The purpose of the “Made in the shade” project was to research and demonstrate the intentional integration of trees and forage-livestock systems in the agroforestry practice known as silvopasture. Silvopasture systems offer tremendous opportunities to improve resource management; increase output and diversity in both woodland and pasture settings; provide multiple conservation services; generate additional short- and long-term economic returns; and enhance aesthetic appeal of farmlands. The project made significant strides to promote sustainable silvopasture systems in the mid-Atlantic, helping establish the region as a test-bed for new silvopasture research and a leader in demonstration and outreach efforts.
This project was conducted to address increasing requests from land owners for information and to begin tearing down key roadblocks to adoption. These include limited data on producer perceptions, system productivity, and economic outputs, and a lack of practical guidelines for establishment and maintenance. Greater effort has been needed to demonstrate and extend information about these systems, which are unfamiliar to most producers and technical service providers.
Surveys of limited resource landowners were conducted during the course of the grant suggest that producers are more aware of environmental issues than their agents and that they view silvopasture favorably, but they have a number of questions surrounding system utility or viability. Research and demonstration sites were prepared or established at three new locations and experiences from these efforts have been informative and will provide a basis for future plantings and management recommendations. Data generated showed the success of growing forages in moderate (between 25 and 50%) shade environments. Species differences were clear and weed abundance at establishment were lower in shaded environments. This research has management implications relative to forage botanical composition, nutritive value, and yield of complex cool season forage mixtures and will inform our recommendations.
Animal performance and behavioral studies demonstrated the over-yielding effects of silvopastures. This management practice can provide similar production to open pastures with the added benefit of trees – as long as shading is not excessive. These studies, conducted in the relatively mild temperate conditions of Virginia’s Ridge/Valley Physiographic region and using sheep, a heat tolerant animal, suggest that silvopastures will likely have even greater potential as a management practice for use with cattle in the hotter regions of the South.
Producer collaboration has been central to demonstrate practices and encourage adoption and they have played a key role in educational efforts. Site visits, field days, conference tours, and NRCS training sessions were among the many methods used to engage producers and technical service providers. Producers also volunteered their time to participate in a forage course , reaching over 100 students in the past three years. In future, we hope to make videos of farmers and farms to give them a larger platform and broader audience for discussing silvopastures in general and their efforts in particular.
The Made in the Shade grant has had a broad impact on our capacity to research and promote silvopasture systems into the future. Both a PhD and an MS student completed their degrees on this project and made multiple presentations in the field and at national meetings. One now is employed as a faculty member and working to develop additional silvopasture research and outreach programming. This program also engaged undergraduate and high school students. All these interactions represent opportunities to shape opinions and efforts of future leaders and scientists.
As work progressed, challenges due to changes in personnel and limits in the available data on hardwood fodder systems encouraged us to alter our plans. Thus, we began to explore the economics of developing pine-based silvopastures. A few versions of enterprise budgets have been created, but we are continuing to refine and debate these models and likely will make them public in a limited manner for the time being.
Efforts to create software that would allow us to put these systems before more people are in process. The current version allows us to “grow” trees but is tied to a specific site. Our vision of developing an app to grow trees on a producer’s “virtual” farm will require greater resources to make this a truly useful tool.
Three extension articles on silvopastures have been developed. Two currently are available from Virginia Tech’s silvopasture website: https://ext.vt.edu/agriculture/silvopasture.html ; the third is in post-production will be available soon. These originally were developed as a series of popular press articles that allowed us to reach a wider audience and to hone and refine our message for the extension publications. We have also developed curriculum for a high school, provided radio and web interviews, spoken and presented at multiple conferences regionally, nationally, and internationally. SARE support has helped us establish a beachhead for future research, training and demonstration efforts and is making it possible for people to learn the benefits of temperate silvopasture systems.
Given these conditions, our project objectives included:
1 – Create and use surveys to study adoption of silvopasture practices among producers and technical service providers. Data generated from survey responses will be used to develop outreach programs.
2 – Determine suitable forage establishment practices for “forest to silvopasture” establishment; and, trees establishment practices for success of hardwood-based “trees into pastures” silvopastures;
3 – Quantify animal and pasture productivity in ~20-year-old, deciduous silvopastures; and,
4 – Couple the establishment information with tree, forage, and livestock production data to estimate the economic value of silvopasture systems
5 – Use research center and on-farm field days to showcase silvopasture management practices and partner with beginning and small farmer and environmental networks and technical services providers to promote these sustainable systems.
6 – Develop web-based delivery tools to disseminate technical and budgeting information to producer and agency communities and create an application that can model developing silvopastures for producers on their farms.
Field studies fig 1 were conducted comparing lamb gains and behavior in different silvopastures in comparison with performance. This work was conducted because little information has been available from planted silvopastures based on hardwood trees that might add additional value to the system. In this case, the trees of interest were black walnut (which provides nuts and quality timber) and honeylocust, a source of edible fodders plus fuel and poles). Our null hypothesis was that animal performance would not differ among systems. Grazing studies (84 d) were conducted in the summers of 2014, 2015, and 2016 in Blacksburg, VA. Treatment paddocks for each experimental unit were small (0.66 ac), so sheep were used for our research. We monitored forage mass pre- and post-graze with each rotation (about once per week) and sampled for nutritive value at bi-weekly intervals. Stocking rates were adjusted for each system (open pasture, walnut silvopasture and honeylocust silvopasture) based on available forage, because we wanted to maintain similar levels of post-graze forage mass. This resulted in lower stocking rates for animals in the walnut silvopastures. Lambs were weighed at 28-d intervals and monitored each week with trail cameras and intra-vaginal thermometry to determine behaviors and body temperatures during the study.
Results and discussion
Forage and animal production
Average pre-graze herbage mass in honeylocust silvopasture was slightly greater (P<0.03) than in open pastures (4470 vs. 4390 lb/ac) and both these systems produced more (P<0. 01) forage than the black walnut silvopastures (3170 lb/ac). Forage in silvopastures had similar crude protein concentrations (average = 14.7%) which was greater (P<0. 01) than in the open systems (13.1%). Forages from the honeylocust silvopasture had lower (P<0.01) neutral detergent fiber concentrations (49.6%) than forages in the black walnut silvopasture and open systems, which did not differ (average = 51.4%). As a percentage, the black walnut silvopastures had (P<0.01) about half the tall fescue found in the other systems and generally had higher levels of orchardgrass and bluegrass. However, this system also had more low-quality forages and weeds than the other system, which were more similar in composition. Over the three years, lamb average daily gains were greater in walnut than in open system (72 vs. 58 g/d), but because of the lower stocking rate in the walnut system, total live weight gain did not differ among systems. Lambs for the study were obtained from commercial herds; lower gains in the first year likely reflect their older age and some health issues that limited our ability to detect differences among treatments – despite greater averages for silvopasture treatments. The observed differences in gain among treatments in year three occurred despite the fact that sheep available that season were hair sheep and likely less stressed than the wool breeds used in previous years. Sheep also may not be the best model for fully demonstrating the potential benefits of silvopasture because they are less sensitive to heat stress than cattle – the most common livestock for the South. Regardless, our data show that these systems do not need to sacrifice animal production, even if forage yields may decline, and they are capable of overyielding relative to open systems through the production of trees and tree products. Additional benefits in terms of sequestered carbon and soil and nutrient conservation warrant exploration to fully account the benefits of this type of management.
Animal behavior and body temperatures
The lack of differences in animal gain and similar forage characteristics among systems points to other factors affecting (and sustaining animal performance) in silvopasture systems that produce less forage. Our analysis of time lapse imagery from trail cameras indicated that lambs spent over 90% of daylight hours within shade cast by trees. When the shade moved, the lambs moved with it. This also reflects a benefit of silvopastures. Distributed trees create distributed shade, reducing the camping that occurs around solitary trees in pastures. Lambs in silvopastures spent more (P≤0.01) time lying down than lambs in the open pastures, the latter spending more (P<0.01) than 2 hours longer standing each day. These data indicate lambs in silvopastures were spending less energy in standing behaviors. Lambs in the black walnut silvopastures spent more (P<0.05) time (488 min/d) grazing than lambs in the honeylocust silvopastures (438 min/d) and lambs in the open pastures (417min/d), perhaps reflecting more limited forage supply. Lambs in silvopasture systems had lower (P≤0.02) temperatures than ewe lambs in the open pastures during the hottest hours (1300–1700 hr) of the day. Ewes in the open pasture experienced more fluctuation in day to nighttime core temperature change (P<0.0001). The significance of the greater temperature fluctuation for lambs in open systems is unclear. For cattle, lower temperatures at night during periods of heat stress is helpful for dissipating heat; thus this may be a coping mechanism for lambs in the open systems. Lambs in the honeylocust silvopastures displayed increasing vaginal temperatures each month, perhaps due to the loss of tree shade over the summer as these trees shed leaves. Clearly, trees in silvopasture systems can alter animal behavior and physiology, and suggest that these changes may compensate for lower forage availability. We interpret the lack of distinct, consistent differences in gain among treatments as a function of our experimental set-up and environment and anticipate future work in larger systems to test the performance and behavior of cattle.
FORAGE ESTABLISHMENT AND PRODUCTION EXPERIMENTS
Forage production in silvopastures can modify production and nutritive value with changes in microclimate. However, these effects are variable given each system’s unique structure and environment. Species diversity can support greater productivity and environmental resilience in grasslands, but this area of effort has received little research for silvopasture systems. Thus, we evaluated the impact of shade levels and mixture complexity on the yield, botanical composition, and nutritive value of forage mixtures (simple = tall fescue and white clover; intermediate = simple + orchardgrass and red clover; and complex = intermediate + Kentucky bluegrass, birdsfoot trefoil, and alfalfa). This work was conducted in a replicated plot trial in Blackstone, VA, in the North-South transition zone fig 2. Given the challenge of having consistent light levels across a silvopasture, we utilized slatted structures to create conditions of 30, 50, and 70% shade relative to a full sun control.
Results and discussion
Forage yield with reduced light environments
Annual yields of cool season forage mixtures were sustained at 70% of full sun but reduced by 22 and 36% at 50 and 70% shade, respectively fig 3. Yields did not differ with mixture complexity. During the hottest part of the year, when cool-season forage growth is reduced, shade tolerant species thrived under the modified microclimate of the 30% shade treatment. Orchardgrass and red clover performed particularly well beneath shade and are recommended for planting systems created though forest conversion. Although orchardgrass is not particularly well-adapted to the transition zone between the northern temperate and southern sub-tropical United States, silvopastures may be an effective way to integrate these shade tolerant cool-season forages into this region’s grazing systems.
Forage nutritive value
In contrast to what is often seen in the literature, shading reduced forage crude protein in spring 2016. However, forage crude protein increased with shading in summer 2015 and fall 2016. These differences indicate that seasonal differences interact with light (and perhaps other microclimatic) conditions. Simple mixtures, which were dominated by fescue, also had lower crude protein in spring and summer of 2016, but otherwise mixture treatments had little effect on protein.
Forage acid detergent fiber concentrations were higher with shade in spring 2015 but declined with shade in summer. The same pattern was observed in spring 2016, but in summer these fibers were highest under moderate (30 and 50%) shade. Similarly, neutral detergent generally was increased with increasing shade and most noticeable at spring harvests each year and summer 2016. When TDN was calculated, levels were lower in spring under shade (each year). Values were slightly greater with shade in summer 2015, but no clear pattern was observed in summer and fall 2016. Preliminary analyses of alkaloid concentrations in tall fescue found lower levels as shade increased to about 50% shade, but not at 70% – and this was not observed in 2016. This work has been slowed by a change in jobs for colleague Teutsch, but we are continuing to work on this area of investigation.
TREE ESTABLISHMENT EXPERIMENTS
Tests of establishment treatments were conducted at one of the demonstration sites. Oak ad black walnut trees were established in ground preparation and protection studies. These studies are on-going looking at tree production but the important and visually distinct features are that unprotected (no tube) walnut trees appeared to tolerate the lack of protection much better than oaks, which are likely subject to greater predation by deer. Predation was a problem even with tree tubes, however, as evident by the “broomy” look of trees browsed just above the top of the tube.
Trainer trees also were planted at this demonstration site and we were surprised by the initial successful planting of bald cypress. Even on uplands in the first year, establishment success was greater than 90%, while “hardy” pitch-loblolly hybrids had about 66% survival. White pine and shortleaf pine had more than 50% mortality, which may represent lower seedling quality or poor handling practices. Re-establishing shortleaf range is of interest to forestry departments in the Southeast and subsequent replantings have been more successful, but white pine has fared poorly, and this in part appears to be a problem of deer predation. These practical experiences are guiding our recommendation to use simple, three-dimensional tree fencing to protect new seedlings where deer pressure is high. We are also less concerned about “weedy” warm-season species such as broomsedge in our tree rows. Our empirical experience indicates these species can hide trees or make browsing them more uncomfortable for deer, or both. Getting producers to tolerate some weeds in fields to during the establishment phase may be another issue, however.
A survey of limited-resource producers from several counties Virginia was conducted to gather their understanding of silvopasture. Participants filled out paper copies that were hand tabulated. The data from the producers is interesting although the sample size (43 producers) was limited. For a bit of comparison here, we have combined data from the producer survey with a survey of extension agents. Agent participants were from several southeastern states and they received and took the survey on-line, prior to receipt of the SARE grant. Survey questions were aggregated by theme into four blocks regarding issues related to the environment, economics, social constraints and technical comprehension of silvopasture. Additional data were gathered on demographics. For some questions, agent data were separated into Virginia and non-Virginia peer groups.
Average age of all respondents was about 50 years and about 40% of producers were female (versus about 22% for agents). About 50% of producers (versus 100% of agents) had four years of college and producer educational backgrounds were largely outside of agriculture or natural resources (76% versus about 40% of agents).
Across all questions, agents generally had stronger opinions than producers. The more circumspect responses of producers was evident in a larger percentage of responses in the “Neither agree nor disagree” category fig 4. This likely reflects the differences in educational backgrounds.
Producers had stronger opinions about environmental issues than agents. More producers than agents saw livestock as contributors to non-point source water pollution, and Virginia agents were more sensitive to this issue than agents outside of Virginia (Table 1). We can only speculate about these differences, but suggest that producers, working with livestock more frequently, may recognize or relate to this issue to a greater degree. In turn, agents from Virginia may be more sensitive given the state’s challenge to deal with Chesapeake Bay water quality. Conversely, Virginia agents were more likely than producers and other state agents (average of 67%) to view silvopasture as a practice that could benefit water quality. Producers and agents alike did not view livestock as a major contributor to greenhouse gases, although producers, again, were both more likely to agree and less likely to see benefits of silvopasture (by sequestering carbon). In contrast to agents, producers did not consider pasture systems to be low in diversity and they saw less benefit of adding trees to increase diversity than agents.
|Table 1. Average of producer and agent responses (%) who responded positively (“agree” or “strongly agree”) to questions about environmental concerns|
|Statement||Va producers||Va agents||Non-Va agents|
|Livestock production is a significant contributor to non-point source water pollution.||58||42||28|
|Silvopasture systems can improve water quality by reducing runoff from pastures into streams.||65||80||69|
|Livestock production is a major contributor to greenhouse gasses.||24||13||6|
|Silvopasture systems can sequester more carbon than open pasture.||41||69||57|
|Pasture systems that produce only livestock are low in biodiversity.||33||93||76|
|Including trees in agriculture systems can benefit biodiversity.||78||81||95|
Both producers and agents alike expressed a great deal of uncertainty about economics of silvopasture systems (Table 2). Almost two thirds expressed no opinion about the ability of silvopastures to produce higher profit margins than traditional forage-livestock systems. A large majority of respondents felt diversity was important for farm financial viability and nearly two third expressed a positive view about the possibility of silvopasture reducing financial risk by diversifying income streams. This suggests a potential point of entry for promoting these systems, but the uncertainties expressed by farmers and agents also reflects the need for continued economic research and modeling.
|Table 2. Distribution of the average of producer and agent responses (as %) among Negative (“disagree or “strongly disagree”), Neutral (“Neither Agree nor Disagree”), and Positive (“agree” or “strongly agree”) categories in response to environmentally-related questions.|
|Tree production and timber value are always diminished when livestock graze in woodlands.||40||37||23|
|Forage livestock systems have low profit margins.||41||35||24|
|Silvopasture systems produce higher profit margins than forage livestock production alone.||15||63||22|
|Diversifying agriculture systems is necessary for farm financial sustainability.||5||10||85|
|Silvopasture systems can help reduce financial risk for producers by diversifying income streams.||4||33||63|
In class instruction was an important component of our educational approach, using real-world examples of silvopasture adoption models. For the past three years, two of our producers volunteered their time to participate in a forage course for Virginia Tech’s two-year associate’s degree program in Agriculture Technology. These producers have taken two very different approaches to silvopasture implementation. One has taken a Ridge/Valley farm in the family for 11 generations and established silvopastures over much of his pastureland by planting trees. This work has been somewhat experimental as he has planted different tree species and configurations to see what will work best for his site and his cattle, and he is the first recipient of NRCS funds for silvopasture implementation in Virginia. The second producer has leased land in the Piedmont that has a loblolly pine stand and sensitive watering sites. He has thinned and improved the stand in the process of creating silvopastures for his growing sheep and goat flock. Each year, both producers have visited to tell 30 to 40 students about their silvopasture operations. Having actual producers who have implemented these different systems talk to the students is particularly effective. The students come from different regions and backgrounds, so showing these different practices better informs them of the multiple types of (and opportunities for) silvopastures. We have reached over 100 students this way over the past 3 years and several have expressed interest and intent to try these practices back at home on the farm. Many of these students also assisted with tree planting at one of our demonstration sites during different years as part of the class lab or service-learning activities.
Educational & Outreach Activities
Estimate of consultations (54) assumes the lead PI and colleagues consulted or answered questions about silvopasture about 1.5 time per month over the course of the project. Given the number of participants, this likely underestimates the total number. Many of the consults involved response to email- or phone-based queries, and several led to on-farm visits.
A curriculum developed during the course of the grant was used in conjunction with a Randolph Henry High School’s agricultural program. Over the course of the grant, several handouts were created for demonstrations and field tours. These were not developed as official fact sheets, however, and that was an oversight on our part.
Five on farm demonstrations of silvopasture were held in the context of extension field days and pasture walks at collaborating farms. Farmer-operators backgrounds included large and small producers with both short- and long-term histories on their properties. One is considered a limited resource landowner. Farm systems were in Ridge/Valley, and Piedmont regions and one farm was in the transitional zone between these ecosystem types. Some farmers had planted trees, while others had thinned forest or stands of plantation pine or both.
Of five published press articles, three were published in Progressive Forage and discussed the rationale for establishing silvopasture, and the methods for doing so when thinning or planting trees. Additional articles were published as updates in the quarterly newsletter of the Virginia Forage and Grassland Council. Progressive Forage articles can be reached at the following links:
Tours involved taking groups on trips to collaborating farms and research sites. Of the seven tours, two were conducted as part of larger annual efforts to showcase timber and forest management practices. Tours of two collaborating farms and three research and demonstration sites occurred in conjunction with the North American Agroforestry Conference. Many impromptu tours to the research sites were not considered in this accounting.
A total of at least 25 webinars, talks, and presentations on our efforts and findings have been held around the region, in the US, and in Europe. A webinar describing agroforestry and silvopasture practices more generally was provided in conjunction with the Small Farmer and Rancher Coalition efforts at Virginia Tech. Oral and poster presentations of research findings have been made by team members and graduate students at meetings of the American Forage and Grassland Council, the American Society of Agronomy, the National Small Farmer Conference, the North American Agroforestry Conference, Virginia Governor’s School, and the World Congress of Silvo-Pastoral Systems over the past three years. This likely misses a few but should provide a sense of the several and wide audiences that were reached.
Workshops involved Joint Education Development training for NRCS and other technical service providers. These efforts occurred both on collaborator’s farm as well as on research stations in Virginia (Southern Piedmont AREC) and Alabama (at the invitation of Auburn University). Three field days which included some focus on silvopastures were held both at the Southern Piedmont and Shenandoah Valley ARECs in Virginia.
We also conducted a radio interview describing our efforts to measure effects of silvopasture on sheep grazing behavior and the recording technology captured to do this was also described.
Interest in silvopastures continues to grow as we work with landowners, extension personnel, and technical service providers. We are getting our results out to producer and scientific communities through extension publications and journal articles, both of which had been missing for researchers and technical service providers. (A “key informant” study we conducted for the forest service indicates most silvopasture knowledge comes from discussion and popular press articles and that there is little solid literature available on the topic.) SARE support has helped us
Economic assessment continues to be a challenge. However, similar weight gains for lambs in silvopastures indicates that introducing trees to pasture systems can increase the total capacity of the landbase by 25% or more and we anticipate this gain will be as good or better with cattle. Silvopastures likely will support greater reproductive success, which may be particularly beneficial for producers with spring-calving herds. Heat stress often reduces conception rates for those animals. A modest 3% improvement in conception rates and calf yield over 20 years of a 30-year production cycle, a farm with an average-sized herd of 35 cows would produce an additional 21 calves. At an average value of $600/calf, these additional animals would return another $12,600 to the farm. Returns are likely to be even greater, however, as research indicates that managing cows in silvopastures reduces calving difficulty and increases calf-weaning weights in spring calving herds – and these systems have potential for timber revenue or on-farm wood use.
Environmental values of silvopasture adoption likely can be accrued through reduced runoff and erosion, reduced stream use by cattle, greater nutrient capture and use efficiency, and increased carbon sequestration. Measure of these effects was beyond the scope of this work. However, the requirements for improved management that occur with silvopasture implementation indicate that these benefits are likely to follow as producers begin adopting these systems.
Silvopastures will likely also accrue greater social and cultural benefits over time as concerns over animal welfare increase. Producers are practical people, but they appreciate aesthetically appealing landscapes and these systems have public (and personal) appeal. In some cases, producers have taken advantage of this management system to provide environmental services (wildlife habitat) that support hunting, reaping rewards for hunting leases in return.
SARE’s support of this project helped us share the silvopasture concept with hundreds of producers and we are getting regular calls for support in helping approach the design and implementation of these systems. At times, field day attendance specifically devoted to silvopasture has not been as large as the lead PI desired, but a colleague noted that we often were getting the early adopters at these sessions who are figuring out where and how to adopt these systems to their farm operations.
One of the most significant impacts that we likely are having is in the training of the next generation of farmers and scientists. Many students have had the opportunity to be on farm with silvopastures or to hear presentations from silvopasture practitioners. In addition, graduate students who have completed their study with silvopastures and continue to explore these systems as new faculty. Introduction to the next generation which is expressing enthusiasm for adoption is perhaps one of the best opportunities for moving this practice forward and onto farms
One observation from this research is that additional work with legumes is warranted. Tree legumes that add value as tree or fodder crops and shelter and forage legumes that perform well in shaded environments offer opportunity to boost system productivity. In our walnut systems, it was not clear if lower forage production occurred due to lower available nitrogen (because this system had less legume). In that system, the question of whether the lower legume presence was due to low light or allelochemicals from the trees was (and is) a question. Although we observed greater gains in honeylocust systems in some years, we were surprised that lamb gains in these systems were not greater. We do, however, have some preliminary evidence that lambs allowed to browse honeylocust leaves have lower levels of intestinal parasites. As well, data from a short term grazing study conducted during the grant period indicated that honeylocust pods can be very nutritious. More work is needed with this species and we continue to receive requests for information on these trees.
Economic work, system evaluation with cattle, and continued training for trainers as well as for producers will be important for understanding and adoption.