Afforestation of an Agricultural Field to Mixed Hardwoods by Direct Seeding (MN DNR)

Overview

This case study is a successful example of afforestation of a mixed hardwood forest stand on a former agricultural field through a technique known as “direct seeding”. https://www.dnr.state.mn.us/treecare/maintenance/collectingseed.html

Direct seeding was used on this site because previous efforts to establish forest by planting seedlings had been unsuccessful due to several factors, including vegetative competition and predation by deer, rabbits and mice.

The direct seeding technique for afforestation of former agricultural fields was pioneered by foresters in the “driftless” region of Iowa, Minnesota and Wisconsin in the late 1990s. Direct seeding involves mechanical and chemical site preparation prior to sowing a variety of seeds to convert an agricultural field to forest.

One strength of direct seeding vs. planting is that it enables superior adaptation to subtle variations in site conditions. When planting a few hundred seedlings/acre at regularly spaced intervals, small variations in site conditions generally are not accounted for. In contrast, direct seeding thousands of seeds of several species over the entire site enables microsite adaptation by species and individuals. 

Notable features of this case study include:

• It provides an opportunity to explore impacts on tree establishment and growth of factors like microsite soil or topography differences, seeding mixes and rates, and cultural techniques.

Specifically, there are patches that have exhibited exceptional oak and black walnut establishment and growth. Did this happen as a result of soil differences, differences in herbicide effectiveness during early grass and weed competition control efforts, sowing rates, genetics, or was it most likely due to a combination of factors?  We share what we were able to learn (and what we were not) about impacts of some of these factors.

• It provides an opportunity to continue learning by documenting current conditions as baseline data to help assess impacts of any future follow-up treatments.

Figure 6: Gathering data in one of the patches with excellent black walnut survival and growth in 2020.

Silviculture Objective(s)

Produce forest products while improving water quality, wildlife habitat, and outdoor recreation opportunities through afforestation of an agricultural field to hardwood forest.

Pre-treatment stand description and condition

Stand establishment and management history: 

The site was an agricultural field growing annual crops before project initiation.

Pre-treatment species composition: 

NA

Pre-treatment growth and stocking: 

NA

Pre-treatment forest health issues: 

NA

Landowner objectives/situation: 

While specific objectives vary from parcel to parcel, lands under the administration of DNR-Forestry are managed in alignment with Section Forest Resource Management Plans (SFRMP) to ensure that state forest management activities meet statewide goals for ecological protection, timber production, wildlife habitat and cultural/recreational values. The DNR assembles teams from the Divisions of Forestry, Fish & Wildlife, and Ecological & Water Resources who work with partners and the public to develop SFRMPs. 

Specific objectives for this project included improving water quality, wildlife habitat, outdoor recreation opportunities and production of forest products.

Silviculture Prescription

Two adjacent agricultural fields were seeded – one in 2005 and one in 2010. The following series of treatments were implemented: 

Treatment	Date	Description	Acres Treated
Kill adjacent boxelder trees	Summer 2005	Kill adjacent boxelder trees to reduce their seeding into fields.	Unknown
Disc field	Fall 2005 Fall 2010	Prepare seedbed by disturbing soil with an agricultural disc.	28 23
Sow seed	Fall 2005 Fall 2010	Sow the following seeds: 2005: A total of 280 bushels of black walnuts and 140 bushels of red oak acorns, and some green ash*.   2010: No records of species and amounts seeded could be found. However, it is clear from seedlings present that red, white and bur oak acorns and black walnuts were seeded. Standard direct seeding rates are: Red oak: 1 to 2 bushels/acre White oak: ¼ to 1 bushel/acre Black walnut: 10 to 20 bushels/acre NOTE: Bur oak was only seeded in 1 small patch, at a much higher-than-standard rate, as a field trial.   The heavy seeds (walnuts and acorns) were dispersed with a fertilizer spreader and disked into the soil from 1” to 3“ deep.   Any lighter seed (ash, maple and all others) was dispersed and dragged lightly to cover it with soil.   *Note: There is no record of it, but the number of green ash seedlings in the portion of the field seeded in 2005 indicate that green ash must also have been seeded then. The 2005 sowing took place before Emerald Ash Borer was established in this area.	28 23
Herbicide release (3 times)	2006 2009 2011	Apply ¾ ounce Oust and 3 ounces Princep herbicide/acre to control grass and weed competition.	28 28 23

What actually happened during the treatment

Heavy natural boxelder seeding from adjacent trees made site operability challenging for follow-up release and tending work.

Post-treatment assessment

In 2020, the site is on its way to becoming a mixed hardwood stand with strong oak and walnut components.  

As shown in Table 1 at the link below, the site averages about 900 stems/acre oak (red and white combined) and 400 stems/acre of black walnut greater than 1 foot tall.

Table 1: Stems/acre of regeneration by species, size class, and plot frequency in 2020.

In addition to the target crop species of oak and black walnut, there are also significant numbers of ash and boxelder, and smaller amounts of a variety of other species.  The naturally seeded hackberry and sugar maple is beneficial in that it will provide some diversity to the future stand. Although we did not pick any up in our plots, there are even a handful of naturally seeded white pine seedlings present.

Notes about ash, elm, and boxelder:

The objectives for including ash in the seeding mixture in early direct seedings were threefold:

1. Help establish “crown closure" and begin shading out grass and weed competition as early as possible.
2. Improve timber quality of high value hardwoods such as oak and walnut by providing side competition to nearby stems and forcing them to grow straighter, and with fewer low limbs.  
3. Due to general ease of establishment, serve as the tree crop in any portions of the stand where higher value species failed to become established. Note: There was no ash sown on the portion seeded in 2010. The portion sown in 2005 that included ash was before EAB was known in this area.

There is also a significant amount of boxelder and elm that seeded in naturally from adjacent trees. Similar to ash, some elm and boxelder during early years of establishment is beneficial to helping achieve earlier crown closure to shade out grass. However, at high levels and left uncontrolled they can have a negative impact by outgrowing and shading out more desirable tree species. Their level of competition to more desirable trees is monitored and controlled, as needed.

While establishment of trees has been good on most of the site, there is tremendous spatial variability in tree survival and growth in 2020 

On some portions of the site, trees have already achieved crown closure and are healthy and growing rapidly. In fact, in 2020 some places would benefit from precommercial thinning work to reduce competition and speed growth of crop trees.

However, on other portions seedlings are growing slowly due to heavy grass competition. Growth and vigor should accelerate considerably after achieving crown closure over the next 5 to 10 years. 

Potential causes of tree survival and growth variability in 2020 

The tree survival and growth variability found on this site is common on direct seedings of former agricultural fields. Many factors may have contributed to the variability, including: 

• Soils
• Topography
• Herbicide effectiveness
• Seed sowing rates
• Seed quality and genetic variability 

We share our thoughts on these factors and the impacts they may have had on tree survival and growth below: 

Soils 

Subtle differences in soil factors such as compaction, structure, nutrient levels and moisture retention can impact tree survival and growth. Spatial variability in soils is commonly cited in agricultural sciences as a cause of different rates of productivity across a site. It is highly likely that it explains at least part of the variability seen in this case study.

Unfortunately, the NCRS soils map shows the same soil series over the entire project area, so it is not refined enough to ascertain subtle soil differences that might have impacts. It would therefore be valuable to do additional soil analysis work in the future on this and/or other former agricultural field direct seeding sites. Soils in areas with good tree growth and survival could be probed and analyzed, and then compared to soils in areas with poor results. 

Topography

The NRCS soils map is not sensitive enough to show subtle differences in elevation, but it is a factor that can have noticeable impacts on soil temperatures and moisture retention and availability, which can impact tree growth. Topography has an impact on tree growth. Specifically, northerly and easterly aspects on even subtle slopes tend to be cooler and moister than places with westerly or southerly slopes, and therefore better mesic tree growth.

We produced a map with 1-foot contour lines, using our highest resolution Lidar elevation data under the hypothesis that very subtle depressions or hills could explain superior tree survival and growth. We did not find such a correlation. Figure 2 at the link below shows 1-meter Digital Elevation Model (DEM) Lidar imagery superimposed over a 2019 color infrared aerial photo of the case study fields and surrounding area. Orange shows highest elevation, then lower in order: yellow, green, light blue and dark blue for the lowest areas. If you look closely, you can see areas in the study fields where the trees are significantly taller than others. At the sensitivity level of the available Lidar, we see no pattern of subtle depressions or hills that appears to relate to these places.

NOTE: We do think it would be worthwhile in the future when even more sensitive Lidar is available to take another look at this.

Figure 2: 1 meter digital elevation model imagery superimposed over 2019 color IR airphoto.

Herbicides

Herbicide effectiveness in controlling vegetative competition (especially grass) in early years of direct seedings can impact tree survival and growth. Application rates would have been within accepted tolerances, so any variability on this site would have been minor. Although probably not a major factor, slight differences in application rates and in herbicide effectiveness due to soil variations could explain some of the variability in tree growth and survival. However, in 2020 there is no way to determine if that was indeed the case on this site. Overall, foresters have observed over the years that effective control of grass competition during early years of direct seedings is critical to successful forest establishment.  

Seed Sowing Rates 

We don’t have detailed seeding records for this site that tell us the exact rates used. We can assume that the rate of seed applied over most of the site was close to recommended levels at that time, as documented in the “Silvicultural Prescription” section. 

Perhaps the most helpful observation we can share on seeding rates is that to the extent practical, more seed is probably better than less. It should be noted that challenges in procuring adequate amounts of seed for direct seedings make it impractical to suggest a large increase in recommended seeding rates for most sites. However, a field trial on this site of much heavier than average acorn seeding was successful in establishing greater oak numbers with better than average growth (see Figures 7 and 8). 

Figure 7: Bur and red oak show excellent growth and survival in 2020 in this patch where seed was sown at exceptionally high rates as a field trial.

Figure 8: Green circle shows location of extremely high acorn seeding rate field trial.

We aren’t sure exactly what rate/acre of acorns was seeded on this small patch, only that one of the foresters overseeing the project recalls that an exceptionally high rate was sown at seeding time as a field trial. The condition of this patch is dense enough in 2020 that it would benefit from a pre-commercial thinning. Acorn availability and cost will limit how much practical application of very high sowing rates could be achieved on future seedings, but generally, more is better than less. 

Seed Quality and Genetic Variability 

Genetic variability and quality of seed is always a factor that impacts tree survival and growth. Some genetic seed variability is beneficial since it is an important factor in achieving the goal of establishing a genetically diverse forest. One great strength of direct seeding vs. planting is that it enables superior adaptation to subtle variations in site conditions. When planting a few hundred seedlings/acre at regularly spaced intervals, small variations in site conditions generally are not accounted for. In contrast, direct seeding thousands of seeds of several species over the entire site enables those species and specimens best suited to take over and thrive in a particular area, based on microsite conditions. Some genetic diversity is a given on direct seedings, when one considers the many trees needed to source the volume of seed required – it has to come from many trees. 

We don’t have records on seed sources for this site that tell us details of exactly where the seed came from. However, we do know that procurement and handling procedures ensured that the seed used came from the either the same seed zone (or at most 50 miles to the south), and that it was viable. 

One interesting observation on this site were some individual 15 plus-foot-tall red oak stems surrounded by much shorter oak and black walnut struggling in heavy grass and goldenrod competition. Figure 9 shows an example of this. Genetic variability may explain the random pattern of thriving oak stems. 

Figure 9: Note the single 15-foot-tall red oak stem surrounded by other stems of boxelder, and much shorter oak and black walnut struggling in heavy grass and goldenrod competition in 2020.

Figure 10: Foreground shows a portion of the site with trees struggling in heavy grass competition. 

Figure 11: Another shot of a portion of the site where black walnut and oak are struggling to compete with grass and goldenrod in 2020.

Although we do not have information on acorn sizes seeded, the phenotypic factor of acorn size for oaks has been shown in studies to impact seeding survival and growth. Specifically, larger acorns tend to produce larger and more vigorous seedlings that compete better than do smaller acorns. 

It is not always operationally feasible to identify and use genetically superior seed, but land managers can encourage seed collection from parent trees that exhibit superior growth and form characteristics and use larger than average acorns.

 Herbaceous plants 

For ground cover in 2020, we observed mostly high cover of sun-loving native weeds, including goldenrod (Solidago sp.) along with some grasses, asters, raspberries, clematis, and wild parsnip.

Plans for future treatments

Monitor the stand through periodic inventory surveys.  Design and execute further planting and regeneration efforts if needed. 

As trees mature and competition reduces a significant number of crowns to less than 1/3 of the bole, a pre-commercial thinning to release the most vigorous and healthy desirable trees will be done.

Costs and economic considerations

Costs

Activity/Expense	When	Total Cost	Cost/acre
Kill adjacent boxelder	2005	1,682	60
Disc field	Fall 2005	2,523	90
Purchase and sow seed	Fall 2005	2,803	100
Herbicide release	2006	2,383	85
Herbicide release	2009	4,205	150
Purchase and sow seed	Fall 2010	?	?
Herbicide release	2011	1,890	68

Other notes

The review and editing assistance provided by Minnesota DNR Ecological Classification System Program Consultant Ashlee Lehner is gratefully acknowledged.

This case study was developed with support from the United States Department of Agriculture's National Institute for Food and Agriculture, Renewable Resources Extension Act. Project #2021-46401-35956, principal investigator Eli Sagor, University of Minnesota.

Summary / lessons learned / additional thoughts

In 2020, the site is on its way to becoming a mixed hardwood stand with strong oak and walnut components 

Use of direct seeding for forest establishment was successful on this site. The site averages about 900 stems/acre oak (red and white combined) and 400 stems/acre of black walnut greater than 1 foot tall. An advantage of direct seeding vs. planting for these former agricultural fields is that it has enabled superior adaptation to subtle variations in site conditions. 

While establishment of trees has been good on most of the site, there is considerable spatial variability in tree survival and growth in 2020 

• On some portions of the site, trees have already achieved crown closure and are healthy and growing rapidly.
• However, on other portions seedlings are growing slowly through heavy grass competition. Growth and vigor should accelerate considerably after achieving crown closure, which should start to happen on most of these areas site over the next 15 years or so. 

The variability found on this site is common on direct seedings of former agricultural fields. Several factors probably contributed to the variability, including soils, topography, herbicide effectiveness and seed sowing rates, quality and genetic variability. While we can’t know the exact impacts of each on this site, we include some related findings and recommendations for future direct seedings below. 

More detailed soils and topography information would be valuable 

Soils: It is likely that spatial variability in soils and topography are partially responsible for different tree survival and growth rates across this site. It would be valuable to do additional soil and Lidar analysis work in the future on this and/or other former agricultural field direct seeding sites. Soils in areas with good tree growth and survival could be probed and analyzed and compared to soils in areas with poor results. 

Topography: The LIDAR map we produced showing 1 foot elevation changes did not show discernable correlation to survival and growth differences visible on aerial photos. We think subtle elevation and aspect differences worthy of further investigation on other direct seeding sites.

Sowing more acorns is probably better than less 

The field trial portion of this site showed that increased dispersal rates of acorns produced greater numbers of established trees with superior growth. Acorn availability and cost will limit practical application of very high sowing rates, but we conclude that generally, more seed is better than less. 

Sowing larger acorns may result in larger, more vigorous seedlings 

We did not learn anything about acorn size in this case study, since our seed collection records are limited, and acorn size is not typically tracked. However, research elsewhere has shown that larger acorns tend to produce larger, more vigorous seedlings that compete better than do smaller acorns. We suggest that it is worthwhile to encourage seed collectors and procurers to collect bigger acorns, to the extent practical.