RESEARCH PHILOSOPHY

The alfalfa leafcutting bee, Megachile rotundata and alfalfa flower.
photo by Karen Strickler

References

More about relationships

References

Thrips research
1997
1998

Some Reductionist research:

Sand for chalcid control?
Comparison of X-ray facilities 

References

References

References

Reductionist and Holistic Approaches to Research

In establishing the Pollination Ecology Program at UI, growers sought solutions to bee management problems that prevent them from producing sustainable bee yields. Traditionally, applied pollination research has focused on methods for rearing bees and controlling bee pests and parasites. Solutions for individual problems like chalkbrood, second generation bees, and pollen balls were sought in isolation from other problems, "holding everything else constant". This approach has led to good solutions, such as Vapona or Para-Gone for parasite control, bleach or paraformaldhyde for chalkbrood control, and various pesticides to control lygus, aphids, and other pests in the field. This is a reductionist approach to research. But from each of these so-called "solutions" additional problems arise: development of pesticide resistance, unwanted bee mortality, risks of toxic exposure to the applicator, increased costs of production.

I seek a paradigm shift in the way that research on alfalfa pollination is approached. I supplement the traditional analytical reductionist approach inherent in the scientific method with a greater emphasis on a synthesizing, holistic approach. Both approaches are important to fully understand a system and to have the greatest long-term impact on managing it, but the holistic approach is often neglected. I focus on systems thinking, emphasizing relationships rather than isolated entities. Instead of seeking solutions to individual problems I am interested in the tradeoffs that develop as one manipulates the system. To do this, my approach starts with modeling to understand how and why the components of the system interact and to identify gaps in our knowledge. My approach continues with monitoring and experimentation, and then returns to modeling to refine our understanding of interactions and to identify new questions to address. Seeking alternate methodology to address questions of interest about the system, and finding ways to explain them to academic and agricultural communities that are unfamiliar with these approaches has been a challenge.

The notion that my research is a paradigm shift comes from Capra, 1982.  Capra, author of "The Tao of Physics", argues that throughout science and society we are seeing a change from linear, mechanistic, reductionist, Cartesian thinking to holistic, systems thinking with an emphasis on interactions rather than isolated entities. I was introduced to Capra's book by Dr. George Bird, the nematologist at Michigan State University. He is an avid proponent of systems thinking, central to the concept of Integrated Pest Management (IPM).

Read more about the importance of relationships and interactions.
Read about Indra's Net.
Kuhn , 1962 first introduced the notion of paradigm shifts in science.  Read his explanation of why scientists in different paradigms have difficulty communicating, from "The Structure of Scientific Revolutions",

Top of page

What do I mean by a "paradigm shift"?

For some time after I was hired, some growers felt, and perhaps still feel, that my research program should focus on solutions to specific problems such as parasite control.  My Department Head did not understand my approach because he asked if I was certain that bee abundance was an important question in the alfalfa seed system.  I was, and I still am. One Parma Field Days (an open house for growers and administrators showcasing our research activities) he apparently heard something that helped him understand. I had been asked by the growers to determine if thrips, an insect that feeds on flowers, was causing loss of flower blossoms in alfalfa fields. In addressing this question, I told visitors that if enough bees are introduced into alfalfa seed fields so that pollination takes place rapidly, within a day or so of flower opening, then the flowers do not last long and thrips populations shouldn't have time to build up in them.  After my presentation my Department Head commented to me "I think I understand now what you are trying to do". This was apparently one of those "ah ha", light bulb moments that we all occasionally experience: a paradigm shift.

A reductionist approach might spend the greatest effort searching for a pesticide to control thrips. The holistic approach might put more effort into determining the underlying cause of the problem, and might seek management strategies that prevent the problem in the first place.

This is not to say that I eschew reductionist or short-term research. I have, in fact, taken on several projects of that kind at the suggestion of various members of the industry. What it means is that I am reluctant to take on projects that seem unlikely to succeed when the system is viewed holistically. Other researchers in my field have already taken a reductionist approach, and have tested many products for parasite and disease control. Products and management techniques are currently available for all mortality factors, though some may be time consuming or uncomfortable. New ideas worth testing are rare. In contrast, an understanding of the dynamics of bloom and its role in pollination and bee reproduction was lacking when I started my research as UI's pollination ecologist. What I have learned by taking this approach is that the dynamics of bloom and its interaction with pollination drives the alfalfa seed production system. We are now in a much better position to understand the tradeoffs inherent between bee management, pest management, and seed production goals.

More about why systems thinking involves a paradigm shift from isee systems.

Top of page

What do I mean when I say that the focus of research should be on understanding tradeoffs rather than solving problems? Here are a few examples. Among the barriers to a sustainable yield of bees are a variety of larval mortality factors that include pollen balls, chalkbrood disease, parasites and predators, second generation, and proportion of females. Take the last factor, proportion of females. Anyone familiar with the biology of tunnel nesting Megachild bees is aware that sex ratios are frequently biased toward males in these bees. Males are usually smaller than females, and there is a well-documented theoretical reason why more males should be produced when females are larger than males (Trivers and Hare 1976, Torchio and Tepedino 1980). So, our ability to alter the sex ratio of these bees is limited. However, we know that sex ratio can be manipulated to some extent, because more females are produced in larger diameter and deeper tunnels (Gerber and Klostermeyer 1972). Polystyrene bee boards generally come in two depths: 3" and 3.75". Based on bee biology, we expect that more female eggs will be laid in 3.75" boards than in 3" boards. Canadian growers use the deeper boards. But 3" boards are recommended for shelters in Northwestern USA. Why not switch to 3.75" boards to get more female offspring? Because pollen balls are worse in the Northwest than in Canada, so it takes longer for the boards to dry in order to be punched out. 3.75" boards are a solution to one problem, but it may not be a solution for the system.

How should we proceed to optimize this tradeoff? We need to know, for example, how much longer it takes to dry 3" vs. 3.75" boards in the USA, and what this costs in time and profits. Is the extra ¾" really a problem? Could a better method of drying the boards be devised? How would the cost compare with the savings in increased female offspring? What about using laminate bee boards? They are deeper than solid boards, and don’t need to be punched out, although they may be more expensive. At a different level, a more long-term solution would be to reduce pollen balls in the Northwest USA. A reduction in pollen balls would not only increase the percent live bees directly, but would also allow growers to use deeper nest blocks and thus get a higher proportion of females.

How do we reduce pollen balls? There is evidence to suggest that pollen balls, chalkbrood, and parasites are high because so many bees are being introduced into alfalfa seed fields. If numbers of bees were reduced, say if half the bees were introduced into fields than are currently used, mortality from all of these problems might be reduced, both because there would be more flower resources available to the bees, and populations in each shelter would be reduced. Would this allow growers to raise a sustainable yield of bees? Perhaps it would. But, consider the tradeoffs. Would profits be the same if half as many bees were introduced? The costs of pollination might decrease; that would be a plus. But profits might go down because fewer flowers might be pollinated at peak bloom. Profits might go down because pest populations would be harder to control, requiring more insecticide sprays and/or causing more insect damage. It would also take longer to pollinate the crop, leaving it vulnerable to shattering late in the season. How do we quantify this risk? What is the optimum solution for maximizing profits? How does the optimum solution change as bee costs and seed prices change?

To understand the tradeoffs and to find the optimum solution under a given set of conditions requires a cooperative, team effort. More research is needed, though that is gradually being accomplished. A better model of the system than mine is needed, one that includes the effects of temperature and pest damage. A thorough economic analysis is needed. This effort requires teamwork, cooperation, and open communication. In my experience this teamwork is rare in the culture of science, as well as in the culture of agriculture. It will require a paradigm shift of both growers and researchers to get there.

Top of page

Sustainability in the alfalfa seed system

In seeking a sustainable yield of bees in their fields, growers are defining the system only as far as their own field boundaries.  Suppose that we could predict the optimum number of bees to maximize profits when pest management costs are included in the analysis. This is the direction that my research has been aimed.  But, is that number of bees sustainable in a given alfalfa seed field? We don't yet know. Suppose it is not? Perhaps when talking about a sustainable yield of bees, we should redefine the boundaries of the system in which bees should be sustainable. If we define the system as the whole Northwest USA, than perhaps one can think in terms of developing a bee industry side-by-side with, but separate from the seed industry. Bees could be raised on separate fields designed to maximize bee yields rather than seed yields. This might mean planting a variety of flower species for pollen and nectar, and a variety of plants for cutting leaves. It might mean creating less dense shelters for the bees. Locally raised bees could be sold to seed growers who need greater bee populations for a lower price than Canadian bees because of lower shipping costs.

Alternatively one can seek sustainability in a larger system, one that includes Canadian seed growing regions. If one does this, then bee yields are already sustainable, baring some new threat to the system.  

How should the system boundaries be defined?  What is practical?  Growers and researchers should discuss this issue.

Top of page

Applied vs. Basic research

My research approach blurs the distinction between applied and basic research. Even basic questions about alfalfa pollination are tied directly to seed yield and may provide guidance for management practices. I have not limited myself to pollinator-plant interactions, but also address interactions between plant, pollinator and pest. I hope that my research will reach a broader audience than is traditional for a pollination ecology program with an agricultural focus. Pollination ecologists who study natural ecosystems should find the insights gained in this agricultural ecosystem (e.g., the negative relationship between flower standing crop and amount of pollination) relevant to natural ecosystems. Conversely, insights from natural systems can improve the management of seed fields. For example, the risk of lower seed yields due to increased self-pollination when floral resources on a plant are abundant, an important topic for natural plant populations in recent years, has practical implications for the timing of bee release in alfalfa seed fields.  The more people know about the alfalfa pollination system, the more resources will become available for optimizing the tradeoffs in the system. 

Criticism, misunderstanding, and controversy surrounding my holistic approach have characterized my program since its inception. I hope that this statement helps to clarify my thinking to those who did not understand.  Comments expressing agreement or disagreement are welcome.

References:

Capra, Fritjof. 1982. "The Turning Point, Science, Society and the Rising Culture".  Bantam books, N. Y.

Kuhn, Thomas, 1962.  "The Structure of Scientific Revolutions".  University of Chicago Press.

Gerber, H. S., and E. C. Klostermeyer. 1972. Factors Affecting the Sex Ratio and Nesting Behavior of the Alfalfa Leafcutter Bee. Washington Agricultural Experiment Station Technical Bulletin 73:1-11.

Torchio, P.F.and Tepedino, V.J. 1980. Sex Ratio, Body Size and Seasonality in a Solitary Bee, Osmia lignaria propinqua Cresson (Hymenoptera: Megachilidae). Evolution 34, 993-1003.

Trivers, R. L., and H. Hare. 1976. Haplodiploidy and the Evolution of the Social Insects. Science 91:249-263.

Articles about systems thinking from isee systems: