Agroforestry research scientists work hard at generating, controlled, replicated "scientific" observations. The day-to-day operation of agroforestry systems used in research also produces practitioner experiences which, although lacking the control of planned experiments, nevertheless provide useful information. I would like to share some of the "scientific" and "practitioner" experiences which I have acquired during 16 years of grazing sheep in conifer/pasture silvopastures.
Agroforest system processes are primarily managed by manipulating vegetation composition and structure. Spatial pattern of trees is an important structural element in silvopastures. Trees are often grouped in clusters or rows in order to facilitate agricultural operations such as spraying, mowing, fertilizer application, and fencing. Although few replicated studies have been published, sufficient information is available to provide some principles for silvopasture design: Pasture production is greater when trees are aggregated together, leaving wide pasture interspaces free of trees, than when trees are spaced equally apart in grids. The influence of tree pattern on pasture production increases as number of trees per acre and individual tree size increases. Tree growth is little affected by pattern, provided that each tree has at least one side exposed to full sun. This means that single rows of trees and double rows of trees should grow at similar rates, while the inner row of triple row plantings will eventually suffer reduced growth from competition.
Surprisingly little has been written about the influence of silvopasture tree pattern on the livestock and native animal components of silvopastures. We have all seen animals seek out shade under trees during hot weather and shelter beneath trees during cold rainy weather. A grove or row of trees provides better protection from the elements than do single trees. However, animals also react to tree pattern in more subtle ways. Many wild animals prefer a habitat which contains protective vegetation where they can shelter from the weather and can hide from predators together with more open feeding areas where preferred food can be found. These conditions are most commonly found at the edge where two plant communities meet. Cluster or row silvopastures have a large amount of forest/pasture edge. This is very attractive to deer, rabbits, and birds which shelter in/under trees and feed in the pasture. In one of our agroforests, deer began to commute in from adjacent oak woodlands to feed in our silvopasture soon after its establishment. Deer use increased as trees grew until they are now living entirely within the silvopasture. Deer damage to young trees has progressed from being primarily browsing damage of newly planted trees to being mechanical damage from rubbing the velvet off their horns in the fall. Because deer appear to select trees between 3 and 6 ft tall for rubbing posts, we expect this damage to end as trees grow.
Planting trees in rows and spraying strips to reduce competition from pasture can expose trees to browsing damage from deer and rabbits. Browsing animals will sometimes follow the rows, eating trees as they go. one year, I debated whether to spray the grass away from one-year-old Douglas-fir trees. I was concerned that the grass, although reducing tree growth, might be concealing the rows of trees from deer. Sure enough, once I sprayed the grass, deer easily found the trees and browsed every one. I would have been better off to accept the grass competition.
Birds such as flickers, robins, and thrushes use our conifer trees as perches. From the tops of tree rows, they can easily survey the pasture alleys for insects, worms, and other food items. Unfortunately, heavy bodied birds frequently break the terminal leaders off trees when attempting to perch on leaders which have not yet become woody. Birds faeces deposited near trees may contain weed seeds. Newly established poison oak under several silvopasture trees most likely arrived by this mechanism.
Livestock are generally easier to herd in row silvopasture than in grid or cluster plantings. Their natural tendency is to drift parallel to barriers such as tree rows. Given a choice, animals will usually head downwind. This makes herding livestock between rows relatively easy. However, they may be reluctant to cross between rows. This is especially true for animals who are used to electric fencing. Modern high voltage fence chargers deliver 5000-7000 volts of charge at high amperage for a very short duration. once animals have experience with electric fences, they are not anxious to repeat it. When in doubt, they often perceive any hard edge as a potential fence line. New Zealand sheep farmers, for instance, have successfully contained livestock by clipping the edge along strips of tall grass to produce a "grass fence". I used to keep our dog out of the vegetable garden by putting up a single piece of string, which he recognized as an electric fence wire. Low elevation remote sensing of our silvopastures shows slightly reduced pasture production in the center of alleys between tree rows. We believe that this results from trampling of livestock who are trying to stay away from what they perceive as a "tree fence". We have attempted to reduce this effect by encouraging sheep to browse the lower tree limbs so that they can easily see under them. A related off shoot of this principle is that you do not want trees planted near gates. I have spent a lot of frustrating time trying to get sheep to move through an open gate. They see the tall grass and trees as a fence line and will not pass through.
Livestock can be extremely difficult to herd in grid or cluster silvopastures. Animals attempt to maintain a personal space between them and a potential predator such as a herder. As you approach them, they will walk away to preserve this space. once this space is reestablished, they will often stop and look around to see that is happening. once livestock pass behind a tree, they lose sight of you. They assume that you can't see them, so lose all interest in being herded. When you suddenly appear around the tree, they are spooked and attempt to flee. This situation becomes really interesting when you are attempting to control a large group of livestock, some of whom can see you, and some who can't.
Tree rows on steep slopes are often planted on the topographic contours. While this is ideal for soil conservation, it may impede livestock movement. Livestock and native ungulates, just like people know better than to walk straight up or down hills. They walk at a slight angle to the hill's contour. These use patterns are usually easily seen in big game or livestock trails. Planting trees along these travel contours should capture most of the soil conservation benefits without impeding animal movement.
In summary, my experience is that silvopastures planted in rows are far superior for livestock production than are either grid or cluster plantings. Trees planted in rows with wide open spaces for pasture production between them, support high forage production and facilitate agricultural operations and animal herding. The large amount of edge created and maintained long into the timber rotation tends to maintain high biodiversity. Deer, rabbits, and perching birds find the structure of cluster and row agroforests appealing. Therefore, one should be prepared for the potential impacts upon trees and pastures which may accompany increased use by wildlife. Double rows should allow for fewer but wider pasture strips between rows without reducing tree production. If planting a new silvopasture today, I would plant trees in offset double rows as shown in Figure 1 below.
Figure 1. Design for a Douglas-fir/grass-clover silvopasture for western Oregon hill lands. It will qualify as a fully stocked forest for regulatory and forestry cost sharing purposes.
Dr. Sharrow is Professor or Range Management and Agroforestry, Department of Rangeland Resources, Oregon State University, Corvallis, OR. 97331
* This article originally appeared in the July 1998 issue of the Temperate Agroforester.