Investigating the Potential of Woodland Silvopasture Systems: Prevalence, Practices, Perceptions and Performance

Progress report for LNC20-433

Project Type: Research and Education
Funds awarded in 2020: $249,731.00
Projected End Date: 10/31/2024
Grant Recipient: University of Missouri Center for Agroforestry
Region: North Central
State: Missouri
Project Coordinator:
Dr. Ashley Conway
University of Missouri Center for Agroforestry
Expand All

Project Information


Silvopasture (integration of forage, livestock, and trees) benefits include increased economic diversity, and in many cases, improved animal performance with increased ecological services. Although there is interest in adopting silvopasture among livestock producers and resource professionals, the diversity of silvopasture systems and outcomes has generated a body of literature that is unable to provide robust and regionally-appropriate recommendations for silvopasture production. The majority of available literature focuses on designed silvopasture, where trees are planted into an existing pasture. Substantial opportunity exists to convert un-managed woodland property with selective thinning, invasive species management and understory development to a silvopasture, thereby potentially improving the ecological health and economic functionality of the area. Currently, there is little information available on silvopasture practices NCR states, particularly with respect to woodlands, and available silvopasture studies may not be meeting the needs of regional producers.

This project will begin methodically addressing the knowledge gaps that may be limiting

silvopasture understanding and adoption in NCR states by: 1) surveying livestock producers in Missouri to assess current woodland management and silvopasture practices and perceptions; 2) developing a silvopasture peer-learning network; 3) evaluating the ecosystem impacts of converting un-managed woodland to a silvopasture system; and 4) assessing the economics of  conversion and estimated forage production to develop a theoretical financial decision-support tool for producers interested in adopting woodland silvopasture. Missouri ranks highest among NCR states for cow herd inventory and 34% of the state is forested, making it an ideal choice for surveying cow/calf producers’ and maximizing outreach impact for this type of silvopasture.

Through direct producer engagement, short-term project outcomes include increased knowledge of silvopasture adoption, management practices, and perceptions among livestock producers, as well as measured economic and ecological metrics of a native woodland silvopasture system. Long-term outcomes include wide-spread rehabilitation of un-managed woodlands to a healthy silvopasture and increased profitability of small-scale livestock producers.

Project Objectives:

Specific objectives: 1) Assess current silvopasture practices and perceptions among grazing livestock producers in Missouri; 1a) Cultivate a peer-learning network of silvopasture practitioners in Missouri and connect to external networks; 2 and 2a) Assess the ecosystem impacts and economics of establishing a woodland silvopasture system to develop recommendations and a financial decision support tool for producers.


Learning outcomes: 1) Increased understanding of the current status of silvopasture practice in the state of Missouri including producer perceptions of benefits and barriers to silvopasture adoption; 2) Quantification of current economic and potential production metrics for a woodland silvopasture system in Missouri.


Over 78% of NCR cow/calf producers are small-scale operators (gross annual farm sales <$250,000), and in
2019, NCR states owned over one third of the US cow herd inventory, according to USDA. These producers
require innovative and cost-effective management practices to improve economic resilience. An exploratory
study of producers and resource professionals in southwest WI indicates interest in silvopasture, but a lack of
information on establishment, economics, and best management practices in the Midwest. These, and similar,
operators are ideally positioned to become more diverse, environmentally sustainable and innovative with
silvopasture management to enhance financial stability, but more information is needed.

Silvopasture production systems offer diversified revenue streams for producers. Previous work has suggested
benefits include improved animal welfare and performance, and ecological services such as increased pasture
diversity, soil and forest health and enhanced wildlife habitat. However, the diversity of silvopastoral systems
has resulted in a wide range of data, which is not always applicable for NCR producers who are interested in
silvopasture adoption. Moreover, information on current silvopasture adoption and practices in NCR states is
limited and/or out of date. In order to develop targeted research questions based on real NCR producer
challenges, there is a need for foundational data collection on the current state of silvopasture production
practices, characterizing the production and economic dynamics of a regionally-specific silvopasture system, and
measuring the system in a controlled setting.

This project will assist small-scale farmers/ranchers by establishing a state-wide peer network of practicing
silvopasture producers. Additionally, applied research questions will be generated from producer input,
generating detailed production and financial return information comparing different management practices,
resulting in a financial decision support tool. Producers who adopt silvopasture can obtain higher profitability
through reduced expenses and diversified revenue streams, and this project will enhance their ability to
optimize system management.



Based on the available data, we hypothesize that woodland grazing practices are already prevalent in the state of Missouri and that interest for woodland silvopasture is high, although practitioners lack technical information on best management practices as they pertain to silvopasture. Furthermore, we hypothesize that selective "conversion" treatments of unmanaged woodlands will shift vegetative functional groups to more biodiverse and functional species, and in the long-term, both thinning and thinning with prescribed burn will produce differentiated ecological states with varying suitability for silvopasture management. 

Materials and methods:

The primary approach for this study will be to build a base of knowledge regarding silvopasture practices and feasibility of woodland silvopasture conversion in the temperate NCR region, beginning with Missouri. The foundational nature of this work will result in critical information necessary to begin designing appropriate long-term silvopasture research sites that will answer questions useful for producers seeking to integrate silvopasture production into their current system.

Specific Aim 1: Evaluate current livestock grazing practices in wooded areas, perceptions of silvopasture, and potential for adoption among livestock owners.

Methods: A targeted paper survey (no more than two pages) will be designed with questions intended to elicit responses on three primary topics: 1) demographic information to establish land ownership, wooded acreage, livestock ownership, and current grazing management practices for livestock; 2) individuals’ understanding of the definition of silvopasture and the perceptions of silvopasture within their established definition, and; 3) characterization of the likelihood of silvopasture adoption given different scenarios, including management recommendations, financial cost-share assistance, and economic and environmental incentives. All questions will be confidential and voluntary, and our statistical model will account for responses left blank for any questions.  

Once developed, this survey will be revised with the input of producer focus group feedback. Three to five voluntary focus groups (a total of at least 15 people) consisting of producers from the target demographic will be assembled either in-person or via video conferencing. Individual feedback followed by group discussion will be used to revise the survey questions to ensure clarity and applicability for producers. These producers will not be included in the final survey results, and will be compensated for their time.

The survey will be disseminated via mail in partnership with collaborators from Missouri Forage and Grasslands Council (MFGC). A total of 3,000 surveys will be distributed through their existing extension, outreach, and organizational networks to generate a sample population of producers potentially engaged in woodland silvopasture. Responses will be anonymous, allocated randomly to blind participant identification numbers that will only be held by our collaborating groups to administer mailings. An opportunity to provide contact information to engage with a silvopasture peer-learning network (PLN) will also be included via a separate mailing card that is mailed with the survey.

Liabilities: Two follow-up reminders will be mailed at 20 day intervals to non-response households. The final follow-up reminder will have information for a separate “non-response” survey, administered and hosted by Qualtrics, an internet survey platform. Non-respondents will have the opportunity to submit this separate shorter survey, which will consist of a brief series of questions designed to identify the reason for non-participation in the survey. This data will not be included in the final statistical evaluation, but will provide descriptive information about the reception of the survey. We are targeting a 30% return rate for the mailed survey, and an additional 15% return rate for the “non-response” survey.

Specific Aim 1a: Utilize information gathered from the survey to develop a peer-learning network (PLN) among livestock producers in Missouri.

Methods: Any interested participants who volunteered their contact information for the PLN will be collected with our collaborating organization, MFGC. The participant information will be collected via separate mailing cards included with the survey, so researchers will not be able to connect survey responses with contact information. Responses will be categorized using identifying factors of either “practicing” or “interested in” silvopasture, and a mailing list will be compiled.

The formalized mailing list will be used to invite producers to register with a hosted network site, and they will have the opportunity to register at a practitioner or an interested party. This will lay the foundation for the PLN, where researchers will facilitate learning opportunities between practitioners and interested parties, disseminate research findings, and offer technical support for all members. A group chat board will provide a communal place for practitioners to answer questions regarding their own experiences, and researchers will identify model farmers to participate in field days, pasture walks, and digital media development.

Liabilities: Although the nature of the PLN will be dependent on the number of responses received, the success of this objective is mutually exclusive from survey response. The intent of this objective is to use survey responses to formalize and enhance participation in a PLN. Currently, a small and informal contact list is held by UMCA for individuals who are interested in learning more about silvopasture. In the event that no survey respondents volunteer contact information for PLN participation, we will begin developing the PLN through social engineering methods of network development.

While the survey responses will be critical to developing appropriate long-term research questions on feasibility and management practices for woodland silvopasture, the development of a permanent study site is necessary to conduct future experiments. The Wurdack Research Center in Cook Station, MO is part of the University of Missouri’s research station network, and consists of approximately 960 acres of upland hardwood forests that are typical of the northeastern area of the Ozark mountain range. Approximately 85 acres have been identified to develop into a long-term study site for woodland silvopasture. The site consists of upland till oak/hickory ecotype, and has remained un-grazed and minimally managed for the recordable history of the site. One of the primary concerns with incorporating cattle into woodlands, even in a silvopasture system, is the potential negative ecosystem impacts. This project will begin the development of the research site by characterizing the current ecosystem profile prior to any management actions, and monitor the changes over the course of the conversion process to quantify short-term and long-term impacts on the woodland ecosystem.

Specific Aim 2: Evaluate designated research site for baseline ecosystem measurements prior to silvopasture conversion, and monitor metrics for three years during the conversion process. Metrics of interest are plant community profile and abundance, with particular interest in quantification of local invasive species, hardwood tree community profile and abundance, existing forage quality and biomass analysis, and soil health metrics (soil organic carbon, bulk density, water infiltration, and microbial activity). Other opportunities exist for separate evaluations of wildlife profiles, insect populations, and water run-off quality, and are being pursued via separate funding avenues.

Methods: Beginning in Year 1 (Nov/Dec 2020), researchers will conduct site surveys using replicated 100’ transects. Measurement locations will be marked as permanent in order to return to the same location for evaluation. The designed site will be separated into 9 paddocks, approximately 8 acres in size, and the sampling locations will be determined such that each paddock has representative permanent measurement sites. Simultaneously, paired woodland areas with similar site characteristics will be evaluated as control areas. No management practices will be applied to these areas. The site is being developed to facilitate several different types of experimental designs and treatment structures for a versatile range of future studies. As such, this initial three-year project will be a completely randomized design with a single-factor treatment structure of two treatments: woodland silvopasture conversation (SILV) and woodland control (CON).

Soil sampling will occur twice every year over the course of the project, once in the late fall (November) prior to ground freeze, and again during the growing season (June). This will remain consistent during the course of the three year study, with the intention of continuing through future projects to ensure long term soil measurement. Soil nutrients will be measured via lab analysis, including carbon, nitrogen, potassium and phosphorus. Soil microbial activity will also be measured. Bulk density and water infiltration will be measured to assess soil physical characteristics effects over time.

Transects will measure tree and plant community profiles. Using random sampling techniques, 25 transects per paddock will be measured and then permanently located for the duration of the study. The locations will be evaluated twice every year (late fall and early summer) for plant and tree community profile and abundance. General categorization of plant material (ie: forbe, forage, hardwood, etc.) will be the primary metric, with additional species identification for invasive species and hardwoods. Frequency of the general plant categories will be analyzed and the proportion of specific invasive and hardwood tree species to the whole community will be evaluated over time.

Management practices that will be applied to the paddocks during the conversion process include thinning the tree stands, building infrastructure (water and fences), and establishing forage. After baseline measurements in November 2020, the paddock boundaries will be marked and the site will be scouted for thinning. The tree selection and thinning process will be well documented for extension purposes. Specific thinning and tree selection criteria will be confirmed after site evaluation with the research team’s forestry expert, but currently the estimated target is thinning to 40% canopy cover.

During the growing season, forage samples will be taken when available. Biomass will be estimated via destructive sampling (total forage clipping within a sample square), and quality samples will be randomly clipped throughout each paddock to evaluate nutrient content. Prior to forage seeding and establishment, the biomass samples will also be used to estimate the forage species profile to evaluate native forage abundance after thinning. Competition between native forages and established introduced forages will be monitored for the duration of the study after paddocks are planted via a broadcast seeder.

Liabilities: At this point, the most critical questions to answer for woodland silvopasture revolve around the impacts on the woodland ecosystem when cattle are introduced into the system. In order to separate between livestock effects and the effects of intentional woodland management, the conversion process must be monitored prior to the inclusion of livestock. Without knowing how the plant community and forage profiles will change with these management practices, we are unable to predict when the paddocks will be able to support cattle. Regular forage measurements will provide the data needed to estimate the best time to include cattle in a converted woodland system; we anticipate that this will be between two to three years. In order to mitigate output liability, this objective is not dependent on the inclusion of livestock in Year 3 of the study. However, if the site is prepared to graze by Year 3, this data will be simultaneously collected under a concurrent project protocol.

Specific Aim 2a: Evaluate the economic costs of converting woodland silvopasture and the value of the forage produced. We anticipated that one of the barriers to silvopasture adoption is the financial investment. As such, this objective will specifically evaluate the costs of woodland conversion and estimate the value of the forage produced to predict profitability for this type of system.

Methods: The real costs of thinning, seeding and infrastructure development will be documented. Additionally, the biomass and quality measurements will be used to estimate the feeding value of the forage produced in a woodland silvopasture. These estimates will be incorporated into existing grazing models to predict livestock performance of different classes of cattle, which will then be validated in future studies. The output for this objective will be used to develop a theoretical decision making support tool for producers to assist with their adoption of silvopasture. This tool will be updated over time with validated data.

Liabilities: This objective is dependent upon the forage quality and biomass estimates that are collected in Specific Aim 2. We anticipate both encouraging a native forage seed bank with thinning and sowing forages will result in a measurable forage stand to collect these estimates. However, it is possible that extreme weather conditions or unforeseen delays will prevent timely forage establishment by the end of the three year period, in which case this objective will be modified. The output will change from a decision making support tool to a producer budget worksheet, which will use our real costs to help producers predict their own potential cost for conversion. This in itself is useful information when approaching outside partners for cost-share funding, and would still be valuable for producers interested in navigating the cost of silvopasture adoption.

Research results and discussion:

Specific Aim 1: Evaluate current livestock grazing practices in wooded areas, perceptions of silvopasture, and potential for adoption among livestock owners.

The final length of the survey for Missouri producers was three pages (double-sided). A total of two participant focus groups were conducted, one in-person and one virtually, for a total of 15 participants. Feedback from these focus groups was incorporated into the development of the survey as anticipated. With the difficulties presented with the COVID-19 pandemic, the decision was made to include an online version of the survey for distribution in addition to mailed surveys. Additionally, the initial co-operator organization sourced for the distribution of the survey had a change of circumstances that did not allow them to participate as initially planned. Instead, we contracted the help of the Farm Journal-Trust in Food organization to assist with the distribution of both the online and mailed survey throughout the state of Missouri.

The online survey was launched Feb 21 2022, and the mailed survey will be distributed by March 15th 2022. Within a week, the online survey has already accumulated 80 responses, with the target being 250 total responses. The survey returned a total of 400 responses, 228 electronic and 172 mailed surveys (response rate 6%; n = 2,994), far exceeding our targeted number of responses. Preliminary data from demographic responses (n = 282) match closely with US Ag Census data, suggesting that the target population of the average Missouri farmer was successfully samples. Additionally, 52.8% of respondents indicated that they would be interested in adopting woodland silvopasture on their land and 46.5% indicated they would not (1% indicated they were unsure). More data analysis is needed to establish relationships between management, perceptions, and barriers to adoptions, but initial responses are encouraging that the survey will adequately reveal current woodland grazing practices among livestock owners in Missouri, and their perceptions of woodland silvopasture. This project is currently being incorporated into a dissertation chapter, as well as an independent manuscript. 

Year 3 data (in preparation)

In addition to the preliminary analysis indicated above, additional relationships have been examined to determine the current practices and motivations of MO producers when it comes to livestock and their woods. Additional key findings that are incorporated into the chapter/manuscript include observations on woodland and grazing management, as well as perceptions of silvopasture adoption barriers, and the critical motivating factors for considering adoption of woodland silvopasture. For example, over 96% of respondents maintained wooded acreage on the property that they owned or managed. Nearly a third of those respondents had wooded acreage on 26-50% of their operational land, and almost 38% of respondents reported the majority of their operational acreage was wooded (greater than 50%) (Table 3).

Additionally, these wooded acreages were characterized as either "moderately wooded" with an estimated canopy cover of 11-25% cover (42% of wooded acreage respondents) and "thickly wooded" with an estimated canopy cover of 26-50% cover (33% of wooded acreage respondents). The more sparse the canopy cover, the more moderate the understory was characterized as, and as the canopy cover increased, the more likely the understory was either bare or dense. Additionally, the more dense the canopy cover was reported to be, the less likely those woods were were be grazed. This suggests some level of woodland management and canopy conditions that are well-suited for the exitance, either current or future, of a silvopasture system, an assumption that is supported by the responses regarding forest management practices. 

Table with descriptive statistics
Description of Tree Canopy Cover and Understory Vegetation Grazed by Livestock

Specific Aim 1a: Utilize information gathered from the survey to develop a peer-learning network (PLN) among livestock producers in Missouri.

The development of the silvopasture PLN will largely be influenced by the participation response of producers from the survey. Over 45 of the 80 responses currently indicated interest in participating in the PLN, which already suggests the beginnings of a robust network of practitioners. However, the survey responses (ie: what estimated proportion of Missouri producers are currently practicing woodland silvopasture, and how many are simply interested in learning more) will provide more information on how best to structure the PLN. 

Specific Aim 2: Evaluate designated research site for baseline ecosystem measurements prior to silvopasture conversion, and monitor metrics for three years during the conversion process.

Due to delays with COVID-19, initial site development for this project did not occur immediately in 2020. However, paddocks have been virtually mapped out with 44 fifth-acre fixed area (FA) plots, and nine future paddocks differentiated. Four FA plots per future 8 acre paddock have been randomized across the landscape to represent the state of each paddock for paddock-level treatments. In addition to the nine paddocks, three unmanaged controls will remain untouched for the duration of the study that fall adjacent to, but outside, the paddock edges. 

Pre-treatment measurements were conducted in each FA plot, which included measuring overstory, midstory, and understory vegetation. Measurements included hemispherical photography to measure canopy cover, tree density, species identification and vegetation categorization by functional group. Preliminary data agree with the initial assessment that the site is a post-successional mesic landscape in an oak-hickory woodland ecotype. Average basal area of the site was 27.1 sq mt/ha (118 sq ft/ac), with an average stocking percent of 72%. As a mean percent of understory cover, graminoid species constituted 2%, forbs constituted 10%, and woody species constituted 16%. Primary tree species consisted of smaller-diameter white oaks (Quercus alba) and slightly larger diameter red oaks (Quercus rubra). Small-diameter hickory (Carya spp.) were also identified. Post-treatment measurements will include these same metrics, in addition to measuring biomass production of planted forages. 

A four color bar graph with tree DBH on the x axis and trees per acre on the y axis. The average of the curve is primarily on the left, with most stems being small diameter, and a long tail going toward the right in the 20-30 inch diameter range. Most species are white oak and red oak.
Overstory diameter distribution by tree species group
Graphic of bar charts describing the mean percent change in cover of plant community groups
Figure 5. Mean change in percent cover (± SEM) by plant functional group per treatment. Values denoted by a different letter indicate significant differences using Tukey's test (α = 0.05).
Bar chart describing mean stems per ha
Figure 4. Mean stems per hectare (± SEM) before (2021) and after (2022) treatment. Within year, values denoted by a different letter indicate significant differences using Tukey's test (α = 0.001).

All paddocks were marked using a method of tree-crop selection, modified for even silvopasture cover, to select trees to keep. The area has been contracted for a commercial timber sale for valuable trees, with any remaining un-marked, un-harvested trees (less than 10" in diameter) being removed using herbicide timber stand improvement practices. Due to the size of the area and some logistical delays for the project due to COVID, the treatment structure has been modified to be able to complete initial data collection on a smaller scale. Additionally, a prescribed burn treatment has been included in the experimental design. This means that, in addition to the three long-term controls for the whole site, the nine paddocks have been randomly assigned to one of three treatments: 1) Untreated control; 2) Mid-story removal of stems 10" or below, and 3) Mid-story removal with prescribed burn. Additionally, each FA plot in the treatment paddocks will have a split-plot treatment of either 1) seeding with a native forage mix or 2) not seeding. 

Year 2 findings:

In the growing season after the treatments had been applied, data were collected on plant community profiles and tree metrics. Although the only statistical difference in plant community cover changes were noted in shrubs, with thinning and burning treatment (RxB+MSR), there were numerical differences in percent cover change. The inclusion of burning numerically increased the cover of forbes when compared to both the thinning and control treatments, and both thinning and thinning with burning appeared to decrease the percent cover of legumes compared to control. More specifically, there was no difference (not pictured) between the seeded and non-seeded split-plot treatments, likely due to the late seeding date. It is possible that this forthcoming year may yield a latent effect on the seeding treatment, such that we will see significant biomass production over the non-seeded areas. 

Figure 4 represents the change in mean stems per acre of trees greater than 10" in diameter. The differences between year 1 and year 2 indicate that the treatment areas did experience a significant reduction in stem density, as expected. 

There were also treatment effects noted on the basal area, particularly in the burned areas, however those areas were already lower in basal area prior to treatment. Similarly, stocking percent exhibited a decrease between year 1 and year 1, but the burned treatments were already different. More telling, the percent in canopy cover difference was significantly reduced in both treatments compared to control, and there was no difference in canopy cover in the first year. This indicates that while total stem and basal area characteristics may have varied between sites, the actual canopy cover (and thus the light getting through the canopy) was significantly effected by both thinning and thinning with burning. 

From this first year of post-treatment data, we have learned several things. 1) Late seeding for native species will not yield significant forage production; 2) Thinning is needed to reduced overstory canopy cover while the addition of burning will reduce the midstory and understory; and 3) repeated burns and further reductions in stand density are likely necessary to see the changes needed to sustain a woodland silvopasture. This coming year, a large scale timber harvest has been completed on the site, so another survey will be done to assess how much more the stand was reduced, what the cover is, and what the effect is to the plant community. Following data collection, a fall burn is planned across the site. Not only will this manage the entire site for the subsequent years of work, but it will create a legacy of treatments to follow up on including light thinning with burning, heavy thinning with burning, and heavy thinning with frequent burning. 

Table quantifying basal area, trees per ha, stocking percent, and canopy cover %
Table 1. Mean basal area, trees per hectare, stocking, and canopy cover (± SEM) by year and treatment.

Specific Aim 2a: Evaluate the economic costs of converting woodland silvopasture and the value of the forage produced.

Costs of this process are carefully being monitored and recorded for the development of a decision support tool, but seasonal growing data will be needed for the completion of this specific aim. 

We are currently evaluating the costs and income from the timber harvest, but are limited in our biomass data to be able to begin development of this tool. We need at least a third, perhaps fourth year of growing data to get estimates of forage production. We are not prepared to modify this objective to a budget sheet at this time, but this decision will be more clear by this time next year. 

Participation Summary
415 Farmers participating in research


Educational approach:

With the survey, results of this part of the project will be able to be summarized and reported on very thoroughly for a variety of audiences, including an extension style report, newsletter and layperson outlets, as well as academic journals. The development of the PLN will also provide a novel opportunity for experiential education about not only the results of the project, but other research information on silvopasture and agroforestry.

This will be combined with field events and demonstrations that are well suited for the long-term site development of the silvopasture. 2021 already hosted several groups who were able to learn about the project and see the site (although no treatments had been applied yet), including stakeholders, practitioners and school children. The value of this site for visual and experiential education is significant, and will continue to be leveraged for the duration of the project and beyond.

Project Activities

Silvopasture Producer Survey Focus Groups

Educational & Outreach Activities

3 Tours

Participation Summary:

20 Farmers participated
20 Ag professionals participated
Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.