Michigan Field Crop Ecological Weed Management: A Decision Support System - Final Report

Disclaimer

This report was prepared by an EPA assistance agreement recipient and represents only the views of the author rather than EPA.

Dale R. Mutch
MSU Extension, W. K. Kellogg Biological Station, Hickory Corners, MI 49060

Adam Davis
USDA-ARS, Weed Ecologist, Urbana, IL 61801

Karen Renner and Christy Sprague
MSU Department of Crop and Soil Sciences, East Lansing, MI 48824

Project Coordinator

Dale R. Mutch, Ph.D.
Extension Specialist
Michigan State University
W. K. Kellogg Biological Station
3700 E. Gull Lake Drive
Hickory Corners, MI 49060
269-671-2412, ext. 224
269-671-4485 (fax)
mutch@msu.edu

Executive Summary

Environmental, economic and human health challenges associated with over-reliance on chemical herbicides have prompted a growing number of producers to consider alternative weed management strategies. Michigan State University Extension (MSUE) has responded to farmer demand for an ecological approach to field crop production and pest management by making a long-term commitment to providing up-to-date, user-friendly information on field crop ecology and management.

Despite the increased availability of this environmentally sound agroecological information for Michigan field crops, many farmers express reservations about reducing their dependence on herbicides. The difficulty of applying ecological weed management concepts has even led some fledgling organic producers to return to herbicides. To promote increased adoption of holistic weed management practices, the MSUE field crop bulletin series must provide growers with a practical framework for implementing them.

This project brought six conventional and six organic farmers, six researchers and five Extension educators together on MSU’s campus to discuss non-chemical weed management strategies. Four intensive meetings were conducted during the late fall and winter of 2003/2004.

During the fall of 2004, an intensive training program for field crop MSUE educators to learn about weed ecology was offered. We conducted three all-day meetings in East Lansing and 18 educators participated. A draft of the Integrated Weed Management: One Year’s SeedingY” (MSUE bulletin E-2931) chapters were presented to the educators for edits and comments. Educators were asked at the last training to schedule winter training programs (Appendix 1—Evaluation).

In the winter of 2005, nine meetings were held across Michigan about E-2931 where MSUE educators hosted programs. More than 600 farmers were trained as a result of this grant, and 145 of those were organic farmers. More than 600 copies of E-2931 have been sold this year.

This proposal leveraged two additional grants, an MSU GREEEN grant for $6,000 to support some of the printing costs, and an NCR-SARE grant for $150,000 to have training programs across the North Central region.

Objectives

Form a working group of Michigan farmers, researchers and Extension agents who are interested in developing weed-suppressive field crop production systems with limited, or no, reliance on herbicides.
Compile working group insights into a text-based, user-friendly, decision-support system, tentatively titled “Michigan Field Crop Ecological Weed Management” (MFCEWM), for adaptive management of specific weed life stages.
Conduct four regional workshops across Michigan to educate farmers and Extension agents in ecological weed management as presented in “Michigan Field Crop Pest Ecology and Management” (MSUE bulletin E-2704) and MFCEWM.
Synthesize regional meeting feedback into a critical evaluation of MFCEWM that identifies points in weed life cycles where basic and applied weed science research should be targeted to develop holistic weed management strategies.

Introduction

Objective 1

Weed management is a challenge in agricultural production systems with minimal reliance on herbicide use. In a national survey of organic farmers (OFRF, 1998), non-chemical weed management was the respondents’ number one research need. The response in the weed science community to such demand has been slow, and inconsistent, with certain weed life stages (e.g. germination and early growth) receiving much more attention than others (e.g., seedbanks). Yet an ecological approach to weed management must exploit weaknesses throughout the weed life to be successful (Mohler, 1993). Weed life cycle management requires the integration of multiple weed management practices (an approach known as ‘many little hammers’) into a coherent system with complementary practices (Liebman and Gallandt, 1997).

Though some farmers have developed ecological weed management systems after several decades of practice (Thompson et al. 1998), many find it difficult to implement the ‘many little hammers’ concept (Jordan et al., 2000). Working groups allow farmers, researchers and Extension agents to work together to solve difficult problems (Jordan et al., 2002). In contrast to the traditional one-way flow of information from researcher to Extension agent to farmer, working groups place researchers and Extension agents in the role of facilitators and fellow learners, opening up possibilities for greater farmer participation in generating new knowledge. This approach has been employed effectively elsewhere in the Midwest to integrate the theory and practice of ecological weed management into successful on-farm applications (Jordan et al., 2000).

To address this objective we conducted four all-day interactive learning sessions in 2004. The first meeting was on January 7, with the rest held on Wednesdays every two weeks, ending on February. We recruited six organic farmers and six conventional farmers, researchers and Extension educators for these meetings. A professional facilitator was hired to conduct the meetings allowing our team the freedom to participate in the discussions. We had a mixture of PowerPoint presentations, small working groups and large total group discussions on non-chemical ways to manage weeds.

It was important to have a professional facilitator involved to help with the different personalities and remain focused on the topic. The first meeting was rather tense as these groups generally aren’t in the same room with each other and many had questions on why the others were there. However, by the end of the fourth meeting, both the organic and conventional farmers bridged their differences and each showed respect for the others’ talents and opinions.

The farmers were chosen for their ability to work in groups and not for their position on agriculture’s future. They all realized that we were paying them for their expertise, wisdom and systems thinking. This group of farmers was outstanding and extremely dedicated to the project.

Objective 2

Many kinds of information must be weighed to come to an action that prevents the build-up of weed populations without resorting to herbicide applications. For example, choice of crops in a rotation sequence may influence seed persistence due to predation, a greater potential for seed decay, or fewer resources for seed development (Liebman and Davis, 2000). Simulation models suggest that managing inputs to and losses from the weed seedbank is the single most important thing a farmer can to manage weed populations. Farmers must also decide which tillage method to use. Tillage depth has ramifications for both seedbank longevity and weed germination, with a trade-off between increased seed longevity, but decreased germination, when weed seeds are placed at greater depths (Mohler, 2001). The appropriate tillage depth for a given situation depends on the composition of the weed flora and characteristics of weed seeds, weed population density, crop competitiveness, crop sequence and weather patterns, among other factors. Cultivation, planting date, row spacing, soil amendments, cultivar selection, nearby vegetation, and landscape factors, can also affect weed life cycles.

Such complexity may tempt farmers to choose one management approach and stick with it. However, adaptive weed management is necessary to optimize weed control in variable environments (Orr et al., 2002) and present weeds with numerous, shifting stresses that prevent one weed species from dominating the agroecosystem (Liebman and Staver, 2001).

Decision-support systems (DSS) can provide a structure upon which farmers can base their decisions, and offer hypotheses for tackling weed management problems for which little information is available. Though DSS exist for conventional weed management programs, such as WeedSOFT, a computer application for herbicide recommendations (Mortensen et al., 2001), none are currently available for non-chemical weed management.

Because of the management expertise and observations gained from their livelihood, farmers are in an excellent position to critically evaluate both their own practices and ecological weed management theory to help formulate DSS for non-chemical weed management in Michigan field crop production systems. To keep the decision-making process transparent and open to feedback, the DSS will be text-based, with annotated links to the underlying ecological theories for each recommendation.

The research team, lead by Dr. Adam Davis, conducted an extensive data search through scientific weed science literature and combined these data with farmer inputs into a bulletin. The book was written and edited with graphics dominating the space.

The new title was Integrated Weed Management “One Year’s Seeding." (Michigan State University Extension bulletin E-2931). It ended up being 112 pages and sells for $10/copy. It can be purchased through the MSU bulletin office at http://web2.msue.msu.edu/bulletins/.

Objective 3

The number of low external input (LEI) and organic producers is growing both nationwide and in Michigan (Geier, 1998; MOFFA, 1999). As the number of acres coming under ecological weed management practices increases, the risks to human health and environmental quality associated with herbicides decrease (Liebman, 2001). Three barriers to the adoption of ecological weed management practices may be fear of failure due to lack of access to agroecological information (Lewis et al., 1997), the lack of agroecological information for specific weed species or life history stages, or a lack of experience in integrating non-chemical weed control tactics into the grower’s current or future production system (Jordan et al., 2000).

Michigan State University Extension (MSUE) has an ongoing commitment to providing agroecological information in a user-friendly format to producers wishing to learn about and adopt more sustainable farming practices. “Michigan Field Crop Pest Ecology and Management” (Cavigelli et al., 2000) is an example of the high-quality, multimedia educational materials that MSUE produces. Regional workshops with growers and MSUE agents will build upon this structure and give participants hands-on experience in using the Michigan field crop ecological weed management DSS. The workshops will encourage critical feedback of the DSS.

Objective 4.

Critical feedback may be this project’s most important product. The process of constructing a DSS and envisioning its use in a variety of scenarios will highlight weed management problems for which there are currently few answers. Some examples of poorly understood areas include:

cropping system influence on weed seed decay
cropping system effects on seed predation
interaction between tillage and seed mortality
genetic shifts in germination timing due to uncontrolled late-emerging weeds.

The DSS will be a work in progress, and investigators will publish a synthesis of the knowledge gaps identified by the DSS working group and workshop participants as a call to action to the weed science research community. Advances in ecological weed management will depend on such bottom-up guidance to ensure that research funds for non-chemical weed control are targeted where growers most need them.

The bulletin went through several extensive edits prior to being available to the public. We conducted three all-day meetings with 17 MSU Extension educators to review chapters and fine-tune PowerPoint presentations. These educators were invaluable to the book.

Nine educators scheduled winter meetings in their areas for training sessions about the book across the state. In the winter of 2005 we presented this book to more than 600 farmers; 145 of them were organic farmers. To date more than 600 copies of the 112-page bulletin have been sold.

We were able to successfully leverage $6,000 from MSU Project GREEEN (Generating Research and Extension to Meet Economic and Environmental Needs) to assist with printing the publication. We also leveraged this PESP grant from an NCR-SARE grant for $150,000 for three years to promote E-2931 throughout the North Central Region.

We were encouraged by Dr. Jim Kells, acting chairperson for MSU Crop and Soil Sciences, to submit E-2931 for the 2005 American Society of Agronomy (ASA) educational materials award program (publications greater than 16 pages). The results will be announced in the fall of 2005.

Approach and Methods

The project will begin by assembling a working group comprising up to 20 farmers, researchers and MSUE agents who are interested in, or already practicing, ecological weed management. Starting in November 2003 and continuing through Feburary 2004, this group will meet monthly for four eight-hour working sessions. Farmers will be paid as active participants, and all participants will be expected to attend each session to maintain continuity. The sessions will cover:

Weed germination/seedling survival/weed-crop interference
Inputs to the weed seedbank
Losses from the weed seedbank
Synthesis and preliminary evaluation of DSS

The emphasis on seedbank dynamics is supported by field experiments and models suggesting that, despite the widespread emphasis on killing weed seedlings, controlling seed production and survival is of paramount importance in regulating weed population size (Jordan et al., 1995; Davis et al., 2003).

Decision-support systems have a long history in agricultural production. One early and widely read version of a DSS was the Farmer’s Almanac, which has been in publication since 1818. This annual volume offers calendar-based prescriptions for timing planting and conducting other field operations. Since personal computers became widely available, computerized decision aids have proliferated. The most widely used weed management DSS in the north central U.S. is WeedSOFT (Mortensen et al., 2001), which uses an internal bioeconomic model to generate herbicide recommendations from information on weed community, weed density and crop characteristics.

This proposal’s investigators want an adaptive management tool that enables farmers to make an educated decision, rather than prescriptions. They will work to reach this through a transparent decision-making process, something that computer applications do not generally offer. Rather than hiding the decision-making algorithm, it will be used as a teaching tool, thus this DSS will be constructed as a decision tree. Decision trees are composed of hierarchical branching pathways with conditional recommendations at each branching point (Chechile and Carlisle, 1991). To avoid visual complexity when there are more than two choices at a given branching point, the decision tree will be divided into sections relevant to given weed life stages and important agronomic operations.

This decision-support system will be offered as a 50-page publication and as a CD-ROM. It will have two levels of detail. At a coarse level, decision support will consist of a separate one- to two-page decision tree for managing each of the four weed life stages: germination, seedling survival, fecundity and seed survival. Each tree will contain branches relevant to corn, soybean, small grains, forage legumes, potatoes and sugarbeets. All branches will be annotated to refer the user to details that will be presented in the publication’s second half. This section will present current information on weed ecology and management pertinent to the choices producers must make. It will also contain a bibliography for users wishing to further educate themselves.

Workshop organizers will start each working group session with a brief presentation drawn from the Michigan Field Crop Pest Ecology and Management (MFCPEM) bulletin and scientific literature. The group will then split into four sub-groups to run weed management scenarios. The sub-groups will focus on assigned situations dealing with specific branches of the decision tree, and then identify and discuss scenarios of their own choosing. Each group will assign one secretary, as records of the scenario process will be important in construction of the decision tree. Finally, the groups will re-convene to present their findings and any identified knowledge gaps.

The working group sessions will progress through the weed life cycle in field crops common to Michigan LEI and organic production systems, such as corn, soybeans and small grain/legume crop mixtures. A sub-group of participants will synthesize the results of each session into decision tree segments and present a draft of the tree to the working group for comment at the final session.

Investigators will revise the decision tree for clarity and accuracy and create annotated links to appendices containing background ecological theories underlying the various management recommendations. The decision tree will then be provided to the graphic designer who developed MFCPEM. He will create a final version for print and CD-ROM publication.

During winter 2004-2005, investigators and selected working group participants will conduct four regional workshops for Michigan growers and MSUE agents who are interested in ecological weed management. To attract a wide audience, Certified Crop Advisor and Restricted Use Pesticide credits will be offered at each workshop.

Each workshop will include a preliminary presentation on the ecological underpinnings of non-chemical weed management, followed by an introduction to the decision tree. Workshop participants will then divide into smaller groups to use the decision tree to address weed management problems. Following this exercise, participants will provide critically important information in the form of brief written critiques of the decision tree, both on ease of use and potential weaknesses.

Finally, the plenary will discuss the decision tree critiques and the outlook for ecological weed management research and applications in the field. The proceedings of these workshops will be synthesized into an article for publication in “Weed Science” highlighting bottom-up recommendations for high-priority research in ecological weed management.

Impact Assessment

This project’s primary goal is to reduce herbicide use in Michigan field crops by creating a decision support system that, 1) encourages growers who are new to ecological weed management to incorporate these practices into their systems, and 2) prevents the loss of certified organic acreage due to grower frustrations with the complexity of weed management in organic systems.

Investigators will conduct written surveys at the outset of the first working group session and at the end of the last working group session to examine whether the process of constructing the decision support system positively affected participants’ desire to continue improving their ecological weed management skills.

Prior to each of the regional workshops participants will be asked to complete a questionnaire that gauges their familiarity with ecological weed management, use of associated techniques during the 2004 field season, or desire to use them in the future.

Following each workshop, participants will complete a second survey to determine if the program increased their knowledge of, and interest in beginning or continuing use of, ecological weed management approaches. Following the 2005 field season, workshop participants will be sent a final survey to determine if they had changed their practices following the training and their satisfaction with the results.

Conclusion

This proposal represents a terrific model for change and adoption. We started the process by recruiting top-notch organic and conventional farmers. Then University researchers and Extension educators were integrated with this group to assure diverse responses would result around non-chemical weed control. Having a professional facilitator really helped these four meetings to fulfill our objectives. Here are a couple of quotes from farmers about the process.

An organic farmer said, “I wasn’t sure why conventional farmers were part of this process at the first meeting, but by the last meeting I realized they have similar weed problems as organic farmers and that we have a lot in common.”

A conventional farmer said, “I thought organic farmers were nuts, but I have really gained a lot of respect for what they are doing after these meetings.”

Following these four meetings, Dr. Davis used the farmers’ comments to compare those to the scientific journal information and 10 chapters for the book were developed. We incorporated farmer comments into the draft of the bulletin. Several of the participating farmers were telephoned and gave us quotes about different non-chemical weed control strategies. These quotes were used to begin each chapter. The chapters include: Chapter 1—Weed life cycles and seed bank dynamics; Chapter 2—Soil properties; Chapter 3—Soil organic amendments; Chapter 4—Tillage; Chapter 5—Integrated crop and weed management; Chapter 6—Crop rotation; Chapter 7—Physical weed management; Chapter 8—Herbicide management; Chapter 9—Biological weed management; Chapter 10—Prevention. Additionally, we developed an intensive weed ecology training program for MSUE educators. Three all-day sessions were scheduled in October 2004.

In October we finished the intensive training program for MSUE educators to learn about weed ecology. We developed PowerPoint presentations around the chapters created to present to the educators. We conducted three all-day meetings in East Lansing and 18 educators participated in these meetings. During the last session, we developed a statewide farmer training schedule.

In February 2005, the Michigan State University Extension bulletin E-2931 was printed and made available. In the winter of 2005 the MSUE educators and MSU researchers held nine farmer meetings across the state and more than 600 farmers attended these programs, and 145 of them were organic. To date we have sold more than 600 copies of the 112-page Integrated Weed Management “One year’s seedingY” (E-2931). This publication has been submitted to the 2005 American Society of Agronomy (ASA) Educational Materials Awards program for publication greater than 16 pages.

Due to the integrated and cooperative effort from this grant, we were able to leverage an MSU GREEEN (Generating Research and Extension to Meet Economic and Environmental Needs) grant for $6,000 to reduce the cost of the publication to $10/copy. Another grant proposal to NCR-SARE was approved for funding at $150,000 for three years to promote E-2931 throughout the North Central region.

Several universities will be using E-2931 for teaching purposes this fall. Therefore, we believe this proposal was extremely successful.

Appendix 1. Evaluation Summary. Integrated Weed Management, Session 1 (3 pp; 13KB )