0564-B1

Refining Agroforestry Systems for Sustainable Agro-Ecosystems in Temperate North America

N.V. Thevathasan^[1] and A.M. Gordon

Abstract

In 1987, the University of Guelph established a large field experiment on 30 ha of prime agricultural land in southern Ontario, Canada. The purpose was to investigate various aspects of intercropping trees with agricultural crops, a common practice in many geographical regions around the world. A variety of spacing, crop compatibility and tree growth and survival experiments were initiated at that time, utilizing ten tree species within the genera Picea, Thuja, Pinus, Juglans, Quercus, Fraxinus, Acer, and Populus. Two between-row spacings (12.5 m, 15 m) and two within-row spacings (5 m, 6.25 m) were utilized in conjunction with all possible combinations of three agricultural crops [soybeans (Glycine max L.), corn (Zea mays L), and either winter wheat (Tritium aestivum L.) or barley (Hordeum vulgare L.)], stratified by approximately six soil types.

Investigations over the last decade have revealed several complementary interactions as a result of ideal tree-crop establishment combinations (e.g. the transfer of nitrogen from fall-shed leaves to adjacent crops with enhanced soil nitrification as the proposed mechanism). In addition, soil organic carbon adjacent to tree rows has increased by over 1%, largely as a result of tree litterfall inputs. Furthermore, it is estimated that intercropping has reduced nitrate loading to adjacent waterways by 50%, a hypothesized function of deep interception by tree roots. We have also noticed increased bird diversity and usage within the intercropped area as compared to mono-cropped adjacent agricultural areas, and have recorded increases in small mammal populations. We speculate that these are indicative of major changes in the flow of energy within the trophic structure identified with intercropping systems. Current investigations address carbon inputs and sequestration, leaching losses and the distribution of earthworms. The results obtained from the above studies are discussed in this paper.

Introduction

Agroforestry is an approach to land-use that incorporates trees into farming systems, and allows for the production of trees and crops or livestock from the same piece of land in order to obtain economic, ecological, environmental and cultural benefits (Gordon and Newman, 1997). In North America, many different types of agroforestry have been employed historically (Gordon et al., 1997), but the vast potential for economic and environmental benefits attributed to agroforestry have yet to be realized on a large scale. The main types of agroforestry systems currently being researched in many areas of North America and in southern Ontario, applicable to portions of the Canadian landscape, would include windbreaks and shelterbelt systems, silvipastoral systems, integrated riparian forest systems, forest farming systems and tree-based intercropping systems.

Interaction in agroforestry is defined as the effect of one component of the system on the performance of another component and/or the overall system (Nair, 1993). Rao et al. (1998) have indicated that the study of interactions in agroforestry systems requires the examination of a number of complex processes, including processes related to soil fertility, competition, microclimate, pest and diseases, soil conservation and allelopathy. Exploitation of positive interactions between the woody (tree) and non-woody (agricultural or annual crop) components and the minimization of negative interactions are the key to the success of tree-based intercropping systems.

In 1987, a tree-based long-term intercropping research experiment was initiated at Guelph, Ontario, Canada using a total of 10 different species of hardwood and coniferous trees, annually intercropped with corn (Zea mays L.), soybeans (Glycine max L.), and either winter wheat (Tritium aestivum L.) or barley (Hordeum vulgare L.). Row spacings of 12.5 and 15 m (within-row spacings of either 3m or 6m) are also incorporated into the design, which covers 30 ha, stratified by approximately six soil types. Conventional standard cultural practices are implemented annually for the respective agricultural crops grown between the widely spaced tree rows.

This paper will deal mainly with the last 7 years of research conducted at the University of Guelph Agroforestry Research Station, identifying and quantifying these interactions so that management strategies can be established that promote complementary interactions and reduce or eliminate negative competitive interactions. The establishment of suitable management strategies will facilitate the adoption of tree-based land-use practices in southern Ontario and other suitable geographical regions in the rest of the province and Canada.

Interactions related to soil fertility

The effects of poplar (hybrid clone DN 177; Populus deltoids x Populus nigra 177) leaf biomass distribution, as a result of litterfall, on soil N transformations and soil organic C was studied from 1993 to 1995. In field experiment 1, poplar leaves were removed after leaf senescence in 1993 and 1994; in experiment 2 leaves were not removed. Poplar leaf biomass distribution showed a distinct pattern, with almost 80% of the leaves falling within 2.5m from the poplar tree row (Table 1).

Table 1. Poplar leaf-biomass distribution in a poplar-barley intercropping system during the 1993 and 1994 growing seasons, University of Guelph Agroforestry Research Station, southern Ontario, Canada (adapted from Thevathasan and Gordon, 1997).

^a 84% of leaf biomass found in the 0-2.5 m zone
^b 82% of leaf biomass found in the 0-2.5 m zone

Differing rates of poplar leaf biomass input across the field created distinct regions with respect to the accumulation of soil nitrogen and carbon. Based on the abundance of these resource pools, the intercropped alley can be divided into three regions: the area close to the poplar tree row (0-2.5 m on either side of the tree row), the middle of the crop alley (2.5-8.0 m from the tree row), and the area furthest away from the tree row (8.0-15.0 m). Observed mean soil nitrate production in the above regions during 1993 (June to August) was 73.1, 41.0 and 34.0 µg 100g^-1 dry soil day^-1 respectively. In 1995, as a result of the removal of poplar leaves from the field for two consecutive years (1993 and 1994), nitrate production values were decreased to 17.6, -2.8 and -1.7 µg 100g^-1 dry soil day^-1 in the same regions, respectively. However, in experiment 2 (June to August 1995, leaves not removed) mean nitrate production in the same regions was 109.4, 15.4 and 5.7 µg 100g^-1 dry soil day^-1 respectively. It appears that the addition of poplar leaves significantly (p < 0.05) affected nitrate production rates, especially in regions close to the tree row and in the middle of the crop alley. It also appears that the major portion of nitrate was released from the labile organic pool (recently added poplar leaf biomass) rather than from the recalcitrant organic pool, since the removal of poplar leaves from the field did not significantly change the soil organic carbon pool over the three-year period. For further details on materials and methods please refer to Thevathasan and Gordon, 1997.

Soil organic carbon (SOC) did not significantly (p > 0.05) change in the indicated regions over the three-year period during which the above study was conducted with recorded SOC means of 3.25, 2.32 and 2.50% respectively (Figure 1). This was to be expected as only 15 to 35 % of added organic residue is actually incorporated into the permanent organic pool (humus) (Brady, 1990). Hence, two years of addition or removal has unlikely affected the total SOC pool close to the poplar tree row.

Figure 1. Variation in soil organic carbon as a function of distance from the tree row in 1993,1994 and 1995 at Guelph, Ontario, Canada (adapted from Thevathasan and Gordon, 1997).

However, it should be emphasized that the high rate of poplar leaf biomass addition (1000 kg C ha^-1year^-1) over a period of 7 to 8 years has resulted in an increase of SOC of approximately 1% close to the tree row; this effect extends into the alley for approximately 4m. This is about a 30-35% increase in SOC close to the tree rows over the given period of time. The build up of soil organic matter under tree canopies and the positive influence of agroforestry tree species in improving soil fertility has been well-reported (Nair, 1993).

Earthworm dynamics under intercropping systems

A study of earthworm population dynamics in a temperate intercropping system was conducted at the University of Guelph Agroforestry Research Station in 1997 and 1998 (Price and Gordon, 1999). Tree species played an important role in determining the spatial and temporal distribution of earthworms within the intercropping system. The effect of tree species on earthworm density and biomass at different periods throughout the year is shown in Table 2. Significant differences (p < 0.05) in earthworm density and biomass were observed between sampling periods and tree species. For example, poplar and ash (Fraxinus americanus L.) tree rows had the greater earthworm densities, possibly due to either greater litter contributions or more rapid decomposition of leaf litter. Earthworm numbers decreased during the summer period but these values were still significantly greater (p <0.05) than those from a comparable conventionally cropped field.

Table 2. Comparison of total mean earthworm numbers (No. m^-2) and biomass (g. m^-2) under an intercropped system and a conventional cropped corn system at the University of Guelph Agroforestry Research Station, Guelph, Ontario (adapted from Price, 1999).

Values followed by the same letter in brackets () within a column are not significantly different (p<0.05)

Values followed by the same letter across a row are not significantly different (p<0.05)

Biological interactions related to avian diversity

A study was conducted on-site by Williams et al. (1995) to investigate the bird use of an intercropped cornfield, a conventional cornfield and an old-field site. The old-field site was comprised of various tall grasses and weeds including goldenrods (Solidago spp.), asters (Aster spp.) and milkweed (Asclepias spp.). Only one species of bird nested in the cornfield and the avian density in the intercropped area was similar to that of the old-field, 7 and 8 species respectively. More species foraged in the intercropped plots (10 species) compared to the cornfield (2 species) and old-field site (8 species).

The study revealed that intercropping provided opportunities for birds to nest and forage that were not available in the monocropped cornfield. The diversity of the breeding population in the intercropped field approached that found in the nearby old-field site, although some of the species were different. The intercropped field also provided foraging opportunities for other species whose diversity and numbers clearly demonstrate the value of the site to local and migrating bird populations.

Biological interactions related to insect diversity

The effects of tree rows on the distribution and diversity of insects in a temperate intercropping system was studied (University of Guelph, Agroforestry Research Site). Arthropod abundance, representation by functional group, and hymenopteran family richness and diversity were all compared between the intercropped and the adjacent monoculture site. This research can further our understanding of intercropping ecology and improve agroforestry system design in terms of entomological concerns.

The results suggests that taxons such as the Opiliones, Dermaptera and Carabidae, which are associated with organic litter and areas that provide shelter during the day were significantly higher in the intercropped system (agroforestry) than in the monoculture system. The abundance of Hymenoptera, and several of its families, was also significantly higher in the intercropped site than in the monocropped site, although no differences were observed in terms of overall family richness and diversity. There were significantly higher numbers of parasitoids and detritivores in the intercropped agroforestry system than in the monoculture system, and the intercropped treatment also supported significantly higher ratio of parasitoids to herbivores. It could be concluded that trees with crops such as corn may improve pest management by providing habitat to augment populations of natural enemies. For more details please refer to Howell (2001).

Interactions related to nutrient and soil conservation

Recent research conducted at the intercropping site has shown that nitrogen leaving the intercropping site as nitrate can be potentially reduced by more than 50% when compared to losses from a monocropped barley field (Thevathasan, 1998). A quantitative assessment of N flow through a tree-based intercropping system and compared to a field crop system will provide useful information on the ability of trees to reduce nitrate leaching. Actual leaching losses have been estimated to be only 9 kg N ha^-1 yr^-1 for the intercropping site although leaching losses in a monocropped field adjacent to the above intercropped field were 20 kg N ha^-1 yr^-1. Thus, the adoption of Intercropping appears to have reduced leaching losses by 11 kg N ha^-1 yr^-1 (Thevathasan, 1998). Understanding nitrogen flow in these systems may lead to reduced nitrate loading to nearby waterways, and may also be useful for future fertilizer management recommendations. Furthermore, with respect to greenhouse gas emissions, less nitrate leaching will reduce potential N₂O emissions into the atmosphere from agricultural fields. The latter scenario is discussed below under the sub-heading ‘Interactions related to climate change mitigation’.

Interactions related to climate change mitigation

In the recent past, many studies have identified tree-based land-use practices as a significant global opportunity to reduce the accumulation of CO₂ in the atmosphere (Dixon, 1995). The United Nations has also estimated that agroforestry based land-use practices on marginal or degraded lands could sequester 0.82-2.2 Pg C per year, globally, over a 50-year period (Dixon et al., 1994). Apart from carbon (C) sequestration, tree-based intercropping systems can also significantly reduce greenhouse gas (GHG) emissions (e.g. nitrogen oxides (NO_x)). Therefore, these systems might potentially have a significant impact on climate change mitigation due to the following reasons: 1) Tree-based intercropping systems can be adopted in agricultural land classes from 1 through 4. Therefore, the land base in Canada that could potentially be brought under tree-based intercropping is substantial which in turn, can have a significant effect on C sequestration and GHG emission reduction; 2) The tree component occupies a part of the land base, reducing the land for agriculture and subsequently reducing the need for inputs such as supplemental nitrogen for crop production. The reduction in N₂O emissions will be directly proportional to the land base occupied by trees; 3) The decrease in nitrogen moving out of the rooting zone will lead to reduced NO_x emissions as a result of denitrification in surface water resources; 4) If the tree species are deciduous, the annual leaf fall will cycle some nitrogen back to the soil reserve. While this nitrogen will be localized to the area close to the tree, it does constitute a quantifiable contribution of nitrogen to the subsequent agricultural crop. Thus, variable application rates of inorganic fertilizer, especially N, can result in reduced environmental losses. Apart from GHG emission reduction, more importantly, C sequestration in agricultural fields can be augmented through this type of land-use practice. Further, annual leaf litter input and fine root turnover can also significantly influence long-term soil organic C dynamics in agricultural fields.

Quantitative measurements on potential N₂O emission reduction and C sequestration were taken at the University of Guelph Agroforestry Research Site in 1999. GHG emission reduction (e.g. N₂O) estimates were derived from previously collected data on this site. Carbon sequestration measurements were collected by destructively sampling a 12-year-old fast-growing fibre tree species (hybrid poplar).

Estimates indicate that the quantity of N recycled to the agricultural zone in the leaf litter from fast-growing deciduous species (e.g. hybrid poplars) can contribute up to 5 kg N ha^-1 yr^-1 (Thevathasan, 1998). This implies that the amount of inorganic fertilizer addition can potentially be reduced by this amount. Cole et al. (1996) suggest that N₂O emission from agricultural land is directly related to the rate of N application and that 1.25% of the N applied is emitted directly from the land as N₂O. As previously indicated, nitrogen leaching losses were only 9 kg N ha^-1 yr^-1 at this site compared to 20 kg N ha^-1 yr^-1 calculated for an adjacent monocropped field, indicating that intercropping has reduced leaching losses by 11 kg N ha^-1 yr^-1 (Thevathasan, 1998). These results suggest that lower additions of inorganic fertilizer in concert with less nitrate leaching losses could lead to a significant reduction in N₂O emission from tree-based intercropping systems.

Quantitative measurements of C sequestered in both above and below-ground woody components of a fast growing fibre tree species at the Agroforestry Research Station, indicate that over a period of 12 years, 9 MT (9000 kg) of C ha^-1 have been sequestered (stems ha^-1 = 111, at a spacing of 15m between rows and 6m within rows). Theoretically, therefore, trees alone have immobilized 36 MT of CO₂ ha^-1 over this 12-year period. In addition to the C sequestered in woody components, C contributions to the soil from annual litterfall alone were 1332 kg C ha^-1. Previous research indicates that the latter annual addition over a period of 10 years in concert with fine root turnover has increased soil organic C by 1% [absolute value] close to the tree rows (Thevathasan and Gordon, 1997). It is important to recognize that trees can significantly impact the C balance when introduced into agricultural fields. In a monocropped agricultural field, annual C input to the soil is in the range of 500 to 700 kg C ha^-1 yr^-1 (P. Voroney, 1999 pers. comm.) whereas in a tree-based intercropping annual C input can be as high as 2500 kg C ha^-1 yr^-1, about 5 times more than that found in monocropped agricultural fields.

Conclusions

In this new millennium, one of our fundamental priorities as policy makers, scientists and resource managers should be to ensure that management recommendations result in sustainable and economical production capabilities. Greater focus must be given to enhancing the attitude that we are essentially stewards of both soil and water resources. The adoption of agroforestry systems, especially intercropping as presented in this paper, shows much promise in this regard.

Among the various alternative agricultural practices currently under consideration in southern Ontario capable of curbing further environmental degradation resulting from traditional practices, tree-based intercropping systems are considered to be a viable option. The ameliorative effects of trees in relation to soil fertility, productivity and nutrient cycling and their filtering ability can be exploited, especially in the context of developing tree-based intercropping systems on both marginal and prime agricultural lands. The success of intercropping depends mainly on the ability of the system components to maximize resource utilization while at the same time maintaining 'complementary' interactions between them. When this occurs, productivity per unit land area is often enhanced resulting in higher economic returns. When components of an intercropping system are very different (e.g. woody and non-woody), the demand for limited resources is staggered in space and time, and resource capture and productivity per unit land area may be maximized.

On a biological level, intercropping increases micro- and macro-faunal diversity and activity, both above- and below-ground. The increased range of faunal activity gives a clear indication of ecosystem ‘health’ within an intercropping system relative to that associated with conventional agricultural practices. From an ecological perspective, intercropping systems trap larger amounts of energy at different trophic levels, demonstrating a higher energy utilization efficiency. In relation to carbon sequestration and greenhouse gas (e.g. N₂O) emission reductions, tree-based cropping systems have the potential to greatly contribute to climate change mitigation. The tangible benefits that are derived from the above described eco-biological processes, along with combined yields obtained from the trees and crops, place this land-use practice above conventional agricultural systems in terms of long term overall productivity.

However, the economics of tree-based intercropping systems need to be examined in more detail. Initial establishment costs and the loss of revenue due to removing cropland from production often deter farmers from adopting these types of systems. Therefore, investigation into policy measures, and/or tax incentives and cost-share programs should be initiated in order to obtain successful adoption rates in southern Ontario in particular and the rest of the country in general.

References

Brady, N.C. (1990). The nature and properties of soils. Macmillan Publishing Company, New York. 621 pp..

Cole, C.V., C. Cerri, K. Minami, A. Mosier, N. Rosengerg and D. Sauerbeck. (1996). Chapter 23. Agricultural options for mitigation of greenhouse gas emissions. p. 745-771. In: R.T. Watson, M.C. Zinyowera and R.H. Moss (eds.). Climate Change 1995. Impacts, Adaptions and mitigation of Climate Change: Scientific Technical Analysis. Published for the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK.

Dixon, R.K. (1995). Agroforestry systems: sources or sinks of greenhouse gases? Agrofor. Syst. 31: 99-116.

Dixon, R.K., J.K. Winjum, K.J. Andrasko, J.J. Lee and P.E. Schroeder. (1994). Integrated land-use systems: Assessment of promising agroforest and alternative land-use practices to enhance carbon conservation and sequestration. Climate Change 27:71-92.

Gordon, A.M. and S.M. Newman (1997). Temperate Agroforestry Systems. CAB International, Wallingford, U.K. 269 pp.

Gordon, A.M., S.M. Newman and P.A. Williams (1997). Temperate agroforestry: An overview. In: Gordon, A.M. and S.M. Newman (eds.). Temperate Agroforestry Systems. 1-8 pp. CAB International, Wallingford, U.K. 269 pp.

Howell, H. (2001), Comparison of arthropod abundance and diversity in intercropping agroforestry and corn monoculture system in southern Ontario. M.Sc.F. Thesis. Faculty of Forestry, University of Toronto.

Nair, P.K.R. (1993). An Introduction to Agroforestry. Kluwer Academic Publishers. 499 pp.

Price, G.W. (1999). Spatial and temporal distribution of earthworms in a temperate intercropping system in southern Ontario. M.Sc. Thesis, Dept. of Environmental Biology, University of Guelph. 131 pp.

Price, G.W. and A.M. Gordon. (1999). Spatial and temporal distribution of earthworms in a temperate intercropping system in southern Ontario, Canada. Agrofor. Syst. 44: 141-149.

Rao, M.R., P.K.R. Nair and C.K. Ong. (1998). Biophysical interactions in tropical agroforestry systems. Agrofor. Syst. 38:3-50.

Thevathasan, N.V. (1998). Nitrogen dynamics and other interactions in a tree-cereal intercropping systems in southern Ontario. Ph.D Thesis. University of Guelph, Ontario, Canada. 230 pp.

Thevathasan, N.V. and A.M. Gordon. (1997). Poplar leaf biomass distribution and nitrogen dynamics in a poplar-barley intercropped system in southern Ontario, Canada. Agrofor. Syst. 37:79-90.

Williams, P.A., H. Koblents and A.M. Gordon. (1995). Bird use of an intercropped corn and old field sin southern Ontario. pp.158-162. In: Ehrenreich, J.H. and D.L. Ehrenreich (eds.), Proceedings of the Fourth North American Agroforestry Conference held July, 1995, Boise, Idaho. 182 pp.