here.  When the RCG is controlled, the area will be treated as in Plan C, with a direct tree seeding and removal of the Box Elder trees.   Plan Details      Photo Gallery

The main result we are looking for – a good stand of trees growing in each treatment area – will begin to be assessed in the spring of 2008 as the trees start to germinate (in Plans A, C and D) or leaf out (in Plan B) and will continue to be assessed over several years as they grow."