scales and provide connectivity in agricultural landscapes at much larger scales (Schultz et al., 2000). Riparian buffers are able to reduce non‐point‐source pollution from agricultural fields through reduced runoff velocity and promotion of infiltration, increased nutrient retention through trees utilizing excess nutrients transported in runoff, and increased sediment deposition on land (Jose, 2009).
In intercropping systems, microclimate modification is common, and although the competition for water, light, and nutrient resources between the tree and crop components is complicated, improved and sustained crop yields have been noted (Thevathasan & Gordon, 2004). Due to enhanced structural diversity within agroforestry systems, microclimatic modifications and therefore plant growing conditions are not homogeneous as they are within conventional agricultural systems. A recent study by Coleman et al. (2020) found that both abiotic (light, soil moisture) and biotic (available soil nutrients) gradients within a 26‐yr‐old tree‐based intercropping system intercropped with concentrated short‐rotation willow (SV1; Salix dasyclados Wimm.) had significant influences on intraspecific variation in crop leaf traits, including increased specific leaf area and crop leaf N concentrations closest to the tree rows. These results contribute to an enhanced understanding of nutrient cycling within agroforestry systems and indicate that tree litter inputs may reduce the need for crop amendments, especially near the tree rows.
The presence of trees within agroforestry systems can also have more indirect influences on nutrient cycling and soil fertility. Price and Gordon (1998) examined the spatial and temporal distribution of earthworms in an 11‐yr‐old tree‐based intercropping system planted with silver maple, white ash (Fraxinus americana L.), and hybrid poplar, in combination with soybean. The researchers found the greatest density of earthworms within the tree rows, with typically decreasing earthworm density towards the middle of the cropping alley. Earthworm density was drastically reduced in the summer, potentially tracking with reduced food availability (litterfall) and reduced soil moisture compared to the spring, and earthworm distribution tended to become more uniform during the summer. The authors found that earthworm density was highest near poplars, providing further evidence of the importance of tree species selection when considering soil fertility and other ecosystem functions.
Many additional goods and services can be provided by the suite of recognized agroforestry practices, including odor control (Tyndall & Grala, 2009), opportunities to embrace integrated pest management systems with reduced pesticide input (Diaz‐Forestier, Gomez, & Montenegro, 2009), and the control of Escherichia coli outbreaks associated with manure application (Dougherty, 2007). Agroforestry systems can also enhance nutrient cycling and nutrient use efficiency with subsequent improvements in downstream water quality and reduced requirements for crop amendments (Jose, 2009). Thevathasan and Gordon (1997) utilized a 7–9‐yr‐old hybrid poplar tree based intercropping system planted with barley and found that mean nitrification rates, N availability, and C content were higher in soils closest to the poplar tree rows compared with the middle of the crop alley. It was also found that soil nitrification rates, soil C, and plant N uptake adjacent to the tree rows were influenced by the leaf biomass inputs of the preceding year, potentially contributing to increased aboveground biomass and greater grain N concentration in the barley intercrop.
In natural systems, a long‐term ecological approach has proven useful to understanding the importance of (a) slow processes that occur on the scale of decades to centuries, (b) processes with high annual variability, (c) rare and unique events, (d) subtle processes, and (e) complex processes with many interacting factors. A long‐term ecological research perspective also holds much potential for helping us understand agroforestry systems. The temporal context provided by engaging in such research can aid us greatly in understanding large‐scale changes in ecosystem processes and thereby reveal the secrecy inherent in what has been termed “the invisible present” (Magnuson, 1990).
Such an approach to understanding the structure and function of agroforestry systems and the relationship of these parameters to net primary productivity is a strong foundation upon which to evaluate the production of ecological goods and services over long periods of time (Gordon & Jose, 2008).
Conclusions
Agroforestry offers a means of regaining some of the structural and functional characteristics that contribute to the sustainability of natural ecosystems that have been lost in the conversion of those ecosystems to homogeneous agroecosystems. An understanding of the structure and function of natural ecosystems is essential to the successful implementation of agroforestry if we wish to create more heterogeneous agroecosystems.
A complete knowledge of the many ecological processes and interactions responsible for a natural system’s sustainability will always elude us—an ecosystem is just too complex. However, perennialism and a high proportion of area in mid‐ to late‐successional states is the usual condition of natural ecosystems and an obvious goal in designing a sustainable agroecosystem. We are making progress in determining how to meet that goal, but there is much left to be learned in both basic and applied ecology. With respect to the inclusion of ecological principles within the broad field of agroforestry, some important areas warranting further research include: (a) the evaluation of net primary productivity (NPP = increment + litterfall + herbivory + mortality) for all types of agroforestry systems in different geographical regions; (b) the continued study of C sequestration and the emission of greenhouse gases such as CO2, N2O, and CH4 in agroforestry systems; (c) the study of belowground interactions and processes in the realm of microbial ecology, root competition, and mycorrhizal associations; (d) the study of both positive and negative interactions among trees, shrubs, grasses, and forbs in agroforestry systems; (e) the implementation of long‐term, system‐level experiments and on‐farm demonstrations including stronger and more specific economic analyses that include the value of all ecological goods and services; (f) the study of albedo—reflectance changes occurring at the landscape level as a result of agroforestry adoption that may have implications for global warming and C sequestration scenarios; (g) a comprehensive evaluation of the biology and economics of agroforestry on a variety of sites; (h) the study of all aspects of silvopastoral systems especially as they relate to potential lowering of wood quality and value, forest regeneration, greenhouse gas emissions, and the issue of animal welfare; (i) a continued screening of useful pharmaceutical and other chemical products from forest farming systems; and (j) a comprehensive evaluation of conservation biology principles—how can we incorporate information gleaned from natural forest systems that remain in our agricultural landscapes into agricultural systems that embrace woody perennials? A better appreciation of the ecosystem and environmental services provided through agroforestry would lead to the development of rigorous environmental and economic assessments and eventually to modification of community tax structures and environmental legislation.
The ecology of highly managed agroforestry systems is becoming better understood although much remains to be done. Two recent texts on the ecology of agroforestry systems (Batish, Kohli, Jose, & Singh, 2008; Jose & Gordon, 2008) present up‐to‐date research results on ecological interactions, belowground ecological processes, resource allocation and partitioning, and the modeling of these in both tropical and temperate agroforestry systems.
Acknowledgments
We wish to sincerely thank and acknowledge and are indebted to R.K. Olson, M.M. Schoeneberger, and S.G. Aschmann for their contributions to parts of this chapter. We have reproduced verbatim or with slight modifications of their writings in the first edition of this chapter. Similarly, we are also indebted to Caron et al. (2008), as some portions in this chapter have been adapted from their publication.
References
1 Anderson, G. W., Moore, R. W. & Jenkins, P. J. (1988). The integration of pasture, livestock and widely‐spaced pine in south west Western Australia. Agroforestry Systems, 6, 195–211. https://doi.org/10.1007/BF02344759
