An e-publication by the World Agroforestry Centre



Chapter 25
Agroforestry in the temperate zone

25.3. Current temperate - zone agroforestry systems

Today, many temperate-zone agroforestry strategies represent extensions of these historical practices, with management techniques modified through research and experience. However, new practices are also developing as landowners in industrialized countries turn to agroforestry as an opportunity to counter problems in both agriculture and forestry. Major food production problems currently include the increasing costs of fossil fuel, farm surpluses, and soil erosion. These problems can be mitigated through less energy-intensive and Hanover, 1987). In the forestry sector, problems such as high plantation-establishment costs, delayed economic returns, and fire risk can be offset by regular revenues from interplanted crops and grazing in the early years of a forest stand.

Since the late 1980s, several compilations covering current temperate-zone agroforestry have become available. The two North American Agroforestry Conferences in 1989 (Williams, 1991) and 1991 (Garrett, 1991), the 1989 International Agroforestry Symposium in Pullman, Washington (Budd et al., 1990) and the reviews by Gold and Hanover (1987), Byington (1990), and Bandolin and Fisher (1991) provide a considerable body of information on agroforestry in North America, and to a limited extent in other temperate countries. Similarly, several reports on agroforestry in Europe are available in the proceedings of the 1989 International Conference on Agroforestry in Edinburgh, UK. (Jarvis, 1991). Agroforestry systems in Australia and New Zealand have also been described in various publications (Anderson et al., 1988; Knowles, 1991). Descriptions of the old agroforestry systems and new developments in the field in China are given by Zou and Sanford (1990) and Zhaohua et al. (1991b). Additionally, numerous reports on agroforestry systems in various parts of the temperate region are available in the recent literature: e.g., Carruthers (1990 - European Community), Newman et al. (1991 - U.K.), Joffre et al. (1988 - Spain), Ovalle et al. (1990 - Chile), Ormazabal (1991 - Chile), Dadhwal etal. (1989 - Himalayan India), and Toky et al. (1989 - Himalayan India). Thus, the management potentials and practices, and structural composition of many, if not most, temperate-zone systems have been well described. Therefore, this chapter will only present a summary analysis of these systems, with emphasis on their potential use and benefits. For this purpose, these systems are grouped and discussed under the following headings: intercropping, silvopastoral, and windbreak systems.

25.3.1. Intercropping under hardwood species

Two major types of hardwood intercropping systems can be differentiated: those with fruit- and nut-producing trees, and those with high value timber species such as poplar. Generally, multicropping offsets plantation establishment costs, allows for more intensive use of both forest and agricultural land (especially close to processing facilities), and reduces cultivation costs of individual crops since cultural operations can be allocated jointly to all crops (Gold and Hanover, 1987).

Perhaps the first, and still one of the best, expositions of the concept of agrisilvicultural systems with fruit and nut trees in North America is advanced by J. Russell Smith in his classic book, Tree Crops: A Permanent Agriculture (Smith, 1950). Based on his travel experience and observations of Mediterranean agriculture, Smith advocated, as early as 1914, North American agricultural systems using nut trees (such as Carya spp. and Juglans spp.), oaks, persimmons (Diospyros spp.), and honeylocust (Gleditsia triacanthos).

Following the Great Depression of the 1930s, work on tree crops commenced especially in the eastern U.S. under the auspices of the Tennessee Valley Authority (TVA), concentrating on black walnut (Juglans nigra), Chinese chestnut (Castanea mollisima), filbert (Corylus spp.), hickories, persimmon, and honeylocust. Unfortunately, the tree crops idea was all but forgotten in the 1950s and 1960s during the post-war economic boom. However, the 1970s saw a renewed interest in tree crops because of the energy crisis, mounting concerns about the high rate of agrochemical and energy use in industrialized agriculture, realization of the adverse effects of soil erosion in row-crop agriculture, and awareness regarding the potential role of trees as an effective component in the overall solution to these problems (Gold and Hanover, 1987).

In commercial fruit orchards, fruit trees are usually widely spaced; orchard practices in the last 100 years often excluded agricultural activities other than forage production for limited grazing within the orchards. The intensive management necessary for fruit production generally concentrates on control of vegetation on the orchard floor, seriously limiting cultural practices for other crops. However, concerns about ecological and economic sustainability are leading many landowners to crop diversification and intercropping within those orchards, with innovations often developed by individual farmers. For example, leeks, corn, and strawberries are grown in peach orchards in Ontario, Canada; oats are grown in some New York apple orchards; and potatoes, grains, soybeans, squash, and peaches have been planted in pecan (Carya illinoensis) orchards in the southern United States (Williams and Gordon, 1991). Approximately 10% of all fruit and nut orchards in Washington State (USA) are intercropped with vegetables for home use, and in another 25% of the orchards cattle or sheep are grazed during part of the year (Lawrence et al., 1992). Intercropping in fruit orchards provides a substantial agroforestry opportunity for documentation, research, extension, and expansion as well as further farmer innovation.

In most fruit orchards, cultivation of vegetable and other crops during the establishment phase reduces the need for vegetation management such as mowing and herbicide application. Fertilizers applied to the orchard trees or vegetables are available to the other crops. Produce from the orchards may be used for home consumption or market sale. As the trees develop and shade the orchard floor, annual crops can be replaced by forage species; at that time the orchard can be opened to grazing as the trees would be large enough to escape damage by animals.

Combinations of fruit trees and other species are traditional practices in regions such as the mid-elevation Himalaya mountains in the Indian subcontinent. For example, in India, citrus is intercropped with winter vegetables and gram (Cicer arietinum) for 2-3 years, and beans and peas are often grown in dwarf-apple orchards for 5-6 years, or beneath apricots, peach, plum, and nectarine for 2-3 years (Tejwani, 1987). Vegetable production is eventually reduced as the fruit trees mature and shade the orchard floor. The main tree species used in these Himalayan agroforestry systems are listed in Table 25.1 (Dadhwal et al., 1989; Toky et al., 1989).

Table 25.1. Common tree species in agroforestry systems in western Himalayas, India.

Source: Adapted from Dadhwal et al. (1989 and Toky et al. (1989).

Far more prevalent than combinations of fruit trees and crops in North America, is multicropping and grazing in black walnut, pecan, and other nut orchards. A significant difference between nut and fruit production is that nut trees may attain larger sizes than the fruit trees and, therefore, require wider spacings with longer intercropping periods. Walnut and pecan trees also represent a major revenue source and are generally harvested for high value wood products after 50 to 80 years.

Black walnut orchards have been more widely studied than any other set of orchard agroforestry practices in North America (Campbell et al., 1991; Garrett and Kurtz, 1983; Garrett et al., 1989; Kurtz et al., 1984; Newman et al., 1991; Noweg and Kurtz, 1987). A current multicropping strategy with black walnut in Missouri employs an initial tree spacing of about 3 by 12 meters (10 by 40 feet), intercropping with combinations of wheat, milo (Sorghum bicolor), and soybeans during the first 10 to 12 years (see Figure 25.1), followed by 10 or more years of cool season forage and limited cattle grazing within the plantation. Walnut production begins 10 to 15 years after planting, but does not reach peak levels until ages 25 to 30 years. Nuts are collected until the trees are harvested between ages 60 and 80 years, depending on site conditions. The nuts are valued both for the meat and the ground shells, which are used as abrasives. Trees are usually pruned at least twice during the first 20 years to promote a clean bole and high quality lumber and veneer products, and one or more thinnings may be used to maintain crown structure for nut production and as an intermediate source of revenue.


Various economic analyses of strategies similar to this have demonstrated that if black walnut were grown solely as a timber crop, it is not likely that the final wood product revenue would justify the expenses for establishing and growing the plantation. However, the addition of nut production generally provides a positive benefit/cost ratio, and intercropping during the early plantation years provides the necessary revenue to make the total program very profitable on many sites (Kurtz et al., 1984). In an economic study of different management regimes of black walnut plantations, Kurtz et al. (1984) found that the internal rate of return (IRR) and present net worth (PNW) were greatest for intensive, multi-crop regimes (Table 25.2). In other words, the more intensive the multiple cropping scheme, the greater the financial return. Even without grazing and nut production, intercropped black walnut (and other hardwoods) is much more attractive financially than wood production alone, assuming moderate to high tree growth rates and reasonable interest rates (Campbell et al., 1991). (For a more detailed discussion of the economics of agroforestry see Chapter 22 and, specifically, Table 22.6 (pp. 399-400) for more information on the black walnut system.) Landowner objectives and site characteristics are also important factors in the success of specific agroforestry schemes. For example, Byington (1990) pointed out that although there is a significant profit margin with multiple crops and black walnut, the high costs and long rotations of such commercial practices may limit their adoption to large farms.

Table 25.2. Present net worth (PNW) and internal rate of return (IRR) of alternative black-walnut management regimes, site index 65 and 80, rotation age 60 years; Missouri, U.S.A.

Note: Site index is an index of site quality for a given tree species indicated by the height (m) of trees at age 50.
Source: Kurtz et al. (1984).

Another, less intensive version of intercropping with nut trees is the low-input management system used with pecans in Kansas, Oklahoma, and Missouri. Natural bottomland forests are cleared of all trees and vegetation except healthy, well-formed pecan trees. A permanent ground cover is established, with regular applications of nitrogen fertilizer (unless a leguminous ground cover is planted), and controlled cattle grazing is used to regulate the groundcover. In addition to revenue from pecans, periodic thinnings provide high value wood products (Reid, 1991). Major limitations of this system include insect infestations, especially in nut crops, variable market prices for the nuts, and increasing growing costs. As with many potential agroforestry systems, a critical factor for the success of the system is the existence of viable markets for the various crop and wood products.

One of the most widely intercropped group of trees is the poplar species (Populus spp.) and their hybrids; these species were traditionally planted for short rotation fiber and fuel production. Poplar plantations in Europe and eastern Canada have been interplanted with corn, potatoes, soybeans, and other cereal and tuber crops, in different temporal sequences, for the first three to six years after tree establishment (Gold and Hanover, 1987). Many of the poplar plantations are only grown for an additional five to ten years after crop harvest before harvesting and establishment of the next rotation. In China, sesame, soybeans, peanuts, cotton, indigo, and various vegetable crops are grown in both hybrid poplar (Figure 25.2) and Paulownia tomentosa plantations (Figure 25.3); the poplars are widely planted in a variety of other crop-border configurations (Farmer, 1992). In Australia, various melon and squash crops are grown for two years, followed by permanent pasture, with cattle grazing on both the pasture and branches lopped from the poplars. Poplar is also frequently planted on plot boundaries of wheat and barley fields in northern India and Pakistan.

Despite the apparent attractiveness of such systems, their success depends on a variety of factors which may or may not be related to increased biological or economic yield. For example, an economic study in northern Italy demonstrated that intercropping provides greater returns than poplar monoculture under all site conditions, as well as greater returns than soybean monocrops when poplar growth rates are high (Carruthers, 1990). This multiple crop system, however, is on the decline for several reasons. Low wood prices and marketing difficulties reduce the potential revenue from timber harvests, regular spraying for persistent poplar diseases has damaged annual crops, and increasing ownership of poplar plantations by part-time farmers or absentee landowners prevents the cultivation necessary for annual crops. Such problems underscore the need to understand owner objectives, infrastructure needs, and joint cultural problems before widespread adoption of agroforestry systems can occur.


25_3_current_temperate_5 25_3_current_temperate_6

25.3.2. Livestock grazing in managed plantations (silvopastoral systems)

The practice of grazing livestock in plantations, especially conifer plantations, has probably been more widely utilized and reviewed than any other agroforestry system in the temperate zone. The approach varies from the relatively simple management system in which livestock are allowed to graze freely in plantations established essentially for timber production, to situations in which trees and pastures are purposely managed to accommodate a long period of carefully controlled livestock production. Although the system occurs in many developed countries, it is most common in North America, Australia, and New Zealand.

In the United States, examples of free grazing in plantations include cattle grazing in industrial pine plantations in the southeast, and sheep grazing in Douglas fir (Pseudotsuga menziesii) and ponderosa pine (Pinus ponderosa) forests in the northwest. In both regions, the primary forage species are natural grasses, herbs, and shrubs. The livestock are generally, but not necessarily, excluded from the plantations during the early years of tree establishment because of possible damage to seedlings. However, even in these early years, livestock may be allowed to graze during seasons when the nonconifer vegetation is more palatable than the seedlings. As the seedlings grow above the height of livestock, the practice becomes more common and less restrictive in terms of animal management. In many plantations, the animals are used as a method of biological control for vegetation that would normally compete with seedlings. Grazing for vegetation management will undoubtedly increase in the future, especially on public lands, as the use of herbicides and fire are restricted due to environmental concerns. Similar systems of livestock grazing management are also common during the summer in the forested mountains of western Canada. Livestock are moved to lower elevations in the winter. In some of these systems, native forages have been improved by prescribed burning, fertilization, or seeding of grass and legumes (Byington, 1990).

The vast majority of research on silvopastoral systems in North America has focused on pine forest with deliberate management of both pasture and trees. These systems are most important in the Southern Coastal Plain under slash pine (Pinus elliottii), and longleaf pine (Pinus palustris); they are popularly known as "pine-and-pasture" or "cattle-under-pine" systems. The earliest studies on pasture improvement in these systems, initiated in the 1940s, indicated that mechanical site preparation and fertilization were essential for forage establishment, and that production of established pasture declined with increasing tree-canopy closure (Lewis and Pearson, 1987). Among the most productive pasture species were Pensacola bahiagrass (Paspalum notatum), annual lespedeza (Lespedeza striata), and white clover (Trifolium repens), with Pensacola bahiagrass being the most shade tolerant.

In the 1950s, a study introducing cattle into pine/pasture mixtures was initiated to compare tree growth with differences in tree spacing, grass species, and fertilization (Lewis and Pearson, 1987). Slash pine seedlings were planted at 3.7 x 3.7 m and 6.1 x 6.1 m spacing, and allowed three years of establishment growth before introduction of Pensacola bahiagrass, Coastal bermuda grass (Cynodon dactylon), or dallisgrass (Paspalum dilatatum). Control plots of uncultivated, unfertilized pine/grass mixtures, in addition to native pastures, were also maintained. Cattle were introduced in the fifth year for annual grazing. The twenty-year results showed that the trees were larger in the fertilized plots; the wider spacing (6.1 x 6.1 m) increased tree diameter and cattle weight gains, but not wood yields; bahiagrass again proved to be the most shade tolerant and high-yielding forage species.

Various tree densities and planting arrangements were also tested as a part of this project. The standard tree-density and arrangement is approximately 1110 trees ha"1 at 2.4 x 3.7 m spacing. For silvopastoral management, the best arrangement was shown to be a double-row configuration of (1.2 x 2.4) x 12.2 m(or(4 x 8) x 40 feet) in terms of both forage production and wood production at mid rotation (Table 25.3, from Lewis and Pearson, 1987). Based on subsequent monitoring of these plots, Sequeira and Gholz (1991) reported that although light penetration and soil temperature were higher in the double-row stands, crown development and stem volumes of trees up to age 18 were superior in single-row stands. The authors suggested that there was great potential for optimizing both tree growth and understory microclimate by joint manipulation of crown structure and stand configuration in silvopastoral systems.

Table 25.3. Average survival, height, diameter at breast-height, basal area, and forage yields at age 13 of slash pine planted in single-row and double-row configurations at 1111 trees/ha, Withlacoochee State Forest, central Florida, USA.

* Means in a column marked with an asterisk are significantly different from the control
treatment (2.4 x 3.7) at the 0.05 level.
Source: Lewis and Pearson (1987).

Injury to and/or mortality of pine seedlings, poor quality of forage, and production of low-quality timber are the major constraints of this system. Delayed introduction of cattle, coupled with controlled stocking rates and improved forage grasses and legumes, are suggested as solutions to the first two problems. Pearson's (1983) analysis of twenty-year research data showed that multiple-use benefits of tree and cattle production and increased flexibility in land management could provide sufficient economic gain to offset the timber-quality problem. Based on sensitivity analyses using various discount rates, Dangerfield and Harwell (1990) also reported that a multiple land-use practice combining trees and grazing in the southeastern United States provided a favorable cash flow to the land user, and mitigated the negative cash-flow periods associated with conventional forestry production.

In general, grazing in plantations with normal spacing for timber production becomes less feasible as trees begin to shade out forage vegetation 5 to 15 years after establishment. Forage production and grazing periods can, however, be extended by either substantially increasing tree spacing and/or altering planting configurations. Although the technical feasibility of altering planting configuration to sustain forage production without reducing timber yield has been adequately demonstrated, the practice has not been widely implemented. The prevailing attitudes of traditional user groups could be one of the major factors that hinder the large-scale adoption of the practice. For example, the manipulation of forest structure for grazing may be viewed as an unnecessary forest management practice by many foresters, landowners, and other natural resource managers who have focused primarily on timber production. They may argue that livestock damage young pine plantations and that livestock managers are not willing to pay an adequate fee for the forage resources. On the other hand, traditional livestock producers contend that grazing provides indirect benefits to timber production on forest land, but are often unwilling to place trees on their pastures. Expanded implementation will probably occur only as private landowners see others purposefully combining pasture, cattle, and timber production (and gaining economic benefits from the system).

In New Zealand, interest in combining pasture and timber production increased in the late 1960s as all suitable land was gradually placed in either agriculture or forestry use (Percival and Knowles, 1983). A drought in 1968 also clarified the role of agroforestry, as farmers sought grazing opportunities in forests, and forest managers realized that grazing livestock would improve access for silvicultural work, reduce fire risk, and provide revenue (Knowles and Cutler, 1980). The interest in continuing this approach was strengthened by the trend towards wider initial spacing, and early pruning and thinning in radiata pine plantations. Considerable research has been done on various aspects of this management system. For example, information has been generated on optimum planting density of trees (to facilitate maximum fodder production without reducing wood yield), weed control measures, evaluation of fodder trees in different management systems, and the use of secondary products such as stems, seeds, and fruit from these trees as potential supplements to traditional forage species (Byington, 1990). Three distinct and viable silvopastoral types have been developed: forest grazing, timberbelts, and trees on pasture. Radiata pine has proved to be the pre-eminent species for profitable agroforestry (Knowles, 1991). Similar efforts with respect to grazing trials have also been conducted in Australia with plantations of eucalyptus (Cook and Grimes, 1977) and radiata pine (Anderson and Batini, 1979; Anderson et al., 1988) (Figure 25.4).

Relatively few studies have been done on grazing under hardwood species, especially in recent years, and most of the earlier research focused on conditions similar to farm woodlots. That research generally demonstrated that hardwood grazing in small areas is not worthwhile from an economic perspective (Byington, 1990). However, there is probably substantial opportunity for grazing in intensively managed hardwood plantations once research has clarified appropriate management strategies.

25.3.3. Windbreaks

As in the tropics, wind erosion is a serious problem in many parts of the temperate zone; the use of windbreaks to protect agricultural fields and homesteads is a common agroforestry practice in those areas (see section 18.5). The greatest benefits from the use of windbreaks occur in areas with winter snow and hot, dry, windy summers as in the Great Plains of the midwestern United States, in Russia, and in China. (Byington, 1990). The Green Great Wall program of China, launched in 1978, is perhaps the longest agroforestry windbreak/shelterbelt project in the world. Its objectives include rehabilitation of wasteland, development of vegetation for the control of sandstorms, and control of soil and water erosion through large-scale afforestation and grassland development. During the first phase (1978-1985), 6.7 million ha of farmland and 3.4 million ha of pastures have been protected through farmland shelterbelts, dune-fixing forests, and other tree-planting activities (Zhaohua et al., 1991b).

The benefits from windbreaks in the temperate zone are similar to those in the tropics. Under normal arid conditions on the U. S. Great Plains, windbreaks modify the microclimate of the protected zone by decreasing wind velocity. Consequently, vertical transport of heat is reduced and humidity is increased behind a windbreak, which generally reduces evapotranspiration. Furthermore, during periods of water stress, stomatal resistances are lower in crops protected by windbreaks than in crops grown in the open. Lower stomatal resistance tends to result in increased photosynthetic rates in the protected area. Air temperatures within the protected zone are generally warmer during the day and cooler at night than in unprotected zones. During the summer, the warmer day temperatures may increase evaporation from plants, but during early spring they may be beneficial for the establishment of most crops (Jensen, 1983). Another microclimatic influence of the windbreaks is the conservation of, or increase in, soil moisture due to more evenly distributed snow and, thus, snowmelt in the spring. These beneficial effects can result in increased crop production in areas protected by windbreaks.

Windbreaks are also likely to have positive impacts on livestock production, although quantitative data to support this conclusion are lacking. This is mainly due to livestock protection from hot winds and dust during summer, and cold winds during winter. Lower wind velocities reduces the effect of wind chill in cold weather and the amount of energy animals need to maintain body temperatures. This, in turn, can reduce feed costs and improve animal production.

The extent to which benefits from windbreaks are realized depends on a number of management and site-related factors. Length, width, shape, and positioning in relation to wind all effect windbreak efficiency. In general, narrow windbreaks composed of three to four rows of trees planted at moderate density, and positioned at an angle as close as possible to 90 to the predominant wind direction are the most efficient. In areas where wind direction changes frequently, it is common to plant windbreaks perpendicular to one another.

The distance between windbreaks is another major factor to be considered in windbreak design. If the height of the windbreak is H, generally, its protective influence extends to areas of up to 20 H distance. Multiple factors, such as soil characteristics, response of crops to protection, and the area of cropland that is lost to windbreaks, can affect the spacing between windbreak lines. On fairly stable soils and for moderately responsive crops such as cereals, the commonly-adopted distance between windbreaks is 15-25 H (Byington, 1990). For forage crops, spacings of 10-14 H may be justified if the additional yield is sufficient to balance the losses from reduced crop production area. The spacing could be profitably decreased even further in highly erosive soils.

Windbreak efficiency also is affected by the type of trees and shrubs planted. Species that can survive and grow in difficult and diverse conditions, while providing needed structure and protection are preferred. Dense crowns, stout boles, retention of lower limbs, and uniform rates of growth are all characteristics conducive to creating effective windbreaks (Byington, 1990). Fast-growing species are desirable for quick establishment and height increment. While some broadleaved species grow faster than conifers, they are usually deciduous; in contrast, conifers are long-lived and, since they retain their foliage, maintain the same density year round. Often, for best results, both conifers and broad-leaf species are grown together in windbreaks. The most commonly used windbreak species in North America include silver maple (Acer saccharinum), saltbush (Atriplex canescens), hackberries (Celtis spp.), Russian olives (Elaeagnus spp.), ash (Fraxinus spp.), honey locust, black walnut, juniper (Juniperus spp.), spruce (Picea spp.), pines, sycamore (Platanus occidentalis), poplar, Douglas-fir, and bur oak (Quercus macrocarpa) (Byington, 1990).

The benefits of windbreaks for agriculture in the temperate zone have long been recognized; consequently, institutions in the U.S., Canada, Europe, Australia, New Zealand, and China are currently involved in windbreak research. Tree improvement and pest management of windbreaks have perhaps received the most research attention. Other research priorities in the past included windbreak establishment and management, analysis of benefits and costs, and quantification of biophysical windbreak effects (Brandle et al., 1988; Hintz and Brandle, 1986). Despite these efforts, significant problems remain: windbreak establishment continues to be difficult; there is a very limited choice of medium-to-tall species that are well adapted and long-lived; better methods are needed for weed control, pest management, and silviculture of the windbreaks; improved understanding of the effects of windbreaks on agricultural crops, especially the benefits and costs of the practice, is necessary; and windbreak design for hilly country is currently inadequate. Despite a long history of windbreaks in land-use systems, major research opportunities remain for this important agroforestry practice.

25.3.4. Other agroforestry practices

Several other agroforestry systems have been important in particular regions of the temperate zone or are likely to become more widely established in the future. The dehesa oak woodlands of Spain and Portugal have provided acorns and forage for grazing animals and a variety of wood products (e.g., timber, charcoal, tannin, and cork) for local inhabitants for centuries (section 25.2). The natural oak woodlands cover approximately 5.5 million hectares in the two countries, with Quercus rotundifolia, Q. suber, and Q. faginea normally providing tree cover of between 5 and 20% (50-100 mature trees ha-1) (Joffre et al., 1988). Grazing and crop cultivation are common under the open oak canopy of this agrosilvopastoral system. Sheep are currently the most common grazing animal (Figure 25.5), although goats, cattle, and pigs are also important components. Grazing management is flexible but includes moving animals to field stubble and fodder sources during dry summer months, with concomitant resting periods for grasslands. In managed dehesas, oaks may be planted where tree cover is insufficient, and established trees are often pruned to improve acorn and wood production (Joffre et al., 1988).

Since the 1950s, traditional dehesa land-use has declined as woodlands have been cleared for agricultural crops or for reforestation. Sheep and goat populations have decreased substantially, and continuous grazing has become more common. Tree management has also declined due to labor shortages. When the cleared agricultural land is abandoned it tends to revert to dense shrublands rather than the previous open woodland. Joffre et al. (1988) concluded that the changes in the last 30-40 years have led to the ecological deterioration of this agroforestry system.

Open woodlands in other Mediterranean countries are also used as silvopastoral systems, with either oaks or carob trees (Ceratonia siliqua) as the dominant species (Joffre et al., 1988). Land management in these systems has not been well developed, nor is it documented in the literature. Similar systems also exist in the Mediterranean-climate regions of California (U.S.) and Chile, and in all these areas, socioeconomic pressures and resultant ecological changes in the last 30-50 years have worked against the maintenance of sustainable agroforestry systems. Increased efforts are needed, and in some places are underway, to develop or maintain the potential of these land-use practices.

Small block plantings of multipurpose trees have been established in many temperate countries for production of fodder, biomass energy, and fuelwood. Additional objectives vary from sundry wood products to soil conservation and water quality protection. Although they are not yet major agroforestry practices in most regions, they offer significant opportunities for expansion and adoption in the future (Barrett and Hanover, 1991). These species may be planted as: small woodlots on farms, biomass energy plantations, strips for soil or water protection, or in a number of other configurations to meet landowner and farmer objectives. Black locust (Robinia pseudoacacia) and various alder (Alnus spp.) species are the most widely preferred species in these MPT systems because of attributes such as nitrogen-fixation, rapid growth, easy establishment from seed, and coppice regeneration (Barrett and Hanover, 1991). These and several other temperate-zone, nitrogen-fixing species are also used for pasture and crop improvement when planted as scattered trees or hedges (Dawson and Paschke, 1991).

Both multiple purpose trees and more traditional wood-product species (e.g., eucalyptus and pines) have been used to mitigate soil conditions that would otherwise restrict agricultural crops. In California and Australia, small plantations or strips of eucalyptus have been planted on agricultural crop land to lower ground water tables and reduce soil salinity. Also in Australia, radiata pine on pasture land has lowered water tables 1.5 m in 10 years (Anderson et al., 1988). The long term benefits from these practices are unknown (Battini et al., 1983; Scherr, 1991). Furthermore, various pines, black locust, and honeylocust have been planted on degraded mine sites as an initial reclamation step in order to prevent soil erosion and create systems suitable for grazing.