Sheep Vegetation Management in Oregon’s Willamette Valley
The forests of Oregon have long been important forage sources for sheep and other livestock. The first domesticated sheep arrived in Oregon with the Spanish conquistadors who established temporary settlements as far North as Vancouver in the late 1700s. By 1850, nine years before Oregon became an official state, there were over 15,000 sheep in the territory (Paul, 1976). Although some of this grazing takes place on dedicated pasture, grazing animals also live and eat in Oregon’s state, federal, and private forests. Today, more than 1/5th of all forestland in the United States is grazed. This land area exceeds all pastures and grazed croplands in the United States (Sharrow et al. 2009). The primary purpose of forest grazing is to make use of available forage and to a lesser extent as vegetation control for tree establishment, fire control, and weed control. Grazing on open range by cattle under low intensive management and extensive native forage use remains the primary approach to forest grazing.
The potential for sheep vegetation management (SVM) within existing hardwood and conifer forestlands is balanced by the concern for grazing’s historic impact on landscapes. The history of open-range grazing in the West provides plenty of evidence for the negative impacts of unmitigated grazing on landscapes including soil erosion and compaction, reduction of desired plant species, and the invasion of weeds associated with overgrazing (Sharrow, 1994). However, evidence exists that well managed grazing can be a force for landscape regeneration even within existing forestlands. Research suggests that vegetation management by sheep can serve as an alternative biological solution to the use of herbicides. Additionally, the successful integration of livestock into forest environments can increase the financial returns for landowners, therefore paying the way for increased investment in land stewardship and bringing neglected plant communities back to active management. Public concerns surrounding the status of Oregon white oak (Quercus garryana) habitats and the impacts of plantation forestry have led to a renewed interest in the beneficial financial and ecological outcomes of silvopasture within both of these forest types. To this end, this review explores the use of sheep as a vegetation management tool to achieve specific desired outcomes for both the oak and Douglas-fir forest types.
The Biogeography of Oak and Douglas-fir Habitats in The Willamette Valley
The Willamette Valley lies within the Willamette ecological province of Western Oregon. The Willamette province is a region bordered by the Cascade and Coast Range mountains to the East and West and the Columbia and Umpqua rivers to the North and South. It is characterized by its maritime Mediterranean climate whose mild temperatures and moisture result in prolific plant growth while arid summers produce unique plant assemblages adapted to extreme dry periods. The Willamette River charts a Northward course through a 20-40 mile wide and 150 mile long floodplain originally composed of a mosaic of riparian forests and wet and upland meadows. Most of the valley floor has been developed and converted to agriculture. Although the drainage basin of the Willamette reaches to the peaks of both mountain ranges, the province its self reaches approximately to the 1,700 foot elevation line in both ranges. At this elevation, forests become dominated by Western hemlock. These trees mark a transition to the Cascade and Coast Range provinces. All together the province covers 6.2 million acres.
A significant portion of the Willamette Valley is in rolling foothills and steep uplands. This landscape, once dominated by white oak (Quercus garryana) savanna, has been converted almost entirely to mixed conifer forests. Written accounts by the first naturalists and pioneers to enter the valley describe a wide savanna landscape of prairies and scattered oak trees. The valley was almost completely clear of fir trees. In fact, so few trees marked the landscape that in many cases, rocks were piled to mark section corners instead of using “witness trees” (Vesely et al. 2004). In summing up observations made in the 1840’s by early settlers, Towle (1974) stated that the valley presented a “remarkably open landscape set between the dense forests of the Cascade and Coast Range.” In reviewing observations made by a 1941 expedition he went on to remark that “the prairie occupied more than one-third more of the valley than did woodland is a fair general estimate.” By reviewing land survey records made when the valley was being subdivided during the 1950s, contemporary researchers have concluded that oak savannas and oak woodlands were the dominant forest type (Habek, 1961). These oak savannas were characterized by their open canopies dominated by trees with full, mushroom shaped crowns. In the land surveys, savannas were classified as areas with trees 50 or more feet apart. Average distance between trees was over 140 feet. Oak savannas were widely distributed in the survey transect with one township being 90% savanna (Anderson et al. 1998). Dense oak forests covered a small fraction of the valley floor and foothills while conifer forests were found only at higher and steeper elevations on either side of the valley (Anderson et al. 1998). Practically speaking, the original conifer distribution occurred outside of the Willamette province and occurred within the Cascade and Coastal ranges. The other plant communities of the Willamette province included riparian forests within the floodplains along creeks and rivers and upland and wet prairies.
2.1 The Role of Succession and Human Management in Maintaining Oak Habitat
Upland forests in the Willamette valley exist in the context of two divergent plant communities. Succession, the progression of plant communities from a pioneer phase of grass and forbs to shrubs through to mature forest, has two very distinct pathways in this region. One pathway leads to dense multi storied mixed conifer forest and the other leads to a prairie plant community of oak, lilies, grass, shrubs and occasional large conifers. Where a landscape sits within this continuum depends in small part on soil type, aspect, depth to bedrock, and hydrology and a very large part on the human relationship with the landscape. Native American fire management for lilies, nuts, game, and fibers has seen the uplands of the Willamette valley in savanna type environment since the last ice age (Boyd, 1999). However, since Euro-American contact in the Americas and the loss of Native American stewardship, plant communities in the Willamette uplands have tended rapidly toward conifer forests. It has been noted that the influence of Euro-Americans on the abundance of oak woodlands predates formal settlement by 200-300 years due to the impacts novel diseases had on human populations (and therefore land management) across North America as early as the 1500s (Apostol et al. 2006). More recent intensive land use and development in the Willamette Valley has led to massive changes in landscape function and composition. The loss of fire as a tool for prairie and savanna management and the conversion of these systems to modern agriculture and forestry use has resulted in one of the most complete conversions of habitat types in North America.
Today, within the Willamette valley, less than 1% of the original white oak ecosystem remains (Vesely et al. 2004). This decline has led to the Nature Conservancy declaring the oak savanna plant community as the most endangered ecosystem in North America (Bell, 2011). The historic plant assemblages and structure of oak savannas were maintained through a combination of fire, drought, soil conditions, herbivory, and flooding. As already discussed, the largest factor in the maintenance of these landscapes however was human management. Specifically the use of prescribed fire as a tool to improve acorn production, spring forage quality, and to assist in the harvest and cultivation of other food, fiber, and medicine species.
Lacking the intermediate level of disturbance within which oak savannas have co-evolved, an oak tree can become outcompeted by other more vigorous forest species like Douglas-fir. Because of its large size and capacity to grow rapidly in a high light environment, Douglas-fir will quickly overtop and shade out oaks. Over time, this encroachment causes oaks to lose their lower branches and their crowns to become vase shaped. Gradually, acorn productivity decreases and oaks fail to establish seedlings (Vesely et al. 2004). This encroachment transforms oak savannas into conifer forests over time.
Today, the restoration of the oak savanna ecosystem is a high priority for restoration practitioners seeking to improve habitat and species diversity in the Pacific Northwest (Apostol et al. 2006). Maintenance of this intermediate succession landscape takes significant investment of time and resources. Given the complexity of prescribed burning projects, mowing and herbicide control of vegetation are the primary tools for maintaining restored oak meadows. Although when listing the management options for oak restoration, reference is often made to the use of sheep and other grazing animals for vegetation control, little specific data has been compiled in regards to the viability of SVM in oak savannas. The following pages attempt to draw a picture of SVM’s potential within these environments.
2.2 Douglas-fir Management in the Willamette Valley
Following the loss of Native American Management and the shift to Euro-American timber and grazing use, the last few centuries have seen a relatively swift conversion of the valley uplands to conifers. Land use laws in Western Oregon essentially mandate the management of the majority of the upland landscape be managed under forestry production. The vast majority of this land use focuses on the production of Douglas-fir (Pseudotsuga menziesii), a fast growing, high-value conifer. Douglas-fir trees are most typically managed in large even age units on 40-60 year rotations. Following timber harvest, Oregon law requires seedling trees to be established in a free-to-grow condition within 2 years.
In recently cleared or newly established conifer plantations throughout the foothills of the Willamette Valley, rampant pioneer species like Scotch broom (Cytisus scoparius), velvet grass (Hulcus lanatus), and Himalayan blackberry (Rubus armeniacus) quickly revegetate in the cleared ground amidst seedling trees and ultimately limit the success of Douglas fir establishment. Additionally, hardwoods like bigleaf maple (Acer macrophylum), and Oregon ash (Fraxinus latifolia) re-sprout from their stumps and throw up significant competing vegetation amidst the seedling trees following clearcutting. In order to establish a “free to grow” condition amongst the tree seedlings, a series of herbicides and adjuvants is used to clear vegetation before and after planting. Typically, three applications of herbicide are used in units facing significant amounts of competing vegetation. While in well-controlled and accessible sites, spray application by backpack crews may be viable, on steep ground and in areas with well-established competitive species, aerial application by helicopter is the norm. Recently, in attempt to make up for dwindling soil fertility, it has also become common to apply nitrogen pellets to the replanted units by helicopter. The application of herbicides and synthetic nutrients in replanted Douglas fir stands has led to increased public concern for environmental and public health and has resulted in a renewed interest in managed grazing as an alternative non-toxic vegetation management tool.
3 Analogous Agroforestry Systems
A good starting point for exploring SVM’s application to Douglas-fir and oak production is to review successful systems already at work in the world. Perhaps the best-known oak/sheep system is the ancient and biodiverse Dehesa System of Spain. This system has maintained the intermediate succession oak habitat for many generations across a significant land base. Located in the Southwestern region of the Iberian Peninsula of Spain and Portugal, Dehesa systems cover 3-4.5 million hectares. With its wet and relatively cold winters and dry summers, this region shares a climate profile similar to the inland prairies of Western Oregon. It has a direct climate analogue with the Umpqua River Basin just South of the Willamette Valley’s headwaters. The Dehesa is an extensively grazed evergreen hardwood savanna ecosystem. Three layers, a tree, shrub, and herbaceous layer are managed for livestock forage production, cork, and other products.
The major livestock species of the Dehesa is the Merino sheep. Other animals including cattle, horses, goats, and pigs are used throughout the growing season and rotated over a series of years to make use of shifting plant resources throughout the growing seasons. Pigs, which forage for acorns in the fall, and sell for up to $50 per pound, are another significant yield. Livestock is utilized to maintain the savanna condition by holding the emergence of brush at bay. Tree production is focused on producing acorns, cork, browse, and fuelwood and prioritizes increasing tree crown cover rather than on stem growth for timber. Tree crown coverage and distribution has been found to be an important factor for wildlife diversity within the Dehesa (Olea et al. 2006). Similar to white oaks in the Willamette valley, which are known to be host to hundreds of insect species and habitat for cavity nesting animals, the oaks of the Dehesa represent a keystone species in the savanna for wildlife health. Pasture management is achieved through shifting grazing and browsing species, rotational management, annual to tri-annual sowings of forage grass and legume species, and occasional inputs of nutrients (Olea et al. 2006). Consistent with the silvopastoral goals of Western Oregon, forage shrubs and trees as well as nuts and berries are an important supplement to livestock during the Summer and Winter hunger periods.
An often cited example of conifer based SVM is found throughout New Zealand, Australia, and Chile. In this case, radiata pine (Pinus radiata) is grown for saw timber and grazed by sheep as a means of harvesting the understory grass crop. This grazing has been seen to not harm young pine seedlings (Sharrow, 2006).
A second example of a Douglas-fir/sheep system is found closer to home on the Crown Lands of Canada. In Canada, an increasing willingness by forest managers to reduce the use of herbicides to control vegetation because of concern for environmental impacts has led to adoption of sheep vegetation management as an alternative biological control strategy in young hybrid spruce (Picea glauca x Picea engelmannii) plantations. A study in the central interior of British Columbia found that SVM could be an effective method for controlling competing vegetation. In this case, shepherded sheep helped the seedling trees achieve a “free to grow” status on slopes as steep as 28%. The modern case of successfully applied SVM on Canadian Crown lands stands out as perhaps the best case for adoption of this practice in Western Oregon.
Considerations for Sheep Vegetation Management
4.1 Foraging Habits
Foraging habits of ruminants follow a spectrum from grazers to intermediate grazers to browsers. Grazers like cows mostly consume grasses while browsers such as goats and mule deer primarily eat nutritious twigs and shrubs. Sheep are intermediate grazers and have nutritional requirements midway between grazers and browsers – they can successfully forage on forbs and grasses as well as shrubs. Sheep have also been found to preferentially consume hardwood species over the less palatable conifer species (Lavendar et al. 1990). While this makes grazing in hardwood settings more challenging, it complements grazing in conifer plantations. For example, Phelps (1979) reported 47% reduction in understory vegetation with little to no browsing on Douglas-fir, silver fir (Abies amabilis), and western hemlock (Tsuga heterophylla). Spruce species in particular have been found to be least likely to be browsed by sheep, even under intense grazing pressure (Adams, 1976).
Compared to most other ruminants, sheep have a limited stomach capacity compared to their total body mass. Because they need to utilize their stomach capacity efficiently, sheep are selective feeders and maintain an optimal diet by foraging broadly for the highest quality foods (Sharrow, 1994). Because sheep will select the most desirable plants, timing grazing to coincide with palatability of vegetation specifically targeted for removal is a main principle in sheep vegetation management. Because of their preference for herbaceous vegetation, sheep are good choices for vegetation control when the establishment and maintenance of more mature or woody plants is desired. Additionally, sheep are capable of maintaining a prairie-like conditions by consuming both herbaceous and young woody plants establishing in open pasture.
4.2 Forage Quality
Sheep used for vegetation management often consume lower quality forage than if allowed to graze freely or on improved pasture. SVM can consequentially result in lower weight gain (Sharrow, 2006). This would be especially true in replanted clearcuts not sown to forage species or in oak restoration projects emphasizing brush control over pasture maintenance. A study conducted in the Cascade Mountains of Washington found that ewes lost 23 lbs. during summer grazing in a Douglas-fir forest (Phelps, 1979). A second study in British Columbia found that sheep grazing a replanted spruce forest gained 68% less weight than sheep grazing improved pasture. Poorest weight gain probably results from sheep being forced to eat mostly woody vegetation whose high tannin content reduces protein availability in the animal’s stomach (Sharrow, 2006).
Adequate monitoring combined with grazing periods adjusted to grass productivity and herd size will lead to improved weight gains. Hedrick (1966) demonstrated that by monitoring for weight gain through the grazing period from the time there is adequate forage until the forage supply drops off and animals ceased to gain, spring growth of rampant vegetation is held at bay while animal productivity is maximized. Once weight and forage reaches a critical threshold, animals can be moved to better pasture if it is available. This period may last about a month in many un-seeded sites in the Willamette foothills (Hedrick, 1966).
Improved weight gains are also achieved by sowing pasture in the form of a legume/grass mix. Restoration minded grazers in oak woodlands may choose to include self-sowing and perennial forb species in their legume grass mixes as a way of increasing native plant and pollinator biodiversity. Another benefit of seeding in forage comes from a process called competitive exclusion, in which one species benefits while outcompeting undesired competitors. One study showed that the combination of seeded grass competition and grazing slowed the establishment of red alder (Alnus rubra) to the extent that, by year 10, Douglas-fir basal area was 50% greater in grazed areas (Sharrow, 1992).
4.3 Sheep Breeds
The characteristics of sheep for Willamette Valley silvopasture need to include cold tolerance, ability to browse woody and course perennial vegetation, quality meat and or wool production, adaptation to steep rugged terrain, and a tendency to form a moving herd in the backcountry.
In her book The Backyard Sheep (2013), Sue Weaver consulted many sources to compile a list of sheep breeds and grouped them by function and yield. Based on this list, if one were to select one breed for trials in the Willamette Valley, it would appear that the Icelandic sheep is particularly well adapted (See Table Section 8.1).
4.3.1 Highlight on Icelandic Sheep
One of the world’s oldest breeds, Icelandic sheep have evolved over 1,100 years under difficult farming conditions. As a result the sheep are sturdy, cold tolerant, and have a strong constitution. They are known to provide a triple yield of meat, fiber, and milk (ISBONA, 2016). A defining quality of the Icelandic breed is the ability to survive on poor-pasture and browse. Perhaps most relevant to vegetation management, selective pressures from their history in the Icelandic backcountry has resulted in their having very large rumens. This trait leads to a higher feed efficiency than other sheep breeds and a higher capacity to browse woody forage (ISBONA, 2016). The ability to add weight on a completely foraged diet positions them for being good sheep for marketing the value added benefit of grass fed meat to local consumers.
4.4 Herd Management
Open range or forest grazing lacks the intensiveness of management and planned interactions that would otherwise characterize it as silvopastoralism. Silvopastoralism is a practice that intentionally integrates trees, livestock, and often improved pastures, into a system of mutually supportive planned interactions (Sharrow et al. 2009). The two most common approaches to silvopastoralism in the United States are integrated forest grazing and silvopasture (Sharrow et al. 2009). Integrated forest grazing uses livestock grazing and browsing to enhance the health and productivity of forestlands. While integrated forest grazing involves two components – livestock and trees, silvopasture involves three by combining trees and livestock with improved pasture. The management goals in both of these systems include fire control, tree establishment, brush control, management of intermediate succession ecosystems, timber, forage, and animal products.
In conifer plantations, the total acreage tends to be large and the terrain uneven. Unless the plantation is relatively small and located on even terrain, herds will likely be maintained by shepherds and not kept in management intensive systems. Sheep travel frequently while grazing so browsing is typically spread evenly across grazed areas (Sharrow, 2006). Additionally, mountain breeds, like Icelandics, as opposed to farm breeds, tend to stay in more uniform herds without breaking up into smaller groups (Sharrow, 2006). Both of these traits complement shepherd management.
In conifer plantations Sharrow (2006) suggests a two pass grazing system when young trees are experiencing competition from both grasses and shrubs. The first pass targets the grass forage in the spring when green growth is most rampant. Spring bud burst on conifers typically coincides with grass and forb growth. At this time, sheep will favor grass and forbs over fresh conifer shoots. Sheep can be quickly moved through plantations as they deplete the available spring grass growth. A second and longer stay can occur after conifers and herbaceous vegetation have matured. This grazing period is meant to target shrubs and young hardwood trees that are generally more palatable to sheep than conifers. In this second phase, the remaining standing herbaceous vegetation can also be trampled.
Management intensive grazing is likely the best practice for smaller acreage savannas. While shepherding may be the best system for extensive oak savanna, this ecosystem rarely occurs in large enough swaths in the Willamette valley to abide this type of extensive management. Also, given the restoration emphasis of savanna grazing, a more targeted species by species approach to vegetation management will likely need to occur. In management intensive grazing, large fields are broken up into smaller paddocks in order to manage timing of grazing and animal impact. These paddocks may be enclosed by permanent or temporary electric fencing. Often, a field will be surrounded by permanent fence and cross-fenced with temporary electric fencing. The animals are moved progressively through the field in timed pulses determined by the stocking rate and the plant recovery period (Butterfield et al. 2006). Mob grazing, in which animals are densely stocked for a brief period in a targeted area, can be utilized as a tool for directing herd impacts like grazing, trampling, and manure deposition in an area where a specific plant or group of plants specifically targeted (Butterfield et al. 2006). Often, management intensive grazing emphasizes high stock densities with animals rotated at frequent intervals. This mimics impacts of grazing animals in native prairie systems with intact predator/prey relationships.
For best results, SVM should be carried out under managed grazing conditions in order to prevent overgrazing of choice species. The time it takes for a plant to regrow to its full energy potential following being bitten is called the recovery period. Overgrazing occurs when a plant is bitten a second time before it has had a chance to regain the store of energy it lost from the first bite (Ekarius, 1999). If a plant is severely overgrazed, often it will weaken and eventually die. Although overgrazing is often caused by too many animals it can also be caused by animals being allowed to selectively feed, even when overstocking is not a concern. The key to preventing overgrazing is to carefully time the grazing of plants so that they have an adequate recovery period (Heckman, 2007). In managed grazing systems, the animals are moved before they have a chance to graze the same plant a second time or begin to favor plants which are meant to be left uneaten. In management intensive grazing systems, the animals are kept out of an established paddock until the plant has gone through its recovery period. Timing is critical because of the need to maintain livestock weight gain while not compromising the regrowth potential of the forage (Heckman, 2007). Recovery periods vary by species and by season. During the highest flush time of the year, the recovery period may be as little as 10 days, during the drier periods of the year, a plant may take 90 days to recover. During the dormant period, it may take more than 180 days for a plant to recover (Ekarius, 1999).
4.5 Tree Protection
Compared to hardwoods, conifer species, because of their low palatability, are less prone to grazing impacts by sheep. Douglas fir, even when browsed to 50% of their vegetation, show little reduction in height or diameter growth over time (Osman et al, 1993). Bark stripping, which is more common with goats or horses than sheep or cattle, typically has a greater effect on productivity over time than defoliation. Bark stripping is more likely to occur on hardwoods than conifers, which have more pitchy bark (Sharrow, 2009). Young hardwood trees are more susceptible to bark stripping than mature trees because young trees lack the corky bark of many older hardwood species.
If hardwood seedling survival is a priority in areas accessible to livestock, plants will need to be protected within sturdy cages to protect them from being eaten or trampled. Mature trees can be harmed by soil compaction or root exposure by animals congregating and grazing under trees. Soil compaction is of greatest concern during the rainy season and on finely textured clay soils. Because of this, in many cases, oak woodlands should not be utilized as livestock overwintering areas. Livestock can take advantage of the microclimate and nutritional benefits of oaks without causing compaction damage if watering facilities, feeding areas, salt block locations, and trees are widely spaced (Vesely, 2004). This layout will encourage animals to spread out and circulate throughout the entire pasture.
4.6 Predator Protection
Livestock guardian dogs are likely the best protection strategy. Mountain breeds of sheep, which stay bunched together as one unit are also more adapted to back-county predator protection.
4.7 Oak Stand Improvement and Resprout Control
The first step in oak restoration typically involves removing competing conifers encroaching on the oak tree canopy. A study conducted in Western Washington by Devine and Harrington (2013) observed the effects of overstory Douglas-fir removal on residual white oak trees. The objective was to assess the 10-year effects of full and partial release treatments on oak height, diameter, and acorn production. It was found that full release treatments were consistently associated with the greatest annual diameter growth. Acorn production was significantly influenced by treatment and year. Although some years had very low production the productive years displayed the highest production in the full release as compared to the partial release treatment (Devine et al. 2013).
Prior to this study, little was known about the effects various intensities of conifer thinning have on the growth of released oak trees. Concern for shock or sun-scald on oaks subjected to a sudden full-release may be assuaged by these results and foresters can forgo the practice of gradual multi-year thinnings. The compatibility of a one-time stand improvement thinning with oak production contributes to greater returns on marketed trees and more rapid transitions to oak woodland or savanna conditions.
Often, both hardwoods and conifers are removed in oak restoration projects. In addition to Douglas-fir as the most common conifer, hardwood tree species often include Oregon ash (Fraxinus latifolia), big leaf maple (Acer macrophyllum), and other subdominant and suppressed oaks. Many hardwoods tend to resprout or coppice from their stumps after being cut. Conventional restoration practices typically include painting each stump with herbicide in order to fully kill the trees and prevent dense thickets of coppiced hardwoods to develop over time.
Hardwoods have been observed to be far more palatable to sheep than conifers (Lavendar et al, 1990). Utilizing sheep to keep the resprouts down may be the first vegetation management value a grazier gets out of a recently thinned oak forest. The picture below shows a photo of controlled re-sprouting with black locust (Robinia pseudoacacia) and blackberries (Rubus armeniacus) on my farm in the Willamette Valley. The sheep have been moved into the understory of a locust grove to prevent further coppicing from the base. The absence of further locust sprouts and blackberries on the left side of the fence after a full season of growth is a testament to the capacity of a small herd of sheep to manage these species.
Image 1: Icelandic Sheep Enclosure on Left Showing 100% Blackberry and Black Locust Resprout Control. Photo by Abel Kloster
4.8 Maintenance of Biodiversity
Utilizing the varied foraging habits of different species can lead to a time-based shift in the abundance of forbs, grasses, and woody vegetation as the animals graze their preferred forages. The benefits of mixed animal systems can be seen in the Dehesa systems where different breeds of animal are used to target different components of the landscape over time. It follows that such management can create a mosaic of plant assemblages at the landscape scale. For example, in the British uplands, mixed grazing systems consisting of sheep and cattle improved animal productivity and growth while supporting more bird and butterfly species than sheep or cattle alone. Cattle are less selective grazers than sheep and can be used to graze some undesired species more readily while increasing wildlife biodiversity (Fraser et al. 2014). The British Uplands, many of which have been maintained by grazing for generations, are internationally important for their unique plant and bird communities. Many of the habitats of the UK’s upland areas have been identified as priority habitats and special areas of conservation in the UK Biodiversity Action Plan (Fraser et al. 2014). The transition from mixed animal grazing systems to intensive sheep production in the uplands has been implicated in dramatic changes in plant assemblages and decreases in plant and animal diversity (Fuller et al. 1999) while low intensity, mixed grazing has been considered beneficial.
Other research corroborates the advantage of grazing’s usefulness in the maintenance of mosaics and biodiversity. In comparing grazing to mowing, Sykora et al. (1990) found that grazers create more micro-sites and openings for new seedling recruitment. Also because sheep prefer specific plants over others, they promote more heterogeneity across a field than a mowing implement (van den Bos and Bakker 1990). In a study conducted in the United Kingdom, researchers seeded five native forbs (Achillea millefolium, Plantago lanceolata, Centaurea nigra, Stachys officinalis, Prunella vulgaris) into a dense stand of grasses (Lolium perenne, Poa spp. and Agrostis spp.). Plots that utilized a mix of sheep grazing and mowing had higher establishment rates than plots that were mowed but not grazed (Jones and Hayes 1999).
For biodiversity management in oak savannas, seasonality of grazing is likely another important factor. Preventing sheep from grazing spring ephemerals, allowing wildflowers to set seed, and limiting herd impacts on soil during the wet season stand out as three timing objectives for herd management in oak savannas. To assist with this time based management, The Savory Institute provides a timed rotation tool called the “Aid Memoir” which assists in planning grazing’s impact on a species by species basis in intensively managed systems (Butterfield et al, 2006). One practice for enhancing biodiversity is to carry out seasonal seed collection for distribution to impacted zones. This practice, sometimes called “Huck and Shuck,” utilizes the mosaic of varying plant assemblages within a field as “mother gardens” for propagule dispersal.
5 Facilitative Interactions Under Sheep Vegetation Management
5.1 Nutrient Cycling
While facilitative interactions between trees and understory forages have been reported in both Douglas-fir plantations and oak savannas, the research is inconclusive. Carlson et al. (1994) studied the plant, soil, and water relations of a Douglas-fir plantation. In this study, a sheep grazed system was undersown with a nitrogen fixing clover (Trifolium subterraneum) and tall fescue (Festuca arundinacea) as a forage crop while a second set of plots was not sown and went ungrazed. It was found that nitrogen-fixing vegetation combined with grazing increases nitrogen uptake by trees. The safety net and nutrient pumping functions of trees are known to be a significant contributor to mineral cycling and subsoil harvesting of surplus nutrients. Moreno et al. (2007) attributed significant nutrient increases under Holm oaks (Quercus ilex) in the Spanish Dehesa to nutrient pumping functions. Given the modern tendency to fertilize conifer plantations, manure inputs from sheep grazing also represent an obvious nutrient cycling service and a secondary function of sheep in these systems.
5.2 Water Use
The majority of oak tree roots within the Dehesa were found to be located below the rooting depth of grasses. This resulted in complementary water and nutrient use during typical rain years but competitive interactions during the more dry years (Moreno et al. 2015). In Iowa, Seopbi et al. (2005) demonstrated that agroforestry buffers planted to Red Oak increased soil bulk density and water infiltration as compared to open fields. The increased infiltration rates may help to mitigate runoff and increase the residence time of water into the dry season within oak savannas. The results of a Douglas fir study near Corvallis, Oregon indicated that grazing of understory vegetation may reduce water stress of trees during dry periods by reducing transpirational water use by the forage plants (Carlson et al. 1994). Hedrick et al. (1966) also observed more abundant soil moisture at the 5 inch and 12 inch depths was found each summer on the grazed plots. This moisture availability was found to correlate well with the removal of palatable herbage by the sheep (Hedrick et al, 1966).
5.3 Soil Compaction
Soil compaction is a significant factor in the Willamette Valley where grazed landscapes experience excessively wet winters and often have heavy clay content. However, Sharrow (2007) observed that sheep grazing on clay soils, even at high stock density, do not interfere with either tree or pasture growth despite the soils displaying increased bulk density and lower rates of soil water infiltration. He hypothesized that the increase in soil micropores from compaction resulted in a greater overall presence of water available to the trees within the soil substrate.
5.4 Tree Growth
In a recent study in British Columbia, a 6 year grazing study found significant increases in internodal length amongst hybrid spruce (Picea glauca x Picea engelmannii) under SVM as compared to trees under no vegetation management. At the end of the study, trees in the grazed blocks were declared “free to grow” while the ungrazed blocks were not anticipated to achieve this status until at least 2019 (Serra et al. 2014). In the 1950s, research was conducted closer to home. In the foothills of the Willamette valley, yearling ewes were grazed for three to four weeks each spring within recently planted Douglas-fir plots. After three years of spring grazing, tree height became significantly greater on the grazed plots and continued to increase though four more years of grazing. The growth differential was lost after three years of non-grazing on the same plots (Hedrick, 1966). Based on the above evidence, it can be assumed that trees in these grazed pastures increased their growth because they have less competition, more retained soil moisture, and more rapid nutrient cycling.
5.5 Forage Growth
Douglas fir growing within the Willamette valley have been observed to have little effect on understory forage production until trees were 9 years old. However by the time the trees are 11 years old, yield had declined significantly to 54% of adjacent pasture (Sharrow, 1991). Competitive exclusion, a process discussed in section 4.2, has been shown to lead to greater tree growth in grazed seeded pastures over ungrazed controls. In one case, where the combination of grass competition and grazing reduced alder (Alnus rubra) establishment, Douglas fir basal area was found to be 50% greater in grazed units (Sharrow, 2006).
In the Dehesa system, increased forage production has been attributed to nutrient pumping, bulk density reductions, improved microclimate, and organic matter deposition under oak trees (Paulo et al. 2016). (See Image 2) Additionally, organic matter turnover due to herd impacts, has been seen to be extremely efficient in the Dehesa where animals can consume up to 85% of the deposited plant mass and effectively increase the organic matter turnover rate up to 24 times faster than in dense forest (Paulo et al. 2016).
Image 2: Forage Growth Under Oaks in the Spanish Dehesa
Conclusion
World food demand is expected to double by 2050 (Godfrey et al. 2010) with the intensification of agriculture having further detrimental effects on habitats and wildlife. Therefore, the goal of increasing productivity while regenerating ecosystems and biodiversity is imperative for the future of global agriculture. A priority for animal production globally is to identify and understand optimal livestock grazing systems that benefit both biodiversity and production (Fraser et al. 2014). In the Willamette Valley, animal integrations into conifer and oak forestry systems represent an opportunity to manage for both increased economic and ecological yields in these landscapes.
Many researchers agree that SVM can be an effective tool for vegetation control when applied during site preparation prior to planting or following planting of young conifer plantations. However, research across the span of ecological regions and tree species has produced mixed results. For instance, studies conducted in Oregon by Sharrow (2009) show that sheep can be effective in Douglas-fir plantations, while Vasiliauskas and Luke (2000) found sheep grazing to be an unsuitable tool in various conifer plantations in Ontario Canada. This disparity in outcomes both regionally and across species, highlights the need for further research at the regional level. However, given the positive results of bioregionally specific research from the Willamette Valley, it is arguable that Douglas-fir SVM is ready to be more actively applied and studied within this region in particular.
Although there is a relative abundance of research in tree, grass, and animal interaction in seedling Douglas-fir SVM, much work remains to be done. Studies comparing herbicide with SVM treatments that examine environmental parameters like soil bulk density, soil conductivity, organic matter content, enzymatic and microorganism activity, and long term vegetative productivity would be particularly revealing.
It is important to note that future adoptions of SVM or indeed any silvicultural practices in industrial forests will require a large leap by land managers: a leap from single crop (trees) production to complex multicrop (trees/forage/livestock) production. This not only requires an increase in management expertise but operational capacity. Other social barriers to adoption are also obvious. For instance, one study by Opio et al. (2001) indicated that the adoption of SVM was hindered by the lack of infrastructure to support sheep farms, the need for long term contracts to ensure forester confidence, and forester perceptions of the profitability of SVM. Comparisons of financial investments and returns between the two management modes would significantly advance the assessment of the viability of SVM.
Despite the grave decline in the range of Oak Savannas in the Willamette Valley, there is an outstanding lack of research associated with the practice of SVM applied to this forestry system. However, the viability of sheep in maintaining oak savannas is well established by real-world examples throughout the world. In the Willamette Valley, there is a need for research on socio-economic factors as well as tree, grass, animal interactions. Additionally, given the fact that one of the top goals for this habitat is the maintenance of the native plant assemblages associated with the oak savanna, much work remains to be done in the establishment of good protocols for animal grazing strategies that enhance rather than degrade biodiversity.
Much work yet remains to establish predictable ecological, economic, and social outcomes for these two practices in the Willamette Valley. Despite the work remaining to be done, existing evidence both within the Willamette Valley and within analogous agroforestry systems throughout the world suggests that SVM is a viable tool in both Douglas-fir and white oak forests in Western Oregon.
Literature Cited
“About Icelandic Sheep.” Icelandic Sheep Breeders of North America. ISBONA, 2016. Web.
Adams, S N. “Sheep Grazing in a Young Sitka Spruce Plantation: A Review.” Journal of Applied Ecology 13 (1976): 507–511. Print.
Anderson, E.W. Borman M.M. Krueger W.C. 1998. Ecological Provinces of Oregon: A treatise on the basic ecological geography of the state. Oregon Agriculture Experiment Station. SR990
Apostol, Dean, and Marcia Sinclair. Restoring the Pacific Northwest: The Art and Science of Ecological Restoration in Cascadia. Washington, DC: Island Press, 2006. Print.
Bell, Dan. “Conservation in the Willamette Valley.” The Nature Conservancy of Oregon. N.p., Oct. 2011. Web.
Boyd, Robert. Indians, fire, and the land in the Pacific Northwest. Oregon State University Press, 1999.
Butterfield, Jody, Sam Bingham, and Allan Savory. Holistic management handbook: healthy land, healthy profits. Island Press, 2006.
Carlson, D H et al. “Plant-Soil-Water Relations in Forestry and Silvopastoral Systems in Oregon.” Agroforestry Systems 25.1 (1994): 1–12. Print.
Devine, Warren D., and Constance A. Harrington. “Restoration release of overtopped Oregon white oak increases 10-year growth and acorn production.” Forest Ecology and Management 291 (2013): 87-95.
Ekarius, Carol . Small Scale Livestock Farming. Massachusetts: Storey Books, 1999. Print.
Fuller RJ, Gough SJ (1999) Changes in Sheep Numbers in Britain: Implications for Bird Populations. Biological Conservation 91: 73-89.
Fraser MD, Moorby JM, Vale JE, Evans DM (2014) Mixed Grazing Systems Promote both Upland Biodiversity and Livestock Production. Plos ONE 9(2): e89054
Godfray HCJ, Beddington JR, Crute IR, Haddad L, Lawrence D, et al. (2010) Food Security: The Challenge of Feeding 9 Billion People. Science 327: 812-818.
Habek, J.R. 1961. The original vegetation of the mid-Willamette Valley, Oregon. Northwest Science 35(2):65-77.
Heckman, Ed. “Management-intensive grazing in Indiana.” (2007).
Hedrick, D W, and R F Keniston. “Grazing and Douglas-Fir Growth in the Oregon White-Oak Forest Type.” Journal of Forestry 64.11 (1966): 735–738. Print.
Jones, A. T., and M. J. Hayes. “Increasing floristic diversity in grassland: the effects of management regime and provenance on species introduction.” Biological Conservation 87.3 (1999): 381-390.
Louise, E., et al. “The development of agroforestry as an integrated land use management strategy.” North American Agroforestry: An Integrated Science and Practice 2nd edition northamericanag (2009): 1-24.
Lavendar, D.P., R. Parish, C.M. Johnson, G. Montgomery, A. Vyse, R.A. Willis and D. Winston. 1990. Regenerating British Columbia’s Forests. University of British Columbia Press, Vancouver. 372 pp.
Moreno, G, and Y. Caceres, Y “System Report: Iberian Dehesas, Spain.” Agforward (2015). Web.
Moreno, Gerardo, and Jose J Obrador. “Effects of Trees and Understory Management on Soil Fertility and Nutritional Statusof Holm Oaks in Spanish Dehesas.” Nutr Cycl Agroecosyst 78 (2007): 253–264. Print.
Olea, L, and A San Miguel-Ayanz. “The Spanish Dehesa. A Traditional Mediterranean Silvopastoral System Linking Production and Nature Conservation.” 21st General Meeting of the European Grassland Federation . Badajoz (Spain): N.p., 2006. Print.
Opio, C., N. Jacob, and D. Khasa. “Factors affecting a sheep vegetation management system in British Columbia, Canada.” Agroforestry systems 53.3 (2001): 305-312.
Paul, Virginia. This Was Sheep Farming . Seattle, Washington: Superior Publishing Company, 1976. Print.
Paulo J A, J Crous-Duran, P N Firmino, S P Faias, and J HN Palma. “System Report: Cork Oak Silvopastoral Systems in Portugal.” Agforward (2016). Web.
Phelps, N L. “Sheep Grazing on Certain Clearcut Harvest Units of the Beckler River Drainage of Western Washington.” M.S. Thesis (1978): 58 pp. Print.
Seopbi, Tshepiso et al. “Influence of Grass and Agroforestry Buffer Strips on Soil Hydraulc Properties for an Albaqualf.” American Journal of Soil Science and Society 69 (2005): 893–901. Print.
Sharrow, Steven. “Applying targeted grazing to coniferous forest management in Western North America.” Targeted Grazing: A Natural Approach to Vegetation Management and Landscape Enhancement. Centennial (CO): American Sheep Industry Association. p (2006): 90-99.
Sharrow, S H, D Brauer, and T R Clason. “Silvopastoral Practices.” North American Agroforestry: An Integrated Science and Practice. Madison, MI: American Soiety of Agronomy, 2009. 105–130. Print.
Sharrow, S H, W C Leininger, and K A Osman. “Sheep Grazing Effects on Coastal Douglas-Fir Forest Growth: A Ten Year Perspective.” Forest Ecology and Management 50 (1992): 75–84. Print.
Sharrow, Steven H. “Sheep as a Silvicultural Management Tool in Temperate Conifer Forest.” Sheep Research Journal Special Issue (1994): 97–104. Print.
Sýkora, Karel V., Geertje van der Krogt, and Jos Rademakers. “Vegetation change on embankments in the south-western part of The Netherlands under the influence of different management practices (in particular sheep grazing).” Biological Conservation 52.1 (1990): 49-81.
Towle, J.C. Woodland in the Willamette Valley: An historical geography. Ph.D. thesis, 1974, University of Oregon, Eugene.
Van den Bos, J., and J. P. Bakker. “The development of vegetation patterns by cattle grazing at low stocking density in the Netherlands.” Biological Conservation 51.4 (1990): 263-272.
Vasiliauskas, Stan, and Alison B. Luke. Testing sheep grazing as a vegetation management tool in Northeastern Ontario. No. TR-041. Northeast Science & Technology, (2000)
Vesely, D. Tucker, G. A Landowners Guide for Restoring and Managing Oregon White Oak Habitats. Pacific Wildlife Research. (2004)
Weaver, Sue Ann. “Backyard Sheep.” Storey Publishing. (2013)
Tables
8.1 Sheep Breed Functions and Uses
Source: The Backyard Sheep by Sue Weaver
BREED DAIRY MEAT WOOL BRUSH CONTROL American Black Belly X Barbados Blackbelly X California Red X X Charollais X X Cheviot X X Corriedale X X Dorper X Dorset X X X Florida Cracker X Friesian X X X Gulf Coast Native X X X Hampshire X X Hog Island X X X Icelandic X X X X Ile de France X X Katahdin X X X Lacaune X Leicester Navajo-Churro X X Rambouillet X Romney X X Royal White X X Santa Cruz X Shetland X Shropshire X X Southdown X X St Croix X X Suffolk X X Texel X X Wiltshire Horn X