ECOAGRICULTURE: a Review and Assessment of Its Scientific Foundations

ECOAGRICULTURE: A Review and Assessment of its Scientific Foundations

Louise E. Buck Thomas A. Gavin David R. Lee Norman T. Uphoff with Diji Chandrasekharan Behr Laurie E. Drinkwater W. Dean Hively Fred R. Werner Sustainable Agriculture and Natural Cornell United States Agency Resource Management University for International Collaborative Research Support Development Program ECOAGRICULTUREGRICULTURE A Review and Assessment of its Scientiﬁ c Foundations

Principal Investigators Louise E. Buck Natural Resources Thomas A. Gavin Conservation Biology David R. Lee Agricultural Economics Norman T. Uphoff International Agriculture and Government with Diji Chandrasekharan Behr Natural Resources Laurie E. Drinkwater Horticulture W. Dean Hively Crop and Soil Science Fred R. Werner Natural Resources

Cornell University Ithaca, New York

This report is submitted by Cornell University to University of Georgia SANREM CRSP under contract # 45586.

The SANREM CRSP is supported by the United States Agency for International Development Cooperative Agreement Number PCE-A-00-98-00019-00, and is managed through the Ofﬁ ce of International Agriculture, College of Agricultural and Environmental Sciences at the University of Georgia, Athens, GA, USA. Table of Contents

Chapter 1: Overview 1 Annex 1-A: Ecoagriculture Assessment Advisors and Recommendations Annex 1-B: Cornell Ecoagriculture Working Group Annex 1-C: External Organizations Consulted Annex 1-D: Experts Consulted

Chapter 2: Opportunities, Concepts and Challenges in Ecoagriculture 8 Annex 2: Time Series Data on World Grain Production total and per capita, World Food Production per capita, Food Prices, Fertilizer Use, and Pesticide Exports, 1961-2003, with comparisons indexed to time series midpoint, 1979-81 average

Chapter 3: Meeting Agricultural Productivity Objectives 26

Chapter 4: Meeting Biodiversity Conservation Objectives 42 Annex 4: Examples of Studies that Examine the Quantitative Relationship Between Various Agricultural Practices and Wild Biodiversity

Chapter 5: Meeting Livelihood Objectives 54 Annex 5: Sample of Economic Models Incorporating Livelihood, Productivity and Ecosystem Elements

Chapter 6: Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 74 Chapter 7: Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 119 Annex 7-A: A Summary of Planning, Monitoring and Evaluation Models for Integrating Multiple Objectives Annex 7-B: Examples of Partnerships and Network Organizations Supporting Ecoagriculture

Chapter 8: Conclusions 135 Chapter 1 Overview

1.1. Introduction these means, largely coincide on a global level with the lands and resources needed to secure protected areas for To deal with global problems of hunger and malnutri- wildlife habitat and species preservation. This realization, natural resource degradation, loss of biodiversity, tion led them to look for and to formulate management and rural decline, alternative visions for land use and strategies that could accommodate both objectives. natural resource management are needed as present practices and systems are making slow or no progress in Their book Ecoagriculture: Strategies to Feed the World achieving these objectives. New combinations of poli- and Save Wild Biodiversity (2002)(2002) proposedproposed thatthat ecoag-ecoag- cies, institutions, technologies and values are required, riculture be accepted and expanded as a set of inclusive linking individual interests and efforts with those of the resource management strategies for landscapes that communities and societies they live in, better balanc- can both produce more food and preserve ecosystem ing the competing aspirations for raising agricultural services, with a special concern for wild biodiversity productivity, ensuring ecological sustainability, and pro- (p. 6). While the strategies identifi ed to date often have moting rural vitality. Growing awareness of the global- been appropriate for smaller-scale units of production, level impacts of current natural resource uses, enhanced relying more on local resources than external ones, through imaging and communication technologies and there is nothing in the concept of ecoagriculture that stimulated by recognition of the increasingly globalized limits it to such production systems. Indeed, a concern nature of socioeconomic systems, has motivated inno- for both food production and the environment means vative management strategies which reconcile increased that larger-scale units, presently highly dependent on agricultural productivity with environmental protection external inputs, should be well within the purview of and greater livelihood opportunities. ecoagriculture. Ecoagriculture is an approach to land use that seeks “to 1.2. Varieties of Ecoagriculture and square the circle,” to achieve productive and sustainable Relation to Biodiversity reconciliations among objectives that are usually in competition. The concept grew out of a comparison and McNeely and Scherr regard ecoagriculture as a consid- convergence of conclusions drawn by Jeffrey McNeely, ered response to the worldwide challenge of meeting the a conservation scientist, and Sara Scherr, an agricultural well-recognized crisis of conserving global biodiversity economist, as they were reviewing spatial information while at the same time producing suffi cient food and developed through the Millennium Ecosystem Assess- livelihoods to support the increasing human population. ment.1 They saw that the lands and resources required It is not a single practice or particular kind of farming for an expanding agriculture to support still-growing system but rather an aggregation of approaches to the human populations, who need to earn their living by production of food, fi ber, and other benefi ts that all aim to integrate biodiversity conservation into agricultural 1 The Millennium Ecosystem Assessment (MEA) is an inter- development efforts. McNeely and Scherr identify national program designed to generate scientifi c information six main strategies for doing this. Three of the six are regarding the consequences of ecosystem change on human concerned with making space for wildlife preservation well-being and options for addressing these changes. It within agricultural landscapes; the other three focus on focuses broadly on global ecosystem health. The objective how to enhance the wildlife habitat value of productive of the MEA is to meet the needs of decision-makers and the public for such information. farm lands themselves. The strategies can be summarized as follows: 1) Creating biodiversity reserves that benefi t local of farms and the prospects for co-management of farming communities. This strategy involves choos- livestock and wildlife. ing areas for protection in places where there are 6) Modifying farm systems to mimic natural ecosys- clear immediate benefi ts, including environmental tems. This strategy focuses on increasing the use of services, livelihood support, farmland protection for perennials to create an agriculture that provides many unique agricultural habitats, and enhancing benefi ts of the environmental services of natural systems, obtainable from protected areas for local farmers including wildlife habitat. Agricultural landscape through market and compensation mechanisms. design is an emerging science for confi guring peren- 2) Developing habitat networks in non-farmed nials, spatially and temporally, to provide desired areas. Improving the biodiversity value of agri- services. New propagation and domestication tech- cultural landscapes will encompass areas in and niques can increase the economic value of perennial around waterways, abandoned fields and forest crops. sites. Other landscape niches for biodiversity may There are many ways in which agricultural practices, on include schoolyards, “sacred groves,” parks, areas a large scale or small scale, affect the fl ora and fauna, around roadways, industrial or hospital sites, and macro to micro, in and around the areas cultivated, agro-ecotourism locations. grazed or fi shed. In general, agricultural practices sim- 3) Reducing land conversion to agriculture by in- plify the landscape relative to the natural condition—a creasing farm productivity. Enhancing agricultural simplifi cation of plant diversity, of physical structure productivity growth and sustainability in high poten- and dimension, and of chemical mosaic. An illuminat- tial areas may encourage reduction or abandonment ing schematic is one developed by Gilbert (1980), of farming in environmentally-sensitive lands. In which shows the hierarchical and interacting nature more marginal lands, technology that enables farm- of organisms in a tropical landscape. The schema has ers to replace shifting cultivation with permanent, been simplifi ed enormously relative to the real world, higher yielding fi elds may allow former fallow land but the complexity remaining is still astounding. For to revert to natural forest, woodland and grassland, contrast, compare this image to a typical agricultural or will prevent further clearing. landscape. 4) Minimizing agricultural pollution. Reducing ag- 1.3. The Development of Ecoagriculture as rochemicals in high-input systems may be achieved an Approach through advances in organic farming, integrated pest management (IPM), and soil conservation. In Ecoagriculture McNeely and Scherr lay out consider- Use of pollutants can be reduced through improved able evidence from numerous case studies, supported effi ciency of pesticide application, use of natural by analyses and proposals about the synergies that may compounds in producing them, and employment of be gained through ecoagricultural practices. Since the IPM techniques. Whole-farm planning and related book was published, a network known as Ecoagriculture waste mitigation methods can reduce their impact Partners (EP) has taken shape (http://www.ecoagri- on biodiversity. culturepartners.org). It draws together 25 institutions 5) Modifying management of soil, water, and vegeta- and some 20 individual partners as well as over 1,000 tion resources. This strategy builds on agroecology, people on a growing and active listserve who agree that conservation agriculture, agroforestry, and sustain- knowledge and understanding need to be advanced to able rangeland and forest management as well as support the extension of ecoagriculture practices on a wildlife biology and ecology to increase farmers’ wider scale. Ecoagriculture Partners and numerous col- natural capital, and thereby long-term fl ows of farm laborators have planned an international Ecoagriculture outputs. Increasing the diversity of crop, tree, and Conference and Practitioners’ Fair to be held in Nairobi livestock components can increase the habitat value in September 2004.

2 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Following the publication of Ecoagriculture, an in- 5) to identify promising institutional opportunities and ternational group of scientists meeting in Switzerland methods for improving knowledge and understand- concluded that an assessment of the scientifi c founda- ing about ecoagriculture. tions for ecoagriculture was warranted. The assess- The report is the product of an inquiry into diverse ment would help clarify misperceptions that might be disciplinary literatures and project initiatives that are forming about the approach and begin building a more related to ecoagricultural land use and consistent with extensive and systematic knowledge base for evaluating its precepts. We have gone beyond the operational the effectiveness of ecoagriculture practices evaluated defi nitions of ecoagriculture in the McNeely-Scherr agronomically, environmentally, economically, and volume to look at the generic confl icts and complemen- socially. Such an assessment should also help deter- tarities among agricultural production, environmental mine priorities for investment in research related to conservation, and livelihood generation. The defi ni- ecoagriculture. tions that McNeely-Scherr put forward were based on Subsequently, the Sustainable Agriculture and Natural induction (from an assemblage of case studies) and Resource Management (SANREM) collaborative re- deduction (making a synthesis of values and goals). search support program funded by USAID arranged for To provide an assessment of ecoagricultural strategies faculty at Cornell University to prepare a review paper in general, we took a few steps back, not limiting the as background for deliberations at the September 2004 literature review and analysis to the parameters set in conference. The Cornell International Institute for Food, the fi rst attempt, by McNeely and Scherr, to formulate Agriculture and Development (CIIFAD) has worked a new fi eld of knowledge and practice. As the review since 1990 on interdisciplinary initiatives to expand the team, coming from a variety of disciplines—agronomy, knowledge and practice for sustainable agricultural and natural resources, agricultural economics, sociology and rural development, with special attention to the joint political science, with inputs from colleagues represent- satisfaction of conservation and development objec- ing still more disciplines—we wanted to take a more tives. This report thus is prepared for the Ecoagriculture encompassing view of this terrain. Conference and Practitioners’ Fair. Our methodology for preparing this report was to work in and with three entities: 1.4. Purpose and Methodology of this Assessment 1) the Ecoagriculture Assessment Team (EAT); The aim of this assessment is to review and summarize 2) the Ecoagriculture Assessment Advisors (EAA); the scientifi c knowledge base for ecoagriculture in the 3) the Ecoagriculture Working Group (EWG). following ways: The EAT was comprised of four Cornell faculty 1) to determine to what extent the concept of ecoagri- members and four research assistants from fi elds and culture has sound scientifi c knowledge and under- disciplines that are instrumental in the development of standing; a sound knowledge base for ecoagriculture: 1) conserva- 2) to evaluate claims that have been made about the tion biology; 2) international agriculture; 3) agricultural benefi ts of ecoagriculture on the basis of scientifi c economics; and 4) natural resources. This group under- evidence; took to plan, research, and produce the report. 3) to characterize capacities for evaluating, with sci- The EAA were nine researchers and practitioners who entifi c rigor, the tradeoffs and synergies among accepted an invitation by the leadership of Ecoagri- objectives addressed within an ecoagriculture frame- culture Partners to provide guidance and assistance work; in shaping the assessment. Each advisor was asked to respond to an initial set of questions concerning what 4) to identify signifi cant gaps in knowledge and cor- they construed to be the most important issues to be responding methodological and institutional limita- addressed in the review. In addition, advisors agreed tions in knowledge generation; and to provide further information upon request. Each

Overview 3 contributed ideas for literature to examine and agreed We liken these objectives to three legs of a stool that to review and comment on a draft report. The names depend on one another to support the ecoagriculture and affi liations of these advisors are provided in Annex enterprise. The stool can only serve its function if all 1-A, as well as their responses to the questions posed three legs are intact and working together. to them. We consider in Chapters 3, 4 and 5 each leg of the stool, The EWG included 27 Cornell faculty and two gradu- characterizing its particular contributions and concerns. ate students who volunteered to provide information Each of these chapters lays conceptual, theoretical, and and guidance to the EAT. Early during the assessment empirical groundwork for assessing how ecoagriculture period, EWG members attended a half-day seminar on alternatives may affect the respective objectives of sus- ecoagriculture presented by Dr. Sara Scherr and facili- tained agricultural productivity, biodiversity conserva- tated by members of the EAT. Copies of the McNeely- tion, and livelihood support in particular situations. Scherr book were provided to EWG members prior to Preceding consideration of the three core concerns, we the seminar so they could engage knowledgeably and discuss in Chapter 2 the concepts, opportunities and critically with issues raised by ecoagriculture concepts challenges in the production of food and ecosystem ser- and practices. The seminar generated substantive vices, biodiversity conservation, and economic evalua- feedback through group processes and prepared EWG tion of multi-objective production systems to improve members to participate in the following assessment. social and human welfare. This places our review of Many were interviewed by members of the EAT to get ecoagriculture in a broader context, offering perspec- their input to the assessment and to identify the most tives on how to know to what extent ecoagricultural relevant scientifi c literature to pursue. Annex 1-B lists approaches, applied globally and in respective regional EWG members and their research interests pertinent settings, are likely to improve or to worsen existing to ecoagriculture. conditions. It informs consideration of the potential In addition to interviewing EWG participants and other value added by these new approaches, and what are the Cornell faculty members, EAT members interviewed uncompensated costs incurred within an ecoagricultural representatives of international organizations based in framework and toolbox. Chapter 2 serves as a reference the Washington, D.C. and New York City areas whose point for understanding and assessing ecoagricultural activities are related in some way to the domain of performance and potential. ecoagriculture. The EAT sought to learn about initia- We follow the three core chapters with an amplifi cation tives that would shed light on the trials, successes, and in Chapter 6 of the emerging opportunities seen within limitations of multi-objective planning and analysis the agricultural sector for practices and farming systems and to characterize those that appear most insightful or that are more eco-friendly, capitalizing on productive promising. Our criterion for inclusion in this purposive potentials embedded in natural processes. Then in sample was that initiatives be concerned with at least Chapter 7 we examine possibilities for integrating the two of the three objectives of ecoagriculture, and ide- three sets of concerns into a coherent analytical whole. ally all three. Annex 1-C lists the external organizations Our focus is on state-of-the-art methods for multi-ob- contacted. Annex 1-D provides names and affi liations jective, multi-scale land-use planning and the potential of internal and external expertise consulted. for integrating scientifi c and planning knowledge into adaptive management frameworks. 1.5. Organization of this Report Chapter 8 offers some conclusions and suggests ways This report is organized around the three interlocking to go forward in further evaluating the potential of objectives of ecoagriculture: ecoagriculture alternatives. Suggestions are provided 1) Agricultural productivity; for advancing research and learning in this domain to expand the empirical basis for assessment, while 2) Biodiversity conservation; and seeking the prospective gains from these promising 3) Livelihood support, poverty reduction, rural vitality, approaches to land use. and contributions to human well-being.

4 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations 1.6. Defi nitions elsewhere. So local and more global biodiversity can be inversely correlated. Numbers alone are no In our efforts to forge a more integrated, interdisciplin- adequate criterion. ary assessment of ecoagriculture, it became clear that one of the hurdles to getting agreement was that key b) Another way in which quantitative measures can terms, particularly biodiversity, agricultural intensifi ca- be misleading is that not all species have the same tion, and degradation meant different things to different biological or ecological value. Many species are people. differentiated from one another in small ways, while other species are very important because (i) they are 1.6.1. Biodiversity unique, perhaps the only surviving species from a previously existing taxonomic group, or (ii) they Inside the fronstispiece of each issue of the journal Con- fi ll a particular ecological niche that has positive servation Biology is the following mission statement implications for other species, for example, within of the society of the same name: “….to help develop a food chain. the scientifi c and technical means for the protection, maintenance, and restoration of life on this planet—its Most biologists are reluctant to make any judgments species, its ecological and evolutionary processes, and that would diminish the importance of any species, its particular and global environment” (emphasis ours). yet almost all agree that some are more important than We subscribe to that goal and detail the following defi ni- others. Thus a practice or farming system that would tion of biodiversity as: all species of native organisms, preserve a “more critical” species even if local biodi- plant and animal, vertebrate and invertebrate, macro- versity were diminished would be positively evaluated. and micro-, as well as the genetic and morphological Conversely, any practice or farming system that com- variation within the populations of each of those spe- promises such a species, even if that farming system is cies. In addition, biodiversity includes the biological benefi cial for biodiversity in some general way, would processes (e.g., predator-prey, plant-pollinator) that be less desirable. produced those organisms and of which those organ- Evaluating the biodiversity associated with a particular isms are a part. To avoid confusion, however, we will agricultural practice or farming system in a certain loca- usually refer to the above as wild biodiversity and to tion thus has qualitative as well as quantitative dimen- the variety of crop and livestock varieties, landraces, sions. Economists have tended to want unambiguous and species as agrobiodiversity. quantitative measures to use in evaluation; environmen- Biodiversity has been popularly understood in quantita- talists usually put more emphasis on qualitative aspects tive terms, perhaps because of the emphasis on preser- when evaluating biodiversity. vation of endangered species. If any species becomes In addition, most conservation biologists, and the public extinct, this reduces biodiversity, framing evaluation in generally, tend to think in terms of preserving species. simple numerical terms. But there are “scale” questions The “species” has long held a central position in biology and “value” questions that quickly make evaluation and, therefore, it is this unit of biodiversity to which more complex. we have been naturally attracted for keeping score of a) It is possible that cutting down an old-growth for- how we are doing. The danger of this focus, however, est, for example, could result in an increase in local is that humans manage or manipulate “habitats,” which biodiversity, letting or getting more plant species contain hundreds of species. Rarely is a single species to grow, and with them possibly a more diverse ar- managed in isolation from the biological community ray of vertebrates, arthropods, etc. But the loss of of which it is a part (although certain exceptions come a few species that are already scarce when the old to mind, e.g., California condor, Gymnogyps califor- forest is cut down could reduce the total number of nianus). We tend to count the number of species that species existing on a more global scale. Indeed, the are threatened or endangered, and we discuss the ef- newly replanted, more-biodiverse landscape could fect of certain land use practices on particular species. be populated with species that are already abundant However, we are really making decisions about how to

Overview 5 change vegetative communities, or habitats, with our produced in a labor-intensive way for example, if agricultural practices, and this has wholesale effects relatively more labor is used per unit of capital or on many species in addition to those that are the im- other inputs, compared to another good produced mediate topic of discussion. We are drawn, therefore, with the same inputs. Intensifi cation within this to making informed decisions about entire landscapes, frame of reference always needs to specify which which contain numerous different habitats. A detailed input is being increased or intensifi ed, relative to explication of the concept of habitat and analytical ap- others, to raise production. proaches to its investigation can be found in Morrison b) In the agricultural sector, the land input is often of et al. (1998). central interest, and production is considered either There is a tension between two views of biodiversity, more intensive (or more extensive) in terms of the as having instrumental vs. ultimate value. The fi rst ratio of non-land inputs relative to land. “Intensive” attributes importance to biodiversity for the multiple production systems concentrate larger quantities of benefi ts and services that it can produce, while the inputs, particularly labor, on a given amount of land. second assigns the existence of biodiversity some incal- “Extensive” production systems, such as mechanized culable, transcendental worth. Persons who take the fi rst agriculture, have usually larger scales of production, perspective are prepared to assign economic values to applying lower amounts of non-land resources per biodiversity and consider tradeoffs against other valued hectare. Even though greater inputs of capital (and outcomes, while persons with the second perspective fuel) are employed, the labor per unit of land is re- tend to resist or discount such assessments. duced and often so are other resources. There is a tenable middle position because even with c) A third use of the term refers to the increased use the fi rst view—that biodiversity is important for the of purchased or external inputs, substituting me- ecological processes and functions it preserves which chanical or chemical inputs for labor. This involves are valuable to people—it is clear that we cannot say expending more capital for equipment and on fuel with any confi dence just which species, or how many, for tractor operations as well as on fertilizer (instead are needed to maintain particular functions (Lawton of compost), and chemical sprays (instead of hand 1999). This gives weight to the second view—which weeding or crop protection). Not using such “mod- accords signifi cant but unknown worth to preserving ern” inputs typically makes agricultural operations biodiversity—even though it does not negate the basis more labor-intensive (in the sense of the fi rst mean- for thinking instrumentally about biodiversity. We have ing discussed). The term “intensifi cation” should not taken either extreme position because neither seems always be used with appropriate specifi cation of what tenable. Instead, we work with both conceptions and resource is the reference point. External-input-inten- valuations, attributing instrumental and intrinsic value sive agriculture is very different from labor-intensive to biodiversity even if their combination is often an production, having different implications for soil and uneasy one. for other kinds of biodiversity. As this third meaning of the term implies, it is not the fact of intensity but 1.6.2. Intensification rather the kind of intensity that is important. Intensifi cation is multi-dimensional and needs to be Economists, when talking about intensifi cation, most identifi ed with regard to specifi c contexts. While the often are referring to meaning (a), while environmental- term typically refers to the amount of inputs devoted ists often are focusing on (b). They may debate about to production, several different frameworks exist from (c). Intensifi cation, in particular in (a) and (b), also which intensifi cation is assessed. relates to the economic concept of total factor produc- a) The most common reference is to the relative amount tivity, which represents the total output that results from of one input e.g., labor, capital, or land—used in a the application of all productive inputs combined. Use particular production process compared to the use of this measure surmounts the well-known limitations of all other inputs. We say that a good or service is in partial productivity measures (labor productivity,

6 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations land productivity, etc.) that are associated with mean- and social deprivation and poverty. While for scientifi c ing (a). However, partial productivity measures are analysis, references to degradation may require some still often used because they give insights into relative explicit biophysical defi nition, when thinking about this factor usage. process and its outcomes, we need to keep the economic and social concomitants and consequences in mind. 1.6.3 Degradation REFERENCES Degradation is another term that has several meanings. From a strict conservation biology point of view, any Gilbert, L. E. 1980. Food web organization and the change in pristine ecosystems represents degradation. conservation of neotropical diversity. In: M. E. Soulé Economists, on the other hand, often think of degrada- and B. A. Wilcox, eds. Conservation Biology: An tion in terms of the decreased capacity of the environ- Evolutionary-Ecological Perspective. Sunderland, ment to meet whatever demands are placed on it. These MA: Sinauer Associates. demands typically stem from two sources: (1) the use of Lawton, J. H. 1999. What to conserve – species or resources in the production of direct economic benefi ts ecosystems? In: Maltby et al., eds., Ecosystem Man- (agriculture, forestry, fi sheries, etc.); and (2) the use agement: Questions for Science and Society, 55-62. of resources for intrinsic welfare-enhancing benefi ts Egham, UK: Royal Holloway Institute for Environ- (wilderness recreation, landscape enhancement, etc.). mental Research, University of London. Soil erosion, for example, corresponds to the fi rst type of demand because it diminishes agricultural productiv- Leakey, R. R. B. 2004. Domestication and marketing ity in direct and measurable ways; the degradation of of novel crops for ecoagriculture. Paper prepared “viewsheds” corresponds to the second type of demand, for International Ecoagriculture Conference and and is typically harder to measure and value, often Practitioners’ Fair, 27 September-1 October, 2004, requiring non-market valuation methods. We consider Nairobi, Kenya. degradation here to be closer to the “economic” view, McNeely, J. A. and S. J. Scherr. 2003. Ecoagriculture: recognizing that pristine ecosystems are unlikely to Strategies to Feed the World and Save Wild Biodi- represent an appropriate counterfactual, especially in versity. Washington, Island Press. agriculture-oriented ecosystems. However, there is a tenable middle position that sees natural ecosystems as Morrison, M. L., B. G. Marcot, and R. W. Mannan. having a multi-faceted productivity that can be com- 1998. Wildlife-Habitat Relationships: Concepts and promised and reduced by many and varied resource Applications, 2nd ed. Madison, WI: University of management practices. Such reductions in any relevant Wisconsin Press. aspect of this productivity, even in an agroecosystem, are what we consider degradation, and may well reduce a system’s sustainability. It is also important to recognize that environmental degradation may be understood in dynamic terms, which is why the degradation of natural resources needs to fi gure more prominently in agricultural planning and evaluation. It is not just the interactions among soil systems, hydrological cycles, vegetative cover, biodiversity, and microclimates that make adverse changes in any one of these signifi cant, because others are then also adversely affected. In fact, biophysical degradation has additional externalities with feedback and interactive effects vis-a-vis malnutrition and hunger, income risks and losses, poor health and declining livelihoods,

Overview 7

ANNEX 1-A: Ecoagriculture Assessment Advisors and Recommendations

Advisor and Issues/opportunities that the assessment should address (ranked) Priority for advancing ecoagriculture science Organization over next five years 1 2 3 Lee R. DeHaan Development of new Do the necessary science Address Perennial seed crop breeding programs must be Plant Breeder plant materials to explore potentials of scientifically the role initiated for use in all major agricultural regions.

The Land Institute, (especially perennials). opportunity #1. of biodiversity in Salina, Kansas agriculture. USA Sandra Gagnon Exploring Linking social aspects to Summarize the existing knowledge and regroup or Agrobiodiversity interconnected habitat conservation, and categorize the different systems (organic, consultant networks for indigenous knowledge ecoagriculture, ecofriendly agriculture, agroforestry,

International Union biodiversity protection. related to incorporation of FAO conservation agriculture…). Many initiatives for the Conservation innovative technique. and philosophies exist that are similar, of Nature and Natural interconnected and in some aspects, redundant. Resources (IUCN), Gland, Switzerland Joanne Gaskell How to balance the Tradeoffs of opportunity Maximizing the Global collaboration and widecast ambitious Researcher provision of food with #1 within cultivated areas services provided thinking like the Millenium Assessment.

International Food need for ecosystem as well as between within cultivated Policy Research services. cultivated and other areas. systems, recognizing Institute (IFPRI), the critical role of Washington, DC food production.

Barbara Gemmill A new model of To better determine what Executive Director intensively-farmed aspects and how

Environment Liaison agriculture is needed, biodiversity can Centre International one that supports contribute to agricultural Nairobi, Kenya biodiversity on-farm to productivity, we need a promote ecosystem much better assessment of services. either crop loss or contributions to crop yields (and this over time, with a strong sustainability aspect).

Annex 1A- 1

Advisor and Issues/opportunities that the assessment should address (ranked) Priority for advancing ecoagriculture science Organization over next five years 1 2 3 Roger Leakey Determine the role of Development of tree Linkages between Developing the integration of production agriculture Professor of biodiversity (planned domestication and agricultural and with wildlife conservation and business development Agroecology and unplanned) in commercialization of commercial for the diversification of agriculture. Our Network

Agroforestry and agroecosystem agroforestry tree products diversification. for Sustainable and Diversified Agriculture may be a Novel Crops Unit, function. as incentive for farmers bit of a model for this. School of Tropical to diversify with Biology, James Cook traditional/ indigenous University, Australia trees. Jules Pretty Getting national Developing better Understanding the More hard evidence on the costs vs. benefits of Professor policies right, scientific understanding roles of social capital ecoagriculture.

Center for removing implicit of the applicability of for promoting Environment and discrimination against ecoagriculture. collective action at Society, and Dept. of sustainable approaches. landscape level. Biological Sciences, University of Essex, UK Thomas P. Tomich What meso-scale Which is better: What policies and Principal Economist ecological services are segregation or integration institutions and Global threatened by of production and influential to land- Coordinator biodiversity reduction? conservation? use change?

Alternatives to Slash- and-Burn Programme (ASB), Nairobi, Kenya

Annex 1A- 2

ANNEX 1-B: Cornell Ecoagriculture Working Group

Name and Title Contact Research Interests

Christopher Barrett [email protected] Poverty-environment Associate Professor, http://www.aem.cornell.edu/profiles/barrett.html linkages; microeconomics Applied Economics and of agricultural development Management David Bates [email protected] Ethnobotany Professor, Plant Biology http://www.plantbio.cornell.edu/people.php?netID=dmb15 W. Ronnie Coffman [email protected] Plant breeding related to Professor and Chair, Plant Breeding; http://www.plbr.cornell.edu/PBBweb/Coffman.html international agriculture Director, International Programs/CALS Paul Curtis [email protected] Wildlife management, pest Assistant Professor, control in agricultural Natural Resources systems

Stephen Degloria [email protected] Geographic information Professor and Chair, www.css.cornell.edu/faculty/degloria.html systems; spatial planning Crop and Soil Sciences and analysis

Laurie Drinkwater [email protected] Ecology of managed Assistant Professor, Horticulture http://www.hort.cornell.edu/department/faculty/drinkwater/index.html ecosystems, nutrient cycling

John Duxbury [email protected] Nutrient transformations; Professor, http://www.css.cornell.edu/faculty/duxbury.html agronomic production Crop and Soil Sciences Lucy Fisher [email protected] Information coordination Extension Associate, and management CIIFAD/Intl. Programs

Theresa Fulton [email protected] Genomics and Director of Outreach, http://www.igd.cornell.edu/staff/Fulton.html bioinformatics in Institute for Genomic Diversity, biodiversity conservation Biotechnology Center and food security

Annex 1B-1

Name and Title Contact Research Interests John Gaunt [email protected] ??? need to put in Senior Research Associate, [email protected] something here Crop and Soil Sciences

Charles Geisler [email protected] Tree tenure, forest tenancy, Associate Professor, http://www.cals.cornell.edu/dept/devsoc/facultyprofile.cfm?FacultyID=42 land tenure and common Rural Sociology property systems

Andre Goncalves [email protected] Agronomy, agroforestry, Graduate Student, ecoagriculture Crop and Soil Sciences

Robert Herdt [email protected] Economics Adjunct Professor, Applied Economics and Management Peter Hobbs [email protected] Conservation agriculture Visiting Professor, http://www.people.cornell.edu/pages/ph14/ Crop and Soil Sciences

Quirine Ketterings [email protected] Nutrient management Assistant Professor, http://www.css.cornell.edu/faculty/ketterings.html Crop and Soil Sciences

Johannes Lehman [email protected] Soil fertility management Assistant, Professor, http://www.css.cornell.edu/faculty/lehmann/index.htm Crop and Soil Sciences

Carl Leopold [email protected] Plant development and seed Professor Emeritus, physiology Boyce Thompson Institute Kenneth Mudge [email protected] Multi-purpose tree Associate Professor, Horticulture http://www.hort.cornell.edu/department/faculty/mudge/index.html propagation and symbiosis Rebecca Nelson [email protected] Plant diseases Associate Professor, http://www.plbr.cornell.edu/PBBweb/nelson.html Plant Pathology

Annex 1B-2

Name and Title Contact Research Interests Alison Power [email protected] Agroecology, insect- Professor, Ecology and Evolutionary http://www.eeb.cornell.edu/power/power.html transmitted diseases of wild Biology and cultivated plants

Susan Riha [email protected] Root, soil, and water Associate Professor, Earth http://www.geo.cornell.edu/eas/index.html relations in agroforestry and and Atmospheric Sciences cropping systems

Tammo Steenhuis [email protected] Water and nutrient transport Professor, Agricultural and http://www.bee.cornell.edu/faculty/faculty-bio.tss1.htm in agroforestry systems Biological Engineering

Margaret Smith [email protected] Crop breeding and genetics Professor, Plant Breeding http://www.plbr.cornell.edu/PBBweb/Smith.html Annex 1-B continued, 4 of 4 pages Janice Thies [email protected] Soil microbial ecology Associate Professor, http://www.css.cornell.edu/faculty/thies/Thies.html Crop and Soil Sciences

David Thurston [email protected] Disease management, Professor Emeritus, http://ppathw3.cals.cornell.edu/ppath/facultyinfo/Thurston.html mulchbased systems Plant Pathology

Terry Tucker [email protected] International extension, Assistant Director, http://ip.cals.cornell.edu/profiles/show_profile.cfm?personid=183 adult education IP/CALS and CIIFAD

Harold van Es [email protected] Soil and water management Professor, http://www.css.cornell.edu/faculty/hmv1/s&wman/harbio.htm Crop and Soil Sciences

David Wolfe [email protected] Environmental physiology, Professor, Horticulture http://www.hort.cornell.edu/department/faculty/wolfe/index.html soil and water management

Annex 1B-3 ANNEX 1C: External Organizations Consulted

Organization Ecoagriculture-related interests Conservation International (CI) Criticisms of ecoagriculture approach; defense of traditional separation of agriculture and conservation; certification programs; integrating conservation and agriculture in business; landscape-level planning with soy-growing farmers to maximize protection of habitat in set-asides. Earth Institute, Columbia University (EI) Soil quality; below-ground biodiversity; linking biodiversity at different levels; conservation and development problems in Africa Food and Agriculture Organization Proposed activities for sustainable land use (FAO) Foundations of Success (FOS) Monitoring and evaluation tools; determining means and goals for measuring biodiversity, and determining the success of projects International Food Policy Research Millennium Assessment; global environmental and agricultural Institute (IFPRI) strategies; using GIS-data to identify win-win areas; models or approaches used to integrate multiple objectives; efforts to look at landscape level issues in plant breeding studies Northern Chiapas Coffee Network Organic coffee; impediments to certification Rainforest Alliance Shade coffee; certification programs

The Nature Conservancy (TNC) Evaluating bioindicators; cattle intensification; organic coffee; conservation measures; need for quantitative data on agriculture’s impact on biodiversity; certified coffee/cacao United Nations Development Program: EI efforts to measure integration of multiple goals in projects related Equator Initiative (UNDP-EI) to ecoagriculture; Equator prize program; combining rural income generation and biodiversity conservation United Nations University: People, Biodiversity conservation and agricultural systems Land-Management and Ecosystem Conservation (UNU- PLEC) U. S. Agency for International USAID involvement and guidance relevant to ecoagriculture Development (USAID) U. S. Department of Agriculture: USDA-CSREES goals and projects relevant to ecoagriculture; CRIS Cooperative State Research, Education agricultural research database; environmental components of US and Extension Service (USDA- agricultural research and extension; CGE and global spatial models to CSREES) monitor and predict environmental; response to agro-economic policy changes U. S. Department of Agriculture: Valuation of ecosystem services; assessing ecological impact of the Economic Research Service (USDA- CRP (Conservation Reserve Program) ERS)

U.S. Department of Agriculture: Natural CEAP (Conservation Effects Assessment Program); NRCS effort to Resource Conservation Service (USDA- evaluate effectiveness of government programs for integrating NRCS) multiple goals; CSP (Conservation Security Program)

Annex 1C - 1 Organization Ecoagriculture-related interests U. S. Environmental Protection Agency EPA efforts and challenges with integrating multiple goals (US-EPA)

Wildlife Conservation Society European experience for conserving wildlife in agricultural systems

Winrock International Winrock efforts relevant to ecoagriculture; future research and practice; low-cost aerial imagery and relevance to ecoagriculture; terminology (in Europe, "ecoagriculture" = organic agriculture); Performance-Based Environmental Policy for Agriculture (PBEPA) program; trade-offs; problems arising from lack of below-ground and plant data World Bank ASB – Alternatives to Slash and Burn program; long-term analysis of trade-offs and linkages; WB efforts to integrate conservation and development; valuation projects; PROFOR activities; payments for ecosystem services; models for planning conservation World Resources Institute (WRI) Global analyses; nutrient trading; stakeholder involvement; analytical methodologies; nutrient management initiative; use of marginal cost curves in environmental evaluations; systems-based analyses; integrating ecosystem services; Millennium Assessment issues; integrating protected areas into larger working landscapes World Wildlife Federation (WWF) Collaborating with producers, e.g. Florida dairy cattle initiative, and Wisconsin certified “healthy” potatoes initiative

Annex 1C - 2

ANNEX 1-D: Experts Consulted

Expert Title* Interviewer Topic Albrecht, Greg Extension Associate, Cornell Nutrient WDH April 1-15, 2004 Integrated nutrient management Mgmt. Program Alvarez, Angel and Northern Chiapas Coffee Network FRW February 24, 2004 Organic coffee, impediments to certification Miguel Gonzalez Hernandez Anderson, Jon USAID – Natural Resource Policy FRW and DCB USAID involvement and guidance relevant to Ecoagriculture Advisor December 3, 2003 Auburn, Jill USDA-CSREES SARE FRW and DCB USDA-CSREES goals and projects relevant to Ecoagriculture; Director/NPL-Sust. Agric. Economic December 2, 2003 CRIS - agricultural research database; environmental & Community Systems components of US agricultural research and extension Barrett, Chris Associate Professor, Applied DCB Biocomplexity research project in Africa; livestock grazing Economics and Management February, 2004 models Bedford, Barbara Senior Research Associate, Natural WDH April 1-15, 2004 Landscape diversity; wetland processes and water Resources management Blockhus, Jill World Bank – Natural Resource DCB February 24, 2004 PROFOR activities Management Specialist Bostick, Katherine WWF – Researcher, Aquaculture and FRW and DCB Collaborating with producers: Florida dairy cattle initiative, Agric. Conservation Strategies Unit December 1, 2003 Wisconsin certified “healthy” potatoes initiative Brown, Douglas Graduate Student, Applied LEB and DCB Spatially-explicit bioeconomic modelling Ronald Economics and Management December, 2003 Chomitz, Ken World Bank – Development Research DCB February 27, 2004 Payments for ecosystem services and models for planning Group conservation Clancy, Kate Winrock – Managing Director, FRW and DCB Terminology ("Ecoag"=organic in Europe); Performance- Wallace Center for Agric. & Environ. December 4, 2003 Based Environmental Policy for Agriculture (PBEPA); Policy assessing trade-offs; lack of below-ground and plant data Coffman,William Professor, Plant Breeding. Director, LEB Biotechnology in cereal and other food crops IP/CALS April, 2004

*Acronyms, initial, and abbreviations are listed at the end of Annex 1D Annex 1D-1

Expert Title* Interviewer Topic Cox, Steve WRI – Executive Director, Global FRW and DCB Global analyses; nutrient trading; stakeholder involvement; Forest Watch December 3, 2003 analytical methodologies, Curtis, Randy The Nature Conservancy FRW and DCB Evaluating bioindicators; conservation measures December 2, 2003 Darwin, Roy USDA – CSREES FRW and DCB CGE and global spatial models to monitor and predict Agricultural Economist December 2, 2003 environmental. responses to agro-economic policy changes Degloria, Stephen Chairperson, Crop and Soil Sciences FRW and DCB GIS and agricultural impact on biodiversity December 17, 2003 Drinkwater, Laurie Professor, Horticulture WDH April 1-15, 2004 Cover cropping; below-ground biodiversity; microbial DCB February, 2004 mediation of nutrient cycles Duxbury, John Professor, Crop and Soil Sciences DCB February, 2004 Sustainable productivity to match global needs; soil health assessments Fernandes, Erick World Bank FRW and DCB ASB – Alternatives to Slash and Burn; long-term analysis of December 1, 2003 trade-offs and linkages Fisher, Lucy Extension Associate International LEB Information systems for tropical soil biology and organic Programs February, 2004 inputs Gaskell, Joanne IFPRI – Research Assistant, DCB December 4, 2003 Millennium Assessment; global environmental and agricultural Environment and Food Production Strategies; using GIS-data to identify win-win areas Tech. Division Geisler, Charles Associate Professor, Development LEB April, 2004 Tenure systems and property rights Sociology Goncalves, Andre Graduate Student, Crop and Soil LEB March, 2004 Ecoagriculture policies and practices in Brazil Sciences Greenhalgh, Suzie WRI – Senior Economist FRW and DCB Nutrient management initiative December 3, 2003 Hansen, Leroy USDA-ERS Agricultural Economist DCB December 29, 2003 CRP and related targeting efforts Hawkes, Christine Post Doctoral Fellow, Univ. of FRW February 25, 2004 Biocomplexity and below-ground biodiversity California, Berkeley

*Acronyms, initial, and abbreviations are listed at the end of Annex 1D Annex 1D-2

Expert Title* Interviewer Topic Hellerstein, Dan USDA-ERS Agricultural Economist DCB December 29, 2003 CRP and related targeting efforts

Herdt, Robert Adjunct Professor, Applied DCB and LEB Economics references; biotechnology Economics and Management December, 2003 Hicks, Frank Rainforest Alliance, Director of FRW and DCB Shade coffee; certification programs Sustainable Agriculture February 17, 2004 Hobbs, Peter Visiting Professor, Crop and Soil LEB, DCB, WDH Conservation tillage/agriculture; soil biological matter Sciences Mar/Apr, 2004 Hoffman, Bill USDA-CSREES Program Specialist, FRW and DCB USDA-CSREES goals and projects relevant to Ecoagriculture; Plant and Animal Systems December 2, 2003 CRIS - agricultural research database; environmental components of US agricultural research and extension Hooper, Michael UNDP - Communication Officer for FRW and DCB EI efforts to measure integration of multiple goals in projects the Equator Initiative (EI) February 16, 2004 related to Ecoagriculture; Equator prize; combining rural income generation and biodiversity conservation Hyberg, Skip USDA -- Economics and Policy FRW April 2, 2004 Assessing ecological impact of CRP (Conservation Reserve Analysis Staff Program) Kadyszewski, John Winrock – Senior Program DCB (phone interview) Low-cost aerial imagery and relevance to Ecoagriculture Coordinator, Ecosystem Services January 7, 2004 Kingsland, Nick The Nature Conservancy (TNC) - FRW and DCB Evaluating bioindicators; cattle intensification; organic coffee Business and Environmental Policy December 2, 2003 Specialist Kiss, Agi World Bank – Senior Ecologist DCB February 24, 2004 WB efforts to integrate conservation and development; valuation projects Lawrence, Patty USDA-NRCS Council of FRW and DCB (email) CEAP (Conservation Effects Assessment Program) Environmental Quality, Criteria and December 25, 2003 Indicators Lehman, Johannes Assistant Professor, Crop and Soil DCB February, 2004 Soil characterization and soil health Sciences

*Acronyms, initial, and abbreviations are listed at the end of Annex 1D Annex 1D-3

Expert Title* Interviewer Topic Luz, Karen The Nature Conservancy – Central FRW and DCB Need for quantitative data on agriculture’s impact on America Division December 2, 2003 biodiversity; certified coffee/cacao

Lynch, Sarah WWF FRW and DCB Collaborating with producers: Florida dairy cattle herd December 1, 2003 management, Wisconsin certified “healthy” potatoes Mausbach, Maurice USDA-NRCS Deputy Chief, NRI – FRW and DCB (email) NRCS effort to evaluate effectiveness of government programs Resource Assessment December 25, 2003 for integrating multiple goals; CSP (Conservation Security Program) McLeod, Donald USDA-CSREES National Program FRW and DCB USDA-CSREES goals and projects relevant to Ecoagriculture; Leader, Economics and Community December 2, 2003 CRIS - agricultural research database; environmental Systems components of US agricultural research and extension Melo, Cristian Graduate Student – Natural Resources FRW March 2, 2004 Eco-certification of bananas in Ecuador Menale, Andy US EPA FRW and DCB Models and efforts to integrate multiple goals; challenges November, 2003 Mudge, Kenneth Associate Professor, Horticulture LEB February 2004 Agroforestry and tree domestication Nelson, Rebecca Associate Professor, Plant Pathology DCB March, 2004 Insect pest management and diversity

Nowierski, Robert USDA-CSREES National Program FRW and DCB USDA-CSREES goals and projects relevant to Ecoagriculture; Leader, Bio-based Pest Management December 2, 2003 CRIS - agricultural research database; environmental components of US agricultural research and extension Pagiola, Stefano World Bank DCB February 24, 2004 Payments for ecosystem services Palm, Cheryl Earth Institute – Senior Research FRW and DCB Soil quality; below-ground biodiversity; linking biodiversity at Scientist February 16, 2004 different levels; development challenges in Africa Parry, Roberta US EPA- Office of Water Quality FRW and DCB (email) Models and efforts to integrate multiple goals; challenges November 28, 2003 Pender, John IFPRI - Senior Research Fellow, FRW and DCB Models or approaches used to integrate multiple objectives Environment and Production December 4, 2003 Technology Div.

*Acronyms, initial, and abbreviations are listed at the end of Annex 1D Annex 1D-4

Expert Title* Interviewer Topic Pereira, Christy USDA – CSREES FRW and DCB USDA-CSREES goals and projects relevant to Ecoagriculture; December 5, 2003 CRIS - agricultural research database; environmental components of US agricultural research and extension Pimentel, David Professor Emeritus, Entomology and FRW February 23, 2004 Issues of soil degradation; soil biodiversity Natural Resources

Piñedo-Vasquez, UNDP (PLEC) – Associate Research DCB (email) PLEC web information Miguel Scientist and Lecturer Pinstrup-Andersen, Professor, Applied Economics and FRW and DCB Expanding organic agriculture to developing countries Per Management and Nutritional Science December 2003

Poe, Greg Assoc. Professor, Applied Economics DCB February, 2004 Environmental economics; valuation tools and Management Powers, Alison Professor, Ecology and Evolutionary FRW and DCB Biodiversity/ecosystem services and agriculture links Biology December 17, 2003 Powers, John US EPA – Economist, Office of FRW and DCB (email) EPA efforts and challenges with integrating multiple goals Water Quality November 28, 2003 Redford, Kent Wildlife Conservation Society FRW March 5, 2004 European experience for conserving wildlife in agricultural systems

Ribaudo, Marc USDA – ERS DCB December 29, 2003 Valuation of ecosystem services; CRP Rice, Dick Conservation International – Chief FRW and DCB Criticisms of Ecoagriculture approach; defense of traditional Economist December 1, 2003 separation of agriculture and conservation Riha, Susan Associate Professor, Earth and WDH April 1-15, 2004 Integrating agricultural intensification and conservation; Atmospheric Sciences environmental thresholds. Salafsky, Nick Foundations of Success FRW and DCB Monitoring and evaluation tools; determining means and goals December, 2003 for measuring biodiversity; determining success of projects Scialabba, Nadia FAO DCB (email) Proposed activities for sustainable land use February 13, 2004

*Acronyms, initial, and abbreviations are listed at the end of Annex 1D Annex 1D-5

Expert Title* Interviewer Topic Sebastian, Kate IFPRI FRW and DCB Millennium Assessment; global environmental and agricultural December 4, 2003 Strategies; using GIS-data to identify win-win areas Semroc, Bambi Conservation International FRW and DCB Certification programs; integrating conservation and December 1, 2003 agriculture in business; landscape-level planning with soy farmers to maximize protection of habitat in set-asides Smale, Melinda IFPRI/IPGRI Research Fellow DCB February 27, 2004 Efforts to look at landscape level issues in plant breeding studies

Steenhuis, Tammo Professor, Agricultural and Biological LEB and DCB Watershed systems Engineering March, 2004 Sydenstricker, John Teaching Coordinator, Rural DCB and LEB Spatially-explicit modeling in forest and agroforestry-based M. Sociology landscapes in the Amazon Thies, Janice Associate Professor, Crop and Soil DCB December 2003 Tools for biodiversity evaluation. Sciences Thurston, David Professor Emeritus, Plant Pathology LEB May, 2004 Agrobiodiversity in traditional agricultural systems Tucker, Terry Associate Director, IP/CALS and LEB May, 2004 International extension communication systems CIIFAD Van Es, Harold Professor, Crop and Soil Sciences WDH April 1-15, 2004 Soil health Warner, Katherine Winrock – Managing Director, DCB December 30, 2003 Winrock efforts relevant to Ecoagriculture; future research and Forestry and Natural Resource Mgmt. practice. Wolcott, Rob WRI FRW and DCB Evaluation of marginal cost curves; systems-based analyses; December 3, 2003 integrating economic services; Millennium assessment; integrating protected areas into larger working landscapes Wolfe, David Professor, Horticulture WDH April 1-15, 2004 Soil health; IPM; cropping system diversity Wood, Stanley IFPRI – Senior Scientist, FRW and DCB Millennium Assessment; global environmental and agricultural Environment and Production December 4, 2003 strategies; using GIS-data to identify win-win areas Technology Division

*Acronyms, initial, and abbreviations are listed at the end of Annex 1D Annex 1D-6

Acronyms, Initials, and Abbreviations

CIIFAD Cornell International Institute for Food, Agriculture and Development CRP Conservation Reserve Program FAO Food and Agriculture Organization IFPRI International Food Policy Research Institute IP/CALS International Programs, College of Agriculture and Life Sciences, Cornell University USAID United States Agency for International Development USDA–CSREES United States Department of Agriculture–Cooperative State Research, Education and Extension Service USDA–ERS USDA–Economic Research Service USDA–NRCS USDA–Natural Resources Conservation Service US EPA US Environmental Protection Agency Winrock Winrock International WRI World Resources Institute WWF World Wildlife Fund

Annex 1D-7

Chapter 2 Opportunities, Concepts, and Challenges in Ecoagriculture

The struggle to maintain [wild] biodiversity is going to be won or lost in agricultural ecosystems. Management of agricultural landscapes will be the litmus test of our ability to conserve species…Integrating human exploitation with conservation through the diversiﬁ cation of types and intensity of land-use is a realistic way of minimizing extinctions in the absence of detailed knowledge of individual species. (McIntyre et al. 1992:606).

2.1 Environmental Concerns 8.1). The growth trend appears somewhat asymptotic, with a dramatic slowdown in the addition of new pro- The history of biodiversity conservation over the past tected areas in recent years to the worldwide protected century has focused on protection of particular geo- list, both in number of accessions and in total acreage. graphic locations or sites (e.g., national wildlife refuges, However, no decline in the creation of new areas has national parks), regulation of activities on public land, probably occurred yet, as 5,000 protected areas that or on protection of particular high-profi le rare species UNEP became aware of after Groombridge and Jenkins (e.g., the Endangered Species Act of 1973). Emphasis published their data were omitted from their fi gure (M. has been placed on the subset of biodiversity known as Cordiner, pers. comm.). So it appears that the era of set- “wildlife,” i.e., vertebrates with obvious recreational or ting aside large areas as parks and reserves is not over, economic value. but the handwriting is on the wall. Even with a resump- The U.S. has provided the basic model for this strat- tion in the number of land set-asides, it is diffi cult to egy, which has since been repeated throughout the imagine that more than a small fraction of the earth’s developed and developing world in various forms. surface will ever be protected in this way. Therefore, Trefethen (1975) describes the American history in those interested in conserving biodiversity should direct eloquent narrative form, and Table 1-1 in Shaw (1985) attention to the much larger area that is and will remain sketches a 350-year history of such initiatives in the in private hands, most of which is under some form of U.S. up to 1982, which was about the time that modern agricultural management, or could be. This point is at conservation biology was born. In the past 20 years, the center of ecoagriculture thinking as proposed by additional legislation and international treaties relevant McNeely and Scherr (2003). to the conservation of biodiversity, e.g., the Convention It is remarkable that conservation biologists have not on International Trade in Endangered Species and the paid more attention to the interface between conserva- Convention on Biological Diversity, have been enacted. tion of wild biodiversity and agriculture until recently, More land has been set aside under strict protection, and given that such a large proportion of species, perhaps the role of non-governmental organizations in conserva- even most of the individuals, reside in land that is man- tion has increased signifi cantly (da Fonseca 2003). aged for agriculture (Western and Pearl 1989; Pimentel The growth of protected areas over 1,000 hectares et al. 1992). This interface is important for two prin- worldwide in IUCN categories I-VI, from 1970 to 2000, cipal reasons: agricultural lands sustain a great deal of is traced in Groombridge and Jenkins (2002:198, Figure the biodiversity in this world that is valued for many

8 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations “non-practical” reasons; and much of the biodiversity natural or anthropogenic event that might wipe out the maintained on these lands, especially micro-biodiver- population. Large agricultural operations, by contrast, sity in the soil, is an important contributor to agricultural are usually managed monoculturally, automatically productivity itself (Wall and Moore 1999; Wardle et al. confl icting with the conservation of biodiversity. 1999; Alkorta et al. 2003). It is almost as though the If neighboring farmers could be enlisted to adopt cer- fi eld of conservation biology, which was “born” about tain agricultural techniques or farming systems with a 1980, needed to hone its theoretical skills by focusing broader spatial perspective, however, it may be possible on relatively simple environments, such as large tracts to regard an entire neighborhood of farms, even small of land containing mostly natural habitat and owned by ones, as a landscape and to manage this as a more or less the government. If one steps outside such an “ecological intact ecosystem. How to get neighboring farmers to laboratory,” one is immediately confronted with large adopt and adapt agricultural practices that are compat- areas of degraded and fragmented habitat, and human ible with the conservation of biodiversity, and how to fi t populations desperately trying to eke out a living by and cumulate these operations into larger spatial units using whatever wild or domestic biodiversity exists so that more biodiversity gets conserved than would there. otherwise be possible, presents a collective challenge Fortunately, concepts and techniques in conservation for agricultural economists, agronomists, conservation biology matured during the 1990s, as indicated by the biologists, policy makers, and concerned citizens. appearance for the fi rst time of several notable texts Domesticated biodiversity, or agrobiodiversity, is im- written for classroom use. It is telling, however, that portant not just for the preservation of ecosystems as not more than a handful of pages in those texts are fundamental parts of the earth’s heritage and for the devoted to the problems of conserving wild biodiver- many environmental services that these ecosystems sity on agricultural land. By our count, out of a total provide, but also for their evident utility to farming en- of 2815 pages collectively in the texts by Fiedler and terprises globally. There are perhaps 30 million species Jain (1992), Primack (1993), Hunter (1996), Meffe and of biota on earth (e.g., Erwin 1982), and most of these Carroll (1997), and Fiedler and Kareiva (1998), only are comprised of numerous varieties (or subspecies), so 34 pages directly discuss biodiversity and agriculture, a that there are tens of millions of varieties and species little over 1 percent. There are, in contrast, many pages of wild biodiversity that are worthy of conservation. and chapters devoted to the biological principles for Domestic biodiversity is an incredibly small proportion designing nature reserves, with many other writings of all the biodiversity on earth. To accept the idea that devoted solely to this topic (e.g., Shafer 1990). it is possible and wise only to conserve domesticated McNeely and Scherr (2003) have, therefore, raised an varieties of plants and livestock on agricultural land important problem that should engage and challenge is to concede the battle for this planet’s unique living conservation biologists as well as agriculturalists. resources without offering any resistance. McNeely and How can conservation of wild biodiversity be made to Scherr (2003) have proposed some concrete steps that work within agricultural systems, given the practical can be taken to turn the current losing tide of extinction. constraints? Because most of the world’s agriculture Are countermeasures against loss of biodiversity under is practiced on operational units with small acreages, the banner of ecoagriculture feasible? Suffi cient? Cost- conserving biodiversity on such parcels sets some upper effective? Sustainable? These are questions that need limit to the body size of animals that could possibly be to be addressed with solid data and rigorous analysis. conserved in these places, because body size is directly The following chapter and chapters as well as materials related to an animal’s requirement for habitat size. in the annexes seek to provide a systematic foundation For plants, small parcels means small population size, for thinking such questions through. Ecoagriculture is which generates the attendant problems of maintaining not a single technology or system but rather, as stated in demographic and genetic viability for those popula- the introduction, a collection of practices and farming tions. For both kinds of populations, small habitat size systems, some old, some new, that have the multiple ob- means greater vulnerability and susceptibility to a single jectives of conserving and even enhancing biodiversity

Opportunities, Concepts, and Challenges in Ecoagriculture 9 at the same time that agronomic, economic and social • It built on a series of technical innovations that objectives of the agricultural sector are met. started in the 18th century, and in some cases well before, such as animal power-based farm implements 2.2 Contemporary Agriculture: and steam and internal-combustion engines to make Accomplishments and Challenges labor more productive and expand the scale of agricultural production. Nobody can know for certain what the structure, technologies and performance of the world’s agricultural • It applied advances in knowledge about soil chemis- sector will be very far into the future. The global popu- try and plant nutrition starting in the 19th century to lation is expected to grow to around 9.3 billion persons overcome soil fertility constraints by the provision by 2050, by which date it may begin stabilizing (FAO of inorganic fertilizers. 2003). Food production will have to be increased sig- • In the 20th century, the protection of crops and nifi cantly not just in order to feed a projected population animals from pests and parasites was enhanced by increase of over 50 percent but also to better meet the a succession of agrochemical products. needs of an estimated 800+ million people who are presently mal- and undernourished, and to satisfy the • Perhaps most dramatically, scientifi c breeding efforts greater demand for food that will be created by rising in the past century improved the genetic potential of incomes. The World Millennium goal of halving the crops and livestock beyond what farmers had accom- number of undernourished people by 2015, at the cur- plished over centuries. Advances in biotechnology rent rate of progress, will not be attained even within have accelerated these efforts in recent years. 50 years. For ecoagricultural methods to succeed, they • Supporting these changes were technical and orga- will have to help meet these increasing food demands nizational innovations for surface and groundwater through signifi cant gains in agricultural productivity, irrigation that overcame the limitations of insuffi cient while at the same time maintaining natural and biotic or irregular natural water supplies where these were resources and satisfactory environmental conditions. a constraint for production. There are also other goals that many would argue the agricultural sector needs to meet, ranging from basic These innovations have progressively enabled farm- livelihood generation to helping assure the continuity ers to raise both the productivity of their land, labor of traditional cultures. and capital and the absolute amounts of food and fi ber produced. The Green Revolution from the mid-1960s Modern agriculture is probably the most successful to the mid-1990s represented a convergence of these combination of technologies ever assembled to address developments. The remarkable success and methods any major human need, in this case, the food necessary of modern agriculture are summarized in Table 1 and to sustain life itself. Figure 1, with the specifi c data given in Annex 2. Be-

Table 1 Changes in Production Indicators by Decade, 1961-2001, in percent World Grain World Grain Fertilizer Pesticide Production Production p/c Use Exports Decade (mmt) (kg) (mmt) (value)

1961-71 48.3% 20.6% 135.5% 93.9% 1971-81 25.3% 4.8% 57.5% 163.7% 1981-91 14.8% -3.0% 17.4% 34.7% 1991-2001 11.1% -3.1% 2.2% 15.5%

Source: Calculated from Worldwatch Institute data archive, same as in Figure 1, data given in Annex 2.

10 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations Figure 1: World Grain and Food Production and Prices, Fertilizer Use, and Pesticide Exports, 1961-2003 (1979-81=100)

tween 1961 and 2000, global per capita food production greatest challenges to food security will continue to ex- went up almost 30 percent, despite a doubling in world ist, account for the greatest share of this increase, with population, with a 60 percent decline in the index of total demand there projected to increase by 49 percent, world food prices (FAO, USDA and other sources, cited or 557 million tons, between 1997 and 2020. Cereal by Worldwatch Institute). Global food production has production in developing countries is not expected to stayed ahead of population growth over recent decades, keep pace with such growth in demand. The net cereal however, this masks wide disparities around the world. deﬁ cit of these countries, which amounted to 103 mil- Some regions—sub-Saharan African and many transi- lion tons, or 9 percent of consumption in 1997-99, is tion economies, for example—face signiﬁ cant threats expected to rise to 265 million tons, or 14 percent of to their food security, and in other areas, the displace- consumption, by 2030. This shortfall will have to be ment of rural smallholders is a major contributor to the met by increasing imports, mostly from industrialized problems associated with rapid urbanization. country and other food exporters (Pinstrup-Andersen What does the future hold? Recent studies by IFPRI and et al. 1999). FAO have projected future food needs, both globally In considering the present and future context for ecoag- and for developing countries, out to the year 2020 and riculture, it is instructive to consider trends in agricul- beyond (Rosegrant et al., 2001). The increase in global tural and food prices. Figure 2 shows long-term price cereal demand is projected to be at least 35 percent, trends from 1960-2000.1 While real price declines are or 654 million tons. Developing countries, where the somewhat different for different products and product

Opportunities, Concepts, and Challenges in Ecoagriculture 11 Figure 2: World Market Prices for Agricultural Commodities, 1960-2000

Source: World Bank 2001a, as cited in World Agriculture: Towards 2015/2030: Summary Report (FAO) (http://www.fao.org/docrep/004/y3557e/y3557e06.htm#f)

classes, the overall trend has been for long-term real price increases) will be infl uenced by the outcome of the price declines. The World Bank’s index of real agricul- current Doha Round of international trade negotiations tural prices decreased by 47 percent between 1980 and as well as progress in regional trade agreements. 2002 (World Bank, 2004). These price declines refl ect Continuation of these real price trends will have at least the fact that world productive capacity in agriculture three important implications for the challenges facing continues to exceed total effective demand—what con- ecoagricultural approaches. First, the adoption of more sumers are willing and able to pay for—notwithstanding environmentally-friendly systems of agricultural pro- the food security problems faced by many people in duction will probably have to occur within a context developing and other nations. of declining global agricultural prices. Continued price Declining real prices refl ect in part the distorting market pressures will reduce farmers’ incentives to innovate impacts of industrialized-country agricultural protec- and aggregate growth rates of crop yields may con- tionism, which generates higher production than would tinue to decline. There will be continued pressure on otherwise occur, with resulting dampening effects on farmers in many areas of the world to consolidate and global prices. Most observers expect real commodity expand their farming operations, as they seek to attain prices to continue their decline, although some predict an economically viable scale of production while facing that this will occur at a slower rate, due to a continued reduced price-cost margins. This has specifi c implica- slowdown in crop yield increases and strong growth in tions for ecoagriculture, since to be broadly effective, demand for meat (Pinstrup-Andersen et al. 1999). The ecoagricultural strategies will have to be successful not extent to which future trade liberalization initiatives only among small-scale producers farming on less-fa- will lead to less extreme rates of price declines (or even vored lands, but among larger-scale producers farming

12 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations in high-productivity regions. Overall, then, these factors mechanisms to address them—including regulation, will create substantial economic challenges for taking private incentive-compatible strategies, and innova- up ecoagricultural practices. tive funding strategies—should also continue to favor ecoagriculture. There are many empirical and policy Second, for individual farmers to surmount declining questions that need to be examined. Given how few agricultural terms of trade they will have to achieve resources and how little attention have been devoted productivity increases through improved technologies, to these new approaches so far, their prospects cannot systems of production and management strategies. be fully evaluated with present information. These strategies must include more effective diversifi cation, particularly into value-added production, as well 2.3 Trends and Problems Bearing on as diversifi cation into off-farm and non-farm income- Current Agricultural Practices producing opportunities (which already account for a sizeable share of farm household income in many rural While trying to meet global food demand and pro- areas). To the extent that ecoagricultural strategies may vide food security, it will be necessary to address have to sacrifi ce some productivity growth in order to many social, environmental, biodiversity and related realize biodiversity preservation and other environmen- externalities facing the agricultural sector in the 21st tal benefi ts, farmers will fi nd it harder to surmount nega- century. The expansion of agriculture, after all, has tive price effects, and the adoption of ecoagricultural served as the main driver behind huge changes to the strategies will be accordingly constrained. Earth’s vegetation. At the end of the twentieth century Finally, declining prices will continue to depress private approximately one-third of the world’s area under veg- and public returns for making investments in Green etation supported domestic plants (i.e. approximately Revolution-type agriculture. Past declines in prices for 40 million square-kilometers under crops and pasture food are one, but not the only, reason why investments grasses) and between 35 percent and 40 percent of the in irrigation and drainage have declined steeply in recent Earth’s terrestrial biological production was used for years. Continued low investments in irrigation systems human needs (Bakkes and van Woerden 1997; Loh will make it more diffi cult to concentrate agricultural 2002). The impact of human development has greatest productivity increases on favored lands, and is likely on aquatic ecosystems, in part due to infrastructure to lead to pressure for food production increases from changes that have accompanied the Green Revolution less-favored, marginal lands. (dam construction, irrigation, etc.). Estimates indicate that over last 30 years freshwater biodiversity has At the same time however, there are other counteract- declined more rapidly than either terrestrial or marine ing forces that should favor the adoption and diffusion biodiversity; see Loh 2002. of ecoagriculture strategies. To the extent that these systems and strategies can capitalize on biological Assuring global food security will continue to depend potentials and processes not yet fully exploited, there to an unknown degree on achieving further gains from could be some productivity bonuses to be realized. energy-intensive, fossil-fuel-dependent agriculture. Where synergies can be identifi ed and utilized, the Modern agriculture will continue to be successful and issue becomes one more of complementarities than needed in many places. But a number of undesirable of tradeoffs. Also, in instances where the continuing externalities, diminishing returns to certain external price-cost squeeze faced by farmers is due primarily inputs, and other challenges will demand attention in to the increasing costs of external inputs (fertilizers, the years ahead. Present methods and technologies for pesticides, etc.), there may be additional economic agriculture should be expected to evolve in response to incentives to adopt ecoagriculture management strate- both foreseeable and unanticipated changes in resource gies. This has been a key impetus, for example, behind availability, governmental regulations, social needs, and the rapid diffusion of zero- and minimum-tillage prac- scientifi c knowledge (Ruttan 2002). tices in many countries. Increased recognition of the The present and future probable situations of the ag- impacts of agricultural externalities and acceptance of ricultural sector need to be looked at with an unsenti-

Opportunities, Concepts, and Challenges in Ecoagriculture 13 mental eye. This is true for current practices and any that there will be demand for an additional 120 million proposed alternatives. We review here in brief some hectares of crop land over the next 30 years, an increase emerging issues for modern agriculture as background of 12.5 percent. Its assessment of the soil, terrain and for considering ecoagricultural opportunities: whether, climate requirements for major crops suggests that as or to what extent, it offers some solutions to existing many as 2.8 billion hectares now uncultivated could and foreseeable challenges facing the sector. be suitable to varying degrees for rainfed production of arable and permanent crops. However, estimates of 2.3.1 Rising Costs of Production potentially cultivable land are often overly optimistic, While the cost of petroleum, the main raw material for and many factors make it improbable that any more most chemical fertilizers and crop protection materi- than a fraction of this land could ever be brought into als, has fl uctuated rather than rising as some critics of intensive production (Young 1999). modern agriculture had predicted, the prices of these First, most of the world’s productive agricultural land iiss agricultural inputs have risen for farmers. This is partly already being used, so additional area must be expected because governments have been withdrawing the input to have lower, and in some cases much lower inherent subsidies they established in the 1960s and 1970s. This fertility. Second, not all of this land is really available trend is due partly to fi scal constraints and partly to a because of the tradeoffs that would be involved in its commitment urged by donors to let resource allocation conversion: the loss of forest cover (on 45 percent of be governed by market mechanisms. potential cultivable land), loss of protected areas (12 Although petroleum price increases will continue to percent), and increasing demand for human habitation fl uctuate, sometimes wildly, global supplies of petro- and settlement (3 percent). Third, land degradation leum are unlikely to become exhausted. However, the resulting from current agricultural practices and their favorable price regime for agricultural inputs that fueled effects – reduced fertility, soil erosion, salinization – are the Green Revolution in the 1960s and 1970s seems leading to a loss of vast areas of arable land each year, unlikely to be restored in the decades ahead. This means and this will continue to subtract many thousands of that the real prices of many external inputs are more hectares of agricultural land each year from available likely to rise than to fall. The decline in fertilizer and supply. Fourth, net gains will be diminished by the fact chemical use refl ected in Table 1 is probably indicative that some current agricultural area, including some of of future trends, driven partly by diminishing returns, the best quality, will continue to be lost to urban expan- discussed below. The costs of water, discussed below, sion. Finally, potential cultivable land is not distributed are also likely to rise in the future in many regions as evenly; more than half of the land in question is in just water supplies dwindle. It is expected that the “price- seven tropical Latin American and sub-Saharan African cost squeeze” facing farmers throughout much of the nations. In other regions, South Asia, the Near East world will continue to be a major problem, as they are and North Africa, for example, close to 90 percent or caught between falling real output prices and increasing more of land suitable for agriculture is already being real input costs. farmed (FAO 2002). So land will become relatively and sometimes absolutely more scarce. 2.3.2 Land Area 2.3.3 Growing Water Limitations Although the overall rate of population growth is slow- ing, the world’s population will continue to expand These are likely to constrain current production technol- through the middle of this century, causing further ogies in the new century more quickly and surely than declines in the per capita availability of arable land. land scarcity. Modern agriculture has very high water The grain production area per person, which was 0.23 requirements. With current irrigation practices, it takes hectares in 1950 and 0.13 hectares in 2000, is projected 16 tons of water to produce a ton of harvested rice; for to decline to 0.08 hectares by 2050, a drop of almost 75 maize, a rainfed crop, 10 tons of water are required to percent within a century (Brown 1999). FAO suggests produce a ton of grain. The exploitation of groundwater sources for irrigated agriculture in the U.S., India, China

14 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations and other countries, which was necessary to accelerate 2.3.4 Natural Resource Degradation the Green Revolution, is now lowering water tables, by a meter or more per annum in some areas, raising costs Modern agriculture faces soil and water constraints in of production and jeopardizing the long-term prospects qualitative as well as quantitative terms. Deep plowing for millions of producers. and over-irrigation, with monocropped soil surfaces unprotected by vegetative cover, are contributing to soil Water-economizing technologies and even such erosion and salinization on a major scale. Fortunately, forgotten methods as water harvesting can alleviate many farmers in the U.S., Europe, Latin America and the squeeze that water scarcity is placing on modern South Asia have started revising their tillage practices, agriculture. But competing demands for water, from with over 72 million hectares in 2002 under various industries and urban agglomerations, will cut into the forms of “conservation tillage” that seek to protect agricultural sector’s share of freshwater supplies in soil resources by physical and biological means, as the future. The most cost-effective opportunities for discussed in the next chapter. The incentives and prac- water acquisition and distribution have already been tices of modern agriculture nevertheless continue to developed, so even if water supply can be expanded, expose millions of other hectares to ongoing threats it will be at higher cost. In 2001-2002, investment in of degradation. major irrigation and drainage schemes were less than $1 billion, compared with about $4 billion in the lat- Water quality is becoming a major concern in many ter 1990s and $11.5 billion 20 years before (constant parts of the world, and should be a concern in even price comparisons). There may be some breakthroughs larger areas, as the use of chemical fertilizers as well as in desalinization or some other new technologies, but agrochemical applications is making some groundwater this is likely only to enhance supply, not lower the cost supplies hazardous to health and even directly toxic in of water. certain areas. There are means and dosages that can minimize such harm, but they are not known widely More water-effi cient production systems will be more enough or regulated effectively. The widespread use of essential in the future. The absorption and retention of wood fuels also represents a source of environmental water in the soil, due to good soil structure associated degradation in many regions. Integrated pest man- with abundant soil life and organic matter, will prob- agement (IPM), one of the strategies associated with ably offer some of the most cost-effective approaches to ecoagriculture that has grown out of experience with mitigating water shortages for agriculture. Reliance on modern agriculture, has demonstrated that reliance on plowing and agrochemical use has diminished soil ca- chemical means can be reduced without production pacities rather than augmenting them. Better (different) losses and that there is economic value in conserving water management of soil and water resources is thus biodiversity within agriculture, discussed in the next essential if the benefi ts of crop genetic improvements chapter and Chapter 6. are to be realized and sustained, directing attention toward ecosystem maintenance rather just than single- 2.3.5 Diminishing Returns or Efficacy of Inputs species enhancement.2 The leading input to modern agriculture has been nitrogen fertilization. In the past 50 years, there has been 2 Matthew McCartney (IWMI) notes that changes in water a seven-fold increase in nitrogen applications, while use and quality are having the greatest impact on aquatic ecosystems as freshwater biodiversity has declined more rap- agricultural production has increased 2.5 times (Tillman idly over the past 30 years than either terrestrial or marine et al. 2002). If further increases in production are to be biodiversity (Loh et al., 2002). To a large extent this is attrib- achieved by applying additional fertilizer, how much utable to modern agricultural technologies, with increased will be needed to get another doubling? In the U.S. offtakes for irrigation, dam construction, and eutrophication “Cornbelt,” the amount of increased output resulting due to fertilizer applications (pers. comm..) from an additional ton of fertilizer has declined over the past 20 years from 15-20 tons of corn to 5-10 tons (P. Muir, web paper). Cassman et al. (1998) report that with

Opportunities, Concepts, and Challenges in Ecoagriculture 15 the current effi ciency of nitrogen fertilizer use in rice The marginal productivity and use of these inputs is production, if we are to achieve the 60 percent increase now also declining on a global scale, like fertilizer. It is in output that everyone expects will be needed in the fortunate that the prospects of crop protection through decades ahead, there will need to be a tripling of nitro- biotechnology investments are looking brighter now as gen fertilizer application. The article did not discuss the chemical strategy is showing declining benefi ts and how such an increase could be justifi ed economically rising costs, particularly in terms of adverse health and or acceptable environmentally. Such an increase defi es environmental externalities. In the U.S., the total use the imagination. And, of course, nitrogen fertilizers are of pesticides increased about 14 times between 1950 very energy intensive in production and contribute to and 2000, while the percent of crop loss due to insects the direct release of carbon dioxide into the atmosphere, increased from 7 percent to 13 percent (Pimentel 1997). exacerbating global greenhouse gas production. While it is true that individual farmers can get benefi ts from specifi c applications, the advantages on a wider The trends in global fertilizer use deserve some elabo- scale and over time are less clear. A “treadmill” effect ration. World fertilizer use peaked in 1989 (see Figure is observed where more applications are needed to keep 1 and Annex 2). This was partly because of fertilizer further crop losses from occurring as natural checks on price increases and declines in the prices farmers receive pests and pathogens get eliminated or suppressed by for their products, but it also refl ects diminishing mar- chemical applications. ginal productivity. Farmers have been learning to use chemical fertilizers more effi ciently in response to cost, Some of the decline in crop response to fertilizer inputs productivity and environmental concerns, and surely is attributable to declining soil organic matter, which millions of farmers can continue to get net benefi ts from is necessary for getting the most productivity out of fertilizer use. But globally, the picture is quite different inorganic nutrients. Remedying low soil organic mat- from 30 years ago, when fertilizer use appeared to be ter is a necessary but not suffi cient condition for better the quickest and surest way to raise production. How crop performance. Monocropping with heavy tillage, the incremental response ratio for fertilizer has changed the central strategy of modern agriculture, has had some for grain production on a global scale is seen in Table 2. dramatic successes in raising yield, but its profi tability Increases in production, as with IPM, are now becoming is often marginal because of input costs, and it has con- associated with reduced fertilizer use. tributed to crop losses from increased pest populations and to increased resistance to chemical controls. These The use of crop-protective agrochemicals was a kind of affect soil biota in ways that make plants more vulner- second wave for the Green Revolution as seen from Ta- able to pest and disease attacks. Increasing application ble 1, with growth in their use highest during the 1970s.

Table 2: World Grain Production, Fertilizer Use, and Grain:Fertilizer Response, 1950-2001 World Grain Increase Fertilizer Increase Incremental Production in Use in Response (mmt) Decade (mmt) Decade Ratio

1950 631 -- 14 -- -- 1961 805 +174 31 +17 10.2 : 1 1969-71 (ave.) 1,116 +311 68 +37 8.4 : 1 1979-81 (ave.) 1,442 +326 116 +48 6.8 : 1 1989-91 (ave.) 1,732 +290 140 +24 12.1 : 1 1999-01 (ave.) 1,885 +153 138 -2 [Not calculable]

Source: Worldwatch Institute, State of the World 1994, p. 185, with data from UNFAO, Interna- tional Fertilizer Industry Association, and USDA; updated with data from WI archive.

16 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations of fertilizers and agrochemicals within monocropped, ture—is the ability of the market to capture the impacts high-external-input systems has become part of the of externalities associated with environmental degra- problem rather than the ideal solution to biotic con- dation resulting from agricultural production and its straints on production. This conclusion may not be a attendant resource use. The valuation of these impacts very gratifying one because so much has been invested and of the contributions of environmental services in the chemical control strategies. But the evidence of more generally is an important methodological and counterproductive use of external inputs is mounting. policy issue that is currently receiving great attention (Pagiola et al. 2002). 2.3.6 Environmental Considerations 2.3.7 Genetic Limitations? Public and policy-makers’ concerns about environmental quality and climate change gave impetus to a Since the initial success of the Green Revolution, based series of international agreements, culminating in the on breeding semi-dwarf varieties of rice and wheat that Kyoto Accord. Momentum for curbing the emission of could utilize higher applications of chemical fertilizer greenhouse gases and making other changes in policy more effectively, the gains from additional genetic im- and behavior was lost when that accord went into limbo. provement have come mostly from breeding pest and However, pressure to avert the negative consequences disease resistance into new varieties. This adds to net of environmental change may well resume. A recent production by reducing crop losses, but the expansion high-level report to the U.S. President has linked such of production possibilities through breeding has slowed. disturbances to national security concerns, e.g., the On-station yields at the International Rice Research effects of rising ocean levels resulting from global Institute in the Philippines, for example, have stagnated warming would submerge large adjacent areas and (Pingali et al. 1997). This could be due at least in part populations. In agriculture, we have seen in recent to declines in soil health and quality after decades of years how increases in “extreme events” can disrupt monocropping and high chemical applications. In any production and threaten food security. Health hazards, case, the genetically-driven part of the Green Revolu- including endocrine disruption from certain chemical tion is not delivering gains as great as before. We saw exposures, are also becoming more of a concern (Col- in Table 1 that the most rapid production increases burn et al. 1996). came during the 1960s when there was not yet much effect from genetic improvements made by the Green Accordingly, agricultural research and development Revolution.3 will need to attend more to their impacts on natural ecosystems and on human health. Agricultural activities The Green Revolution’s gains came from increased use such as irrigated rice and cattle production that raise of inputs applied to varieties that were bred to be more the concentration of methane in the atmosphere will responsive to fertilizer, agrochemicals, and irrigation. come under more scrutiny. How chemical fertilizers As seen above, the use of these inputs is becoming contribute to nitrous oxide emissions and the build-up more expensive, less productive at the margin, less of nitrates in groundwater needs also to be considered, attractive because of negative externalities, or simply along with the economic and environmental costs of less possible because of limited new opportunities. the energy required to product them. In the future, agriculturalists in all roles and positions will need to think 3 Determining the exact area under high-yielding varieties more about reducing negative environmental impacts is diffi cult because of different criteria, but starting in 1965, and enhancing positive results like increasing carbon the area under modern varieties in Asia reached about 12 million ha by 1971. This expanded to over 40 million by sequestration. Whenever the full net benefi ts and costs 1981, to over 70 million by 1991, and to over 80 million by of agricultural practices are assessed, environmental 2001 (calculated from data in Table 34 of http://www.irri. impacts will get increasing attention. org/science/ricestat ricestat). As can be seen from Table 1, the most rapid gains in grain production came during the fi rst A continuing challenge in assessing these net benefi ts decade of the Green Revolution, a period when fi eld impact and costs—and one of direct relevance to ecoagricul- began only in mid-decade.

Opportunities, Concepts, and Challenges in Ecoagriculture 17 The long-term viability of agricultural development Various concepts and practices that can contribute to strategies based primarily on varietal improvement is what is being called “ecoagriculture” are emerging. No becoming less certain. assumption or argument is made that ecoagriculture will replace what is presently understood as modern There are substantial expectations that biotechnology agriculture. Rather, we consider in this paper how an ex- with new techniques for genetic modifi cation will rein- pansion of opportunities and diversifi cation of methods vigorate crop and animal breeding programs. Field-ap- that are becoming more evident could serve multiple plicable results hold great promise in many cases, but objectives. As discussed in the next section, we assume are only beginning to be realized. While some increase that optimization among these several objectives is the in regulation and scrutiny may be justifi ed, across-the- goal rather than maximizing any single criterion such as board attempts to close the door on genetic modifi cation yield or profi t or employment or number of species. work could foreclose opportunities that are benefi cial for the poor and/or the environment. Ecoagriculture is 2.4 Tradeoffs and Synergies with not a substitute for efforts to improve the genetic po- Ecoagriculture tentials of crops and animals. When managing plants and livestock under what are thought to be optimal Whether at local, national or global levels, much is conditions, one always wants to have the best available expected of the world’s existing agricultural and food genetic base to get the best return from other inputs. systems. We ask that they feed us and feed the world, Biotechnology has potential to contribute to biodi- that they reduce hunger and solve malnutrition, that versity conservation, environmental enhancement, they generate sustainable incomes and livelihoods for ensuring food security and reducing poverty, but there millions of households. In some countries, we ask that are offsetting concerns at all levels. The scope of bio- these systems contribute to export growth strategies. In technology to increase food production, lower food many countries, we ask that they reduce poverty and prices, and decrease hunger and malnutrition is great, contribute to economic and social equity. Finally, we but the large costs associated with biotechnological in- increasingly ask that they do all these things in ways novation and its possible risks and hazards need to be that are environmentally friendly and do not threaten reckoned with. We also need to consider how widely biodiversity. cost-intensive technologies can be made available to An obvious question is: are we asking too much? poor agriculturalists, who may be late adopters and Are any systems of agricultural production capable suffer from price declines. There is growing evidence to meeting all of these social goals simultaneously? of environmental benefi ts from such technologies as Surely some tradeoffs are necessarily involved. To what Bt cotton (see, for example, Pray et al. 2001). But extent, then, can the goals of food security, economic these benefi ts must be weighed against other environ- growth, poverty alleviation, biodiversity preservation, mental concerns such as crop-weed hybridization and and environmental enhancement be achieved synergisti- unanticipated transfer of genetic traits. Agricultural cally, or are tradeoffs unavaoidable in pursuing these biotechnology methods for overcoming limitations to objectives? conventional genetic improvement efforts are certainly In assessing ecoagriculture, one needs answers to these worth pursuing, but most agree that biotechnology is questions because the success of many ecoagricul- not a panacea, and its time frame for results is uncer- tural strategies relies on being able to achieve multiple tain. Where there are methods for raising production economic and social goals with synergistic effect. By that are lower-cost and more environmentally friendly defi nition, ecoagriculture requires systems of sustain- these should be considered. These could come within able agricultural and natural resource management that the small-farm sector or the commercial sector. In either “simultaneously enhance productivity, rural livelihoods, situation, productivity gains are crucial, particularly as ecosystem services and biodiversity” (http://www. the pressures and opportunities of global markets make ecoagriculturepartners.org/whatis.htm). One must ask: competition for low-cost production greater. what is the record regarding the simultaneous accom- plishment of these joint goals?

18 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations One need not search far to identify a large relevant Kuznets Curve (EKC). The EKC relates economic and literature. Since the 1970s, a number of international environmental outcomes in a way similar to the original conferences, white papers, and reports have addressed Kuznets curve that related income inequality with per these issues, including UNESCO’s Man and the Bio- capita income levels. It posits an “inverted-U” shape sphere Program (1987), the World Commission on relating environmental degradation on the Y-axis with Environment and Development (the Brundtland Report, income levels on the X-axis (Figure 3). At relatively low 1987), various prominent reports by international en- levels of national income, it is said that environmental vironmental organizations (e.g., IUCN/UNEP/WWF degradation (e.g., deforestation, air pollution, destruc- 1980), and the World Bank’s 1992 review of the status tion of biodiversity) will be relatively low but will of the global economy and environment. begin to increase as incomes rise, until at some higher income level, environmental degradation reaches a peak The actual empirical evidence from these and more for that country and levels off and then declines (Pan- recent sources that can demonstrate the complementar- ayotou 1995). The proposed reasons for this inverted-U ity of economic and environmental outcomes is mixed. relationship are many, including structural economic A recent volume reviewing major global initatives change ﬁ rst toward manufacturing, and then away from addressing poverty/production/environment linkages manufacturing toward cleaner service and information concluded that: industries; the development of cleaner technologies in Often…the existence of synergies appears richer, industrialized economies; increasing consumer accepted on faith, rather than concluded as demands (with increasing income) for environmental a result of careful analysis, research and goods and embodied services; enforcement of environ- observation… [T]radeoffs often, although not mental regulations that results in a cleaner environment; always, characterize the simultaneous pursuit and other factors. of development goals....Simple assertions of The EKC is an attractive, even seductive, notion be- complementarities in the realization of multiple cause it implies that societies can “have their cake goals have, in many instances, been shown to and eat it too.” If, at some level of national income, be unrealistic and overly simplistic or, at best, environmental degradation can be expected to decline to pertain to mostly long-run and aggregate with further increases in income, the challenge is to get relationships (Lee et al. 2001). beyond that threshold level as quickly as possible, as- Does the strategy of ecoagriculture hold promise of suming that economic growth will (eventually) produce addressing these multiple goals meaningfully and at a cleaner environment. That has led to various studies a large scale, or is it a strategy that is only applicable designed to identify the level of income at which such in isolated instances, under special circumstances, and with limited generalizability? Figure 3. Environmental Kuznets Curve The literature on tradeoffs versus complementarities in natural resource management is a large one and draws from many distinct sources. We have provided a detailed review of this literature elsewhere (Lee, et

al. 2001). It is possible to discern at least two distinct

Degradation l l

schools of thought that are relevant here, one focused at l

a a

t t n

the aggregate level, and the second at the micro (farm n

e e m

and household) level. m

n n

o o

r r i

At the aggregate level, much of the debate regarding i

v v n

whether the pursuit of economic and environmental n

E E objectives can be synergistic or not is incorporated in the evidence on what is referred to as the Environmental Income

Opportunities, Concepts, and Challenges in Ecoagriculture 19 an ‘environmental transition’ will occur. By implication, The same study suggested that “There do exist many im- policies to control or minimize degradation along the portant examples demonstrating the potential synergies way become unnecessary or a distraction, according to between pursuit of household food security, economic such an analysis. growth and environmental sustainability objectives through agricultural intensifi cation… [However,] un- Empirical results regarding the extent to which an in- der most circumstances, agricultural intensifi cation is verted-U relationship explains the observed patterns in necessary but not suffi cient to achieving food security, environmental pollution and degradation are decidedly poverty alleviation and environmental goals.” Thus, it mixed. In general, EKC relationships appear to be most appears that, at the micro level as well, the generaliz- consistent with the data for explaining the concentration ability of the synergy hypothesis which is relevant to of urban air pollutants, where effects are localized, pol- ecoagriculture is limited. lution measures are standardized, and the complications introduced by confounding factors are minimized. EKC However, in searching for factors and patterns that can relationships have been unsupported when trying to ex- lead to synergistic solutions like ecoagriculture, the plain measures that increase or decrease monotonically, same study identifi ed a number of conditioning factors, e.g., solid waste generation or the percent of population the outcomes of which may increase the likelihood that without access to public sanitation facilities or potable a given strategy may achieve results consistent with water. Much research has been conducted looking at hoped for synergies (see also Vosti and Reardon 1997). EKC relationships for deforestation, and these results These conditioning factors include: are quite disparate, with little empirical support overall • Population pressure which may induce intensifi ca- for the hypothesized inverted-U (Koop and Tole 1999). tion and land use changes in either a productive, In summary, at the aggregate level, empirical support sustainable fashion (via “Boserupian intensifi ca- for economic-environmental synergies as represented tion”) or in an environmentally degrading manner; by the Environmental Kuznets Curve are not conclusive or persuasive. • Technological change, particularly to the extent that it is labor-intensive and not labor-saving; At the micro level, there is a broad and growing literature that addresses the question of synergies vs. • Agroecological conditions that capitalize on bio- tradeoffs in economic and environmental relationships logical processes and interactions such as nutrient at farm and household levels. This body of research is cycling, soil water retention, or reduced costs of reviewed in Chapter 5 and will be only broadly charac- production, e.g., no-till; terized here. Much of this literature has emanated from • Labor-market conditions, especially as they infl u- extensive, multi-year and multi-site applied research ence seasonal labor demand, off-farm employment projects with the goal of analyzing the linkages and opportunities, and through these factors, help provide tradeoffs among food production, economic returns, capital for households to invest in technologies and and varying measures of environmental outcomes. A management practices that make their agriculture recent comprehensive synthesis of this literature as of more sustainable; the late 1990’s concluded that, • Infrastructure, roads and market access that help …the common assumption of inherent comple- farm households diversify production patterns, in- mentarity between agricultural intensifi cation, crease incomes, and reduce risk, increasing access economic development and environmental to input and product markets; and goals does not appear to stand up well to empirical scrutiny. Whatever synergies may • Policy, land tenure and property rights changes that exist across space and in the long run can be enable producers to make the most of market-based complicated by many factors at the micro level reforms. and in the short run (Lee et al. 2001). Depending on the joint outcomes of these (and other) factors, a specifi c strategy for agricultural intensifi ca-

20 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations tion may prove successful or unsuccessful. The “good lations. In strictly economic terms, one should apply news” is that synergistic outcomes are possible; the ”bad discounting to comparably evaluate streams of benefi t news” is that this is often diffi cult and dependent upon and costs. However, even if greater net benefi t over positive interactions (feedback) and outcomes among time are calculated for organic methods, for example, multiple criteria. farmers need to be able to cover their costs for any transition period involving short-term losses. From a For assessing whether synergistic outcomes are societal perspective, since there are positive externali- achieved or achievable, it is important to identify sev- ties from making this transition, for example, in terms eral cross-cutting criteria that arise from the assessment of groundwater quality, discounting of future benefi ts of specifi c strategies. These criteria are important to could undervalue the true future value of improved consider when assessing specifi c ecoagricultural tech- environmental and human health. nologies and systems. The fi rst requires identifying whether specifi c systems take into account all relevant Another issue relates to scale. Benefi ts and externalities externalities, i.e., benefi ts and costs accruing to indi- may differ between those experienced by small, individuals, enterprises, communities, and society beyond vidual farming operations and the aggregated operations the private farming operation. These are important for of large numbers of small producers. Aggregation and making overall assessments on behalf of society, yet scale effects are clearly demonstrated at the market they are generally ignored in the private assessments level, where the productivity gains achieved by an in- that individual decision-makers make concerning dividual producer may give rewards in the marketplace. resource use. In developing countries, it is often the But if an entire community or region shares in these, case that economic development priorities are more the increased production may move output prices down- strongly felt and weighed than broader environmental ward, mitigating the aggregate benefi ts to individuals and sustainability perspectives to which outsiders, gov- from productivity-enhancing farming practices, even ernment agencies, and NGOs, give more precedence. if they and others in society are better off as a result of Reconciling these different objectives complicates any the changes in production technology. Scale is impor- summative evaluation. tant in other ways as well. Biodiversity preservation, for example, is unlikely to be assured by more benign To the extent that externalities exist—e.g., downstream practices just at the farm scale; it is only likely to result sedimentation and water quality deterioration that if the scale of these practices is large enough to protect may accompany erosive upstream tillage and farming distinct animal and plant populations and the ecosys- practices—it may be necessary to develop innovative tems they are part of in a sustainable fashion. mechanisms for those who gain from such practices to compensate those who lose from environmentally- Scale—often discussed in terms of “scaling up”—is detrimental practices so that societally (and environ- important at political and project levels. Much of the mentally) desirable practices can be promoted. Part scholarly and action-oriented research on sustainable of this assessment may involve developing full and agricultural systems has been conducted at the farm and comprehensive accounting of environmental damages community level. But increasingly, donors, policymak- generated by private resource practices and/or altering ers and international organizations want to know if the policy mechanisms such as agricultural subsidies that results achieved at a micro level are generalizable to exacerbate the effects of these practices. regional or national levels. The interest of such organizations is typically in achieving signifi cant impact, Some externalities have an intrinsic time lag, with the and this must be demonstrated by the application and incidence of costs and/or benefi ts delayed, which leads success of technologies, livelihood and development to current resource allocation decisions being subopti- strategies beyond the individual farm, household or mal. For example, switching from chemical to organic community level. This necessitates attention from the methods of fertilization may give some loss of yield onset of research and development projects to the scal- for three to fi ve years while the soil with more organic ability of results and to the broader impacts of micro- matter recovers previously suppressed microbial popu- oriented applied research.

Opportunities, Concepts, and Challenges in Ecoagriculture 21 All of these elements—both the conditioning factors Brush, S.B., ed. 2000. Genes in the Field: Nn-farm and the cross-cutting elements identifi ed above—are conservation of crop diversity. Boca Ratan, FL: important to the consideration of ecoagricultural in- Lewis Publishers. novations and strategies in this report. As indicated Brussaard, L., T. W. Kuyper, W. A. M. Didden, R. G. in Chapter 1, this report is not attempting to evaluate M. de Goede, and J. Bloem. 2004. Biological soil ecoagriculture per se. Rather, we are considering what quality from biomass to biodiversity: Importance are the necessary conditions under which ecoagriculture and resilience to management stress and disturbance. can succeed? What would have to occur if the strate- In: Managing Soil Quality: Challenges in Modern gies reviewed in the following chapters are to address Agriculture, eds. P. Schjønning, S. Elmholt and B. satisfactorily an integrated set of food production, biodi- T. Christensen, 139-161. New York: CAB Interna- versity and related goals? In addressing these questions, tional. we will show the importance of conditioning factors and related concerns like scalability and externalities when Callicott, J. B. and K. Mumford. 1997. Ecological sus- evaluating the costs and benefi ts of these strategies. tainability as a conservation concept. Conservation These analytical focuses provide a framework within Biology 11: 32-40. which to identify the relevant criteria for success of Cassman, K. G., S. Peng, D. C. Olk, J. K. Ladha, W. ecoagriculture and the concerns which ecoagriculture Reichardt, A. Dobermann, and U. Singh. 1998. strategies must satisfy if they are to become widely Opportunities for increased nitrogen-use effi ciency accepted, adopted and adapted as part of the inventory from improved resource management in irrigated rice of techniques, systems and management strategies that systems. Field Crops Research 56: 7-39. farmers will employ. Colburn, T., D. Dumanoski, and J. P. Myers. 1996, Our REFERENCES Stolen Future. New York: Dutton.

Adams, G. A. and D. H. Wall. 2000. Biodiversity above Collins, W. W. and C. O. Qualset, eds. 1999. Biodi- and below the surface of soils and sediments: Link- versity in Agroecosystems. Boca Ratan, FL: CRC ages and implications for global change. Bioscience Press. 50: 1043-1048. Costanza, R., R. d’Arge, R. de Groot, S. Farber, M. Alkorta, I., I. Albizu, and C. Garbisu. 2003. Biodiversity Grasso, B. Hannon, K. Limburg, S. Naeem, R. V. and agroecosystems. Biodiversity and Conservation O’Neill, J. Paruelo, R. G. Raskin, P. Sutton, and 12: 2521-2522. M. van den Beltet. 1997. The value of the world’s ecosystem services and natural capital. Nature 387: Altieri, M. A. 2004. Linking ecologists and traditional 253-260. farmers in the search for sustainable agriculture. Frontiers in Ecology and the Environment 22:: 335-5- da Fonseca, G. A. B. 2003. Conservation science and 42. NGOs. Conservation Biology 17: 345-347. Bakkes, A.J. and van Woerden, J.J. 1997. The future of Daily, G. C. and K. Ellison. 2002. The New Economy of the Global Environment – UNEP. Nature: The Quest to Make Conservation Proﬁ table. Washington, DC: Island Press/Shearwater Books. Borlaug, N. and C. J. Dowswell. 1995. The importance of agriculture and a stable adequate food supply to Daily, G. C., J. S. Reichert and J. P. Myers, eds. 1997. the well-being of humankind. In: Research in Domes- Nature’s Services: Societal Dependence on Natural tic and International Agribuisness Management, ed. Ecosystems. Washington, DC: Island Press. R. A. Goldberg, 1-31. Greenwich, CT: JAI Press. Edwards, P. J., J. Kollmann, and D. Wood. 1999. De- Brown, L. 1999. Feeding nine billion. In: State of the terminants of agrobiodiversity in the agricultural World 1999. New York: Oxford University Press, in landscape. In: Agrobiodiversity: Characterization, association with Worldwatch Institute. Utilization and Management, eds. D. Wood and J.

22 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations M. Lenné, 183-210. Wallingford, UK: CABI Pub- IUCN/UNEP/WWF. 1980 World Conservation Strat- lishing. egy: Living Resource Conservation for Sustainable Development. Gland, Switzerland: World Conser- Erwin, T. L. 1982. Tropical forests: Their richness in vation Union, United Nations Environment Pro- Coleoptera and other arthropod species. Coleopter- gramme, and World Wildlife Fund. ists Bulletin 36: 74-75. Jackson, D. L. and J. L. Jackson, eds. 2002. The Farm FAO. 2002. World Agriculture: Towards 2015/2030, as Natural Habitat: Reconnecting Food Systems with Summary report. http://www.fao.org/ docrep/ 004/ Ecosystems. Washington, DC: Island Press. y3557e/y3557e00.htm For Tables: http://www.fao. org/docrep/004/y3557e/ y3557e04.htm#TopOfPage Kaihura, F. and M. Stocking, eds. 2003. Agricultural (viewed 04/20/04) Biodiversity in Smallholder Farms in East Africa. Tokyo: U.N. University Press. FAO. 2003. Biodiversity and the Ecosystem Approach in Agriculture, Forestry and Fisheries. FAO Inter- Koop, G. and L. Tole. 1999. Is there an environmental departmental Working Group on Biological Diversity Kuznets Curve for deforestation?, Journal of Devel- for Food and Agriculture, Rome. opment Economics 58: 231-244. Fiedler, P. L. and S.K. Jain, eds. 1992. Conservation Koyama, Osamu. 1998. Projecting the Future World Biology: The Theory and Practice of Nature Conser- Food Situation. http://ss.jircas.affrc.go.jp/ kanko/ vation, Preservation and Management. New York: newsletter/nl1998/no.15/04koyamc.htm (viewed Kluwer Academic Publishers. 04/20/04) Fiedler, P. L. and P. M. Kareiva, eds. 1998. Conserva- Kreuss, A. and T. Tscharntke. 1994. Habitat fragmen- tion Biology for the Coming Decade. New York: tation, species loss, and biological control. Science Chapman and Hall. 264: 1581-1584. Folgarait, P. J. 1998. Ant biodiversity and its relation- Lee, D. R., and C. B. Barrett, eds. 2001. Tradeoffs or ship to ecosystem functioning: A review. Biodiversity Synergies? Agricultural Intensifi cation, Economic and Conservation 7: 1221-1244. Development and the Environment. Wallingford, UK: CAB International. Groombridge, B. and M. D. Jenkins. 2002. World Atlas of Biodiversity: Earth’s Living Resources in the 21st Lee, D. R., P. J. Ferraro and C. B. Barrett. 2001. In- Century. Published with the Natural History Museum troduction: Changing perspectives on agricultural of London, IUCN, WCI, UNEDP, WWF and WRI. intensifi cation, economic development and the en- New York: Chapman and Hall. vironment. In: Tradeoffs or Synergies? Agricultural Intensifi cation, Economic Development and the En- Hansson, L., L. Fahrig and G. Merriam, eds. 1995. vironment, eds. D. R. Lee and C. B. Barrett, 1-16. Mosaic Landscapes and Ecological Processes. New Wallingford, UK: CAB International. York: Chapman and Hall. Lennon, J. J., P. Koleff, J. J. D. Greenwood, and K. J. Hooper, D. U., D. E. Bignell, V. K. Brown, L. Brus- Gaston. 2004. Contribution of rarity and common- saard, J. M. Dangerfi eld, D. H. Wall, D. A. Wardle, ness to patterns of species richness. Ecology Letters D. C. Coleman, K. E. Giller, P. Lavelle, W. H. Van 7: 81-87. der Putten, P. C. de Ruiter, J. Rusek, W. L. Silver, J. M. Tiedje and V. Wolters. 2000. Interactions be- Letourneau, D. K. 1997. Plant-arthropod interactions tween aboveground and belowground biodiversity in agroecosystems. In: Ecology in agriculture. ed. in terrestrial ecosystems: Patterns, mechanisms, and L. E. Jackson, 239-290. San Diego, CA: Academic feedbacks. Bioscience 50: 1049-1061. Press. Hunter, M. L. J. 1996. Fundamentals of Conservation Loh, J. (ed.) 2002. Living Planet Report. Gland, Swit- Biology. Cambridge, UK: Blackwell Science. zerland: World Wildlife Fund. http://resolver.library. cornell.edu/misc/4666626

Opportunities, Concepts, and Challenges in Ecoagriculture 23 Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Pingali, P. L., M. Hossain and R. V. Gerpacio. 1997. Grime, A. Hector, D. U. Hooper, M. A. Huston, D. Asian Rice Bowls: The Returning Crisis? Walling- Raffaelli, B. Schmid, D. Tilman, and D. A. Wardle. ford, UK: CAB International, in association with the 2001. Biodiversity and ecosystem functioning: Cur- International Rice Research Institute. rent knowledge and future challenges. Science 294: Pinstrup-Andersen, P., R. Pandya-Lorch, and M. W. 804-808. Rosegrant. 1999. World Food Prospects: Critical Is- Luck, G. W., G. C. Daily, and P. R. Ehrlich. 2003. sues for the Early Twenty-First Century. Washington Population diversity and ecosystem services. Trends DC: International Food Policy Research Institute. in Ecology & Evolution 18: 331-336. http://www.ifpri.org/pubs/fps/fps29.htm (viewed 04/20/04) McIntyre, S., G. W. Barrett, R. L. Kitching, and H. F. Recher. 1992. Species triage: Seeing beyond wound- Pray, C., D. Ma, J. Huang and F. Qiao, “Impact of Bt ed rhinos. Conservation Biology 6: 604-607. Cotton in China,” World Development 29(2001): 813-825. Meffe, G. K. and C. R. Carroll. 1997. Principles of Conservation Biology. Sunderland, MA: Sinauer Primack, R. B. 1993. Essentials of Conservation Biol- Associates. ogy. Sunderland, MA: Sinauer Associates. O’Riordan, T. and S. Stoll-Kleemann, eds. 2002. Bio- Rosegrant, M. W., M. S. Paisner, S. Meijer, and J. Witcov- diversity, Sustainability and Human Communities: er. 2001. Global Food Projections to 2020: Emerging Protecting beyond the Protected. Cambridge, UK: Trends and Alternative Futures. Washington DC: Inter- Cambridge University Press. national Food Policy Research Institute. http://www. ifpri.org/pubs/books/gfp/gfp.pdf (viewed 04/20/04) Ostfeld, R. S. and K. LoGiudice. 2003. Community disassembly, biodiversity loss, and the erosion of an ecosystem service. Ecology 84: 1421-1427. Ruttan. V. W. 2002. Productivity growth in world agriculture: Sources and constraints. Journal of Pagiola, S., J. Bishop, and N. Landell-Mills. 2002. Sell- Economic Perspectives 16: 161-184. ing Forest Environmental Services: Market-based Mechanisms for Conservation and Development. Shafer, C. L. 1990. Nature Reserves: Island Theory and London: Earthscan. Conservation Practice. Washington, DC: Smithson- ian Institution Press. Paoletti, M. G. and D. Pimentel, eds. 1992. Biotic Di- versity in Agroecosystems. Amsterdam: Elsevier. Shaw, J. H. 1985. Introduction to Wildlife Management. New York: McGraw-Hill. Panayoutou, T. 1995. Environmental degradation at different stages of economic development. In: Beyond Soule, J. D. 2002. Conservation and agriculture as Rio: The Environmental Crisis and Sustainable Live- neighbors. In: The Farm as Natural Habitat: Recon- lihoods in the Third World, eds. I. Ahmed and J.A. necting Food Systems with Ecosystems, eds. D. L. Doeleman, 13-36. New York: St. Martin’s Press. Jackson and J. L. Jackson, 169-188. Washington, DC: Island Press. Pimentel, D. 1997. Techniques for Reducing Pesticides: Environmental and Economic Beneﬁ ts. Chichester, Srivastava, J. P., N. J. H. Smith, and D. A. Forno. 1999. UK: John Wiley. Integrating Biodiversity in Agricultural Intensiﬁ cation: Toward Sound Practices. Washington, DC: Pimentel, D., U. Stachow, D. Takacs, H. Brubaker, A. World Bank. Dumas, M. J. O’Neil, D. Onsi, and D. Corzilius. 1992. Conserving Biological Diversity in Agricul- Thrupp, L. A. 1998. Cultivating Diversity: Agrobiodi- tural/Forestry Systems. BioScience 42: 354-362. versity and Food Security. Washington, DC: World Resources Institute.

24 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Tillman. D., K. G. Cassman, P. A Matson, R. Naylor, World Bank. 2004. Global Economic Prospects 2004: and S. Polasky. 2002. Agricultural sustainability Realizing the Development Promise of the Doha and intensive production practices. Nature 418: Agenda. http://www.worldbank.org/prospects/ 671-677. gep2004/appendix2.pdf (viewed 04/20/04) Toet, S., L. Huibers, R. S. P. Van Logtestijn, and J. T. A. Young, A. 1999. Is there really spare land? A critique of Verhoeven. 2003. Denitrifi cation in the periphyton estimates of available cultivable land in developing associated with plant shoots and in the sediment of countries. Environment, Development and Sustain- a wetland system supplied with sewage treatment ability 1: 3-18. plant effl uent. Hydrobiologia 501: 29-44. Trefethen, J. B. 1975. An American Crusade for Wild- life. New York: Winchester Press and the Boone and Crockett Club. UNESCO. 1987. A Practical Guide to MAB (Man and the Biosphere). Paris: U.N. Educational Scientifi c and Cultural Organization. Vosti, S.A. and T. Reardon, eds. 1997. Sustainability, Growth and Poverty Alleviation: a Policy and Agro- ecological Perspective, Baltimore: Johns Hopkins University Press. Wall, D. H. and J. C. Moore. 1999. Interactions underground. Bioscience 49: 109-117. Wardle, D. A., K. E. Giller, and G. M. Barker. 1999. The regulation and functional signifi cance of soil biodiversity in agroecosystems. In: Agrobiodiversity: Characterization, Utilization and Management. eds. D. Wood and J. M. Lenné, 87-121. New York: CABI Publishing. WCED. 1987. Our Common Future. New York: Ox- ford University Press, for the World Commission on Environment and Development. Western, D. and M. Pearl, eds. 1989. Conservation for the Twenty-First Century. Oxford, UK: Oxford University Press. Wilcox, B. A. and D. D. Murphy. 1985. Conservation strategy: The effects of fragmentation on extinction. American Naturalist 125: 879-887. Wood, D. and J. M. Lenné, eds. 1999. Agrobiodiversity: Characterization, Utilization and Management. Wallingford, UK: CAB International. World Bank. 1992. World Development Report 1992: Development and the Environment. New York: Ox- ford University Press.

Opportunities, Concepts, and Challenges in Ecoagriculture 25 ANNEX 2: Time Series Data on World Grain Production total and per capita, World Food Production per capita, Food Prices, Fertilizer Use, and Pesticide Exports, 1961-2003, with comparisons indexed to time series midpoint, 1979-81 average

World World Food World Food Fertilizer Pesticide Year Index World Grain p/c Index Index Index Index Index Grain Prices Production p/c Use Exports (mmt) (kg) (1960=100) (1979-81= 100) (mmt) ($mill)

1961 805 55.8 261 80.6 100.1 112.7 84.1 89.5 31 27 1.32 18 1962 858 59.5 273 84.3 100.5 113.2 84.9 90.3 34 30 1.51 21 1963 867 60.6 270 83.3 107.6 121.2 85.3 90.7 38 33 1.64 22 1964 914 63.4 279 86.1 109.2 123.0 86.3 91.8 42 37 1.81 25 1965 914 63.4 273 84.3 105.4 118.7 85.7 91.2 47 40 1.63 22 1966 992 68.8 290 89.5 108.1 121.7 87.6 93.2 52 45 1.95 27 1967 1032 71.6 296 91.4 107.1 120.6 89.1 94.8 56 48 2.04 28 1968 1065 73.9 299 92.3 103.0 116.0 89.7 95.4 60 52 2.14 29 1969 1068 74.1 296 91.4 99.6 112.2 88.1 93.7 63 55 2.30 31 1970 1087 75.4 293 90.4 98.5 110.9 89.0 94.7 69 60 2.42 33 1971 1194 82.8 315 97.2 96.7 108.9 90.0 95.7 73 64 2.56 35 1972 1156 80.2 300 92.6 95.7 107.8 87.4 93.0 79 69 2.80 38 1973 1246 86.4 316 97.5 148.8 169.0 90.5 96.3 85 74 3.67 50 1974 1216 84.3 303 93.5 151.2 170.3 90.1 95.6 82 71 5.05 69 1975 1241 86.1 303 93.5 109.2 123.0 90.8 96.6 91 79 5.60 76 1976 1348 93.5 324 100.0 101.0 113.7 91.8 97.7 95 83 5.01 68 1977 1333 92.4 315 97.2 89.1 100.3 91.9 97.8 101 88 5.70 78 1978 1454 100.8 338 104.3 87.9 99.0 94.7 100.7 109 95 6.76 92 1979 1413 98.0 323 99.7 90.5 101.9 94.2 100.2 112 98 7.36 100 1980 1418 98.3 319 98.4 89.5 100.8 93.2 99.1 117 102 7.88 108 1981 1496 103.7 330 101.9 86.3 97.2 94.7 100.7 115 100 6.75 92 1982 1552 107.6 337 104.0 74.4 83.8 96.2 102.3 115 100 6.39 67 1983 1478 102.5 315 97.2 82.2 92.6 95.0 101.1 126 110 6.59 90 1984 1632 113.2 342 105.6 83.3 93.8 98.1 104.6 131 114 6.89 94 1985 1665 115.5 343 105.9 69.0 77.7 98.5 104.8 129 112 6.74 92 1986 1678 116.4 340 104.9 51.4 57.9 99.1 105.4 133 116 7.35 100 1987 1618 112.2 322 99.4 49.1 55.3 98.0 104.3 139 121 8.17 111 1988 1565 108.5 307 94.8 58.9 66.3 97.9 104.1 145 126 8.69 119 1989 1700 117.9 328 101.2 60.3 67.9 99.9 106.3 143 125 9.14 125 1990 1779 123.4 337 104.0 51.7 58.2 100.8 107.2 138 120 9.51 130 1991 1717 119.1 320 98.8 50.1 56.4 99.4 105.7 135 118 9.09 124 1992 1797 124.6 330 101.9 49.1 55.3 100.6 107.0 125 109 8.96 122 1993 1727 119.8 312 96.3 48.6 54.7 100.0 106.4 120 105 9.02 123 1994 1777 123.2 317 97.8 49.3 55.5 101.6 108.1 112 106 10.39 142 1995 1715 118.9 301 92.9 49.4 55.6 102.3 108.8 130 113 11.92 163 1996 1882 130.5 326 100.6 58.1 65.4 105.1 111.8 135 118 12.61 172 1997 1902 131.9 326 100.6 54.5 61.4 106.3 113.1 137 119 11.68 159 1998 1891 131.1 319 98.5 48.5 54.6 106.8 113.6 138 120 12.27 167 1999 1882 130.5 314 96.9 42.1 47.4 108.3 115.2 140 122 11.87 162 2000 1864 129.3 307 94.8 43.0 48.4 108.2 115.1 135 118 11.52 157 2001 1908 132.3 310 95.7 138 120 10.50 139 2002 1837 127.4 295 91.0 10.91 149 2003 1874 130.0 298 92.0 Sources: Indexes are calculated from Worldwatch Institute's data archive collected from FAO, USDA and other international sources (same as for Figure 1). There are no figures worldwide on pesticide use, so pesticide exports are tracked as a proxy. This may understate use since domestic production has increased in a number of countries.

Annex 2 - 1 Chapter 3

Meeting Agricultural Productivity Objectives

3.1 Considerations for Assessing output is not just a variable. A substantial increase is a Ecoagricultural Alternatives sine qua non for meeting the other two goals. Because ecoagriculture has multiple objectives and they However, this does not justify a “production fi rst” or a are not necessarily closely correlated, the challenge is “production only” strategy. Food consumption short- to fi nd some optimization among them. Where there are ages are most directly a consequence not of inadequate confl icts or contradictions among the goals, one should supplies but of poverty and a lack of purchasing power. focus on what tradeoffs can give the best joint outcome. Nobody should think that production by itself will To the extent that there are compatibilities and even satisfy human needs. Rather, it is a necessary but not synergies among the objectives, or these can be created, suffi cient condition for dealing with these. It can be it will be possible to attain more satisfactory outcomes argued, however, that there is a certain primacy as well since the objectives can then be met better respectively as urgency for raising agricultural productivity, albeit in by approaching them in a collective manner. environmentally-friendly ways and with linkages and conditions that ensure that enough of the food produced The task of achieving an optimum in this case dictates gets to those who need it most.1 a kind of minimax strategy, where one tries to satisfy each goal as fully as possible but subject to the attain- ment and maintenance of certain minimum levels for 1 We are not addressing here the issues raised both the others. Of course, the acceptable minimum can be pointedly and persuasively by Smil (2004), that part quite high, e.g., the food production that is needed to of the challenge of satisfying both food production and meet the nutritional requirements of a still-growing environmental objectives turns on dealing with the world population, especially since there are at least 800 problem of overconsumption, no longer just in richer countries but also within the middle-income and some million persons who are seriously undernourished at lower income countries. The shift to a, “Western” diet as present. Estimates of the increase needed vary, but most household incomes increase, with more consumption agree that in the next three to four decades, at least a 50 of meats, fats, sugars and salt, not only contributes percent increase in production should be attained. to growing and serious health problems associated with obesity (heart disease, diabetes, etc.), but it places Failure to accomplish this particular objective will put unnecessary burdens on the world’s land and water uncontrollable pressures on natural and biological re- resources. This is a more serious problem to the extent sources that are already under threat of loss. Moreover, that the food consumed comes from high-intensity stagnating or decelerating food production will under- animal production operations that very ineffi ciently mine efforts to create livelihoods and improve people’s convert vegetable foods into high-quality protein. Smil’s arguments about the irrationality of policies well-being, particularly for the poor, infl ating food that subsidize such patterns of production and con- prices and directing resources away from investments sumption are compelling, but rectifying these involves that could help create employment. So there are strong both altering dietary preferences and political reforms reasons why continuing improvement in agricultural that are beyond the scope of this paper.

26 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations The world’s capacity to produce more food to meet organisms, rather than on one particular species (crop higher levels of demand is not in question. The eco- or animal) at a time. It seeks to form synergies among nomic and environmental costs of doing so can be quite them that can be captured in production processes. Also, substantial, however, and need to be reckoned with. The agroecology explicitly embraces multiple objectives, situation of the poor will not be improved by produc- including benign impacts on ecosystems and contribu- ing more food at higher cost, higher because the new tions to social and cultural values, rather than direct all technologies require more capital investment or because efforts toward single goals such as yield or profi tability more marginal, less accessible land and water resources (Altieri 1987). are utilized. Moreover, in terms of economic growth and Proponents of agroecology claim that higher outputs livelihood expansion, if food production becomes more can be achieved by reducing rather than by increasing costly, this will direct resources away from investment certain inputs, such as inorganic fertilizers and agro- in creating jobs and wealth and meeting other needs. So, chemical biocides, as well as with less use of fossil- as suggested in the previous chapter, fi nding ways to fuel energy. By capitalizing on synergistic biological raise productivity in the agricultural sector is essential processes that can be nurtured within agroecosystems, to be able to meet all three objectives in positive-sum fostering benefi cial interactions among species, espe- ways. cially in the soil, they seek to reduce the “environmental There are two general directions in which agricultural footprint” of agricultural production. This, to be sure, production can develop in the 21st century, although makes little sense from an input-centered perspective the overall pattern of investment and activity will, of on agriculture that gives little thought or credit to what course, be some combination. This mix will respond plants and animals, in conjunction with their growth to the relationships among factors of production deter- environment, can do for themselves. The GxE effect mined by their relative availability, cost and produc- (genetics times environment), which stresses interac- tivity (Hayami and Ruttan 1985), responding also to tion and interdependence, is downplayed in more linear changes in technology and to certain political, social models that regard genes as directly causal. and cultural considerations. For agroecological approaches to be feasible underpin- The approach of the Green Revolution, which greatly nings for ecoagriculture, they will have to demonstrate increased food availability and lowered food costs, was that they can achieve higher agricultural productivity (a) to make improvements in the genetic potentials of with constant or often lower external input methods. plants and animals, and then (b) to increase inputs—of A substitution of one approach by the other is not the fertilizers, feeds, water, chemical treatments, etc. The issue. On theoretical grounds, optimization reasoning genetic improvements made were ones that in particular suggests some compromise and combination of ap- increased plants’ and animals’ effi ciency of converting proaches. Pragmatically, it never happens that whole inputs into outputs. Within such a paradigm, increas- systems of production get replaced by other systems, ing outputs is typically a function of investing in more except perhaps over many years, partly because there inputs. This implies that lower external input strategies are powerful factors that inhibit change. What is rel- must generally give lower output, as contended by the evant to consider is to what extent, and where, and how most outspoken advocates of “modern agriculture.” quickly, ecoagricultural systems will become accepted Avery (1995) even contends that the environment will and utilized. Possibly, ecoagriculture will be limited to benefi t from the use of more rather than fewer chemical areas that have presently relatively low productivity, inputs for agricultural production. where modern agriculture (plowing, chemical amendments, etc.) has not been performing very well anyway, Quite a different view of how agriculture is best and leaving the bulk of food production to higher-input most productively practiced has emerged over the past methods in other, more-favored areas. Or over time, several decades, loosely termed agroecology. This ecoagricultural elements may be integrated into existing differs from modern agriculture by focusing on com- modes of production, much as integrated pest manage- binations and communities of plants, animals and soil ment (IPM) and conservation tillage have become part

Meeting Agricultural Productivity Objectives 27 of modern agriculture over the last two decades after a centuries-old trend for farmers to avail themselves earlier skepticism. of external inputs, particularly to enhance the energy resources they can invest in their production, thereby Evaluation of alternative practices, though they are reducing or making more effi cient the labor and land formulated in terms of kind, needs to become one of resources at their disposal. degree. To what extent can lower external input agriculture meet agricultural production needs? This will Some argue that this trend is inexorable and that any vary according to physical environments, crops, and agricultural system that requires more labor inputs marketing and other infrastructure. That ecoagricul- therefore cannot become widely used. However, more ture may not be a superior or acceptable strategy in important than labor-intensity for decision-making, all cases does not invalidate it for some or many areas at least by households that depend particularly upon where the alternative of high-external-input agriculture their labor power for income and well-being, are con- may be less suitable, for various reasons. And possi- siderations of labor productivity. What are the returns bly over time, as resistance based on present interests to labor? Anything that earns them more output per and thinking diminishes, these practices will become hour or day of labor has an attractive logic, especially “mainstreamed.” if it reduces their cash costs, as fi nancial resources are typically a binding constraint. In making comparisons, it is important that the high external input agricultural alternative be assessed fully Also, where land is relatively scarce, what are the and fairly, including, importantly, a consideration of ex- returns to land? More intensive management of agri- ternalities. Decision-making by individual households cultural production generally increases the returns to typically deals with prevailing market prices for inputs land, although some of the gains from increases in land and outputs, calculating private net benefi ts, or costs, productivity may be offset by reductions in labor pro- without regard to externalities and the “social costs” ductivity, unless there are new technologies available resulting from production. From a societal point of to offset these. Throughout much of the world, smaller view, however, when deciding about policies, subsidies, farming operations have higher output per hectare, as infrastructure investments, etc., it will be very important larger operations optimize income by more “extensive” to take all externalities, positive and negative, into ac- strategies, which means less investment of resources count. Where the interests and conclusions of private per hectare. as distinguishable from public actors diverge, the latter A further consideration that will weigh ever more need to accommodate the former, at least in the short heavily in 21st century agricultural decision-making run. However, any disparities between the two sets of is returns to water, which will often become a more conclusions should be addressed as a matter of public limiting resource than land, labor or capital. Agricul- policy, with efforts made to reshape public policy and tural practices and systems that conserve water and redirect private choices so that through incentives, use it more productively, such as mulching or perma- regulations, etc., there is a convergence of private (in- nent ground cover, will be more and more attractive. dividual) and public (societal) net benefi ts. Nobody can know for certain what will be the effects When we refer to inputs as being high or low, all kinds of anticipated climate change, but higher temperatures of inputs need to be assessed. What are referred to as and shifting rainfall patterns are likely to put stress on “low-input” agricultural production strategies usually both irrigated and rainfed systems of production. This refer to purchased or external inputs, ones exogenous will make different crops, and land and water manage- to the farming system, not generated within it. What is ment practices more attractive in the future as modern often meant is “low external-input production strate- agriculture is “thirstier” than many alternatives. gies” which often have higher inputs in terms of labor There is also a time dimension that needs to be incor- and management intensity, employing more “imported” porated in any evaluations. When agriculture has been inputs to get higher productivity from household re- conducted with heavy tillage and/or with applications sources or to reduce the need for them. There has been of fertilizers and agrochemicals, the soil systems on

28 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations which agriculture is based have diminished nutrient ricultural sectors around the world in this 21st century resources such as nitrogen and reduced populations will increasingly be challenged to justify themselves in of soil biota that support and protect plant growth as terms of how they contribute to these three objectives, discussed in Chapter 6. Modern agricultural practices not just to outright production. create a kind of dependence on continuing tillage and chemical applications because intensively-exploited 3.2 Addressing Tradeoffs and soil systems become diminished in their capacities to Complementarities sustain crop yields without continuing manipulation of 3.2.1 Biodiversity or substitution for biological processes. Switching to production systems that are less intensive Gains in the productivity of agricultural lands that Green with respect to external inputs usually requires some Revolution and related technologies made in the latter period of “transition” to restore more favorable biotic part of the 20th century probably slowed the conver- conditions in the soil. Yields may decline at fi rst when sion of forest and other uncultivated areas to arable use, external inputs are reduced. The question for farmers thereby making a contribution to the conservation of is how deep and how long such reductions are, and can biodiversity (Dowswell and Borlaug 1995). There are, they be reduced in their severity and duration? Further, however, some analysts who object that the methods can farmers manage such transitions within their own used had many unintended or uncounted consequences, resource base, or do they need external fi nancial support particularly with regard to soil and water quality, so that to make the transition to a lower-external-input mode the total effect was not as benign as Green Revolution of production? Such support is a kind of subsidization proponents suggest (e.g., Altieri 1987). to put production onto a lower-cost track in the future. We leave to others the task of sorting out how much It is not very different in form but quite different in debit should be placed against the credits attributable to function from previous public sector subsidization of the Green Revolution in this regard. There are debates agricultural inputs such as fertilizers or pesticides to over how much benefi t was actually created since one get farmers to switch onto higher-cost tracks. However, should not attribute all gains in production since 1965 the benefi t-cost ratios and sustainability can be much to the Green Revolution. The proper comparison is more favorable because costs and vulnerabilities are with any incremental productivity gains attributable lowered. to its new technologies over and above some baseline These are diffi cult, multifaceted issues, often requir- rate of improvement that might have been expected ing more information than decision-makers, private to continue. Sorting this out is complex. Probably the and public, presently have in hand. With regard to new technologies made some positive contributions to ecoagricultural land use, which is our concern here, the maintenance of vulnerable ecosystems and to the it is important to stress that any evaluations should species richness associated with them. But our task is to be undertaken with a commitment to rely on empiri- look ahead, rather than back, considering what are the cal knowledge, fi tted into defensible models or other most promising options now in this new century. kinds of systematic analysis, that are explicit about The most obvious way in which some if not all eco- any value assumptions being made and not driven agriculture practices can benefi t from conservation by preconceptions that either favor or dismiss certain of biodiversity, so that agricultural productivity and modes of agriculture. Both modern agriculture and any also livelihoods are improved, is through preservation ecoagricultural alternatives should be even-handedly of a more diverse gene pool for the species of crops evaluated, looking at all external as well as internalized and animals being grown and of their near-relatives. costs and benefi ts of production. Analysis of any and all Some biotechnology visionaries envision a time when production systems should weigh their respective con- genes can be engineered rather than just relocated, tributions to agricultural productivity, to biodiversity but for some time to come, plant and animal breeders preservation and other environmental conservations, will probably have to rely on the genetic material that and to household livelihood and well-being, since ag- nature has provided in both variety and abundance,

Meeting Agricultural Productivity Objectives 29 recognizing that some or much of this is in danger of a substantial portion of the photosynthate produced in becoming lost. Biotechnology opens up opportunities their canopies is not widely appreciated, though well to utilize genes from a wide variety of species for any documented in the scientifi c literature.3 The carbohy- particular crop or animal improvement effort. But the drates, amino acids and other organic compounds put natural pool remains the base for innovation. Practices into the rhizosphere provide energy and other substanc- and benefi ts associated with this are considered in es for the bacteria and fungi living in, on and around Chapter 6, Section 6.2.1. the roots. These microbes in turn are food for larger organisms like protozoa and nematodes, which in turn One facet of biodiversity that is given too little attention get preyed upon by collembola, mites and various other is the importance of this within soil systems that directly arthropods. All of these, as well as plant roots, benefi t support plant life and, indirectly, all animal life includ- from the activities of earthworms and other fauna that ing human beings. Modern analytical methods such make the soil better aerated, better aggregated, more wa- as terminal restriction fragment length polymorphism ter-retentive, suppressive of pathogens, etc. Plants are (T-RLFP) are enabling microbiologists to gain more thus more active than passive, intricately interdependent precise knowledge about the populations of microor- with their environment. The disciplinary designation ganisms living and functioning in the soil.2 The soil has of “plant pathology” has focused scientifi c attention been a kind of “black box” about which we have had on the negative roles of microorganisms even though only limited knowledge based on those organisms that pathogens are a minority of soil biota, for the most part are culturable, a small fraction of the total. Indeed, a held in check by other organisms in the subterranean large part of our soil research has been conducted under domain. As seen in Chapter 6, this understanding of soil axenic conditions, i.e., after sterilization or fumigation and crop disease and health is gaining wider apprecia- has eliminated all living organisms, so that resulting tion and acceptance. explanations are framed in physico-chemical terms, minimizing or ignoring the biological dimensions of No matter how much effort is made to “industrialize” soil functioning. This gives a truncated understanding agriculture through engineering and chemical interven- of how soil systems actually perform. It is like trying to tions, it remains a thoroughly biological enterprise. understand the functioning of the human body through Chemical and mechanical inputs can make agriculture anatomy only, without studying its physiology, since more productive and predictable to a degree, but it studying sterilized soil samples is more like dissecting will always be vulnerable to “extreme events,” i.e., a cadaver than appraising the ways in which living droughts, fl oods, heat waves, cold spells, as well as pest organisms function. or disease attacks. The things that are included under the heading of “biotic and abiotic stresses” underscore What is widely known though still incompletely under- how dependent food production systems are on a multi- stood is that the fertility and sustainability of soil sys- faceted environment that needs to stay within certain tems depend on a complex and intricate food chain/food ranges of temperature, moisture, etc. if agriculture is web underground, well described by Wardle (2002). to be successful. These affect not only the plants and This differs from the common implicit view of soil as an inert medium for anchoring plants and a repository 3 See, for example, the encyclopedic treatments of this for (our) nutrients and water that are taken up by plant subject in Pinton et al. (2001) and Waisel et al. (2002). roots. That plants exude into the soil through their roots In general, about 40-60 percent of photosynthate goes into the roots, much of it used for metabolism; but a 2 This kind of analysis has showed, for example, that similar proportion gets exuded into the rhizosphere when nitrogen fertilizer is added to the soil, the expres- around the roots, where soil biota are orders of mag- sion of the nifH gene supporting biological nitrogen nitude more than in bulk soil. Good examples of this fi xation by endophytic bacteria living the plant roots literature on the symbiotic relationships between soil is inhibited and reduced (Tan et al. 2003). This general organisms and plant roots are Bonkowski (2004), effect of inorganic fertilizer inhibiting soil microbial Brimecomb et al.(2001), Dakota and Phillips (2002), processes including biological nitrogen fi xation has Frankenberger and Arshad (1995), and Gyaneshwar been reported in the literature for some time. et al. (2002).

30 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations animals of immediate concern but the other organisms can be made more productive than monocropped ones, that have major effects, positive and/or negative, on especially if full-cost accounting is done, calculating productive outcomes. the costs of soil loss through erosion and of desiltation, for example. A small change in rice cropping systems With enough expenditure of energy and other external in China showed that just by interplanting two different inputs, agricultural operations can withstand the effects varieties of rice, one high-yielding but susceptible to of severe environmental fl uctuations, but these are high- blast infection with the other a local variety resistant cost solutions except where the environment is naturally to blast, production could be increased by 89 percent, benign and reliable. Unfortunately, we appear to be with reduced costs of crop protection (Zhu et al. 2000). entering a period when “extreme events” are becoming The benefi ts of reduced chemical applications for the more common. Agriculture is sustainable only where, biodiversity and health of the soil were not included in for the most part, “normal” environmental processes these calculations. prevail or production systems are very resilient. Agro- ecosystems are better able to withstand the effects of Agricultural practices and cropping systems commonly biotic or abiotic “shocks” when they are more diverse, reduce the diversity of plant and/or soil biota, but they both in the crops that compose the farming system and can enhance or sustain ecological services that con- in the total ensemble of species, plant, animal and mi- tribute to biodiversity beyond the cultivated areas at crobial, that constitute the agroecosystem, and when soil least indirectly. Ecosystem services depend on a high systems are able to withstand various stresses. Much of degree of biodiversity if not necessarily on maximum modern agriculture has been premised on the economies biodiversity. Many farmers and even decision-makers of scale that can come with monoculture. But this has its who may place no particular value on biodiversity per own vulnerabilities which diversity within and among se can be motivated to protect this for the sake of the species, above- and below-ground, can mitigate. ecosystem services that maintain the quantity and quality of natural resources—water, soil, air. 3.2.2 Ecosystem Services It is easier—though still often quite diffi cult—to put a There is a more obvious convergence of interests be- value on these services than on the biodiversity that sup- tween agriculturalists and environmentalists with regard ports them, by estimating the value of soil lost through to the functions that ecosystems perform in terms of the erosion and of crops foregone for lack of water, the costs hydrological cycle and purifying water, land and air of desilting rivers and reservoirs and of health problems through biological processes (Section 6.2.8 in Chapter resulting from impure air or water, etc. Economists and 6). Intact ecosystems within watersheds help to capture others have attempted to attach monetary values to and store water in the ground, vegetation and microbial such benefi ts, comparing them with the costs or other populations, at the same time fi ltering and detoxifying forgone benefi ts of achieving them. This effort has had it so that more and better-quality water is available for some success and is the subject of considerable current agricultural, domestic and other uses throughout the interest, though it remains still inadequate for a fully year. Solid and fl uid wastes get decomposed and can be satisfactory assessment as discussed in Chapter 5. made less objectionable (and often benefi cial) through microbial activity. The maintenance of vegetative 3.3 Considering Ecoagricultural ground cover reduces soil erosion and loss from water Approaches runoff and wind, keeping land more productive and It is appropriate to think of ecoagriculture in terms of preserving water source quality. This also minimizes a variety of approaches rather than as a single thing or the costs of waterway, canal and reservoir siltation. even single characterization. In Chapter 1, we reviewed Both agricultural productivity and livelihood creation the ecoagricultural strategies that McNealy and Scherr and maintenance are enhanced by having intact and (2003) proposed as exemplifying this emergent ap- functioning ecosystems. Agricultural production sys- proach. All can contribute to agricultural productivity tems based on diversity from intercropping or rotation while enhancing biodiversity either directly or indi-

Meeting Agricultural Productivity Objectives 31 rectly. Although the term “agroecology” already has next generation, resilience in the face of shocks, and widespread use, the concept of ecoagriculture has merit robustness on response to the vagaries of climate—these by focusing on various kinds of agriculture, rather than are all valid objectives. on what appears to be a certain kind of ecology. Producing a surplus for market sales is always desirable As suggested in Chapter 2, the approaches that come for farm households, and this is important for countries under this rubric should not be regarded as some kind of as a whole, which need to assure the food supply for ur- primitive or backward agriculture because they are not ban populations. But for many countries, the agricultural “modern,” i.e., not do not rely on engineering, chemi- sector currently is not self-supporting, with hundreds of cal or genetic improvements. What comes under the thousands or even millions of food-defi cit households. heading of ecoagriculture can be considered as “post- This slows overall national growth because they are modern” agriculture in that it builds upon the science not yet productive enough to add much to net national and experience of “modern” agriculture as it reorients output. Larger-scale, currently more productive farming agricultural development efforts in directions that are operations may be very successful for themselves and more suited to the requirements of the 21st century in for the national agricultural sector, but to the extent that terms of factor productivity. In this sense, it may be their smaller-scale sector counterparts lack productive seen as the most modern kind of agriculture, drawing alternative opportunities for their underemployed labor, heavily on recent advances in the biological sciences this subtracts from rather than adds to national wealth. about how soil systems function as the foundation for Poverty reduction would thus benefi t the whole nation, all agriculture (Chapter 6). not just those identifi ed as poor. This is why a concern with livelihoods ranks alongside the other two criteria The reservoir of practices that can be drawn on for for successful agricultural development. ecoagricultural development is based on knowledge that has been developed in the past 10-20 years or more There have been many advances in knowledge and by researchers in many countries. These innovations practice in the last decade or two in a number of areas often have come from working closely with farmers that can contribute to effective ecoagriculture initia- and NGOs involved in innovations to resolve problems tives, building on an accumulating scientifi c base but or constraints that they encountered when using what that warrant much more research both for evaluation are regarded as modern agricultural methods under and for improvement. As a foundation for this report, various circumstances. The innovations are responsive we undertook a review of component elements for to variations in types of soil, topography of landscape, ecoagriculture, surveying literature that has appeared size of holding, labor, cash or other limitations, pest and just in the last fi ve years in the main relevant scientifi c disease problems, market failures, etc. They are thus journals and discussing with faculty who are most as a rule more adaptive strategies, formulated induc- knowledgeable about these areas of research and ap- tively, rather than universal solutions based on general plication to get their insights. The results of this effort to scientifi c principles deductively applied and validated summarize the state-of-the-art are presented in Chapter on a few (small) locations operating under near-ideal 6. Here we summarize some of the key fi ndings of this conditions, as much of contemporary research-driven review, having already referred to current thinking on innovation in agriculture. agrobiodiversity conservation and utilization above.

One common characteristic of ecoagricultural ap- 3.3.1 Organic Systems of Production with proaches is that these farming practices and cropping Improved Soil Health and Biodiversity systems have a plurality of objectives, not just aiming to achieve the highest yield possible or the greatest net One of the factors driving modern agriculture with its earnings per crop or highest present-year proﬁ t margins. dependence on engineering and chemical inputs has Ensuring food security, maintaining stable as well as been the achievement of higher returns to capital and increased income over time, enhancing the natural labor by displacing and reducing labor. One of the ar- resource base to assure productivity for this and the guments against “organic” agriculture has been that it

32 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations is too labor-intensive and thus is not competitive with been shown able, in some contexts, of doubling rice mechanized agriculture relying on fertilizers for nutri- yields or more without requiring any chemical inputs, ents, rather than mobilizing them from soil processes, and saving water at the same time. It has been shown and on chemicals for pest and disease control. Organic that in hilly areas of Central America, maize and bean farming methods fi nd that chemical inputs can be reproduction can be doubled or more with simple meth- duced and eventually eliminated. While there may be ods, not using mechanization or chemical fertilizers. some yield loss initially as the biotic systems below- The latter can add to yield, but chicken manure with and above-ground adjust (recuperate), cash outlays are green manure and cover crops can give better results reduced, so profi tability may suffer little or not at all. As (Bunch 2002). In these and other cases, greater labor- labor productivity increases with mastery of techniques intensity frequently substitutes for external-input use, and development of labor-saving practices and imple- creating an additional set of challenges that may offset ments, costs of production are further reduced.4 the gains achieved from lower external-input use. The question from a food-production standpoint is There is one branch of organic agriculture that is based whether aggregate output from such systems can com- on philosophical principles, while another branch is pare with modern production methods. In the domain more based on scientifi c understanding of soil processes of horticulture, it is found that once soil systems have and capacities. Both use similar practices that mobilize been enriched by various organic practices, yields can ecosystem services, so their results are quite similar. certainly be higher. A feature article in Nature (April 22, There are other branches that use organic practices for 2004) discussed the growing acceptance of “organic” more mundane reasons, either because environmental methods and principles by conventional science and protection regulations restrict their use of chemicals, commercial agriculture, not to mention large-scale food because of adverse effects on groundwater and/or wild- companies. One factor that draws them together is a life, or because given rising costs of purchased inputs growing appreciation of the contribution of soil biol- (fuel, fertilizer, etc.) and premium prices received for ogy to agricultural production, discussed in numerous chemical-free produce these alternative production places throughout this report. methods are becoming more profi table. The further question is whether organic methods can Nature (April 22, 2004) reports that worldwide the satisfy the much larger basic demand for the staple crops demand for organic products is rising about 20 percent that supply most of the calories in our diet, particularly per annum. It concludes that the evidence on whether for the poor. Conservation (no-till) agriculture (Sec- organic food products are healthier for consumption is tion 6.2.9 in Chapter 6) is not necessarily organic, but too mixed to draw any fi rm conclusion (although the it is evolving in that direction, reducing the need for evaluations are mixed between showing health benefi ts chemical fertilizers and crop protection as soil quality and not, so probably there will ultimately be agreement is improved. In the U.S., over 30 percent of arable land on some benefi ts but maybe not as great as proponents is now under some form of conservation tillage, most suggest). In terms of impact on the environment, there of this for staple crops. Moreover, the System of Rice are probable benefi ts from improved water, air and soil Intensifi cation (SRI) (Section 6.2.10 in Chapter 6), quality, though some studies do not show any signifi - though not necessarily an “organic” methodology, has cant improvement. The durability of benefi ts cannot be known because there are few long-term studies. On 4 One concern is that labor-intensity can be a disin- productivity effects, a 21-year Swiss study concluded centive for adoption of these practices. We discuss that organic fi elds produce, on average, 20 percent less this important constraint further in the fi nal section than conventional fi elds, though another long-term of this chapter. On the other hand, labor intensive study in the U.S. found no difference, with a 20-40 production creates employment for the poor, who are most in need of livelihood enhancement. A reduction in percent advantage for organic fi elds in drought years, labor-intensity which is a good thing for agricultural probably because root systems are more developed and production considerations has negative consequences soil systems more robust. for livelihoods.

Meeting Agricultural Productivity Objectives 33 A continuing point of controversy, noted in the Nature culture technology and policy, because trees were the review, is whether organic farming can acquire suf- province of another discipline—forestry—and usually fi cient nitrogen, without using fertilizers, to achieve a different government department. and sustain the yields needed to feed large populations. Since then, the science of agroforestry has explored the If one compares organic matter inputs with chemical advantages of combined production systems involv- fertilizer, the former do not contain equivalent nitrogen. ing trees, crops and animals and has generated a host However, proponents of organic agriculture see this of insights into how woody perennials interact with comparison as spurious, because the function of com- other elements of agricultural systems to affect their post and other organic inputs is to serve as a substrate productivity and sustainability. A focus of research has for soil biota more than to nourish plants directly. They been to determine the agroecological conditions under advise: instead of feeding the plant, feed the soil, and which particular agroforestry practices will generate the soil will feed the plant. This thinking is supported multiple private and public benefi ts, and to understand by what is being learned about “soil health” and below- the constraints in realizing this potential. ground biodiversity (see Sections 6.3.6 and 6.3.7 in Chapter 6). To understand and evaluate organic methods It is now well established that the key to effective thus invites a paradigm shift in thinking, though proper practice is to design and manage systems that optimize evaluation needs then to be as empirical and systematic complementary interactions and limit competitive ones. as any other made in agricultural science. The initial promise of agroforestry stemmed from the realization that many early successional tree spe- There is still more unknown than known about the cies were fast-growing and could deliver a variety of interactions among plants, soil, nutrients and microor- products and services that small farms needed, quickly ganisms. But enough scientifi cally-respectable research and often simultaneously. Food, fodder, fi rewood and is accumulating, documenting the contributions of building materials could be produced by the same or a microorganisms to soil fertility and sustainability as complementary selection of trees that also improved soil well as to plant performance, that previous dismissals fertility and moisture retention. Over time research has of organic agriculture as outside the realm of science produced a better understanding of the ecological, cli- are no longer sustainable. This does not mean that all matic and management conditions under which various conventional agriculture can, should or will ever be combinations of trees, annual crops, and animals can or given up. The adverse effects of chemical-based crop cannot generate multiple benefi ts on a sustained basis. production and protection are often not very serious, In 1999 a publication series on “Advances in Agroecol- or there may be little alternative to them for meeting ogy” released a state-of-the-art volume, Agroforestry in immediate food needs. As said before, ecoagriculture is Sustainable Agricultural Systems, edited by L. Buck, J. not intended to be a monolithic replacement for present Lassoie, and E. C. M. Fernandes. The material in this agricultural systems and practices. The conclusion from compendium showed how agroforestry can contrib- our review of scientifi c understandings is that there are ute to sustainable agricultural productivity. We have good justifi cations for more of agricultural production drawn upon selected information from this volume to to be moving in this direction. In doing so, it can have characterize the conceptual foundations of agroforestry benefi cial effects for biodiversity and for livelihoods. that are particularly relevant to the assessment of eco- 3.3.2 Agroforestry agriculture. In Chapter 6 we amplify how knowledge of agroforestry contributes to an understanding of the The fi eld of agroforestry arose in the late 1970s to make potential for ecoagriculture. explicit the socioeconomic roles and biophysical func- A conceptual foundation upon which tropical agrofor- tions of trees within tropical farming systems. Prior to estry research was initiated is that trees help maintain this perceptual reorientation, trees and shrubs on farms soil fertility and support the growth of associated crops and in farmscapes were not “seen” by mainstream (Rao et al., 1997). Today a substantial body of knowl- agronomists and other scientists and professionals who edge is available on the role of nutrient cycling in the had instrumental roles in shaping international agri-

34 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations maintenance of soil fertility in agroforestry systems. in hedgerows/shelterbelts maintained along the edges Nair et al. (1999) reviewed four major categories of of cropland and pastureland, while simultaneously agroforestry systems that occur across the four major reducing soil erosion and moisture loss. They point agroecological/geographical zones of the tropics and out also the roles of riparian tree buffers in reducing demonstrated that agroforestry systems can provide the negative effects of sediment and chemical run-off nitrogen for crop production in all of them (hedgerow on aquatic life systems. Their conclusions, like those intercropping, parklands, improved fallows, and shaded of many other researchers, are that agroforestry can perennial crop systems) under specifi ed conditions. provide the means to increase biomass, and in turn, The review revealed that agroforestry systems are not improve food crops and livestock productivity, while capable, however, of providing suffi cient amounts of simultaneously upgrading the productivity of degraded phosphorus to maintain crop yields. The authors con- soils. They provide evidence of a doubling of grain cluded that while the basic tree-mediated processes of crop yields in some agroforestry systems and of a 60 nitrogen fi xation, production and decomposition of tree percent increase of animal products while protecting biomass, and nutrient uptake from deep soil horizons are the soil from erosion. measurable, a lack of appropriate research methodologies prevents adequate determination of the dynamics 3.3.3 Conservation Agriculture of these processes under fi eld conditions, thus making The concept and practice of Conservation Agriculture it diffi cult to predict the success of management strate- has evolved out of a variety of zero-till, no-till and low- gies applying these processes. till practices over the past 30 years, as noted in Section Another cornerstone of the fi eld of agroforestry re- 6.2.9 of Chapter 6. Three decades ago, the idea that one search is that there are numerous wild species of trees could grow fi eld crops more productively and more and shrubs that have potential for being domesticated profi tably without plowing was for most farmers and in agroforestry systems, to exponentially expand the scientists an absurdity—about as sensible as propos- timber and non-timber benefi ts of trees to farmers. ing that rice could produce better in unfl ooded paddies The domestication and commercialization of trees is than in continuously fl ooded paddies, discussed in the viewed by Sanchez and Leakey (1997) as essential following section. Yet, experience and scientifi c evalu- to achieving agroforestry’s potential to balance food ations have shown that conservation tillage—ceasing security with the utilization of natural resources within tillage and maintaining continuous vegetative cover on an ecological framework “akin to the normal dynamics fi elds—has many advantages for enhancing production of natural ecosystems.” Pointing to important tree- with lower inputs and sustainable yield, also making domestication successes worldwide, and with explicit agricultural domains more hospitable to biodiversity attention to the nutritional value of new or improved above and below-ground (Calegari 2002). tree products, Leakey and Tomich (1999) project high Agronomists have long known the negative conse- returns on research investment in domestication. They quences of plowing for many years: increased wind and lay out a carefully reasoned agenda for expanding the water erosion, soil compaction, oxidation of soil organic efforts of public research institutions to domesticate matter, loss of nitrogen, adverse impacts on both aerobic more tree species in order to promote economic growth and anaerobic soil organisms and loss of biodiversity with enhanced food security and reduced poverty, while among microorganisms, and loss of aggregation and preserving the possible environmental services of wild desirable structural properties affecting, among other species. things, water infi ltration rates (Brady and Weil 2002). A third defi ning principle of agroforestry, in addition However, the pragmatic need to control weeds made to roles of trees in soil and crop improvement and the plowing an accepted practice, discouraging consider- direct benefi ts of tree products to households, concerns ation of alternatives, particularly investigation of other its relationship to the conservation of biodiversity. Pi- means for weed control that could overall enhance mentel and Wightman (1999) provide evidence of wild desirable forms of biological activity in and on the soil. biodiversity conservation in multispecies gardens and Avoiding plowing can cut erosion dramatically and

Meeting Agricultural Productivity Objectives 35 encourages great soil microbial diversity and activity. In Chapter 6 (Section 6.2.10), we report briefl y on the It also avoids the destruction of biologically-induced most demonstrative examples of integrated resource soil aggregation and the ripping up of vast but unseen management, achieving agricultural objectives more networks in the soil of the hyphae of mycorrhizal fungi effi ciently and benefi cially as well as making the agri- which assist plant roots in acquiring water and nutrients cultural environment more hospitable to biodiversity, from a greater volume of soil. also improving livelihoods and well-being. We go into one example at some length here because it shows how The initial experimentation with no-till has evolved more ecologically-grounded crop management can into more complex systems for managing plants, soil, achieve greater outputs by using less external inputs, water and nutrients, including particularly permanent a strategy that will make ecoagriculture a more fea- vegetative cover as CIRAD researchers working with sible proposition because the benefi ts of positive-sum colleagues in Brazil, Vietnam, Madagascar, Tunisia complementarities can be enlisted to serve multiple and other countries have done. It can sometimes take objectives rather than have to wrestle with zero-sum a few years to build up the fertility that has been lost tradeoffs. through tillage practices, but higher levels of production are attainable with these methods and at lower cost (P. The System of Rice Intensifi cation (SRI) developed in Hobbs, pers. comm.). Madagascar some 20 years ago and now is spreading around the world, with positive results demonstrated The CIRAD experimentation in Brazil has led to farm- in about 20 countries. SRI often can double the yields ing systems where yields of maize and soybeans can of irrigated rice when SRI methods are used as recom- be raised by 50 percent with a 50 percent reduction in mended. This increase can be achieved at lower cost the use of external inputs, mobilizing nutrients already because: in the soil but unavailable with current practices and then recycling them. These methods adapted to the • Farmers do not need to purchase and use any new production of unirrigated (rainfed) rice have led to seeds; the methods have worked with any and all average yields over 8 tons per hectare, more than ob- varieties so far, eliciting a more productive phenotype tained in most irrigated rice production systems (Seguy from any rice genotype. et al. 2003). Some soil amendments can be used with • Farmers do not have to apply chemical fertilizer; conservation agriculture as it is not an “organic” pro- although this gives positive results, home-made duction system. However, it evolves in that direction compost usually gives better results at lower cost. as soil biodiversity is enhanced with these alternative practices. • Neither are agrochemicals usually needed, since as a rule, SRI rice is enough healthier and resistant to 3.3.4 Integrated Resource Management pest and disease attacks that chemical protection is not economic. conservation agriculture is an example of a broader set of agricultural production systems that can be charac- • Irrigation water can usually be reduced by about 40- terized as “integrated resource management,” which 60 percent; fi elds are not kept continuously fl ooded includes what is called integrated nutrient management. during the vegetative growth phase. Basically, integrated resource management refers to With costs of production being reduced while yield changes in the direct management of plants, soil, water goes up, profi tability is enhanced at the same time that and nutrients, which indirectly affect the abundance adverse environmental impacts are diminished. SRI and diversity of soil organisms to obtain higher yields permits less water to be withdrawn from surface or and greater production effi ciency through synergistic groundwater supplies, and soil and water quality are effects among these resources. Where animals are part improved when fertilizers and agrochemicals can be of the farming system, they are additional resources to reduced. be managed in complementary ways. This all sounds too good to be true, and there are objections from within the scientifi c community, as noted in

36 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Chapter 6. But the objections are for the most part based methods because of this crop’s genetic potential for on limited or incorrect information and on questionable profuse tillering. We have seen that SRI ideas applied assumptions. Once the methods applying SRI insights to upland (rainfed) rice in the Philippines could give and principles have been adapted to local conditions yields averaging over 7 tons per hectare, three to four and are popularized, many farmers appreciate their times more than usually obtained without irrigation. advantages, even if some scientists hold back. Examples SRI challenges standard thinking because its methods from two countries show the scope for achieving ag- produce more output with less total inputs. In fact, ricultural improvements with SRI. Additional country initially SRI requires more labor input, while farmers experiences are summarized in Chapter 6. are learning its methods, but once these have been mastered, many farmers fi nd that SRI can even become • In Cambodia, where only 28 farmers could be per- labor-saving for them.6 suaded (by the NGO known as CEDAC) to try SRI methods in 2000, by 2003 this number had reached How is this possible? The SRI experience suggests that 9,100, and this year, the number is expected to be it is possible to have the kind of complementarity and between 40,000 and 50,000. Farmers who have tried synergy in farming systems that can obviate the need the methods for three years have seen their yields go for tradeoffs, having both more agricultural output and from 1.34 tons per hectare to 2.75 tons per hectare; conservation of biodiversity: their costs of production fall from 231,300 riels per • With SRI, by transplanting very young seedlings hectare to 113,140 riels per hectare, and their net (8-12 days old instead of 3-4 weeks old), the plants' household income from rice go from 499,900 riels potential for greater tillering and root growth is pre- to 879,800 (Tech, 2004). served. This can be explained in physiological terms • The International Water Management Institute by an understanding of phyllochrons (see Section (IWMI) evaluated SRI methods in Sri Lanka with a 6.2.10 in Chapter 6). SRI plants have 30-50 tillers, survey of 60 SRI farmers randomly selected in two sometimes even 80 to 100 or more, with correspond- districts and 60 non-SRI farmers. The fi rst group ingly larger root system—when other conducive SRI were not using all the recommended practices, or practices are also used.7 using them fully, but still averaged a 50 percent in- • With SRI, rice is planted much more sparsely, leaving crease in yield. There was a 90 percent increase in wide spacing between plants. Instead of having 3-6 the productivity of water used as applications could plants in a hill, single plants are set out in the fi eld, be reduced. Labor productivity went up 50 percent not in a row but in a grid pattern with spacing of in the dry season and 62 percent in the dry season. Profi tability of rice production (riels per hectare) 6 In the fi rst year, labor requirements can be 20-50 went up 83 percent fi guring labor costs at the pre- percent more. But in a Cambodia evaluation, 55 per- vailing off-farm wage rate and 206 percent if family cent of the farmers who had three years of experience labor inputs were not costed (Namara et al. 2003).5 with SRI reported that it is easier to practice than their conventional methods; only 18 percent considered it SRI principles can probably be extrapolated or adapted more diffi cult, and 27 percent said there was no real for other crops, but rice may be more responsive to its difference (sample size N = 120) (Tech, 2004). In an evaluation done with Madagascar farmers (N = 108), 5 Of particular interest was that rice farmers using it was found that SRI methods were labor-saving by conventional methods were experiencing net economic the fourth year (Barrett et al. 2003). losses from their production in 28 percent of seasons; 7 Farmers are beginning to experiment with direct SRI farmers only in 4 percent, so given the lower cost seeding to save the labor required for nursery-making requirements and higher yields, SRI was also less and transplanting, and this is also producing good risky. Contrary to the conclusion from some research results. Transplanting is not necessary with SRI; what in Madagascar (Moser and Barnett 2003), the IWMI is necessary is to avoid trauma to rice plant roots after evaluation found that poorer farmers were as likely they have begun their vegetative growth, i.e., beyond to adopt SRI as richer ones, and were more likely to about the 15th day (the start of the 4th phyllochron of continue using it. growth).

Meeting Agricultural Productivity Objectives 37 25x25 cm or wider. This gives plants more exposure which means upper leaves are “subsidizing” lower to solar radiation and helps achieve “the border ef- leaves. fect” for the whole fi eld, not just along the edges.8 • Less irrigation water means that plant roots do • With SRI, rice paddies are not kept continuously not degenerate, preserving their ability to take up fl ooded during the period of vegetative growth, nutrients throughout the growth cycle rather than only kept moist, and intermittently dried to induce begin senescing after panicle initiation. Alternating deeper root growth. When roots are continuously wet and dry soil conditions also contributes to soil hypoxic, they begin degenerating, so that by the biodiversity by getting both aerobic and anaerobic time of fl owering, when grain production begins, as biota and giving plants the support of a wider range many as 3/4 of the roots can have degenerated. This of organisms. impairs plants' capacity for accessing soil nutrients, • Less or no application of chemical fertilizer and and makes the addition of chemical fertilizer all the agrochemicals contributes to more abundant and more necessary. With SRI methods, it takes fi ve to diverse soil biotic communities which perform a six times more force (kilogram per plant) to uproot variety of services for plants. These include be- rice plants because of their larger and healthier root sides biological nitrogen fi xation, also phosphorus systems. solubilization, greater nutrient uptake, suppression • With SRI, it is recommended that compost be added of soil pathogens, inducing systemic resistance, to the fi eld to support larger and more diverse popula- producing phytohormones to stimulate root growth tions of soil biota, in the idiom of organic farming, and other services listed in Sections 6.2.6 and 6.2.7 feeding the soil rather than the plants. Chemical in Chapter 6. fertilizer enhances yield with SRI, but not as much as The case of SRI has been elaborated at length here will organic fertilization, because the latter serves as because it shows new possibilities for developing ag- a better substrate for microbial populations. Further, ricultural production systems that are less dependent the application of inorganic nitrogen suppresses the on external inputs and better able to match and benefi t expression of biological nitrogen-fi xing genes in from the processes that occur in natural systems. This endophytic bacteria that live in rice roots (Tan et al. has been the guiding principle of CIRAD’s research 2003) to establish more productive and sustainable farming Thus, one can show with scientifi c explanations and systems in a wide variety of countries, mimicking as validation why “less” can produce “more.” much as possible the ground cover and nutrient cycling of forest ecosystems. • Transplanting small, younger plants can preserve greater plant potential for tiller and root growth for This has been thought to be not productive enough to physiological reasons that are documented in the meet world food needs. But there is mounting evidence literature.9 that these agroecological or ecoagricultural systems are often competitive in terms of sheer output, and more • Fewer plants per hill and per m2 can produce more rice because the close spacing of typical rice production can reduce radiation in the lower part of the 9 Phyllochrons, similar to degree-days but better canopy even below the threshold for photosynthesis, grounded in physiology, were fi rst documented in research by T. Katayama in the 1920s and 1930s. He did not publish his results, however, until 1951, after 8 Scientists have recognized for years that rice plants World War II, and his book was never translated into on the edges of fi elds grow stronger and produce more. English. Phyllochrons are still not much known among When taking samples to estimate yield, these must be rice scientists, but they are studied by wheat scientists taken from the center of the fi eld “to avoid the border (see special issue of Crop Science, 35:1, 1995) and many effect.” However, this effect is only undesirable if it forage scientists since this analysis applies for all biases estimates; agronomically, it is something to gramineae species. It is well-known by rice scientists be desired. in Japan (Matsuo et al. 1997).

38 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations benefi cial in terms of profi tability because costs of pro- embrace polycropping in preference to monocropping, duction are lowered. There is not enough length or depth reduce soil disturbance and erosion, and so forth—are of experience to make any fi rm judgments about the likely to reduce the impairment of soil biodiversity and sustainability of such systems, but we need to remember to provide more hospitable habitats for a wide range of that the modern, high-external inputs systems are not species. Thus, while no studies that we could fi nd have themselves prime candidates for sustainability. calibrated biodiversity effects of these alternatives, general knowledge of ecological principles and dynamics It would be a mistaken and dubious enterprise to suggests that the effects of these farming practices and construct agricultural systems that cannot utilize any systems should be benign or even positive. external inputs, and ecoagriculture does not take a doc- trinaire position on this. It is not intrinsically “organic,” There are a number of different kinds of biodiversity even though it benefi ts from the insights and practices to be evaluated and supported, as discussed in the next that have come from organic agricultural production. chapter. Few if any kinds of agriculture are going to be SRI is pragmatically rather than necessarily organic. solutions for protecting a particular endangered species If any nutrients become limiting, such as phosphorus, of any genus. Conversion of wild lands to cultivation they can be added, probably with little adverse effect or grazing invariably will reduce the biodiversity that on soil biota if genuine defi cits are being remedied. It existed before agriculture was introduced. But if the is amendments in excess of plant and other biotic needs agriculture itself has some degree of diversity, and fi elds that create problems of soil and water pollution and of have borders, hedgerows, windbreaks, etc. the level of inhibiting microorganisms. conservation can be reasonably high, and certainly more favorable than conventional practices. 3.4 The Scientifi c Basis for Ecoagricultural Approaches Even an agricultural sector as modern and bountiful as California’s depends for its productivity and sustain- More detailed information and discussion of the body ability on the vast array of organisms of myriad species of experience and literature that informs the consider- that live in, on and around the fi elds and pastures that ation of ecoagricultural alternatives presented here is produce crops and animals (Qualset, 1995). While it given in Chapter 6. Unfortunately, there is very little has been thought most effi cient to focus research and literature that provides any specifi cs on the linkages and resources on one species at a time, agriculture remains effects between agricultural practices and conservation a thoroughly biological enterprise, where interactions of biodiversity, discussed in the next chapter. This is and symbioses among species are basic to each spe- diffi cult to offer any overarching conclusions on the cies’ success. issue of tradeoffs vs. complementarities with regard to Ecoagriculture seeks to capitalize on such dynamics, agroecological options. In general, the practices and which give it the chance, even with low inputs, to systems associated with ecoagriculture should be more achieve high levels of productivity and be economi- favorable for livelihood generation from the standpoint cally profi table. What is not known, given the limited of reducing external-input-dependent modes of produc- experience with ecoagricultural approaches and even tion that often are favored. The latter have typically less evaluation research done on them, is how far they tended to be labor-replacing or -displacing, through can be developed as an alternative to conventional the mechanization of operations or the substitution of agriculture, and in what and how many places. The chemicals for manual labor. We have not found evidence scientifi c basis for pursuing ecoagriculture in many if that agroecological approaches have adverse impacts not all places appears quite sound, even if it still needs on livelihood creation, and they should, in general, be and warrants further elaboration. supportive of more and more diverse income streams for households most in need of income. One important constraint to the broad applicability and dissemination of agroecological and ecoagricultural Agricultural practices such as those reviewed in Chapter approaches, as previously mentioned, relates to labor 6—which reduce water extraction and/or chemical use, demands and labor use. Very often, it is the substitution

Meeting Agricultural Productivity Objectives 39 of labor inputs for externally-purchased inputs in these Domestic and International Agribusiness Manage- systems that makes comparable, or even increased, ment, ed. R. A. Goldberg, 1-31. Greenwich, CT: productivity levels achievable. These strategies yield JAI Press. signifi cant benefi ts in economizing on the use of often Brady, N. C. and R. R. Weil. 2002. The Nature and unaffordable external inputs and achieving higher labor Properties of Soils, 13th edition. Upper Indian River, use among household members, as well as by increas- NJ: Prentice Hall. ing local and regional demand for labor. However there are numerous disadvantages that must be kept in mind. Brimecombe, M., F. A. DeLeij, and J. M. Lynch. 2001. Family labor resources and local labor markets may The effect of root exudates on rhizosphere microbial be incapable of supplying the required labor demands, populations. In: The Rhizosphere: Biochemistry and especially on a seasonal basis. Labor may have higher Organic Substances at the Soil-Plant Interface, R. opportunity costs especially in peri-urban areas and Pinton, Z. Varanini, and P. Nannipieri, eds., 95-140. other areas of relatively good infrastructure, and may New York: Marcel Dekker. thus be unavailable at the lower levels of remuneration Bunch, Roland. 2002. Increasing productivity through common to agricultural employment. Finally, agricul- agroecological approaches in Central America: Ex- tural labor is typically hard and arduous work, and the periences from hillside agriculture. In: Agroecologi- prospect of doing more of it may be unappealing to cal Innovations: Increasing Food Production with many farm households. These and other limitations Participatory Development, N. Uphoff, ed., 162-172. of agroecological approaches (documented in Lee London: Earthscan Publications. and Ruben 2001), are important to keep in mind when evaluating the offsetting benefi ts of low external-in- Calegari, Ademir. 2002. The spread and benefi ts of put systems. Empirical evidence on the adoption and no-till agriculture in Paraná State, Brazil. In: Agro- subsequent disadoption of agroecological approaches ecological Innovations: Increasing Food Production has confi rmed that higher labor demands and related with Participatory Development, N. Uphoff, ed., labor market issues are among the primary causes of 187-202. London: Earthscan Publications. disadoption where it has occurred (Neill and Lee 2001; Dakora, F. D., and D. A. Phillips. 2002. Root exudates Moser and Barrett 2003). as mediators of mineral acquisition in low-nutrient environments. Plant and Soil 245, 35-47. REFERENCES Frankenberger, W. T. and M. Arshad. 1995. Phytohor- Altieri, M. A. 1987. Agroecology: The Scientifi c Basis of mones in Soils: Microbial Production and Function. Alternative Agriculture. Boulder: Westview Press. Marcel Dekker, New York. Avery, D. 1995. Saving the Planet with Pesticides and Gyaneshwar, P., G. Naresh Kumar, L. J. Parekh and P. S. Plastics: The Environmental Triumph of High-Yield Poole. 2002. Role of soil microorganisms in improving Farming. The Hudson Institute, Indianapolis, IN. P nutrition in plants. Plant and Soil, 245, 83-93. Barrett, C. B., C. M. Moser, J. Barison, and O. V. Hayami, Y. and V. W. Ruttan. 1985. Agricultural Innova- McHugh. 2004. Better technologies, better plots or tion. Johns Hopkins University Press, Baltimore, MD. better farmers? Identifying changes in productivity Leakey, R. R. B., and T. P. Tomich. 1999. Domestica- and risk among Malagasy rice farmers. Agricultural tion of tropical trees: From biology to economics Systems, forthcoming. and policy. In: L. E. Buck, J. P. Lassoie and E. C. Bonkowski, M. 2004. Protozoa and plant growth: The M. Fernandes, eds., Agroforestry in Sustainable microbial loop in soil revisited. New Phytologist, Agricultural Systems, pp. 319-338. Boca Raton, FL: 162, 617-631. Lewis Publishers. Borlaug, N. and C. J. Dowswell. 1995. The importance Lee, D. R., and R. Ruben. 2001. Labor use in sustainable of agriculture and a stable adequate food supply agricultural systems. Paper presented at XXIV Con- to the well-being of humankind. In: Research in

40 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations ference of the International Association of Agricul- ricultural productivity. California Agriculture, tural Economists, August, 2001, Berlin, Germany. 49, 45-49. Matsuo, Takane, Yuzo Futsuhara, Fumio Kukuchi, and Rao, M.R., Nair, P.K.R.and C.K Ong, 1997. Biophysi- Hikoyuki Yamaguchi. 1997. Science of the Rice cal interactions in tropical agroforestry systems. Plant, 3 volumes. Tokyo: Food and Agriculture Agroforestry Systems 38, 3-49. Policy Research Center. Sanchez, P.A. and Leakey, R.R.B, 1997. Land use trans- McNealy, J. and S. Scherr. 2002. Eco-Agriculture: formation in Africa: Three determinants for balanc- Strategies to Feed the World and Save Wild Biodi- ing food security with natural resources utilization. versity. Island Press, Washington, DC. European Journal of Agronomy. Moser, C. M. and C. B. Barrett. 2003. The disappointing Seguy, L., S. Bouzinac, A. C. Maronezzi, E. Scopel, J. adoption dynamics of a yield-increasing, low-exter- L. Belot, and J. Martin (2003). The success of no- nal-input technology: The case of SRI in Madagas- tillage with cover crops for savannah regions: From car. Agricultural Systems, 76, 1085-1100. destructive agriculture with soil tillage to sustainable agriculture with direct seeding, mulch-based Nair, P.K.R., Buresh, R.J., Mugendi, D.N. and C.R. systems—20 years of research of CIRAD and its Latt, 1999. Nutrient cycling in tropical agroforestry Brazilian partners in the cerrados region of Brazil. systems. In: Buck, L.E., J.P. Lassoie and E.C.M. Proceedings of the II. World Congress on Conser- Fernandes (eds) Agroforestry in Sustainable Agri- vation Agriculture, Iguassu Falls, Brazil, August, cultural Systems. Lewis Publishers, Boca Raton. 310-321. (Chapter 1). Smil, V. 2004. Food production and the environment: Namara, R. E., P. Weligamage, and R. Barker. 2003. Understanding the challenge, encountering the lim- The prospects of System of Rice Intensifi cation its, and reducing the waste. Paper prepared for the adoption in Sri Lanka: A microeconomic and so- World Bank, March. ciopolitical perspective. Paper prepared by staff of the International Water Management Institute, Tan, Z., T. Hurek, and B. Reinhold-Hurek. 2003. Effect Colombo, for presentation at International Confer- of N-fertilization, plant genotype and environmental ence on Irrigation and Drainage (ICID) conference, conditions on nifH gene pools in roots of rice. Envi- November, Taipei. ronmental Microbiology 5, 1009-1015. Neill, S. and D. R. Lee. 2001. Explaining the adoption Tech, C. 2004. Ecological System of Rice Intensifi - and disadoption of sustainable agriculture: The case cation (SRI) Impact Assessment. CEDAC Field of cover crops in Northern Honduras. Economic De- Document, February. Cambodia Center for the velopment and Cultural Change 49: 793-820. Study and Development of Agriculture CEDAC), Phnom Penh. Pimentel, D. and A. Wightman. 1999. Economic and environmental benefi ts of agroforestry in food and Uphoff, N., ed. 2002. Agroecological Innovations: fuelwood production. In: L. E. Buck, J. P. Lassoie and Increasing Food Production with Participatory De- E. C. M. Fernandes, eds., Agroforestry in Sustainable velopment. Earthscan Publications, London. Agricultural Systems, pp. 295-318. Boca Raton, FL: Wardle, D. A. (2002). Communities and Ecosystems: Lewis Publishers. Linking the Aboveground and Belowground Compo- Pinton, R., Z. Varanini and P. Nannipieri, eds. 2001. nents. Princeton University Press, Princeton, NJ. The Rhizosphere: Biochemistry and Organic Sub- Waisel, Y., A. Eshel, and U. Kafkaf, eds. 2002. Plant stances at the Soil-Plant Interface. New York: Roots: The Hidden Half, thirdthird edition.edition. MarcelMarcel Marcel Dekker. Dekker, New York. Qualset, C. O., P. E., P. E. McGuire, and M. L. Zhu, Y., et al. 2000. Genetic diversity and disease con- Warburton. 1995. ‘Agrobiodiversity’ key to ag- trol in rice. Nature 406, 718-722.

Meeting Agricultural Productivity Objectives 41 Chapter 4 Meeting Biodiversity Conservation Objectives

In 1981, as I emerged from a Ph.D. program in ecology, I sought real-world applications of the principles of ecosystem functions and ecological interactions I had been studying. Two areas of particular interest emerged: conserving biodiversity and enhancing sustainability of agriculture. I spent the next twenty years bouncing back and forth between them, often wondering why it seemed so hard to bring the two together, despite a common dependence on ecological processes for long-term success. (Soule 2002:169)

4.1. Considerations for Assessing are critical. However, this is a statement of who we are. Conservation of Biodiversity in The question to be addressed in this report is: what are Agriculture we going to do? Chapter 4 highlights what is known and what is not known about the varied relationships Assessments regarding biodiversity are fraught with between agricultural practices and wild biodiversity. We emotional pitfalls, misunderstandings of goals, terms, build upon a selection of studies that have examined ex- and concepts and a dichotomy of perspectives that sets actly these relationships. But before starting, a number “nature fi rst” against “people fi rst.” We believe that of issues should be clarifi ed with regard to assessing much of this controversy is a confusion more often the conservation of biodiversity within agricultural based on a mistrust of the perceived philosophy of the settings: spatial scale; ecosystem health; domesticated “other” camp than on real disagreements over issues. varieties; and the overlap between wild biodiversity Over the past 20 years we have been involved in numer- and agrobiodiversity. ous discussions and graduate seminars at Cornell where the policy debate quickly polarized into one where pro- 4.1.1. Spatial Scale ponents for biodiversity (usually biologists) argued with persons focusing on the plight of poor people trying to We are trying to evaluate/assess agricultural systems/ make a living off of the land (most often agronomists practices that maintain as much of existing biodiversity or sociologists). To conclude that biologists do not care locally, regionally, or globally as possible. But clarifi - about poor rural people or that social scientists studying cation is needed regarding quantity vs. quality of wild agriculture do not care about the world’s natural biota biodiversity. The hierarchy of local to global diversity is too simply stated, and certainly wrong. is important. For example, it is possible to increase aggregate biodiversity locally through various land-use The interdisciplinary team preparing this report con- practices, and the introduction of non-native species, siders the conservation of wild biodiversity extremely while simultaneously making it impossible for certain important. We also consider the situation of poor people “sensitive” species to survive there (Sax and Gaines trying to make a living from agriculture dire. And we 2003). If these sensitive species are the ones that tend believe that the demands for an adequate supply of to be lost in nearly every local situation, they become healthy, safe food for the increasing world population rare regionally, e.g., gray wolves (Canis lupus) or glob-

42 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations ally (Spix’s macaws, Cyanopsitta spixii). Therefore, if There are many studies associating biodiversity in toto one takes a larger spatial perspective, more biodiversity with agricultural systems’ productivity and resilience, may be conserved globally if certain local practices but the question is not easily resolved because of the encourage the conservation of certain sensitive species complexity and multidimensionality of ecosystems even at the expense of more common species (Lennon and the fact that biodiversity even in undisturbed eco- et al. 2004). The claim that certain agricultural prac- systems is optimized rather than maximized. Multidi- tices actually increase local biodiversity may be true, mensional optima are devilishly diffi cult to identify. but in some cases, the biota that they benefi t may not The relationships need not be symmetrical, as loss of be the biota that need aggrandizement. This point has biodiversity appears more disruptive than its enhance- been made in an agroforestry context by Schroth et al. ment from a stable system. A high research priority is (2004a). So claims that particular agricultural prac- to resolve uncertainties regarding such relationships, tices “increase biodiversity” need to be made clear: particularly between the diversity of soil biota and ag- enhancing local biodiversity that may augment local ricultural productivity (Johnson et al. 1996; Schlapfer agricultural productivity is quite different from conserv- and Schmid 1999; Loreau 2000; Griffi ths et al. 2001; ing particular species as part of local biodiversity that Bardgett 2002; Catovsky et al. 2002; Anderson and actually contributes to global or regional biodiversity Weigel 2003). in need of protection. 4.1.2. Ecosystem Health For a very concrete example, we probably have more American robins (Turdus migratorius) and northern The quantity of biodiversity is often not the most rel- cardinals (Cardinalis cardinalis) in the northeastern evant concept (Callicott and Mumford 1997). A more or U.S. than we did 300 years ago, because most “local” less complete assemblage of the native biota, interacting land-use practices inadvertently favor these species, but and functioning normally, represents ecological integ- we probably have far fewer hermit thrushes (Catharus rity, and this in turn contributes to ecosystem integrity. guttatus) and scarlet tanagers (PirangaPiranga olivaceaolivacea), which As species are lost from a system, that ecosystem’s require large areas of moderately old, contiguous for- integrity is increasingly compromised. True ecosystem est. The global conservation status of a species should integrity is likely only to be found in protected areas inform any assessment of a local practice that can affect free of most kinds of human use and, therefore, it is not a globally rare species or process if one wants to claim a realistic goal for agricultural systems. On the other that a particular local agricultural practice benefi ts hand, ecosystem health, which is determined primar- biodiversity. The term “biodiversity” includes millions ily by whether ecological sustainability is possible, is of species; some of these species are in great need of a reasonable goal for any biological system, including help, while many are not. those that are agricultural. On the other hand, the conservation of biodiversity that Ecosystem health focuses more on the thermodynamic is locally, regionally, or globally rare may not be what is properties of the system, on “multiscaled interacting being claimed. Maybe the claim is simply that one local processes, such as photosynthesis, energy transfer practice or system maintains more biodiversity than from one trophic level to the next, and nutrient cy- an alternative practice on that site. This claim is also cling” (Callicott and Mumford 1997: 37). There are relevant and important because when the conservation a myriad of human-created biological systems, e.g., of rare species is not at stake, it is usually preferable to tree farms, hayfi elds, and rubber plantations, that may have more biodiversity, even if common, than less. It exhibit ecosystem health even if not ecosystem integrity is generally believed that, all else being equal, having because they are missing most of the species native to more biodiversity contributes to greater agricultural that ecosystem type that were originally present there. productivity than an alternative system that sustains Stated logically, ecosystem health is a necessary, but less biodiversity on the same site. insuffi cient condition for ecological integrity, while ecological integrity is a suffi cient, but not necessary condition for ecosystem health.

Meeting Biodiversity Conservation Objectives 43 4.1.3. Domesticated Varieties adoption of practices that contribute to below-ground diversity creates healthier populations of below- Domesticated varieties are undoubtedly important for ground wild biodiversity, which leads to increased agricultural productivity now and in the future. Support crop production and a diversity of above-ground wild for this view is now legion (e.g., Thrupp 1998; Collins biodiversity—there would be evident incentives for and Qualset 1999; Srivastava et al. 1999; Wood and devising production strategies that benefi t farmers and Lenné 1999; Brush 2000; Kaihura and Stocking 2003). at the same time conserve wild biodiversity in agricul- The genetic and morphological diversity present in the tural landscapes. However, such a relationship is very various breeds and varieties of domesticated plants and complex and apparently equivocal, and some links are animals, although perhaps no more than a few thousand diffi cult to measure, let alone see (Adams and Wall years old, should be conserved whenever possible. In 2000; Hooper et al. 2000). Research that makes these fact, farmers in tropical countries often require, even links clearer will have many benefi ts. demand, that they have access to certain varieties of plant forms that increase their options for dealing with 4.2. A Survey of Studies on Agricultural future uncertainty (S. Padulosi, pers. comm.). Practices and Conservation of It is fairly common to fi nd statements of the following Biodiversity sort: “Traditional agriculture conserves agrobiodiver- Expanding agriculture is often cited as one of the sity and safeguards reservoirs of genetic diversity” greatest threats to wild biodiversity. The ecoagricul- (Altieri 2004: 35). But Altieri is referring here to ture approach derives from the proposition that some agrobiodiversity, not to wild biodiversity, and he is agricultural practices, at least in some contexts, can presumably referring to the genetic diversity found in maintain wild biodiversity within an area or landscape. crop varieties that farmers have chosen to cultivate in We sampled the literature for studies that would provide their fi elds. Those concerned with conserving agrobio- evidence supporting ecoagricultural claims about the diversity see their goal challenged by the infl uence of potential of agriculture to conserve wild biodiversity. seed producers and “modern agriculture” that encourage farmers to plant still fewer varieties with the promise 4.2.1. Loss of Biodiversity due to Monocropping of of higher yields, thus narrowing the genetic pool for Large Areas domesticated crops. However, the protection of wild biodiversity on agricultural lands seems an even more It is abundantly clear that agricultural practices have diffi cult and precarious task. been instrumental in reducing the amount of high-quality habitat for wild biodiversity. This is especially true 4.1.4. Overlap between Wild Biodiversity and of certain crops, e.g., bananas, palm oil, and pineapple, Agrobiodiversity grown on large acreages in tropical lowlands, where wild biodiversity would have been plentiful. Some The emphasis on biodiversity that is important to agri- of these areas would almost certainly be considered cultural production, i.e., agrobiodiversity, is different biodiversity hotspots today if millions of hectares had from the focus on wild biodiversity as defi ned for ecoag- not already been cleared for this type of monoculture riculture (McNeely and Scherr 2003). However, an area agriculture. A review of major crops and their impact where these two categories of biodiversity converge is on native fauna and fl ora has been undertaken by Clay in the soil, which on some farms may contain as much (2004:166), who suggests, for example, that “It is quite wild biodiversity as a tropical rain forest. likely that the production of sugarcane has caused a Soil organisms are known to create a healthy soil envi- greater loss of biodiversity on the planet than any other ronment conducive to better crop productivity (Brus- single agricultural crop.” saard et al. 2004), so below-ground biodiversity is likely Why is it that wildlife hate monocropping? An oblique to be encouraged by farmers and land managers through but meaningful answer goes something like this: “If you compatible agricultural practices whenever possible. build it, they will come.” The vast majority of herbi- If this link were well and widely understood—that the

44 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations vores are insects, and in fact, the number of species of ers around the world farm relatively small parcels of insects on earth probably exceeds that of all other life land for their subsistence. And because of the need to forms combined (Erwin 1982). If a ﬁ eld or a landscape hedge the risks of crop failure and assure household is comprised of one species of plant, there is a limit to survivability, farming systems are often quite diverse. the number of species of herbivores that will be attracted This means that the potential to improve productivity, to feed or oviposit eggs on those plants, even assuming economic viability, and conserve wild biodiversity on that no chemical control by humans is involved. these parcels is high.

Herbivorous insects have evolved the ability to process 4.2.2. Quantitative Studies of the Effects of the chemical compounds in the plants on which they Agricultural Practices on Wild Biodiversity feed, and especially in the tropics, each plant species represents a different suite of compounds to which We examined the agricultural and biological literature insects must adapt. For each plant species, there is for studies that document a correlation between the some fi nite number of herbivores capable of feeding implementation of a particular agricultural strategy, or on that plant. As the number of plant species in an area suite of practices, with some measure of wild biodi- increases, the number of herbivores likely to be found versity. We surveyed peer-reviewed journals, govern- among them also increases. With an increase in her- ment reports, in-house documents on experiments or bivores comes an increase in predators and parasites, fi eld trials, and other sources of written information each adapted to feeding on or parasitizing a subset of from studies that provide evidence—or an absence the herbivores found in this system. of evidence—for a claim that a particular agricultural method conserves biodiversity. Results were included It is really simple arithmetic. More species of plants only where the investigators actually quantifi ed the results in more species of herbivores, which attracts amount or kinds of biodiversity. We organized these more species of predators, parasites, and parasitoids, studies within a matrix depicting the range of both etc. As the number of invertebrate species increases, agricultural techniques and measures of biodiversity, the area becomes more attractive to vertebrates that e.g., taxa or habitat conserved. feed on them, e.g., birds, bats, frogs, and a variety of mammals. So as the number of plant species increases, We also interviewed numerous individuals responsible so does the number and variety of vertebrates that feed for, or knowledgeable about, agricultural practices and directly or indirectly on plants. their effects on biodiversity. We received many e-mails containing suggestions, references, people to contact, This considers only feeding behavior. The physical philosophical concerns, and other factors to consider. structure of the vegetative community is also important. This section summarizes the results of our literature If you have many layers of vegetation reaching from review, drawing also on any other sources of informa- the ground to many meters high, including herbaceous tion that would temper or bolster our statements based ground cover, shrubs, small trees, and canopy trees, on this literature. there are diverse assemblages of organisms that have adapted to each of these strata. With these vegetative Our sample includes 79 studies that met these criteria, strata come many more nooks and crannies, e.g., hollow although this set of studies is by no means exhaustive trees, decaying logs, and leaf masses on tree branches, (Annex 4). It should be viewed as a representative that attract and support animals to breed and nest there. sample of the kinds of results being produced by those As the arithmetic continues, biodiversity increases. It who study agriculture and its effects on wild biodi- is expected in wildlife biology that if the appropriate versity. The table is organized by the single variables habitat is available for any organisms that are adapted examined by these experimental studies, which is the to using it, they will appear in time, assuming they still approach generally used in fi eld studies, i.e., all other exist somewhere in that landscape. variables are isolated in an attempt to get an answer about the effect of perennial crops vs. annual crops on Most of the world’s agriculture is not comprised of large birds, for example. The number of column headings acreages of a single crop. Millions of small landhold-

Meeting Biodiversity Conservation Objectives 45 (n = 18) suggests the complexity of this issue for the reduction or elimination of those chemicals results real world, where several of these variables may be in higher below-ground biodiversity. extant and interacting in complex ways on the same Insofar as soil biodiversity is associated with ag- site. To make matters more complicated, the calculus of ricultural productivity, studies should be done that a particular set of interacting variables is likely to have directly assess net costs or benefi ts of alternative different effects on different taxonomic or functional production practices using chemical fertilizers. What groups of wild biodiversity. increased production costs result from chemically- Nevertheless, there are quantifi able results. We found intensive farming, or lower value of production that 9 of the 18 ecoagricultural strategies that we sur- when soil is degraded biologically? Once soil is veyed were found by at least three authors to affect being chemically fertilized, continued applications diversity of at least three taxa. The strategy most often are usually needed to sustain or raise production. correlated with the conservation of wild biodiver- (More inorganic nutrients are needed to compensate sity was the maintenance of adjacent hedgerows or for loss of organically-mobilized nutrients.) Is the woodlots. Eighteen of the 24 studies addressing that increased production from fertilizers worth more strategy documented positive correlations with eight than the possibly foregone income (higher cost of taxa plus the conservation of natural habitat. Organic production with eventually lower yield)? We need agriculture was correlated with increased diversity of reliable knowledge comparing “with” and “without” seven taxa plus habitat in eight studies. Shaded tropical systems of production, and not just in a single year agriculture, especially coffee and cacao, were found to (when the effects of previous chemical applications have higher species richness of three taxa by eight dif- are still strong) but over some number of years. This ferent studies. Most importantly, this literature matrix is a controversial subject because modern agriculture points out some patterns in the research and reveals is wedded to the use of fertilizers, but a biological some important gaps. It should therefore be useful for and an ecological perspective raise questions about strengthening the call for more research and focusing both profi tability and productivity over time, seeing it on the direct conservation benefi ts of sustainable a fi nancial value in soil biodiversity. agriculture practices. For example, many proscriptions are likely to be criticized as site-specifi c, taxon-specifi c, 4.2.3. Issues of Measurement or agronomic-specifi c, and some of those that measured To aid in the discussion of biodiversity conservation, biodiversity across taxa found inconsistent or even in- further work needs to be done to refi ne methodolo- verse correlations. Few studies included economic or gies for measuring biodiversity. Standard indices for productivity data, or linked directly to parallel studies biodiversity should be developed that allow for on- with such data. And most studies relied on local species site measurements that are relevant to regional and richness without any direct measure of impacts on or global biodiversity. These could include: 1) weighted implications for regional biodiversity. analyses that incorporate the conservation value of More broadly, a few high priority knowledge gaps endemic, rare, or threatened species; 2) measure- merit mention. First, the current debate over the link- ments of reproductive success, rather than simple age between diversity and ecosystem functioning or presence vs. absence counts; or 3) assessment of the ecosystem stability is not likely to be settled soon, impact of any local species assemblage on regional and yet important advances are being made. One of or global biodiversity. Better assessments of biodi- the most important aspects of this debate, regarding versity will be a vast improvement over the current the nexus between agriculture and conservation, is common practice of simply counting the number of the agricultural value of below-ground biodiversity. species present. Several studies have demonstrated that application of Further, creating an index or other methodology is agricultural chemicals, especially pesticides, herbicides, only the fi rst step. Improved measurements will be etc., signifi cantly reduces soil biodiversity, while the valuable only if they are widely adopted. For ex-

46 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations ample, conservation biologists have developed sophis- 4.3.1. Landscape-Level Recommendations ticated means to classify habitats. They quantify char- The following propositions are reasonable ones to acteristics such as vegetation type, spatial components guide recommendations for the management of any (size, confi guration, location within the broader matrix), agricultural landscape. and the importance to wildlife to assess the value of a particular area to conservation. Some conservationists 1) SIZE: A larger habitat patch is better than a smaller already incorporate these factors into their projects. habitat patch. Assuming that the two patches are the Conservation International (CI) is working with soy- same shape, the smaller patch will have a greater pro- bean farmers in the Brazilian Cerrado to protect the portion of its area affected by any “edge effects.” rapidly disappearing forests there. Farmers are legally 2) HETEROGENEITY: Habitat patches that are spa- required to set aside a certain percentage of their land tially and temporally heterogeneous, i.e., in terms for conservation. CI attempts to classify and prioritize of soils and abiotic factors, are usually superior to remaining forest areas and encourages farmers to set habitat patches that are homogeneous, especially aside those parcels that are most threatened or most with respect to plant species and morphological valuable for regional conservation of wild biodiversity types, i.e., herbs, shrubs, and trees. A diversity of (Conservation International 2003; B. Semroc, pers. edaphic conditions tends to result in a greater diver- comm.). Too many projects at present claim a conser- sity of plant species, which should result in turn in a vation benefi t when any “habitat” is protected, with greater diversity of animal species. little mention of what that habitat actually represents, or what species it supports. 3) SETTING: The landscape context within which the habitat patch sits is important; the less “hostile” 4.3. Principles of Habitat Management for the landscape is in general, the more effective that Maintaining Biodiversity habitat patch will be in maintaining biodiversity. For example, a patch of mature forest is more likely to Habitat fragmentation is the most serious threat to bio- maintain its biodiversity if surrounded by a young logical diversity and is the primary cause of the present secondary woodland than if surrounded by cattle extinction crisis (Wilcox and Murphy 1985). pasture. It is impossible to study every agricultural practice 4) CONNECTEDNESS: Connection among habitat and to evaluate all of their effects on biodiversity. patches is generally advantageous, so that the issue Fortunately, enough work has been done to produce of corridors must be addressed. Habitat connections some biological generalizations that seem to be true among patches are almost certainly benefi cial for the over most geographical areas (e.g., Hansson et al. dispersal of some species, though perhaps negative 1995). Biologists always end up concluding that it is for others, and neutral for most. However, the weight the quantity and quality of habitat that is the limiting of critical thinking on this issue still recommends factor for biodiversity. The following principles of re- having connections whenever possible. In addition, serve design (Meffe and Carroll 1997) can be adapted the connection itself represents habitat to some wild to thinking about agricultural landscapes, by replacing species. the word “reserve” with the word “habitat patch.” It is our belief that if you asked a representative sample 5) PREFERENCE FOR UNDISTURBED AREAS: of ecologists or conservation biologists about the fol- Natural physical units (e.g., ridges, canyons, drain- lowing statements, most would agree, for example, age basins) and modifi ed landscape elements (e.g., that larger habitat patches are better at protecting wild roads, cities, agricultural fi elds) should be identi- biodiversity than smaller habitat patches. fi ed. Having a diversity of natural elements in any landscape will enhance its biodiversity value, while modifi ed elements detract from that value. Distur- bance of almost any kind encourages the invasion of non-native species and weedy species. This does

Meeting Biodiversity Conservation Objectives 47 not mean that modifi ed landscapes are invariably Because biodiversity is multi-dimensional (4.1.2), hostile to biodiversity but rather that having more evaluating its extent and status, as well as changes in undisturbed areas confers biodiversity benefi t. these, is challenging. If the processes and dynamics that produce these outcomes are regarded as a kind of 6) MODERATION OF SHARP CONTRASTS: Buffer “black box,” making assessments becomes that much zones that “soften” the ecotone between the habitat more diffi cult. Accordingly, it is important for environ- patch and modifi ed landscape elements should be mentalists and agriculturalists to have some agreement encouraged. All natural habitat patches are infl u- on the “intermediate” factors so that these can inform enced by the matrix in which they are found; the less evaluations, knowing something about mechanisms difference there is between the patch and its matrix, and causal infl uences for what are admittedly very the better. complicated and inexact outcomes. 4.3.2. Additional Principles 4.4. Intensifi cation of Agriculture and the In addition, four general principles of good conserva- Conservation of Biodiversity tion management can be suggested (Meffe and Carroll 1997): The primary question here is whether intensifi cation of agriculture in some locations saves natural habitat 1) Critical ecological processes and biodiversity com- elsewhere from being converted to food production, an position must be maintained. Of course, these have argument put forth by Pagiola and Kellenberg (1997) to be identifi ed before they can be maintained, and and Borlaug (1998). This issue was a major concern at determining what process is critical is not usually the DIVERSITAS workshop/scoping meeting held in simple. Alexandria, Egypt in May 2004. However, there was no 2) External threats must be minimized and external agreement on a defi nition of the term “intensifi cation.” benefi ts maximized. The greatest external threat Some believed it should be defi ned by the level of inputs to a patch of natural habitat, aside from destroying per hectare and others by the yield per hectare. The term it outright, is the invasion of non-native species or can be and is used differently by different disciplines some form of pollution. Indigenous people often and individuals. We presented the three main uses of the represent an external threat to the remaining habitat term at the end of Chapter 1. A World Bank publication patch, but their knowledge of the local fauna and states: “A basic principle of biodiversity-friendly policy fl ora also represents a benefi t. reforms is to discourage agricultural extensifi cation and encourage agricultural intensifi cation.” (Pagiola and 3) Evolutionary processes must be preserved. This is Kellenberg 1997: viii, emphasis in original). a diffi cult principle to implement, because of the long time horizon that must be adopted. It is always If by intensifi cation, one means that the agricultural a worthy goal, however, and can be approached by fi elds are so heavily manipulated physically and chemi- maintaining as many of the biological elements of cally to increase yields at the expense of nearly all wild the system as possible. Refer to the goal statement biodiversity on that site, then the entire justifi cation in for conservation biology in section 1.6.1. biodiversity terms must rest on this increase in yield saving natural habitat elsewhere since the production 4) Management must be adaptive and minimally intru- process in this manner reduces or degrades habitat in sive. This really applies particularly to natural areas and near the site of production. There may be many that are actively managed for wild biodiversity, but reasons why natural habitats are still being converted the general advice is wise. Tread softly, proceed besides the need to produce more food, so modern, slowly, and evaluate what you do as you go. energy- and chemical-intensive agriculture should not These various propositions and principles provide some be blamed for all habitat loss. However, because so theoretical foundations for assessing different agricul- much of the world’s land area is employed in agricultural systems/practices with regard to their contribution tural production, it seems unnecessary and unwise to to, or compatibility with, biodiversity conservation. “write off” wild biodiversity on the 40 percent of the

48 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations world’s surface in agriculture with the hopes that the For example, wetlands are known to remove excess remaining natural habitats will become repositories for nitrate deposited from the atmosphere or moving from all the remaining fauna and fl ora on earth not exploited agricultural fi elds before it reaches streams and lakes. for agricultural purposes.1 If nothing else, agricultural However, without the presence of denitrifying bacteria, lands are potential connections among the remaining the removal does not occur (Toet et al. 2003). Simply natural habitats (see item 4 in 4.3.1). Therefore one looking at two wetlands, one with and one without these can argue that this function, at a minimum, should be important bacteria, would not reveal which wetland encouraged (Kreuss and Tscharntke 1994). contains these essential organisms. Simply observing some habitat or agricultural system does not reveal the On the other hand, if intensifi cation means increas- critical evidence on whether that system has the health ing yields by some means other than the high-input, or integrity to provide ecosystem services other than monoculture approach, e.g., by crop diversifi cation, the most basic, i.e., all plants remove carbon dioxide then there is potential for on-site conservation of wild from the air. Rigorous measurements or studies need to biodiversity as well as agrobiodiversity. A recent study be conducted. Although Annex 4 lists examples where in northern Greece found yet again that the species rich- agricultural practices are compatible with or even ness of woody plants was the best overall bioindicator conserve various forms of wild biodiversity, it would for several groups of plant and animal biodiversity (Kati be extremely valuable to produce a comparable table et al. 2004). This was compatible with the diversifi ed showing documented examples of how wild biodiver- agriculture practiced in that region. One can say that if sity specifi cally aids crop production. intensifi cation involves diversifying crop varieties, with encouragement of a diversity of native woody plants, Although much is still to be learned about this, many then there can be overall increase in species diversity. are sure that the ecosystem services provided by wild Again, the exact approach to “intensifi cation” makes a biodiversity are advantageous to those trying to produce difference in biodiversity outcomes. On this, Edwards food (Paoletti and Pimentel 1992; Letourneau 1997; et al. (1999) offer useful insights. FAO 2003). However, the debate continues over how much and what kinds of biodiversity are necessary to 4.4.1. Ecosystem Services enable ecosystems to function properly. As land use But are there stronger reasons for agriculture to foster intensifi es, Loreau et al. (2001) believe that a larger populations of wild biodiversity on agricultural lands? pool of species is required for proper assembly and That nature can provide benefi ts to humans simply functioning of ecosystems. The hypothesis cannot be by allowing ecosystems to function naturally is now rejected however that a few dominant species can pro- commonly accepted (Costanza et al. 1997; Daily et al. vide suffi cient functional diversity to explain primary 1997; Daily and Ellison 2002). Ecosystem services are production in grassland ecosystems. “ecological processes that benefi t people” (Luck 2003). Again, much of the purported value of biodiversity in These include maintaining hydrologic and nutrient this regard is below-ground. In addition to the usual sus- cycles, providing benefi cial interspecifi c relationships pects that are known to be important to soil health (e.g., (pollination, pest control), and sequestering, fi ltering bacteria, fungi, earthworms, etc.), ants, which spend or degrading contaminants, to name a few. For these time below and above ground are also important. Ants services to be available, viable populations of native score second in animal turbation only to earthworms, biota need to be present and fulfi lling their role in that but ants have a wider geographical distribution (Fol- system (Ostfeld 2003). garait 1998). In Argentina, for example, Camponotus punctulatus move 2100 kilograms per hectare per year 1 This discussion of agriculture is, obviously, focused on of soil in sown pastures to construct their mounds. On fi eld-crop production, where the issue of intensifi cation the negative side of the ledger, most forms of agriculture and monoculture is most evident; there are also some or other disturbances examined (except fi re in Austra- intensive uses of aquatic and marine ecosystems. The lia) have resulted in a decrease in ant species richness same issues apply in those exploited areas. (Folgarait 1998, Table 2).

Meeting Biodiversity Conservation Objectives 49 4.5. Conserving Wild Biodiversity in of America (Christensen et al. 1996; Dale et al. 2000). Agricultural Landscapes Understanding what needs to be done to conserve wild biodiversity in agricultural landscapes is diffi cult, but The results of studies listed in Annex 4 show that enacting the appropriate changes on the ground without some agricultural approaches are more benign than compromising economic and productivity gains will others with respect to conserving wild biodiversity. be even more diffi cult. Any gain in conserving wild For example, the most studies that we found actually biodiversity without signifi cant losses in the other documenting conservation benefi ts were studies on two spheres should be reason to cheer, and how to the value of maintaining hedgerows, windbreaks, or accomplish that constitutes the greatest information natural habitat adjacent to agricultural fi elds (also see gap in ecoagriculture. As seen in Chapter 3, there are Schroth et al. 2004b). Similarly, organic agriculture, and promising areas of research and practice, but selecting, shaded tropical agriculture have been well-documented adapting, and welding them into large-scale application to maintain more species at all or nearly all taxonomic remains a challenge to scientists, policy-makers, and levels than the standard modern practices that these agriculturalists alike. ecoagriculture strategies replace. However, those are simply the practices that have been best studied. Perhaps we need to think even further out of the Therefore we conclude that there should be increased box. Would it not be possible to create entirely new research on other practices, rather than endorse the habitats on agricultural land, which encompass both practices we know most about. Further, most of these cultivated and wild plants, herbaceous and woody studies have relied on simple on-site species counts, plants in edible, medicinal, and structural forms, highly and therefore are missing a key facet of biodiversity cultivated areas and those that are much less so, with conservation: impacts on regional or global diversity. a diversity of domesticated livestock along with wild More studies are needed to provide this link to deter- vertebrates and invertebrates? It would mean thinking mine whether agricultural systems implementing these about landscapes anew, almost as if they were devoid strategies reduce or encourage the conversion of more of life, like many degraded agricultural environments, natural habitat. and intentionally designing landscapes that purpose- fully include wild and agrobiodiversity, with humans This should be no surprise that different agricultural who live off the products of this design? practices should result in different effects on native fl ora and fauna in localities, given that those practices foster Such an Edenesque result would require community and different sets of ecological interactions. Determining population ecologists, agronomists, landscape ecolo- such results is demanding in terms of data and better gists, and even landscape architects, horticulturists, methodologies, but even doing so is a long way from city and regional planners, working closely with local demonstrating that introducing “better” agricultural residents who have indigenous and practical knowledge. practices on farms will have signifi cant consequences The novel result might protect more biodiversity than for local or regional biodiversity. In The Netherlands, the present piecemeal, “hold onto what you can afford for example, agri-environment schemes have been to” approach espoused now. Many people spend their established, but overall landscapes are so affected by lives trying to design towns and cities that are physi- intensive agricultural uses in general that the on-farm cally and psychologically healthy for their inhabitants, schemes have had little positive effect thus far on wild and that function effi ciently. Why should a diversity biodiversity relative to nearby fi elds not under the of scientists and farmers not attempt to design agri- conservation schemes (Kleijn et al. 2004). cultural landscapes from the ground up with the same kind of goals? Actually, this idea is decades old. It was There is a danger in concentrating too intently on indi- proposed by Wilson and Willis (1975) as a way to at vidual fi elds, or plots or crops. The context within which least think about designing future nature reserves in a those units exist is extremely important. The principles degraded world. They termed this approach “applied governing these relations have been well articulated in biogeography.” To initiate such a process would require recent reviews sanctioned by the Ecological Society a great deal of knowledge about species interactions,

50 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations the compatibility of domestic and wild species of plants Clay, J. 2004. World Agriculture and the Environment: and animals, the appropriate spatial confi guration of A Commodity-by-Commodity Guide to Impacts and land uses for maintaining viable populations of wild Practices. Washington, DC: Island Press. species in an agricultural matrix, etc. The questions Christensen, N.L., A. M. Bartuska, J. H. Brown, S. Car- that need to be answered are not entirely new, but the penter, C. D’Antonio, R. Francis, J. F. Franklin, J. A. novel context within which these questions would be MacMahon, R. F. Noss, D. J. Parsons, C. H. Peterson, asked might lead in unexpected directions. M. G. Turner, and R. G. Woodmansee. 1996. The re- REFERENCES port of the Ecological Society of America Committee on the Scientifi c Basis for Ecosystem Management. Adams, G. A. and D. H. Wall. 2000. Biodiversity above Ecological Applications 6: 665-691. and below the surface of soils and sediments: Link- Collins, W. W. and C. O. Qualset, eds. 1999. Biodi- ages and implications for global change. Bioscience. versity in Agroecosystems. Boca Raton, FL: CRC 50:1043-1048. Press. Altieri, M.A. 2004. Linking ecologists and traditional Conservation International. 2003. The Center for Envi- farmers in the search for sustainable agriculture. ronmental Leadership in Business. Washington, DC: Frontiers in Ecology and the Environment. 2: 35- Conservation International. 42. Costanza, R., R. d’Arge, R. de Groot, S. Farber, M. Anderson, T. H. and H. J. Weigel. 2003. On the current Grasso, B. Hannon, K. Limburg, S. Naeem, R. V. debate about soil biodiversity. Landbauforsch. Volk. O’Neill, J. Paruelo, R. G. Raskin, P. Sutton and 53: 223-233. M. van den Beltet. 1997. The value of the world’s Bardgett, R .D. 2002. Causes and consequences of bio- ecosystem services and natural capital. Nature. 387: logical diversity in soil. Zoology. 105: 367-374. 253-260. Brush, S. B.,ed. 2000. Genes in the Field: On-farm Daily, G. C. and K. Ellison. 2002. The New Economy of Conservation of Crop Diversity. Boca Raton, FL: Nature: The Quest to Make Conservation Profi table. Lewis Publishers. Washington, D.C.: Island Press/Shearwater Books. Brussaard, L., T. W. Kuyper, W. A. M. Didden, R. G. Daily, G. C., J. S. Reichert and J. P. Myers, eds. 1997. M. de Goede and J. Bloem. 2004. Biological soil Nature’s Services: Societal Dependence on Natural quality from biomass to biodiversity: Importance Ecosystems. Washington, DC: Island Press. and resilience to management stress and disturbance. Dale, V. H., S. Brown, R. A. Haeuber, N. T. Hobbs, N. In: P. Schjønning, S. Elmholt and B. T. Christensen, Huntly, R. J. Naiman, W. E. Riebsame, M. G. Turner, eds., Managing Soil Quality: Challenges in Mod- and T. J. Valone. 2000. Ecological principles and ern Agriculture, 139-161. New York, NY: CABI guidelines for managing the use of land. Ecological Publishing. Applications 10: 639-670. Bullock, J. M., R. F. Pywell, M. J. W. Burke and K. J. Edwards, P. J., J. Kollmann and D. Wood. 1999. Deter- Walker. 2001. Restoration of biodiversity enhances minants of agrobiodiversity in the agricultural land- agricultural production. Ecological Letters. 4: 185- scape. In: W. D. and J. M. Lenné, eds., Agrobiodiver- 189. sity: Characterization, Utilization and Management, Callicott, J.B. and K. Mumford. 1997. Ecological sus- 183-210. Wallingford, UK: CAB International. tainability as a conservation concept. Conservation Erwin, T. I. 1982. Tropical forests: Their richness in Biology. 11:32-40. Coleoptera and other arthropod species. Coleopter- Catovsky, S., M.A. Bradford and A. Hector. 2002. Bio- ists’ Bulletin 36: 74-75. diversity and ecosystem productivity: Implications for carbon storage. Oikos. 97: 443-448.

Meeting Biodiversity Conservation Objectives 51 FAO. 2003. Biodiversity and the Ecosystem Approach Kreuss, A. and T. Tscharntke. 1994. Habitat fragmen- in Agriculture, Forestry and Fisheries. Rome: Inter- tation, species loss, and biological control. Science. departmental Working Group on Biological Diversity 264: 1581-1584. for Food and Agriculture, Food and Agriculture Lennon, J. J., P. Koleff, J. J. D. Greenwood and K. J. Organization. Gaston. 2004. Contribution of rarity and common- Folgarait, P.J. 1998. Ant biodiversity and its relationship ness to patterns of species richness. Ecology Letters. to ecosystem functioning: A review. Biodiversity and 7: 81-87. Conservation. 7:1221-1244. Letourneau, D. K. 1997. Plant-arthropod interactions Griffi ths, B.S., K. Ritz, R. Wheatley, H.L. Kuan, B. in agroecosystems. In: L. E. Jackson, ed., Ecology in Boag, S. Christensen, F. Ekelund, S.J. Sorensen, S. Agriculture, 239-290. San Diego: Academic Press. Muller and J. Bloem. 2001. An examination of the Loreau, M. 2000. Biodiversity and ecosystem function- biodiversity-ecosystem function relationship in ar- ing: Recent theoretical advances. Oikos. 91: 3-17. able soil microbial communities. Soil Biology and Biochemistry. 33: 1713-1722. Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J.P. Grime, A. Hector, D. U. Hooper, M. A. Huston, D. Hansson, L., L. Fahrig and G. Merriam, eds. 1995. Raffaelli, B. Schmid, D. Tilman and D. A. Wardle. Mosaic Landscapes and Ecological Processes. New 2001. Biodiversity and ecosystem functioning: Cur- York: Chapman and Hall. rent knowledge and future challenges. Science. 294: Hooper, D. U., D. E. Bignell, V. K. Brown, L. Brus- 804-808. saard, J. M. Dangerfi eld, D. H. Wall, D. A. Wardle, Luck, G. W., G. C. Daily and P. R. Ehrlich. 2003. Popu- D. C. Coleman, K. E. Giller, P. Lavelle, W. H. Van lation diversity and ecosystem services. Trends in der Putten, P. C. de Ruiter, J. Rusek, W. L. Silver, Ecology and Evolution. 18: 331-336. J. M. Tiedje and V. Wolters. 2000. Interactions between aboveground and belowground biodiversity McNeely, J. A. and S. J. Scherr. 2003. Ecoagriculture: in terrestrial ecosystems: Patterns, mechanisms, and Strategies to Feed the World and Save Wild Biodi- feedbacks. Bioscience. 50: 1049-1061. versity. Washington, DC: Island Press. Johnson, K. H., K. A. Vogt, H. J. Clark, O. J. Schmitz Meffe, G. K. and C. R. Carroll. 1997. Principles of and D. J. Vogt. 1996. Biodiversity and the productiv- Conservation Biology. Sunderland, MA: Sinauer ity and stability of ecosystems. Trends in Ecological Associates. Evolution. 11: 372-377. Ostfeld, R. S. and K. LoGiudice. 2003. Community Kaihura, F. and M. Stocking, eds. 2003. Agricultural disassembly, biodiversity loss, and the erosion of an Biodiversity in Smallholder Farms in East Africa. ecosystem service. Ecology. 84: 1421-1427. Tokyo: U.N. University Press. Pagiola, S., J. Kellenberg, L. Vidaeus and J. Srivastava. Kati, V., P. Devillers, M. Dufrêne, A. Legakis, D. Vokou, 1997. Mainstreaming Biodiversity in Agricultural and P. Lebrun. 2004. Testing the value of six taxo- Development. Environment Paper 15. Washington, nomic groups as biodiversity indicators at a local DC: World Bank. scale. Conservation Biology. 18: 667-675. Paoletti, M. G. and D. Pimentel,eds. 1992. Biotic Diver- Kleijn, D., F. Berendse, R. Smit, N. Gilissen, J. Smit, sity in Agroecosystems. Amsterdam: Elsevier. B. Brak, and R. Groeneveld. 2004. Ecological ef- Pimentel, D., C. Wilson, et al. 1997. Economic and fectiveness of agri-environment schemes in different environmental benefi ts of biodiversity. Bioscience. agricultural landscapes in The Netherlands. Conser- 47: 747-757. vation Biology. 18: 775-786. Sax, D. F. and S. D. Gaines. 2003. Species diversity: From global decreases to local increases. Trends in Ecology and Evolution. 18: 561-566.

52 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Schlapfer, F. and B. Schmid. 1999. Ecosystem effects of biodiversity: A classifi cation of hypotheses and exploration of empirical results. Ecological Applica- tions. 9: 893-912. G. Schroth, G. A. B. da Fonseca, C. A. Harvey, H. L. Vasconcelos, C. Gascon, and A-M. N. Izac, eds. 2004a. Agroforestry and Biodiversity Conservation in Tropical Landscapes. Washington, DC: Island Press. Schroth, G., G. A. B. da Fonseca, C. A. Harvey, H. L. Vasconcelos, C. Gascon, and A-M. N. Izac. 2004b. Introduction: The role of agroforestry in biodiversity conservation in tropical landscapes. In: G. Schroth, G. A. B. da Fonseca, C. A. Harvey, H. L. Vasconce- los, C. Gascon, and A-M. N. Izac, eds. Agroforestry and Biodiversity Conservation in Tropical Land- scapes, 1-14. Washington, DC: Island Press. Soule, J.D. 2002. Conservation and agriculture as neighbors. In: D. Jackson and L. Jackson, eds., The Farm as Natural Habitat: Reconnecting Food Sys- tems with Ecosystems, 169-188. Washington, DC: Island Press. Srivastava, J. P., N. J. H. Smith and D. A. Forno. 1999. Integrating Biodiversity in Agricultural Intensifi cation: Toward Sound Practices. Washington, DC: World Bank. Thrupp, L. A. 1998. Cultivating Diversity: Agrobiodi- versity and Food Security. Washington, DC: World Resources Institute. Tilman, D. 1999. The ecological consequences of changes in biodiversity: A search for general principles. Ecology 80: 1455-1474. Toet, S., L. Huibers, R. S. P. Van Logtestijn and J. T. A. Verhoeven. 2003. Denitrifi cation in the periphyton associated with plant shoots and in the sediment of a wetland system supplied with sewage treatment plant effl uent. Hydrobiologia. 501: 29-44. Wilcox, B. A. and D. D. Murphy. 1985. Conservation strategy: The effects of fragmentation on extinction. American Naturalist. 125: 879-887. Wood, D. and J.M. Lenné, eds. 1999. Agrobiodiversity: Characterization, Utilitization and Management. Wallingford, UK: CAB International.

Meeting Biodiversity Conservation Objectives 53 ANNEX 4: Examples of Studies that Examine the Quantitative Relationship Between Various Agricultural Practices and Wild Biodiversity

Ecoagricultural strategy: agricultural practice claimed to conserve biodiversity Other Taxa or surrogate of Other Other mixed Perennial Low- No- Increased Shaded tropical ag. Trees in pastures grassland diversity measured agroforestry cropping crops tillage tillage fallow mgt Mammals Birds 76, 75, 34, 58, 54, 46 28, 39 11 72 57 Ants 52, 53, 54, 3 24 Other vert.species Soil invert.species 44 63 43 Other invert.species 54, but no 60 24 65, 1 68, 42 68 20, 45, 56 Soil microbes 68 27 68 Soil biomass 27 74 74 Soil organic matter 64 72, 32 74 74 Trees 64 8 Other plants 64 39 10, 7

Ecoagricultural strategy: agricultural practice claimed to conserve biodiversity Other Post- Lower Taxa or surrogate of Reduced Managed Eco-certi- Adjacent hedgerow, Landscape level Organic Euro harvest intensity diversity measured chemicals flooding fication forest. practice innovation treatment ag. Mammals 67, 16; but no 13 Birds 57, 69 9, 10, 29 10, 15 6, 26, 25, 71 48, 57 37, 19, 48, 13 59, 66, 48, 57 Ants 53 17, 26 24, 3 Other verts. 36 31 Soil inverts. 50 49 6 50 24, 13, 2, 4, 56, 60, 66, 18, 32, 73, 41, but Other inverts. 4 but no 73 70, 22, 41 mixed 44, no 42 Soil microbes 50, 14, 18 49 50 Soil biomass 55 74 Soil organic matter 61 74 62 Trees but no 35 38, 8, 30; but no 13 30 37, 78, 30, 12; but no Other plants 10 5, 73 10 77, 12 77, 66, 30, 73 13 Annex 4- 1 Literature Matrix Index

Code Citatio

1 Abate, T. 1991. Intercropping and weeding effects on some natural enemies of African Bollworm Heliothis armigera hbn. Lepidoptera Noctuidae in bean fields. Journal of Applied Entomology 112(1):38-42. 2 Abensperg-Traun, M. and G.T. Smith. 1999. How small is too small for small animals? Four terrestrial arthropod species in different-sized remnant woodlands in agricultural Western Australia. Biodiversity and Conservation 8:709-726. 3 Armbrecht, I. and I. Perfecto. 2003. Litter-twig dwelling ant species richness and predation potential within a forest fragment and neighboring coffee plantations of contrasting habitat quality in Mexico. Agricultural Ecosystems and Environment 97(1-3):107-115. 4 Asteraki, E.J., C.B. Hanks and R.O. Clements. 1992. The impact of the chemical removal of the hedge-base flora on the community structure of carabid beetles Col. Carabidae and spiders Araneae of the field and hedge bottom. Journal of Applied Entomology 113(4):398-406. 5 Aude, E., K. Tybirk, and M.B. Pedersen. 2003. Vegetation diversity of conventional and organic hedgerows in Denmark. Agricultural Ecosystems and Environment.1-13. 6 Ausden, M., W.J. Sutherland and R. James. 2001. The effects of flooding lowland wet grassland on soil macroinvertebrate prey of breeding wading birds. Journal of Applied Ecology 38(2):320-338. 7 Austrheim, G. and O. Eriksson. 2001. Plant species diversity and grazing in the Scandinavian mountains - patterns and processes at different spatial scales. Ecography. 24(6):683-695. 8 Backes, M.M. 2001. The role of indigenous trees for the conservation of biocultural diversity in traditional agroforestry land use systems: The Bungoma case study: In-situ conservation of indigenous tree species. Agroforestry Systems. 52(2):119-132. 9 Beecher, N.A., R.J. Johnson, J.R. Brandle, R.M. Case and L.J. Young. 2002. Agroecology of birds in organic and nonorganic farmland. Conservation Biology 16(6):1620-1631. 10 Berendse, F., M.J.M. Oomes, H.J. Altena and W.T. Elberse. 1992. Experiments on the restoration of species-rich meadows in the Netherlands. Biological Conservation 62(1):59-65. 11 Bird, J.A., G.S. Pettygrove and J.M. Eadie. 2000. The impact of waterfowl foraging on the decomposition of rice straw: Mutual benefits for rice growers and waterfowl. Journal of Applied Ecology 37:728-741. 12 Boutin, C., B. Jobin, C. Bonger and L. Choini. 2002. Plant diversity in three types of hedgerows adjacent to cropfields. Biodiversity and Conservation. 12:1-25. 13 Burel, F., J. Baudry, A. Butet, P. Clergeau, Y. Delettre, D. Le Coeur, F. Dubs, N. Morvan, G. Paillat and S. Petit. 1998. Comparative biodiversity along a gradient of agricultural landscapes. Acta Oecologica. 19(1):47-60. 14 Caballero-Mellado, J. and E. Martinez-Romero. 1999. Soil fertilization limits the genetic diversity of Rhizobium in bean nodules. Symbiosis. 26(2):111-121.

Annex 4- 2 15 Cuenca, G., R. Herrera and E. Meneses. 1991. Vesicular-arbuscular mycorrhizae in cacao plantations in Venezuela. Acta-Cientifica-Venezolana. 42(3):153-159. 16 Daily, G.C., G. Ceballos, J. Pacheco, G. Suzan and A. Sanchez-Azofeifa. 2003. Countryside biogeography of neotropical mammals: Conservation opportunities in agricultural landscapes of Costa Rica. Conservation Biology 17(6):1814-1826. 17 Day, J.H. and M.A. Colwell. 1998. Waterbird communities in rice fields subjected to different post-harvest treatments. Colonial Waterbirds. 21:185-197. 18 den Belder, E., J. Elderson, W.J. van den Brink and G. Schelling. 2002. Effect of woodlots on thrips density in leek fields: a landscape analysis. Agricultural Ecosystems and Environment 91(1-3):139-145. 19 Deschenes, M., L. Belanger and J.F. Giroux. 2003. Use of farmland riparian strips by declining and crop damaging birds. Agricultural Ecosystems and Environment 95(2-3):567-577. 20 Di Giulio, M., P.J. Edwards and E. Meister. 2001. Enhancing insect diversity in agricultural grasslands: the roles of management and landscape structure. Journal of Applied Ecology 38(2):310-319. 21 Donald, P.F., R.E. Green and M.F. Heath. 2001. Agricultural intensification and the collapse of Europe's farmland bird populations. Proceedings of the Royal Society of London B. 268:25-29. 22 Dover, J., T. Sparks, S. Clarke, K. Gobbett and S. Glossop. 2000. Linear features and butterflies: the importance of green lanes. Agricultural Ecosystems and Environment 80:227-242. 23 Duelli, P. and M.K. Obrist 1998. In search of the best correlates for local organismal biodiversity in cultivated areas. Biodiversity and Conservation. 7(3):297-309. 24 Dunn, R.R. 2000. Isolated trees as foci of diversity in active and fallow fields. Biological Conservation 95(3):317-321. 25 Elphick, C.S. and L.W. Oring. 1998. Winter management of Californian rice fields for waterbirds. Journal of Applied Ecology. 35(95-108). 26 Elphick, C.S. and L.W. Oring. 2003. Conservation implications of flooding rice fields on winter waterbird communities. Agriculture, Ecosystems and Environment. 94:17-29. 27 Ferguson, H.J. and R.M. McPherson. 1985. Abundance and diversity of adult carabidae in four soybean cropping systems in Virginia, USA. Journal of Entomological Science. 20(2):163-171. 28 Fischer, J. and D.B. Lindenmayer. 2002. The conservation value of paddock trees for birds in a variegated landscape in southern New South Wales. 1. Species composition and site occupancy patterns. Biodiversity and Conservation. 12:807-832. 29 Freemark, K.E. and D.A. Kirk. 2001. Birds on organic and conventional farms in Ontario: partitioning effects of habitat and practices on species composition and abundance. Biological Conservation 101(3):337-350. 30 Freemark, K.E., C. Boutin and C.J. Keddy. 2002. Importance of Farmland Habitats for Conservation of Plant Species. Conservation Biology. 16(2):399-412. 31 Glor, R.E., A.S. Flecker, M.F. Benard and A.G. Power. 2001. Lizard diversity and agricultural disturbance in a Caribbean forest landscape. Biodiversity and Conservation. 10(5):711-723. Annex 4- 3 32 Glover, J. 2003. Characteristics of Annual vs Perennial Systems. The Land Institute. Salina, Kansas. 33 Gobbi, J.A. 2000. Is biodiversity-friendly coffee financially viable? An analysis of five different coffee production systems in western El Salvador. Ecological Economics. 33(2):267-281. 34 Greenberg, R., P. Bichier and A.C. Angon. 2000. The conservation value for birds of cacao plantations with diverse planted shade in Tabasco, Mexico. Animal Conservation. 3:105-112. 35 Gullison, R.E. 2003. Does forest certification conserve biodiversity? Oryx. 36(2):153-165. 36 Hamer, A.J., J.A. Makings, S.J. Lane and M.J. Mahony. 2004. Amphibian decline and fertilizers used on agricultural land in southe-eastern Australia. Agricultural Ecosystems and Environment 102:299-304. 37 Harvey, C.A. 2000. Windbreaks enhance seed dispersal into agricultural landscapes in Monteverde, Costa Rica. Ecological Applications 10(155- 173). 38 Harvey, C.A. and W.A. Haber. 1998. Remnant trees and the conservation of biodiversity in Costa Rican pastures. Agroforestry Systems. 44(1):37-68. 39 Harvey, C.A., C.F. Guindon, W.A. Haber, D.H. DeRosier and K.G. Murray. 2000. The importance of forest patches, isolated trees and agricultural windbreaks for local and regional biodiversity: the case of Monteverde, Costa Rica. XXI IUFRO World Congress, 7-12 August 2000, Subplenary sessions,. Kuala Lumpur, Malaysia. Vol 1. 787-798. 40 Hietala-Koivu, R., T. Jarvenpaa and J. Helenius. 2004. Value of semi-natural areas as biodiversity indicators in agricultural landscapes. Agricultural Ecosystems and Environment 101(1):9-19. 41 Horner-Devine, M.C., G.C. Daily, P.R. Ehrlich and C.L. Boggs. 2003. Countryside biogeography of tropical butterflies. Conservation Biology 17(1):168-177. 42 Jeannert, P., B. Schüpbach and H. Luka. 2003. Quantifying the impact of landscape and habitat features on biodiversity in cultivated landscapes. Agricultural Ecosystems and Environment 98:311-320. 43 Kandji, S.T., C. Ogol and A. Albrecht. 2001. Diversity of plant-parasitic nematodes and their relationships with some soil physico-chemical characteristics in improved fallows in western Kenya. Applied Soil Ecology 18(2):143-157. 44 Klein, A.M., I. Steffan-Dewenter, D. Buchori and T. Tscharntke. 2002. Effects of Land-Use Intensity in Tropical Agroforestry Systems on Coffee Flower-Visiting and Trap-Nesting Bees and Wasps. Conservation Biology. 16(4):1003-1014. 45 Kruess, A. and T. Tscharntke. 2002. Grazing Intensity and the Diversity of Grasshoppers, Butterflies, and Trap-Nesting Bees and Wasps. Conservation Biology. 16(6):1570. 46 Lindell, C. and M. Smith. 2003. Nesting bird species in sun coffee, pasture, and understory forest in southern Costa Rica. Biodiversity and Conservation. 12(3):423-440. 47 Lohr, L. 2002. Benefits of U.S. Organic Agriculture. Thesis (Ph.D). - University of Georgia. Agricultural and Applied Economics. Athens, Georgia.

Annex 4- 4 48 Luck, G.W. and G.C. Daily. 2003. Tropical countryside bird assemblages: Richness, composition, and foraging differ by landscape context. Ecological Applications 13(1):235-247. 49 Mäder, P., A. Fliebach, D. Dubois, L. Gunst, P. Fried and U. Niggli. 2002. Soil Fertility and Biodiversity in Organic Farming. Science. 296(5573):1694-1697. 50 Neher, D.A. 1999. Soil community composition and ecosystem processes - Comparing agricultural ecosystems with natural ecosystems. Agroforestry Systems. 45(1-3):159-185. 51 Ovenden, G.N., A.R.H. Swash and D. Smallshire. 1998. Agri-environment schemes and their contribution to the conservation of biodiversity in England. Journal of Applied Ecology 35(6):955-960. 52 Perfecto, I. and J. Vandermeer. 2002. Quality of agroecological matrix in a tropical montane landscape: Ants in coffee plantations in southern Mexico. Conservation Biology 16(1):174-182. 53 Perfecto, I. and R. Snelling. 1995. Biodiversity and the Transformation of a Tropical Agroecosystem - Ants in Coffee Plantations. Ecological Applications 5(4):1084-1097. 54 Perfecto, I., A. Mas, T. Dietsch and J. Vandermeer. 2003. Conservation of biodiversity in coffee agroecosystems: a tri-taxa comparison in southern Mexico. Biodiversity and Conservation 12(6):1239-1252. 55 Petersen, C., L. Drinkwater and P. Wagone. 1999. The Rodale Institute Farming Systems Trial: The first 15 years. 40. Rodale Institute, Kutztown, PA. 56 Petit, S. and M.B. Usher. 1998. Biodiversity in agricultural landscapes: the ground beetle communities of woody uncultivated habitats. Biodiversity and Conservation. 7(12):1549-1561. 57 Ratcliffe, C.S. and T.M. Crowe. 2001. The effects of agriculture and the availability of edge habitat on populations of Helmeted Guineafowl Numida meleagris and on the diversity and composition of associated bird assemblages in KwaZulu-Natal province, South Africa. Biodiversity and Conservation. 10(12):2109-2127. 58 Reitsma, R., J.D. Parrish and W. McLarney. 2001. The role of cacao plantations in maintaining forest avian diversity in southeastern Costa Rica. Agroforestry Systems. 53(2):185-193. 59 Remsen, J.V., M.M. Swan, S.W. Cardiff and K.V. Rosenberg. 1991. The importance of the rice-growing region of south-central Louisiana to winter populations of shorebirds, raptors, waders, and other birds. Journal of Louisiana Ornithology. 1:35-47. 60 Ricketts, T.H., G.C. Daily, P.R. Ehrlich and J.P. Fay. 2001. Countryside Biogeography of Moths in a Fragmented Landscape: Biodiversity in Native and Agricultural Habitats. Conservation Biology. 15(2):378-388. 61 Rodale Institute, The. 2000. Farming Systems Trial: The First 15 Years. Rodale Institute, Kutztown, PA. 62 SARE. 2003. Good Bedfellows: Cattle, Pecan Trees an Environmentally Sound Mix. as cited in SARE Internal Document - 2003 annual report. Sustainable Agriculture Research and Education, Cooperative State Research, Education, and Extension Service, US Department of Agriculture. Washington, DC.

Annex 4- 5 63 Schmidt, O., R.O. Clements and G. Donaldson. 2003. Why do cereal-legume intercrops support large earthworm populations? Appl. Soil Ecology 22(2):181-190. 64 Siebert, S.F. 2002. From shade- to sun-grown perennial crops in Sulawesi, Indonesia: implications for biodiversity conservation and soil fertility. Biodiversity and Conservation. 11(11):1889-1902. 65 Stinner, B.R. and A. Tonhasca, Jr. 1991. Influences of shelterbelts and other alternate vegetation in agricultural landscapes on invertebrate ecology. USDA - USFS. 17-23. 66 Svensson, B.S.B., K. Vessby and A. Glimsky. 2001. Plants, insects and birds in semi-natural pastures in relation to local habitat and landscape factors. Biodiversity and Conservation. 11:1839-1863. 67 Tattersall, F.H., D.W. Macdonald, B.J. Hart, P. Johnson, W. Manley and R. Feber. 2002. Is habitat linearity important for small mammal communities on farmland? Journal of Applied Ecology 39(4):643-652. 68 Tonhasca, A., Jr. and B.R. Stinner. 1991. Effects of strip intercropping and no-tillage on some pests and beneficial invertebrates of corn in Ohio. Environmental Entomology. 20(5):1251-1258. 69 Tourenq, C., N. Sadoul, N. Beck, F. Mesleard and J.L. Martin. 2003. Effects of cropping practices on the use of rice fields by waterbirds in the Camargue, France. Agricultural Ecosystems and Environment 95(2-3):543-549. 70 Usher, M.B. and S.W.J. Keiller. 1998. The macrolepidoptera of farm woodlands: determinants of diversity and community structure. Biodiversity and Conservation. 7(6):725-748. 71 van Groenigen, J.W., E. G. Burns, J. M. Eadie, W. R. Horwath, and C. van Kessel. 2003. Effects of foraging waterfowl in winter flooded rice fields on weed stress and residue decomposition. Agriculture, Ecosystems and Environment. 95:289-296. 72 Volk, T.A. 2004. Sustainability of Short-Rotation Woody Crops Grown for Bioenergy and Bioproducts. seminar presentation, Cornell University Department of Soils and Crop Sciences. Thursday, February 12, 2004. 73 Weibull, A.C., Ö. Östman and Å. Granqvist. 2003. Species richness in agroecosystems: the effect of landscape, habitat and farm management. Biodiversity and Conservation. 12(7):1335-1355. 74 Wells, M. 2001. The Sustainability of Rotational Grazing and On-Farm Processing for United States Dairies. University of Edinburgh. 67. 75 Wunderle, J.M. 1999. Avian distribution in Dominican shade coffee plantations: Area and habitat relationships. J. Field Ornithol. 70(1):58-70. 76 Wunderle, J.M. and S.C. Latta. 1998. Avian resource use in Dominican shade coffee plantations. Wilson Bull. 110(2):271-281. 77 Zechmeister, H.G. and D. Moser. 2001. The influence of agricultural land-use intensity on bryophyte species richness. Biodiversity and Conservation. 11:1609-1625. 78 Zechmeister, H.G., A. Tribsch, D. Moser, J. Peterseil and T. Wrbka. 2003. Biodiversity 'hot spots' for bryophytes in landscapes dominated by agriculture in Austria. Agricultural Ecosystems and Environment 94(2):159-167.

Annex 4- 6 Chapter 5

Meeting Livelihood Objectives

5.1 Introduction management and soil-health related practices are being advocated, yet the economic viability of these Enhancing ecosystem and environmental outcomes to- systems remains uncertain (Gobbi 2000; Lyngbaek, et gether with improved food security and rural livelihoods al. 2001). Research that directly evaluates alternative is one of the central tenets of sustainable development in livelihood strategies which simultaneously conserve rural areas. Specifi cally, ecoagriculture strategies seek wild biodiversity is in its infancy (Carpentier et al. to conserve wild biodiversity in agricultural landscapes 2000; Williams et al. 2001; Stolton 2002). Syntheses of while enhancing agricultural and food production and research on agricultural intensifi cation have identifi ed rural welfare (McNeely and Scherr 2003). To be sustain- conditions under which synergies, rather than tradeoffs, able, conservation of wild biodiversity in productive may emerge that simultaneously achieve diverse social landscapes should not compromise the enhancement objectives, consistent with ecoagriculture strategies of economic well-being, locally or globally. The chal- (Vosti and Reardon 1997; Lee et al. 2001; Margolius lenge facing ecoagriculture strategies, therefore, is et al. 2001; Jagger and Pender 2003). that of generating complementary outcomes among three goals: (i) improving household livelihood objec- These and related outcomes raise numerous policy tives such as food security and income generation; (ii) questions. In general, are ecoagriculture strategies best achieving productive land uses (agriculture, forestry, characterized by tradeoffs or synergies among multiple etc.) that contribute to household economic (and other) goals? More broadly, is it more effi cient to spatially objectives; and (iii) achieving biodiversity conservation segregate or to integrate biodiversity conservation, that generates environmental benefi ts. food production and local economic development? In integrative conservation and development efforts, The simultaneous achievement of these outcomes is how are the costs and benefi ts distributed with respect a challenge, conceptually and in practice. To date, to economic growth, employment, and the livelihood empirical analyses of alternative land use practices concerns of small landowners? Such questions raise the have tended to reveal an inconclusive relationship issue, more generally, as to how tradeoffs or comple- among achieving livelihood enhancement, biodiver- mentarities between objectives may be identifi ed and sity conservation, and productive land uses. In one assessed. prototypical example, household level research in two settlements in the western Brazilian Amazon found Economic analysis can contribute in multiple ways that among alternative crop, livestock, and extractive to a framework for assessing possible tradeoffs and activities, improved pasture-livestock systems involved complementarities among the objectives of different low labor requirements and were the most profi table ecoagriculture strategies. An ideal livelihood-oriented for households in terms of returns to labor, but ranked framework would employ multiple measures of house- low in terms of carbon sequestration (Vosti et al. 2002). hold welfare, including cash income, health, food secu- The authors also concluded in this case that “agricul- rity, and risk minimization, and would recognize that tural intensifi cation appears to be at odds…with plant non-monetary factors (e.g., social capital and norms) biodiversity” (p. 258). Elsewhere, improved nutrient also infl uence decision-making (Ashley 2000). Be-

54 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations cause ecoagriculture strategies involve landscape level the economic implications of alternative land use and changes and multiple objectives, economic analysis technological practices given a household’s orientation should reveal the distribution of monetary (and non- toward alternative welfare-maximizing objectives (i.e., monetary) costs and benefi ts associated with different achieving food security, minimizing risk, or generating practices among different stakeholders. In addition, income). And it can simulate the farm and household- analysis should ideally assess the contribution of dif- level effects of policy changes that may modify output ferent market and non-market mechanisms in equitably and input prices, input availability and costs, or new redistributing costs and benefi ts (e.g., compensating legislation that may infl uence resource management farmers for conserving biodiversity). At a macro level, decisions. economic analysis can bring considerations regard- At regional and national levels, the concerns of policy- ing trade policies, local, regional and national market makers are often somewhat different. Regional adop- arrangements, and incentives, subsidies and technol- tion profi les, aggregate market impacts, and regional ogy adoption measures into the analytical framework. environmental effects resulting from the introduction Achieving all, or even most, of these diverse objectives of new technologies or systems are often of central in a single study however is a tall order. interest. National sector-level effects of domestic and In this chapter we examine how economic analyses international policy changes are of great concern since centered on achieving livelihood objectives can assist these effects defi ne the policy environment within in determining whether ecoagriculture is a suitable op- which households operate. Ideally, policymakers will tion for a particular landscape, what factors infl uence need to assess how ecoagriculture approaches perform the viability of ecoagriculture at the local and regional against other management alternatives in terms of level, and what policy conditions enable and obstruct diverse measures of economic growth, employment ecoagricultural approaches to land use. We begin by dis- generation, equity concerns, and poverty alleviation. cussing some key livelihood and policy considerations Economic analysis is also useful in estimating the net that pertain when assessing complementarities versus present value of different activities at different geo- tradeoffs among the three objectives of ecoagriculture. graphic scales, time periods, and under varying price We also describe, on a selective basis, the characteris- regimes. Alternative enterprises can be evaluated at tics of existing and emergent models and methods that social versus private prices, leading to an assessment would enhance an ecoagriculture analytical framework. of the social opportunity costs of different management We review quantitative and mixed models for examin- decisions. The effects of management decisions on ing and understanding key relationships. We conclude national poverty alleviation and equity goals can also this chapter by discussing the strengths and limitations be assessed. Biodiversity concerns can be addressed of existing models and identify areas requiring addi- by better understanding the externalities of resource tional research to strengthen the economic knowledge management decisions and their internalized effects in base underlying ecoagricultural strategies. household or government decision-making. Similarly, economic models may contribute to understanding how 5.2 Considerations in Assessing the macroeconomic, trade and sector policies affect the vi- Livelihood Impacts of Ecoagriculture ability of ecoagriculture strategies. The potential contributions of economic analysis to a Few individual studies can make progress using all, or framework for examining ecoagriculture approaches even most, of these criteria. Yet, ecoagriculture strate- will vary depending on the objective of the research gies inherently have multi-dimensional objectives, and project or development program. Some generalizations thus need to be evaluated by multiple criteria. Criteria are possible, however. At the household level, eco- that are central will differ by type and scale of tech- nomic analysis can identify and incorporate fi nancial, nology, production system, ecoregion, national policy demographic, biological, management and institutional objectives, and dominant household concerns. factors that inform decision-making. It can also reveal

Meeting Livelihood Objectives 55 5.2.1 Economic Viability and Related Concerns 5.2.2 Equity and Poverty

The economic viability of land management practices Equity and poverty concerns are central to ecoagricul- is a necessary but not suffi cient condition for adoption ture strategies for many reasons. Perhaps most centrally, in most cases. As Cary and Wilkenson (1997) conclude, rural poverty is centered in many of the same regions “…the best way to increase the use of conservation where wild biodiversity is richest and also under threat practices to overcome land degradation …will be to en- (Nelson, et al. 1997, cited in McNeely and Scherr sure the practices are economically profi table.” Perhaps 2003). Thus if one is to simultaneously address biodi- the most widely used conservation agriculture practice, versity and livelihood needs, one must necessarily deal zero (and minimum) tillage, has been widely shown to with poverty and equity issues. Alternatively, one can be economically profi table and has spread to millions contrast the Green Revolution, in which benefi ts were of hectares of North and South America and elsewhere, focused in highly productive areas, and ecoagriculture with increased yields of maize and soybeans, while strategies, which are often explicitly aimed at address- fertilizer and herbicide applications have been sharply ing the needs of marginal farmers and rural households reduced (FAO, 2001). However, other characteristics living in less-favored lands. In the latter context, the of management practices, farm households, and the distinction between “welfare poverty”—measured land itself can also infl uence adoption and implementa- by income, consumption or nutrition indicators—and tion. Yield, uncertainty, and health effects of proposed “investment poverty” is relevant (Reardon and Vosti, practices, and decision-makers’ objectives regarding 1995). As Reardon and Vosti suggest, the latter should food security, minimizing risks, and discount rates refl ect the ability of the household to invest in resource (e.g., preference for short- versus long-term benefi ts) improvements that “maintain or enhance the quantity will infl uence land management decisions. Practices and quality of the resource base—to forestall or reverse that require upfront investments and that may involve resource degradation.” This is an important distinc- a transition period during which returns are low (for tion, because investment-poor households will lack example, tree-based systems) may be unattractive and the resources and ability to “invest their way out of” have to surmount additional impediments to achieve poverty, even though in some cases they may not be widespread adoption. poor by conventional income-based standards. Produc- tion strategies that entail signifi cant fi xed costs and a Empirical research has shown that a wide set of house- minimum effi cient scale of investment can be beyond hold characteristics, land and agroecosystem factors, the reach of poor households. Risk and subsistence con- and external policy factors will also infl uence the adop- straints may further discourage these households from tion of land management practices (for example, see accumulating assets and increasing productivity. These Neill and Lee 2001). For a subsistence household, pro- concerns hold particular importance for conservation duction and consumption decisions are non-separable agriculture and ecoagriculture strategies that require and will result in different optimal input choices than investment of householders’ time, marketed surplus, or would occur if only production or consumption objec- other household resources in order to reap benefi ts that tives were present. Similarly, a risk-averse household may primarily accrue over the long run. with limited access to credit will respond to biophysical factors differently than a risk neutral household. Labor Equity concerns are important for other reasons as well. shortages or varying opportunity costs of labor will also Biodiversity conservation efforts are unlikely to prove alter the appeal of certain land use practices (see Chap- sustainable if the net benefi ts accruing from projects and ter 3). Adoption of conservation practices is different programs are distributed inequitably. Increased equity from the adoption of conventional technology because is likely to reduce the number and severity of resource these practices often result in externalities (positive confl icts. The net benefi ts resulting from biodiversity and negative), can involve high transaction costs if conservation may indeed differ across stakeholder cooperative action is required across a large area, and groups, particularly those that are heterogeneous. In the benefi ts may only accrue after a long period of time some cases, sharp inequities in the incidence of benefi ts (Marsh and Pannell, 1998). may lead to technological failure. In semi-arid India,

56 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations for example, Kerr and Sanghi (1992) found that soil of diversifi cation efforts and accompanying land use and water conservation technologies such as contour practices. or graded bunds that resulted in unevenly distributed Effective marketing services are particularly important benefi ts and costs and that required group action were when the diversifi cation is into high-value cash crop not undertaken. In this case, “equity becomes a prereq- production (Kherallah and Gruhn 2001). Effective mar- uisite to effi ciency”, the authors concluded. keting strategies may seek to take advantage of market 5.2.3 Markets premiums for agricultural products grown in specifi c conditions or for “niche” markets. Examples include Markets profoundly infl uence the economic viability shade-grown coffee, organic produce, and genetically- of ecoagriculture, as with all types of agricultural sys- modifi ed organism-free products. But successfully tems. Product market characteristics dictate the prices pursuing these options and obtaining premium prices received by farmers, the variability of those prices, and may require special management expertise, marketing the levels of marketing margins that infl uence profi t- services or market information. Niche markets often ability levels. Factor market characteristics dictate the suffer, for example, from the lack of a “critical mass” of cost of inputs and thus the viability of land management marketable production at the regional level to achieve practices. Transaction costs associated with marketing an effi ciently functioning market. Inadequate storage will depend on market characteristics, including market facilities and scale of production in input markets such proximity, scale economies in marketing and distribu- as those for mechanical services may further limit tion, and the availability of market information (Omamo production capabilities. Another market dimension, 1998). Current and anticipated output and input prices product certifi cation, is particularly important for niche are crucially important in infl uencing production deci- markets, and can be costly (Gobbi 2000). Certifi ca- sions, land allocation patterns, and investments in new tion is done on a crop or product basis, and is usually technologies. To cite just one of many such examples, granted for one or two products grown within complex Baltas and Korka (2002), in a study of land use practices systems (e.g., coffee or cocoa, within a shade grown in 76 villages in northwestern Greece, show that land coffee or cocoa system). And certifi cation or ecolabel- is more likely to be allocated to the land use practice ing policies alone may be inadequate for stimulating with the higher level and lower variability of expected demand. They may benefi t from being associated with returns. efforts to educate consumers regarding relevant produc- To the extent that ecoagriculture strategies are practiced tion processes (Nunes and Riyanto 2004). But not all in less-favored areas distant from central city markets producers have the management abilities or particular and without good infrastructure, they may be par- expertise to be successful in pursuing high-value pro- ticularly susceptible to market imperfections. Effective duction strategies. product diversifi cation, for example, is a particularly 5.2.4 Internalization of Externalities important strategy in areas where crop monocultures are not competitive. Ecoagriculture strategies such A signifi cant component of ecoagriculture is the con- as agroforestry-based systems and agrosilvopastoral servation of biodiversity. Achieving this goal results in systems are inherently multi-product systems, neces- positive externalities: pollination services, soil health sitating effi ciently functioning markets for multiple and fertility, maintaining genetic potential, controlling commodities. If these markets are present, this may pest populations, etc. To achieve biological conser- encourage strategies such as the domestication of high- vation, households bearing the costs of preservation value non-timber forest products within either enriched may need to be compensated, especially in situations forest fallows or other forms of multistrata agroforestry in which they are producing a public good for which (Leakey 2001). Without access to competitive markets compensation through standard market channels may be and the availability of storage, transport, and communi- impossible. There are different possible arrangements cation infrastructure, farmers and traders cannot market for market- and payment-based compensation for eco- their products effectively, limiting the effectiveness system services. These mechanisms can generate incen-

Meeting Livelihood Objectives 57 tives to conserve biodiversity and maintain ecosystem key challenge in this fi eld, then, is that of identifying, services in productive landscapes; the mechanisms can quantifying, and valuing the actual services provided by be spatially explicit and have landscape-level effects. the ecosystem, and then determining how to structure a sustainably effective incentive-based payment system The mechanisms for paying for ecosystem services around those services. A similar diffi culty arises with range widely. They include government programs such performance-based payments, where it is necessary to as the Conservation Reserve Program in the United identify performance measures that are measurable at States (Feather et al., 1999) and the Environmental a reasonable cost and clearly linked with management Quality Incentives Program in the U.S., similar pro- decisions. grams in the European Union, and other direct payments programs supporting conservation practices (Ferraro 5.2.5 Property Rights and Kiss 2002). Among developing countries, Costa Rica’s payments for environmental services (PSA) Property rights assign rights to individuals to streams programs are perhaps the most elaborate, compensat- of benefi ts. Benefi ts can be generated from agricultural ing farmers and landowners for environmental services land, but include all land-based natural resources (Place including carbon sequestration, watershed protection, and Swallow 2002). There is widespread evidence biodiversity conservation, and the provision of scenic in the literature of causal linkages between effective beauty (Zbinden and Lee 2005). Other mechanisms are property rights and the adoption of technology (Feder market-based such as certifi cation (e.g., for shade grown et al. 1985). Property rights are especially important in coffee) and tradeable permits (Greenhalgh and Sauer situations employing technologies with long-term cost 2003; Chomitz et al. 2004). Payments can be made for and benefi t streams, such as the adoption of agroforestry a wide set of ecosystem services such as water quality, systems, soil conservation structures, and managed water fl ow, water retention, soil erosion mitigation, and fallows systems. In these and similar cases, assured habitat creation. A recent review identifi ed roughly 300 access to future benefi ts is necessary to induce current examples of such mechanisms worldwide (Landell- period investments. Gebremedhin and Swinton (2003), Mills and Porras 2002, cited in Pagiola et al. 2002b). for example, found that secure land tenure arrange- Currently, there are 20 countries that engage in some ments were among the most important determinants of form of payments (primarily direct payments and some soil conservation investments in the Tigray region of subsidy program) for conservation (Ferraro 2004). Ethiopia. Property rights can thus positively infl uence incentives to adopt new land management strategies, Payments for ecosystem services have the potential especially in areas with an existing high degree of inse- for signifi cantly advancing conservation efforts. The cure rights (Place and Swallow 2002).1 Property rights popularity of these mechanisms is based on the premise affect access to information, wealth, risk management that they can directly and cost effectively address en- and credit, which also in turn infl uence technology vironmental goals and use economic instruments, such adoption. Property regimes can also buffer environ- as taxes or incentive payments, to stimulate voluntary mental risks and price fl uctuations. landholder compliance (Stoms et al. 2004). These approaches avoid far more complicated and management- Indigenous tenure systems and property rights institu- intensive options, such as integrated conservation and tions in Asian and African societies often provide suf- development programs, which often seek to advance fi cient incentives to adopt technology (Knox McCulloch conservation goals indirectly. What is lacking in eco- et al. 1998). Common property regimes accommodate system service payments programs is the scientifi c basis multiple users beyond the household level, and there- to substantiate the contribution of improved practices fore are better able to distribute benefi ts equitably on biodiversity conservation or ecosystem health (per- (Knox and Meinzen-Dick 1999). However, in many sonal communications with Pagiola 2004 and Chomitz 2004). Operational issues such as program design and 1 The exception is when the costs of investments are very low fi nancing, assuring sustainability, and balancing equity or the expected profi ts are high, in such cases property rights and environmental goals also present challenges. The are not as relevant (Place and Swallow 2002).

58 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations locations, these systems are challenged by state owner- welfare (Poteete and Ostrom 2003). Increased social ship and private property tenure. A recent comparative capital can thus contribute to improving livelihoods, review of land tenure and forest management in seven as evident in 69 villages in Rajasthan, India where Asian and African countries found that: 1) state owner- local development was positively associated with a ship is the most inappropriate land ownership system; combination of social capital and capable government 2) common property systems were effective when agency (Krishna 2003). The viability of management key forest resources were minor products, and 3) that approaches that require coordination and cooperation high-value tree production is less amenable to common among land users also positively benefi ts from social property management (Otsuka and Place 2001). The capital. authors concluded that social forestry projects should be redesigned to retain community management of for- 5.2.7 Macro and Trade Policies est production, but strengthen individual incentives for Government policies have distorted agricultural per- the management of high-value trees. Property rights are formance worldwide. Subsidies, protectionist trade also shaped by technology adoption. Tree planting has policies, currency overvaluation, and land reforms are been widely cited as one investment that confers strong examples of macro-level policies that have strongly land rights to individuals (Fortmann and Bruce (1988), shaped resource use and agricultural production (World Suyanto et al. (1999), Snyder (1996) and Baland et al. Bank 2003). For example, agricultural price subsidies (1999) as cited in Place and Swallow 2000). and price supports have supported the expansion of large-scale cropping systems in industrialized countries, 5.2.6 Social Capital and Collective Action thereby reducing global commodity prices and making Social capital is “the structure of relations between it diffi cult for developing country producers to compete actors and among actors” that encourages productive in these markets. This issue is a central factor in the activities” (Pretty 2002). Social capital is based on rela- impasse in the current Doha Round of trade negotiations tions of trust that reduce the cost of collective action. under the auspices of the World Trade Organization. Reciprocity and exchange among group members gen- Were industrialized countries to signifi cantly reduce erates trust and confi dence in collective efforts (Pretty agricultural subsidies, those countries would likely lose and Ward 2001). Institutions, including common rules, a substantial degree of trade competitiveness. To the norms and sanctions that elevate group interests above extent that marginally productive areas in industrial- those of individuals, further strengthen obligations ized countries would go out of production, biodiversity among people. This engenders positive environmental would benefi t. Conversely, projected increases in global outcomes (Pretty 2002). Varughese (1999, as cited in commodity prices would likely benefi t developing Poteete and Ostrom 2003) found that in Nepal, forest countries that have a disproportionate reliance on ag- conditions were more closely correlated with collective riculture. These price increases would likely improve action than with population pressure. Chakrabarti et al. the economic prospects for many rural households in (2001, cited in Poteete and Ostrom 2003) also found a developing countries. It is not clear that environmental positive correlation between forest conditions and level and biodiversity objectives would necessarily be real- of collective action. ized by a simple shifting of commodity production from industrialized to developing countries, where environ- Social capital can benefi t the implementation of land mental laws and regulations are typically less stringent management practices that enhance agricultural produc- and less likely to be enforced. At the same time, the tivity and conserve biodiversity (Pretty 2002). In the prospect of improved economic returns would create Peruvian Altiplano, social and human capital displays options for developing country producers that could a clear association with the choice of sustainable ag- potentially strengthen ecoagriculture strategies. ricultural practices (Swinton and Quiroz 2003). It can engender synergy between productive land uses and The global reduction in trade barriers in the past two ecosystem integrity (Flora 1995; White 1995; Krishna decades has also had a signifi cant effect in some coun- and Uphoff 1999; Dougill 2001) and infl uence local tries expanding their exports of “non-traditional” com-

Meeting Livelihood Objectives 59 modities, including the high-value products previously will need to take into consideration activities and in- mentioned. These products have great potential as part vestments in neighboring plots and, where relevant, the of export diversifi cation in many countries, and promo- broader landscape to fully reap the environmental and tion of non-traditional, value-added exports is a widely biodiversity benefi ts of alternative land use practices. touted economic growth strategy. But taking advantage Similarly, for governments, selection of ecological of increasing global demand for these products is not reserves, environmental enforcement, and research easy. Success in non-traditional agricultural export in- and information exchange all require optimal alloca- dustries typically requires superior management exper- tion of limited funds (Polasky et al. 2001). Intangible tise, good marketing and transportation infrastructure, benefi ts or indirect costs associated with investments close attention to global price and market trends, and in conservation have to be made explicit and, wherein pursuing modern business and marketing practices. possible, measured when comparing different invest- These requirements are often outside the realm of ments or policy measures for promoting biodiversity smallholder agriculturalists. To achieve success, eco- conservation on productive landscapes. Governmental agriculture strategies that seek to serve the international and non-governmental organizations investing in market must not only succeed on the production side, biodiversity conservation and social welfare need to but in marketing and product distribution. stimulate cost-effective approaches that effi ciently achieve the objectives of the initiative. The reduction of trade barriers has the potential to further increase production, trade, and consumers’ All of the aforementioned factors are important consid- welfare. However, trade also can negatively affect erations when designing and promoting land manage- resource use and lead to environmental degradation if ment practices with multiple objectives. The viability production with negative environmental externalities of a particular land management strategy under a par- shifts to countries with relaxed or poorly enforced en- ticular combination of factors cannot be transferred to vironmental regulations, and/or if those countries don’t a different context without research. Model outcomes, enforce the regulations they have. A favorable policy therefore, are important for specifi c conditions. Model context for ecoagriculture will require measures that interpretation will also depend on which factors are internalize negative externalities and reward sustain- incorporated and how effectively they complement able resource use. reality.

5.2.8 Investment in Conservation 5.3 Assessing Tradeoffs and Synergies in Ecoagriculture Strategies As we have seen, there is widespread evidence that economic viability is a crucial incentive for the widespread Approaches that examine the role of the abovemen- adoption of most agricultural and natural resource man- tioned factors on land use and that evaluate tradeoffs agement practices. Thus, investments in conservation and synergies among biodiversity, productivity and of wild biodiversity on productive lands must gener- livelihood objectives, can improve stakeholders’ ate greater returns than alternative investments. For a understanding of ecoagriculture strategies. Ideally, private landowner, ecoagriculture will be economically models are able to integrate variables measuring bio- rational if the current and discounted future expected logical diversity, recognize the heterogeneity among benefi ts derived from management practices exceed and within stakeholder groups, capture the spatial and their costs (including opportunity costs). temporal scale of the activities, incorporate feedback Investment in appropriate land management practices loops, address risk and uncertainty, and model dynamic across a landscape is important to achieve wildlife con- interactions among alternative land and resource uses. servation. Ecoagriculture strategies require landowners Examination of household decision-making should also to make fi nancial and capital commitments. Suitability be based on theoretically sound frameworks. Moreover, of an investment will depend on private and social costs approaches should recognize that the farm household is and benefi ts. Returns on private investments, however, the main decision-making units regarding land use and technology adoption. However, understanding market

60 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations outcomes and addressing policymakers’ interests often the analytical approach. This section briefl y describes require analysis at a more aggregative level. some of the basic characteristics of major quantitative modeling approaches and illustrates how they are cur- Meeting all of these criteria is clearly a major challenge, rently being applied. particularly for ecoagriculture-related research where knowledge generated in terms of multiple dimensions 5.3.1.1 Bioeconomic Models and indicators is required. It is the rare analytical or modeling approach that successfully addresses even a Bioeconomic models link biophysical and decision- few of these criteria. Typically, this necessitates trad- making processes at the household, village and regional eoffs among modeling criteria depending on what are levels (Kruseman and van Keulen 2001). Biological judged to be the key issues and constraints in specifi c variables can include soil- related parameters, water empirical situations. Model design and application will quality measures, number of different species per area also vary depending on the ecoagriculture strategy of or variables representing biological diversity. Bioeco- particular concern. Ideally, modeling approaches should nomic models often include biophysical information also be comprehensible to a range of stakeholders, in- as constraints (Jensen et al. 2001); employ measures of cluding private and public decision-makers. land suitability; make assumptions regarding impacts of practices on biodiversity (Fleming and Milneb 2003); This section reviews modeling approaches of potential and develop sustainability indicators as measures of relevance to the analysis of ecoagriculture. The dis- environmental impact (Zander and Kachele 1999). cussion is by necessity selective; no attempt is made Economic and biological components can be defi ned to to be wholly inclusive. The particular emphasis is on best represent the relevant characteristics of the prob- quantitative modeling approaches that are capable of ad- lem at hand. Bioeconomic models are most commonly dressing tradeoffs and synergies among the biophysical developed at the farm and household levels. They often and economic components of a system. A few examples have separate modules that specify farm household of mixed approaches are also discussed. Mixed models preferences, technological choice, and sustainability incorporate both quantitative and qualitative informa- parameters (Heerink et al. 2001). They can be solved tion and examine relationships between biological through linear programming, dynamic programming, and social components of land use practices, or they and other related approaches. examine synergies and trade-offs between productivity, biological diversity, and welfare. Details regarding Bioeconomic models can reveal tradeoffs and synergis- the models referenced in this section and other related tic relationships between the biophysical and economic models are in given in Annex 5. dimensions of the system, or among the biological elements. Bioeconomic models at the household level 5.3.1 Quantitative Modeling Approaches often incorporate farm household modeling that accounts for natural resource endowments, inputs and Quantitative approaches for assessing the interaction labor allocation decisions, and output choices and between ecological and economic phenomenon are nu- consumption preferences under different conditions of merous and often quite sophisticated. Models typically market development (Heerink et al. 2001). Biophysi- link biophysical and economic factors at varying scales, cal information is linked to the production side of the simulate land use changes at farm or regional landscape farm household model. These approaches are also scales, or may use spatially explicit information incor- used to understand the impact of international trade on porating GIS-based approaches. Bioeconomic models production, marketing, and land conservation decisions and systems models are among the promising vehicles (Barbier and Schulz 1997). for analyzing biophysical-economic interactions. These, and other, modeling approaches vary widely—by level Bioeconomic models have been developed to assess the of decision-making, level of detail in the data, scale economic impacts of a broad range of technology adop- of analysis, and which variables are considered as ex- tion and policy scenarios (de la Briere 2001; Barbier et ogenous versus endogenous (e.g., determined) within al. 2003). Okuma et al. (2002), to cite just one example,

Meeting Livelihood Objectives 61 used a dynamic utility-maximizing bioeconomic model incorrect conclusions about the viability of alternative to assess, ex ante, the likely impact of adopting multiple management strategies (Brown 2003). This is certainly technologies on crop-livestock systems under various the case for ecoagriculture, as in other applications. In policy scenarios in the Ginchi watershed of Ethiopia. ecoagriculture, the mosaic of land-uses across a land- They found that under the existing policy conditions, scape affects the viability of integrating agricultural soil conservation measures may not be a profi table productivity with conservation of habitat for wildlife venture and that tree planting or conservation measures and improved livelihoods. were not adopted despite their profi tability. Instead the Advancement in spatially explicity modeling has mostly need for food self-suffi ciency resulted in fallowing in occurred in non-economic fi elds of study. Geographers the crop-rotation pattern. and biophysical scientists have taken the lead in de- Examination of policy impacts on biophysical param- veloping spatially explicit models of land use change. eters, such as soil erosion and carbon sequestration, is There are three categories of spatially explicit non-eco- also possible (Barbier and Bergeron 1999; Schipper et nomic models – simulation models, estimation models, al. 2001). Schipper et al. (2001), for example, use an and hybrid approaches2 (Irwin and Geoghegan 2001). elaborate static bioeconomic model (SOLUS) to explore Change is stimulated using vast amounts of spatially the aggregate effect of policy measures on both effi - disaggregated land use and land cover data, typically ciency and non-economic and environmental sustain- obtained from satellite imagery. These models are not ability objectives at a regional level in the Atlantic Zone typically behavioral or economic models and gener- of Costa Rica. The model incorporated sustainability ally use an “ad hoc approach to identifying physical and environmental parameters relevant to the region in variables that represent the outcomes of economic the optimization. The authors found that technological and social processes” (Irwin and Geoghegan 2001). improvements result in increased economic surplus Spatially explicit non-economic simulation models of and reduced environmental degradation, more so than land use change often build on mathematical models restrictions on soil nutrient depletion. in which the behavior of a system is generated by a set of probabilistic or deterministic rules (Irwin and Bioeconomic models are limited in how they represent Geoghegan 2001).3 Although these models have made biophysical parameters. In the case of soil, some ap- a signifi cant contribution to land use/land cover mod- proaches incorporate information on erosion, repre- eling, they have been limited in capturing the causal senting changes in soil components at a geographical relationship between individual choice and land use scale, but do not adequately capture the temporal scale outcomes (ibid.). or dynamic changes in soil conditions (Lehman 2004, pers. comm.). Representations of how a specifi c practice Work in environmental economics has focused on de- affects biological diversity may be based on inappro- veloping economic models of individual landowners’ priate assumptions. Likewise, the use of available crop decisions within a spatially explicit framework (Irwin growth models does not capture the dynamics of multiple cropping systems (Heerink et al. 2001). Another 2 Hybrid models involve estimating parameters with simula- major challenge for these approaches is aggregating tions. up to the regional or watershed levels which are often 3 Much of the work in this area identifi es transition rules by of greatest to policymakers. The regional scope of the quantifying the resulting pattern. The explanatory power of SOLUS model cited above is unusual. Additional re- these models is demonstrated by showing how the hypoth- search is needed on these and various other aspects of esized interaction effect is similar to the resulting spatial bioeconomic modeling. evolution of the land use pattern (Irwin and Geoghegan 2001). Moreover, these models are based on change rules 5.3.1.2 Spatially Explicit Models that are based on the cell’s attributes and states of surrounding cells. The use of only these attributes overlooks a host Ignoring the spatial dimension in resource management of other factors across a landscape that change land use problems that are inherently spatial in nature can lead to patterns (Irwin and Geoghegan, 2001).

62 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations and Geoghegan 2001). Spatially explicit economic model converts a spatial-temporally explicit resource models have focused largely on deforestation. These extraction model into a nonseparable household model models use economic theory and the empirical literature applicable to subsistence agriculture. The spatial dimen- to identify variables to include in land use conversion sion is made explicit by recognizing that households’ models and to identify causal relationships (Chomitz decisions regarding which parcel to fallow versus to and Grey 1996; Pfaff 1999). They estimate the param- cultivate varies with spatial factors (e.g., location of eter values to help understand the factors infl uencing plot and spatial variation in soil conditions) (Brown land uses rather than predicting micro-level changes 2003). This type of modeling approach is similar to a in the spatial pattern of the landscape. For example, systems modeling approach that is spatially explicit. It Chomitz and Grey (1996) found that in Belize roads suffers from the same limitations as the spatially explicit infl uence the probability of land being used for com- economic models discussed above and those associated mercial agriculture because road access affects market with systems approaches. access4. Spatially explicit economic models, such as those discussed above, require creative use of spatial 5.3.1.3 Systems Analysis Frameworks data in models and improved empirical methods, in- Models that integrate distinct disciplinary frameworks, cluding those being developed in spatial econometrics for purposes of this report, are identifi ed as systems (Irwin and Geoghegan 2001). analysis frameworks. Systems analyses are based on More recently, spatially explicit economic models have theoretical models or operate around posited causal been developed to improve targeting in the retirement relationships among elements of a system. They often of agricultural lands, such as the U.S. Conservation involve hierarchical systems, with separate models for Reserve Program (Newbold and Weinburg 2003). biophysical, socioeconomic, and policy components Newbold and Weinburg developed a theoretical model connected through input-output relationships, tenure that uses spatially explicit models of production func- arrangements, or markets. The components can be tions in an optimization framework. The objective spatially explicit, include macroeconomic components, is to maximize environmental benefi ts with a budget and employ multiple data sources. constraint. In this model environmental benefi ts are Loosely coupled or open models have distinct disciplin- being considered with different weights added to each. ary components executed independently, with outputs Using information on the environmental benefi ts, the from each component feeding directly into other models benefi t-cost ratios are calculated for each spatial cell (Antle and Stoorvogel 2003). A closed model is one that could be conserved. The cells with highest ben- in which the processes in distinct submodels interact efi t-cost ratios are conserved and the benefi t-cost ratio simultaneously, rather than recursively, as in the case of for neighboring cells are updated. The process is then loosely coupled models. This can involve decomposing repeated till the budget is completely expended. Exten- each of the models into a sequence of sub-processes sions of these models would be useful for examining that are loosely coupled, with feedback among the ecoagriculture initiatives that create additional habitat submodels. “Another approach is to simulate the two for wildlife because of the inherently spatial nature of models sequentially, once for each of the shorter time habitat creation. steps, each time using data from one model to initial- Bioeconomic models have been extended to incorporate ize the other model up to that point in time” (Antle and spatial elements that inform resource use decisions. Stoorvogel 2003). Antle and Stoorvogel (2003) applied Brown (2003) reviews examples of such models these two different systems models to identify the eco- and also provides a detailed description of a spatial- nomic, environmental and human health tradeoffs in temporally explicit model of shifting cultivation and Ecuador’s potato-pasture production system. With the forest cover dynamics in the Congo Basin. The latter loosely coupled model they found that farmers shifted towards potato production as the price of potatoes 4 The fi ndings, however, need to be interpreted with caution increased relative to milk. This resulted in increased because there were endogeneity problems. pesticide use and environmental and human health ef-

Meeting Livelihood Objectives 63 fects associated with using pesticides. In contrast, the Systems models that explicitly examine wildlife habi- closed systems approach revealed substantially differ- tat are limited. An example is TAMARIN, a systems ent estimates of impacts at sites where the feedbacks model being pilot tested in the Atlantic Rainforest area were strongest. of Brazil (Stoms et al. 2004). TAMARIN evaluates and compares different landscape confi gurations for A systems analytical framework is helpful in examining achieving conservation goals at the regional scale. The the interactions among different systems components. model sets general environmental goals and then uses Simulations are used to assess how modifi cations in one economic instruments to induce voluntary compliance component alter other elements. Systems models can be to achieve these goals. TAMARIN simulates a range of used in various contexts that are relevant to assessing programs involving economic instruments with fl exible ecoagriculture strategies, such as in analyzing: factors specifi cation of program eligibility, payment rules and infl uencing households’ decisions across different land budget for the program. The model uses three layers of use options (Legg and Brown 2003a); alternatives for spatial information in GIS—current land cover, ecologi- simulating land use in unobserved situations (Antle and cally distinctive sub-areas, and the opportunity cost of Capalbo 2001); examining tradeoffs between environ- conservation expressed as a value which is assigned mental and economic indicators resulting from differ- to each point on the landscape. The main objective of ent policy and technology scenarios (Stoorvogel et al. the model is to assist planning conservation of wildlife 2003); assessing potential outcomes from simultaneous habitat. changes in land use at a landscape scale (van Noordwijk 2002); and identifying the impact of policies on bio- A key challenge in using systems analysis approaches logical diversity (Stoms et al. 2004). Systems analyses is parameterization of the components of the models. are potentially suitable for examining ecoagriculture These models are typically highly data intensive. Sen- strategies that enhance habitat value of farmlands and sitivity analysis is also important in systems analyses to efforts to increase habitat for wild biodiversity. assess the threshold levels for the different elements of the system. There are different user-friendly interfaces Research and development of systems analytical frame- for conducting a sensitivity analysis for a systems ap- works has furthered the understanding of interactions proach (e.g., STELLA software). among the social, economic, and ecosystem service components of systems. The Alternatives to Slash and Systems analytical frameworks that capture the interac- Burn (ASB) program managed by the International tions of various systems elements provide a promising Centre for Agroforestry Research (ICRAF) has sup- framework for examining ecoagriculture. Hierarchical ported the development of systems models within the models that link together different disciplinary models SIMILE environment. SIMILE links systems dynamics either through open or closed loops are suitable for this modeling with spatially explicit landscapes (Legg and purpose. These frameworks are fl exible and different Robiglio, 2001). It is a series of biophysical models disciplinary models can be adapted to the local context. of farming systems linked through land tenure rela- Furthermore, they are relatively easy to comprehend tionships to models of villages and households which and can generate policy recommendations. incorporate kinship and other linkages (ibid.). SIMILE models activities and landscapes and evaluates the im- 5.3.1.4 Macro-Level and CGE Models pacts of a range of interventions and policy options on Macro-level models can be used to assess the effect rates of forest conversion to agriculture. FALLOW (van of policy changes, such as trade liberalization, on land Noordwijk 2002) extends the above systems frame- uses and the implications for biodiversity conservation. work and includes plant biodiversity. This model was Much of the recent interest in these approaches has developed to predict changes in food self-suffi ciency, been stimulated by understanding the effects of major soil fertility, carbon stocks, plant species richness and trade liberalization initiatives (WTO, NAFTA) on land watershed functions at a landscape scale with changes and resource use. In terms of approaches particularly in population and land use, and has been applied in relevant to analyzing ecoagriculture strategies, one Indonesia.

64 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations subset of “trade-conservation” models links two-coun- 2003). Mixed models are capable of using both qualita- try, two-good trade models with species area curves to tive and quantitative information, although the former examine how trade policies affect the area of land used must sometimes be incorporated in ad hoc fashion. Such for agriculture, timber and conservation. These models models can include a broad range of social variables, have production, species assemblages, and consumer such as social capital, property rights, and village-level preference components. The model is solved for species information. These models can also incorporate spatial assemblages and the maximization of consumer utility information, including climate variables, land uses, and under different trade regimes (Polasky et al. 2003). To levels of soil erosion (Pender 2003). make models tractable, assumptions such as the incom- Mixed models can generate insights into how policy patibility of productive land uses and the conservation changes affect land use practices and household welfare, of native biological diversity may be necessary (Polasky and can provide information on feasible development et al. 2003). options and resultant biophysical changes. Pender et al. Computable General Equilibrium (CGE) models are (2001) identifi ed the major pathways of development useful in capturing the second-round effects of land that have been occurring in Honduras, their causes use changes. CGE models applied at a regional level and implications for agricultural productivity, natural are useful in the ecoagriculture context for assessing resource sustainability and poverty. He found that the impact of trade or domestic policies. However, development pathways affected changes in agriculture few CGE models addressing questions related to ag- and resource management but not changes in poverty riculture and land use explicitly integrate biophysical (also see Pender 2003). The horticulture, coffee, and or biodiversity components. CGE models often use non-farm employment development pathways resulted different land classes, in which biological/biodiversity in favorable productivity outcomes but their implica- properties are implicit rather than explicit. Spatially tions for resource conditions are mixed. These models explicit information can be linked with a CGE model might be modifi ed to include additional pertinent spatial to generate information on how large-scale phenomena, information. For example, spatially explicit information such as climate change, affect the integrity of region- can be incorporated in a dynamic fashion to capture the ally specifi c ecosystems (Darwin et al. 1996). Darwin interactions between the socioeconomic, productivity, et al. (1996) used land classes in a model that links and biological elements of the system. Bolwig et al. GIS with a CGE model to assess how global changes (2003) applied such a model in Uganda to estimate the in climate, human populations and international trade potential consumer and producer benefi ts, and export policies might affect land area under tropical forests revenues of simulated changes at the subnational and at the regional scale. National-level CGE models and national levels. The results revealed that increased pro- multi-market models can be used to analyze the impact duction in cotton increased benefi ts but the production of economic policy reforms on prices faced by farm was occurring in areas that were not close to the markets households (Dyer et al. 2001). They can also assess (Bolwig et al. 2003). how land use practices affect national-level measures Mixed models have distinct benefi ts and disadvantages. of development (Lewandrowski et al. 1999). These For example, they are not constrained by having to link models could be altered to examine issues at scales of data collected at different temporal and spatial scales. interest to ecoagriculture strategies. However, these approaches may not capture the full 5.3.2 Mixed Models Used to Evaluate Tradeoffs interactions among the various biological and social elements of a system. Moreover, a further key limitation is Tradeoffs and synergies can be examined using models that these approaches are time- and resource-intensive, that analyze ecological and social elements of a land and are dependent on the collection and incorporation of use practice independently and that incorporate theoreti- extensive empirical data. Despite these limitations, the cally sound assumptions and/or information about re- application of these models may usefully be extended lated practices (Dougill et al., 2001; Jagger and Pender, to various ecoagriculture strategies.

Meeting Livelihood Objectives 65 5.4 How Economic Evaluation Can ing efforts. To do this rigorously requires detailed data Contribute to Making Sound Choices of very different types, collected over multiple years, and analyzed using sophisticated methodologies, some This brief review has identifi ed some of the analytical examples of which have been reviewed here. These approaches that have been used to examine the multi- analytical and data requirements do not lend them- dimensional aspects—economic, production, and selves well to “one-off” research efforts, but rather to environmental—of alternative production systems long-term, well-funded research programs incorporat- relevant to ecoagriculture. For the purposes of specifi c ing multiple disciplinary approaches. In the past, such production strategies, the strength of these approaches research programs have understandably been mostly is in their varying potential to: 1) identify specifi c focused on micro-level production systems (individual measurable outcomes that are consistent with the multi- farms or communities, several communities, or the sub- dimensional indicators of ecoagriculture strategies; and watershed level), where conditions are more likely to be 2) evaluate specifi c tradeoffs and synergistic relation- homogeneous, and analysis is more tractable. However, ships between and among these alternative outcomes. this has meant the ability to address the aggregate-level Regarding the latter, ecoagriculture has the potential market and policy issues of primary interest to policy- to benefi t from achieving synergistic outcomes from makers is often limited. One of the key challenges for many sources, including: ecoagriculture, from an analytical perspective, is to • Increased effi ciency of input use; make the most of past and emerging research efforts, recognizing that the ability to rigorously assess the • Synergies between component inputs; tradeoffs and complementarities among multiple objec- • Substituting natural capital for fi nancial capital; tives is necessarily limited to relatively few strategies, and that the conditions favoring an “ideal” research • More effi cient spatial organization; environment arise only infrequently. • Improved input performance; Formal analysis and modeling of ecoagriculture should • Economies of scale through farmer collaboration; build on and extend current efforts to capture various • Benefi ts to farming from wild species or revegeta- elements at the landscape scale. Efforts to assess the im- tion. pact of land management on ecosystem services should better capture the spatial and temporal element of these Whether it is able to do so will require signifi cant prog- services. Such models could benefi t from strengthening ress in applied research along the lines outlined here. In their bioeconomic component to incorporate appropri- some limited cases, modeling efforts have prioritized ate biophysical and biodiversity variables, such as soil the assessment of tradeoffs from the outset, making erosion or soil nutrient parameters, species diversity these approaches particularly promising (for examples, measures (e.g., the Shannon index, or intra- or interspe- see numerous chapters in Lee and Barrett 2001). The cies diversity), or information on water quality. approaches reviewed here can also provide insights regarding the effectiveness of policies, incentives for Better understanding the tradeoffs and synergies adoption, and effects of different land management resulting from ecoagriculture strategies given posi- approaches. Such analyses will assist in designing of tive ecosystem externalities would provide valuable mechanisms to compensate individuals for conserving insights. Moreover, the prospect of designing appro- biodiversity in productive landscapes. priate payments and compensation schemes around these externalities is likely to be of growing interest to It is important not to underestimate the analytical both policy-makers and resource owners and manag- and data-related challenges posed by ecoagriculture ers. The TAMARIN model (see Annex 5) provides research. The explicitly multi-dimensional nature of an interesting example of such an effort. TAMARIN ecoagriculture strategies means that economic, agro- is a model that explicitly considers wild biodiversity nomic, and biodiversity elements should ideally be (mega-fauna) and the economic incentives necessary for simultaneously incorporated in analytical and model- creating necessary habitat, and is based on economic

66 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations and ecological principles. Adjustments to this model, modeling approaches and used in meaningful fashion including improvement of its economic sub-module, both for research and policy analysis. Research makes could extend its relevance to ecoagriculture. clear that different environmental indicators can respond differently to economic and production deci- Extending research fi ndings from one scale (e.g., the sions. Biodiversity outcomes associated with identical farm or household) to a larger scale (e.g., the watershed production strategies may be dramatically different for or landscape) must be carefully done since simple plant biodiversity compared with soil micro-fauna com- aggregation is often not appropriate, either in terms pared with large mammals. And these outcomes may of market outcomes (e.g., price determination) or bio- be very different from other environmental outcomes physical effects (e.g., soil erosion impacts). Ultimately, (soil erosion, land deforested, carbon sequestration, a focus on spatial dimensions should lend insights into etc.). We need better tools, measures, and indicators to the crosscutting issue of “separation versus integra- incorporate in modeling efforts to identify alternative tion” identifi ed earlier in this chapter. Ecoagriculture outcomes within the broader objective of biodiversity strategies are built on the assumption that integrative conservation. approaches to simultaneously realizing economic, production, and biodiversity goals are optimal, or at least A deeper understanding of factors that inform house- preferred. But integrative strategies employed at the holds’ decision-making across multiple objectives and micro level may not necessarily yield optimal results their implications for ecoagriculture would be benefi - when considered at a more aggregative level. Future cial. Models used to predict or examine the viability research directed at this issue of the scale and spatial of ecoagriculture need to more adequately represent dimensions of ecoagriculture strategies is required to whether or not the decision-maker is updating their conclusively demonstrate the viability of these strate- knowledge regarding the system, and how this knowl- gies under varying biological and market conditions. edge changes with spatial aggregation, or dispersion, of habitat. The role of social capital in shaping decision- Implementation of ecoagriculture strategies would making, mitigating risks, and ensuring the necessary also benefi t from linking improved quantitative and conditions at the landscape level needs to be better mixed models of land use with planning processes. represented in models of ecoagriculture. A comprehensive understanding of the tradeoffs and synergies among the livelihoods, biodiversity, and Improved representations of the biophysical component agricultural productivity components of ecoagriculture of ecoagriculture systems would strengthen ecoagri- should inform participatory planning processes and as- culture analyses. An ideal approach would use simple sist decision-making. Economic analysis could forecast yet theoretically sound measures of the biophysical the impact of proposed ecoagriculture strategies and/or system. Using these measures, the biophysical compo- assess its feasibility. The analyses could identify the dis- nents could be examined at the appropriate geographic tribution of costs and benefi ts or more accurately assess scale and models then capture the temporal scale at the viability of specifi c land use strategies given local which biophysical changes affect land managers. conditions and the political context. Such information The biophysical component of the model should be would assist in effectively targeting ecoagriculture. dynamic and, therefore, evolve with the land management practices. 5.4.1 Directions for Future Research Relatively few of the models examining the impacts In addition to the above, there are several areas of ad- of land use changes consider the negative externali- ditional research that could enhance our understanding ties faced by land managers. For example, a farmer’s of the tradeoffs and synergies resulting from ecoagri- selection and planting of perennials on the boundary culture. Given the focus of ecoagriculture strategies of his/her property can affect the crop productivity on the conservation of wild biodiversity, perhaps the of the neighboring farms (Jagger and Pender 2003). greatest need is in developing measures and indicators Nutrient leaching can affect soil health in neighboring of wild biodiversity that can be easily incorporated in plots. GIS-based models should be able to capture these

Meeting Livelihood Objectives 67 effects or internalize this information in measures of Antle, J.M., and J. Stoorvogel. 2003. Incorporating land suitability. This will be crucial to comprehend the Systems Dynamics and Spatial Heterogeneity in landscape scale effect of ecoagriculture strategies. Integrated Assessment of Agricultural Production Systems. In Working Paper, Department of Agri- Most of the models reviewed in this chapter are not cultural Economics and Economics, Montana State restricted to a particular scale of agricultural activ- University. Bozeman, Montana, USA. ity. This is important because “large-scale, high-tech agricultural producers have considerable scope for Antle, J.M., and S.M. Capalbo. 2001. Econometric- ecoagriculture” (McNeely and Scherr 2003, p. 108). Process Models for Integrated Assessment of Agri- However, the applicability of ecoagriculture will vary cultural Production Systems. American Journal of in different land management situations, including Agricultural Economics 83 (2):389–401. those defi ned by scale. Monocropping tends to prevail Ashley, C. 2000. Applying Livelihood Approaches to at the large-scale end of agricultural systems, and we Natural Resource Management Initiatives: Experi- need to better understand how measures of biodiversity ences in Namibia and Kenya. London, Overseas change as scale varies. Empirical research should also Development Institute: 30. be conducted for systems with distinct characteristics including land size, risk preference, and knowledge. Baltas N.C., and O. Korka. 2002. Modeling Farmers’ Such research would assist in classifying what factors Land Use Decisions. Applied Economics Let- affect the suitability of ecoagriculture strategies and ters 9:7 453-457. improve targeting of ecoagriculture approaches. Barbier, B., and G. Bergeron. 1999. Impact of policy Comprehending the impact of land management interventions on land management in Honduras: practices on livelihoods would benefi t from multidis- results of a bioeconomic model. Review of Details ciplinary research approaches that can examine the in Matrix. Agricultural Systems 60:1-16. relevant components of the system at multiple scales Barbier, B., R.R. Hearne, J.M. Gonzalez, A. Nelson, and integrate this information for decision-making and and O.M. Castaneda. 2003. Trade-offs Between policy formulation. Examples of approaches include the Economic Effi ciency and Contamination by Coffee Alternatives to Slash and Burn program of the CGIAR, Processing A Bioeconomic Model at the Watershed and the SCRIPs initiative at the International Food Level in Honduras. Paper presented at 25th Inter- Policy Research Institute. national Conference of Agricultural Economists (IAAE), Durban, South Africa. REFERENCES Barbier, E.B., and C.-E. Schulz. 1997. Wildlife, bio- Abiye, A.A., and J.B. Aune. 2001. Technical Options diversity and trade. Environment and Development for Agricultural Development in the Ethiopian High- Economics 2: 145-172. lands: A Model of Crop-Livestock Interactions. In Economic Policy and Sustainable Land Use. Recent Bolwig, S., S. Wood, and J. Chamberlin. 2003. A Advances in Quantitative Analysis for Developing Spatially-Based Planning Framework for Sustain- Countries, edited by N. Heerink, H. van Keulen and able Rural Livelihoods and Land Uses in Uganda. M. Kuiper. Heidelberg: Physica-Verlag. Washington DC: International Food Policy Research Institute. Albersen, P.J., and S. Laixiang. 2001. Land Degrada- tion as a Transformation Process in an Intertemporal Bontkes, T.S. 2001. Agricultural Prices and Land Welfare Optimisation Framework. In Economic Degradation in Koutiala, Mali: A Regional Simula- Policy and Sustainable Land Use. Recent Advances tion Model Based on Farmers’ Decision Rules. In in Quantitative Analysis for Developing Countries, Economic Policy and Sustainable Land Use. Recent edited by N. Heerink, H. van Keulen and M. Kuiper. Advances in Quantitative Analysis for Developing Heidelberg: Physica-Verlag. Countries, edited by N. Heerink, H. van Keulen and M. Kuiper. Heidelberg: Physica-Verlag.

68 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Brown, D.R. 2003. A Spatio-Temporally Explicit in the Dominican Republic Highlands: Household Model of Land Use Decisions: Shifting Cultivation Probit and Duration Models. In Economic Policy and and Forest Cover Dynamics in the Congo Basin. Sustainable Land Use. Recent Advances in Quantita- Ithaca, NY. tive Analysis for Developing Countries, edited by N. Heerink, H. v. Keulen and M. Kuiper. Heidelberg: Cacho, O. 2001. An analysis of externalities in agro- Physica-Verlag. forestry systems in the presence of land degradation. Ecological Economics 39 (1):131-143. Dougill, A., J. Soussan, O. Springate-Baginski, N.P. Yadav, O. Dev, and A. Hurford. 2001. Impacts of Carpentier, C.L., Stephen A. Vosti, and J. Witcover. community forestry on farming system sustainability 2000. Intensifi ed production systems on western in the Middle Hills of Nepal. Land Degradation & Brazilian Amazon settlement farms: could they save Development 12 (3):261-276. the forest? Agriculture Ecosystems and Environment 82:73-88. Dyer, G., A. Yunez-Naude, and J.E. Taylor. 2001. Ef- fects of Land Degradation in a Diversifi ed Economy Cary, J. and R. Wilkenson 1997. Perceived Profi tability with Local Staple and Labour Markets: A Village- and Farmers’ Conservation Behaviour. Journal of Town CGE Analysis from Mexico. In Economic Agricultural Economics 48(1): 13-21. Policy and Sustainable Land Use. Recent Advances Chakrabarti, M., S. Khaling, J. Bhattacharya, S. Ranjan in Quantitative Analysis for Developing Countries, Ghosh, A. Sarkar, and A. Sarkar. 2001. Functioning edited by N. Heerink, H. van Keulen and M. Kuiper. of Joint Forest Management in the Forests of North Heidelberg: Physica-Verlag. Bengal: Observation from 12 IFRI Sites. Paper FAO. 2001. Conservation Agriculture Study. Rome, presented at a Dissemination Seminar, St. Joseph’s Food and Agriculture Organization. http://www.fao. College, Darjeeling, India, December 2001. org/DOCREP/004/Y271E/y271e00.htm#toc. Chomitz, K., and D.A. Gray. 1995. Roads, Lands, Mar- Feather, P., D. Hellerstein, and L. Hansen. 1999. Eco- kets, and Deforestation. A Spatial Model of Land Use nomic Valuation of Environmental Benefi ts and the in Belize. Washington DC: The World Bank. Targeting of Conservation Programs. The Case of the Chomitz, K. M. 2004. Payments for Ecosystem Ser- CRP. Washington, DC: USDA Economic Research vices. World Bank. Washington, DC. Personal Service. Communication. Feder, G., R. E. Just, D. Zilberman. 1985. “Adoption of Chomitz, K.M., T.S. Thomas, and A.S. Brandáo. 2004. Agricultural Innovations in Developing Countries: Creating Markets for Habitat Conservation When A Survey.” Economic Development and Cultural Habitats are Heterogeneous. Washington DC: The Change 33(2): 255-298. World Bank. Ferraro, P. 2004. Global Conservation Payment Initia- Crissman, C.C., J.M. Antle, and J.J. Stoorvogel. 2001. tives. http://epp.gsu.edu/pferraro/special/special.htm Tradeoffs in Agriculture, the Environment and 2004 [cited March 15 2004]. Human Health: Decision Support for Policy and Ferraro, P.J., and A. Kiss. 2002. Direct Payments to Technology Managers. In Tradeoffs or Synergies? Conserve Biodiversity. Science 298 (5599):1718- Agricultural Intensifi cation, Economic Development 1719. and the Environment, edited by D. R. Lee and C. B. Barrett: CABI Publishing. Fleminga, E., and M. Milneb. 2003. Bioeconomic modelling of the production and export of cocoa for price Darwin, R., M. Tsigas, J. Lewandrowski, and A. Ra- policy analysis in Papua New Guinea. Agricultural neses. 1996. Land Use and Cover in Ecological Systems 76:483–505. Economics. Ecological Economics 17:157-181. Flora, C.B. 1995. Social Capital and Sustainability: de la Brière, B. 2001. Imperfect Food Markets and Agriculture and Communities in the Great Plains Household Adoption of Soil Conservation Practices

Meeting Livelihood Objectives 69 and Corn Belt. Research in Rural Sociology and Kherallah, M., and P. Gruhn. 2001. Market Feasibility Development: A Research Annual 6: 227-246. of Land Use Options for Uganda. Washington, DC: International Food Policy Research Institute. Gebremedhin, B, and S.M. Swinton. 2003. Investment in Soil Conservation in Northern Ethiopia: The Role Knox McCulloch, A., R. Meinzen-Dick, et al. 1998. of Land Tenure Security and Public Programs. Ag- Property Rights, Collective Action and Technologies ricultural Economics 29: 1 69-84 for Natural Resource Management: A Conceptual Framework. International Food Policy Research Gobbi, J.A. 2000. Is biodiversity-friendly coffee fi nan- Institute, SP-PRCA Working Paper No. 1. Wash- cially viable? An analysis of fi ve different coffee ington DC. production systems in western El Salvador. Ecologi- cal Economics 33 (2):267-281. Knox, A. and R. Meinzen-Dick 1999. Property Rights, Collective Actions and Technologies for Natural Re- Greenhalgh, S. and A. Sauer 2003. “Awakening the source Management. CGIAR Systemwide Program Dead Zone: An Investment for Agriculture, Water on Collective Action and Property Rights. Policy Quality, and Climate Change.” WRI Issue Brief. Brief No.1. Washington DC. Washington, DC: World Resources Institute. Krishna, A. 2003. Understanding, Measuring and Uti- Heerink, N., A. Kuyvenhoven, and M.S.v. Wijk. 2001. lizing Social Capital: Clarifying Concepts and Pre- Economic Policy Reforms and Sustainable Land Use senting a Field Application from India. Washington, in Developing Countries: Issues and Approaches. In D.C.: CAPRi, IFPRI. Economic Policy and Sustainable Land Use. Recent Advances in Quantitative Analysis for Developing Krishna, A., and N. Uphoff. 1999. Mapping and Mea- Countries, edited by N. Heerink, H. van Keulen and suring Social Capital: A Conceptual and Empirical M. Kuiper. Heidelberg: Physica-Verlag. Study of Collective Action for Conserving and De- veloping Watersheds in Rajasthan, India. Washing- Irwin, E.G., and J. Geoghegan. 2001. Theory, data ton: World Bank - Social Capital Initiative Working methods: developing spatially explicit economic Paper No. 13. models of land use change. Agriculture Ecosystems and Environment 85 (1/3):7-23. Kruseman, G. and H. van Keulen. 2001. Soil Degra- dation and Agricultural Production: Economic and Jagger, P., and J. Pender. 2003. The Role of Trees for Biophysical Approaches. Economic Policy and Sus- Sustainable Management of Less-Favored Lands: tainable Land Use. Recent Advances in Quantitative The Case of Eucalyptus in Ethiopia. Review of In Analysis for Developing Countries. N. Heerink, H. Matrix. Forest Policy and Economics 5:83-95. van Keulen and M. Kuiper. Heidelberg, Physica- Jansen, H.G.P., R.A. Schipper, P. Roebeling, B.E. H., Verlag: 21-48. H. Hengsdijk, B.B. A.M., and A. Nieuwenhuyse. Leakey, R.R.B. 2001. Win:Win Landuse Strategies for 2001. Alternative Approaches to the Economics of Africa: 1. Building on Experience with Agroforests Soil Nutrient Depletion in Costa Rica: Exploratory, in Asia and Latin America, International Forestry Predictive and Normative Bio-Economic Models. In Review 3: 1-10. Economic Policy and Sustainable Land Use. Recent Advances in Quantitative Analysis for Developing Lee, D. R., and C. B. Barrett, eds. 2001. Tradeoffs or Countries, edited by N. Heerink, H. van Keulen and Synergies? Agricultural Intensifi cation, Economic M. Kuiper. Heidelberg: Physica-Verlag. Development and the Environment. Wallingford, UK: CABI International. Kerr, J. and N. K. Sanghi. 1992. Indigenous Soil and Water Conservation in India’s Semi-Arid Tropics. Legg, C. and D. R. Brown. 2003a. Issues in Modeling IIED Gatekeepter Series. London: 30. Decision-Making Dynamics Along A Gradient of Agricultural Intensifi cation in the Humid Forest of Cameroon. Manuscript submitted to Environment and Development Economics Journal.

70 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Legg, C. and D. R. Brown 2003b. Modeling the Dy- Economics Association Annual Meeting, Montreal, namics of Coupled Human and Natural Systems Canada. http://agecon.lib.umn.edu/cgi-bin/pdf_view. Along a Gradient of Agricultural Intensifi cation in pl?paperid=9057&ftype=.pdf. the Humid Forest of Cameroon. Paper presented at Nunes, P. A. L. D. and Y. E. Riyanto. 2004. Informa- Conference on Reconciling Rural Poverty Reduc- tion as a Regulatory Instrument to Price Biodiversity tion with Renewable Resources, Cornell University, Benefi ts:Certifi cation and Eco Labeling Policy Prac- Ithaca, NY. tices. Paper presented at International Conference Legg, C., and V. Robiglio. Spatially Explicit Modelling on Economics of Sustainable Forest Management, of Landscape Change at the Humid Forest Margin University of Toronto, Toronto, Canada. in Cameroon. Paper presented at Workshop on Inte- Okumu, B. N., M.A. Jabbar, D. Colman, N. Russell. grated Management for Sustainable Agriculture, 2002. A Bio-economic Model of Integrated Crop- Forestry and Fisheries, August, 2001, CIAT, Livestock Farming Systems: The Case of the Ginchi http://www.ciat.cgiar.org/inrm/ Cali, Colombia. Watershed in Ethiopia. In Natural Resources Man- workshop2001/docs/titles/3-1CPaperCLegg.pdf. agement in African Agriculture. C. B. Barrett, F. Lehman, J. 2004. Soil Modeling. Cornell University, Place and A. A. Aboud, CABI International. Ithaca, NY. Personal Communication. Omamo, S. W. 1998. Transport Costs and Smallholder Lewandrowski, J., R. F. Darwin, et al. 1999. Estimat- Cropping Choices: An Application to Siaya District, ing Costs of Protecting Global Ecosystem Diversity. Kenya. American Journal of Agricultural Economics Ecological Economics 29: 111-125. 80(1): 116-123. Lyngbæk, A. E., R. G. Muschler, F.L. Sinclair. 2001. Pagiola, S. 2004. Payments for Ecosystem Services. Productivity and Profi tability of Multistrata Organic World Bank. Washington DC. Personal Communi- Versus Conventional Coffee Farms in Costa Rica. cation. Agroforestry Systems 53: 205–213. Pagiola, S., N. Landell-Mills, J. Bishop. 2002. Mak- Margoluis, R., V. Russell, M. Gonzalez, O. Rojas, J. ing Market-based Mechanisms Work for Forest Maddaleno, G. Madrid, D. Kaimowitz. 2001. Maxi- and People. Selling Forest Environmental Services. mum Yield? Sustainable Agriculture as a Tool for Market-based Mechanisms for Conservation and Conservation. Washington, DC, Biodiversity Support Development. S. Pagiola, J. Bishop and N. Landell- Program, World Wildlife Federation. Mills. Sterling, VA, Earthscan: 261-289. Marsh, S. P. and D. J. Pannell. 1998. What We Think We Pender, J. 2003. Development Pathways for Hillsides Know About Extension and Why Its Not Enough for and Highlands: Some Lessons from Central America Landcare. SEA Working Paper 97/03. http://www. and East Africa. Washington DC, IFPRI. Manuscript general.uwa.edu.au/u/dpannell/dpap973f.htm. Submitted to Special Issue of Food Policy on Less Favored Areas. McNeely, J. A. and S. J. Scherr. 2003. Ecoagriculture: Strategies to Feed the World and Save Wild Biodi- Pender, J., P. Jagger, E. Nkonya, D. Sserunkuuma. 2001. versity. Washington DC: Island Press. Development Pathways and Land Management in Uganda: Causes and Implications. Washington DC, Neill, S. and D. R. Lee. 2001. Explaining the Adop- IFPRI: 118. tion and Disadoption of Sustainable Agriculture: The Case of Cover Crops in Northern Honduras. Pender, J., S. Scherr, G. Durón. 2001. Pathways of De- Economic Development and Cultural Change 49: velopment in the Hillsides of Honduras: Causes and 793-820. Implications for Agricultural Production, Poverty, and Sustainable Resource Use. Chapter 10 in D.R. Newbold, S. C. and M. Weinberg. 2003. Wetlands, Lee and C.B. Barrett, eds., Tradeoffs or Synergies? Wildlife, and Water Quality: Targeting and Trade Agricultural Intensifi cation, Economic Develop- Offs. Paper presented at American Agricultural

Meeting Livelihood Objectives 71 ment and the Environment. Wallingford, UK: CABI Romano, C. B. 2001. Soil Conservation and Imper- International. fect Labour Markets in El Salvador: An Empirical Application of a Dynamic Control Model of Farm Pfaff, A. S. P. 1999. What Drives Deforestation in the Production. Economic Policy and Sustainable Land Brazilian Amazon? Evidence from Satellite and Use. Recent Advances in Quantitative Analysis for Socioeconomic Data. Journal of Environmental Developing Countries. N. Heerink, v. K. Herman and Economics and Management 37: 26 - 43. K. Marijke. Heidelberg, Physica-Verlag: 115-134. Place, F. and B. Swallow. 2000. Assessing the Rela- Schipper, R. A., Hans G.P. Jansen, et al. 2001. Integrated tionships Between Property Rights and Technology Bioeconomic Land-Use Models: An Analysis of Pol- Adoption in Smallholder Agriculture: A Review of icy Issues in the Atlantic Zone of Costa Rica. Chapter Issues and Empirical Methods. CAPRi Working 14 in D. R. Lee and C. B. Barrett, eds. Tradeoffs or Paper 2. Washington DC. Synergies? Agricultural Intensiﬁ cation, Economic Place, F. and B. Swallow. 2002. Assessing the Rela- Development and the Environment. Wallingford, tionships Between Property Rights and Technology UK: CABI International, 267-284. Adoption in Smallholder Agriculture: Issues and Shiferaw, B., S. Holder, B. A.M. Bouman, H. Hengs- Empirical Methods. Innovations in Natural Resource dijk, A. Nieuwenhuyse, F. Sáenz. 2001. Population Management. The Role of Property Rights and Col- Pressure and Land Degradation in the Ethiopian lective Action in Developing Countries. R. Meinzen- Highalnds: A Bioeconomic Model with Endogenous Dick, A. Knox, F. Place and B. Swallow. Baltimore: Soil Degradation. Economic Policy and Sustainable The Johns Hopkins University Press: 45-72. Land Use. Recent Advances in Quantitative Analysis Polasky, S., C. Costello, C. McAusland. 2003. On for Developing Countries. N. Heerink, H. van Keulen Trade, Land-Use, and Biodiversity. St. Paul, MN. and M. Kuiper. Heidelberg, Physica-Verlag: 73-92. http://ﬁ esta.bren.ucsb.edu/~costello/research/papers/ Skonhoft, A. and C. W. Armstrong. 2003. Conservation OnTradeFinal2.pdf. of Wildlife. A Bioeconomic Model of a Wildlife Re- Polasky, S., J. D. Camm, B. Garber-Yonts. 2001. Se- serve Under the Pressure of Habitat Destruction and lecting Biological Reserves Cost-Effectively: An Harvesting Outside the Reserve. NCFC Working Pa- Application to Terrestrial Vertebrate Conservation per Series in Economics and Management, Depart- in Oregon. Land Economics 77(1): 68-78. ment of Economics and Management, Norwegian Poteete, A. and E. Ostrom. 2003. In Pursuit of Compa- College of Fishery Science, University of Tromso. rable Concepts and Data About Collective Action. Tromso: 19. http://www.nfh.uit.no/working_papers/ Washington DC, CAPRi-IFPRI: 33. ifo/skonhoft_og_armstrong20030.pdf. Pretty, J. 2002. People, Livelihoods and Collective Stolton, S. 2002. Organic Agriculture and Biodiversity. Action in Biodiversity Management. Biodiversity, Bristol, UK, IFOAM: 31. http://www.ifoam.org/dos- Sustainability and Human Communities. Protecting sier/biodiversity.pdf. Beyond the Protected. T. O’Riordan and S. Stoll-Kl- Stoms, D. M., K. M. Chomitz, et al. 2004. TAMARIN: eemann. Cambridge: Cambridge University Press: A Landscape Framework for Evaluating Economic 61-86. Incentives for Rainforest Restoration. Landscape Pretty, J. and H. Ward. 2001. Social Capital and the and Urban Planning 68: 95-108. Environment. World Development 29(2): 209-227. Stoorvogel, J.J., J. M. Antle, C. C. Crissman, W. Bowen. Reardon, T. and S. A. Vosti. 1995. Links Between 2003. The Tradeoff Analysis Model: Integrated Rural Poverty and the Environment in Developing Bio-Physical and Economic Modeling of Agricul- Countries: Asset Categories and Investment Poverty. tural Production Systems. Paper for submission to World Development 23(9): 1495-1506. Agricultural Systems, URL: http://www.tradeoffs. montana.edu/pdf/agsystems02.pdf

72 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Stoorvogel, J.J., A. J.M., C. C. Crissman, W. Bowen. of Wildlife Preservation in Agriculture. European 2001. The Tradeoff Analysis Model Version 3.1: A Review of Agricultural Economics 26(4): 443-460. Policy Decision Support System for Agriculture. Zander, P. and H. Kachele 1999. Modelling Multiple User Guide. Wageningen, Laboratory of Soil Sci- Objectives of Land Use for Sustainable Develop- ence and Geology, Wageningen University, The ment. Agricultural Systems 59(3): 311-325. Netherlands: 89. Zbinden, S. and D.R. Lee. 2005. Paying for Environ- Swinton, S. M. and R. Quiroz. 2003. Poverty and the mental Services: An Analysis of Participation in Deterioration of Natural Soil Capital in the Peruvian Costa Rica’s PSA Program. World Development, Altiplano. Environment, Development and Sustain- forthcoming. ability 5(3-4): 477-490(14). van Noordwijk, M. 2002. Scaling Trade-offs Between Crop Productivity, Carbon Stocks and Biodiversity in Shifting Cultivation Landscape Mosaics: The FALLOW Model. Ecological Modelling 149(1-2): 113-126. Varughese, G. 1999. Villagers, Bureaucrats, and Forests in Nepal: Designing Governance for a Complex Resource. Ph.D. diss. Bloomington, IN.: Indiana University. Vosti, S. A. and T. Reardon. 1997. Sustainability, Growth, and Poverty Alleviation: A Policy and Agro- ecological Perspective. Baltimore and London: John Hopkins University Press. Vosti, S. A., Julie Witcover, C. L. Carpentier. 2002. Agricultural Intensifi cation by Smallholders in the Western Brazilian Amazon from Deforestation to Sustainable Land Use. Washington, D.C., Interna- tional Food Policy Research Institute. White, T. 1995. The Emergence and Evolution of Col- lective Action: Lessons from Watershed Management in Haiti. World Development 23(10): 1683-1698. Williams, J. et. al. 2001. The Contribution of Mid- to Low-Rainfall Forestry and Agroforestry to Contrib- ute to Greenhouse and Natural Resource Manage- ment Outcomes. Australian Greenhouse Offi ce/ Mur- ray-Darling Basin Commission, Canberra.: 72 p. World Bank. 2003. Global Prospects Realizing the De- velopment Promise of the Doha Agenda. Washington DC: 44. http://www.farmfoundation.org/projects/ documents/Lewisspeechreport.pdf. Wossink, A., J. van Wenum, C. Jurgens. 1999. Co-or- dinating Economic, Behavioral and Spatial Aspects

Meeting Livelihood Objectives 73 ANNEX 5: Sample of Economic Models Incorporating Livelihood, Productivity and Ecosystem Elements

Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

BIOECONOMIC MODELS (Static models)

Barbier and To determine Na- The model has a biological component, which The popula- This paper provides a theoretical repre- Schulz, 1997 the optimal wild tional examines the change in total species stock tion dynam- sentation of the models. In solving these (Bio- and natural scale based on aggregate biological growth rate ics of the models for the closed economy, it is evi- economic species exploi- across species and expansion in number of species is dent that the conventional bioeconomic model of tation levels species as size of habitat changes. The model incorporated models that do not factor in the opportu- wildlife ex- factoring in op- assumes habitat conversion occurs because of in the bio- nity cost of preserving habitat or the value ploitation portunity cost of other economic activities and all output is con- logical com- of biodiversity may over-estimate optimal and trade) habitat conser- sumed. The model takes into consideration the ponent of long-run levels of wildlife species and vation, and social utility function of the country and exam- the model habitat. In the case of the open economy wider social ines the trade-off between harvesting or utiliza- model it is unclear, when compared to the value of biodi- tion of wild species from conserved natural closed economy, whether the long-run versity. Also, to habitat versus conversion of habitat of land to optimal level of species and habitat con- examine the produce other commodities. Examines two version would be higher. However, com- impact of trade scenarios - (i) a closed economy where the parative statics suggest that trade inter- on optimal ex- objective is to maximize the present value of ventions (made to internalize global val- ploitation of future welfare, and (ii) an open economy in ues of biodiversity) can be counterproduc- species and which the wildlife products can be exported. tive whereas an alternative policy of an habitat in the The latter case has a trade balance equation international transfer of funds would lead long-term and a new definition of total domestic con- to greater long-run conservation of spe- sumption. The objective is to maximize open cies and habitat. economy welfare. The key output of this model is the optimal long-run levels of wildlife and habitat exploitation.

Barbier et al, To identify the Water Bioeconomic model solved with mathematical Incorporated Application to watershed in Honduras: 2003 (Trade- optimal location shed programming. Model minimizes annualized with consid- with collective processing and change in off between for coffee proc- costs for a sub-watershed subject to con- eration of location of processing plant - improved economic essing plants straints (processing capacity, processing quan- water de- efficiency is possible. This will require efficiency along a river by tity, water availability, effluent concentration, mand and providing adequate incentive for produc- and con- examining and investment). Uses GIS to situate the dif- effluents ers to participate and may require coffee tamination) transport, vari- ferent plants and determine which ones would must be be- cooperative assistance. able and fixed minimize cost for the whole sub-watershed. low the pre-

Annex 5 - 1 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

costs Also tests the effect of different coffee premi- determined ums and effluent level restrictions on the loca- maximum tion of the processing plant. The key outputs effluent con- include graphs that show the relationship be- centration tween quintals of coffee processed and water contamination for each of the different locations of the processing plant. It is also possible to generate outputs for different coffee premiums.

de la Briere, Analyze the de- Farm- Intertemporal behavioral model of households' Soil fertility In the application to a watershed in the 2001 (Probit terminants of house labor allocation to agriculture, soil conserva- is incorpo- Dominican Republic: 1. Welfare measures and duration adoption and hold tion, and labor market activities in the context rated as a such as food subsidies are effective in models) maintenance of level of household specific market failures (in this constraint promoting conservation, 2. Food self- soil conserva- case food market). The household maximizes sufficiency induces soil conservation, 3. tion an additively separable utility function over Actual land ownership favors mainte- three periods and over food and income, given nance of soil conservation, 4. Large land a discount rate, the production technologies for holdings are associated with lower soil food and soil fertility, wage rate, land and conservation adoption and maintenance, household labor availability. Using information 5. Less education results in earlier aban- from comparative statics, the model estimates donment of soil conservation, 6. Soil ero- adoption of soil conservation for two scenarios sion on farmers' plots is not totally (without subsidy and with subsidy but for all stopped. households). The model uses duration analyses to assess whether the households continue to practice soil conservation beyond the short-term. The key output of the model can be used to improve targeting of technology transfer.

Jansen et To predict the Farm Bioeconomic model (Technical Coefficient Incorporated Applied to the Atlantic Zone of Costa al., 2001 short-term (less scale Generators (TCG), Linear programming (LP), as con- Rica: Introduction of technological change (UNA-DLV) than 5 years) and econometrics, GIS). Uses individual farm straints results in: small and medium farms in- effects of agri- re- models for representative farms. TCGs are (e.g., include crease cash crop production; haciendas cultural policies gional used to generate coefficients of different crop the expected extend pasture at the expense of frontier on farmers' land scale and livestock production systems and tech- monetary forest areas; banana plantations increase

Annex 5 - 2 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

use decisions nologies. Uses partial model results for the value of nu- use of less suitable soils. For overall re- different farm types to determine total regional trient losses gion with adoption of alternative technol- product supply and factor demand through in farm in- ogy - expansion of crop and pasture is at weighted aggregation. The GIS is used to ar- come objec- the expense of agricultural frontier forests; chive and manipulate geo-referenced data and tive). there is increase in economic surplus and present the results spatially. The key output reduced nitrogen depletion. includes technological options and farmers' reactions and measures of (partial and aggregate) policy effectiveness.

Jansen et To explore the Re- Bioeconomic model (includes: Technical Coef- Sustainabil- Applied to the Atlantic Zone of Costa al., 2001; aggregate effect gional ficient Generators (TCG) for production tech- ity and envi- Rica. Looked at different scenarios and Schipper et of policy meas- scale nologies; linear programming (LP) to select ronmental identified the impact of technology and al., 2001, ures on both optimal combination of production systems; parameters policy on ecological and environmental (Sustainable efficiency and econometrics; and GIS). It is based on the Re- relevant to measures. Found that the improved tech- Options for non-economic gional Economic and Agricultural Land Use the region nology scenario does increase economic Land Use and environ- Model (REALM) which identifies optimal com- are incorpo- surplus and less environmental degrada- (SOLUS)) mental sustain- bination of production systems by maximizing rated in the tion. Similarly restrictions on soil nutrient ability objectives economic surplus generated by the agricultural optimization. depletion have improved benefits but at a regional sector. The latter is formulated as a multi- The model lower economic benefits than technologi- level. market structure. It is a single year model. The assesses cal progress. key outputs include quantification of trade-offs the effect of between economic and sustainability objec- these re- tives which are identified by running the model strictions on for different scenarios or by simulating alterna- economic tive land use systems. surplus.

Skonhoft To understand Re- The model has two components: an ecological The popula- This is a theoretical model that assumes and Arm- the effect of gional and an economic model. The ecological model tion dynam- there is no illegal harvesting occurring strong, 2003 habitat conver- scale captures the change in stock of wildlife within ics of the within the reserve. It finds that land con- (Conserva- sion and har- and outside the reserve. The model uses one species is version triggered by improved profitability tion in re- vesting outside stock of wildlife to represent the whole game incorporated in agriculture does not necessarily have a serves) reserves affects population. This stock is divided into two sub- in the eco- negative impact on the wildlife population conservation populations that are connected by dispersion. logical com- in the reserve. The model also finds that within reserve The economic second component of the model ponent of when harvesting occurs outside the re- looks at a single owner and assumes that the model. serve, there is a smaller stock of wildlife

Annex 5 - 3 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

maximization of current equilibrium net bene- within the reserve. However, when har- fits steers land use and harvesting policy. The vesting occurs and harvesting is profit- benefits include both consumptive and non- able, the impact on stock population in the consumptive value of wildlife. The key output is reserve is ambiguous because the owner an identification of wildlife population levels, has an incentive to create habitat. given different conditions.

Zander and To develop sus- Farm- The model simulates the influence of prices Practical This paper provides a theoretical repre- Kachele, tainable land level and policy regulations on the farmers' decision- measures sentation of the model. 1999 (Multi- use options making and the effect of resulting agricultural on the envi- objective practices on the indicators of sustainability. ronment are decision The model involves a hierarchical structure of incorporated support tool economic and ecological modules. The first using select for agroeco- level develops technical coefficients, the sec- sustainability system ond level calculates the economic coefficients indicators. manage- of site-specific production techniques, the third The paper ment) level evaluates the ecological effects of these talks about techniques, the fourth generates the linear six of them: programming model, and the fifth level starts potential for the subprogram that solves the equation sys- nitrogen tem, analyzes the data and transfers it to GIS. leaching, There is a feedback between the outputs of the potential model and the goal setting. The key output is a wind, water regional sector model which provides overall erosion, pro- economic evaluation, trade-off scenarios, and tection of regional land use patterns which provide an wild flora, overall evaluation of the ecological condition. disturbance index for amphibians and an index for par- tridges.

BIOECONOMIC MODELS (Dynamic models)

Annex 5 - 4 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Albersen To identify and Plot The welfare model is a T-period optimization One of the This study provides a theoretical repre- and Sun, simulate socially and model in which 1. every consumer maximizes vectors in the sentation of the model. Mentions that the 2001 (Social desirable and na- the T-period sum of their own utilities subject transformation transformation function (where land is welfare economically tional to intertemporal budget constraint and dis- function repre- transformed from one state to another) maximiza- efficient trajec- count rate, 2. every producer maximizes his sents the can be incorporated in the profit maximi- tion) tories of invest- own profit for each time period, subject to characteristic zation model of producers, or embedded ment and re- technology and natural condition constraints, of the land in an intertemporal welfare optimization source use for 3. the commodity markets clear, and 4. re- (both inherent framework or other (partial) intertemporal the future. source utilization matches resource availabil- characteristics frameworks that use the transformation ity. The model develops an agent-specific and character- function in combination with resource transformation function that describes the set istics resulting management. This approach can also be of technologically efficient production func- from the deci- used to examine multiple inputs and out- tions. A cost of transformation and degrada- sion makers' puts tion induced transformation process are also behavior. modeled and incorporated into the decision- making process of the economic agent. A key output of this model is a set of investments at the individual and government level.

Barbier and To examine the Wa- The model uses linear programming to A soil erosion The model is applied to a microwatershed Bergeron, effects of state tershe maximize aggregate utility of a whole micro- equation. Yield in Honduras. Finds that changes in mi- 1999 (Dy- policies on d watershed over a five year planning horizon. and erosion crowatershed are due mainly to changes namic/ re- farmers' in- scale The results of the first year of planning hori- parameters in population rather than changes in pol- cursive lin- comes and zon become the initial resources of a new are from bio- icy. Labor is the constraining factor. Ar- ear pro- natural resource model that is solved for the following five physical model eas with low population and high popula- gramming) conditions years. The process is moved one year for- - EPIC (Ero- tion farmers employ soil conservation ward and repeated until the full time period of sion Productiv- measures. Areas with medium population the simulation has been covered. It accounts ity Impact Cal- (when expansion to other land areas for for the whole watershed, and identifies two culator). The agriculture is still feasible) farmers do not social groups (farmers and ranchers). The microwater- invest in soil conservation. elements in the model include a population shed is also and labor, partitioning the watershed, crop divided into production, product allocation, livestock pro- different land duction, soil erosion, perennial and forests, categories EPIC model. Different policies are examined. The output includes simulations of land use across watershed, income per person by

Annex 5 - 5 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

group, simulation of crop productivity, and conservation measures. This is done for the different policies.

Bontkes, To provide Farm Model is an intertemporal simulation model. Incorporated The model is applied in Mali where the 2001 (Re- qualitative in- and The model divides up farms and soils by type as soil fertility base case involves decreasing soil or- gional simu- sight into the re- and the number of farms and soils in each in step two of ganic matter content when no additional lation model) dynamics of gional types is known. The simulation proceeds by the simulation. measures are taken, leading to decreas- agricultural de- determining: 1. the areas per crop and inputs ing millet yields while maize yields in- velopment applied per farm type, 2. availability of nutri- crease as a result of application of fertil- ents for crop uptake, 3. crop production (in- izer and animal manure. Scenario 1 is a cluding pasture), 4. animal production based change in prices for cotton and com- on available quantity and quality of feed, 5. pound fertilizer. The increase in cotton cereal prices, 6. sale or purchase of animals price initially increases area under cotton, based on farm income, 7. changes in number but this eventually decreases as soil or- of farms per farm type and the soil fertility ganic matter decreases. A change in fer- characteristics per farm and soil type. This tilizer price does not have a major impact new situation (7) is used to as the initial con- on cotton area. These changes result in dition for the subsequent year. This informa- reduced soil organic matter, but benefit tion is aggregated to the regional level. large farmers in terms of development. Farmers’ decision-making is simulated using decision-rules that are derived from interpretations of interviews and literature and some socio-economic research. The model produces continuous input-output relationships.

Fleming and To capture the Indus- This is a dynamic stochastic simulation mod- This is incor- Applied to PNG: Have runs that are de- Milneb, 2003 lagged effects try (no ule composed of eight modules for estima- porated in the terministic and stochastic for the various (Dynamic of government geo- tion: economic surplus; area of cocoa trees land suitability, policy options. There is a range of im- stochastic policies, exoge- graphi (with separate modules for each cocoa tree but not really pacts. The most positive impact on profit simulation of nous factors c variety); land suitability; input–output rela- considered in results from a devaluation of local cur- production- and decisions scale) tions; break-even values; export volumes and the analysis rency and reduction in opportunity cost of export mar- made by cocoa values; export prices; and domestic price labor input. Neither short-term price sup- keting) producers on formation. (did not use a dynamic optimiza- port or price stabilization are preferred the industry tion model to estimate the area replanted (from a profit standpoint) to the results of

Annex 5 - 6 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

under trees). The economic surplus module the base-case. A planting subsidy pro- measures the welfare changes. The output is vides a small gain over the base case. discounted economic surplus under different pricing policies for both smallholders and es- tate owners.

Okumu et al, To assess, ex Wa- The model uses a dynamic non-linear Captures the The model is applied to Ginchi watershed 2002 (Inte- ante, the likely tershe mathematical programming framework. The biophysical in Ethiopia: Soil conservation measures grated crop- impact of multi- d model optimizes an aggregate watershed component via may not be a profitable venture given ex- livestock ple technology scale utility function-comprising production, con- the soil loss isting policy conditions (for both long and farming sys- adoption on sumption, profit and leisure. This optimiza- yield decline short term) instead crop rotation is pre- tem) crop-livestock tion function is indirectly linked to the bio- equation. ferred. Under policy of adoption of tech- systems under physical aspects of the watershed through an nology find that tree planting or conserva- various of policy exponential soil-loss yield-decline equation tion measures are not adopted despite scenarios with single-year time lags. And cumulative their profitability. Instead the need for soil losses in previous periods determine food self-sufficiency results in fallowing in yields in the next period. For each location in the crop-rotation pattern. Also improved the watershed, the model calculates the op- human-nutrition goals could be achieved timal fertilizer and dung application rates for simultaneously by improving land user every crop activity and selects the most prof- rights that facilitate the multiple-adoption itable crops for cultivation. Relative prices, of commercial trees, new high-yielding costs and yields are adjusted for the effects crop varieties and increased fertilizer ap- of erosion in each period. The key output of plication. the model is a quantification of tradeoffs in the achievement of indicators of human and biophysical welfare given the current and simulated policy environment and institutional settings.

Romano, To analyze how Farm This dynamic farm profit maximization model Soil quality is Application in El Salvador: For house- 2001 (Dy- lack of access house- analyzes the importance of different factors the index of holds without access to markets, the off- namic con- to labor markets hold on the use of soil conservation practices. The soil character- farm employment is negatively related to trol model of affects soil man- farm production is determined only by family istics and is soil conservation and significant. It is not farm produc- agement deci- labor and soil quality. A farmer maximizes incorporated significant for those with access to labor. tion) sions present value of production with the following as a con- Rural market conditions, therefore, affect

Annex 5 - 7 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

constraints: soil quality, labor availability and straint. behavioral responses of farm households initial soil quality. In the empirical model pro- with respect to soil conservation. duction is a function of farm capital owned, land size operated, labor used in productive activities on farm, soil quality index, and index of total factor productivity. The model separates farmers that have access to labor markets and those that do not. The model uses logit regression. The output provides information on the relationship between off- farm employment and use of soil conservation measure.

Shiferaw et To examine the Farm In this model a farm household maximizes Soil erosion This model is applied in Ethiopia. The al., 2001 inter-linkages house- welfare given available biophysical, human, and nutrient findings for better off households: When (Endoge- between popu- hold capital and information resources, and insti- depletion is family labor is abundant, land is scarce nous soil lation pressure tutional constraints. This bioeconomic model estimated for and labor-intensive conservation tech- degradation) and poverty, uses a non-separable farm household model the different nologies are available, capital market im- their impact on to determine the production, consumption land types that perfections encourage investment in miti- household wel- and investment decisions interdependently in are examined. gating soil erosion. When land is abun- fare and land an intertemporal setting. The decision vari- This in turn dant and shadow value of labor is high, management, ables include crop and area choice by land determines household does not invest in conserva- and the conse- type, levels of fertilizer use, allocation of land, crop yield. tion of soil. For poor households the quent pathways labor and traction power in different seasons, model shows that scarcity of labor relative of development seasonal labor and oxen hiring, livestock to land discourages investment in soil in a low poten- production, selling and buying crops and conservation. For poor, land-scarce, labor tial rural econ- livestock, buying of farm inputs, choice of rich households, the investment in farm omy. consumption good, savings and borrowing of conservation will depend on the off-farm credit, choice of level of soil conservation opportunities. i.e., When markets are investments. A calibrated model is used to imperfect, poverty in key assets (e.g., run simulations. The key output is a privately oxen and labor) limit the ability or the will- optimal land management practice under ingness of a household to invest in con- population pressure. servation.

Annex 5 - 8 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

SPATIAL MODELS

Chomitz et To determine Re- This is an optimization model where the prob- Focuses This is a theoretical model that is applied al., 2004 what the trading gional lem is to optimize the benefits from the land by mainly on to the situation in Minas-Gerais, Brazil. (Transfer- domain should scale determining how much land to leave in forest, forest cover. The findings are that compared to the able devel- be for a TDR how much land to retire from agriculture and, in base scenario of command and control, opment program so as the transferable development rights scenario, the municipal level trading offers small rights (TDR) to minimize cost selling permits. The model simulates different changes in compliance cost. But when program) of compliance policy scenarios, in each of which the equilib- biome-basin level trading is allowed. The while resulting rium of supply and demand for land is deter- trading among landowners with forest in environmen- mined. Also have biodiversity priority areas that area only results in a notable decrease in tally preferable are distinguished and the value of land is from compliance cost. Forest-deficient house- landscapes land sale values. The model estimates the re- holds capture the savings. Similarly, duction in compliance cost and the rents when trading is allowed in the 'forest first' earned by the suppliers of legal reserves. Key scenario (after trading forest area, can output of this model are maps regarding areas trade land retired from agriculture), the where forest cover would be preserved and the social cost of compliance is reduced (less economic and environmental impact under al- than forest only) and again the savings ternative trading domains. accrue largely to forest-deficient households.

Chomitz and To explore the Na- This model assumes that land will tend to be Returns to This model is applied to Belize. The Gray, 1996 trade-offs be- tional devoted to its highest-value use, taking into activities on model reveals that (i) market access and (Spatial tween develop- scale account tenure and other constraints. The the land are distance to roads strongly affect the Model of ment and envi- value of a plot depends on land's physical pro- a function of probability of agricultural use, especially Land Use) ronmental dam- ductivity for that use and the farm-gate prices land suitabil- for commercial agriculture, (ii) high age posed by of relevant inputs. A reduced-form, multinomial ity for the slopes, poor drainage, and low soil fertility road building logit specification of this model calculates im- activity discourage both commercial and semi- plicit values of land in alternative uses as a subsistence agriculture and (iii) semi- function of land location and characteristics. subsistence agriculture is especially sen- The resulting equations can then be used pre- sitive to soil acidity and lack of nitrogen diction or analysis. (highlighting that subsistence farmers are able to judge the soil).

Annex 5 - 9 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Newbold Prioritization of Re- This model uses spatially explicit models of Included in This model has not been applied to a and Wein- sites for conser- gional production functions in an optimization frame- the biologi- specific context. However, it should be burg, 2003 vation activities scale work. The objective is to maximize environ- cal benefits. noted that the model can be modified by (Targeting taking into ac- mental benefits with a budget constraint (and E.g., mallard attaching weights to different environ- conservation count the cost have multiple environmental benefits being abundance mental benefits and to species to reflect programs) and trade-offs considered with different weights added to is a measure their value to society or use a measure of between differ- each). The model has three components, the of habitat taxonomic diversity (Polasky et al, 2001 ent environ- first two are the models of the environmental quality, and as cited in Newbold and Weinburg, 2003). mental services benefits, and the third is the cost equation. Us- reduction of One of the distinguishing features of this ing the equations for the two environmental nitrogen approach is that it takes into considera- benefits, the benefit-cost ratios are calculated loading is a tion the spatial pattern of different land for each spatial cell that could be restored (as- measure of uses and its impact on environmental suming no restoration). The cells with highest water quality benefits. The estimation of environmental benefit-cost ratios are restored and the benefit- benefits benefits could be further refined. cost ratio for cells in which the interaction are updated. The process is then repeated till the budget is completely expended. The key output is a map of sites that should be targeted for restoration.

Pfaff, 1999 Identify factors Plot This approach uses an economic land use Soil charac- Applied to the Brazilian Amazon, this (Economic affecting defor- and model. It allocates land to different land uses to teristics and model finds that own and neighboring model of estration na- maximize profit. The decision model involves vegetation county paved roads increase deforesta- deforesta- tional identifying whether the land use of clearing for- type (cer- tion while distance to national market has tion) scales est results in greater returns than not clearing rado) are an inverse relationship with deforestation. the land. Uses GIS, and spatial data as ex- explanatory Soil quality is positively related to defor- planatory variables for deforestation at the variables estation and low clearing cost of cerrado county level. The output includes findings from can reduce pressure on forest areas, re- a regression analysis. ducing forest clearing.

Annex 5 - 10 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Wossink et To examine the Re- Uses a mathematical formulation of a network Used in the This model is applied to farms in Haar- al., 1999 economics of gional problem. The model has a perception and con- measure of lemmermeer in the Netherlands. Based (Network joint spatial pro- scale joint analysis component through which land biological on the perception analysis, it is found that design mod- duction of wild- managers state their preferences for land use output, using farmers in the region prefer margins in eling and life and agricul- alternatives. The model, using the findings from the most regular crops than grass strops or fallow conjoint ture at the re- the preference and conjoint analysis, examines extensive strips. The perception analysis reveals 4 analysis) gional level and select alternatives. The model minimizes costs species key attributes for compatibility of un- incorporating subject to various constraints such as connec- based state sprayed crop edge with farm organiza- farmers percep- tivity, allowable gap lengths, and other corridor indicator tion: (i) width of the margin, (ii) the type of tion and prefer- requirements (e.g., the cost of leaving un- (available for compensation payment scheme for im- ences sprayed edges is calculated using partial agriculture in plementing the unsprayed crop edges, budgeting and programming techniques). Uses the Nether- (iii) guidance, and (iv) where margins a GIS model (ECONET4) for the empirical ap- lands) should be included in the rotation. Using plication of this model. In ECONET4, can use these findings a spatial model is used to either distance or edge weights (the latter is simulate cost and benefits of 4 different better for selection of cost effective network). strategies of network design. The authors The model provides a spatial representation of find that using selective control lowers a set of edges and land use practices on these cost of wildlife preservation. edges that satisfy the corridor requirements and the cost and wildlife preservation benefits associated with each alternative.

SYSTEMS ANALYSIS FRAMEWORKS

Antle and To simulate de- Farm This model combines econometric production Recognizes Is able to simulate land use in unob- Capalbo, cision-making scale model represented by a system of supply and the spatial served situations. Also finds that when 2001 both within and demand equations and process-based repre- variation in spatial variability in returns is simulated, (Economet- outside of ob- sentation of discrete land use decisions repre- physical the distribution of net returns varies ac- ric process served data in a sented by equations of profit maximization with conditions in cording to the productivity level of the models) manner that is a discrete step function regarding crop choice. the produc- sub-area zones. This results in a non- consistent with The econometric production model calculates tion function linear characterization of supply response economic theory expected net return in the simulation of land and in the (which differs from the constant-elasticity and the bio- use decision. The land use decision for a site is crop models supply function) physical con- made by comparing expected returns for differ- straints and ent activities. For the selected activity the model generates factor demand equations and

Annex 5 - 11 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

processes. determines the input use. This is then integrated into a biophysical model. Land use and other decisions are used to initialize computa- tion of expected returns for subsequent period. The output is a joint distribution of outcomes in the population of land units or farms. It also provides mean net returns for the different price scenarios of the sub-areas via the production function. In this model the economic decisions are based on the spatial and temporal distributions of expected returns.

Carpentier Objective of ASB Multi- Different technical models (bioeconomic, The models The findings vary based on the area et al., 2000; framework is to ple agroecological, etc.) are used to estimate the on carbon where the studies are done. For example, Vosti et al., evaluate alterna- scale elements that go into this matrix. The local sequestra- Carpentier et al (2000) examined the 2002; van tive land-use s. The benefits are measured via productivity and tion, and adoption of four types of intensification Noordwijk, systems based ac- profitability, the regional ones are measured productivity and predicted their economic and envi- 2002; on their ability to tions based on watershed functions and the global take into ronmental impacts using a farm level bio- Suyamto et address interna- for benefits are measured through carbon seques- account the economic linear programming model. al., 2003 tional environ- chang tration and biodiversity. The matrix has a col- biophysical They found that although intensified land (Alternative mental con- e are umn of different (relevant) land uses in the re- elements uses on all cleared lands (including pas- to slash and cerns, agro- at the gion. These are compared along a set of eco- ture) had a low deforestation rate it re- burn trade- nomic sustain- farm, logical, economic, agronomic, and policy crite- sulted in the least amount of preserved off matrix) ability issues, and ria that are relevant to policy-makers and the forest after 25 years. In contrast, intensifi- and farmer na- farmers who would adopt the technology. The cation on forested land, via low-impact adoption con- tional findings from the research are incorporated into forest management, slowed the defores- cerns. and a trade-off matrix. No weighting is given to the tation rate, but deforestation did not stop inter- different criteria in the matrix, therefore those unless timber prices were equal to or na- using the matrix information can make assess greater than a particular amount. The au- tional which land use would be most suitable given thors concluded that in the long run, there policy different objectives. is a trade-off between farm income and levels forest preserved, which results from intensification of land uses on the cleared land. Other findings are discussed in this table (see Suyamto et al., 2003)

Annex 5 - 12 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Crissman et To operational- Re- The approach begins with stakeholders identi- As environ- The approach was applied in northern al,2001; ize an ecore- gional fying the critical dimensions of concern (the mental at- Ecuador and Northern Peru where the Stoorvogel gional approach scale sustainability criteria). These criteria form the tributes of farmers use a range of highly toxic et al., 2001; by linking disci- with basis for formulating hypotheses regarding the land chemicals for pest management. The Antle and plinary models of farm trade-offs between competing objectives. The which in turn scenarios were increase pesticide price Stoorvogel, agricultural pro- land analytical model is based on the Tradeoff affect pro- or IPM. The trade-offs examined were 2003; Stoor- duction and en- as Analysis Model, which is an integrated GIS- duction and those between agricultural production and vogel et al., vironmental unit of based biophysical and economic modeling sys- manage- pesticide loading or neurological damage. 2003 health impacts. analy- tem. This model simulates land use and input ment prac- The model finds that the tax or price pol- (Trade-off This enables an sis use decisions. The model has prices, policy tices and icy typically resulted in movement along a analysis ap- assessment of and attributes of human, physical and biologi- agricultural tradeoff curve and therefore resulted in proach) technology and cal populations affecting land use and input- output the same relationship, whereas the IPM policy analysis. use decisions at the field scale. These deci- solution resulted in improvement in one sions in turn determine agricultural output, en- factor while the other factor was held vironmental impacts and health effects. The constant. analysis results in a joint distribution of management practices, environmental impacts and health outcomes for each unit in production as a function of prices and policy parameters. The outputs include two-dimensional trade-offs curves in terms of agricultural output, environmental quality indicators and health indicators.

Legg and To model the Vil- This is a spatially explicit system-dynamics This is in- Land use patterns change with differ- Brown household level lage / model. Involves mapping the area, location, cluded in the ences in population density and market (2003a); decision-making house ownership and cultivation state of fields. This land patch access. The age of the forest or fallow Legg and for selection of hold model follows the FLORES-type SIMILE mode - with was the most important criteria for choice Brown fallow patches model. There is multiple-instance land patch information of location for new fields. Time required (2003b); and for cultivation model that contains a sub-model for each agri- re: crop sys- for vegetation cleared to dry was an im- Brown cultural system. These sub-models are models tems, and portant criteria. The presence of indicator (2003) (Dy- for the crops and other elements characteristic rainfall species (for fertility), and (for forest melon namic spa- of the system. The multiple-instance household (which influ- fields) tree size were important. With the tially explicit model has sub-models for population, labor ences pro- exception of monoculture fields, the size village availability, stocks and decision-making. Re- ductivity) of the field is determined by food needs of model) spondents scored each field type based on the household and is limited by labor. In- criteria to reveal the importance of these crite- tensification of land occurs for a variety of ria for decision-making. The selection of the reasons, not only the availability of land.

Annex 5 - 13 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

patches was then based on a "suitability index". These findings differ slightly for each vil- An optimization model with a time step of half a lage. This information is then used for the year was used to rank the suitability index of simulation to determine patch selec- each land patch for a household. A key output tion/patch size. is a map of distribution of different patches, and therefore information on what lands are left fallow and for how long and what is cultivated in different places.

Stoms et al., To evaluate and Re- TAMARIN builds on idea of setting general en- Incorporated No specific findings - this is a planning 2004 compare differ- gional vironmental goals and then using economic in the pa- tool. This has currently being pilot-tested (TAMARIN) ent landscape scale instruments to induce voluntary compliance. It rameter set- for the Atlantic Rainforest area of Brazil. configurations to simulates a range of programs involving eco- ting of a sce- achieve conser- nomic instruments with flexible specification of nario. vation goals - program eligibility, payment rules and budget recognizing key for the program. The model allows for detailed principles (see specification of spatially varying opportunity comments) costs. The model uses three layers of spatial information in GIS - current land cover, ecologically distinctive sub-areas, and opportunity cost of conservation expressed as a value, which is assigned to each point on the landscape. There are four steps: defining the parameters of the scenario, selecting the sites for the portfolio, projecting future land use in the region based on portfolio, and evaluating the effects on the social, economic, and landscape configuration factors. van Noord- To explore the Land- The simulation model predicts food self- Includes Suyamto et al (2003) application in Su- wijk, 2002; simultaneous scape sufficiency, soil fertility, carbon stocks, plant carbon stock matra found that a 25 percent forest Suyamto et impacts of or species richness and watershed functions on within forest, cover of riparian forest has the lowest al., 2003 changes in land water- the basis of a number of biophysical and man- agroforest, sediment load into rivers, compared to (Forest, use at a land- shed agement parameters for a 100 grid-cell land- fallow area, allocating forest to steepest slopes or Agroforest, scape scale - scape. The model also includes options for and below ridge tops. The farmers' response to price

Annex 5 - 14 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Low-value examines the scale harvesting forest products and for changing the ground. As- shocks depends on their adaptive capac- Lands or watershed func- food crop based agricultural system into an sesses ity to experiment and reorganize land use Waste tions within a agroforest or tree-crop system. It includes a change in patterns. van Noordwijk highlights the (FALLOW) dynamic land- simple water balance that predicts the resulting biodiversity importance of interpreting field-derived model) scape impact on the landscape using certain rules. from leaving data in the context that they were derived The key outputs are trade-off curves. An addi- land fallow. rather than as system properties. The tional output in van Noordwijk (2002) is a set of Also Incor- number of trade-offs between productiv- parameter values for the critical point at which porates bio- ity, carbon stock and biodiversity depend the watershed functions 'crash'. physical on the scale of model application, the in- elements in ternal variability of the landscape, and the the land use transient behavior under changes in land part - e.g., use intensity. soil erosion, sedimentation, storm flow, etc.

MACRO-LEVEL AND CGE MODELS

Darwin et Look at how Re- This model uses the Future Agriculture Re- Environ- The following situations are simulated in al., 1996 global changes gional sources Model (FARM) to evaluate impacts of mental this paper: climate change with changes (Impact of in climate, hu- scale global climate change on agricultural systems framework - in water supply and distribution of land in Land cover man populations worldwide. This model has a GIS and a com- enters into different land classes, population growth change and international putable general equilibrium (CGE) component. the produc- with regional increases in population and (GIS-FARM trade policies The GIS links climate variables to land and wa- tion function: labor, and trade policy (in the agricultural model - might affect ter resources in FARM's environmental frame- has climate, sector) by adjusting difference in pro- CGE)) tropical forests work. The model also uses land class informa- length of ducer price and foreign market prices tion. The economic framework links land, wa- growing (i.e. duties/tariffs). Global climate change ter, and other primary factors with production, season, run- threatens the health and integrity of trade and consumption of 13 commodities in off, water tropical forest ecosystems and the biodi- eight regions. The model uses revenue from information, versity within them. As more area is con- sale of these primary production factors to pur- land class verted to agricultural purposes, the for- chase consumer goods. Multiple situations are information. estlands in moist tropical regions will de- simulated. The outputs provide information on cline slightly (therefore deregulation of land use and land cover change in moist tropi- agricultural trade will pose a small cal areas under simulated scenarios. threat).

Annex 5 - 15 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Dyer et al., To capture the Village This village-town CGE model is built on a se- These are The model is applied in Mexico. It reveals 2001 (Vil- "second and level ries of models of agricultural households en- modeled that the high degree of diversification in lage-Town higher-round" (re- gaged in crop production and other economic implicitly - the households shields them somewhat CGE Analy- effect of land gional) activities, including migration. The farm house- for example, from land degradation. This model is sis) degradation on holds maximize utility, defined on consumption changes in useful to understand the effect in situa- households goods and savings. The general equilibrium the bio- tion of market imperfections. For exam- other than those equations local market-clearing conditions for physical ele- ple, when the main staple crop's market directly affected. factors and goods, a village-town savings- ments (i.e. in the village is closed (i.e., high transac- investment balance, and a trade-balance equa- land degra- tion costs to trade with town), land deg- tion. The model provides information on the dation) re- radation results in a change in endoge- percentage change in fixed wage and endoge- sults in nous price and may result in more inten- nous wage in the different scenarios. changes in sive staple production and therefore productivity more degradation. of crops. Or land is retired because it is degraded.

Lewandrows To assess how Global The model uses the Future Agricultural Re- Incorporated The findings from this model include: ki et al., setting aside of scale sources Model (FARM), which consists of GIS this in the protecting five percent of the world's land 1999 (GIS- land reduces the and a computable general equilibrium model land classes area under this strategy reduces global FARM gross domestic (CGE model). It presents sectoral and regional identified for GDP (1990) by approx. $45.5 billion. In- (CGE) product and trade-offs associated with incremental increase set aside - creasing the land protected to 10 and model for gross world in the global land area set aside to protect bio- tundra, for- 15% increases the cost to $93.3 billion cost of land product, and to logical resources. The CGE model extends the est tundra, and $143.8 billion. Most of the global set aside) examine the im- Global Trade Analysis Project (GTAP) model tropical sea- costs (45%) fall on the densely populated pact of a 10 per- by including heterogeneous land endowments, sonal forest, and highly developed regions of Japan cent global re- introduction of water as primary input in the etc. and EC. The United States will have to tirement plan on crop, livestock and service sector and modeling bear 15% of the cost. regional crop, crop production as a multi-output sector. Pro- livestock and duction is determined by profit maximization forestry sectors and assumes competitive markets. To simulate implementation of land set aside, 5% to 15% of land is removed from economic use within each region-land class using 1% increments. The model output provides information on the

Annex 5 - 16 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

regional economic and land use impact of land retirement at the different percentage levels.

Polasky et To identify cost Re- The model address a budge constrained Uses bio- This model is applied to conserve terres- al., 2001 effective strate- gional maximal covering local problem. The model geographic trial vertebrate species in Oregon, USA. (Selecting gies for conserv- scale uses an integer programming formulation of data that The model reveals that the level of cov- Biological ing maximum reserve site selection. The model maximizes describes erage is less costly using a budget- Reserves) number of ter- the number of species included in the network the range of constrained approach than a site- restrial species of reserves subject to two constraints. The first each spe- constrained approach. For conserving given a conser- is that a species is not included if sites in which cies in order 350 species, the budget-constrained ap- vation budget it occurs are not selected and a budget con- to know if a proach for site selection lowers costs by straint (with expenditure being measures as the species if 10 percent. Moreover, covering the ma- opportunity cost of land). The output is a cost- found in cer- jority of the species costs less than cov- coverage curve and a map of which sites tain sites. ering the remaining 10 percent of spe- should be selected. The output can often have cies. The model can be modified to multiple combinations of sites that generate an maximize a more complex function than optimal solution. number of species (e.g., a function that recognizes species value, etc.). This analysis is a first step in an effort to establish reserves.

Polasky et To analyze the The model is a two-good two-country trade The model A theoretical application of the model al., 2003 effects of trade model. The two goods in the model, grain and uses spe- compares an autarky versus free trade (Trade and on land use and timber, are each produced by a fixed ratio of cies-area situation. The effect of trade on global Biodiversity trace the likely labor and land. Only certain lands (habitat curves biodiversity depends on the degree to model) effects of land types) are capable of producing each good. It species endemism in each country. The use changes on is assumed that once land is converted to a number of native species surviving in biodiversity productive use, it is incapable of supporting country 1 in a free-trade equilibrium de- native biological diversity. Land that is not con- pends on the distribution of species verted for production remains as natural habitat among habitat types and how this capable of supporting species. The number of changes with productive use of land. remaining species remain is determined by a Citizens utility levels under autarky or species-area curve relationship. Other assump- free trade depends the relative strength tions include that consumers in each country of their preferences for species conser-

Annex 5 - 17 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

have identical Cobb-Douglas preferences over vation versus consumption of private timber and grain, and have separable utility goods. The model reveals that the in- over consumption goods and species conser- creased specialization associated with vation. The model maximizes social utility func- trade can have important consequences tion. The utility function values both local spe- for patterns of habitat conversion. On the cies conserved and global species conserved. other hand, trade may increase welfare in situations with low endemism or with low importance of species conservation relative to consumption of private goods.

MIXED MODELS

Bolwig et al., To estimate the Re- DREAM is a menu-driven software for evaluat- Agroclimatic Application in Uganda: Finds that in- 2003 (Dy- potential con- gional ing economic impact of agricultural research suitability creased production in cotton does result namic Re- sumer and pro- or na- and development. It is a single-commodity (based on in increased benefits and that this is oc- search ducer benefits, tional model designed to measure economic returns length of curring in areas which are not close to Evaluation and export reve- scale in predefined regions to commodity-oriented growing pe- the markets. for Man- nues of simu- research under different market conditions. Its riod, not on agement lated changes at application can be made spatially explicit - as is soil or soil (DREAM)) the subnational the case here. The output includes maps that pH, etc.) is and national provide information on spatial distribution of incorporated levels. The esti- producer surplus from the changes in produc- in the defini- mated impacts tivity. tion of the are always rela- land classes tive to a baseline (‘no change’) scenario

Annex 5 - 18 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

Cacho, 2001 To determine the Wa- Maximizes the discounted value of net revenue Incorporated The model is applied to dryland salinity (Analysis of optimal land use tershe over a rotation cycle. In the calculation of the in the yield agriculture in Australia. The model uses externalities) mix between d net revenue, it includes the monetary benefits function a 30-year rotation. In the model the value forestry and ag- scale of the agricultural crop and the monetary bene- which is de- of forest externality is critically affected riculture in the fits of the forestry operation. This is based on pendent on by both the initial state of the land and presence of for- the marginal benefit of forestry being equal to the state of the discount rate. In the multiple rotation est externalities. the marginal benefit of agriculture. The value of the land example, the optimal decision rule spe- Emphasis here the forest benefit is the improved agricultural (and recog- cies the area of trees to plant as a func- is on converting yield because of the trees impacts on the bio- nizes that tion of the state of the land at the time of agriculture to physical elements and the benefits of selling the state of the decision. Also the level at which the forestry - im- the trees at the end of the rotation - does this the land de- water table should be kept, depends on proved land profor one 30 year rotation and for five 30 year pends on the discount rate. Therefore optimality ductivity in agro- rotations. This is to capture the context of sus- the previous does not imply sustainability (because forestry system tainability. The model output provides graphical state of based on discount rate, could have water information on changes in discounted net pre- land) levels that are higher than is sustain- sent values as the area under forest changes, able). assisting to identify the optimal points in the graphs.

Feather et To use non- Na- This method uses spatially disaggregated static Integrated in Application in the United States: conclu- al.,1999 market valuation tional models applicable across a wide geographic the model as sions include that the benefits of fresh- (Non-market techniques to scale area. It estimates the benefits of CRP looking part of step water recreation is greater than those valuation) quantify the en- at recreational uses. The model uses a higher two - where associated with pheasant hunting. How- vironmental resolution than regions for the geographic and the physical ever change in consumer surplus is benefits and im- behavioral data. The three models use different effects are greater for the pheasant hunting than the prove targeting activities in the calculation of the environmental translated freshwater recreation (suggesting that farm acres amenity. Each model is a three step model: (i) into biologi- CRP provides a greater contribution to determining how the CRP acreage creates cal results the benefits of pheasant hunting). This physical effects, (ii) translating these physical information is then used as baseline for a effects into biological results, and (iii) examin- simulated targeting mechanism ing how the biological results affect consumer welfare. For steps (i) and (ii) assumptions or indicators or extrapolations are used because of limited information. Step (iii) is the economic calculation of consumer surplus. The three models illustrate different means of accounting for variations in the price and quality of the site.

Annex 5 - 19 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

The model output reveals the changes in level of consumer surplus associated with CRP. It also provides information on the relative value of the different recreational uses.

Gobbi, 2000; To evaluate the Farm The model is used for the different production Biological Gobbi (2000) finds that the net present similar study financial viability plot systems. The model 1. estimates parameters and bio- value of biodiversity friendly (BF) coffee is done by of investing in of production and sale for the typical farm for physical fac- (except for commercial polyculture) is Lyngbæk et the conversion each production system, 2. computes invest- tors are not more sensitive to declines in yields than al, 2001 of a coffee plan- ments necessary to certify the farm as Biodi- explicitly anticipated premium levels. In terms of (Benefit-cost tation to a biodiversity Friendly, 3. estimates the production measured. It mean net present value (NPV) it de- analysis of versity friendly costs and sales of the farm once certified, 4. is implicitly creases in the following order: farms with certified coffee planta- uses Monte Carlo approach to incorporate risk assumed monoculture (above 1200m), farm with crop) tion. for production and price variables, 5. estimates that the dif- traditional polyculture, farm with techni- expected net present value considering the ferent pro- cally appropriate shade at more than situation certified as biodiversity friendly versus duction sys- 1200m, farm with technically appropriate non-certified. Also examines the sensitivity of tems have shade at less than 1200m, and farm with investment to declines in production due to varying im- commercial polyculture. Small farmers shade effects and to declines in premium lev- pacts for may need additional help and incentives els. The output provides a financial analysis of biodiversity. to adopt the BF certification criteria (if adopting biodiversity friendly crop production they are not starting with low shade under the different production systems. cover).

Jagger and To do an ex ante Farm The economic component of the analysis in- The ecologi- This form of analysis is applied in Tigray, Pender, benefit-cost level volves a multiple year cost-benefit analysis cal compo- Ethiopia. The study reveals that the main 2003 (Bene- analysis based (therefore with discount rate). Estimate a base nent of the factor influencing household or commu- fit cost on community case scenario and then consider the influence analytical nity decisions to invest in tree growing analysis of and village level of variable harvesting periods and potential method in- are the costs and returns of the invest- crop use) survey data to crop losses (due to nutrient and water uptake volves re- ment. The harvesting period and the op- illustrate the ef- by eucalyptus) related to negative externalities. viewing lit- portunity cost of land (especially where fect of planting Benefit cost analysis is based on simple pa- erature rele- eucalyptus is planted on cropland) af- eucalyptus. To rameters and sensitivity analysis is used to re- vant to fects the rate of return. Management of make policy rec- flect the key variables hypothesized to influ- Eucalyptus woodlots by villages or individuals has ommendations. ence returns to investment for tree planting. cultivation in higher rate of return than those managed Have institutional barriers to obtaining rights to Northern by higher administrative level. The find-

Annex 5 - 20 Ecosystem Source (Name of elements approach) Application Scale Characterization of methods and outputs in model Findings

harvest both timber and NTFP, therefore esti- Ethiopia ings however do not involve a detailed mate potential rather than actual benefits. The based on understanding of the markets. model estimates private internal rates of return studies in estimates for the different scenarios. areas that are topog- raphically and climati- cally similar.

Pender et To identify the Com- Identification of the pathways of development No biophysi- Pender et al (1999) apply this approach al., 1999; major pathways munity involves classifying the primary and secondary cal or biodi- to 48 communities in central Honduras. Pender et of development level occupations in the communities. Then empiri- versity vari- They have six pathways of development al., 2001; that have been cal analysis involves comparing descriptive ables in- - 1. Basic grain expansion communities, Pender, occurring, their statistics across pathways, econometric analy- cluded. 2. Basic grains stagnation communities, 2003 (De- causes and im- sis, and qualitative information from the survey. These con- 3. Coffee expansion communities, 4. velopment plications for Looked at econometric analyses to identify fac- siderations Horticultural expansion communities, 5. pathways) agricultural pro- tors affecting pathway, factors affecting house- are included Forest specialization communities, 6. ductivity, natural hold responses, factors affecting collective re- by examin- non-farm employment communities. The resource sus- sources, and factors affecting outcomes. They ing whether pathways are distinguished by compara- tainability and estimate direct and indirect impacts of marginal conservation tive advantage, including agriculture po- poverty changes in factors that overlap in the different measures tential, population density, access to regressions on various outcome measures. are adopted markets and technology. Changes in The model provides an understanding of the (e.g., mini- poverty are not pathway dependent, but key factors that explain the development path- mum till, agriculture and resource management ways. incorpora- are. In the case of horticulture, coffee tion of crop and non-farm employment pathways, residue, productivity outcomes are more favor- etc). able, but the implications for resource conditions are mixed.

Annex 5 - 21 Chapter 6

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems

6.1 Introduction areas (see Annex I-C). Key points from the interviews are included in this chapter, along with supporting There is a burden of proof on proponents of ecoagri- materials derived from recently published literature. culture, as suggested in Chapter 2, to show that eco- These provided the empirical basis for the presentation logically-based land-use and resource management in Chapter 3. We have cast our net widely, looking for practices can achieve enough complementarities and any and all documented fi ndings that could bear on this synergies, while avoiding tradeoffs, so that agricultural subject. The categories of knowledge and experience productivity is enhanced while environmental objec- reviewed in this chapter are: tives are also achieved. If the farming systems proposed under the heading of ecoagriculture are to be attractive 1. Agrobiodiversity conservation and utilization; and acceptable, and thus to help meet world food needs 2. Organic agricultural production systems; and conserve biodiversity, they will have to demonstrate 3. Agroforestry; positive-sum dynamics rather than just zero-sum (or worse, negative-sum) outcomes. 4. Systems approaches to pest management; To assess the opportunities for capitalizing on such posi- 5. Integrated nutrient management; tive-sum dynamics, we reviewed the state of knowledge 6. Soil health; and practice in a number of areas that can be characterized as integrated agricultural management systems. 7. Contributions of below-ground biodiversity to sus- The topics chosen cover the breadth of research that tainable crop production; underlies promising innovations in agricultural science 8. Management of the hydrological cycle; and production. Proponents of these various systems are 9. Conservation tillage/conservation agriculture; and often yet not aware of ecoagriculture as an overarching, emerging fi eld, but their methods enhance and exploit 10. System of Rice Intensifi cation. natural ecosystem processes in ways that can support Because many of these areas of agricultural endeavor agricultural biodiversity, resource conservation, and are relatively recent, there is not in all cases an extensive sustainable food production. They can contribute to literature to draw on. This means that claims should be ecoagriculture especially if links can be made to protec- tentative, pending more confi rmatory evidence. We re- tion of wild biodiversity. port here on things that most scientists who have worked In-depth interviews were conducted with Cornell fac- in the area for some years agree on, though there is not ulty and staff having expertise in these different topical unanimity on all points. Where there is disagreement,

74 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations we note this, trying to be fair to all informed points of Secondary environmental benefi ts are increasingly be- view. ing focused on in scientifi c assessments of the agricultural sector (NRC 2003). This poses diffi culties when 6.2 Assessing Ecoagricultural Alternatives doing economic evaluations because these benefi ts often are “externalities,” not credited (or debited) to the How feasible is it to achieve greater long-term agricul- cropping system adopted. Evaluations of ecoagricultural sustainability and food supply by relying more on tural alternatives—to assess whether they are desirable above-ground and below-ground biological processes, or not from a societal perspective—should internalize e.g., interactions among fl ora and fauna, and organic any externalities, both positive and/or negative, that can matter decomposition, to promote agroecosystem be clearly attributed to the system or practice. Where benefi ts such as nutrient cycling and disease and pest net benefi ts are demonstrable, it becomes incumbent control in a manner that produces abundant crops from upon government decision-makers to fi nd incentives ecologically diverse, robust, and productive farmland and/or restrictions that make agricultural practices environments? which are favorable in economic, social and/or envi- This question summarizes the central opportunities and ronmental terms more attractive to resource managers challenges for ecoagriculture. The same question ap- and the public. plies, with appropriate changes in wording, to livestock Different evaluation criteria are needed for different production. The answer which emerged from the inter- kinds of agricultural enterprises, and there can be no sin- views and literature review is that increasing attention to gle solution or best-system for all of agriculture. Being natural and biological cycles, which are manifestations concerned as we are with food security and livelihood of biocomplexity, can contribute substantially to the generation as well as biodiversity conservation, we are improvement of nutrient and water cycling, pest con- particularly concerned with the incentives and benefi ts trol, and crop and livestock productivity. By mobilizing that affect smaller producers in the agricultural sector, these processes in support of agricultural production, and with households that derive their livelihoods mostly the effi ciency and sustainability of operations can be from agriculture, rather than primarily with larger pro- improved at the same time that environments for bio- ducers. However, the latter make a major contribution diversity are enhanced, moving agricultural production to helping feed urban populations, and their impacts parameters rather than simply making tradeoffs within on the natural resource base can be greater over time them. because of their reliance on mechanization and agro- Analyses of cropping systems diversity and biological chemicals. So we reiterate what has been said already, function require the use of multiple research techniques, that ecoagriculture pertains to—has potential relevance including both traditional factorial experiments under for—the whole agricultural sector, not just smallholder controlled conditions and on-farm systems-related production units. Inasmuch as ecoagriculture is based monitoring studies (Drinkwater 2002). Identifying on sound scientifi c principles, it should be available and critical parameters that refl ect soil health and system useful to larger farming operations as well. productivity is essential for knowing what is happening Modern agricultural systems are facing instances of within an agricultural system (and why), not just what declining yield and factor productivity with increas- results are obtainable with certain methods. Fortunately, ing frequency, due to the consequences of intensive monitoring studies and simulation models are becom- management, including compaction of soil and organic ing more robust in their capacity to evaluate complex matter depletion (Wolfe, pers.comm.). Integrated ecosystems. Many of the tools needed for monitoring and logical management practices can overcome the poor evaluation are still under development. Periodic and yields associated with such depleted systems and can location-specifi c evaluation should become a refer- address ecological weaknesses associated with inten- ence point for helping farmers choose from among a sively-managed conventional agriculture systems that number of improved management practices that ad- have forgone the benefi ts of free or low-cost ecologi- dress their particular needs and opportunities (H. van cal services. What are the available means for tapping Es, interview).

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 75 these potentials? The following sections review what limited farmers, such as chemical signaling to insect has become or is becoming known about them. predators (Lewis et al. 1997), success in polyculture (Cox et al. 2002), and tolerance of nutrient limitation 6.3 Agrobiodiversity Conservation And (Kamara et al. 2003). Utilization Well-designed strategies for the conservation and pro- The promotion of agrobiodiversity often focuses prin- motion of agrobiodiversity would include: cipally on the conservation of crop genetic resources • in-situ conservation of agrobiodiversity, including (Paroda et al. 1999; Hammer et al. 2003). This is expect- habitat protection of wild populations of fl ora and ed to have benefi ts for future improvements in crop or fauna, with maintenance of native species and tra- animal genetic potential whether through conventional ditional agroecosystems; breeding or transgenic programs. This gives proponents of the most modern agricultural development a stake in • ex-situ germplasm conservation, for development farming systems that preserve agrobiodiversity, and not and introduction of improved varieties; and sus- just for presently commercially exploitable species or tainable uses of biodiversity, including agricultural their close relatives. Advances in technology have made systems management and scientifi c research (Long it possible to derive benefi ts from the genetic potential et al. 2003). of quite varied species (Tanksley and McCouch 1997). To this, we would add the identifi cation, evaluation and So the stake of high-tech agricultural science in the aims conservation of soil organisms, from microorganisms and achievement of ecoagriculture, with its emphasis to macrofauna, classifi ed in functional terms, for the on preservation of agrobiodiversity, is growing. The sake of maintaining soil health and fertility over time. in-situ conservation of wild crop relatives and non-crop This could be subsumed under the fi rst point above, but organisms is recognized as critical to the maintenance the unseen aspects of agrobiodiversity have been too of long-term agricultural sustainability (Meilleur and often and easily overlooked, so explicit attention needs Hodgkin 2004). to be directed toward the preservation of subterranean On-farm conservation of biodiversity is a stated prior- biodiversity. ity of the Biodiversity Convention from the 1992 Rio The importance of agrobiodiversity is ubiquitous. Conference, but implementation of agrobiodiversity One of most striking examples of this is in the state of conservation policies have focused primarily on creat- California, which has one of the most productive agri- ing and operating ex-situ repositories for crop genetics cultural sectors in the world. The sector is thought of as (Wood and Lenné 1997). While genetic resources are well-endowed because of the fertility of California soils, an important part of raising and sustaining agricultural the abundance of capital invested in the sector, and the production, breeding and biotechnology should not be education and talent of its labor and management. But the sole or primary concern of agrobiodiversity conser- the success of all California’s productive enterprises vation. Management of agricultural genetic resources depends, as mentioned in Chapter 2, on a diversity of that promotes eventual uniformity of crop genetics biodiversity that extends from crops and livestock to could give no option but to preserve genetic material their wild relatives, and beyond this to a vast array of only in repositories, but this would become narrower pollinators, symbionts, competitors, pests, parasites, and narrower in its base and in its potential. predators, and biological control agents, as well as soil organisms of myriad sorts (Qualset et al. 1995). 6.3.1 Strategies for More Diverse Genetic Resources Long-term food security, even in a system as modern as California’s, depends fundamentally on the conser- Along with the conservation of genetic resources, vation and utilization of genetic resources and native agricultural sustainability requires a wider focus on habitats so that the intricate and far-fl ung web of plant, breeding for diverse crop traits that are often ignored but animal and microbial organisms that help produce that that could be of particular value for smaller, resource- state’s food and fi ber remains intact. Desired commer-

76 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations cial crops are not obtainable from the land, and certainly • Productivity of pastures has been shown to increase not at acceptable cost, without maintaining the vitality with increasing species diversity, with the greatest of the complex ecological systems within which agri- gains occurring with the incorporation of three to cultural production processes are nested . four species, with diminishing though still positive gains beyond this (Fick interview). Studies in Eu- 6.3.2 Related Considerations ropean grasslands have similarly observed a correla- Some specifi c observations that emerged from our tion, though not necessarily linear, between species review of the literature in this area include: biodiversity and productivity (Hector et al. 1999). • A spatially and temporally diverse set of crops will • Agrobiodiversity can increase wildlife habitat, pro- have a rooting distribution that can exploit different ductivity, and resource conservation (Alkorta et al. soil resource pools (Jama et al. 1997). Total produc- 2003). As the same time, agriculturally biodiverse tivity from multiple crops is greater than with single farming systems can maximize ecosystem services, crops, provided that appropriate weed controls can contributing to food security and promoting agricul- be maintained. Currently, the most common controls tural sustainability (Thrupp 2000). are chemical, but biological means are becoming bet- 6.4 Organic Agricultural Production ter known and more widely used, such as mulches, Systems crop rotations, and modifi ed spacing and timing of cultural practices. Organic agriculture is quickly emerging as a popular • Introduced cash crops have dramatically reduced and productive farming enterprise. For many decades, agrobiodiversity in many areas, even as remote as in the face of the development of intensive “Green Rev- the cold desert valleys of the Himalayas (Kuniyal olution” agricultural systems, organic agriculture often 2004). This has adverse consequences over time for considered a fringe enterprise, based on “intuition” at the sustainability and resilience of farming systems, best and “junk science” at worst. Yet, in the past 10-20 with eventually rising costs. Enhanced food security years, the changing context in which modern agriculture is more likely with a mixture of traditional and in- is practiced and an accumulation of scientifi c evidence troduced crops. (e.g., Kumar et al. 2004) have widely reoriented thinking. In a feature article considering this trend, Nature • Agricultural diversifi cation in the Philippines has noted that: been found to increase food security, decrease the [i]n the past, organic agriculture has been set risk of temporary food shortages, and improve squarely against intensive farming and chemi- nutrition (Frei and Becker 2004). The increase of cal-based agribusiness… in the media, these production and income with highly-specialized arguments rage more fi ercely today than ever production systems did not, in this study, enhance before. Yet behind the harsh rhetoric, a little- the well-being of farm households even if there noticed convergence of views is taking place. were short-term economic benefi ts as assessed in For decades, the study of organic farming sat at the fringes of the green revolution in aggregated econonomic terms. agriculture, as intensive techniques marched • To measure the impact of disturbance on diversity, it across the world, sending yields skyrocketing. is necessary to identify and monitor species assem- But mainstream agronomists are becoming concerned about the long-term sustainability blages, e.g., functional groups, not just individual of this approach, and are increasingly focusing species (Belaoussoff et al. 2003). A study in Guelph, on soil integrity…. Mainstream agronomists Ontario, concluded that diversity indices were not now acknowledge, for example, that intensive useful for detecting the possible effect of disturbance farming reduces biodiversity, encourages ir- on assemblages of carabid beetles, which feed on reversible soil erosion, and generates run-off insects that from a human perspective are mostly that is awash with harmful chemicals, including nitrates from fertilizers that can devastate harmful. Such effects are easily overlooked. aquatic ecosystems. (April 22, 2004, p. 792).

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 77 6.4.1 Difficulties in Evaluation organic (non-chemical) production methods is usually not as great in the fi rst year of use as in subsequent The focus of discussion is turning, reinforced by con- years, as organic matter and soil populations increase sumer preferences for chemical-free food supplies, from over time with the addition of biomass amendments and whether or not organic farming is preferable, to whether plant roots’ exudation. Accordingly, long-term trials it can be productive enough to compete economically are needed for valid comparisons of the productivity of with conventional practices. Given that costs of produc- alternative production systems when comparing organic tion can be lowered by avoiding or reducing chemical with conventional practices. inputs, the evaluation often turns on the cost and productivity of labor and capital inputs in these alterna- 6.4.2 Measurement of Effects Associated with tive systems, as well as whether consumers will pay Organic Agriculture for the added value of food considered to be healthier. There is little debate about the desirability of reducing One of the most relevant assessments is by Mäder et agrochemical usage, even if there is not agreement on al. (2002) who contrasted organic and conventional differences in nutritional value and taste. management systems in a long-term experiment, using identical crop rotations, varieties, and tillage for all There is little disagreement that organic production is, treatments. They documented a 20 percent decrease in in principle, more compatible with the conservation yield associated with organic practice, along with 30- of biodiversity, including wild biodiversity, because 54 percent reductions in nutrient inputs and 97 percent of effects of chemical use beyond the farmed fi eld on reduction in pesticide use. The organic plots exhibited fi sh, birds, reptiles and other fauna. The question is not increases of 10-60 percent in nutrient-use effi ciency, easily resolved, however, whether the differences are soil fertility, phosphorus cycling, and aggregate stabil- signifi cant because all agricultural practices to some ity; of 40 percent in biomass and in mycorrhizal sym- extent reduce biodiversity compared with uncultivated bioses; of 130-320 percent in microbial and earthworm or ungrazed conditions. decomposition; and of 200 percent in biodiversity and One diffi culty in making evaluations is that when arthropod abundance. practicing organic agriculture within a larger landscape Organic systems are commonly associated with in- where there can be impact from nearby chemically-increased populations and increased diversity of carabid tensive or biotechnological agricultural practices, the beetles, weeds, earthworms, and soil microbes (Heyer et transfer of pollen, dust or water from surrounding farms al. 2003). Ecoagricultural research and experimentation can affect organic crops. The extent of such effects is can be expected to reduce the costs presently associated still a matter of controversy, but positive and negative with organic systems (increased management complex- impacts do not appear to be symmetrical in that it is ity, reduced yields, and some tillage-based erosion) and easier to move from a more to a less biodiverse situa- to increase benefi ts derived from capitalizing on more tion than the reverse. robust ecosystems processes (reduced inputs and input An additional diffi culty in evaluating organic agriculture costs; increased profi tability; weed, pest, and disease is that the balance of costs and benefi ts changes over control). Whether yield will necessarily decline with time in a path different from conventional agriculture. organic methods remains a matter of contention. The The latter is more often subject to diminishing returns rice production system discussed in Section 6.12, for and increasing costs of production as the effi cacy of example, shows increases in output associated with a chemical fertilizer and agrochemical sprays declines, reduction in chemical inputs when other methods that so that more of these production inputs must be used to favor soil biological activity are introduced. maintain output levels. This is often referred to as “the 1 chemical treadmill.” Conversely, especially on soils 1 Recall from Chapter 2 that pesticide use in the U.S., which that have been exposed to inorganic fertilizers and vari- has gone up 10-fold since World War II, has nevertheless ous chemical biocides, which have a depressing effect seen the percent of crop losses due to insect damage increase on populations of soil organisms, the productivity of from 7 percent to 13 percent (Pimentel 1997).

78 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations 6.4.3 Improving Genetic Resources for Organic rous, potassium, calcium, and magnesium to the soil, Agriculture increased soil organic matter and available phosphorous and potassium, and increased nitrogen storage (Clark Organic farms presently rely largely upon crop variet- et al. 1998). Overall, organic practices resulted in ies bred for conventional farming systems. There is a increased carbon and larger pools of stored nutrients, need for crop (and animal) breeding programs that focus owing to the effects of organic fertilizers and cover on organic alternative systems, so that the plants and cropping. These are, however, still fragmentary research animals being managed are ones better able to exploit results. Only a fraction of the research effort previously naturally available nutrient pools and biologically di- devoted to chemical-based agricultural practices has verse systems (VanBueren et al. 2002). Development gone into studying the chemical, physical and biologi- of improved varieties for use in organic systems should cal dynamics and effects in soil systems under organic enhance the profi tability of organic farming beyond farming practices. Most of what has been measured present levels. and documented so far has pointed toward soil and Crops bred in this manner may also benefi t conventional above-ground environments that are more hospitable to farming systems, because they are likely to exploit a biodiversity, supporting more robust and cost-effi cient broader diversity of available resources. Objections production, even if not always giving the highest yields have been raised to crop or animal breeding programs in agronomic terms. that, to obtain the highest possible yields, have made selections based on “ideal” conditions, producing 6.4.5. Organic Agriculture Research Needs plants and animals that are accordingly more dependent For the potential of organic agriculture to be both more on external nutrient provision and protection against (optimally) productive in economic and food-supply pathogens. crops or animals bred to thrive under more terms and more benefi cial for maintaining biodiversity, normal conditions their there are a number of areas in which research should be undertaken. Faculty working with organic agriculture 6.4.4. Assessing Systematic Effects identifi ed the following focuses for more systematic “Organic farming systems aim at resilience and buff- investigation: ering capacity in the farm-ecosystem by stimulating • Improving integrated nutrient management recom- internal self-regulation through functional biodiversity mendations for organic systems; in and above the soil, instead of external regulation through chemical protectants” (Van Bueren et al. 2002). • Developing improved crop varieties for organic Measures of system robustness (soil health, below- systems; ground biodiversity, agricultural diversity, contribution • Development of low- and zero- tillage organic farm- of community structure to pest and disease resistance, ing systems; nutrient cycling, and crop productivity) would favor • Overcoming the yield defi cit between organic and the goals decreasing inputs, decreasing pollution and 2 resource loss, maximizing productivity, and maximizing conventional practices; ecosystem services. • Refi nement of measures for evaluating the benefi ts Labile pools of soil organic matter (SOM) are defi nitely of organic systems: system robustness and function; affected by long-term management practices, with or- characteristics of and impacts on biological systems; ganic systems supporting a larger microbial biomass soil health; temporal trends in organic matter, nutri- and more active decompositional processes (Fliessbach ent availability and productivity; and and Mäder 2000). Biologically-active soil organic mat- 2 Note that the system of rice intensifi cation, discussed in ter is positively affected by cover cropping and organic 6.12, has demonstrated in many countries the ability, without nutrient inputs (Wander et al. 1994). reliance on fertilizers or chemical inputs, to raise production substantially, even in the fi rst season of use. Thus, it should Transition from conventional to organic farming sys- not be assumed that organic production methods will always tems resulted in increased inputs of carbon, phospho- give lower yields.

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 79 • Improving the understanding of below-ground pro- This year’s World Agroforestry Congress (WAC) is fo- cesses, interactions among soil chemical, physical cusing on the maturing of agroforestry as a sustainable and biological components, and contribution of land-use strategy over the past 25 years. The keynote natural processes (disease resistance, nutrient avail- address (Nair 2004) documents the solid scientifi c ability, etc.) to crop yield. foundations and notable knowledge gains that have been made concerning the nature, extent, and processes As noted above, there has been comparatively little of how trees and crops when growing together. Their research undertaken on organic agricultural systems multiple interactions affect their respective and mu- and practices, refl ecting the prevailing scientifi c para- tual productivity and alter or sustain the environment digm and the availability of private-sector support for at levels ranging from individual plants and plots to research. As noted at the start of this section, after years ecosystems and regions. of separation and antagonism, there is now a growing convergence of interest concerning conventional and The keynote concludes that the rigorous scientifi c organic agriculture. The label “organic” has often been methods now available for doing biophysical and so- expressive of values more than empirical differences, cioeconomic research and for measuring the benefi ts of with critics no less than proponents basing arguments on agroforestry now enable us to make a host of defi nitive, conventional wisdom rather than on objective study. scientifi cally valid statements about the roles and potentials of agroforestry. Nair challenges the agroforestry Fortunately, this polarization appears to be waning, research community to capitalize on these opportunities as many agriculturalists—scientists as well as prac- and work with the rest of the scientifi c world and society titioners—turn their attention more and more to soils to realize the unique potentials offered by agroforestry and below-ground processes (e.g., Kumar et al. 2004), strategies for addressing the problems of food security recognizing that these are the foundation for all agri- and environmental protection. culture, including livestock production. What nurtures and conserves healthy soil systems, with their immense 6.5.1 Contributions of Agroforestry to Multiple biodiversity, is good for long-term agriculture and thus Objectives, including Wild Biodiversity for food security and livelihood generation. The numerous ways that agroforestry can contribute 6.5 Agroforestry to achievement of the Millenium Development Goals of the United Nations have been documented by the Agroforestry, like ecoagriculture, is a land-use strategy Director-General of the World Agroforestry Center with the triple objectives of agricultural production, (Garrity 2004). Among the most promising pathways natural resource conservation, and livelihood enhance- for increasing on-farm food production and income ment. Since its inception as a fi eld of research and a through agroforestry are: fertilizer tree systems, tree focus of development intervention from the late 1970s, cropping, and improved tree product processing. aspirations and claims about agroforestry have centered on the conjunction of ecological, economic and social These advances can be helpful also in addressing the benefi ts that can come from diversifying production need for more enterprise opportunities on small-scale through the integration of trees and other large woody farms, more equitable returns to small-scale farmers, perennials into farming operations. reducing child malnutrition, and alleviating national tree product defi cits. In addition to food, fertilizer and In recent years, agroforestry research has begun to adopt wood products, which can enhance livelihood oppor- a landscape perspective. This is attractive because it tunities, there are mechanisms being developed that enables scientists to scale up their results once they have would compensate the rural poor for environmental a reasonable understanding of small-scale processes. services that their agroforestry practices provide to This is important because the functions of tree cover society. Although these social benefi ts have not yet manifest at landscape, regional and global scales as a explicitly included wild biodiversity, as the value of result of larger-scale patterns and processes (Schroth this resource gains appreciation, movement toward and Sinclair 2003).

80 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations ecological service valuation and compensation is likely regions, such as shortening fallow periods or reduc- to lead to this outcome. tion and simplifi cation of shade canopies in tree-crop plantations, generally reduces the habitat value of these A recent book entitled Agroforestry and Biodiversity systems for native species. Conservation in Tropical Landscapes (Schroth et al. 2004), edited by a large team of distinguished scientists, Schroth et al. (2004) highlight the key role of manage- explores in-depth how agroforestry practices can help ment for providing habitat for various forest species. promote biodiversity conservation in human-dominated This is true whether windbreaks, hedgerows and in- landscapes, synthesizing the current state of knowledge tercropped plantations are being managed on a short in the fi eld and identifying areas where further research rotational or a semi-permanent basis, no matter what is needed. This volume will be of great value to those the structural complexity and diversity of their shade who are designing ecoagricultural strategies in tropi- canopy and understory or the degree of weeding, pol- cal regions or trying to identify knowledge gaps and larding and pesticide use (p. 494). research needs. Furthermore the abundance and diversity of plant and Schroth and associates explore three hypotheses on how animal species will be greatly infl uenced by the size of agroforestry can help to conserve tropical biodiversity: the agroforestry system and its location, particularly 1) by reducing the economic pressures to deforest re- its proximity and degree of connectivity to remaining maining forest land and to degrade forest through the forest cover. The authors also emphasize the role of unsustainable extraction of its resources; 2) by pro- hunting pressure on the animal diversity and abundance viding suitable habitat for forest-dependent plant and of agroforestry, observing that present levels of wildlife animal communities; and 3) by creating a biodiversity- consumption in many tropical regions are unsustainable, friendly matrix of regulated land-use areas to facilitate regardless of the land-use regime. movement between existing patches of natural habitat Schroth et al. endorse “matrix management” in the de- and to buffer them against more hostile land uses. sign and implementation of biodiversity conservation The volume brings together much research-based evi- strategies. The book explores the role of agroforestry dence for how specifi c agroforestry practices, applied in providing a smooth transition between open agri- in particular economic conditions with respect to land, cultural areas and forest boundaries, one that reduces labor and capital availability, and in the presence of edge effects and the incursion of fi re into forest areas, specifi c institutional support factors, can maintain or and further provides connectivity between patches of enhance the abundance and/or diversity of wild species primary habitat. The authors conclude that evidence in a land use system. Not surprisingly, it concludes that supporting the hypothesis that the conservation value of there are a multiple biological, socio-economic and habitat fragments is greater if they are imbedded within institutional factors that condition the effective inte- a matrix of agroforestry than if they are located in less gration of agroforestry into each of these biodiversity diversifi ed, structurally simpler agricultural land uses is conservation strategies. still mainly indirect. But they note that direct evidence is slowly accumulating (p. 495). Regarding agroforestry as habitat for native plant and animal species in agricultural landscapes, Schroth and 6.5.2 Issues of Genetic Diversity associates (2004) explore the relationships between land-use intensifi cation and the suppression of wild The issues of biological and genetic diversity manage- species. They conclude that where high diversity of ment in agroforestry are extremely complex (Atta-Krah wild plant and animal species occurs within agrofor- et al. 2004). Recent studies on the functional elements estry systems or landscapes, this is usually the result of traditional agroforestry systems and modern tech- of extensive management or temporary abandonment nologies demonstrate that the practice of agroforestry of cultivated areas rather than specifi c management to has been a system for the management and conserva- promote its persistence (p. 492). They stress that in- tion of diversity. Yet agroforestry research, over time, creasing intensifi cation of land-use practices in tropical has deemphasized the biological diversity issue in its

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 81 evaluation. Since farmers value diversity and manage ing better the spatial and temporal variation in the agroforestry from this perspective, notwithstanding the composition and function of forest gardens (complex presumed benefi ts to society, Atta-Krah and his co-au- agroforests) should support broader applications of thors propose cross-disciplinary research to advance our multi-functional agroforestry systems in a variety of understanding of species and genetic diversity at both settings. A thorough review of biodiversity in complex inter-and intra-specifi c levels. Such knowledge should tropical agroforests is offered in Chapter 10 of Schroth inform to strategies for enhancing both agrobiodiversity et al. (2004). and wild biodiversity within agroforestry systems. An analogous agroforestry system in temperate North Certain agroforestry practices stand out among those America is known as “forest farming” (Hill and Buck that are best suited to harbor and foster biological di- 2000). This system integrates high-value under-story versity. Home gardens, known for the effi cient nutrient and mid-story crops into mixed deciduous hardwood cycling offered by their multi-species composition, stands to improve the short-term economic returns to have been documented throughout the tropical world the forests that are being managed also for long-term to conserve biodiversity, while providing relatively timber production and wildlife habitat. By improving secure livelihood support through product diversifi ca- the value of standing forest through production of tion (Kumar and Nair 2004). Given their proximity to native specialty products including medicinal herbs, the homestead, home gardens are not likely to offer mushrooms, fruits, nuts, syrups and craft materials, it suitable habitat for many human-adverse wild spe- is hypothesized that economic pressures to convert such cies. However, the relatively sustainable productivity forests to less “habitat-friendly” uses will be reduced. that these complex systems exhibit may contribute to To date, studies have not been conducted to test this intensifi cation strategies that can serve to release land hypothesis, however. from agricultural production elsewhere for the benefi t of wildlife conservation there. Also, not all fl ora and fauna 6.5.4. Implications for Biodiversity that need preservation are adverse to human proximity Biodiversity conservation in agroforestry systems has so long as their growing environment is favorable and been the explicit focus of several studies that have they are not impinged upon. shown higher biodiversity in agroforestry systems than in other agricultural systems (Griffi th 2000). In some 6.5.3 Agroforestry Variations cases, the levels of species richness are equivalent to “Forest gardens,” sometimes referred to as “nature-anal- those of forests (Estrada et al. 1993; Perfecto et al. ogous agroforestry systems,” are often reconstructed 1996). However, there are many contingencies affect- natural forests in which wild and cultivated plants ing these relationships. Perfecto et al. (2003) examined co-exist, thus integrating production and biodiversity the species richness of birds, fruit-eating butterfl ies values (Wiersum 2004). Known also as “complex and ground-foraging ants along a coffee intensifi ca- agroforests” (Michon and de Foresta 1999), these eco- tion gradient represented by a reduction in the number logically sustainable systems have been documented of species of shade trees and the percentage of shade to “conserve a good portion of the original biodiver- cover in coffee plantations in Southern Mexico. They sity” (p. 395). They are often quite dynamic in species found that responses of the three taxa differed along composition, in response to changing socioeconomic this intensifi cation gradient, and there was no correla- conditions. tion between them. There is agreement within the agroforestry research This suggests the importance of distinguishing among community that these complex systems have not been different levels of shade, and also that different taxa adequately researched, relative to the benefi ts that they may respond to habitat changes at different scales. could come from more thorough knowledge. Special- Considering economic, institutional and biological fac- ists who study these systems argue that understand- tors together, Current et al. (1995) found that farmers in

82 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Central America are able to exercise long-term preser- identifying, monitoring and controlling the populations vation of their farms as well as of the biodiversity as- of pests as well as benefi cial species. sociated with them, when they have legally recognized land tenure, access to bank credit, and an agricultural 6.5.5. Issues of Domestication management scheme that includes a complex mosaic of Tree domestication is widely perceived as an area of fruit trees, timber trees and annual crops that can offset agroforestry practice and science that can substantially ecological and economic risks. enhance the productivity and profi tability—and thus Huang et al. (2000) have assessed methods for quan- the adoptability—of agroforestry systems (Leakey and tifying the effects of agroforestry on biodiversity Tomich 1999; Simon et al. 2000; Leakey et al. 2002). conservation. Their interdisciplinary team developed Domestication involves the selection, propagation a functional-group-based Analytical Hierarchy Process and management of trees to adapt them to agricultural (AHP) and a distinctness index of functional groups. systems. This could give value to a wider range of spe- With these tools they ranked the effectiveness of cies than currently managed in intensive, monocropped various agroforestry practices and plantations for the systems of production. This research-supported process protection of natural forests with a long-term view. The will be most effi cient and effective if domestication authors propose that the AHP and functional-groups initiatives are farmer-driven and market-led, according index will be useful for the integrated planning and to Simon and Leakey (2004). evaluation of agroforestry management in biodiversity These authors propose that the objective of agroforestry conservation areas, in addition to improving knowledge domestication should not be just to select super-species on the potential roles of agroforestry in biodiversity or provenances, but should include the promotion of conservation. genetic diversity and matching intra-specifi c diver- Agroforestry has been characterized by Shanker and sity to the needs of farmers, markets and plurality of Solanki (2000) as an “eco-friendly land use system environments. With different farmers using different that is favorable for the management of insect pests and provenances, on a landscape basis, there could be sig- benefi cial insects,” affi rming its contribution to agro- nifi cant diversity within any particular species that is biodiversity. These authors show how the biodiversity dominantly in use. Young and Boyle (2000) point out found in particular agroforestry systems is amenable to that out-crossing will result in gene-mixing in seed incorporation into systems of integrated pest manage- production, with a continuous creation of new hybrids ment (IPM), especially reliance on biopesticides and and complexes that would broaden the genetic base. biological control, practices discussed in the following section. These techniques in turn provide an ecologi- 6.6 Systems Approaches To Pest cally-favorable setting for economic ventures such as Management apiculture, sericulture and lac culture, which they cal- It is a common observation, often considered a truism, culate can be highly profi table. that complex, healthy landscapes resist the spread of A fuller exploration of agroforestry IPM systems in pests and diseases. That undesired organisms and patho- temperate countries of Africa, Asia and Latin America gens exist in almost all situations is true, but their mul- has been provided by Dix et al. (1999). They focus on tiplication and dominance is usually a result of certain the importance of incorporating a blend of short- and imbalances: changes in rainfall or temperature patterns long-term, multiple-generation techniques for managing from the usual or normal; changes in the combination pests in multiple crops. They also point out the role of of nutrients available, including certain scarcities or managing “edge vegetation” in diverse agroecosystems surfeits; changes in the nutrients available at cell level, so as to enhance native natural enemies of pests. This making for malnourished cells more vulnerable to pest international team of scientists proposed an agenda for and disease attack.3 improving IPM practices within agroforestry systems, Maintaining robust agricultural production systems focusing on knowledge needed and techniques for with diverse components is one way to facilitate natural

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 83 control of pests and diseases that adversely affect crops preventive, pre-emptive and prophylactic measures as and animals in farming systems. Natural habitats associ- well. Precise application and careful timing can reduce ated with farming systems can enhance agrobiodiversity the negative effects of chemical use, as can the replace- and promote benefi cial organisms that combat, check ment or more-toxic with less-toxic chemicals. But these or control pests or pathogens. It is, of course, possible are not optimizing solutions, as therapeutic chemical that such habitats can harbor pests and pathogens that treatments lead to the build-up of toxic products, pest have adverse effects on crops and cause unacceptable resistance, secondary pests, and pest recurrence. losses. This is one of many instances where location- and system-specifi c, knowledge is needed to make 6.6.1.2 Revised Approaches optimizing decisions. What our growing understanding These improved methods, which include integrated of agroecology is pointing to is to give up “shoot fi rst pest management (IPM), biocontrol, and biotechnol- and ask questions afterwards” approach to crop and ogy, can be effective by themselves. But they work animal protection. best if utilized in ways that work with ecosystem strengths, not simply in a therapeutic manner. A broad 6.6.1 IPM Alternatives interpretation of IPM can comprise all aspects of a A focus on integrated ecological management can pro- systems approach to sustainable pest management. vide long-term, lower-cost and sustainable pest control, However, in practice, IPM has often been made into a in contrast to the chemical interventions that attempt monitoring program where chemicals are used on an to create an empty ecological niche, making it devoid as-needed basis as determined by economic thresholds. of pests but also devoid of natural processes that limit While this may reduce chemical application rates and pest abundance (McRoberts et al. 2003). So far, most improve farm profi tability, it does not promote more IPM approaches have been only partial. We use the term sustainable systems of natural pest control. Similarly, “systems approaches to pest management” because this although biocontrol—introduction—introduction ofof pest-specificpest-specific refers to managing whole agricultural systems for the control organisms—has often been very successful in objective of reducing damage by pests and diseases, controlling certain pests of some perennial crops, it rather than “managing the pests” per se, as the term has been less effective with annual crops. For the lat- IPM implies. ter, treatment is mostly a therapeutic rear-and-release approach, not focusing on how to strengthen natural A volume written for the National Academy of Sciences biocontrol capacities in the agroecosystem. Biotechnol- (Lewis et al. 1997) has assessed the potential benefi ts ogy holds defi nite promise for controlling pests and of total systems pest management, comparing three diseases, but the most successful applications will be approaches: to strengthen systemic resistance to pests, rather than 6.6.1.1 The Therapeutic Approach develop biopesticides. It should be possible to integrate biotechnology into sustainable, ecologically-based pest Treat pest outbreaks with countermeasures, such as management strategies. chemical sprays. The logic of this approach leads to 6.6.1.3 Recommended Approach 3 This refers to the theory of “trophobiosis,” which maintains that organisms what are optimally nourished are better able Lewis et al. (1997) endorse ecosystem management, to withstand or resist predation, infection, etc. This is on its with a focus on crop attributes and multitrophic interface a reasonable proposition, but it has been slow to gain scientifi c acceptance. It has been implicitly the foundation actions and therapeutic applications carefully used to of FAO’s very successful IPM program, which maintains that have minimal disruption of ecosystem dynamics. They the best way to reduce pest and disease damage is to “grow recommend solutions that achieve net benefi ts at the healthy plants,” i.e., to provide spacing, moisture regimes, total ecosystem level by harnessing inherent strengths etc. that optimize growth. Under such conditions, damage for controlling pests and diseases within specifi c ecosys- and losses remain below the economic threshold where it tems, and that work in cooperation with natural systems would be profi table to intervene with chemical prophylaxis or control.

84 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations rather than compete with them. The authors elaborate 6.6.2 Conditions Affecting IPM Effectiveness on this alternative with these recommendations: Crop heterogeneity is one solution to the vulnerability • Ecosystem management: Redirect pest management of monocultured crops to disease. Intra-specifi c crop to incorporate year-round soil, weed, crop, water, and diversifi cation provides an ecological approach to dis- associated practices at farm and community levels, ease control, as was demonstrated by the interplanting considering the effects of these practices on the of rice varieties susceptible and resistant to blast in overall fauna, nutritional state, and balance of local China (Zhu et al. 2000). ecosystems. Develop multiple-function interventions such as cover cropping that provide nutrients, IPM, which has been very successful in protecting Asian erosion control and habitat for benefi cial organisms. rice crops, has been less successful in sub-Saharan Af- Integrate cover-cropping with conservation tillage rica (Orr 2003). This is apparently because economic and crop rotation. Avoid large-scale monocropping incentives are less when agricultural productivity is which creates conditions for pest and disease suc- limited by depleted soil nutrients and soil biota and cess. Manage fi eld boundaries to promote habitat for there is less immediate response to reduced chemi- natural pests. Try to benefi t from synergies within cal application, changing the economic incentives for natural processes.• using methods that are more labor-intensive even if capital-saving (Orr and Ritchie 2004). IPM in Africa • Crop attributes: Consider crop plants as active par- may therefore make sense only as part of a holistic farm ticipants in multitrophic patterns of crop-soil-water- management strategy that fi rst focuses on overcoming nutrient interactions together with a multiplicity of nutrient limitations and restoring soil biodiversity. soil organisms, and do not neglect these traits when breeding for crop fi tness and production. Investigate In the United States, IPM implementation and progres- how soil, nutrition, and water affect the expression sion toward national program objectives is monitored of traits that control a crop’s interactions with pests and documented using the USDA Performance Plan- and pest-control agents.• ning and Reporting System (http://www.pprs.info) on a state-by-state basis. The National Roadmap for IPM • Therapeutic interventions: Use external inputs with (2003) presents an overall goal of reducing the use of the aim of bringing pest organisms and pathogens highly-toxic, broad-spectrum pesticides, but otherwise into acceptable bounds with as little ecological does not seek to limit pesticide usage, focusing on the disruption as possible, rather than attempting to number of farms implementing IPM as a measure of completely remove the organism. Avoid toxic, broad- success. From this data source, we know that: spectrum pesticides, and use therapeutics only as a secondary tool, to be used in association with the • IPM was being implemented on about 71 percent of promotion of robust natural defenses against pests. U.S. cropland in 2000, compared with 40 percent in 1994; application of the highest-toxicity pesticides Lewis et al. (1997) conclude from their review of al- had declined by 13 percent over this period. ternatives that: • Overall pesticide application had, however, increased Truly satisfactory solutions to pest problems by 4 percent, indicating that IPM as currently prac- will require a shift to understanding and promoting naturally occurring biological agents ticed is not targeting reduced pesticide usage as a and other inherent strengths as components of priority. total agricultural ecosystems and designing our The research needs highlighted in the National Road- cropping systems so that these natural forces map for IPM focus on precision-use of pesticides, keep the pests within acceptable bounds…. A fundamental shift to a total-system approach for not mentioning ecosystem or natural controls. Issues crop protection is urgently needed to resolve identifi ed include the regulation of pesticides, ground- escalating economic and environmental conse- water contamination, and evolving pest resistance. quences of combating agricultural pests. IPM management strategies are usually subdivided by crop type, rather than by farming system, which

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 85 defl ects attention from the evaluation of inter-crop more effective than aggregated diversifi cation, e.g., pest dynamics. The U.S. national IPM program could intercropping or fallow strips, in promoting spider abun- provide a powerful boost to adopting ecoagricultural dance. In-fi eld habitat, more than fi eld-edge habitat, management techniques if research were prioritized seems to be critical for supporting in-fi eld pest control to address total-systems pest control with the goal of by spiders, while diversity at the landscape level pro- reducing pesticide application while strengthening motes airborne sources of colonization. These results natural ecosystem processes for pest control. imply that healthy populations of insect predators are best achieved through the promotion of high crop-resi- For improving the precision management of chemical due, low-spray systems. or biological agents at IPM thresholds, data on climate (temperature and water balance) can be used to predict Predatory carabid beetles, another benefi cial generalist the outcome of competition between crops and weeds. insect, have been found to be healthier and larger and This allows farmers to live with a higher threshold of to experience increased reproductive success in small tolerance for weed occurrence under conditions favor- fi elds with large perimeters and in landscapes contain- able to crop growth (Susan Riha, pers. comm.). As ing a higher percentage of perennial crops (Bommarco more such research is done, looking at when and how 1998). Of fi ve monitored farms, the three that were pests and diseases contribute to crop losses in a multi- managed organically exhibited greater spatial heteroge- factorial manner, it should be possible to make further neity, crop diversity, and percentage of perennial crops. refi nements that are benefi cial both to farmers and to While this evidence is not extensive, it was carefully the environment. gathered and is consistent with many other studies.

6.6.3 Multi-Species Contributions to Pest Control Food availability is an important habitat-limiting factor for generalist arthropod predators, including carabids, Plant diversity at plot, fi eld, and landscape scales can cicindelids, mantids, and web-building spiders (Bom- promote more stable environmental systems, with natu- marco 1998). Diverse, low-spray agricultural systems ral processes controlling the spread of disease and the that foster non-pest insect communities in fi elds and development of resistant pest varieties. Diversity can fi eld-edges are therefore expected to provide more provide a dampening effect on the spread of pests and robust non-chemical control of pests than will large, diseases, through natural competition with a diverse uniform, low-residue, annual monocultures because the biota. Release of natural biocontrol agents has provided latter are more likely to maintain signifi cant populations very successful control of pests in certain cases, particu- of predatory insects. larly for perennial crops such as cassava (Sseruwagi et al. 2004). Large-scale “rear-and-release” programs for 6.6.4 Agricultural Effects on Field Margins annual crops, however, are not as effective if the biocon- Field boundaries are important semi-natural environ- trol agent is not able to establish a permanent popula- ments in agricultural environments. While it is true tion due to unavailability of suitable habitat or climate that cultivated fi elds may “encroach” on natural en- severity. With more knowledge, it should be possible to vironments, the converse of this, given the way that devise conservation strategies that support agricultural ecosystems work, is that the latter in turn “intrude” on benefi cials, identifying and developing them at both the the former, unless inhibited by chemical or other prac- fi eld and landscape scales (Weibull et al. 2002). tices. Plant assemblages in the herb layer of hedgerows Sunderland and Samu (2000) have provided a review respond to a complex set of environmental variables that of the effect of agricultural diversification on the are related to the diversity within the associated farming abundance and pest-control potential of spiders, which systems. Close collaboration between ecologists and are generalist predators that have been demonstrated agronomists is necessary to develop fi eld-boundary signifi cant pest-control benefi ts. “Interspersed diver- management and planning strategies that optimize the sifi cation,” including undersowing, partial weediness, presence of non-cultivated fl ora and fauna that benefi t mulching, and reduced tillage, has been shown to be cultivated crops (LeCoeur et al. 2002).

86 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Agricultural operations such as chemical application, biotypes (Hald 2002). This points out that border areas, can certainly affect the fl ora of fi eld margins, with im- while important for conserving soil resources, may not plications for the spread of pests and benefi cials. Fauna play a strong role in conservation of wild biodiversity. are of course also affected. Improved management Perhaps they represent instead a fertile niche within strategies, including maintenance of grass or wildfl ower agricultural systems for biomass production. border-strips, have a generally positive effect on popu- Regardless of farming system, farmers can, if they want lations of benefi cial species. Biodiversity of the fi eld to, reduce their environmental impact on near-fi eld margins is increasingly recognized as having important areas by reducing the intensity of their management implications for conservation of wild biodiversity, as at the edges of fi elds. Herbicide, pesticide and manure discussed by Marshall and Moonen (2002) for European applications can be halted a few meters from the fi eld agriculture. edge to provide for a transitional zone, and care can be A comparison of fi eld-margin border plots sown with taken so that a meter of unplowed ground is left along grasses found increased occurrence of non-sown spe- the edge of drainage ditches to avoid erosion. cies under organic management. Naturally-regenerating Much remains to be learned about managing pest and fi eld-border plots (not seeded) took longer to establish diseases in crops (and animals) in ways that effectively groundcover, but eventually demonstrated increased capitalize upon the control potentials of natural eco- species diversity and increased presence of Coleop- systems within which the particular agroecosystem tera. Field margins sown with a complex seed mixture, has been fi tted and must somehow operate. There will including forbs, under organic management, provided continue to be a place for chemical controls in certain the greatest control of undesirable species, along with places, for certain crops, and under various conditions, increased invertebrate abundance and diversity (As- such as where there are labor constraints or where previ- teraki et al. 2004). ous management has depleted biological resources. In Organic farming practices were found to have sig- the latter instance, chemical control measures will do nifi cantly reduced adverse impact on hedge-bottom little or nothing to improve the long-term sustainability vegetation, compared to conventional practices, with and profi tability of farming operations. an overall higher number of species found. Hedgerow Systemic approaches to integrated pest (and disease) species diversity, however, was greater in the conven- management are gaining a more reliable knowledge tional system, perhaps because of higher extinction rates base as more researchers investigate the complexity and resulting from pesticide drift (Aude et al. 2003). robustness of ecosystem dynamics. This is probably the Creating 3-meter unsprayed buffer zones along fi eld area within modern agriculture where scientifi c thinking margins was shown to be an effective method for reduc- has shifted most dramatically over the past 20 years, ing pesticide contamination of fi eld-border ditches. This from high confi dence in chemical controls to a growing reduced risks to aquatic organisms, increased vegetation skepticism and a search for more biologically-based diversity and abundance, had signifi cant increases in alternatives. The search for these alternative methods is phytophageous insects, and increased foraging by an in- driven as much by the negative incentive of economic sectivorous bird (deSnoo 1999; deSnoo and vanderPoll and environmental costs as it is by protective benefi ts, 1999). A cost-benefi t analysis showed that this practice but the latter are increasingly demonstrated. was feasible for winter wheat and potatoes, but was too Fortunately, approaches can be combined to make costly for sugar beets. them mutually more effective. Schroth et al. (2000) Maintaining 2-meter uncultivated borders along streams have shown positive feedbacks between agroforestry is required in Denmark, to stabilize stream banks and systems and pest/disease management. The planting of reduce erosion. Botanical quality of these agricultural windbreaks and shade trees and the cultivation of crop border areas was found to be low when compared and tree mixtures can have a demonstrable impact on to natural-grassland stream-borders, however; many the incidence and control of pests and diseases. Thus, species were associated with eutrophic and productive in thinking about ecoagriculture it is important to think

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 87 about overall strategies, rather than just the adoption • Nutrient loss in form of non-point-source pollu- and application of a particular approach. tion. External effects of unbalanced systems include: 6.7 Integrated Nutrient Management • Nutrient and sediment pollution of downstream Strategies built around integrated nutrient management ecosystems; are increasingly used to overcome productivity limitations exhibited by degraded tropical soils. They can also • Confl icts with neighbors; and used to decrease downstream effects of over-fertilized • Depletion of water table and stream fl ow. agricultural systems by increasing internal cycling of nutrients and minimizing nutrient import. In both cases, These are impacts that can be found anywhere nutrient biological productivity forms the basis for improving imbalance occurs in soil systems. The solutions to these yields of crops and animal products, improving farm problems will, of course, be location- and site-specifi c, profi tability, reducing nutrient losses, and increasing responding to the infl uences of soil types, topography, system sustainability. climate, cost structures, etc. We consider experience in dealing with nutrient imbalances and fi nding integrated Integrated management strategies are showing great solutions in contrasting agroecosystems. potential to increase system yield through increased reliance on biological processes, discussed also in Section 6.7.1 Improving Nutrient Use Efficiency in New 6.9 below. These strategies depend on the identifi cation York State of ecological and biophysical processes that support yield and ecosystem functions. Once weak linkages Agriculture is increasingly called upon to provide have been identifi ed, management practices can be environmental services (NRC 2003). This especially adapted to compensate for these constraints or gaps in true where the agricultural sector is able to produce nutrient cycling, using a variety of site-specifi c tools surpluses, and the societal benefi ts from food and fi ber (Albrecht and van Es interviews). Management of the production are diminishing relative to costs. Intensive agrobiological system can restore system balance either input of agricultural nutrients can increase crop and by increasing inputs to compensate for weaknesses in livestock productivity, but over-fertilization can lead nutrient cycling, or by decreasing the exports from to saturation of soil-nutrient binding capacity, and cycles, often having the effect of reducing pollution. increased risk of nutrient pollution. Reduction of negative downstream effects from agriculture (in terms of Through more precise management of nutrient inputs nutrient, chemical and biological pollution, soil erosion, and by ensuring biological resource compatibility, nu- and water consumption) is increasingly regarded as a trient cycling can be improved to become both more critical component of healthy landscapes and productive productive and more conserving. Then when nutri- ecological systems. ent balance is improved to satisfy crop and livestock requirements, nutrients are produced on-farm to the On dairy farms in New York State, imports of phospho- greatest extent possible, and nutrient limitations and/or rous and nitrogen are often 60 to 70 percent greater than build-up of excess nutrients are minimized. the amount of nutrients exported in milk and animal sales (Klausner et al. 1998). This leads to excessive Internal effects of unbalanced systems include: levels of soil nutrients and increased risk of nutrient • Nutrient depletion; pollution. The recent application of the Clean Water Act to agriculture has further motivated farmers, regulators, • Soil degradation following compaction and loss of researchers, extension educators, and farm advisors to organic matter; develop approaches for more effi cient nutrient use. • Unwanted nutrient accumulation; The Agricultural Environmental Management (AEM) • Saturation of the soil’s nutrient-holding capacity; program in New York State, supported by Cornell, has and developed nutrient management planning processes and

88 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations certifi cation programs to help farmers improve environ- production was increased by 13 percent, as farm profi t- mental performance and comply with laws, such as the ability went up and nutrient imports were substantially Concentrated Animal Feeding Operation (CAFO) regu- reduced following the implementation of integrated lation (New York State Soil and Water Conservation nutrient management planning (Klausner et al. 1998). Committee 2004; USEPA 2002). Effective development This is an example of how reduced inputs can produce and implementation of farm-specifi c integrated nutrient more output. Producers value cuNMPS as an effective management plans relies upon careful characterization tool for managing their farms profi tably, sustainably, of the farm system. Computer software packages are and with environmental responsibility (Cerosaletti et employed by farmers and their advisors to characterize al. 2003; McMahon and McMahon 2002; Tylutki et complex production and environmental systems and al. 2003). guide implementation CUNMPS does not explicitly consider wild biodiver-biodiver- An integrated nutrient management software package, sity, and crop diversity is generally not increased on the Cornell University Nutrient Management Planning participating farms (Albrecht interview). However, System (cuNMPS ) is now being used in New York to because robust biological cycles are intentionally and increase farm productivity while decreasing nutrient holistically managed to increase productivity and imports and exports (Fox et al. 2002). This results in increase ecosystem services, this integrated nutrient improved nutrient cycling and decreased incidence of management program provides a good example of nutrient pollution for groundwater and surface sources successful ecoagricultural management. The benefi ts (NMPWT 2004). The cuNMPS is a software suite and for the farm are clear, documented by McMahon and knowledge base for precision nutrient management McMahon (2002), and the positive externalities associ- planning across livestock and fi eld crop systems on ated with increased nutrient conservation can have a integrated farms.4 signifi cant benefi cial impact on downstream habitats (streams, lakes, reservoirs, and estuaries), promoting Coordinated precision management of each step of the wild biodiversity, healthy aquatic ecosystems, fi sheries dairy-farm nutrient cycle (feed inputs, in-cow process- productivity, clean water supply, and numerous other es, manure management, crop production) strengthens environmental benefi ts. the effi ciency of the biological cycle and reduces the need for imported feed and fertilizer. Productivity and • Solutions: Minimize external inputs; maximize in- profi tability are increased by maximizing the on-farm ternal cycling; connect manure application to crop production of high-quality forage and minimizing need production; connect crop production to livestock for purchased inputs. On one monitored farm, milk feeding; maximize profi table use of excess nutrients by composting and selling manure or by otherwise 4 The cuNMPS is currently comprised of Cornell Cropware for capturing nutrients for use within productive sys- crop nutrient management and the Cornell Net Carbohydrate tems. and Protein System (CNCPS) for herd nutrient management. Cropware is a product of the Nutrient Management Spear • Tools: Integrated nutrient management; Whole Farm Program (NMSP 2004; http://nmsp.css.cornell.edu) in the Planning; identifi cation and dissemination of on- Cornell Department of Crop and Soil Sciences and CNCPS is farm best management practices; using nitrogen and developed and delivered by the Cornell Department of Animal phosphorous indices to refl ect soil chemical status Science (CNCPS 2004; http://www.cncps.cornell.edu). The and hydrological controls on nutrient loss, and to developers of cuNMPS collaborate with the broader nutrient avoid over-fertilization without reducing crop yields. management community through the Cornell CALS Nutrient Management for Dairy and Livestock Farms Program Work Complex, integrated nutrient management systems Team (NMPWT 2004; http://www.inmpwt.cce.cornell.edu). benefi t from the use of computerized modeling tools The NMPWT is comprised of farmers, researchers, extension such as cuNMPS. educators, government agents, and private-sector agricultural consultants, and it serves as a forum for communication and development of research and extension proposals to address agricultural environmental issues.

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 89 6.7.2 Dealing with Productivity Limitations in Sub- can be offset by manuring, mulching, and minimum- Saharan Africa tillage, but crop yields may not respond unless any phosphorus defi ciencies are also addressed (Roose and Depleted soil fertility is the fundamental biophysical Barthes 2001). constraint on many soils in sub-Saharan Africa, with profound ramifi cations agricultural productivity and In eastern and southern Africa, an integrated program of human well-being. Traditional practices such as long- nutrient input management has substantially increased term fallows have broken down in the face of population soil fertility and maize yields (Sanchez et al. 2001; pressure and poor infrastructure (Sanchez et al. 2001). Jama et al. 1998a, 1998b; Jama et al. 1997). It combines Annual nutrient depletion rates are high, 20 kilograms rock phosphate applications, Tithonia biomass transfer, nitrogen per hectare, 3 kilograms phosphorous per leguminous cover cropping, and improved fallows. Si- hectare, and 15 kilograms potassium per hectare multaneous implementation of erosion control methods (AEE 2000). Chemical fertilizer is often prohibitively such as use of grass strips is often necessary to conserve expensive due to poor infrastructure, and fertilizer ef- the invested resource capital. Only after soil fertility fectiveness is diminished by soil degradation following is restored can additional best-management practices loss of organic matter. Low-input nutrient management such as improved crop genetics, IPM and agrodiversi- systems cannot overcome severe soil nutrient defi cien- fi cation be employed in an effective manner Sanchez cies unless external nutrient inputs are fi rst provided. et al. 2001). In Togo, integrated use of leguminous cover crops and phosphorus inputs resulted in increased In degraded systems, biophysical factors such as com- nitrogen use effi ciency and maize yield, according to paction, depletion, nutrient sequestration, and salt ac- Fofana et al. (2004), while in Zimbabwe, researchers cumulation can have a controlling effect on soil quality, concluded that cover crops alone are not suffi cient to such that yields dwindle even if conventional (Green support high yields unless some fertilizer is also used Revolution) interventions are employed, i.e., farmers (Chikowo et al. 2004). just use improved genetic materials and increase fertilization (Wolfe, pers. comm.). There are numerous cases of improving crop yield with simple enrichment of soil organic matter. Bio- Integrated management of inorganic and organic nu- mass transfer of Glyricidia has increased income and trient inputs, in combination with erosion control and maintained productivity of onions, cabbage and maize crop diversifi cation/intensifi cation, shows great promise when used in place of inorganic fertilizer in Zambia for restoring the productivity of degraded African soils (Kuntashula et al. 2004). Improved fallow systems (Sanchez et al. 2001; AEE 2000). Nutrient “re-capital- using Glyricidia also increased maize yields in Mali ization” can jump-start the nutrient cycle and promote (Kaya and Nair 2001). Short-duration Sesbania ffallowsallows increased yields from a more robust system (Sanchez similarly increased maize yields in Malawi; when a Ses- et al. 2001). Without integrated management of nutrient bania fallow replaced a Cajanus catch crop, however, stocks and cycles, benefi cial techniques such as cover there was a resulting loss in pigeonpea harvest (Ikerra cropping for biological nitrogen fi xation have only et al. 2001). Integrated nutrient management has also limited impact (Sanginga 2003), but with integrated shown promise for increasing upland rice production management, system productivity can be restored. in Tanzania (Meertens 2003). Loss of organic matter can create severe limitations on Various approaches to integrated nutrient manage- nutrient availability in low-fertility tropical soils, and ment in sub-Saharan Africa have had positive effects, effective soil and nutrient management systems focus while differing considerably in their effectiveness and on increasing soil organic matter contents through the resource requirements (Place et al. 2003). Systems of application of manures, composts and biomass (Ganry integrated nutrient management show great promise 2001). Incorporating crop residues can provide a short- for improving crop yields and smallholder income. term organic matter pool that greatly increases fertilizer It is recommended that the focus of research remain effi ciency (Diop 2002; Yamoah et al. 2002). Loss of on improving the productivity of smallholder farming organic carbon to mineralization, erosion, and leaching

90 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations systems, rather than promoting centralized management 6.8 Soil Health of large acreages (Sanchez et al. 2001), if intensifi cation is expected to balance the effects of extensifi cation. An A “healthy” soil has more capacity to support plant and example relating the adoption of large-scale soybean other biotic growth, to sustain ecosystem processes, to farming in Brazil to worker displacement and increased sequester and decompose pollutants, and to resist dis- deforestation in neighboring regions is presented in ease (Doran and Zeiss 2000). Measures of soil health McNeely and Scherr (2002). are diverse, depending upon the setting, soils, and management goals. Management strategies are similarly Increased soil fertility can help resist the spread of diverse, including cover crops, improved rotations, invasive weeds such as Striga and Imperata which are reduced tillage, and organic practices. The term “soil promoted by conditions unfavorable for other plants health” is used here interchangeably with “soil quality,” (Abunyewa and Padi 2003; Gacheru and Rao 2001; although the latter has often been used by researchers Sauerborn et al., 2003), particularly if crop rotation is that did not consider soil biological processes. also practiced (Oswald and Ransom 2001). CIRAD researchers have found that the presence and spread of The concept of “soil health” requires integrated consid- Striga are indicative of low soil organic matter, and that eration of physical, chemical, and biological processes this crop-strangling weed can be controlled by raising contributing to crop productivity and ecosystem func- SOM (Olivier Husson, CIRAD, pers. comm.). tion. The assessment of soil quality, and its change over time, is a primary indicator of sustainable land Simple and effective methods for evaluating ecological management (Doran 2002). Knowledge of temporal health and farming systems productivity are necessary changes in soil properties is necessary to the evaluation to identify obstacles to farm sustainability in sub-Sa- of management practices (Arshad and Martin 2002). haran Africa (Lightfoot and Noble 2001). Crews and Critical values for soil quality indicators (organic mat- Peoples (2004), when comparing legume and fertil- ter content, topsoil depth, water infi ltration rate, soil izer sources of nitrogen, conclude that legume-based aggregate stability, macro- and meso-porosity, nutri- systems have the potential to be more environmentally ent and pollutant concentration, microbial respiration, sustainable, and that many countries have the capac- nematode diversity, crop yield) all need to be identi- ity to reduce or eliminate agricultural dependence on fi ed on a site- or system-specifi c basis (Ekschmitt et synthetic nitrogen fertilizers. (This is seen also in the al. 2003). discussion of rice production in Section 6.12.) Chemical and textural/structural analyses of soil are The rice-wheat rotational cropping systems in India widely available and form the backbone of conven- are exhibiting a decline in productivity, low nitrogen- tional agricultural methods. Soil systems, however, use effi ciency, and declining soil health (Yadav et al. are composed of interactive chemical, physical, and 2003). A fi eld experiment demonstrated that inclusion biological components (Karlen et al. 2003; van Es, of a biological nitrogen-fi xing component (cowpea) in pers. comm.). Measures of soil quality in physical and the crop rotation increased wheat yield by 20 percent biological terms are currently the focus of considerable and also soil organic matter and nitrogen-use effi ciency. research, following on advances in the understanding These kinds of documented improvements from capital- of underground processes.5 izing on natural ecosystem processes have prompted increasing interest among scientists and practitioners Simple, cheap and effective measures of soil health in pushing and possibly expanding the limits of what are very much in demand, to provide tools for growers can be achieved agriculturally without relying on synthetic inputs. These explorations relate closely to the 5 At Cornell, a Soil Health Program Work Team is currently following subject. developing methodologies for measuring physical and biological soil properties that will be offered to New York State growers through extension and outreach channels (SHPWT 2004; Wolfe et al. 2003; Wolfe et al. 20043) through extension and outreach.

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 91 enabling them to identify factors limiting productivity biologically active soil organic matter (Haynes and and to monitor changes in soil health following the Beare 1997; Scott 1998), indicating a potential for agro- adoption of improved practices (van Es, interview). diversity practices to affect soil quality and microbial Important measures that have been identifi ed so far processes. Plants have a direct effect on soil structure include: standard chemical and textural analyses (phos- through physical and chemical processes as well as from phorous, potassium, micronutrients, pH, organic matter root exudation and rhizodeposition (Angers and Caron content, cation exchange capacity; soil type/texture); 1998). It is, of course, true that long-term and acceler- physical analysis (bulk density; macropore and meso- ated nutrient extraction can lead to soil degradation pore porosity; permeability at standardized moisture and depletion of nutrient capital. However, that such content; wet and dry aggregate stability); and biological effects are not seen under natural systems of vegetation analysis (microbial abundance, biomass and respira- raises interesting questions. Long-term experiments tion; nematode diversity). Rapid tests for biologically will be needed to evaluate cropping systems’ effects on active organic matter (potassium permanganate) and soil health, crop yield, soil fertility, and sustainability soil chemical properties (hyperspectral analysis) are (Haefele et al. 2003). also under evaluation (Shindlebeck, pers. comm.; van How far agricultural production can be intensifi ed Es, interview). through the exploitation of above-ground and below- Soil health and landscape diversity are components ground ecological interactions to sustain soil fertility of the Healthy Landscapes Initiative that is currently has not been systematically evaluated. As researchers addressing the human, ecosystem and socio-economic delve more deeply into the parameters and dynamics health of regions in fi ve Eastern European countries of soil health, the recuperative capacities of natural (van Es and Huska 2001). While soil quality is acknowl- systems become more evident and impressive, as does edged to be a critical parameter for maintaining agri- the interrelation of agricultural productivity and natural cultural productivity, and related research has increased ecosystem processes (Wander and Drinkwater 2000). exponentially in the last decade, there is no agreement While previously it has been often considered that cul- yet on the effectiveness of soil health indices as a tool tivated systems are strict alternatives to (and replace) for agricultural management, because they are multi- natural ones, there can be considerable compatibility dimensional and have limited fi eld validation (Delgado between productive agricultural systems and healthy and Cox 2003; Karlen et al. 2003; Letey et al. 2003). and robust ecosystems, with the maintenance of healthy soils being a link and common denominator. Land quality indicators have been developed for African and Dutch systems by Bouma (2002), who relied upon 6.9 Contributions of Below-Ground crop simulation to develop a ratio of observed-to-poten- Biodiversity to Sustainable Production tial crop yield that was used to evaluate sustainability of production. Sanchez et al. (2003) have developed There is increasing awareness of the integral role of soil classifi cation systems for fertility and for monitoring biology in biochemical and nutrient fl ow processes on tropical soil health and capability that appear to be ef- a landscape scale (SWCS 2000). Soil biology is thus fective. An integrated soil quality index developed in emerging as an increasingly active focus of agricul- China has been shown to vary signifi cantly among land tural research, due to the contributions that microbial use in Sichuan Province, ranging from 0.028 (cultivated processes can make to soil health, nutrient cycling, land) to 0.80 (natural forestland). The low value for and disease resistance. A diverse and robust microbial cultivated soils seemed to predominantly refl ect poor community can contribute to ecological equilibrium, values measured for soil bulk density, a parameter that for example, such that the spread of diseases caused can refl ect many years of tillage or low soil organic by fungal pathogens is resisted through direct competi- matter (Fu et al. 2003). tion (Hobbs interview). Microbial processes can play Experimental results have demonstrated that crop spe- a ubiquitous role in regulating the conservation and cies have differential effects on aggregate stability and availability of plant nutrients (Scott 1998).

92 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Below-ground processes are highly interactive, and they long-term nutrient release, and nutrient conservation are in turn affected by, and affect, the complex ecologi- (Eghball et al. 2004). This is because the nutrients are cal interactions that occur above-ground, as well docu- not supplied directly to the plants, but rather are utilized mented and described by Wardle (2002). Biodiversity by soil fl ora and fauna, from micro to macro, which in soils is normally extremely high (Wardle and Giller in turn increase nutrient availability for plants. This is 1996), but it can be compromised and reduced by till- done through mineralization, solubilization, physical age, continuous saturation (hypoxia), or certain chemi- transport (in the case of mycorrhizal fungi and sidero- cal concentrations, occurring naturally or man-made. phores), and other processes. Here we focus on microbial fl ora and fauna, though one Microbial biomass plays an important role in regu- should have an appreciation for all of the creature act- lating the turnover of fi ne particulate organic matter ing in intra- and inter-species communities, presented (Fliessbach and Mäder 2000). Labile pools of soil vividly and informatively by Wolfe (2001). organic matter are distinctly affected by long-term soil The Soil Quality Institute of the Natural Resources Con- management (particularly tillage) and some have been servation Service has published a Soil Biology Primer shown to support a larger microbial biomass and more that aims to have cross-national validity (SWCS 2002). active decompositional processes. While inorganic soil Soil food webs are extremely intricate and ramifi ed. nutrient amendments can enhance microbial biomass They include plants and especially their root systems, by providing more nutrients to soil communities, they which exude into the rhizosphere around the roots can also have negative effects such as suppressing the a share of the carbohydrates, amino acids and other genetic expression of nitrogenase and phosphatase, compounds synthesized within the plant; organic matter enzymes that fi x nitrogen and solubilize phosphorous from deceased organisms of all kinds;6 bacteria, fungi, (Tan et al. 2003; Turner and Haygarth 2001). This di- protozoa, nematodes, arthropods, earthworms, and minishes the “natural” supply of these nutrients for the many other species of macrofauna (SWCS 2000). The soil and plants. On the other hand, nutrients provided in main soil biological parameters currently measurable organic form are less likely to have effects and thus can are: (1) microbial biomass, (2) microfl ora composi- have a nutritional multiplier effect beyond their initial tion, (3) mineralization processes, and (4) synthesiz- amount, by supporting larger microbial biomass and ing processes (Filip 2002). Critical limits within these more active decompositional processes. parameters vary with soil type and also with climatic, Spatial variability in soils has been shown to be ex- topographical, and moisture-related parameters. tremely high for physical properties (texture, bulk 6.9.1 Nutrient Cycling in Soil Systems as an density), chemical properties (moisture, pH, carbon, Ecosystem Service nitrogen) and biological attributes (microbial biomass; microbial population size; and potential respiration, Organic farming systems can often be said to “feed the nitrification and nitrogen mineralization), even in soil, not the crop,” because they rely upon biological fi elds that have been continuously monocropped for uptake and decomposition to provide plant available many years (Robertson and Klingersmith 1997). This nutrients, rather than the application of readily avail- diversity presents challenges for the development of able fertilizers. Nutrient inputs in the form of manures, sampling methodologies that accurately and adequately composts and cover crops, although less in chemical refl ect the impact of experimental treatments. Well- fertilizer, have been shown to promote below-ground designed interdisciplinary studies are necessary to diversity, carbon accumulation, improved soil structure, characterize the linkages among below-ground communities that regulate important ecosystem processes 6 Conservation tillage resources on the internet include: Rolf (Wall and Moore 1999). Derpsch (http://www.rolf-derpsch.com); the Conservation Application of chemical fertilizers can provide a reason- Tillage Technology Center (http://www.ctic.purdue.edu/ able supplement to nutrient-defi cient soils to jump-start CTIC/CTIC.html); and the Rice-Wheat Consortium (RWC) (http://www.rwc-prism.cgiar.org/rwc). nutrient cycling (Sanchez et al. 2001), but manage-

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 93 ment of biological cycles is a necessary component ily retained in the soil, will greatly reduce the of sustainable nutrient management. Factors such as need for surplus nutrient additions. (Drinkwater maintaining soil moisture through mulching and low- 2004) ering soil temperature through cover crops can have a Reliance upon biological processes to provide plant large impact on nutrient conservation and management nutrients requires an increased understanding of be- processes below-ground (Sanginga et al. 2003). low-ground nutrient cycling, such as short-term carbon Interaction among organisms can provide a slow release dynamics following cover cropping (Puget and Drink- of plant-available nutrients to crops through decomposi- water 2001) and the infl uence of legume-based crop- tion of organic matter. In contrast, chemical fertilizers ping systems on conservation of carbon and nitrogen release soluble nutrients quickly, increasing the risk of (Drinkwater et al. 1998). nutrient loss to leaching (Drinkwater, interview). Also, 6.9.2 The Rhizosphere as an Ecosystem within there is much attention given in the soil literature to Ecosystems what is called “immobilization” (uptake) of nutrients by soil organisms, but less to the return provision of The last decade has seen signifi cant advances in the these to plants and other organisms when the consum- understanding of rhizosphere processes. Plants have ers become deceased. This can happen very quickly been shown to support a diversity of microbial organ- through predation or senescence, in which case these isms in the rhizosphere, primarily by releasing carbon nutrients become available. “Immobilization” sounds from their roots (Cheng et al. 2003; Hamilton and Frank like a negative process, but in fact, such tying up of 2001). The diverse rhizosphere community scavenges nutrients in soil spaces prevents or at least reduces their nutrients effectively and releases plant-available nu- leaching or any other loss to the system. trients upon senescence. The below-ground biota can An integrated nutrient management strategy based upon be adversely affected by tillage, chemical fertilizers, ecological concepts can make signifi cant improve- pesticides, crop diseases, poor crop health, and poor soil ments in nitrogen-use effi ciency by timing microbial tilth can, on the other hand, be enhanced through crop decomposition to release nutrients during periods of diversifi cation, cover cropping, organic matter inputs, crop growth and by sequestering nutrients that are not and other management practices. mined by the crop. Measures that refl ect the health and activity of below- Rather than focusing on soluble, inorganic ground communities are not unambiguous or always plant-available nutrient pools, an ecosystem- correlated. Microbial respiration, while easily mea- based approach seeks to optimize organic and sured, does not always correlate to the robustness of mineral reservoirs with longer residence times ecosystem functioning (Cadet et al. 2004), and studies that can be accessed through microbially- of the effects of herbicides on microbial populations and plant-mediated processes. This requires deliberate use of varied nutrient sources and should examine, for example, community dynamics strategic increases in plant diversity to restore and soil health in addition to respiration. Sensitive in- desired ecosystem functions such as nutrient dicators of soil biological health may include the rate and soil retention, internal cycling capacity, of nitrogen turnover (Schloter et al. 2003), enzymatic or aggregation. Breeding for cultivars and as- activity (Schloter et al. 2003), nematode community sociated microorganisms that do not require composition (Cadet et al. 2004), the ratio of benefi - surplus nutrient additions is critical if plant- and microbially-mediated ecosystem processes cial-to-parasitic nematodes, decomposition rate (Wolfe such as mineralization-immobilization, biologi- interview), abundance of predatory nematodes (Ferris cal weathering, and carbon sequestration are et al. 2004), the ergesterol-to-biomass ratio (Dinesh et to be harnessed. Integrated management of al. 2003), and mycorrhizal abundance. biogeochemical processes that regulate the cycling of nutrients and carbon, combined with increased reservoirs that are more read-

94 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations 6.9.3 Some Effects of Soil System Dynamics isms.The prospects for successful ecoagriculture hinge very much on a better understanding and use of subter- Some results from fi eld-based studies of below-ground ranean biodiversity. Much research is needed, however, dynamics and their impacts on agricultural production as below-ground communities and processes remain and/or biodiversity include: inadequately understood and are challenging to study. • In South African sugar cane plantations, high productivity patches corresponded to changes in nematode 6.10 Management of the Hydrological communities and soil nutrient availability, but not Cycle to changes in microbial respiration (Cadet et al. Water is conventionally regarded as an input to ag- 2004). riculture, not as something endogenous to crop and • In Belgium, chemical fertilizer application was as- animal production systems. But water is produced by sociated with lower methanotroph richness when the interactions among plants, soil and microorganisms, compared to use of organic fertilizers (Seghers et affected by temperatures and other climatic factors al. 2004). This supports the results of other studies, themselves refl ecting land use and vegetative cover. which have shown that the methanotrophic com- Ground cover affects rates of evapotranspiration that munity is negatively affected by the application determine how much of water remains in the soil and of chemical fertilizers. Results of the study, which how much exits to the atmosphere. Soil that is depleted used genetic markers to identify bacterial and fungal of biological life lacks water-holding capacity. Thus, species, demonstrated that fertilizer applications water is more appropriately viewed not as an “input,” affected both the bulk-soil and endophytic soil mi- which refl ects an industrial manner of thinking, but as crobial communities, but that herbicide usage did a component (indeed, a sine qua non) of living sys- not have an observable effect. tems—ubiquitous but variable, from the cellular level up to the atmosphere. • A Swedish study (Dahlin et al. 1997) demonstrated that sewage sludge application, with associated low Water, which along with carbon is the most common levels of heavy metal contamination, resulted in constituent living organisms, is essential for both ag- moderate impairment of microbial function, using ricultural production and wildlife habitat through the a broad range of measures of microbial function. maintenance of sustainable and semi-natural water • A study in India (Dinesh et al. 2003) demonstrated supplies. There can be no agriculture of any sort, eco- signifi cant decreases in microbial activity following or otherwise, without water. The question is whether deforestation of a tropical forest area, likely due to better management of water resources can contribute reductions in organic matter. Measures used suc- to agricultural enterprises that are more productive, cessfully included ergesterol-to-biomass ratio and more hydrologically sustainable, and more supportive metabolic quotient. of diverse natural habitats. Improvements in agricultural resource management that increase the water-use • A California study demonstrated that maintenance of effi ciency of crops or livestock should be regarded as conditions for biological activity (fall irrigation and benefi cial for biodiversity, including in many instances carbon addition) resulted in increased abundance of wild biodiversity, because this will reduce the impact of benefi cial nematodes in the following year’s tomato agriculture on natural hydrological cycles. In dryland crop. This result demonstrated the potential for spe- areas where pumping rates are high and water tables are cifi c agrosystem management to support organisms declining, reduction of crop evapotranspiration is also a benefi cial for agricultural purposes (Ferris et al., desirable objective from the standpoint of biodiversity 2004). conservation. These and other studies have shown the potential for Problems resulting from unsustainable water use are managing plants, soil, water, and nutrients to promote most dramatically seen where extraction from aqui- the abundance and diversity of benefi cial soil organ- fers, rivers, and reservoirs reduces fl ows to streams,

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 95 groundwater, and wetlands to such an extent that natural groundwater depletion if and to the extent that it reduces habitats are degraded or lost, with resulting species evapotranspiration. This important consideration is loss of fi shes, plants, and other aquatic organisms. often neglected by large-scale irrigation management This confl ict is seen increasingly in the western United projects. Governments can spend large sums on making States, but also occurs in India, China and elsewhere management changes that do not result in any reduction with less public attention. In areas with only shallow in groundwater depletion. wells to support village water supplies, receding water The fundamental constraint to be reckoned with is the tables are also a matter of great concern. Improvements water demands of the particular crop sequence (Kendy in hydrological modeling are now allowing for more et al. 2004). Similarly, other techniques that improve complex management of water resources, such that use irrigation effi ciency by reducing percolation, e.g., lining of water resources can be more carefully managed (T. canals and implementing precision irrigation to the crop Steenhuis, pers. comm.). root zone, may have little effect on the overall hydro- 6.10.1 Effects of Changes in Water Supply and logical mass balance and should not be relied upon to Evapotranspiration stop groundwater recession (Kendy et al. 2002).

Most agricultural thinking looks to irrigation as a means 6.10.2 Interactions with Cropping Patterns of providing water for plant growth whenever water Ecological imbalances in water-defi cit areas can be is in short supply at certain times or throughout the redressed most directly by shifting to crop species that growing period. Improved irrigation management com- require less water. Water-use effi ciency can be improved monly focuses on reducing groundwater extraction rates by changing crop species, such as the replacement of (supply management) by pumping only as much water alfalfa with cereal-legume intercropping systems on the as can be used by the irrigated crops, and by reducing southern high plains of the U.S. (Lauriault and Kirksey evaporation through precision irrigation. Groundwater 2004). In the case of the North China Plains, water depletion, however, is the result of the mass balance budget modeling using 38 years of historic climate data between infl ows (recharge) and outfl ows (pumping). In identifi ed the replacement of irrigated winter wheat areas with shallow unrestricted aquifers, water pumped with a more drought-tolerant crop within the present in excess of crop requirements drains back into the wheat-corn rotational farming system as the most ef- groundwater supply so it is not really lost. fective method of stabilizing water extractions where If pumping is reduced without reducing total crop the water table is dropping as much as 1 m or more per transpiration, precision water application will, in many year (Kendy et al. 2004). It appears that if only cotton cases, reduce pumping rates without reducing the rate and millet had been grown, instead of corn and wheat, of water table recession, because the recharge of excess the water table could have remained stable. The cur- irrigated water declines. Precision water management rent water table decline in North China is putting all might, in some cases, actually increase the overall organisms within the region, plant, animal, human, and groundwater depletion rate. This would happen if a microbial, at risk. reduction in the water required per hectare encouraged This research points out the risk of growing crops with farmers to increase their area under irrigated crops (and a high water demand, without considering the effect thus increase evapotranspiration), assuming that there that irrigation will have on the regional hydrologic had been some actual water-saving. cycle. Certainly water could be extracted from ground Care must be taken to evaluate hydrological dynamics or surface water sources and applied through irrigation and water balances when assessing farming system infrastructure to meet the demands of a high-water-de- sustainability. Some water “savings” that appear real manding cropping system. But this can have adverse are not, when evaluated on a watershed scale. Currently, effects on the hydrological cycle An ecoagricultural ap- many irrigation design efforts focus on maximizing proach consider developing highly productive cropping water-use efficiency. This strategy will only slow

96 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations systems that more closely mimic the evapotranspiration in some ways a more fi nite resource than land. It has the rates of the native vegetation. advantage of possible re-use, if quality is not degraded by use, but as population rises, its availability per capita A review of African cropping systems that maximize will continue to decline (Elphick and Oring 1998). water-use effi ciency has been provided by Van Du- ivenbooden et al. (2000). Windbreaks if planted and Efforts to make production systems more water-effi - maintained can reduce crop evapotranspiration sig- cient should include measures that promote sustainable nifi cantly. In severely water-limited areas, a desirable management of the water balance; maintenance of soil practice such as cover-cropping for carbon and nitrogen organic matter and increases in infi ltration and reten- production could be ecologically unsound if it increases tion of water into and near the root zone; plus enhanced overall evapotranspiration. Whenever water is a limit- soil life. The latter requires water, but maintains and ing factor, it is wiser to adapt the cropping pattern to recycles soil moisture as an intrinsic quality. This is the conditions than to assume that increasing supply evident when one considers the desiccation of soil that will solve the problem. There are millions of hectares has been lost all of its living organisms. A synergy is of soil on six continents no longer productive because seen in the interrelationship of soil health with water of salinization due to irrigation practices that did not retention and use-effi ciency. adequately account for and respect hydrological cycling Ecoagriculture will need to fi nd ways to economize on and soil-water relations. water use if it is going to meet food production needs Water budgeting is an important tool for regions that without compromising the biodiversity of agricultural are water-limited. Analysis of groundwater balances in and wild areas. Fortunately, both of the next two focuses Indonesia has identifi ed management practices that im- of discussion have the potential to increase yield while proved water supply and downstream water use during reducing water use, whether in rain-fed or irrigated dry months, while also permitting increased irrigation agricultural systems, thereby fulfi lling the objective of during the wet months (Peranginangin et al. in press). “more crop per drop” that is becoming an aspiration more and more widely shared. By maintaining hydrological adequacy and diversity, conservation of on-farm wetland areas can strengthen 6.11. Conservation Tillage/Conservation ecosystem functions such as fi ltering of agricultural Agriculture pollutants and preserving of wild biodiversity (Bedford interview). Hydrological modifi cations to farmland, Conservation tillage systems, variously known as “low- such as artifi cial drainage, can result in nutrient pollu- till,” “zero-till” or “no-till,” avoid plowing in order to tion and degradation of wildlife habitat, so such changes maintain soil cover, improve soil structure, and control should not be made without considering external erosion. By including also the improved management of effects. The agricultural sector is a major consumer crop residues, water resources and agricultural inputs, of water throughout much of the world, and agricul- conservation tillage has evolved into a set of ideas and tural impacts on the surrounding landscape should be techniques now characterized by FAO as “conservation carefully considered, so that crop productivity can be agriculture.” It thus has many facets that make it quite sustained, and so that wetland and aquatic habitats, so consistent with ecoagricultural objectives. It should important to wild biodiversity and healthy landscapes, not be regarded, however, as a total system. Rather it are maintained. should be considered and practiced as one component As water becomes an increasingly scarce resource in of a larger program of healthy landscape management this world, the demands of the agricultural system need that can address issues of agricultural sustainability and to be modifi ed and reduced through a combination of avoidance of pollution (P. Hobbs, interview). increasing the water-use effi ciency of current cropping The basic features of conservation agriculture include systems and changing the cropping systems themselves the following (Hobbs and Gupta 2004): to promote more sustainable water usage. Fresh water is

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 97 • Little or no soil disturbance, thereby promoting 6.10.1 Herbicide Use and Direct Seeding growth of native soil microorganisms and fauna, and As currently practiced, no-till systems have the tradeoff soil biological processes; of an increased dependency on the use of herbicides. • Ground cover from using green-manure cover crops South Asian farmers, presently wedded to the rice- and/or previous-crop residues will reduce erosion wheat rotational farming system, have not developed dramatically, and lowers soil temperature with ben- more varied rotational systems that enable them to efi cial effects for soil organisms; control weeds through a well-planned succession of • Crop rotation, helps control pests, diseases, and crops as is being done in Latin America, nor have they weeds; developed implements for cutting plants in optimal ways so that chemical killing of the plants becomes un- • IPM practices integrated into farming system, saving necessary, also done in Latin America (Calegari 2002). costs and enhancing biological processes gener- There is an increased risk with some no-till systems of ally; soil nutrient loss through percolation due to better soil • No more burning of crop residues, which improves aggregation which increases macropores. Also, without air quality; plowing in the spring, soil temperatures will be lower, and this can delay germination in some cases. • Use of agrochemicals is more effi cient and often reduced, with a defi nite reduction in the use of fossil CIRAD in Brazil has developed and is promoting direct fuels; seeding into vegetative cover. This increases biological productivity and plant nutrition by trying to mimic natu- • Increased profi tability, due to lower costs of pro- ral forest conditions, having a robust method for weed duction, usually accompanied by yield increases, and erosion control plus improvements in soil structure although these may not come immediately; a few and nutrient use effi ciency (Seguy et al. 2003a, 2003b). years may be needed for soil quality to build up with No-till agriculture adoption in Brazil has been linked to the cessation of plowing; increased soil organic matter and increased soil fertil- • Profi tability can often be further enhanced by stop- ity, particularly at shallow (0-5 centimeters) soil depths ping tillage, permitting farmers to plant their crop (Machado and Silva 2001). Rainfed rice yields in the earlier, closer to the optimum time; and range of 8-9 tons per hectare, well above usual irrigated • Water-use effi ciency is improved with ground cover yields, are being achieved in Brazil within 4-5 years of and no plowing. using these no-till methods, and production of maize and soybeans has been increased by about 50 percent In South Asia, zero tillage for wheat following the har- with a similar reduction in costs of production. vest of rice in rice-wheat cropping systems, promoted by CIMMYT, has been very successful and is being Conservation agriculture is spreading rapidly, covering adopted more and more widely (Hobbs and Gupta more than 70 million hectares by 2002. What has been 2004). From practically no area under no-till in 1997, its demonstrated in the last three decades is that a system use in South Asia has expanded to over a million hect- of agriculture that “imitates” nature can in fact be ares without any signifi cant investment in agricultural productive and profi table. Direct seeding mulch-based extension. The advantages that the new set of cultural cropping systems have been increasing in popularity practices bring to farmers has been enough to persuade in Latin America, accounting already for 50 percent them to take up conservation agriculture, aided to be of cropped acres in Brazil, Paraguay, and Argentina sure by the recent design and availability of appropriate (Scopel et al. undated). implements for planting through ground cover. This is not surprising given that changing management practices has had such a positive impact on profi tability. The reduced consumption (combustion) of fossil fuels adds to air quality, as noted above, while reduced ero-

98 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations sion and application of chemical fertilizers enhances 6.10.3 Other Issues Relating to Conservation water quality. Both create an environment more suitable Agriculture for sustaining biodiversity. Unlike many conservation Certain reduced-tillage methods such as raised beds and agriculture technologies that are labor intensive, low ridged till provide alternative methods of cultivation and zero tillage systems often demand less labor. The that can be effective at conserving water and reducing use of herbicides remains a controversial aspect of tillage impacts (Sayre and Hobbs 2003). The Land In- conservation agriculture where these are employed, stitute in Salina, Kansas, is working to breed perennial however. crops for grain production, seeding them in diverse The control of weeds has been a principal reason for groupings that mimic vegetation complexes found in plowing through the years, and it is the main constraint natural prairies. This work focuses on hybridizing an- when moving to no-till agriculture. In many regions, nual crops with perennial relatives, and on selecting for adoption of reduced tillage agricultural systems has high grain production in existing perennials. Cox et al. been made possible by the development of herbicide (2002) advocate the development of a large and com- resistant-crop varieties. Although the environmental prehensive breeding program for these purposes, based impact of herbicides has been diminished in recent on their initial experience. Perennial grain cropping years with different chemical formulations, attention systems are an example of a “natural systems” approach must be paid to possible secondary chemical effects on to agriculture, where intact natural ecosystems are used soil microbial communities, mycorrhizal symbioses, as models for the design of nature-based agricultural and soil health. systems that harness ecosystem functions to support 7 Recent advances show great potential for weed and pest production and pest control (Jackson 2002). control using non-chemical means, after the transition to Thirty years ago, practicing cropped agriculture a reduced tillage system has been made. These systems without plowing was practically unthinkable. No-till rely upon groundcover, crop rotation, cover cropping, agriculture was equated by some critics with atavistic, and reduced soil disturbance to control weeds. As with “dibble stick” cultivation. What has been demonstrated other kinds of ecoagriculture, we see here that biological in the last three decades is that a system of agriculture methods can be invoked to solve constraints in ways that “imitates” nature can in fact be both productive that are environmentally benign. and profi table. Conservation agriculture is getting comparable or often better yields. Even more often, 6.10.2 Effects on Nutrient Availability and it is producing higher net incomes for farmers, who Management see their costs of production decline, which makes it a Multi-function cover crops lead to increased nutrient- stable innovation. use effi ciency and economic and agronomic benefi ts In the systems developed by CIRAD, within a few to the farmer. Increased systems complexity, however, years, the yields are usually higher, which makes them leads to increased complexity of management. Par- preferable on both economic and agronomic grounds, ticipatory research and farmer involvement is critical and certainly they are more benign for biodiversity, to the development of effi cient and practical systems wild and otherwise, than the more chemical-input- that farmers can and will manage effectively. Research intensive systems that they are replacing. Most of the should focus on how better to understand and enhance conservation agriculture systems developed initially soil biological processes, reducing dependence on ex- were the temperate zones of the U.S. and the Southern ternal inputs, and integrating livestock effectively into Cone countries in Latin America. The CIRAD systems, the new farming systems that emerge under the rubric of conservation agriculture. 7 Conservation tillage resources on the internet include: Rolf Derpsch (http://www.rolf-derpsch.com); the Conservation Tillage Technology Center (http://www.ctic.purdue.edu/ CTIC/CTIC.html); and the Rice-Wheat Consortium (RWC) (http://www.rwc-prism.cgiar.org/rwc).

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 99 mentioned above, have been devised mostly for tropi- scientifi c literature and that should have some gener- cal areas, although now they are adapted also for dryer, alizability. Nothing defi nitive can be said yet about its Mediterranean-type climates like Tunisia and southern sustainability, but many farmers who have used SRI France. methods for up to 10 years report that their high yields have been maintained and even increased by continu- But there are drawbacks. While conservation tillage is ing to enhance soil organic matter rather than rely on amenable to mechanized agriculture and can economize inorganic nutrients for soil fertility. on labor inputs, many other conservation agriculture technologies and practices require additional labor in- As with all agricultural practices, there is considerable puts, which may be unavailable or unaffordable at the variation in the yields from resulting SRI practices prevailing agricultural wage, especially on a seasonal within and between countries. Agricultural differs from basis (Lee and Ruben, 2001). In addition, conservation industrial production in that the outputs associated with agriculture innovations, like the next one discussed, a given set and level of inputs can be quite variable have had diffi culty attracting research funding because because biological processes are involved. Also, the they are often relatively simple, not requiring high-tech methods are not similarly productive everywhere. There research of the sort that is most interesting to donor are some climates or soil conditions where the results agencies or to scientifi c journals that control the pro- are disappointing, as with any agricultural technology. fessional rewards of publication. The demonstrated While some of the yields reported with SRI are quite environmental benefi ts, with higher production at lower spectacular, it is the changes in average yield that are cost, should make them of much interest to the scientifi c more important, usually 50-100 percent with reduced and donor communities. inputs. In Chapter 3, the fi ndings of two evaluations of SRI 6.12 The System of Rice Intensifi cation were summarized, one undertaken by GTZ in Cambodia Ecoagriculture needs to demonstrate that there can and the other by IWMI in Sri Lanka. They were done be production gains, not just tradeoffs, resulting from with randomly selected samples of farmers who are us- agricultural methods that are more supportive of bio- ing SRI (at least partially) or who are not. The results of diversity, as suggested in Chapter 2. An example of extensive data gathering and analysis confi rmed what such a system with evident positive-sum effects is the has been reported from less systematic and detailed System of Rice Intensifi cation (SRI). Although still evaluations for a number of years. So there should no controversial in some scientifi c circles (Dobermann longer be much question whether SRI methods work. 2004; Sheehy et al. 2004; Surridge 2004), the use of SRI The interesting questions are “how?” and “why?” as is rapidly spreading. Five years ago, SRI was known answers may give leads for making agriculture in the only in Madagascar, where it had been developed in the 21st century more productive, profi table, and environ- early 1980s by Fr. Henri de Laulanié (Laulanié 1993). mentally benign. Today its potential to raise yields and factor productivity while reducing inputs (including water) and costs 6.12.1 Getting More Output from Fewer External of production with benefi cial environmental effects has Inputs been seen in at least 19 countries in Asia, Africa and SRI goes against the usual logic that more or “better” Latin America (Stoop et al. 2002; Uphoff et al. 2002; inputs are required to get more output. SRI gains come Uphoff 2003, 2004). with less or no use of chemical fertilizers and agrochem- Because scientifi c work on SRI began only in the last ical protection, with 30-50 percent less use of water, fi ve years, it is too early to know how broadly its prin- with fewer plants/m2 and by starting with younger and ciples and practices can be adapted to other crops with smaller plants when transplanting. No new varieties are similar success. The more that is learned about SRI, required as the methods have evoked more productive however, the more evident are a number of principles phenotypes from any and all rice genotypes used thus responsible for SRI results that are supported by current far. The “trick” is giving plants a better environment to grow in.

100 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations First, wider spacing of plants generalizes “the edge ef- soil organic matter and providing nutrients in organic fect” for the whole fi eld.8 Second, by alternately fl ood- forms have already been discussed (6.8. and 6.9.) While ing and drying rice fi elds for periods of 3-6 days or just SRI is not organic in principle, it often becomes organic applying small amounts of water daily to keep the soil for pragmatic reasons. moist but not saturated—not keeping rice paddies con- The SRI methodology is quite different from the para- tinuously fl ooded, as has been the practice for most rice digm of the Green Revolution, which raised grain yields cultivation for centuries—the plants have an aerobic (a) by improving the genetic potential of crop plants, (or an alternating aerobic-anaerobic) soil environment, particularly their effi ciency in utilizing higher inputs of quite different from the hypoxic one that is created by fertilizer and water, and then (b) by increasing these in- conventional water management practices. puts. More chemical means of protection had to be used When rice roots are continuously hypoxic, they begin because as is well-known, high applications of nitrogen degenerating, so that by the time of fl owering, when fertilizer not only make plants more liable to lodging, grain production begins, as many as three-fourths of the because tillers and root systems have less strength, but roots can have degenerated (Kar et al. 1974). It is true also more vulnerable to insects and pathogens. that rice plants can survive under fl ooded conditions, SRI makes neither kind of change, utilizing either which has made it possible them to grow in environ- modern or traditional varieties, and reducing rather than ments where other crops could not. But they do not increasing external inputs. SRI is essentially a different thrive. It can be demonstrated (and has been shown by way of managing plants, soil, water and nutrients and is replicated factorial trials) that especially when using thus variant on a generic agroecological strategy, similar other growth-promoting practices, the maintenance of to that adopted by CIRAD in its direct-seeding-through- mostly aerobic conditions is more favorable for rice vegetative-cover methodology but for different condi- production. tions and different crops. These four kinds of inputs When rice roots are kept fl ooded and die back during are managed in ways that make plant roots larger and the growth cycle, the plants must necessarily rely upon more effective. These changes also, partly because of chemical fertilization because they do not have root the larger roots produce more exudates, induce defi nite systems that can access a greater volume of nutrients, changes in soil biological activity. and a greater variety of micronutrients. SRI was origi- An interesting phenotypical change that results with nally developed using chemical fertilizer, which was the SRI practices, which makes it of scientifi c interest, is norm in Madagascar. But when government subsidies that a long-standing constraint on rice productivity—an were removed and small farmers could no longer afford inverse correlation between panicle number and panicle to buy fertilizer, Fr. de Laulanié began using compost, size normally observed with fl ooded rice (Ying et al. and found that the results were even better, as well as 1997)—is broken. A positive correlation is generally cheaper for cash-strapped farmers. The advantages of observed instead, as SRI plants have both more panicles organic fertilization, shown in replicated trials, are not (grain-bearing tillers) and larger panicles (more grains unique to SRI, of course. The advantages of increasing per panicle). This is what makes possible the yield increases in multiples rather than increments. Only 8 When taking crop-cutting samples to measure (estimate) the latter are possible if the usual negative relationship the yield of rice or any other crop, people are always told to (diminishing returns) holds. When SRI plants have take them from the middle of the fi eld—“to avoid the edge larger root systems supporting bigger canopies, and effect.” One gets unrepresentative samples from the edges of vice versa, there is positive feedback with the plants a fi eld because it is known that plant growth is more robust there, where plants have more exposure to sun and light and becoming open rather than closed systems. less root competition. With SRI, by planting in a square pattern, 25x25 centimeters (or wider as soil quality improves 6.12.2. Biological Dynamics with repeated use of the system), and only one plant per hill instead of 3-6 in each clump, the growing environment for The results of SRI above-ground are easy to see -- rice plants is greatly changed. increased tillering where single plants have 50 to 100

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 101 tillers or even more, compared with 15-30 for clumps other services (Dobbelaere et al. 2003). The application of conventionally-grown, i.e., crowded and fl ooded, of inorganic nitrogen fertilizer can have a suppressive rice. Having fewer plants increases their exposure to effect on these organisms, as Tan et al. (2003) have the sun’s radiation and to circulating air. Measurements shown with very sophisticated analysis reported in made at the Indonesian rice research center at Sukaman- Chapter 3; the application of inorganic nitrogen to soil di have found that illumination at lower levels within affects the expression of nitrogen-fi xing genes in endo- the canopy of conventionally-grown (close-planted) rice phytic bacteria that live in rice roots. This has also been was not high enough to support photosynthesis, so that seen in research by colleagues in Madagascar. lower leaves were in effect being “subsidized” by the Research from the University of Antananarivo has photosynthesis of upper-level leaves; with wider SRI shown that SRI practices are associated with large spacing, all leaves received enough light for photosyn- increases both in endophytic populations of a well- thesis (Dr. Anischan Gani, pers. comm.). known nitrogen-fi xing bacteria Azospirillum in plant Larger, more active root systems contribute to superior roots and in crop yield (and tillering). Table 6.1 shows plant performance, although plant breeders preoccupied the results of replicated trials at the Beforona experi- with increasing plants’ harvest index (the percent of ment station. Similar increases in yield have been seen total biomass ending up in the edible portion of the on clay soil with SRI methods plus compost elsewhere plant) have considered roots “a waste.” This is closed- in Madagascar so the yield results was not surprising system thinking. Root growth in all plants is encouraged (Randriamiharisoa and Uphoff 2002). What was sig- by wider spacing and more aerated soil, especially if nifi cant to see were the 17- and 21-fold increases in more organic matter provided through compost. SRI endophytic populations when SRI methods were used root systems require much more force to uproot, as and, even better, when compost was added to the soil. much as 10 times more kilograms per plant, compared Students of microbiology know that such populations with conventionally-grown rice plants whose systems can easily vary by several orders of magnitude, and are smaller and necrotic. These differences can be seen that the application of inorganic fertilizer has negative when roots are inspected, and they can be measured effects for soil organisms (in these trials, reducing the precisely with some effort.9 What is not so easy to population of Azospirillum by 40 percent.).10 Interest- ascertain is the effects and benefi ts of having more ingly, the Azospirillum populations varied only within abundant and diverse populations of soil biota, most the roots, as the concentrations of this bacteria did not of which are invisible to the human eye. vary in the rhizosphere soil, remaining around 25x103 ml across the six sets of trials. SRI supports the advice given by organic farmers, “feed the soil, and let it feed the plant.” While chemical fertil- No conclusion should be inferred the increase in Azo- izer enhances yield with SRI methods, organic fertiliza- spirillum populations in rice roots was itself or alone tion raises yield even more, because the latter provides responsible for the measured increases in yield. This better substrate for microbial populations. In section particular bacteria was studied because it is a known 6.9, we considered how various in the rhizosphere, on nitrogen-fi xing agent and could be reliably measured and around the roots, contribute to plant nutrition and at laboratory facilities available within Madagascar. It health. We did not discuss much the contributions also may be an “indicator” species for what is going on in of endophytic microorganisms such as bacteria species the rhizosphere and in roots more generally, since it is that live inside root tissues and fi x nitrogen and provide known that effects on growth are more usually the result of microbial ecology rather than just the infl uence of 9 Barison (2002) found that root length density (cm/cm3) at a depth of 40-50 centimeters was on average 0.23 for SRI 10 The work involved in setting up replicated trials was such plants, and 0.06 for conventionally-grown plants. At maturity, that only the second and fourth treatments done on clay soil root-pulling resistance (kg/plant) was 55.19 kg per SRI clump were repeated on loamy, less fertile soil. The results of these (a single plant) and 20.67 kilograms per clump for conven- trials were also consistent with previous research fi ndings tionally-grown rice (4-6 plants) (Tables 13 and 14). on soil microbiology.

102 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations Table 6.1. Yields and Tiller Formation Associated with Populations of Azospirillum in Rice Plant Roots in Response to Different Plant, Soil, Water and Nutrient Management Conditions Azospirillum Tillers Paddy count in roots per Yield (103/mg) Plant (t/ha)

CLAY SOIL Traditional cultivation, no amendments 65 17 1.8

SRI cultivation with no amendments 1,100 45 6.1

SRI cultivation with NPK amendments 450 68 9.0

SRI cultivation with compost amendments 1,400 78 10.5

LOAM SOIL

SRI cultivation with no amendments 75 32 2.1

SRI cultivation with compost amendments 2,000 47 6.6

Data from Raobelison (2000), reported by Randriamiharisoa (2002). an individual species. meso and macro. These live in, on and around the plants’ roots providing a variety of benefi ts and services in Research on the various SRI processes and the syner- terms of plant nutrition and protection, including: gies among them is still in its early stages, with more questions than answers. What can be said with some • Nitrogen fi xation by both endophytic and free-as- confi dence is that it may be possible to produce “more” sociated diazotrophs Döbereiner 1987; Boddy et al. from “less,” that is, to get more outputs with reduced 1995); external inputs. For this to occur, there must be greater • Phosphorous solubilization by many different bacte- contributions from soil systems, and the plants’ own ria and fungi (Turner and Haygarth 2001; Gyanesh- genetic operations need to convert inputs into outputs war et al. 2002); more effi ciently. We have some evidence for the latter from a study in Madagascar where an analysis of SRI • Greater access to nutrients and water from a larger and non-SRI plants, using QUEFTS modeling tech- volume of soil due to the services of mycorrhizal niques, found that paddy yield in response to compa- fungi (Martin et al. 2001); rable levels of nitrogen, phosphorous, and potassium • Making more nitrogen available to plant roots plateaued around 10 tons per hectare with SRI-grown protozoa when they “graze” bacteria that live on plants, whereas it leveled off about 5 tons per hectare the exudation from plant roots, because they have a for plants grown by conventional methods—by the lower carbon:nitrogen requirement than the bacteria same farmers, and on the same farms (N=108) (Barison they consume, excrete “excess” nitrogen; nematodes 2002). in turn, consume protozoa within the underground SRI does not require the addition of compost to the soil soil food web and contribute nitrogen in the same to get higher yields. However, yields are enhanced and way (Bonkowski 2004); made more sustainable by increasing organic matter in • Producing phytohormones auxins, cytokinins and the soil to support a vast array of soil organisms, micro, other biochemical compounds that promote root

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 103 growth and benefi t plants in many other ways (Fran- for nothing” is diffi cult to accept. SRI does initially kenberger and Arshad 1995; Kapulnik and Okon require more labor and certain much more knowledge 2001); and skill, so it is not exactly a “free lunch,” but the returns to land, labor, capital and water are all increased • Protecting plants from the effects of pathogens concurrently with SRI, something hardly seen before. through competition, production of antibiotic protec- It has been objected that there are no scientifi c studies tion, and induced systemic resistance (ISR) (Doeb- testing and validating the results reported. There are a belaere et al. 2003). number of detailed theses and studies in French and All of these processes are supported the exudation from Chinese, but only few in English. It has been diffi cult plant roots of carbohydrates, amino acids, vitamins, to get agricultural scientists involved in studying and enzymes and other compounds into the root zone, plus evaluating SRI because it represents such a different rhizodeposition of dead root cells (Brimecombe et al. paradigm from mainsteam modern agriculture.12 This 2001; Neumann and Römheld 2001). The combined puts SRI in a Catch-22 position since those most will- effects of these processes is to assist plants in growing ing to give it enough credence to try the methods have faster, larger and remain healthier. Specifi cally with been NGOs, farmers, and social scientists, all interested regard to rice plants, Yanni et al. (2001) have docu- in innovations that would be particularly appropriate mented the extent to which rhizobacteria in their root for small and marginal farmers and for benefi ting the zone increase the protein content as well as rice yield environment. We will review some of the critiques made per hectare. against SRI as they are similar to ones made against There is a vast literature on these relationships, brought other agroecological innovations. Each of these has together in books such as Pinton et al. (2001), Waisel to be evaluated on its own merits, to be sure, and with et al. (2002) and Wardle (2002), that examine what regard to particular places and seasons since one must goes on in the rhizosphere and the rest of the soil. The be careful about generalizing any biological processes incorporation of decades of research by microbiolo- and potentials. However, some general principles can gists on plant-soil-organism-nutrient relationships into usually be educed. mainstream agronomic literature has been slow, but it 6.12.3.1 Unconfirmed Field Observations (UFOs) is increasing. While these relations are acknowledged, most attention is still given to chemical and physical This is how Cassman and Sinclair (2004) have char- relationships in the soil, e.g., Brady and Weil (2002).11 acterized and dismissed SRI because (not enough) As noted already, much of soil and crop research has results have not appeared in peer-reviewed journals. tried to “control for,” by eliminating the effects of There are evaluations such as reported in Chapter 3 microorganisms and other creatures in the soil. SRI with unreproachable methodology that confi rm the dramatically calls attention to the biological dimension reports on SRI performance that have been coming in of crop production and sustainability, supporting a more for half a dozen years. Leading rice research institu- “biocentric” understanding of soil systems. tions in China, India and Indonesia, the three largest rice-producing countries and previously leaders in the 6.12.3 Critiques of SRI Green Revolution, have now validated SRI methods to As noted above, SRI is still controversial in some their satisfaction and are recommending them for dis- scientifi c circles, where the idea of “getting something semination. Probably the most eminent rice scientist in the world, Prof. Yuan Long Ping, known as “the father of hybrid rice” and co-recipient of the 2004 World Food 11 However, as noted in Chapter 3, the amount of attention in the 13th edition of the major text on soil and crop science 12 Sinclair (2004) begins his critique of SRI by saying that has much more detail on soil biological factors than earlier even “Discussion of the system of rice intensifi cation (SRI) editions, and Chapters 11 and 12 represent an excellent in- is unfortunate because it implies SRI merits serious consid- troduction to ‘life in the soil’ and its implications for a more eration.” No consideration is given of the extensive evidence productive, sustainable agriculture. of SRI benefi ts such as summarized in Chapter 3.

104 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Prize, has been working with SRI since 2000 and has Raju in West Godavari, Andhra Pradesh, planted a large been promoting its development in China. The Sichuan fi eld of 44 hectares with SRI methods, and harvested a Academy of Agricultural Sciences has evaluated SRI yield of over 10 tons per hectare, having organized and since 2001 and reported 20-50 percent increases in the managed the operations with the skill that made him a already high rice yields obtained with modern inputs successful commercial producer (Dr. A. Satyanarayana, and hybrid varieties (Zheng et al. 2004). Director of Extension, pers. comm.). Where there are labor supply constraints, such methods may not be Ironically, when SRI methods have been tried on re- feasible, but where they are raising labor productivity, search stations, the yields are usually lower than on there will be incentives to fi gure out how to utilize them. farmers’ fi elds, reversing the common situation where This applies to other agroecological innovations. farmers fi nd it diffi cult to replicate good results that researchers have obtained. IRRI’s fi rst SRI trial yielded Where labor productivity can be raised by better use of 1.44 tons per hectare, and the next season, the yield was other resources, mobilizing biological support, adoption 3 tons per hectare (Rickman 2003). SRI trials conducted and spread should become feasible. An evaluation of by the Philippine Department of Agriculture’s Agricul- SRI in Cambodia among 120 farmers who have used tural Training Institute in southern Mindanao averaged it for three years found that 55 percent considered the 12 tons per hectare (ATI 2002.) Why the difference? It methods requiring less labor and effort; only 18 percent probably has something to do with the abundance and reported that it was more diffi cult (Tech 2004). The diversity of soil biota, probably adversely affected by GTZ evaluation of SRI in Cambodia, where farmers are decades of monocropping and heavy application of gaining experience and confi dence with it, analysis of agrochemicals of many kinds. If the development of actual labor expenditure found SRI to be labor-neutral, SRI had been left to scientists working on research with some benefi t from when labor requirements are stations, its benefi ts and opportunities would probably reduced or increased (Anthofer 2004). There should never have emerged. be no limitations on such innovations according to scale. If productivity gains come from capitalization 6.12.3.2 Limitation to Small Scale upon biological potentials, these should be accessible Because SRI was developed to be of benefi t to small on various scales. farmers in Madagascar, enabling them to feed their 6.12.3.3. Limitation to Certain “Niches” families without continuing to encroach on remaining forest ecosystems through slash-and-burn cultivation, This has been concluded by Dobermann (2004), that it has been most demonstrated and most successful on SRI is indeed a productive innovation, but only for a small scale. Its initial labor-intensity means that it is certain soil (e.g., high Fe content) or other conditions. diffi cult to adopt directly on a large scale. It has been This claim has been made without systematic evidence. argued against SRI that it cannot have a signifi cant As farmers have seen the benefi ts of SRI in Cambodia, impact on world food supplies for this reason. In fact, it has been spreading throughout the country, and this because there are many millions of small farmers in year is expected to be used by 40-50,000 farmers. The the world, an innovation that is particularly suitable noted rural development consultant, Roland Bunch, for them could cover a large share of production areas. on a consulting assignment in Cambodia for the NGO There is nothing wrong with an innovation that meets ADRA, reported that 100 farmers in one village had the food security needs of the poor directly. That it may tried SRI but only with assurances from ADRA that it not be appropriate for all farmers is no reason to pass would compensate them for any shortfalls in production up the benefi ts it could provide to those for whom it because their average yield was very low, only 1 ton is suitable. per hectare. When Bunch visited the village in May, 2003, he was told that the average SRI yield had been As farmer interest, confi dence and experience with SRI 2.5 tons per hectare, with not a single farmer asking for increases, it is being used on a larger scale. This past any compensation (Bunch, pers. comm.). In this case, an rabi season (2003-2004), one large farmer, N. V. R. C. environment that had not been predicted to respond to

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 105 the new methods (farmers there had little water control) though some will be normal with any innovation, as not gave a very good result even though the methods were all may benefi t or get suffi cient benefi t from it. not fully applied. The Moser-Barrett study usefully called attention to When SRI was introduced into Andhra Pradesh, India, the fact that the very poor might not be able to adopt the state agricultural university and extension service SRI because their income liquidity is so low that they oversaw 300 on-farm trials across all 22 districts of that cannot afford to invest more labor in an innovation even state. The average SRI yields were 8.34 tons per hectare if it would yield them net economic benefi ts. And, of compared with 4.89 tons per hectare with same variety course, the additional labor requirements themselves on adjoining plots, measured by government personnel. may represent a critical constraint to adoption in some In fi ve districts, the average was over 10 tons per hect- settings. The IWMI evaluation of SRI in Sri Lanka, on are, and only on saline soils was a substantial increase the other hand, found no difference in SRI adoption not seen. Thus, so far there is no basis for concluding by richer and poorer farmers, and it even found that that SRI benefi ts will be limited to only certain agro- the poor were less likely to disadopt once they started ecosystem conditions.13 The point here is that previous (Namara et al. 2004). experience and scientifi c fi ndings based on different growing conditions may not be valid for practices and 6.12.3.5 Soil Depletion farming systems that change the growing environment. A different critique of SRI, which at least implicitly Biological processes can change parameters, so that an acknowledges its productivity gains, is that such high empirical approach should be taken to evaluation rather yields will quickly deplete the soil unless nutrient than rely too heavily on a priori reasoning. amendments are made. As noted above, SRI and re- 6.12.3.4 Disadoption lated agroecological improvements are not necessarily adverse to inorganic soil amendments. If a particular One study in Madagascar found that the rate of disadop- defi ciency develops, such as phosphorous, there is tion of SRI practices as rather high, 40 percent (Moser nothing wrong with remedying this. What is suggested, and Barrett 2003). This may have been true for the fi ve though, is that there should be a burden of proof on this villages studied, but there is evidence from the same recommendation, that it will make net and sustainable part of the country that SRI adoption has proceeded improvements to soil fertility. With biologically savvy very rapidly, even without systematic extension efforts, plant, soil, water and nutrient management practices, it the area of SRI in a French irrigation project expanding is being shown that soil systems have more capacity to from 35 hectares to 543 hectares over fi ve years (Hirsch mobilize “unavailable” nutrients and make them avail- 2000). Average SRI yields were 8.55 tons per hectare, able for plant nutrition than the mainstream approach compared with 3.77 tons per hectare for the govern- to agricultural improvement allows. ment-recommended “modern” system of rice produc- Nitrogen, the main plant nutrient requirement, can be tion, and 2.36 tons per hectare for peasant practice. The provided free from the virtually unlimited atmospheric spread of SRI in Cambodia, reported above, indicates supplies through a variety of biological pathways. There that disadoption need not be a signifi cant problem, are vast amounts of phosphorus, often considered the most limiting nutrient in many soil systems, complexed 13 First-season trials in Mozambique with SRI methods on in a wide variety of organic and inorganic molecules saline soils have produced yields as good or better (up to 2.5 that can be accessed by soil microbes under supportive times more) without chemical fertilizer and irrigation than the current average yield on good soils with fertilizer and conditions. There is bound to be disagreement for some irrigation, 3 tons per hectare (Maria Zelia Meneta, pers. time as to how much, and whether adequate, supplies comm.). A second season of trials is currently underway to can be mobilized by these processes. check out these results further. But it appears that even soil salinity may not be a major constraint if soil organic matter We do not want to present any overly optimistic con- is increased suffi ciently. clusions. Our task here is not to recommend alternative practices. Rather, we are reporting opportunities that

106 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations warrant further study and testing in the fi eld, to ascertain in Brazil, where the introduction of low-tillage soy- their potentials and their limitations. The conclusion beans has caught on rapidly, leading to enhanced soil we can confi dently offer is that the closed-system, conservation and productivity on cultivated areas, but zero-sum thinking about soil-plant nutrition is prob- leading also to the centralization of ownership, with a ably too reductionist and abiological. The adequacy or corresponding migration of thousands of unemployed inadequacy of soil nutrient supplies should be treated agricultural smallholders to less productive areas which as an empirical matter without presumptions that ex- were subsequently deforested (McNeely and Scherr clude examination of different plant-soil-water-nutrient 2002). This is why any evaluations need to take a sys- management systems that could, by enhancing and temic perspective, i.e., a “general equilibrium” analysis, mobilizing soil biological capacities, that could make not just a partial-equilibrium, ceteris paribus analysis, agricultural systems more productive, profi table and as discussed in Chapter 7. sustainable, with positive externalities for the conservation of biodiversity. This extended discussion of SRI has 6.14 Possible Contributions of not been intended to promote its adoption but to open Biotechnology up consideration of ways of practicing agriculture that In thinking about ecoagricultural alternatives, we should could be supportive of the new directions proposed for consider what contribution, if any, biotechnology make ecoagriculture. to the conservation of biodiversity. Many claims are 6.13 Discussion made to this effect. As a rule, advances in biotechnology have been attempted within the paradigm of the Green Twenty years ago, two axioms of agriculture were (a) Revolution, seeking to make inputs more productive that good yields of fi eld crops depend on plowing, and rather than to minimize dependence on external factors (b) the best yields from rice depend on maintaining of production. Researchers have mostly looked for solu- constant fl ooding of fi elds. Both of these are being tions that bypass natural system constraints rather than proven wrong by the experience with conservation to work within them. This often neglects potentially agriculture and SRI. In both cases, higher production benefi cial natural system contributions that could be is being achieved with a reduction in most if not all strengthened or reinforced inputs, contradicting the principle that to get more, you Future increases in crop yield will be benefi cial if must invest more. This truism comes from the realm of they also provide secondary benefi ts, such as greater industrial production, as for generations we have tried suppression of weed growth due to direct competition to “industrialize” agriculture. Yet, with a selective har- by a robust crop. Great care must be taken to evaluate nessing of biological processes and potentials, as seen secondary effects of biotechnology, however, because not only with SRI and conservation agriculture, but also risks can be increased. Critiques of genetic-modifi ca- agroforestry, IPM, integrated nutrient management, and tion (GM) technology have often turned not on the GM other ecologically-grounded approaches to agriculture, per se but on the effects of associated chemical inputs, the sector is being reoriented to its biological origins. such as the evaluation published by a committee of This is what ecoagriculture needs and encourages. the Royal Society in the U.K. in October, 2003, which With a more biologically-based agriculture, inten- critiques adverse environmental impacts on biodiversity sifi cation becomes feasible because it can be more from the herbicides which Roundup-ready maize was economically productive as well as environmentally engineered to tolerate. benign. By increasing production on areas that are suit- Certain intensive and disruptive practices, e.g., glypho- able for such cultivation, the advantages of exploiting sate application to herbicide-resistant crop varieties, currently uncultivated areas should diminish, reducing can facilitate the implementation of some resource- threats to the future for wild biodiversity. We need to conserving agricultural practices, e.g., soil-conserving note, however, that intensifi cation may be necessary minimum-tillage practices, by overcoming certain limi- but not suffi cient for purposes of conservation. We tations on productivity, in this instance, weed control. need to avoid tradeoffs such as have been experienced

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 107 Thus, blanket judgments about any particular category resource use, and maintenance of wild biodiversity. of practice are less tenable than more qualifi ed, well- However, biotechnological research could make a con- informed conclusions that address concrete situations tribution by taking a broader view, rather than focusing and needs in their entirety. In general, it can be said that just on fi xing fl aws with present practices. It cannot give the sustainability of production can be greater with less up the reductionist and mechanistic methodologies that adverse impact on biotic resources, by developing inte- molecular biology requires. This is a strength whose grated cropping systems that rely upon cover cropping limits are still not known. It is also a weakness whenever and crop rotation, for example, to maximize ecosystem this focus is on single organisms without reference to contributions to fertility management and pest control. what these organisms do to others and how they are in The use of chemical controls in such a situation is better turn affected by others. regarded as an interim solution than a fi nal one. REFERENCES Care should be taken not to overexploit natural cycles that are most easily understood while neglecting impor- Abunyewa, A. A., and F. K. Padi. 2003. Changes in tant feedback effects that could tap broader ecosystem soil fertility and Striga hermonthica prevalence as- benefi ts such as disease resistance and sequestration sociated with legume and cereal cultivation in the of carbon. Providing ecosystem services through inte- Sudan savannah zone of Ghana. Land Degradation grated management can protect soil health, downstream and Development 14: 335-343. water sources, and the natural biota. There are multiple AEE. 2003. Preface to special issue on “Balanced nu- advantages from engaging endogenous biological trient management systems for cropping systems in cycles that support agricultural systems and practices the tropics: From concept to practice.” Agriculture, so that reliance on exogenous inputs can be reduced, Ecosystems and Environment 100: 99-102. promoting healthier soils and more robust crops with concomitant economic profi tability. Alkorta, I., I. Albizu, and C. Garbisu. 2003. Biodiversity and agroecosystems. Biodiversity and Conservation Given the large fi nancial interests at stake with bio- 12: 2521-2522. technological development, the economic benefi ts to agrobusinesses are likely to remain a powerful force Angers, D. A., and J. Caron. 1998. Plant-induced in future decision-making processes. Great care must changes in soil structure: Processes and feedbacks. be taken to ensure adequate scientifi c oversight and Biogeochemistry 42: 55-72. evaluation of proposed biotechnological solutions, and Anthofer, Jürgen et al. 2004. Evaluation of the System to try in all cases to promote the use of natural systems of Rice Intensifi cation (SRI) in Cambodia, Febru- whenever possible. Otherwise, exploitation of the ag- ary-April 2004. Report prepared for GTZ, Phnom ricultural landscape without regard for the complexity Penh. of natural processes can lead to degraded systems and Arshad, M. A., and S. Martin. 2002. Identifying critical dwindling yields, ever more dependent upon external limits for soil quality indicators in agroecosystems. inputs, with increased potential for system collapse. Agriculture, Ecosystems and Environment 88: 153- This would have adverse effects not only for wild bio- 160. diversity but for human communities as well. Asteraki, E. J., B. J. Hart, T. C. Ings, and W. J. Manley. Ecoagricultural approaches appear to provide viable 2004. Factors infl uencing the plant and invertebrate alternatives that are based upon scientifi c understanding diversity of arable fi eld margins. Agriculture, Eco- and management of complex biological processes to systems and Environment 102 :219-231. promote more abundant crop yields, more sustainable ATI. 2002. SRI Field Day Report, October 28, 2002, word processed report. Cotobato, Philippines: Agricultural Training Institute, Department of Ag- riculture.

108 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Atta-Krah, K., R. Kindt, J.N. Skilton and W. Amaral. Organic Substances at the Soil-Plant Interface, 95- 2004. Managing biological and genetic diversity in 140. Marcel Dekker, New York. tropical agroforestry. In: P. K. R. Nair, M. R. Rao Cadet, P., L. Guichaoua, and V. W. Spaull. 2004. Nema- and L. E. Buck, eds., New Vistas in Agroforestry. todes, bacterial activity, soil characteristics and plant Dordrecht, Netherlands: Kluwer. 183-194. growth associated with termitaria in a sugarcane fi eld Aude, E., K. Tybirk, and M. B. Pedersen. 2003. Vegeta- in South Africa. Applied Soil Ecology 25: 193-206. tion diversity of conventional and organic hedgerows Calegari, A. 2002. The spread and benefi ts of no-till in Denmark. Agriculture, Ecosystems and Environ- agriculture in Paraná state, Brazil. In: N. Uphoff, ment 99: 135-147. ed., Agroecological Innovations: Increasing Food Barison, Joeli. 2002. Nutrient-use effi ciency and nutri- Production with Participatory Development, 187- ent uptake in conventional and intensive (SRI) rice 202. Earthscan, London. cultivation systems in Madagascar. Master’s thesis, Cassman, K. G., and T. Sinclair. 2004. Agronomic Crop and Soil Science Department, Cornell Univer- UFOs? ACSSA [Agronomy, Crop and Soil Science sity, Ithaca, NY. Alert], March 2004. Belaoussoff, S., P. G. Kevan, S. Murphy, and C. Swan- Cerosaletti, P. E., D. G. Fox, L. E. Chase and D. C. Fra- ton. 2003. Assessing tillage disturbance on assem- zier. 2003. Precision feed management and its role blages of ground beetles (Coleoptera: Carabidae) in watershed protection. In: Proceedings of Cornell by using a range of ecological indices. Biodiversity Nutrition Conference for Feed Manufacturers, Syra- and Conservation 12: 851-882. cuse, NY, 251-259. Cornell University,University, Ithaca, NY.NY. Boddy, R. M., O. C. de Oliviera, S. Urquiaga, V. M. Cheng, W., D. W. Johnson, and S. Fu. 2003. Rhizo- Reis, F. L. de Olivares, V. L. D. Baldini, and J. sphere effects on decomposition: Controls of plant Döbereiner. 1995. Biological nitrogen fi xation as- species, phenology, and fertilization. Soil Science sociated with sugar cane and rice: Contributions Society of America Journal 67: 1418-1427. and prospects for improvement. Plant and Soil 174:195-209. Chikowo, R., P. Mapfumo, P. Nyamugafata, and K. E. Giller. 2004. Maize productivity and mineral N Bommarco, R. 1998. Reproduction and energy reserves dynamics following different soil fertility manage- of a predatory carabid beetle relative to agroeco- ment practices on a depleted sandy soil in Zimbabwe. system complexity. Ecological Applications 8: Agriculture, Ecosystems and Environment 102: 846-853. 119-131. Bonkowski, M. 2004. Protozoa and plant growth: The Clark, M. S., W. R. Horwath, C. Shennan, and K. M. microbial loop in soil revisited. New Phytologist, Scow. 1998. Changes in soil chemical properties re- 162, 617-631. sulting from organic and low-input farming practices. Bouma, J. 2002. Land quality indicators of sustainable Agronomy Journal 90: 662-671. land management. Agriculture, Ecosystems and En- CNCPS. 2004. Cornell Net Carbohydrate and Protein vironment 88: 129-136. System. Available from http://www.cncps.cornell. Brady, N. C., and R. Weil. 2002. The Nature and Prop- edu/cncps/description.htm. erties of Soils, 13th edition. Prentice-Hall, Upper Cox, T. S., M. Bender, C. Picone, D. L. V. Tassel, J. Saddle, NJ. B. Holland, E. C. Brummer, B. E. Zoeller, A. H. Brimecombe, M., F. A. DeLeij, and J. M. Lynch. 2001. Paterson, and W. Jackson. 2002. Breeding peren- The effect of root exudates on rhizosphere microbial nial grain crops. Critical Reviews in Plant Sciences populations. In: R. Pinton, Z. Varanini, and P. Nan- 21: 59-91. nipieri, eds., The Rhizosphere: Biochemistry and

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 109 Crews, T. E., and M. B. Peoples. 2004. Legume versus Döbereiner, J. 1987. Nitrogen-Fixing Bacteria in Non- fertilizer sources of nitrogen: Ecological tradeoffs Leguminous Crop Plants. Berlin: Springer Verlag. and human needs. Agriculture, Ecosystems and Doberman, A. 2004. A critical assessment of the system Environment 102: 279-297. of rice intensifi cation (SRI). Agricultural Systems Current, D., E. Lutz, and S. J. Scherr, eds. 1995. Costs, 79(3):261-281. Benefits and Farmer Adoption of Agroforestry: Doran, J. W. 2002. Soil health and global sustainability: Project Experience in Central America and the Carib- Translating science into practice. Agriculture, Eco- bean. World Bank Environment Paper Number 14. systems and Environment 88: 119-127.119-127. Washington, DC: World Bank Doran, J. W., and M. R. Zeiss. 2000. Soil health and Dahlin, S., E. Witter, A. Martensson, A. Turner, and sustainability: Managing the biotic component of soil E. Baath. 1997. Where’s the limit? Changes in the quality. Applied Soil Ecology 15: 3-11. microbiological properties of agricultural soils at low levels of metal contamination. Soil Biology and Drinkwater, L. E., M. W. Wagoner and M. Sarrantonio. Biochemistry 29: 1405-1415. 1998. Legume-based systems have reduced losses of nitrogen and carbon. Nature 396: 262-265. Delgado, J. A., and C. A. Cox. 2003. Soil quality: Simi- lar goals but different pathways. Journal of Soil and Drinkwater, L. E. 2002. Cropping systems research: Water Conservation 58: 170. Reconsidering agricultural experimental approaches. HortTechnology 12: 355-361. deSnoo, G. R. 1999. Unsprayed fi eld margins: Effects on environment, biodiversity and agricultural prac- Drinkwater, L. E. 2004. Improving nitrogen fertilizer tice. Landscape and Urban Planning 46: 151-160. use effi ciency through an ecosystem-based approach. In: A. R. Mosier, J. K. Sayers, and J. R., Freney, eds., deSnoo, G. R., and R. J. v. d. Poll. 1999. Effect of Agriculture and the Nitrogen Cycle: Assessing the herbicide drift on adjacent boundary vegetation. Impacts of Fertilizer Use on Food Production and Agriculture, Ecosystems and Environment 73: 1-6. the Environment. SCOPE Nitrogen Fertilizer RAP, Dinesh, R., S. G. Chaudhuri, A. N. Ganeshamurthy, in press. and C. Dey. 2003. Changes in soil microbial indices Eghball, B., D. Ginting, and J. E. Gilley. 2004. Residual and their relationships following deforestation and effects of manure and compost applications on corn cultivation in wet tropical forests. Applied Soil Ecol- production and soil properties. Agronomy Journal ogy 24: 17-26. 96:442-447. Diop, A. 2002. Management of organic inputs to in- Ekschmitt, K., T. Stierhof, J. Dauber, K. Kreimes, and crease food production in Senegal. In: N. Uphoff, V. Wolters. 2003. On the quality of soil biodiver- ed., Agroecological Innovations: Increasing Food sity indicators: Abiotic and biotic parameters as Production with Participatory Development, 125- predictors of soil faunal richness at different spatial 137, Earthscan, London. scales. Agriculture, Ecosystems and Environment Dix, M. E., B. Bishaw, S. W. Workman, M. R. Barnhart, 98: 273-284. N. B. Klopfenstein, and A. M. Dix. 1999. In: L. E. Elphick, C. S. and L. W. Oring. 1998. Winter man- Buck, J. P. Lassoie and E. C. M. Fernandes, eds., agement of Californian rice fi elds for waterbirds. Agroforestry in Sustainable Agricultural Systems, Journal of Applied Ecology. 35: 95-108. 131-156. Boca Raton, FL: Lewis Publishers. Ferris, H., R. C. Venette, and K. M. Scow. 2004. Soil Doebbelaere, S., J. Vanderleyden, and Y. Okon. 2003. management to enhance bacterivore and fungivore Plant growth-promoting effects of diazotrophs in nematode populations and their nitrogen mineralisa- the rhizosphere. Critical Review in Plant Science tion function. Applied Soil Ecology 25: 19-35. 22: 107-149.

110 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Filip, Z. 2002. International approach to assessing soil in Agroforestry, 5-17. Dordrecht, Netherlands: quality by ecologically-related biological param- Kluwer. eters. Agriculture, Ecosystems and Environment Griffi th, D. M. 2000. Agroforestry: A refuge for tropi- 88: 169-174. cal biodiversity after fi re. Conservation Biology 14: Fliessbach, A., and P. Mäder. 2000. Microbial biomass 325-326. and size-density fractions differ between soils of Gyaneshwar, P., G. Naresh Kumar, L. J. Parekh and P. S. organic and conventional agricultural systems. Soil Poole. 2002. Role of soil microorganisms in improv- Biology and Biochemistry 32: 757-768. ing P nutrition in plants. Plant and Soil 245, 83-93. Fofana, B., H. Bremen, R. J. Carsky, H. Van Reuler, Haefele, S. M., M. C. S. Wopereis, A. M. Schloebohm, A. F. Tamelokpo, and D. Gnakpenou. 2004. Using and H. Wiechmann. 2004. Long-term fertility experi- mucuna and P fertilizer to increase maize grain yield ments for irrigated rice in the West African Sahel: and N fertilizer use effi ciency in the coastal savanna Effect on soil characteristics. Field Crops Research of Togo. Nutrient Cycling in Agroecosystems 68: 85: 61-77. 213-222. Hald, A. B. 2002. Impact of agricultural fi elds on veg- Fox, D. G., T. P. Tylutki, G. L. Albrecht, P. E.Cerosaletti, etation of stream border ecotones in Denmark. Agri- and L.O. Tedeschi. 2002. Environmental protection culture, Ecosystems and Environment 89: 127-135. and the Cornell University Nutrient Management Planning System. In: Proceedings of Cornell Nutri- Hamilton, E. W., and D. A. Frank. 2001. Can plants tion Conference for Feed Manufacturers, Syracuse, stimulate soil microbes and their own nutrient sup- NY, 79-98. Cornell University,University, Ithaca, NY.NY. ply? Evidence from a grazing tolerant grass. Ecology 82: 2397-2402. Frankenberger, W. T., and M. Arshad. 1995. Phytohor- mones in Soils: Microbial Production and Function. Hammer, K., N. Arrowsmith, and T. Gladis. 2003. Marcel Dekker, New York. Agrobiodiversity with emphasis on plant genetic resources. Naturwissenschaften 90: 241-250. Frei, M., and K. Becker. 2004. Agro-biodiversity in subsistence-oriented farming system in a Philippine Haynes, R. J., and M. H. Beare. 1997. Infl uence of six upland region: Nutritional considerations. Biodiver- crop species on aggregate stability and some labile sity and Conservation 13: 1591-1610. organic matter fractions. Soil Biology and Biochem- istry 29: 1647-1653. Fu, B., S. Liu, Y. Lu, L. Chen, K. Ma and G. Liu. 2003. Comparing the soil quality changes of different land Hector, A., B. Schmid, C. Beierkuhn-Lein, M. C. uses determined by two quantitative methods. Jour- Caldeira, M. Diemer, P.G. Dimitrakopoulos, J. A. nal of Environmental Sciences 15: 167-172. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P. Högberg, K. Huss-Danell, J. Joshi, A. Jumpponen, Gacheru, E., and M. R. Rao. 2001. Managing Striga C. Körner, P. W. Leadley, M. Loreau, A. Minns, C. P. infestation on maize using organic and inorganic H. Mulder, G. O’Donovan, S. J. Otway, J. S. Pereira, nutrient sources in western Kenya. International A. Prinz, D. J. Read, M. Scherer-Lorenzen, E. D. Journal of Pest Management 47: 233-239. Shulze, A. S. D. Siamantziouras, E. M. Spehn, A. C. Ganry, F., C. Feller, J. M. Harmand, and H. Guibert. Terry, A. Y. Troumbis, F. I. Woodward, S. Yachi, and 2001. Management of soil organic matter in semiarid J. H. Lawton. 1999. Plant diversity and productivity Africa for annual cropping systems. Nutrient Cycling experiments in European grasslands. Science 286: in Agroecosystems 61: 105-118. 1123-1127. Garrity, D. P. 2004. Agroforestry and the achievement Heyer, W., K. J. Hülsbergen, C. Wittmann, S. Papaja, of the Millennium Development Goals. In: P. K. R. and O. Christen. 2003. Field related organisms as Nair, M. R. Rao, and L. E. Buck, eds, New Vistas possible indicators for evaluation of land use inten-

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 111 sity. Agriculture, Ecosystems and Environment 98: L.) breeding lines derived from a low nitrogen maize 453-462. population. Journal of Agricultural Science 141: 221-230. Hill, D. B., and L. E. Buck. 2000. Forest farming practices. In: H. E. Garrett, W. J. Rietveld, and R. Kapulnik, Y., and Y. Okon. 2001. Plant growth pro- F. Fisher, eds., North American Agroforestry: An motion by rhizosphere bacteria. In: Plant Roots: Integrated Science and Practice, 283-320. American The Hidden Half, YY.. WWaisel,aisel, eds., 869-885. Marcel Society of Agronomy, Madison, WI. Dekker, New York. Hirsch, R. 2000. La Riziculture Malgache Revisitée: Kar, S., Varade, S. B., Subramanyam, T. K., and Diagnostic et Perspectives (1993-99). Report to Ghildyal, B. P. 1974. Nature and growth pattern of Agence Française de Développement, Départment rice root system under submerged and unsaturated des Politiques et des Etudes, Antananarivo. conditions. Il Riso (Italy) 23: 173-179. Hobbs, P., and R. Gupta. 2004. Problems and challenges Karlen, D. L., S. S. Andrews, B. J. Weinhold, and J. W. of no-till farming for the rice-wheat systems of the Doran. 2003. Soil quality: Humankind’s foundation Indo-Gangetic Plains in South Asia. In: R. Lal, P. H. for survival. Journal of Soil and Water Conservation Hobbs, N. Uphoff, and D. O. Hansen, Sustainable 58: 171-180. Agriculture and the International Rice-Wheat Sys- Kaya, B., and P. K. R. Nair. 2001. Soil fertility and crop tem, 101-120. New York: Marcel Dekker. yields under improved-fallow systems in southern Huang, W., O. Luukkanen, S. Johanson, V. Kaarakka, S. Mali. Agroforestry Systems 52: 1-11. Räisänen, and V. Vihemäki. 2002. Agroforestry for Kendy, E., P. Gerard-Marchant, M. T. Walter, Y. Zhang, biodiversity conservation of nature reserves: Func- C. Lieu, and T. S. Steenhuis. 2002. A soil-water-bal- tional group identifi cation and analysis. Agroforestry ance approach to quantifying groundwater recharge Systems 55: 65-72. from irrigated cropland in the North China Plan. Ikerra, S. T., J. A. Maghembe, P. C. Smithson, and R. J. Hydrological Processes 17: 2011-2031. Buresh. 2001. Dry-season sesbania fallows and their Kendy, E., Y. Zahng, C. Liu, and T. S. Steenhuis. 2004. infl uence on nitrogen availability and maize yields Groundwater recharge in irrigated cropland in the in Malawi. Agroforestry Systems 52: 13-21. North China Plain: Case study of Luancheng County, Jackson, W. 2002. Natural systems agriculture: A truly Hebei Province, 1949-2000. Hydrological Processes, radical alternative. Agriculture, Ecosystems and special issue. In press. Environment 88: 111-117. Klausner, S. D., D. G. Fox, C. N. Rasmussen, R. E. Jama, B., R. A. Swinkels, and R. J. Buresh. 1997. Ag- Pitt, T. P. Tylutki, P. E. Wright, L. E. Chase, and W. ronomic and economic evaluation of organic and C. Stone. 1998. Improving dairy farm sustainability inorganic sources of phosphorus in western Kenya. I: An approach to animal and crop nutrient manage- Agronomy Journal 89: 597-604. ment planning. Journal of Production Agriculture 11: 225-233. Jama, B., R. J. Buresh, J. K. Ndufa, and K. D. Shepherd. 1998a. Vertical distribution of roots and soil nitrate: Kumar, B. M., and P. K. R. Nair. 2004. The enigma of Tree species and phosphorus effects. Soil Science tropical homegardens. In: P. K. R. Nair, M. R. Rao, Society of America Journal 62: 280-286. and L. E. Buck, eds., New Vistas in Agroforestry, 135-152. Dordrecht, Netherlands: Kluwer. Jama, B., R. J. Buresh, and F. M. Place. 1998b. Sesbania tree fallows on phosphorus-defi cient sites: Maize Kumar, V., D. J. Mills, J. D. Anderson, and A. K. yield and fi nancial benefi t. Agronomy Journal 90: Mattoo. 2004. An alternative agricultural system is 717-726. defi ned by a distinct expression profi le of select gene transcripts and proteins. PNAS: Proceedings of the Kamara, A. Y., J. G. Kling, A. Menkir, and O. Ibikunle. National Academy of Sciences 10: 10535-10540. 2003. Agronomic performance of maize (Zea mays

112 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Kuniyal, J. C., S. C. R. Vishvakarma, and G. S. Singh. Lightfoot, C., and R. Noble. 2001. Tracking the eco- 2004. Changing crop biodiversity and resource use logical soundness of farming systems: Instruments effi ciency of traditional versus introduced crops in and indicators. Journal of Sustainable Agriculture the cold desert of the northwestern Indian Himalaya: 19: 9-29. A case of the Lahaul valley. Biodiversity and Con- Long, C.-L., H. Li, Z. Ouyang, X. Yang, Q. Li, and B. servation 13: 1271-1304. Trangmar. 2003. Strategies for agrobiodiversity con- Kuntashula, E., P.L. Mafongoya, G. Sileshi, and S. servation and promotion: A case from Yunnan, China. Lungu. 2004. Potential of biomass transfer tech- Biodiversity and Conservation 12: 1145-1156. nologies in sustaining vegetable production in the Machado, P. L. O., and C. A. Silva. 2001. Soil man- wetlands (dambos) of Eastern Zambia. Experimental agement under no-tillage systems in the tropics Agriculture 40: 37-51. with special reference to Brazil. Nutrient Cycling in Laulanié, H. 1983. Le système de riziculture intensive Agroecosystems 61: 119-130. malgache. Tropicultura (Brussels) 11: 110-114. Mäder, P., A. Fliessbach, D. Dubois, and L. Gunst. 2002. Lauriault, L. M., and R. E. Kirksey. 2004. Yield and Soil fertility and biodiversity in organic farming. nutritive value of irrigated winter cereal forage grass- Science 296: 1694-1697. legume intercrops in the Southern High Plains, USA. Marshall, E. J. P., and A. C. Moonen. 2002. Field mar- Agronomy Journal 96: 352-358. gins in northern Europe: Their functions and interac- Leakey, R. R. B., A. R. Atangana, E. Kengni, A. N. Wa- tions with agriculture. Agriculture, Ecosystems and ruhiu, C. Usoro, P. O. Anegbeh, and Z. Tchoundjeu. Environment 89: 5-21. 2002. Domestication of Dacryodes edulis in west Martin, F. M., S. Perotto, and P. Bonfante. 2001. My- and central Africa: Characterization of forest genetic corrhizal fungi: A fungal community at the interface variation. Forest, Trees and Livelihoods 12: 57-71. between soil and roots. In: R. Pinton, Z. Varanini, and Leakey, R. R. B., and T. P. Tomich. 1999. Domestication P. Nannipieri, eds., The Rhizosphere: Biochemistry of Tropical Trees: From Biology to Economics and and Organic Substances at the Soil-Plant Interface, Policy. Boca Raton, FL: CRC Press. 263-296. New York: Marcel Dekker. LeCoeur, D., J. Baudry, F. Burel, and C. Thenail. 2002. McMahon, M. and E. McMahon. 2002. CUNMPS:NMPS: Why and how we should study fi eld boundary biodi- A Producer Perspective. Available from the CALS versity in an agrarian landscape context. Agriculture, Nutrient Management for Dairy and Livestock Ecosystems and Environment 89: 23-40. Farms Program Work Team (www.inmpwt.cce. cornell.edu) Lee, D., and R. Ruben. 2001. Labor use in sustainable agricultural systems. Paper presented at XXIV Con- McNeely, J. A., and S. J. Scherr. 2003. Ecoagriculture: ference of the International Association of Agricul- Strategies to Feed the World and Save Wild Biodi- tural Economists, August, 2001, Berlin, Germany. versity. Washington, DC: Island Press. Letey, J., R. E. Sojka, D. R. Upchurch, D. K. Cassel, McRoberts, N., G. Hughes, and S. Savary. 2003. Inte- K. R. Olson, W. A. Payne, S. E. Petrie, G. H. Price, grated approaches to understanding and control of R. J. Reginato, H. D. Scott, P. J. Smethurst, and G. diseases and pests in fi eld crops. Australasian Plant B. Triplett. 2003. Defi ciencies in the soil quality Pathology 32: 167-180. concept and its application. Journal of Soil and Water Meertens, H. C. C. 2003. The prospects for integrated Conservation 58: 180-187. nutrient management for sustainable rainfed lowland Lewis, W. J., J. C. Van Lenteren, S. C. Pathak, and J. H. rice production in Sukumaland, Tanzania. Nutrient Tumlinson. 1997. A total system approach to sustain- Cycling in Agroecosystems 65: 163-171. able pest management. PNAS: Proceedings of the National Academy of Sciences 94: 12243-12248.

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 113 Meilleur, B. A., and T. Hodgkin. 2004. In situ con- Orr, A. 2003. Integrated pest management for resource- servation of crop wild relatives: Status and trends. poor African farmers: Is the emperor naked? World Biodiversity and Conservation 13: 663-684. Development 31: 831-845. Michon, G., and H. d. Foresta. 1999. Agroforests: In- Orr, A., and J. M. Ritchie. 2004. Learning from failure: corporating a forest vision in agroforestry. In: L. E. Smallholder farming systems and IPM in Malawi. Buck, J. P. Lassoie, and E. C. M. Fernandes, eds., Agricultural Systems 79: 31-54. Agroforestry in Sustainable Agricultural Systems, Oswald, A., and J. K. Ransom. 2001. Striga control and 381-406. Boca Raton, FL: Lewis Publishers. improved farm productivity using crop rotation. Crop Moser, C. M., and C. B. Barrett. 2003. The disap- Protection 20: 113-120. pointing adoption dynamics of a yield-increasing, Peranginangin, N., R. Sakthivadivel, N. R. Scott, E. low external-input technology: The case of SRI in Kendy, and T. S. Steenhuis. in press. Water ac- Madagascar. Agricultural Systems 76, 1085-1100. counting for conjunctive groundwater/surface water Nair, P. K. R., M. R. Rao, and L.E. Buck, eds. 2004. management: Case of the Singkarak-Ombilin river New Vistas in Agroforestry. Dordrecht, Netherlands: basin, Indonesia. Journal of Hydrology. Kluwer. Perfecto, I., A. Mas, T. Dietsch, and J. Vandermeer. Namara, R. E., P. Weligamage, and R. Barker. 2003. 2003. Conservation of biodiversity in coffee agroeco- Prospects for Adopting the System of Rice Intensifi - systems: A tri-taxa comparison in southern Mexico. cation in Sri Lanka: A Socioeconomic Assessment. Biodiversity and Conservation 12: 1239-1252. Research Report No. 75. Colombo, Sri Lanka: Inter- Pimentel, D. P. 1997. Techniques for Reducing Pes- national Irrigation Management Institute. ticides: Environmental and Economic Benefits. National Roadmap for IPM, 2004. http://www.ipmcen- Chichester, UK: John Wiley. ters.org/IPMRoadMap.pdf. Pinton, R., Z. Varanini, and P. Nannipieri, eds. 2001. Neumann, G. and V.Römheld. 2001. The release of The Rhizosphere: Biochemistry and Organic Sub- root exudates as affected by the plant’s physiological stances at the Soil-Plant Interface. New York: status. In: R. Pinton, Z. Varanini, and P. Nannipieri, Marcel Dekker. eds., The Rhizosphere: Biochemistry and Organic Place, F., C. B. Barrett, H. A. Freeman, J. J. Ramisch, Substances at the Soil-Plant Interface, 41-93. New and B. Vanlauwe. 2003. Prospects for integrated soil York: Marcel Dekker. fertility management using organic and inorganic in- New York State Soil and Water Conservation Commit- puts: Evidence from smallholder African agricultural tee. 2004. Agricultural Environmental Management. systems. Food Policy 28: 365-378. Available at: http://www.agmkt.state.ny.us/soilwater/ Puget, P. and L. E. Drinkwater. 2001. Short-term AEM/AEM.html Accessed May 2004. dynamics of root and shoot-derived carbon from a NMSP. 2004. Nutrient Management Spear Program. leguminous green manure. Soil Science Society of Available from: http://nmsp.css.cornell.edu/publica- America Journal 65: 771-779. tions/index.asp. Qualset, C. O., P. E. McGuire, and M. L. Warburton. NMPWT. 2004. CALS Integrated Nutrient Management 1995. ‘Agrobiodiversity’ is key to agricultural pro- Program Work Team. Available from: http://www. ductivity. California Agriculture 49: 45-49. inmpwt.cce.cornell.edu/. Randriamiharisoa, Robert. 2002. Research results on NRC. 2003. Frontiers in Agricultural Research. Food, biological nitrogen fi xation with the System of Rice Health, Environment, and Communities. National Intensifi cation. In: Assessments of the System of Rice Academic Press, Washington, DC. Intensification: Proceedings of an international conference, Sanya, China, April 1-4, 2002, eds. N. Uphoff et al., 148-157. Ithaca, NY: CIIFAD.

114 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Randriamiharisoa, Robert, and Norman Uphoff. 2002. Sayre, K. D., and P. R. Hobbs. 2004. The raised-bed sys- Factorial trials evaluating the separate and com- tem of cultivation for irrigated production conditions. bined effects of SRI practices. In: Assessments of In: R. Lal, P. Hobbs, N. Uphoff and D.O. Hansen, the System of Rice Intensifi cation: Proceedings of eds., Sustainable Agriculture and the Rice-Wheat an international conference, Sanya, China, April System, 337-355. New York: Marcel Dekker. 1-4, 2002, eds. N. Uphoff et al., 40-46. Ithaca, NY: Schloter, M., O. Dilly, and J. C. Munch. 2003. Indicators CIIFAD. for evaluating soil quality. Agriculture, Ecosystems Raobelison, F. D. E. 2000. Suive experimental de la and Environment 98: 255-262. teneur en elements fertilisants liberes par la fumure Schroth, G., and F. L. Sinclair. 2003. Impacts of trees on organique au cours du cycle de developpement du the fertility of agricultural soils. In: G. Schroth and riz: Cas du SRI au CDIA, Beforona. Memoire de F. L. Sinclair, eds., Trees, Crops and Soil Fertility. Fin d’Etudes, Faculty of Agriculture, University of Wallingford, UK: CABI International. Antananarivo, Madagascar. Schroth, G., G. A. B da Fonseca, C. A. Harvey, C. Robertson, G. P., and K. M. Klingensmith. 1997. Soil Gascon, H. L. Vasconcelos, A-M. Izac, eds. 2004. resources, microbial activity and primary produc- Agroforestry and Biodiversity Conservation in Tropi- tion across an agricultural ecosystem. Ecological cal Landscapes. Covela, CA: Island Press. Applications 7: 158-170. Scott, N. A. 1998. Soil aggregation and organic matter Roose, E., and B. Barthes. 2001. Organic matter man- mineralization in forests and grasslands: Plant spe- agement for soil conservation and productivity. cies effects. Soil Science Society of America Journal Nutrient Cycling in Agroecosystems 61: 159-170. 62:1081-1089. Sanchez, P. A., B. Jama, A. I. Niang, and C. A. Palm. Seghers, D., L. Wittebolle, E. M. Top, and W. Vers- 2001. Soil fertility, small-farm intensifi cation and traete. 2003. Impact of agricultural practices on the the environment in Africa. In: C. B. Barrett and Zea mays L. endophytic community. Applied and D. R. Lee, eds., Tradeoffs or Synergies? Agricul- Environmental Microbiology 70: 1475-1482. tural Intensifi cation, Economic Development and the Environment, 325-344. Wallingford, UK: CAB Seguy, L., S. Bouzinac, E. Scopel, and F. Ribeiro. International. 2003a. New concepts for sustainable management of cultivated soils through direct seeding mulch-based Sanchez, P. A., C. A. Palm, and S. W. Buol. 2003. cropping systems: The CIRAD experience, part- Fertility capability soil classifi cation: A tool to help nership and networks. Proceedings of the II. World assess soil quality in the tropics. Geoderma 114: Congress on Conservation Agriculture, Iguassu 157-185. Falls, Brazil, August, 284-285. Sanginga, N. 2003. Role of biological nitrogen fi xation Seguy, L., S. Bouzinac, A. C. Maronezzi, E. Scopel, in legume-based cropping systems: A case study of J. L. Belot, and J. Martin. 2003b. The success of West Africa farming systems. Plant and Soil 252: no-tillage with cover crops for savannah regions: 25-39. From destructive agriculture with soil tillage to Sanginga, N., O. Lyasse, and J. Diels. 2003. Balanced sustainable agriculture with direct seeding, mulch- nutrient management systems for cropping systems based systems—20 years of research of CIRAD in the tropics: From concept to practice. Agriculture, and its Brazilian partners in the cerrados region of Ecosystems and Environment 100: 99-102. Brazil. Proceedings of the II. World Congress on Conservation Agriculture, Iguassu Falls, Brazil, Sauerborn, J., B. Kranz, and H. Mercer-Quarshie. 2003. August, 310-321. Organic amendments mitigate heterotrophic weed infestation in savannah agriculture. Applied Soil Shanker, C., and K. R. Solanki. 2000. Agroforestry: An Ecology 23: 81-186. ecofriendly land-use system for insect management. Outlook on Agriculture 29: 91-96.

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 115 Sheehy, J. E., S. Peng, A. Dobermann, P. L. Mitchell, Tech, C. 2004. Ecological System of Rice Intensifi ca- A. Ferrer, J. Yang, Y. Zou, X. Zhong, J. Huang. 2004. tion (SRI) Impact Assessment. CEDAC Field Docu- Fantastic yields in the system of rice intensifi cation: ment. Phnom Penh: Cambodia Center for the Study Fact or fallacy? Field Crops Research 88: 1-8. and Development of Agriculture. Simon, A. J., and R. R. B. Leakey. 2004. Tree domesti- Thrupp, L. A. 2000. Linking agricultural biodiversity cation in tropical agroforestry. In: P. K. R. Nair, M. and food security: The valuable role of agrobiodi- R. Rao, and L. E. Buck, eds., New Vistas in Agrofor- versity for sustainable agriculture. International estry. Dordrecht, Netherlands: Kluwer. Affairs 76: 265. Simons, A. J., Z. Tchoundjeu, M. Munjuga, W. J., I. Turner, B. and P. Haygarth. 2001. Phosphorus solubili- Dawson, S. Ruigi, A. Lengkeek, and A. Degrande. zation in rewetted soils. Nature 411, 258. 2000. The future of trees is on farms: Tree domes- Tylutki, T. P., D. G. Fox, and M. McMahon. 2003. tication in Africa. In: Forest and Society: The Role Implementation of the cuNMPS: Development and of Research. XXI IUFRO World Congress, Kuala Evaluation of Alternatives. Department of Animal Lumpur, Malaysia. Science, Cornell University, Ithaca NY. Sinclair, Thomas R. 2004. Agronomic UFOs waste Uphoff, N., ed. 2002. Agroecological Innovations: valuable scientifi c resources. Rice Today 3: 43. Increasing Food Production with Participatory Sseruwagi, P., W. S. Sserubombwe, J. P. Legg, J. Ndun- Development. London: Earthscan. guru, and J. M. Thresh. 2004. Methods of surveying Uphoff, N., E. C. M. Fernandes, Yuan L.P., Peng J., the incidence and severity of cassava mosaic disease S. Rafaralahy, and J. Rabenandrasana, eds. 2002. and whitefl y vector populations on cassava in Africa: Assessments of the System of Rice Intensifi cation: A review. Virus Research 100: 129-142. Proceedings of an International Conference, Sanya, Stoop, Willem, Norman Uphoff, and Amir Kassam China, April 1-4, 2002. Cornell International Institute (2002). A review of agricultural research issues for Food, Agriculture and Development, Ithaca, NY. raised by the System of Rice Intensifi cation (SRI) Available from: http://ciifad.cornell.edu/sri/ from Madagascar: Opportunities for improving farm- Uphoff, N. 2003. Higher yields with fewer external ing systems for resource-poor farmers. Agricultural inputs? The System of Rice Intensifi cation and po- Systems 71: 249-274. tential contributions to agricultural sustainability. Sunderland, K., and F. Samu. 2000. Effects of agricul- International Journal of Agricultural Sustainability, tural diversifi cation on the abundance, distribution, 1: 38-50. and pest control potential of spiders: A review. Ento- Uphoff, N. 2004. SRI—The System of Rice Intensi- mologia Experimentalis et Applicata 95: 1-13. fi cation: An opportunity for raising productivity in Surridge, C. 2004. Feast or famine? Nature 428, 360- the 21st century. Paper for International Year of Rice 361. conference, FAO, Rome, February 12-13. SWCS. 2002. The Soil and Water Conservation Society. USEPA. 2002. Concentrated Animal Feeding Opera- Available from. http://www.swcs.org. tions Proposed Rule. Available from: http://cfpub. epa.gov/npdes/afo/cafofi nalrule.cfm. Accessed May Tan, Z., T. Hurek, and B. Reinhold-Hurek. 2003. Effect 2004. of N-fertilization, plant genotype and environmental conditions on nifH gene pools in roots of rice. Envi- Van Bueren, E. T. L., P. C. Struik, and E. Jacobsen. ronmental Microbiology 5:1009-1015. 2002. Ecological concepts in organic farming and their consequences for an organic crop ideotype. Tanksley, S. D., and S. R. McCouch. 1997. Seed banks Netherlands Journal of Agricultural Science 50: and molecular maps: Unlocking genetic potentials 1-26. from the wild. Science 277: 1063-1066.

116 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Van Duivenbooden N., M. Pala and C.Studer. 2000. Wolfe, D. 2001. Tales from the Underground: A Natu- Cropping systems and crop complementarity in dry- ral History of the Subterranean World. Cambridge, land agriculture to increase soil water use effi ciency: MA: Perseus. a review. Netherlands Journal of Agricultural Sci- Yadav, R. L., V. K. Singh, B. S. Dwivedi, and A. K. ence 48 (3-4): 213-236 Shukla. 2003. Wheat productivity and N use-ef- van Es, H., and D. Huska. 2000. Conference Proceed- fi ciency as infl uenced by inclusion of cowpea as ings. In: Environmental Management of the Rural a grain legume in a rice-wheat system. Journal of Landscape in Central and Eastern Europe. Podban- Agricultural Science 141: 213-220. ské, Slovakia. Yamoah, C. F., A. Bationo, B. Shapiro, and S. Koala. Wall, D. H., and J. C. Moore. 1999. Interactions under- 2002. Trend and stability analyses of millet yields ground. BioScience 49: 109-117. treated with fertilizer and crop residues in the Sahel. Field Crops Research 75: 53-62. Waisel, Y., A. Eshel, and U. Kafkafi , eds. 2002. Plant Roots: The Hidden Half, 3rd ed. New York:York: Marcel Yanni, Y. G. et al. (2001). The benefi cial plant growth- Dekker. promoting association of Rhizobium leguminosarum bv. trifolii with rice roots. Australian Journal of Plant Wander, M. M., and L. E. Drinkwater. 2000. Fostering Physiology 28: 845-870. soil stewardship through soil quality assessment. Applied Soil Ecology 15: 61-73. Ying, J., S. Peng, Q. He, H. Yang, C. Yang, R. M. Visperas, and K. G. Cassman. 1998. Comparison Wander, M. M., S. J. Traina, B. R. Stinner, and S. E. of high-yield rice in tropical and subtropical envi- Peters. 1994. Organic and conventional management ronments: I: Determinants of grain and dry matter effects of biologically active soil organic matter yields. Field Crops Research 57: 71-84. pools. Soil Science Society of America Journal 58: 1130-1139. Young, A. G. and T. Boyle. 2000. Forest fragmentation. In: A. G. Young and T. Boyle, eds., Forest Conser- Wardle, D. A. 2002. Communities and Ecosystems: vation Genetics: Principles and Practice, 123-134. Linking the Aboveground and Belowground Compo- Collingwood, Australia: CSIRO Publishing and CAB nents. Princeton University Press, Princeton, NJ. International. Wardle, D. A., and K. E. Giller. 1996. The quest for a Yuan, L. P. 2002. A scientist’s perspective on experi- contemporary ecological dimension to soil biology. ence with SRI in China for raising the yields of super Soil Biology and Biochemistry 28: 1549-1554. hybrid rice. In: N. Uphoff, E. C. M. Fernandes, L. P. Weibull, A.-C., Ö. Östman, and A. Granqvist. 2003. Yuan, J. Peng. S. Rafaralahy, and J. Rabenandrasana, Species richness in agroecosystems: The effect of Assessments of the System of Rice Intensifi cation: landscape, habitat and farm management. Biodiver- Proceedings of an International Conference, Sanya, sity and Conservation 12: 1335-1355. China, April 1-4, 2002. Cornell International Insti- Wiersum, K. F. 2004. Forest gardens as an ‘intermedi- tute for Food, Agriculture and Development, Ithaca, ate’ land-use system in the nature-culture continuum: NY, 23-25. Characteristics and future potential. In: P. K. R. Zheng, J., X. Lu, J. Xinlu, and Y. Tang. 2004. The Nair, M. R. Rao, and L. E. Buck, eds., New Vistas System of Rice Intensifi cation (SRI) for getting su- in Agroforestry,y, 123-124. Dordrecht, Netherlands: per-high yields of rice in the Sichuan Basin. Report Kluwer. to the Sichuan Academy of Agricultural Sciences, Wood, D., and J. M. Lenné. 1997. The conservation of Chengdu, China. agrobiodiversity on-farm: Questioning the emerg- Zhu, Y., H. Chen, J. Fan, Y. Wang, Y. Li, J. Chen, J. Fan, ing paradigm. Soil Biodiversity and Conservation S. Yang, L. Hu, H. Leung, T. W. Mew, P. S. Teng, Z. 6: 109-129. Wang, and C. C. Mundt. 2000. Genetic diversity and disease control in rice. Nature 406: 718-722.

Contributions of Agricultural Biodiversity and Natural Ecosystem Processes to Sustainable Agricultural Systems 117 ACKNOWLEDGMENTS

We would like to thank a number of colleagues for providing ideas, insights and scientiﬁ c literature references for putting together this overview of emerging thinking and practice in the agricultural sector: Greg Albrecht, Cornell Nutrient Management SPEAR Program: integrated nutrient management; Barbara Bedford, Department of Natural Resources: landscape diversity, wetland processes and water management; Gary Fick. Department of Crop and Soil Sciences: pasture species diversity and productivity; Peter Hobbs, Department of Crop and Soil Sciences: conservation agriculture; Susan Riha, Department of Earth and Atmospheric Sciences: integrating agricultural intensiﬁ cation and conservation; environmental thresholds; predicting maize/weed competition using climate data; Tammo Steenhuis, Department of Biological and Environmental Engineering: practical methods of achieving sustainable water balances in irrigated areas, hydrological modeling; Harold van Es, Department of Crop and Soil Sciences: soil health; David Wolfe, Department of Horticulture: soil health, IPM, cropping system diversity.

118 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations Chapter 7 Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives

7.1. Introduction wildlife within agricultural landscapes2 involve the management of biophysical and social processes at The deliberate integration of wild biodiversity con- multiple geographic and temporal scales.3 Planning, servation objectives with agriculture and livelihood monitoring and evaluation of landscape-level initiatives improvement goals distinguishes ecoagriculture from needs to account for the multiple scales at which the other land-use frameworks. Informed by the principle various activities and processes occur. Ecoagriculture 1 of ecological integrity, ecoagriculture requires coor- strategists need to fi nd ways to link the scale of man- dinated farm- and landscape-level strategies that ac- agement with the scale of the environmental issue that commodate the requirements of diverse assemblages is being addressed. It will be important therefore, to of plant, animal and human populations. It will be ground monitoring and evaluation schemes in “hierar- impossible to accurately predict the effects of all the chy theory,” which maintains that to understand pro- numerous interactions of interest within this complex cesses occurring at any particular scale, insights from sphere of ecological and human activity, in advance of both the level above and the level below are required launching an ecoagriculture initiative in a particular (Campbell et al. 2001). In both planning and evaluation, setting. Therefore, the implementation of ecoagricul- information will be needed about the interactions that ture strategies will need to be adaptive, improved by occur within and among systems that are functioning learning through iterative planning, monitoring, and at these different scales. evaluation activities similar to what needs to be done in ecosystem management more generally (Walters 2 and Holing 1990). To respond to this reality, it will be The strategies associated with making space for wildlife within agricultural landscapes include: (i) creating biodi- important to design ecoagriculture initiatives so that versity reserves that benefi t local farming communities, (ii) they can assimilate feedback and new information developing habitat networks in nonfarmed areas, and (iii) about the dynamic social and ecological conditions of reducing land conversion by increasing farm productivity the landscapes in which they are established. (McNeely and Scherr 2003). 3 The three general ecoagriculture strategies that Mc- For example, the different subsystems associated with modifi cations of soil, water and vegetative resources intended to Neely and Scherr (2003) propose to make space for enhance the habitat value of farmlands operate at different spatial and temporal scales. Soil management usually occurs 1 Ecological integrity “includes maintaining viable popula- at a smaller scale than that occupied by the production system tions of native species, representation of ecosystem types of which it is a part. Yet the biophysical processes linked across their natural range of variation, maintaining ecologi- with soil health change faster than production-system-level cal processes, management over the long term, and accom- processes (Izac 2003). modating human use within these constraints” (Grumbine 1996).

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 119 This chapter addresses two central questions related to sionals from different disciplines to seek integrative ap- the potential for adaptive management of ecoagriculture proaches for understanding how and why landscapes are strategies and initiatives: dynamic and ever-changing (Holling 1998). Adaptive management is a prominent translation of this underunder-- i) What is known about landscape-level planning, standing into practice. This approach frames problems monitoring and evaluation that can be applied to in multidimensional contexts, extends analyses of ecoagriculture to help translate the principles of the natural resource systems beyond political boundaries, framework into practice; and and examines interactions at multiple scales and across ii) How can the scientifi c knowledge that will be called scales. Biological and social data, robust monitoring upon to support the design and development of ef- systems, and information feedback loops are built into fective ecoagriculture practices be integrated into this approach. Adaptive management treats human ac- landscape planning and management frameworks? tivity as part of the system and recognizes that learning Our aim here is to provide a basis for assessing current is a social process. Accordingly, it fosters partnerships capacities to analyze and manage the highly integrative and provides for negotiation to meet both conservation systems as required for successful ecoagriculture strate- and human needs. gies, and for determining additional knowledge needs Numerous factors stimulated the shift toward a land- and corresponding research priorities in this arena. scape-scale, coupled framework for natural resource The chapter considers concepts and principles that are management that blends biological and human dimen- suitable for guiding multi-objective land use planning, sions. Appreciation for the hierarchical, i.e., nested, monitoring and evaluation processes. Within this realm, interactions within a landscape underscores the need it identifi es frameworks, models and methods that have to preserve ecosystems so as to maintain valuable and demonstrated some promise for integrating planning, vulnerable ecological processes (Izac 2003). The dy- science and management to promote ecologically and namic nature of ecosystems has made it evident also economically sustainable land use, including initiatives that complete information regarding any system will that link monitoring and evaluation with decision-mak- be unachievable (Holling 1998). However, recognizing. ing the various subsystems within any ecosystem, and We begin with a brief overview of ecosystem-based ap- the myriad of interactions among them, can facilitate proaches and discuss their relevance to ecoagriculture. the examination of its functions and the identifi cation We then review frameworks for planning, monitoring of factors infl uencing change. This understanding was and evaluation that are conceptually consistent with further reinforced by evidence of interactions among the these approaches.4 We present distinguishing char- social, political, economic, and ecological elements of acteristics of selected methods that are applicable at the system (Vosti et al. 2002; Jagger and Pender 2003); different spatial and temporal scales, and how they the growing recognition that biophysical systems ex- may be used for integrative, adaptive collaborative tend beyond political and property boundaries (Loomis management. Finally, we comment upon the challenges 2002); and the orientation of new initiatives toward and areas needing further research in planning, monitor- integrated objectives (Izac and Sanchez 1999). ing, evaluation to support the effective management of Ecosystem-based approaches can certainly be applied integrative systems. to the sustainable management of agricultural lands. Research has revealed how natural and human elements 7.2. Ecosystem-Based Approaches of a landscape infl uence agricultural productivity (e.g., Recognizing the importance of interactions between van Noordwijk 2002). Growing evidence of the impact different elements of a land-use system has led profes- of agriculture on biodiversity, and vice versa, and of the importance of agrobiodiversity in ecosystem function- 4 Some of the models discussed in this chapter have been ing has underscored the realization that agriculture does highlighted in Chapter 5 already for their value in under- not occur in isolation from its surroundings (Gleismann standing tradeoffs between or among the three pillars of 1998; Brookfi eld 2001; Jackson and Jackson 2002). ecoagriculture.

120 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Therefore, accounting for interactions among the dif- with varying values and frames of reference (Schelhas ferent elements within agricultural systems and between et al. 2001: xxv). Furthermore, the trend toward de- agricultural and non-agricultural systems is crucial to centralization of decision-making for natural resource ensuring sustainable productivity. management (Agrawal 2002; Ribot 2002) and toward more diverse governance systems will expand the 7.2.1 Adaptive Management range of claims on resources and magnify the power Adaptive management uses policy interventions as dynamics associated with their control. Both latent and “treatments” to test hypotheses about the effects they evident confl icts are likely to ensue, underscoring the will have on ecosystem function and performance. need to facilitate joint understandings among actors and Adaptive management emerged once it was accepted to negotiate concurrence on how best to proceed. This that knowledge of complex systems will always and requires a greater degree and variety of participation by invariably be incomplete. The approach is more than stakeholders than is usually provided for (and achieved) informed “trial-and-error” (Lee 1999). It uses the best in conventional land-use planning and management available knowledge of the context and the system to processes. generate “a risk-averse, ‘best guess’ management strat- Adaptive collaborative management (ACM) expands egy” (Resilience Alliance 2004). As new information is upon adaptive management to satisfy these conditions generated, the “best guess” strategy gets revised. Adap- (Buck et al. 2001; Salafsky et al. 2001). ACM involves tive management, therefore, emphasizes continuous multi-stakeholder learning processes that do not follow learning and improving stakeholders’ understanding of a pre-determined course. It engages stakeholders in a complex system as it evolves (Holling 1998; Salafsky social learning through participation; considers mis- et al. 2001). The adaptive approach gives new meanings takes, errors and/or failures to be sources of lessons to conservation strategies, making them “bioregional and information; and uses different methods to generate in scope and collaborative in governance, as well as knowledge that “keeps pace with ecosystem change adaptive in managerial perspective” (Lee 1999). resulting naturally or from expanding human activity” Conservation organizations agree that any framework (Schelhas et al. 2001: xx). ACM identifi es strategies for for designing effective initiatives requires planning of conserving biodiversity based on scientifi c knowledge both the actions and the monitoring and evaluation, in an adaptive management framework with participa- then implementing both and analyzing data from these tory decision-making through a range of collaborative activities, using the results to revise and improve the processes (Schelhas et al. 2001). initiative as experience unfolds (Conservation Measures 7.3 Planning and Implementation Partnership 2003). It may be expected, for the sake of Frameworks Consistent with Ecosystem- utilizing specialized expertise, that these responsibili- Based Approaches ties will be delegated to different agencies or actors. However, in the normal course of events, this will Ideal approaches to the planning, monitoring, and not produce integrated activities or much cumulative evaluating of ecoagriculture-based land-use systems and corrected impact. Accordingly, the integration of would link, in an iterative manner, analytical models these roles is as important as carrying out the respec- with planning and implementation frameworks that are tive activities themselves if management is to be truly consistent with ecosystem-based goals and approaches. adaptive. Conceptually, these approaches would use the types of analytical models discussed in Section 5.3.1 to iden- 7.2.2. Adaptive Collaborative Management tify trade-offs among wild biodiversity conservation, Ecoagriculture thinking derives from ecosystem agricultural productivity, and rural livelihoods. This in- management principles. One can anticipate that the formation would feed into implementation frameworks application of these principles will transpire on large and models for planning, monitoring and evaluating human-occupied landscapes that include stakeholders ecoagriculture land use, such as those described in this section. New fi ndings from the planning, monitoring,

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 121 and evaluation processes would, in turn, be used to up- interactions within and among different ecosystems date the analytical models and fuel successive iterations that together constitute a landscape, therefore, requires of analysis, planning, monitoring and evaluation. understanding the fl owsows of biological and non-biological resources across it. Planning and implementation frameworks that are consistent with ecosystem-based approaches include Spatially-explicit models use simulation and visualiza- spatially-explicit planning frameworks (Dolman et al. tion tools to capture how changes in policy, regulations, 2001; Hawkins and Selman 2002); frameworks that in- demographics and/or land management practices alter tegrate planning, monitoring and evaluation (Campbell characteristics of that landscape. For example, spatially- et al. 2001; Conservation Measures Partnership 2003); explicit modeling is central to the Gap Analysis Pro- and systems-based monitoring and evaluation (Bellamy gram (GAP), a nationwide effort in the United States to et al. 2001). Each of these frameworks has an analytical comprehensively inventory and computerize the kinds component embedded within it. The analytical models and geographic distributions of species of plants and build on different combinations of the principles dis- animals that contribute to biodiversity (Smith 1998). cussed in Chapters 3, 4 and 5 of this report. This section Conservation scientists, economists and planners are reviews characteristics of these frameworks that pertain using information from GAP to identify biodiversity most directly to the particular challenges and opportuni- hotspots and to assess the impacts of specifi c policies. ties of ecoagricultural land-use approaches. The overlap In Europe, approaches that integrate landscape ecol- among the different frameworks refl ects a convergence ogy measures, such as landscape stabilization and new in thinking about complex natural resource management landscape creation based on focal species, are being issues. Annex 7A presents concrete applications of these used (Hawkins and Selman 2002). How to integrate a frameworks in summary form. participatory approach into these planning procedures is still evolving (Ford and McConnell 2001). 7.3.1 Spatially-Explicit Frameworks Quantitative simulation models such as those discussed Spatially-explicit research and planning frameworks are in Section 5.3.1 are key components of spatially-explicit crucial for conserving wild biodiversity within produc- planning frameworks. These models defi ne relation- tive landscapes. These frameworks use biological and ships among system components and can represent social science concepts and methods to model and map relevant changes across large geographic scales (and in the characteristics of a specifi c system, exploring the some cases, temporal scales). Simulation models can be impacts of changes at the macro- and micro-scale. They used to optimize land use as attempted in a watershed also engage stakeholders in designing land management in the Lake Erhai basin of China (Wang et al. 2004). strategies, in assessing the suitability of specifi c poli- Or they can be used to develop scenarios and engage cies, and/or validating models. stakeholders in designing, validating and/or utilizing the A landscape is a heterogeneous area within which outputs from models. In Sweden, for example, simula- the units of analysis, evaluation and action can be tion models were used to identify optimal strategies for denominated as “patches,” basic units of land use that minimizing pollution of water bodies from agricultural are relatively homogeneous (Liu and Taylor 2002). Its practices (Arheimer et al. 2004). Information generated boundaries are set by some agreement among relevant from the various steps in spatial planning can be used actors that the units within it have enough interaction subsequently to refi ne the simulation model (Ellis et and interdependence that it is both illuminating and al. 2000). practical to consider the patches within this particular Stakeholder participation in spatially-explicit planning area concurrently and as a set rather than in associa- varies as to timing and method. Decision-makers can tion with some other set of patches. The functionality help to defi ne the starting conditions (Arheimer et al. of a landscape is maintained through fl ows of matter, 2004) or stakeholders can be involved to specify goals energy, and organisms across patches through processes at the outset of the process, also shaping the set of ac- such as migration and dispersal. Modeling the complex ceptable combinations of policy changes and new man-

122 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations agement practices. Trade-off assessments have applied out tradeoffs between them; (iv) the characteristics of the latter approach (see Crissman et al. 2001, in Annex the research site; and (v) the challenge of maintaining 5). Stakeholders can be involved, also, in validating integration with numerous components and interactions the accuracy of the spatial model as exemplifi ed by (Campbell et al. 2001). an effort to monitor land- cover change in Amazonia Bellamy et al. (2001) characterize a monitoring and (Sydenstricker-Neto et al. 2004). Or they can select evaluation system-based approach used in Australia management options based on simulated outputs. In a that accounts for each of these factors. The approach spatially-explicit planning framework for sustainable evaluates natural resource management initiatives as rural livelihoods and land uses in Uganda, stakehold- a system that links the objectives and rationale of the ers interacted with researchers to defi ne and evaluate program or policy with performance on the ground. The alternate rural development pathways based on prior systems approach has to: identifi cation of feasible intervention opportunities (Bolwig et al. 2003). Regardless of when stakehold- • assess public and private investment in integrated ers become involved, effective facilitation by persons approaches to natural resource management; responsible for and skilled in encouraging dialogue • identify social, economic, institutional, environmen- and giving it shared meaning is critical to getting suc- tal and technological factors that infl uence natural cessful and encompassing participation. Facilitators resource management initiatives; must overcome obstructions to open and constructive • interaction among the different actors. develop suitable performance criteria for assessing the potential impacts and infl uences of a resource Ecoagriculture strategies for creating wildlife habitat in management approach on institutional arrangements agricultural landscapes will benefi t from frameworks and society more broadly; that encompass changes at large scales across different • political and property boundaries. Spatially-explicit identify the outcomes and expectations of an inte- planning models can create the needed interface be- grated resource management initiative; and tween research and practice while being accessible to • establish guidelines and techniques for identifying stakeholders and using their input in the design and/or progress toward agreed objectives and outcomes refi nement of models. For ecoagricultural initiatives, (Bellamy et al. 2001). the participatory component will need to be tailored Bellamy et al. (2001) have developed the approach to the temporal and geographical scales of particular to evaluate a variety of natural resource management initiatives, as well as to the unique and variable char- initiatives including a community-based Integrated acteristics of the political system and stakeholders Catchment Management process; a community-based involved. resource information center; and a decision-support 7.3.2 Systems-Based Monitoring and Evaluation system for sustainable grazing management. These Framework evaluations have been implemented and coordinated by a research organization together with other stake- Monitoring and evaluation of biodiversity conserva- holders. The approach of the Bellamy group appears tion initiatives should begin at the start of a program particularly relevant to the challenges of ecoagriculture and generate feedback into planning and implementa- since it can be implemented at any scale, it engages tion mechanisms (Conservation Measures Partnership stakeholders with diverse capacities and interests, and 2003). When integrating conservation and development it is founded on rigorous scientifi c methodology. at multiple scales, processes of monitoring and evaluation must accommodate: (i) multiple scales of interac- 7.3.3 Frameworks that Integrate Planning, tions and responses; (ii) the non-linearity and time lags Monitoring and Evaluation in complex systems; (iii) multiple stakeholders’ often Appreciating the need in natural resource management contrasting objectives that make it diffi cult to identify to blend hard and soft sciences (e.g., positivist and con- common research and management aims and to sort

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 123 structivist approaches) has led to an experimental-sci- target and further that the underlying conceptual model ence approach to planning, monitoring and evaluation. employed be able to reveal direct threats, indirect threats Monitoring and evaluation are integrated with planning and opportunities for meeting the goal. Scientists’ and through the use of information gained through moni- stakeholders’ interests and perspectives will also inform toring to refi ne or revise processes and objectives of the targets and objectives.5 An ideal approach would planning. Adaptive management and ACM approaches, have an in-built “learning program” that promotes discussed above, are outcomes of this direction in systematic learning from the actions undertaken to thinking. The implementation of such approaches is determine how and why they contribute to specifi c challenging, however, and few efforts have completely conservation and development objectives (Salafsky and embodied the key principles (Lee 1999). Margoluis 1999). The program is an iterative process that engenders continuous learning and improvement The Conservation Measures Partnership (2003) is test- of outcomes. ing a framework for translating adaptive management into practice in the arena of biodiversity conservation. Integrated planning, monitoring and evaluation frame- Annex 7-B expands on this and other partnership and works require new processes for researching complex network initiatives. The framework is applicable at any and coupled natural and human systems. Integrated scale and to any set of actions, ranging from a single natural resource management (INRM) is an example project or program to coordination among several of innovative efforts designed to embed research in a programs and regional initiatives. The elements of the multi-stakeholder learning process. INRM is oriented framework include: toward augmenting social, physical, human, natural and fi nancial capital (CGIAR 2002). It involves orienting • defi ning objectives that are relevant to a given con- the objectives of research toward these ends, adding text; weight to participatory approaches for implementing • planning both actions and their monitoring and evalu- research, using guiding principles for research that ation; broaden the temporal and spatial scales of analysis, and • implementing the actions and their monitoring and using a variety of analytical tools (Sayer and Campbell evaluation; 2001). • analyzing available data to evaluate the effectiveness INRM brings together different forms of knowledge of the activities; to identify the key problems from the relevant spatial and temporal scales and the underlying forces affecting • using these results to modify actions to maximize them (Izac and Sanchez 1999). In INRM, interdisciplin- desired impacts; ary research identifi es a range of economically viable • communicating the results to key external and inter- land-use practices that rehabilitate and/or strengthen nal audiences; and ecosystem function in agroecosystems. The trade-offs associated with the different land-use systems are • iterating the process with the objective of improv- reviewed, using inputs from stakeholders. It also uses ing the actions (Conservation Measures Partnership research fi ndings to examine trade-offs between per- 2003). spectives and interests of different stakeholders and to Principles that guide the whole framework include shape partnerships (CGIAR 2002). stakeholder involvement, a clearly defi ned timeline, and The Alternatives to Slash and Burn (ASB) program is allocation of adequate fi nancial resources and human an example of INRM research (see Annex 7-B). This capacity for each element. Integrated planning, monitoring and evaluation frame- 5 Often there are multiple targets related to both conserving works use a variety of models to guide the selection of biodiversity and improving local welfare. In these cases, actions and analyses. Salafsky et al. (2002) maintain Salafsky et al. (2002) recommend developing separate con- that the approach should be tailored to the conservation ceptual models for each target.

124 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations program uses interdisciplinary research to evaluate the tially-based planning framework for sustainable rural performance of different land uses according to global livelihoods and land use also incorporates monitoring environmental measures, agronomic sustainability, and evaluation and creates an iterative process by which smallholders’ socioeconomic interests, and policy and strategies for poverty alleviation and sustainable natu- institutional measures. For example, in the Brazilian ral-resource-based development are adapted (Bolwig Amazon, ASB examined the adoption of four types of et al. 2003). intensifi cation and their economic and environmental impacts using a farm-level bioeconomic linear program- 7.4 Planning, Monitoring, and Evaluating ming model. The study revealed that there is a trade-off Ecoagriculture between farm income and forest preserved as a result of Planning, monitoring and evaluation of ecoagricultural the intensifi cation of land uses on the cleared land. Of land use need to embody a constructivist perspective, the four intensifi cation types, intensifi cation on forested being sensitive to cross-scale impacts of activities and land, i.e., low-impact forest management, slowed the interconnected with a systems perspective. Methods deforestation rate, but did not stop it unless timber prices will need to be tailored always to the ecoagriculture exceeded a certain value (Carpentier et al. 2000). strategy employed. Depending on the strategy of any Integrated natural resource management embodies the particular initiative, different actors need to be brought principles of ACM by ensuring that: (i) management together during various stages of the planning, monitor- approaches are adaptive; (ii) INRM research is trans- ing and evaluation to collaborate and coordinate actions, ferred into practice; and (iii) the approach provides for, and to exchange information and knowledge regarding and is based upon, negotiation among all stakeholders. elements of the system. In addition, feedback from these INRM creates an interface between science and practice processes should inform learning and adaptation of the through its use of stakeholder knowledge in defi ning approach being used. The how and when of planning, the main problems and constraints in the system. As- monitoring and evaluation are important for learning sociated with this process is empowerment of relevant about systems and improving practice (Douthwaite et al. stakeholders and addressing their confl icting interests 2002). The use of appropriate tools and techniques will (CGIAR 2002). assist in gaining a greater understanding of the system, Approaches that integrate planning, monitoring and more learning, and further improvement of appropriate evaluation are being encouraged in community-based strategies (Saterson et al. 1999). natural resource management (Baron 1998; Margoluis 7.4.1 Important Considerations and Salafsky 1998; Johnson 1999).6 Other efforts to integrate planning, monitoring and evaluation include 7.4.1.1.4.1.1 Scale the ecoregion-based planning process in Madagascar, which uses a collaborative process for identifying the Considerations of scale should guide the planning, mon- actions and strategies for integrating biodiversity con- itoring and evaluation of ecoagriculture initiatives even servation and with livelihood enhancement (Cowles though it makes these activities more complex. Three et al. 2001).7 In Uganda, the implementation of a spa- relevant scale considerations have been distinguished: scaling-out, scaling-up, and spatial scaling-up.

6 Community-based natural resource management is the • Scaling-out involvesinvolves spreadingspreading innovationinnovation fromfrom management of natural resources under a detailed plan farmer-to-farmer, community-to-community, and developed and agreed to by all concerned stakeholders. within the same stakeholder groups. Communities are the primary implementers of the plan, • Scaling-up results when there is institutional expan- with assistance and monitoring by external organizations (USAID 2000, as cited in CBNRM Net http://www. cbnrm. sion from grassroots organizations to involve policy net/resources/terminology/cbnrm.html). makers, donors, development institutions, and other 7 The pilot phase of this effort was too brief to implement stakeholders able to create an enabling environment monitoring and evaluation. for change.

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 125 • Spatial scaling-up involves the widening of scale the system through judicious use of technologies and of operation from, for example, experimental plot landscape planning (Scherr 2004, pers. comm.). to farm and to watershed (Douthwaite et al. 2003). 7.4.1.3 Criteria and Indicators Scholars and practitioners of ecoagriculture need to be careful not to mismatch the scale of assessment with the The selection of suitable criteria and indicators for scale of management as these are not always ideally the monitoring ecoagriculture will be challenging because same, and to select and tailor tools accordingly (Cash there are multiple stakeholders, and they have often and Moser 2000). diverging interests. A participatory assessment process for identifying the indicators to use and what would be 7.4.1.2 Characterization of Landscapes considered a favorable outcome can overcome some of these obstacles of differing interests (Campbell et al. Knowledge of the condition of agroecosystems and of 2001). Monitoring and evaluation approaches should the landscapes in which they are embedded is essential employ techniques that create conducive conditions for for the effective planning and evaluating of ecoagricul- stakeholders to engage in the defi nition of criteria and ture-based initiatives. As a fi rst step, the location and selection of indicators. extent of current land uses need to be determined (Wood et al. 2000). A realistic characterization and classifi ca- The criteria used will infl uence monitoring and evalu- tion of landscapes is needed to facilitate the planning ation fi ndings, which in turn inform the selection of of corridors and other habitat confi gurations for wild geographic areas for intervention, adaptation of in- biodiversity and for the enhancement of ecosystem terventions, and the associated learning (Perez and services that are important to local communities and Tschinkel 2003). The selection of suitable criteria and wider areas. indicators should be practical and scientifi cally justifi ed. Criteria used by land managers and scientists may differ Spatial information, discussed in Section 7.4.2.4, because of their different knowledge and understanding increasingly is used to characterize land cover. Clas- of the system (Oba and Kotila 2001). In such cases, it sification and interpretation of spatial information is important to facilitate the formulation of criteria and makes clear the extent of different land uses including indicators jointly through the use of suitable decision- agriculture. It is common to fi nd differences, however, support techniques. in the extent of agricultural land use that is predicted by common global land-cover databases and higher- 8 resolution spatial information (Wood et al. 2000). These Sebastian et al. (2000) have refi ned the characterization discrepancies arise from under-reporting of agricultural of landscapes to recognize the extent of agriculture in landscapes where it is not dominant, in part by distinguishing area by satellite remote-sensing, due in part to this among three agricultural intensities. In this characterization, technology’s inability to discriminate certain agricul- the agricultural-extent descriptions include: (i) agriculture- tural crops and pastures that appear similar to natural dominated landscapes, with more than 60 percent under forests, woodlands and grasslands, and partly due to its agriculture; (ii) agriculture and natural vegetation mosaic, limitations in identifying agriculture within landscapes with 40-60 percent land under agriculture; (iii) landscapes where this is less than 30 percent of the land cover. with agriculture that is not a dominant class, with 30-40 percent under agriculture; and (iv) other vegetated land Efforts to plan and evaluate ecoagricultural land-use cover, with less than 30 percent under agriculture. McNeely initiatives would benefi t from more systematic char- and Scherr (2003) use a similar typology in their discussions acterization of landscape types for this purpose. The of ecoagriculture. The typology being used distinguishes the following landscape types: (i) urban/agriculture dominated typology should capture both spatial and seasonal landscapes, with over 80 percent of land under crops or variations in vegetation cover and represent agriculture infrastructure; (ii) agriculture-dominated landscapes, with realistically, where it is not the dominant land cover.8 60-80 percent of land under crops; (iii) agriculture-nature The latter is important because landscapes in which mosaics, with 30-60 percent of land under crops; and (iv) cropped area is not dominant can make greater con- nature-dominated mosaics, with less than 30 percent of land tributions to maintaining diverse wild biodiversity in under crops (Scherr 2004, pers. comm.).

126 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations 7.4.1.4 Collaboration and Learning 7.4.2.1 Scenario-Based Methods

Translating ecoagriculture visions into reality will de- Developing implementable visions for ecoagriculture in pend on sustained and effective collaboration between a particular landscape setting will require that multiple scientists of different disciplines, and among scholars, stakeholders be involved in strategizing. Scenario meth- donors, policy makers, practitioners, land managers, ods will be useful in facilitating this process. Scenario and other key stakeholders. Stimulating collaboration, methods are a category of techniques associated with however, requires building trust, partnerships, and such “visioning” processes (Wollenberg et al. 2001). networks that minimize associated transaction costs They can be used to help stimulate creative ways of and elicit collective action. Organizational support thinking about alternative futures. These methods are and suitable institutional arrangements for information suitable for planning when uncertainty and complexity exchange can enhance the creation of horizontal and are high. Future visioning can be used to reveal differ- vertical partnerships and networks. These in turn can ent perspectives regarding a situation and/or divergent facilitate the transfer of information, coordination of interests. It can also provide a starting place for negotia- research and actions, and learning. tion among preferences and for aggregation of different Negotiation will be important to reach collaborative ideas. Future visioning can also foster learning if it is arrangements, to address trade-offs associated with situated within a decision-making context. ecoagriculture, and to adapt strategies. Stakeholder To be used most effectively, scenario-based approaches groups engaged in negotiation will have different un- should have clearly defi ned purposes and involve facili- derstandings of how their and others’ actions infl uence tators who have extensively researched the context and outcomes and, accordingly, they will start with varying understand the feasibility of potential outcomes. The interpretations. Research and development organiza- approach can be applied at any scale and in different tions can provide tools to help facilitate stakeholder contexts. However, the scale, context, and objectives consultation and negotiation processes (van Noordwijk will infl uence the specifi c selection of tools to be used. et al. 2001). The use of multiple tools and approaches At a large landscape scale, for example, Geographic to generate information and data in a manner that stake- Information Systems (GIS) may be used to simulate holders consider reliable can contribute to dialogue and conception and evaluations of different scenarios, as learning about the system and identifi cation of suitable was the case in Oxfordshire, where GIS simulations actions. were used to elicit farmers’ perspectives on whole- landscape planning (Dolman et al. 2001). 7.4.2 Approaches and Tools 7.4.2.2 Decision-Support Systems Effective approaches and tools for guiding the development of ecoagriculture land use will integrate Decision-support techniques can facilitate the selection disciplinary and multidisciplinary variables, highlight and prioritization among multiple planning objectives as key issues, create an interface between science and well as evaluation criteria. These techniques are based practice, and facilitate the design of intervention strate- on systematic pair-wise comparisons and/or ranking of gies. Techniques associated with confl ict management, objectives and criteria by which land-use alternatives participatory action research, participatory research ap- are evaluated. Examples of decision-support techniques praisal, facilitation, and negotiation will be invaluable. include: analytic hierarchy processes and voting theory, A few such approaches and tools are identifi ed below, multiple attribute utility approaches, and SWOT meth- that can be useful in addressing the challenges of scale, ods (assessing strengths, weaknesses, opportunities and uncertainty, learning, and coordination among multiple threats) integrated with an analytic hierarchy process stakeholders. Annex 7-A presents additional informa- (Pesonen et al. 2001; Laukkanen et al. 2002; Prato tion on these and related approaches and tools. 2003). In Finland, the Finnish Forest and Park Service (FFPS), which employs now a participatory planning process, tested a decision-support approach to establish

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 127 priorities among different management scenarios in the ommendations regarding land use and natural resource state forests of Western Finland. An Analytical Hierar- management. chy Process in Strengths Weaknesses, Opportunities, Geographic Information Systems have become impor- and Threats Analysis (A’WOT) was used to evaluate tant mechanisms for storing, manipulating, analyzing four alternative management strategies that met differ- and integrating both spatial and non-spatial data (Liu ent land-use requirements in this area. The strategies and Taylor 2002). They are becoming widely used allocated varying amounts of land among three differ- for planning purposes and increasingly in collabora- ent uses, recreation, habitat protection, and production. tion with land managers (Ford and McConnell 2001; Using A’WOT, the strategy that allocated the most land Brown et al. 2002; Sedogo and Groten 2002; Parisi et to recreation received highest priority followed by the al. 2003; Sydenstricker-Neto et al. 2004). The Florida protection strategy (Pesonen et al. 2001). Agroforestry Decision-Support System is an example of Simple decision-support tools are helpful for aggregat- a spatially-explicit planning tool. It informs landowners ing the indicators of multiple monitoring criteria in and extension agents about the potentials of agroforestry an integrative manner. Possible mechanisms include for a particular area with certain features and guides creating simple additive indices, deriving compound the selection of appropriate tree or shrub species to be variables using principle component analysis, or using used in given local conditions (Ellis et al. 2000). The canonical correlations to combine indicators across Threat Identifi cation Model is another spatial tool for scales (Salafsky and Margoluis 1999; Campbell et al. assessing agricultural land management sustainability 2001). For example, an additive index of capital asset at suitable land unit scales (Smith et al. 2000). indicators (created by adding together the values for TAMARIN is an example of a spatially-explicit plan- natural capital, physical capital and fi nancial capital) ning tool used to evaluate and compare economic can be used to compare three different land management compensation mechanisms for achieving specifi c biodi- scenarios and determine which one generates the great- versity conservation goals by examining the landscape est asset value (Campbell et al. 2001). Radar diagrams confi gurations resulting from a compensation scheme and two-dimensional plots can also be used to visualize (Stoms et al. 2004).9 This tool could be particularly changes in indicators (Campbell et al. 2001). The selec- useful in formulating and monitoring ecoagriculture tion of an appropriate mechanism should complement strategies that develop habitat networks in non-farmed the objective for the monitoring and evaluation and the areas and create biodiversity reserves. TAMARIN cur- choices of data. rently is being piloted in the Atlantic rainforest of Brazil 7.4.2.4 Spatial Information Generation with government agencies, donors and non-government stakeholders (Stoms et al. 2004). Remote-sensing tools, including satellite imagery and aerial photography, allow researchers or managers to 7.4.2.5. Community-Based Data Collection collect spatially detailed information without direct Ecoagriculture strategies for making more space avail- physical contact (Liu and Taylor 2002). Satellite imag- able for wildlife within agricultural landscapes require ery is increasingly becoming affordable, and the resolu- information about patterns of land use and interaction tion of images is improving greatly. Aerial imagery is among elements in the landscape (McNeely and Scherr also increasingly more affordable, and in some cases 2003). Planning, monitoring and evaluating these strate- resolution is suffi cient to eliminate the need for valida- gies requires methods that encompass research fi ndings tion of information through fi eld-based data collection across large geographic scales and can manage multiple (Kadyszewski 2004, pers. comm.). The accuracy of observations. In addition to remote-sensing, fi eld-based remote-sensing data is still variable and depends on mechanisms for data collection are crucial to validate spectral, spatial, temporal, and radiometric resolution spatial information and monitor biological changes at (Degloria 2004, pers. comm.). Inaccurate interpretation a large geographic and temporal scale. of remote-sensing data can result in inappropriate rec- 9 This model is reviewed in Annex 5.

128 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations Citizen-based data collection is becoming a more agement composed of farmers, local government and widely accepted approach for obtaining social, bio- pest management agency representatives. Information physical, economic, and ecological information that regarding ideal management was elicited from these can be used for scientifi c studies, policy analysis and actors and other stakeholders. This and other sources advocacy (Fleming and Henkel 2001) Citizen-based of information are managed via the Internet in a user- data collection, or ‘citizen science’ as it is often called, friendly format that allows for addition and extraction is widely used in ornithological research and ecological of new and relevant information (Allen et al. 2001). monitoring (Alliance for Chesapeake Bay 2004; Cornell Lab of Ornithology 2004). Citizen science involves 7.5. Challenges and Future Directions building community capacity to undertake monitoring, Efforts to successfully implement landscape-level through a sequence of activity: visions for ecoagriculture will require deliberate goal- • enabling citizens to understand the measures that setting and strategy formulation. Integrated models of represent the health of their local environment; the system based on empirical evidence of relationships between biological and social dimensions, relating • training individuals and groups interested in moni- resource availability to stakeholder interests and ca- toring through a step-by-step process from learning pacities, should inform goal-setting. Strategy selection what to monitor to how to do it; should be based on knowledge about the main threats • entering data in a distributed network and sharing to achieving the goals that is obtained from predict- them with others; and ing, monitoring and evaluating the effectiveness of a • analyzing the data and taking appropriate next steps particular strategy for mitigating threats to biodiversity (Wildlife Habitat Canada 2004). and livelihood objectives. Citizen science can be used to obtain data on those vari- We have described in this and previous chapters, various ables for which data collection techniques are simple, tools and approaches that may be used and/or modi- standardized, transferable, and inexpensive, and which fi ed to plan, monitor and evaluate efforts to achieve are rewarding to individuals. The approach can be used the multiple objectives of ecoagriculture in a certain to collect fi eld-level data across a wide geographic area setting. In bringing together information on different and for an extended period of time. Citizen science ap- elements of the system, the methods should account for proaches have been used in developing a biodiversity the importance of scale. Scale should be treated in an it- registrar in India and in monitoring water quality and erative manner, leading to nested processes of planning, fi sh populations in the Northwest United States (Gadgil monitoring and evaluation that rely on scale-sensitive et al. 2000; Alliance for Chesapeake Bay 2004). information about activities and their effects.

7.4.2.6 Integrative Knowledge Management 7.5.1 Expected Difficulties

Monitoring and evaluating ecoagriculture practices Adaptive, collaborative planning, monitoring, evalua- will benefi t from knowledge-management techniques tion, and management, while theoretically grounded, are that facilitate relevant learning from these activities. diffi cult to implement. Lack of conducive institutional Integrated System of Knowledge Management (ISKM) and policy contexts for adaptive management, including is an example of such an approach. ISKM uses work- divergent interpretations of its meaning and confl icting shops, interviews, visual representation of ideas (with power differentials among prospective actors, are two software such as VENSIM) and community dialogue key problems. In addition, incompatibilities between to facilitate communication among stakeholders and to project-funding timeframes and the time required for stimulate collaborative planning (Bosch et al. 2003). institutionalizing an adaptive process, limited funding Allen et al. (2001) apply this approach to bovine Tb for applied interdisciplinary research, and data-inten- vector management in North Canterbury, New Zealand. sive models, all create obstacles. The authors worked with an advisory group in pest man-

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 129 In ecoagriculture, the issue of scale generates additional other stakeholders. The last decade has witnessed the challenges. Transferring information from one scale to formation of new partnerships and network-based another is not a linear process. Whether or not a process initiatives that are focused on integrating science and at one scale occurs at another scale depends on how the practice with a multidisciplinary orientation. Two associated biophysical and social sub-processes change broad categories of partnerships can be distinguished: at the new spatial scale (van Noordwijk et al. 2001). (i) organizational partnerships and (ii) information Similarly, rules or relationships that are applicable at dissemination, partnership-building and coordination one scale may not transcend scales, and approaches networks. Organizational partnerships are composed “successful” at one scale may have a neutral or nega- of a well-defi ned group of non-governmental, gov- tive impact at another scale (Lovell et al. 2002). Van ernmental and/or research organizations that work Noordwijk et al. (2001) have provided the example collaboratively to address specifi c issues. An example of erosion at the plot level resulting in impoverished is the recently formed Conservation Measures Partner- soils in one place but enriching them in another, with ship which is a joint venture of conservation NGOs and relatively little soil actually being redistributed on a other collaborators who have come together to work on larger scale. For biodiversity, scaling-up is complicated designing, managing and measuring impacts of their because taxonomic or genetic diversity at any scale conservation-oriented activities. depends both on richness at a smaller scale and on the In contrast, information dissemination, partnership- level of similarity between units at the same scale (van building and coordination networks are created around Noordwijk 2001). a broad area of concern. Networks are open to all A consequence of this scaling challenge is that stake- interested parties, tend to be internet-based, provide a holders have available limited scientifi c information forum for exchanging information regarding current and regarding interactions between different elements of future activities, coordinate research, and/or the build- complex land-use systems and biodiversity, especially ing of new alliances. Networks can vary in character, at the landscape scale. Approaches and models used depending on how structured they are and whether a in planning, monitoring and evaluating integrated sys- facilitator is involved. The People Land Management tems are often confi ned by the availability of data and and Ecosystem Services (PLEC) is a prominent example resources, and by the need to make models tractable of a network that aims to identify, test and promote and valid. As a result, few models have incorporated locally-developed, multi-objective land management the temporal element of systems or the dynamic nature practices. This and other examples of networks and of interactions within these systems which are needed partnerships that appear relevant to ecoagriculture are to approximate real-world outcomes. Nevertheless, described in Annex 7B. among the efforts that are in place, progress is being Ecoagriculture Partners, formed in 2002, serves to made to better understand relationships between causes catalyze interactions among scientists of different dis- and effects. ciplines and key stakeholders at multiple levels. It could 7.5.2 Future Directions be benefi cial to the future of ecoagriculture if Ecoagri- culture Partners were to support both an organizational Ecoagriculture is confronted by a split between the partnership component and a network component. The philosophies, understanding and approach of the sci- organizational partnerships could be comprised of a entists and managers involved in wildlife conservation well-defi ned group of organizations that would address and those active with agriculture production. Forging key concerns or issues surrounding ecoagriculture suffi cient common ground between these two cultures practice and research. The network component would to initiate viable ecoagriculture plans and projects will be a vehicle for disseminating information on research put state-of-the-art collaboration and social-learning and practice in this area while stimulating information approaches to the test. exchange, formation of new partnerships, and social Partnerships and networks can engender collabora- learning. Key to its success will be transparency and tion among scientists of different disciplines and with participation (Bruce et al. 2002).

130 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations REFERENCES Brookfi eld, H. 2001. Exploring Agrobiodiversity. New York: Columbia University Press. Agrawal, A. 2002. The regulator community: Decen- tralization and the environment in the Van Panchayat Brown, E. M., O. Y. Da, A.J. Asher, and J.F. Bartholic. (Forest Councils) of Kumaon, India. Mountain Re- 2002. Interactive distributed conservation planning. search and Development 21: 208-211.208-211. Journal of the American Water Resources Associa- tion. 38: 895-903. Allen, W., O. Bosch, M. Kilvington, J. Oliver, and M. Gilbert. 2001. Benefi ts of collaborative learning for Brown, F. I. and A. S. Alechandre. 1995. Empowering environmental management: Applying the Integrated local communities in land-use management: The Systems for Knowledge Management approach to Chico Mendes Extractive Reserve. Cultural Survival support animal pest control. Environmental Manage- Quarterly 18: 54-59. ment 27: 215-223. Bruce, J., C. Valdivia, U. Tan-Kim-Yong, and S. O. Alliance for the Chesapeake Bay. 2004. Citizen Keya. 2002. Report of the First External Review of Monitoring. http://www.acb-online.org/project. the System-Wide Programme on Collective Action cfm?vid=87 (visited May 15, 2004). and Property Rights (CAPRi). Washington, DC: Interim Science Council Secretariat, Food and Ag- Arheimer, B., G. Torstensson, and H. B. Wittgren. 2004. riculture Organization of the United Nations. Landscape planning to reduce coastal eutrophication: Agricultural practices and constructed wetlands. Campbell, B., J. A. Sayer, P. Frost, S. Vermeulen, M. Landscape and Urban Planning 67: 205-215. Ruiz Pérez, A. Cunningham, and R. Prabhu. 2001. Assessing the performance of natural resource sys- Baron, N. 1998. Keeping watch: Experiences from the tems. Conservation Ecology 5: 22. [online]. fi eld in community-based monitoring. Lessons from the Field. Washington, DC: Biodiversity Support Carpentier, C. L., S. A. Vosti, and J. Witcover. 2000. Network. Intensifi ed production systems on Western Brazil- ian Amazon settlement farms: Could they save the Bellamy, J. A., D. H. Walker, G. T. McDonald, and G. forest? Agriculture, Ecosystems and Environment J. Syme. 2001. A systems approach to the evaluation 82: 73-88. of natural resource management initiatives. Journal of Environmental Management 63: 407-423. Cash, D. W. and C. Moser. 2000. Linking global and local scales: Designing dynamic assessment and Bentrup, G. and T. Leininger. 2002. Agroforestry: Map- management processes. Global Environmental ping the way with GIS. Journal of Soil and Water Change 10: 109-120. Conservation 57: 148A- 153A. CGIAR. 2002. Current Defi nition of INRM. http://www. Bocco, G., F. Rosete, P. Bettinger, and A. Valzquez. inrm.cgiar.org/All%20about%20INRM/INRM.def. 2001. Developing a GIS program in rural Mexico: htm (viewed April 25, 2004). Community participation = success. Journal of Forestry 99: 14-19. Conservation Measures Partnership. 2003. Open standards for the practice of conservation. Draft Version Bolwig, S., S. Wood, and J. Chamberlin. 2003. A 1. Washington, DC: Foundations of Success: 22. spatially-based planning framework for sustainable rural livelihoods and land use in Uganda. Report Cornell Lab of Ornithology. 2003. The Power of Citi- No. 158. Washington, DC: International Food Policy zen Science. http://birds.cornell.edu/ LabPrograms/ Research Institute. CitSci/ (viewed April 27, 2004). Bosch, O. J. H., A. H. Ross, and R. J. S. Beeton. 2003. Cowles, P. D., S. Rakotoarisoa, H. Rasolonirinamanana, Integrating science and management through col- and V. Rasaoromanana. 2001. Facilitation, participa- laborative learning and better information manage- tion, and learning in an ecoregion-based planning ment. Systems Research and Behavioral Science process: The case of AGERAS in Toliara, Mada- 20: 107-118. gascar. In: Biological Diversity: Balancing Interests

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 131 Through Adaptive Collaborative Management, eds. Crops and Soil Fertility Concepts and Research L. E. Buck, C. C. Geisler, J. Schelhas and E. Wol- Methods, eds G. Schroth and F. L. Sinclair. Wall- lenberg, 407-422. Boca Raton, FL: CRC Press. ingford, UK: CAB International. Dolman, P. M., A. Lovett, T. O’Riordan, and D. Cobb. Izac, A.-M. N. and P. A. Sanchez. 1999. Towards a Natu- 2001. Designing whole landscapes. Landscape Re- ral Resource Management Paradigm for International search 26: 305-335. Agriculture: Example of Agroforestry Research, International Centre for Research in Agroforestry Douthwaite, B., R. Delve, J. Ekboir, and S. Twomlow. (ICRAF). http://www.inrm.cgiar.org/documents/use- 2002. Contending with complexity: The role of ful_docs.htm. evaluation in successful INRM. Paper prepared on behalf of the INRM-Impact Community of Practice Jackson, D. L. and L. L. Jackson, eds. 2002. The Farm for presentation at the INRM Workshop, ICARDA, as Natural Habitat. Washington, DC: Island Press. Aleppo, Syria, September 16-19, 2002. Jagger, P. and J. Pender. 2003. The role of trees for Douthwaite, B., T. Kuby, E. van de Fliert, and S. Schulz. sustainable management of less-favored lands: The 2003. Impact pathway evaluation: An approach for case of eucalyptus in Ethiopia. Forest Policy and achieving and attributing impact in complex systems. Economics 5: 83-95. Agricultural Systems 78: 243-265. Johnson, A. 1999. Measuring our success: One team’s Ellis, E. A., P. K. R. Nair, P. E. Linehan, H. W. Beck, experience in monitoring the Crater Mountain and C. A. Blanche. 2000. A GIS-based database Wildlife Management Area Project in Papua New management application for agroforestry planning Guinea. In: Lessons from the Field. Washington, DC: and tree selection. Computers and Electronics in Biodiversity Support Program. Agriculture 27: 41-55. Laukkanen, S., A. Kangas, and J. Kangas. 2002. Apply- Fleming, B. and D. Henkel. 2001. Community-based ing voting theory in natural resource management: ecological monitoring: A rapid appraisal approach. A case of multiple-criteria group decision support. Journal of the American Planning Association 67: Journal of Environmental Management 64: 127- 456-466. 137. Ford, R. and W. J. McConnel 2001. Linking geomat- Lee, K. N. 1999. Appraising adaptive Management. ics and participation to manage natural resources Conservation Ecology 3: 3. [online]. in Madagascar. In: Biological Diversity. Balancing Liu, J. and W. W. Taylor. 2002. Coupling landscape Interests Through Adaptive Collaborative Manage- ecology with natural resource management: Para- ment, eds. L. E. Buck, C. C. Geisler, J. Schelhas and digm shifts and new approaches. In: Intergrating E. Wollenberg, 385-405. New York: CRC Press. Landscape Ecology into Natural Resource Manage- Gadgil, M., P. R. S. Rao, G. Utkarsh, P. Pramod, and A. ment, eds. J. Liu and W. W. Taylor, 3-19. Cambridge, Chhatre. 2000. New meanings for old knowledge: UK: Cambridge University Press. The People’s Biodiversity Registers Program. Eco- Loomis, J. B. 2002. Integrated Public Lands Manage- logical Applications 10: 1307-1317. ment. New York: Columbia University Press. Gleismann, S. R. 1998. Agroecology: Ecological Pro- Lovell, C., A. Mandondo, and P. Moriarty. 2002. The cesses in Sustainable Agriculture. Chelsea, MI: Ann question of scale in integrated natural resource man- Arbor Press. agement. Conservation Ecology 5: 25 [online]. Hawkins, V. and P. Selman. 2002. Landscape-scale Margoluis, R. and N. Salafsky. 1998. Measures of Suc- planning: Exploring alternative land use scenarios. cess: Designing, Managing, and Monitoring Conser- Landscape and Urban Planning 60: 211-224.211-224. vation and Development Projects. Washington, DC: Izac, A. M. N. 2003. Economic aspects of soil fertility World Wildlife Fund. management and agroforestry practices. In: Trees,

132 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations McNeely, J. A. and S. J. Scherr 2003. Ecoagriculture: Development Programs to Maximize Results and Strategies to Feed the World and Save Wild Biodi- Learning. Washington, DC: Biodiversity Support versity. Washington, DC: Island Press. Program. Messing, I. and M. H. Hoang-Fagerstrom. 2001. Using Salafsky, N., R. Margoluis, and K. Redford. 2001. farmers’ knowledge for deﬁ ning criteria for land Adaptive Management: A Tool for Conservation qualities in biophysical land evaluation. Land Deg- Practitioners. Washington, DC: Biodiversity Sup- radation and Development 12: 541-553. port Program. Oba, G. and D. G. Kotile. 2001. Assessments of land- Saterson, K., R. Margoluis, and N. Salafsky. 1999. scape-level degradation in Southern Ethiopia: Pas- Measuring Conservation Impact: An Interdisciplin- toralists versus ecologists. Land Degradation and ary Approach to Project Monitoring and Evaluation. Development 12: 461-475. Washington, DC: World Wildlife Fund. Parisi, D., M. Taquino, S. M. Grice, and D. A. Gill. Sayer, J. A., and B. Campbell. 2001. Research to in- 2003. Promoting environmental democracy using tegrate productivity enhancement, environmental GIS as a means to integrate community into the EPA- protection, and human development. Conservation BASINS approach. Society and Natural Resources Ecology 5: 32 http://www/consecol.org/vol5/iss2/ 16: 205-219. art32. Perez, C. and H. Tschinkel. 2003. Improving Watershed Schelhas, J., L. E. Buck, and C.C. Geisler. 2001. In- Management in Developing Countries: A Framework troduction: The challenge of adaptive collaborative for Prioritising Sites and Practices. London: Agricul- management. In: Biological Diversity: Balancing tural Research and Extension Network (AgREN) UK Interests Through Adaptive Collaborative Manage- Department for International Development, 1-20. ment, eds. L. E. Buck, C. C. Geisler, J. Schelhas and E. Wollenberg, xix-xxxv. Boca Raton, FL: CRC Pesonen, M., M. Kurttila, J. Kangas, M. Kajanus, and Press. P. Heinonen. 2001. Assessing the priorities using A’WOT among resource management strategies at Scherr, S.J. 2004. Review Comments. Personal Com- the Finnish Forest and Park Service. Forest Science munication, July 23, 2004. 47: 534-541. Sedogo, L. G. and S. M. E. Groten. 2002. Integration of Prato, T. 2003. Multiple-attribute evaluation of eco- local participatory and regional planning: A GIS data system management for the Missouri River system. aggregation procedure. GeoJournal 56: 69-81. Ecological Economics 45: 297-309. Sheil, D., R. K. Puri, et al. 2002. Exploring Bio- Resilience Alliance 2004. Adaptive Management. logical Diversity, Environment and Local People’s http://resalliance.org/ev_en.php?ID=1073_ Perspective in Forest Landscapes: Methods for a 201&ID2=DO_TOPIC Multidisciplinary Landscape Assessment. Bogor, Indonesia: Centre for International Forestry Research Ribot, J. C. 2002. Democratic Decentralization of Natu- (CIFOR). ral Resources: Institutionalizing Popular Participa- tion. Washington, DC: World Resources Institute. Smith, C. 1998. Gap Analysis Overview. http://www. dnr.cornell.edu/gap/overview.html (viewed April Salafsky, N. and E. Wollenberg. 2000. Linking liveli- 24, 2004). hoods and conservation: A conceptual framework and scale for assessing the integration of human Smith, C. S., G. T. McDonald, and R. N. Thwaites. needs and biodiversity. World Development 28: 2000. TIM: Assessing the sustainability of agricul- 1421-1438. tural land management. Journal of Environmental Management 60: 267-288. Salafsky, N. and R. Margoluis. 1999. Greater Than the Sum of Their Parts: Designing Conservation and Stoms, D. M., K. M. Chomitz, and F. W. Davis. 2004. TAMARIN: A landscape framework for evaluating

Integrating Agricultural Productivity, Biodiversity Conservation, and Livelihood Objectives 133 economic incentives for rainforest restoration. Land- scape and Urban Planning 68: 95-108. Sydenstricker-Neto, J., A. W. Parmenter, and S. De- Gloria. 2004. Participatory reference data collection methods for accuracy assessment of land-cover change maps. In: Remote Sensing and GIS Accuracy Assessment, eds. R. S. Lunetta and J. G. Lyon. 75-90. Boca Raton, FL: CRC Press. van Noordwijk, M. 2002. Scaling trade-offs between crop productivity, carbon stocks and biodiversity in shifting cultivation landscape mosaics: The FAL- LOW model. Ecological Modelling 149: 113-126.113-126. van Noordwijk, M., T. P. Tomich, and B. Verbist. 2001. Negotiation support models for integrated resource management in tropical forest margins. Conservation Ecology 5: 21-39. Vosti, S. A., J. Witcover, and C. L. Carpentier. 2002. Agricultural Intensiﬁ cation by Smallholders in the Western Brazilian Amazon From Deforestation to Sustainable Land Use. Washington, DC: Interna- tional Food Policy Research Institute. Walters, C. J. and C. S. Holling 1990. Large-scale management experiments and learning by doing. Ecology 71: 2060-2068. Wang, X., S. Yu, and G. H. Huang. 2004. Land allocation based on integrated GIS-optimization modeling at a watershed level. Landscape and Urban Planning 66: 61-74. Wildlife Habitat Canada. 2004. Citizen Science. http:// www.whc.org/citizen_science.htm (viewed April 27, 2004). Wollenberg, E., D. Edmunds, and L. E. Buck. 2001. An- ticipating change: Scenarios as a tool for increasing adaptivity in multistakeholder settings. In: Biologi- cal Diversity: Balancing Interests Through Adaptive Collaborative Management, eds. L. E. Buck, C. C. Geisler, J. Schelhas and E. Wollenberg, 329-347. New York: CRC Press. Wood, S., K. Sebastian, and S. J. Scherr. 2000. Pilot Analysis of Global Ecosystems: Agroecosystems. Washington, DC: International Food Policy Research Institute and World Resources Institute.

134 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations ANNEX 7-A: A Summary of Planning, Monitoring and Evaluation Models for Integrating Multiple Objectives

Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Scenario-Based Methods

Arheimer et Landscape Identify an effec- Land- Involve stakeholders in creating Stakeholders are in- Arheimer et al (2004) applied this approach to a al. (2004) Planning tive way for reduc- scape different management scenar- volved in defining the catchment in Southern Sweden. The scenario with Actor ing pollution from includ- ios using the actor game. Use a management scenarios modeling revealed that possible modifications Game non-point source ing mul- quantitative model to simulate and can see the future in agricultural practices (such as timing of fer- pollution tiple changes in pollution levels, impact of these scenar- tilization and plowing, changed crop cultivation) farms using these different land man- ios. This information, could reduce the nitrogen load to the sea by agement scenarios, and identify along with additional some 30%, while wetland construction would optimal management strategy. discussion, is used to only reduce the original load by approximately identify an optimal 5%. Therefore changes in agricultural practices management practice. can be the most effective and less expensive way to reduce nitrogen transport from land to the sea.

Dolman et Whole Land- Improve planning Land- Farmers provided information This approach does not Dolman et al. (2001) pilot-tested this approach al. (2001) scape Plan- of farm lands for scape regarding management. This have a dynamic inter- in Oxfordshire, UK. Future scenarios of inte- ning amenity, biodiver- includ- information is used to build face. It uses informa- grated whole-landscape management were sity and other envi- ing mul- scenarios developed along with tion regarding practice developed with different levels of amenity, environmental benefits tiple surveys of biological and bio- to create the models, ronmental and biodiversity benefits. Of the dif- farms physical variables. Using visu- but there are not ferent scenarios considered, the reactions to alization tools (3-D visualization mechanisms for dia- buffer strips round streams and field margin and 2-D GIS maps of changes), logue among stake- prescriptions were universally favorable. All the scientists validate the relation- holders. farmers were willing to enter into such ar- ships included in the model. rangements if an appropriate compensation Farmers are then invited to was provided. However, the majority wanted to share their perspectives regard- see only select hedges and field margins al- ing different scenarios. tered. In contrast, the scenario involving restoration of wetlands, extensive grassland and seasonal flooding to formerly dry arable land within the floodplain was not as easily accepted because it involved significant changes in resource management.

Annex 7A - 1 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Wollenberg Future Sce- Identify or assess Multiple The process of constructing Stakeholders are in- et al. narios feasible decision scales future scenarios involves: (i) volved in generating (2001) options in natural defining the purpose of the and creating the sce- resource man- scenario, (ii) collecting the nec- narios, which can be agement essary information regarding linked with model or the system, its structure and statistical forecasting drivers of change, (iii) generat- methods as appropriate ing the scenarios, and (iv) discussing and evaluating the scenarios.

Decision-Support Systems

Laukkanen et Multicrite- Support group Multiple Multi-criteria approval involves Decision-makers (and In Finland, this approach was used to facilitate al. (2002) ria Ap- decisions associ- scales multiple steps: (i) determine stakeholders) can be planning the management of a forest area co- proval ated with multiple- natural resource management involved in determining owned by multiple owners. The owners had objective natural alternatives and criteria by the criteria and relevant different, and in some cases, multiple objec- resource man- which alternatives will be com- alternatives. Research- tives, including: timber production, preservation agement and pared, (ii) have decision- based information or of scenic beauty, habitat preservation, biodiver- planning (in an makers rank the criteria in simulation models are sity conservation, and wild berry yield. The mul- objective manner) terms of importance, and (iii) used to determine how tiple-criteria approval tool was used to select determine which alternative is alternatives would per- between 20 alternative forest plans for a 10- favored for each criterion. For form according to these year planning period. Software was used to each step, can use different criteria. determine how the different plans performed approaches, e.g., pair-wise according to measurable criteria. Two plans comparison can be used for were selected through this process, one maxi- step (iii). mized net income and scenic beauty, the other maximized net income and berry yield (Lauk- kanen et al. 2002).

Annex 7A - 2 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Pesonen et Analytical Improve the quan- Multiple A' WOT involves: (i) a SWOT Stakeholders can be The Finnish Forest and Park Service (FFPS) al. (2001) Hierarchy titative basis of scales analysis of the relevant factors involved in designing has adopted a participatory planning process. Process in strategic planning identified in the internal and the strategies to be In western Finland there are differing interests Strengths processes external environment of plan- evaluated using in the forest lands managed by FFPS. Citizens Weak- ning, (ii) a pair-wise comparison A'WOT. They share are most interested in recreation and nature nesses, between factors in each SWOT their perceptions of the protection, and the FFPS manages forest for Opportuni- group (i.e., strength, weakness, importance of different wood production. A'WOT was used to evaluate ties and opportunity and threat), (iii) a SWOT factors and four alternative strategies meeting land-use Threats pair-wise comparison between groups. Research and requirements in this area, with varying amounts Analysis SWOT groups, and (iv) pair- existing knowledge are of land allocated to the different uses. The (A' WOT) wise comparison between al- used to identify the strategy that allocated most land to recreation ternative strategies subject to relevant factors. received highest priority followed by the 'protec- all SWOT factors. tion strategy' (Pesonen et al., 2001).

Prato (2003) Multiple Evaluate alterna- Multiple This approach involves identify- Stakeholders compare Prato (2003) used this approach to compare Attribute tive management scales ing management objectives and alternatives based on five management approaches developed by the Evaluation systems alternatives, attributes of objec- their preferences for Army Corps of Engineers for the Missouri River. tives, and assigning weights to attributes. The value of He used hypothetical weighting schemes for these. For weights uses a hier- the attributes under the ten attributes to evaluate the options. Utility archical approach in which different management scores for the alternatives were obtained using weights are first assigned to alternatives is esti- a linear additive utility function. These scores objectives and then the weight mated using technical indicated that the modified conservation plan, for each objective is allocated models (e.g., CEN- one that incorporates adaptive management to the attributes that describe TURY model) or by with increased drought-prevention measures, the objective. The attribute using expert-based changes in dam releases and unbalanced lev- weights and values are com- methods. els in the upper three reservoirs, was preferred bined in a utility function that is to the current water control plan with the neu- used to score the alternatives. tral, pro-recreation/fish and wildlife, and pro-fish Scores are used to rank alter- and wildlife weights. natives.

Annex 7A - 3 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Approaches for Community-Based Data Collection

Alliance of Citizen Sci- Monitor biological Multiple Involve citizen volunteers in Data collected via citi- The Alliance Citizen Monitoring program is a Chesa- ence factors in defined scales: monitoring of biological factors zen monitoring are regional network of trained volunteers who per- peake Bay ecosystems local using pre-designed tools and used by scientists to form weekly water-quality tests. The information (2004); water- guides for monitoring. understand patterns of is used to track the condition of waters flowing Cornell Lab shed to change in natural re- toward the Chesapeake Bay. The Cornell Lab of Ornithol- eco- sources. The findings of Ornithology uses data collected from citizen ogy, region are made available to a monitors to monitor bird populations, migra- (2004); broad audience to tions, behavior, and disease. Wildlife Habitat Wildlife stimulate learning of Canada uses citizen monitors regarding the Habitat the ongoing natural status of streams, wildlife populations and inva- Canada, resource and biological sive species. (2004) processes.

Gadgil et People's Document the Multiple Engage local communities, This approach empow- During 1996–1998, 52 People's Biodiversity al. (2000) Biodiversity status of biodiver- scales: interviewing key informants and ers local communities Registers were prepared from village clusters Register sity local to groups, map the study site by valuing their tradi- distributed in eight states and union territories regional landscape; visit representative tional knowledge for of India, representing a broad range of ecologi- elements of this landscape with comprehending cal and social regimes. These registers re- key individuals; discuss re- changes in natural sys- vealed a generally declining productivity and source use at the study area tems. diversity of living resources outside of inten- with villagers and outsiders sively-managed ecosystems. They also docu- familiar with the resource base. mented a gradual disappearance of practical ecological knowledge and erosion of traditions of sustainable use and conservation.

Annex 7A - 4 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Approaches for Knowledge Management and Stimulating Collaborative Learning

Allen et al. Integrated Improve informa- Multiple Multiple stakeholders are in- Practice informs the Allen et al (2001) apply this approach to bovine (2001) System of tion exchange scales volved. Their involvement starts research priorities. Also Tb vector management in North Canterbury, Knowledge among stake- from the beginning (defining the scientific informa- New Zealand. They work with an existing advi- Management holders (can be scope) and is maintained tion and the practice- sory group in pest management composed of (ISKM) regarding specific through the whole process. based knowledge are farmers, local government and pest manage- issues or more Create a user-friendly internet- used together in deci- ment agency representatives. Information re- broad considera- based interface for information sion-making. garding ideal management is elicited from the tions. exchange as part of the infor- stakeholders. This information, along with other mation exchange and man- sources, is managed via the Internet in a user- agement. friendly format that allows for addition and extraction of new and relevant information.

Bosch et Integrated Integrate science Multiple Workshops, interviews, visual A key dimension of this Bosch et al (2003) deal with weed management al. (2003) System of and management scales representations of ideas (using process is adaptive in New Zealand using a systems framework. Knowledge by improving un- VENSIM), and community dia- management and on- They facilitate community dialogue and access Management derstanding be- logue for collaborative planning. going knowledge- local and scientific information to identify best (ISKM) tween scientists The information is used to facili- building. This approach management practices. They use the internet to and managers tate dialogue among stake- is based on the princi- manage and allow information exchange. The holders who are crucial to prob- ples of experiential same authors, in another example, use VEN- lem-solving in this approach. learning and systems SIM and INFLUENCE software to show com- thinking. It is cyclical monalities and differences in stakeholders' un- and iterative. derstanding of hydrological issues, and use this as a starting point for dialogue.

Sheil et al. Multidiscipli- Generate informa- Multiple Community meeting; commu- Use local knowledge This effort is ongoing in Indonesia. (2002) nary Land- tion that can be villages nity mapping; community-based regarding uses of forest scape As- used to make bet- data collection; transects; eth- resources, and scien- sessment ter decisions re- noecology; GIS. The commu- tific knowledge regard- garding resource nity is involved in conducting ing other resources, management research, as field guides and together for better re- for local information. These source management. inputs are used to guide later stages.

Annex 7A - 5 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Approaches and Tools for Landscape Level Planning

Bentrup Agroforestry Determine the Multi- GIS, with spatially-explicit soil, The conditions for Bentrup and Leininger (2002) use this approach and Lein- Mapping best location to county, weather and slope/aspect in- growing certain species to determine suitability of agroforestry systems inger grow certain spe- sub- formation, and species re- are known from prac- using both single species and multiple species (2002) cies in agrofor- water- quirements. tice and research. The in Nebraska. They found that decorative willows estry systems shed selection of species to will tolerate most soils, while only a few areas in cultivate in agroforestry south eastern Nebraska are suitable for grow- systems uses market ing mushrooms, medicinal herbs and other information. high-value specialty products.

Brown et Interactive Facilitate informa- Multiple Identify the data that conserva- Information from the This project was conducted in the United al. (2002) Distributed tion exchange scales tion planners require; then de- GIS tool can be used to States, specifically in Michigan, linked with fed- Conservation among stake- velop a prototype to bring these identify high-risk areas, eral/state programs. Planning holders in process spatial data to planners over and then planners can the web. This tool was built on focus on human activi- an interactive GIS format. The ties and best manage- key is providing the information ment practices to ad- at the right scale. Aerial pho- dress the problem. tography images could be help- These areas would ful. Can have web-based mod- need to be assisted ules into which higher resolu- through extension, tion information is entered (e.g., education and aware- regarding a specific field) and ness-raising programs, can generate information that is and outreach efforts. of relevance to planners (e.g., soil erosion, etc.).

Ellis et al. Florida Agro- Determine the Multiple GIS integrated with soil and Practice and research (2000) forestry Deci- potential for spe- scales climate information, and com- are used to identify sion Support cific agroforestry bined with a plant database. suitable conditions for System systems and spe- species. Queries are cies used to identify the suitability of agroforestry and specific species.

Annex 7A - 6 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Ford and Geomatics Facilitate commu- Local to Low-cost air photography and The community uses Geomatics was used in the fourth phase of a McConnell and Participa- nity-based con- regional computer-mapping technology, the aerial photography five- phase participatory planning process in (2001) tion Approach servation planning facilitation of dialogue. to discuss various dif- Madagascar around Mantadia National Park. ferent land uses for The geomatic images provided a basis for dis- village-based land-use cussion in the village-based land-use planning planning, seeking to and resulted in the creation of four new 'user integrate their objec- groups' in the buffer zone. It was also agreed in tives into the regional the co-management plan of the national park co-management plan. that 50 percent of park revenues would be invested in joint development projects. and that people would not expand their slash-and-burn activities any further into the park.

Parisi et al. GIS inte- Integrate informa- Water- GIS is used to provide informa- Incorporation of human This approach was applied to the Upper Pearl (2003) grated with tion on the human shed tion regarding the locality of information shows dif- River Basin in Mississippi, revealing differences Environ- dimensions of human settlement, the level of ferences in human rela- between communities situated in the upstream mental Pro- land use patterns social capital, the economic tionships with natural watershed and those living in the downstream tection capacity of the community, and resources given the watershed. Communities in the former were Agency's demographic characteristics. characteristics of the more rural-oriented, had lower levels of human Better As- community from the capital, smaller labor forces, higher unemploy- sessment demographic and eco- ment rates, higher poverty rates, and the local Science Inte- nomic information in- economies were predominantly based on ex- grating Point corporated. tractive industries (agriculture, mining, forestry, and Nonpoint These differences in and fishing) compared to those in the down- Sources relationships between stream watershed. (EPA-BASIN) people and their environment need to be considered in water management policy- making.

Annex 7A - 7 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Sedogo Geoinforma- Integrate local Local to GIS, local PRA surveys, struc- Biophysical information This approach was applied in Burkina Faso, and Groten tion Approach participatory land regional tured systems approach (for available as GIS layers where local-level data and regional data were (2002) management in integration of the data). Users' is integrated with stake- integrated in GIS. Using the information col- regional planning perceptions help to inform a holder information re- lected via PRAs at the local level, different ori- with GIS conceptual model of system. garding practice. entations can be given to the regional planning process. For example, local- level information reveals that almost half of the province where the study is conducted is facing either a potential labor shortage for implementing land management activities or potential local-to-regional planning conflicts. Less than 25% are facing both constraints. Building on these results, regional planners can explore different scenarios to support local plans.

Smith et al. Threat Identi- Assess the unsus- Multiple (i) Identify and rate potential Information from this The model was tested in North Queensland, (2000) fication Model tainability of agri- scales hazards to land-use sustainabil- model improves land- Australia, using the case of sugar growers in cultural land man- ity (e.g., to productivity), (ii) use planning as com- the area. According to the TIM, if the most agement ex ante identify land management op- pared to the traditional common land-management practices used for tions available to land users, land evaluation ap- sugar production in the sub-catchment were (iii) identify the relationship be- proach. Ex ante infor- implemented on the example land units, land tween these land management mation regarding the degradation (soil structural decline, soil erosion, practices and hazards, (iv) rate land-degradation haz- soil organic matter decline, reduced soil water the suitability of land manage- ards can assist in modi- retention) would be likely. To maintain the sus- ment practice on land units fying practices and tainability of the farm, the farmer should identify based on previous information, adopting more suitable alternatives to disc plowing and disc harrowing, and (v) identify potential secon- land-management op- using trucks in harvesting operations, and culti- dary hazards to sustainability tions. This approach vating differently to achieve pest, weed and from the practice. This is done requires information disease control. by linking a relational database pertaining to the rela- (e.g., Microsoft Access) to GIS tionships between land management practices and land-degradation hazards, as well as on the reversibility of hazards.

Annex 7A - 8 Name of Approach/ Approach Involved and/or Research-Practice Source Tool Application Scale Tools Used Interface Outcome of Application

Syden- Participatory Determine land Large Start with digital images from The visual map engen- This approach was applied in a large coloniza- stricker- Reference cover change in forest satellites. For the reference ders discussion and tion area of Amazonia, Brazil. The accuracy of Neto et al. Data Collec- colonization area; area data collection, field data are awareness regarding land cover classes was between 85 and 89 (2004) tion engage commu- collected by local individuals to land cover changes. percent. Another outcome was farmer empow- nity stakeholders assist in the development of Farmers' realization erment. Farmers gained a greater appreciation in processes of spectral models of each land regarding the land for the development patterns that they were not mapping and in cover type for image classifica- cover changes over aware of. Also, they felt that a deeper under- assessing the tion, and interviews. time stimulates debate standing of what was happening in their area accuracy of maps, on the incentives for would enable them to better respond to local and in evaluating forest conversion ver- needs and contribute to state-wide discussions relevance of maps sus the constraints on promoting environmental sustainability. for understanding imposed by the agricul- community-based tural systems adopted land use dynam- by farmers. ics.

Wang et al. Integrated Allocate land spa- Water- GIS integrated with optimization This approach presents Wang et al. (2004) applied this approach in the (2004) GIS and Op- tially to optimize shed model. a scientific approach to Lake Erhai basin of China. According to the timization land use formulating policies and optimization results, the land use expansion Modeling strategies of environ- includes paddy farming by 1.561 km2, vegeta- mental management. ble farming by 0.037 km2 and forest by 13.948 Interpretation of the km2. The land uses to be reduced included dry environmental planning land farming by 1.110 km2, industry by 0.001 results can be used as km2, and barren land by 14.435 km2. GIS is a policy-support docu- used to determine where these changes should ment for government be made in the landscape. For example, in the authorities and indus- case of sub-area 1, the GIS-optimization model tries that have direct or recommends a reduction of industrial and dry- indirect connections to land farming uses and the expansion of paddy land development and and vegetable farming area. water- quality management.

Annex 7A - 9 ANNEX 7-B: Examples of Partnerships and Network Organizations Supporting Ecoagriculture

Geographic Name Objective Coverage URL

Organizational Partnerships

Millennium The Millennium Ecosystem Assessment (MA) is an international program designed to Global http://www.millenniumassessment.org/e Ecosystem meet the needs of decision makers and the public for scientific information concerning n/index.aspx Assess- the consequences of ecosystem change for human well-being and options for respond- ment (MA) ing to those changes. The MA focuses on ecosystem services (the benefits that people obtain from ecosystems), how changes in ecosystem services have affected human well- being, how ecosystem changes may affect people in future decades, and response options that might be adopted at local, national or global scales to improve ecosystem management and thereby contribute to human well-being and poverty alleviation. The assessments are done by scientists in national and international research organizations. MA uses the web to exchange information, is product-oriented, and the activities associated with the assessment will be repeated every five years.

Alternatives ASB is a global partnership of over 50 institutions with a shared interest in conserving Brazil, http://www.asb.cgiar.org/home.htm to Slash- forests and reducing poverty in the humid tropics. ASB was founded in 1994 as a sys- Cameroon, Indonesia and-Burn tem-wide program of the Consultative Group on International Agricultural Research (ASB) (CGIAR) with the objective of mitigating destructive shifting cultivation by addressing the underlying social and ecological causes and reducing damage to forests by promoting sustainable management of areas adjacent to the forests. ASB is convened by the Nai- robi-based World Agroforestry Centre (ICRAF) and is governed by a global steering group of 12 representatives from participating institutions.

IUCN Sus- The Sustainable Use Initiative is a global technical effort to increase knowledge and un- Global http://www.iucn.org/themes/ssc/susg/sui tainable derstanding of factors that influence the sustainability of natural resource use. The initia- .html Use Initia- tive is made up of 16 decentralized networks of regional Sustainable Use Specialist tive (USI) Groups (SUSGs). The SUSG mission is to promote conservation of biodiversity and alleviate poverty by: (i) improving understanding of social and biological factors that enhance sustainable use of wild living resources, (ii) promoting understanding to IUCN's members and decision-makers and others, and (iii) assisting IUCN members, partner organizations and government in applying this understanding. These groups are made up of practitioners who analyze and compare local use systems through case studies, regional re-

Annex 7B - 1 Geographic Name Objective Coverage URL

views, workshops and symposia.

Program on PROFOR is a multi-donor partnership formed to pursue a shared goal of enhancing for- Global http://www.profor.info/ Forests ests' contribution to poverty reduction, sustainable development, and protection of envi- (PROFOR) ronmental services. Through improved knowledge and approaches for sustainable forest management (SFM), PROFOR seeks to encourage the transition to a more socially and environmentally sustainable forest sector supported by sound policies and institutions that take a holistic approach to forest conservation and management. PROFOR fosters such policies and institutions through support to participatory processes, such as national forest programs, and knowledge generation in four key thematic areas: forest governance, forests' contribution to livelihoods of the rural poor, mitigation of adverse cross- sectoral impacts on forests, and innovative approaches to financing SFM.

Conserva- The Conservation Measures Partnership (CMP) is a joint venture of conservation NGOs Global not available tion Meas- and other collaborators who have come together to work on issues related to impact ures Part- assessment and accountability in an effort to find better ways to design, manage and nership measure the impact of their conservation actions. CMP's mission is to improve the prac- (CMP) tice of biodiversity conservation by developing and promoting common standards and an auditing mechanism for the process of conservation and measuring conservation impact. Each organization within CMP has biodiversity conservation as its primary goal, focuses on field-based conservation actions, and is working to develop better approaches to project design, management and assessment. The core members of CMP include African Wildlife Foundation, Conservation International, The Nature Conservancy, Wildlife Con- servation Society, and the World Wildlife Fund. Collaborating organizations include En- terprise Works Worldwide, World Commission on Protected Areas/IUCN, and Founda- tions of Success. FOS serves as the CMP coordinator, carrying out the day-to-day management of CMP and facilitating the completion of various technical projects.

Annex 7B - 2 Geographic Name Objective Coverage URL

System for START establishes and fosters regional networks of collaborating scientists and institu- Global http://www.start.org/ Analysis, tions in developing countries. The networks conduct research on regional aspects of Research environmental change, assess the impacts and vulnerabilities resulting from these and Train- changes, and provide information to policy-makers. START acts to enhance the scientific ing capacity of developing countries to address the complex process of environmental (START) change and degradation through a wide variety of training and career development programs.

Networks for Information Dissemination, Partnership Formation, Social Learning, and Coordination

Collective CAPRi is a web-based network that disseminates information among researchers, practi- Global http://www.capri.cgiar.org/ Action and tioners and other individuals. It fosters research and promotes collaboration on institu- Property tional aspects of natural resource management between Future Harvest Centers, Na- Rights tional Agricultural Research Institutes, and others. CAPRi intends to contribute to policies (CAPRi) and practices that alleviate rural poverty by analyzing and disseminating knowledge on the ways that collective action and property rights institutions influence the efficiency, equity, and sustainability of natural resource use.

USAID- FRAME is a community of experts and practitioners active in the management of Africa’s Africa http://www.frameweb.org based Net- natural resources. It is also a knowledge-based tool that shares this community’s les- work of sons, best practices, and solutions. FRAME creates an opportunity for analysts and de- Resources cision-makers to think strategically about environmental and natural resource manage- for Africa ment issues as they relate to the challenge of sustainable development in Africa. It builds (FRAME) a dynamic that brings together people who rarely interact, catalyzing new working relationships that facilitate the exchange of information and broadening of perspectives. FRAME is designed to facilitate the increased use of up-to-date information by environment/NRM decision-makers and practitioners as they analyze issues, plan strategically, and advocate their positions. FRAME seeks to generate and provide information (from within and outside Africa) to help answer strategic questions facing environment and natural resources management in Africa: (i) What are the key environmental and natural resource management issues confronting Africa? (ii) Where are people addressing these issues in innovative and effective ways? (iii) What factors helped people achieve this progress? And (iv) What will it take to achieve broad-based changes in the management of the environment and natural resources which are required to support sustainable development in Africa?

Annex 7B - 3 Geographic Name Objective Coverage URL

People, PLEC aims to identify, test and promote locally developed management practices that: (i) Global http://www.unu.edu/env/plec/index.htm Land Man- combine traditional knowledge and approaches with new knowledge and technologies; agement, and (ii) embrace ecosystem functions and processes for enhancing livelihoods, princi- and Eco- pally through optimal degree of structural, spatial, temporal, trophic, species and genetic system diversity. The project involves local farmers and scientists in setting up demonstration Services sites in critical ecosystems and areas of globally significant biodiversity. The PLEC net- (PLEC) work uniquely provides for South-to-South cooperation and South-to-North arrangements. The project is organized into a network of locally-based clusters and representa- tively-diverse regions that have been established in Africa (Ghana, Guinea, Kenya, Tan- zania, Uganda), Asia-Pacific (China, Thailand, Papua New Guinea), and Latin America (Brazil, Jamaica, Peru, Mexico) with participation of scientists from the North (currently United States, Japan, Australia, and Britain). Each cluster is multidisciplinary, based in a national organization in collaboration with several institutions.

Interna- The International Sustainability Indicators Network is a member-driven organization that USA http://www.sustainabilityindicators.org tional Sus- provides people working on sustainability indicators with a method of communicating tainability with and learning from each other. Through listserv discussions, virtual and in-person Indicators meetings, and special programs and trainings, the Network facilitates shared learning Network and development among sustainability indicators practitioners and others. This website (ISIN) also has links to ongoing efforts to set up indicators.

Sustainable This is an electronic newsletter that has information for community based pro- USA http://www.sustainable.org Communi- grams/activities/ efforts to protect and manage resources. Community involvement in ties natural resource management is a major area of focus. This section presents various approaches and techniques used successfully in different communities to protect and restore their natural resources. Under each of the resources (biodiversity, land and agriculture, water, etc.) there is a list of institutions that are working on the issue and also a few case studies and recommended reading. The objective of this network is to widely disseminate information regarding various approaches and techniques successfully used by different communities to protect and restore their natural resources.

Annex 7B - 4 Chapter 8

Conclusions

8.1 Biodiversity Contributions biodiversity conservation and utilization, conservation tillage, organic production systems, systems approaches The concept of ecoagriculture as a multi-objective to pest management, integrated nutrient management, land-use strategy that can deliver agricultural produc- soil health, below-ground biodiversity in agricultural tivity, livelihood support and biodiversity conservation systems, management of the hydrological cycle, and benefi ts at farm and landscape levels has a foundation the system of rice intensifi cationBiological dynamics in scientifi c knowledge and understanding. This as- give ecoagriculture a chance, even with low external sessment has found that there is considerable evidence, inputs, to achieve reasonably high levels of productivity some of it rigorously quantifi ed, that a variety of agri- and to be economically profi table. Further, synergies cultural practices throughout the world provide habitat are possible between agrobiodiversity and wild biodi- for locally and globally important species of wildlife versity especially considering the dynamics of ecosys- at the same time they produce food and other benefi ts, tems underground. What is not known, given limited including livelihood creation. Practices that deliver experience with ecoagriculture practices and even less habitat benefi ts most consistently to the broadest spec- with evaluation research, is how far these emerging ap- trum of taxa, according to our analysis of a sample of proaches can be developed to improve upon, or become studies from the recent literature, are hedgerows and an alternative to conventional agriculture, and in what woodlots adjacent to farm fi elds, organic production and how many places. systems, and shaded tropical tree crops, especially coffee and cacao (Chapter 4). 8.2 Agricultural Sector Contributions From many of the studies examined, it can also be in- Four conclusions can be drawn from our review of ferred that “biodiversity-friendly” agricultural practices agricultural literature and experience. First, there are were also delivering substantial livelihood support. substantial opportunities for agriculture in the future However, few studies addressed this aspect directly, and to become more productive, profi table, sustainable and thus they offered little insight or evidence on how socio- environmentally friendly by relying more on energy and economically viable or how sustainable the practices be- nutrients made available through biological processes. ing evaluated may be. This is an important limitation to These processes can be favorably affected by making much of the biophysical literature, and one that should changes in management practices for soil, water, plants, be addressed to strengthen the scientifi c underpinnings animals, and nutrients. This trend will not replace all of ecoagricultural land use in the future. use of chemical amendments, but the presumption ap- Our assessment has identified and characterized a pears to be shifting from chemicals substituting for or variety of agricultural production approaches that are compensating for biologybiology,, to chemicals supplementing consistent with the premises of ecoagriculture and that biology. This is the direction in which we see 21st cen- illustrate the potential for “positive-sum” relationships tury agriculture moving. Ecoagriculture to be successful between agricultural productivity and ecosystem func- and to spread will need to have strong foundations in tions. Evidence for achieving synergistic relationships agricultural science and practice that are both “good through effective management of interactions among agriculture” and “good ecological stewardship.” biological components of agricultural production sys- Second, two of the longest-held and most fi rmly be- tems is found in the literature on agroforestry, agro- lieved precepts in agriculture—that plowing isis requiredrequired

Conclusions 135 for crop cultivation and best results, and that continuous and potentials and become cost-saving, more profi table fl ooding ofof ricerice ppaddiesaddies iiss nnecessaryecessary fforor tthehe hhighestighest and environmentally benign. A present constraint on production—are being proven wrong by experience such systems is often their labor-intensity, as they often with conservation agriculture and the System of Rice require additional labor inputs that may not be avail- Intensifi cation. These methodologies are getting very able at prevailing wage rates or within the household. benefi cial, cost-effective results by changing the way Work is underway modifying many systems to account that plants, soil, water and nutrients are managed. The for this constraint, and labor intensity has the benefi t reasons for this are fully explainable by what is already of creating employment and livelihood opportunities. known and accepted within the crop and soil sciences, But it remains an important constraint. How much of but an old paradigm of production has often kept sci- a constraint labor-intensity is will depend in large part entists and practitioners from seeing and accepting the on what happens to labor productivity. New practices new opportunities when fi rst presented, even evoking and farming systems that raise the productivity of la- strong resistance at fi rst. We do not know in how many bor, as well as other resources, will become attractive other respects conventional wisdom subtly incorporated and certainly gain acceptance. Ecoagriculture will not into science is limiting future opportunities that would succeed if it is labor-intensive without giving farmers be supportive of ecoagriculture. and agricultural sectors higher total factor productivity, including greater labor productivity. This is a tall order, Third, biotechnology is opening up many new pos- but current scientifi c knowledge and practical experi- sibilities, but so far it has been applied mostly within ence suggest that it is feasible in many cases. existing Green-Revolution production paradigms, not being connected to the emerging agroecological para- 8.3 Livelihood Contributions digm which focuses on ensembles of organisms and on their interactions rather than on species in isolation. Our review of the literature on economic and liveli- USDA research reported in Chapter 3 (Kumar et al. hood considerations has yielded several conclusions 2004) shows how the most advanced techniques for relevant to the socio-economic dimensions of research genetic analysis can be used to illuminate how chang- relevant to ecoagriculture. To begin, ecoagriculture is ing plants’ environmental conditions, particularly in the inherently a multi-objective strategy in which liveli- rhizosphere, can give better results, higher production hood generation must be achieved simultaneous with and more tolerance to disease. It does this by identifying crop productivity and biodiversity conservation. This which genes express themselves differently and trigger means that evaluating the multi-dimensional outcomes different physiological processes and responses when, in of empirical research is central to evaluating ecoagri- this case, a leguminous mulch was used with less urea, culture strategies; yet, in fact, most relevant research compared with plastic mulch and higher N fertilization. to date has yielded results and outcomes that are par- The organic agriculture adage that instead of “feeding tial in nature, only infrequently addressing all three the plant” we should “feed the soil and the soil will feed dimensions of ecoagriculture. This makes it diffi cult the plant” is gaining scientifi c respectability. These are at present to assess comprehensively the tradeoffs and complicated processes, and scientifi c knowledge is still complementarities that arise in ecoagriculture strate- accumulating, not fi nal, while some of what we now gies because the relevant research is so sparse. Rigor- know is being unlearned. Linking biotechnology and ously and comprehensively addressing the research agroecology, with special interest in soil biology and challenges introduced by ecoagriculture in enough ecology, including more study of plant roots and the real-world contexts to frame generalizable conclusions rhizosphere, is a promising new frontier (Science June about its viability will require a much more concerted 11, 2004) that would benefi t ecoagriculture. approach among researchers to broaden their analytical frameworks and outcomes. Lastly, farming systems research has progressed to produce more scientifi cally informed analyses and rec- Second, in the limited instances where multi-dimen- ommendations for farming systems such as agroforestry sional results are rigorously derived and available, the and crop rotation that capitalize on biological processes research suggests that tradeoffs are at least as common

136 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations as synergies in accounting for the relationships among at the farm, household, or community level, at most production, livelihood, and biodiversity objectives. the sub-watershed level. This is understandable, and In other words, achieving results with respect to one results from many reasons: the core interests of many objective—whether higher crop production or produc- researchers and development practitioners in conduct- tivity, or enhanced household livelihoods, or preserving micro-level research; the complexity of farms and ing biodiversity—more often than not comes at the households and the decision-making underlying these expense of one or more of the other objectives. This units; the relative expense and analytical complexity is not compatible with what ecoagricultural strategies of conducting research at more aggregative levels; and seek. We have raised, but certainly not answered, the so forth. But the focus on micro-level research often closely related question of “separation versus integra- fails to incorporate effects that may be only realized tion” which is also highly relevant in this context. at a broader scale, yet which may become critical to To the extent that situations exist where production, decision-making in farms and households. livelihood, and biodiversity objectives can be equally For example, much of marketed farm production in (or better) achieved in spatially distinct contexts, then any country goes toward feeding urban residents. But the central premise underlying ecoagriculture may be those consumers may be indifferent to consuming do- fl awed in these cases. Given that the research record mestically-produced foodstuffs versus imports, and will on all of these issues is sparse, it will be necessary to often choose the cheapest alternative. Focusing simply expand not only the scope of relevant research but the on increasing domestic production or productivity in specifi c applications (geographic, systems, scale, etc.) an ecoagriculture (or any other) context may “win the in order to derive a better sense as to the conditions battle and lose the war” if we fail to consider what is (economic, biological, climatic, etc.) under which happening in international markets and among urban specifi c ecoagricultural strategies may be expected to consumers. A second example is the familiar economic thrive and succeed. problem of “immizerizing growth,” wherein increased Third, one of the central notions behind ecoagriculture production may result in lower prices and offsetting is that of addressing externalities—including but not terms-of-trade effects that may counteract the economic limited to biodiversity preservation—that may be in- incentives that originally gave rise to increased produc- adequately or ineffectively refl ected in current private tion. This has been common, for example, in non-tra- resource-use decisions. As we have seen, much of the ditional export markets, such as those for horticultural current interest in valuing ecosystem services is based products, where simultaneously increased production on the idea of incorporating previously ignored exter- in many countries has driven down market prices for nalities into private decision-making by attempting these products, jeopardizing the diversifi cation strate- to value the underlying services that have previously gies that gave rise to production in the fi rst place. The been treated as public goods, often resulting in resource point here is that scale and aggregation are important. overuse, degradation, and unacceptable social costs. Deriving highly successful ecoagriculture solutions that Whether ecoagriculture is in the long-run successful meet production, livelihood, and biodiversity objectives in helping preserve biodiversity will depend on better in confi ned micro-level situations may still be problem- success than has been achieved to date in valuing these atic when scaled up geographically or aggregated up and other ecosystem services and developing replicable, in regional and international markets. Researchers and institutionally sustainable, adequately funded, incen- policy-makers must be aware of these limitations. tive-compatible programs that can adequately address A further tradeoff issue that deserves more attention, issues of externalities. is whether, or to what extent, the intensifi cation of Finally, it is important to mention a broader point agricultural production in areas better-endowed for regarding scale, “scaling up,” and aggregation that agricultural production and raising yield there will in has not been explicitly elaborated in this report. fact reduce pressure on the more fragile, “marginal” Much relevant research pertaining to ecoagriculture, areas. It has been argued that millions of hectares of if not directly labeled as such, has been conducted biodiversity-rich lands have been preserved because

Conclusions 137 of the productivity gains achieved through the Green within a particular area and time frame, but also within Revolution. This is certainly true to some extent, but each domain. Should we be predominantly concerned it is not clear by how much. This logic applied to pe- with conserving agrobiodiversity or wild biodiversity ripheral zones around protected areas, for example, in a certain area or with a certain farming system, for can be defeated if the higher productivity promoted example? Perhaps a tougher choice is whether species around a park draws more population toward the vul- richness is an appropriate measure of biodiversity when nerable natural resources. The apprehension that this it may represent a greater abundance of common spe- will happen has led toward more ecoregionally-defi ned cies at the expense of a few rare or threatened ones? conservation efforts, for example, in Madagascar, seek- Trade-offs between livelihood benefi ts will pit present ing to promote economic development and opportunity vs. future, local vs. global, along with more qualitative on a broader landscape or regional scale, not just in (moral) questions such as to whether it is suffi cient proximity to vulnerable areas. Though the existence of simply for populations to survive or should we be sat- these tradeoff is logically appealing, there is little direct isfi ed only if they thrive? To what extent should some evidence showing any 1:1 substitution of production stakeholders be allowed to profi t at the expense of increase in favored areas offsetting demand for equal other actors and elements of an ecoagricultural system? production increase in areas where biodiversity is still These are all diffi cult questions, which are often hard rich. This adds to the case for examining ecoagricultural to address in an informed or authoritative way through alternatives and promoting viable ones, not following existing institutions and decision-making processes. a dichotomous strategy to conserve wild biodiversity It is important to the future success of ecoagriculture but rather a more integrated one. as an approach to land-use analysis and practice that planners, managers and policy-makers remain realistic 8.4 Combining an Integrative Vision with Making Diffi cult Choices about how much any particular system can deliver. Ex- alting ecoagriculture as a simple answer will limit its While there is reason to conclude that notions of eco- viability as an integrative vision and concept. Therefore, agriculture are scientifi cally grounded, the foundation robust models and methods need to be designed and is hardly rock-solid. Indeed, our assessment uncovered used to systematically and reliably assess synergies vs. rather few rigorous studies that have examined crop tradoffs in ecoagriculture systems. This work remains productivity, biodiversity conservation, and livelihood to be done. support as joint outputs of an agricultural system. The At its present stage of development, ecoagricultural assessment has highlighted notable examples of a few thinking is perhaps at best an attractive visioning process studies that did assess simultaneously the levels, and for improving land use. Already this vision has identi- interactions among, productivity, poverty alleviation, fi ed a good number of land-use systems throughout the and ecological services such as carbon sequestration world that appear to be managed, whether incidentally and/or hydrological function, producing at least some or deliberately, for agricultural productivity, liveli- intermediate if not final results. Integrating direct hood support, and conservation of biological diversity.diversity. measures of biodiversity conservation into these mul- As an emergent interdisciplinary fi eld for the applied tivariate, integrated studies, however, remains a task biological and social sciences, ecoagriculture is gain- of the future. ing the attention of ecologists, biologists, economists, This assessment concludes that although there is scope political scientists and sociologists as well as planning for synergistic effects and positive-sum interactions and management specialists who are prepared to work between and among the three legs of the ecoagricultural within various “systems” frameworks to understand “stool,” trade-offs are likely to be the more prevalent interactions that affect the three essential outputs of outcome and decision rule at present and for some time any ecoagriculture system. Furthermore, agronomists, into the future. Diffi cult choices will be necessary not plant and animal breeders, plant pathologists, hydrolo- only among or between agricultural productivity, local gists, geographers and scientists of other disciplines livelihood sustainability, and biodiversity conservation

138 Ecoagriculture: A Review and Assessment of its Scientifi c Foundations are inclined and equipped to contribute to the pursuit 8.4.1. Bridging Gaps in Thinking and Analysis of ecoagriculture goals. The efforts of a few pioneering scientists who have As global society comes to terms with the need to forged ahead in the domain of ecoagriculture notwith- reconcile expanding human activity with the survival standing, our assessment has detected a continuing and well-being of other life forms on the planet, and division between mainstream agricultural and con- as institutions charged with fi nding mutually satisfac- servation scientists. Their different worldviews lead tory solutions to human welfare and biodiversity con- to different and often divergent objectives, study sites servation needs gain popular and political footholds, and conclusions concerning how best to develop, use we can anticipate that interest in scientifi cally-sound and manage renewable natural resources. While this knowledge and information about ecoagriculture will situation limits progress at present, relative to the expand. To help accelerate these conditions, concerned challenges that opportunities for ecoagriculture pose scientists should seek funding and use the resources to the scientifi c community, we anticipate that gaps at their disposal to advance research within the arena and differences will be overcome in time. The process of ecoagriculture. Indications of movement in this will accelerate as appreciation is gained for the global direction are the inclusion of a suite of ecoagriculture conservation community’s increasing inclination to sessions in the World Agroforestry Congress held in encompass agricultural lands and “working landscapes” Orlando, Florida, June 2004, and the interest of a large in its conservation strategies. Furthermore, as agricul- number of agriculture scientists and conservation biolo- tural scientists trained to think agroecologically gain gists in participating in the International Ecoagriculture maturity and status within their organizations, and as Conference and Practitioners’ Fair, to take place in cases are highlighted of where conservation programs Nairobi, Kenya, September 2004. and practices generate sustainable economic benefi ts This review and assessment of the scientifi c founda- for local people, making conservation more generally tions for ecoagricultural land use has shown that there acceptable and attractive, ecoagriculture approaches is a tremendous amount of literature and web-based will advance information on this subject as well as numerous projects To bridge the chasm in a timely manner, however, and knowledgeable people who can contribute to the un- will require concerted efforts on the part of numerous derstanding needed to advance this fi eld of knowledge individuals and insti tutions working together. This and practice. The fl edgling fi eld is not unifi ed, however. assessment has tried to address key defi nitional and To harness these resources and improve access, more conceptual issues up front, in particular biodiversity, deliberate coordination mechanisms are needed. The intensifi cation, and trade-offs vs. synergies. Misunder- formation and mission of Ecoagriculture Partners (EP) standings regarding these terms have slowed progress is a signifi cant move in this direction. EP, whose core toward collaboration among the diverse people and partners include leadership of the World Agroforestry organizations working in this area. We hope that this Center, IUCN, and the NGO Forest Trends, appears assessment has contributed also by identifying the well-suited to stimulate the high level of organization types of partnership and network organizations that are and authority that will be required to generate sus- demonstrating effectiveness in meeting challenges of tainable research and development momentum from integration and collaboration. the initial fl ush of enthusiasm for ecoagriculture that this assessment has documented. The outcome of the 8.4.2. Engaging Multiple Disciplines planned international conference in September will give Understanding the numerous scale issues that ecoagri- a better reading of how far and how fast this collabora- culture poses presents both methodological and cross- tive effort will proceed. EP needs to remain especially disciplinary challenges. This assessment has noted the attentive to the need to channel fi ndings from research importance of accounting for scale effects in different and development initiatives through its portals to keep ways: 1) spatial scale: the need to link scales of analysis everyone interested adequately informed. and action according to the issue involved and the units

Conclusions 139 of measurement and management; 2) aggregation and and used in a meaningful fashion both for research scaling up: larger spatial units present challenges of and policy analysis. In addition to better measures and combining and integrating efforts, including the diffi - indicators, we need better modeling methodologies culties of collective action; and 3) measurement scales: that can evaluate alternative outcomes with respect to recognizing that the relations among scales of analysis the objective of biodiversity conservation. With these are not linear and must deal with emergent properties at measures and tools in hand, biodiversity conservation the same time that good reductionist analysis is carried values can be integrated into some of the numerous out. These issues will have to be reckoned with in any extant models of agriculture and natural resource land forthcoming evaluations of ecoagricultural land use. use and economics, and others could be adapted for the expanded purpose. There are numerous pathways into the interdisciplinary fi eld of ecoagriculture. Landscape ecology – the study Related to this, we need to know how to measure and of the interactions between the temporal and spatial assess biodiversity in general more quickly, effi ciently aspects of a landscape and its fl ora, fauna and cultural and accurately in the fi eld. For this we need to develop components (Smithers 2004)—is probably best suited protocols for the widespread use of these methods. to organize the fi eld’s core knowledge domain. Land- We need also to know more about corridors through scape ecologists generally can be expected to have the agricultural landscapes. Do they work, and under what background and orientation to promote communica- design parameters and management conditions? There tion, interdisciplinary research, and the development of has been a lot of discussion of corridors as valuable for knowledge and interaction between scientists and those biodiversity shaped by thinking and experience with engaged in the planning and management of landscapes. large fauna. A corridor that goes from high elevation Conservation biology, ecological economics, and agri- to low elevation and back to high elevation may be cultural sciences need to be securely positioned within tenable for mobile mammals, but it has little relevance integrative frameworks that landscape ecologists, plan- for plants. ners, and managers take initiative to shape. In addition, more needs to be known about the relation- 8.5 Research Needs and Opportunities ships between below-ground and above-ground biodiversity. Related to this, are some general rules of farm Long-term, landscape-scale experiments are needed management that are almost always correlated with that encompass communities of farms and attempt to enhancement of wild biodiversity? Fewer chemicals, improve economic viability, agricultural productivity less tillage, better soil health, more tree cover? and conservation of biodiversity, simultaneously in aggregative and cumulative ways. The landscape of farms, We would like to see a table constructed of currently which would encompass a variety of experiments on documented examples of how wild biodiversity spe- the productivity effects of different methods, would be cifi cally contributes to or enhances crop and animal managed to see to what extent species of plants or ver- production, and a protocol for documenting additional tebrates that were not there at present could be induced examples from around the world. Building up a sys- to colonize or re-colonize the area, or whether existing tematic knowledge base on these relationships would diversity can be maintained over long periods of time. strengthen the case for making eco-friendly transitions While such outcomes, if achieved, would be good for in farming systems. biodiversity, what specifi c changes would they induce, An important question that needs answering is: what over time, and would this be good for the other two legs can particular agricultural practices and/or landscapes of the stool: how, why, and under what conditions? do for regionally or globally rare species of fauna or Perhaps the greatest near-term current research need fl ora? Also, how do terrestrial practices affect aquatic is to develop measures and indicators of wild biodi- and/or marine ecosystems? Any habitat manipulation versity, with respect to its variety and abundance, that will inevitably favor and disfavor some particular can be easily incorporated into modeling approaches species. Given that any agricultural practice is bound

140 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations to help some species and displace others, it would be instructive to have a list of species that are known to be extremely important in terms of ecosystem services, or that are rare regionally or globally, or that are good indicator-species for other groups of species. At the same time we need to identify which agricultural practices and agricultural landscape-design features can actually help these selected species. It will be insightful, then, to evaluate how livelihood support indicators correspond to these relationships. Ecoagriculture Partners is uncovering a vast network of agricultural practices and practitioners around the world that are allied with ecoagriculture thinking. The proclaimed interest of many hundreds of organizations in joining EP to gain recognition and momentum for their activities suggests a powerful opportunity to establish a global, practitioner-based ecoagricultural monitoring system. The leadership of EP will presumably use the upcoming conference as an opportunity to mobilize the scientists and organizational leaders in attendance to initiate a strategy for more systematically measuring and evaluating ecoagricultural phenomena so that governments, donor agencies, NGOs, local government bodies and, most of all, rural communities and resource-users can make better informed decisions about how they can “eat their current cake” and still “have a desirable future cake” as well.

REFERENCES

Kumar, V., D. J. Mills, J. D. Anderson, and A. K. Mattoo. 2004. An alternative agricultural system is deﬁ ned by a distinct expression of select gene transcripts and proteins. PNAS: Proceedings of the National Academy of Sciences 101: 10535-10540 (July 20). Smithers, R, 2004. Landscape ecology of trees and forests. Proceedings of the 12th Annual IALE (UK) conference 21-24 June, 2004. Lincolnshire, UK: International Association for Landscape Ecology (UK) and Woodland Trust.

Conclusions 141 142 Ecoagriculture: A Review and Assessment of its Scientiﬁ c Foundations