VIEWPOINT Agroforestry, Horticulture, and the Evolution of Cropping Systems Kenneth R. Tourjee Center for Agroforestry, University of Missouri–Columbia, Columbia, MO 65211-7270 John M. Shopland Horticulture and Agroforestry Research Center, University of Missouri–Columbia, 10 Research Center Road, New Franklin, MO 65274-9727 Michele Warmund Department of Horticulture, University of Missouri–Columbia, Columbia, MO 65211-7140

The Association for Temperate Agroforestry Understanding this relationship requires ex- attention on genetic diversity and plant do- defines agroforestry as an intensive land man- amining the roles of silviculture and horticul- mestication. The genetic diversity of horticul- agement system that optimizes the benefits ture in agriculture. tural and agronomic crops is usually defined in from the biological interactions created when terms of genotypes or highly selected popula- trees and/or shrubs are deliberately combined AGRICULTURE tions. Forestry generally defines genetic di- with crops and/or livestock (Slusher et al., versity at the species and provenance levels; a 1997). Agroforestry is becoming increasingly Agriculture has been broadly defined as forest stand of even a single species usually important in the United States, and in other the technology of raising plants and animals contains much more genetic diversity than temperate regions of the world, as an attractive (Janick, 1986). Janick (1986) defined horti- does the typical orchard. We can infer from option for small landowners wishing to farm culture as that part of plant agriculture con- this that forestry and horticulture/agronomy marginal land or manage forested tracts more cerned with so-called garden crops, agronomy operate at opposite ends of the domestication intensively. The several research centers in the with field crops, and forestry with forest trees scale, since domestication is an evolutionary United States committed to agroforestry re- and their products. The species and commod- process driven by selection. Oak (Quercus sp. search are one measure of its importance. ity of interest determine classification into one L.) timber, pecan [Carya illinoinensis These include the Univ. of Missouri Center for of these three areas. The distinguishing char- (Wangenh.) K.] nuts, and apple (Malus Agroforestry (Columbia), the U.S. Dept. of acteristics of each area are derived from their ×domestica Borkh.) fruit are three commodi- Agriculture (USDA)–Agricultural Research respective commodities. These characteristics ties that show this relationship between do- Service Dale Bumpers Small Farms Research are determined by species biology and the mestication and agricultural status. Oaks are Center (Booneville, Ark.), and the USDA– cultural practices used in commodity produc- relatively unselected and are under the pur- Forest Service National Agroforestry Center tion, and therefore suggest a forestry : agricul- view of forestry. Apples have a long history of (Lincoln, Nebr.). ture dichotomy (Spedding, 1988). This di- crop evolution and are clearly a horticultural The objective of this paper is to relate chotomy is best discussed in terms of ecologi- crop. Pecan represents a transitional stage temperate agroforestry to horticulture in terms cal and genetic systems. between these two domestication endpoints. of cropping systems evolution. We view this Currently, horticulturists are developing in- evolution as a movement from natural for- ECOLOGICAL PERSPECTIVE tensive management practices and selecting est ecosystems toward agroecosystems. pecan cultivars, although wild germplasm is Agroforestry is a blending of its constituent An ecological perspective on the differ- still an important contributor to crop produc- (agriculture and forestry) characteristics (see, ences between forest and agricultural systems tion (William Reid, personal communication). e.g., Fig. 2.1 in Nair, 1993). Therefore, these focuses attention on management decisions. characteristics should be observable in each of Key management practices of horticulturists AGROFORESTRY the five temperate agroforestry systems—al- and agronomists (fertilization, irrigation, pes- ley cropping (Garrett and McGraw, 1999), ticide/herbicide treatment, and tillage) are more Agroforestry’s identity relies less on spe- forest farming (Hill and Buck, 1999), riparian intensive than those of foresters (pruning, seed- cific commodities and more on how these end- buffer strips (Schultz et al., 1999), silvopastures ing/planting, and thinning). Moreover, agro- products are produced than does the identity of (Clason et al., 1999), and windbreaks (Brandle ecosystems are manipulated to a much greater either horticulture or forestry (Anderson and et al., 1999). Consequently, both horticultural extent than forest ecosystems. Traditionally, Sinclair, 1993). Consequently, its definition (i.e., agricultural) and silvicultural research agricultural science attempts to maximize pro- must reflect a commodity-free but discipline- are required to develop productive and sus- ductivity, often with intensive inputs of labor centered approach. Alley cropping can serve tainable agroforestry cropping systems. How- and resources, by creating ‘domesticated en- as a model for examining the nature of ever, the expertise exists within horticulture to vironments’ [environments modified to suit agroforestry. Based on the preceding discus- study the interaction between trees and other the purposes of a grower—e.g., greenhouse sion, the extent to which ecological principles crops (e.g., Hogue and Neilson, 1987). There- rose (Rosa sp.) production]. Today, however, are applied to reduce the need for external fore, the relationship between silviculture and there is great interest in developing and using inputs, and how much domestication is ex- horticulture in agroforestry is not necessarily agricultural methods that are less intensive pressed by the germplasm, determines whether demarcated along expertise in managing the [e.g., no-till soybean (Glycine max L.) produc- an alley cropping system is identified as productivity of woody vs. nonwoody plants. tion]. Silviculture, on the other hand, has been agroforestry. From this perspective, intercrop- developed to manage natural woodlands for ping hybrid corn (Zea mays L.) between rows optimal production, although current trends of an almond [Prunus dulcis (Mill.) D.A. are toward a greater reliance on intensive Webb] cultivar is not agroforestry—although Received for publication 13 Apr. 1998. Accepted plantation forestry. it may be a profitable agricultural intercrop- for publication 26 Aug. 1998. We gratefully ac- knowledge the editorial assistance of Lyn Mortimer. ping system. Both corn and almond grown in The cost of publishing this paper was defrayed in GENETIC PERSPECTIVE monoculture are agricultural crops selected part by the payment of page charges. Under postal from elite germplasm pools and their cultiva- regulations, this paper therefore must be hereby A genetic perspective on the differences tion requires a high level of inputs. Intercrop- marked advertisement solely to indicate this fact. between forests and agroecosystems focuses ping them does not introduce a forestry com-

22 HORTSCIENCE, VOL. 34(1), FEBRUARY 1999 ponent to the system from either an ecological a selection for improved production efficiency. (provenances) of ginseng germplasm compat- or genetic perspective. Harvesting agricultural products (outputs) from ible with pine stands and developing cultural Agroforestry cropping systems occupy an a system requires that inputs be increased. methods to produce high-grade timber directly intermediate position, between natural and This is a distinguishing feature in comparing employ forestry concepts. ‘Silvi/horticultur- agroecosystems, on a scale measuring domes- natural with agroecosystems. Further fine tun- ists’ would maximize the efficiency of man- tication of the production environment. This is ing the management of these cropping systems agement operations for the alley cropping sys- a result of blending agriculture with forestry. will increasingly depend upon improving the tem. Agroforestry breeders would develop These systems should be designed to mimic at efficiency of input use—a horticultural more blueberry cultivars with increased shade toler- least some processes of natural forest ecosys- than a silvicultural research theme. However, ance, ginseng cultivars with increased toler- tems to obtain benefits that must otherwise cropping systems derived from agroforestry ance to acidity, and pine cultivars that produce accrue to the system externally. Woodmansee will remain different from those developed quality wood and pine straw with growth char- (1984) compared natural with agricultural eco- from a strictly agricultural standpoint. This is acteristics responsive to the intercropping en- systems, and concluded that natural ecosys- because the forestry component of agroforestry vironment. tems have developed several mechanisms to ensures that all of the implications of the Alley cropping hay between rows of black sustain productivity that could be exploited in question “Does a hectare of cropland equal a walnut trees grown for nut production is an properly designed agricultural systems. The hectare of wild host plant?” are addressed as example that shows the role of germplasm vegetative canopies and litter of natural eco- they relate to cropping system design (Janzen, domestication in defining agroforestry. Black systems reduce soil temperatures during the 1986). These include the ecological mecha- walnut is a heterozygous forest tree found in season of most biotic activity, thereby reduc- nisms described by Woodmansee (1984) and heterogeneous populations. It is in the early ing the rates of microbial activity and water the benefits obtained from the support a di- stages of domestication and horticulturists have evaporation. Decreasing the evaporation rate verse plant community provides to predator not yet defined advanced cultural practices. prolongs soil water availability. The lower soil insect populations. Altieri (1991) provides Breeding for improved cultivars would rely temperatures also decrease rates of decompo- examples demonstrating the use of biodiversity heavily upon the concept of provenance. Se- sition and mineralization. The net effect of in agroecosystems to control pest populations. lection against unwanted variation (e.g., vari- these processes is to synchronize plant and An evolution toward horticultural research ability in time to nut maturity) would probably microbial activity and reduce gaseous and is evident also from a genetic viewpoint. For be as important as improving mean perfor- leaching losses from excess mineralization. example, black walnut (Juglans nigra L.) is a mance. Therefore, breeding programs for black Another feature of natural ecosystems is that a forest species widely used in alley cropping in walnut would be similar to Duvik’s prebreeding large part of their nutrient content is retained the Midwest and parts of Canada. In the future, scheme for domestication of wild germplasm within living tissues. Nutrients immobilized black walnut cultivars will bear little resem- (Duvik, 1990). He proposed that prebreeding in living biomass are not vulnerable to erosion blance to their wild relatives. Nut and timber programs should be established to make the and leaching loss. Furthermore, Woodmansee production and quality will greatly exceed that wild germplasm of established crops more noted that the rooting structure of natural eco- found in native forest trees. The ultimate de- accessible and useful to breeders. systems displays greater heterogeneity than velopment of the cropping system will rely on Cropping systems coevolve with their does that found in typical agricultural systems, horticultural research in the sense that its germ- germplasm because crop genotypes are se- thus reducing the potential for nutrient loss plasm will be highly selected and further ge- lected to respond to specific cultural practices. through leaching. netic gains generally will be made without The development of the mechanically har- Agroforestry research focuses on acceler- recourse to wild germplasm. Also, genotypes vested tomato is an excellent example of this ating domestication through the development will be selected to interact positively with process (Dickman, 1978). Once a cropping of advanced germplasm, cultural practices, specific cultural practices. system is developed and established on a wide and orchard design to exploit these ecological scale, market forces maintain it, and large mechanisms of natural ecosystems. Domesti- CROPPING SYSTEMS fluctuations of the cropping system will be cation is directed toward the development of curtailed. The development of agroforestry cropping systems that draw on the diversity We define agroforestry cropping systems practices, therefore, is more likely to begin inherent in a forest environment to create an through the tension between forestry’s ap- with wild germplasm without well-defined agroecosystem with a reduced need for exter- proach to mix naturally compatible indigenous cultural practices. Simply interplanting culti- nal energy inputs. Therefore, ecological con- species with horticulture’s charge to manipu- vars of a woody agricultural crop with those of siderations (i.e., understanding the relation- late the environment for maximum productiv- another agricultural crop should not be consid- ships among component crops of the system), ity of perhaps disparate and exotic species. ered agroforestry because this ignores the co- as well as the cash value of the companion Intercropping blueberry (Vaccinum sp. L.) and evolution of crops and cultural practices within crop, are critical aspects in system design. ginseng (Panax quinquefolium L.) with pine cropping systems. It is unlikely to result in These relationships include enhancing pest (Pinus sp. L.) is an example of an alley crop- optimal productivity for either crop without management, minimizing energy inputs, ping system that emphasizes the use of eco- management practices designed to enhance checking soil erosion, and increasing the ben- logical principles in its design. Wild blueberry beneficial interactions between system com- efits of mutualism (e.g., through nitrogen fixa- and ginseng both evolved in a woodland envi- ponents (van Noordwijk, 1996). tion, mycorrhizal compatibility, shading, etc.). ronment as understory plants exploiting niches Also, the effects of allelopathy and competi- different from those of their companion over- CONCLUSIONS tion for light, moisture, and nutrients must be story trees. Potential commodities from this gauged. system are blueberries, ginseng botanicals, Agroforestry cropping systems are mod- pine straw, and timber. Pine straw litter func- eled after forest ecosystems and agroforestry HORTICULTURE tions as acidic mulch (blueberry thrives on research has an ecological and genetic basis. acidic soils) that improves soil structure. The These two disciplines relate agroforestry re- Cropping systems research developed trees provide a shade environment conducive search topics to those of horticulture and silvi- through agroforestry becomes increasingly to ginseng production. Optimization of this culture. Ecology brings an emphasis on sys- horticultural as the system evolves, since cropping system’s efficiency would depend tems science and modeling. Genetics provides agroforestry defined in this way is an interme- on research that is both horticultural and silvi- an understanding of genotype-by-genotype diate stage in the domestication process. As cultural. Designing the orchard to maximize interaction and selection. Therefore, the cropping system becomes more widely the opportunity for light interception by the agroforestry research is constrained to a greater adopted, the economic aspect of the system blueberry bushes while maintaining produc- extent than is horticultural research, by those increasingly dominates. Markets are estab- tivity from the other components would draw issues addressing development of cropping lished and competition among farms result in on horticultural concepts. Identifying sources practices that emphasize minimum resource

HORTSCIENCE, VOL. 34(1), FEBRUARY 1999 23