In recent debate concerning the relationship between agriculture and biodiversity a rather contentious issue seems to have emerged, the contradictory normatives involved in the so-called land sparing versus land sharing (wildlife-friendly farming) positions. This debate is not very useful if your concern is either the production of food or the conservation of biodiversity. On the other hand, if the concern is with the expansion of lands devoted to the industrial model of agriculture with the immense profits due to agents such as Syngenta, Monsanto, ADM and the other giants, this framing of the debate is indeed quite useful. Indeed, it is not surprising that it is in Europe that this debate has largely been concentrated, with Monsanto especially anxious to obtain access to European markets for their GMO products. The land sparing debate is a winner for them and, as has been noted by Chomsky and others before, setting the terms of the debate is far more important than the debate itself. And the land/sparing versus land/sharing framework is a marvelous example of setting the terms of the debate so that one side is almost certainly going to win (interestingly, even with this framing, it is not a slam dunk).
The notion that the framework you chose can frequently generate enormous bias is an accepted fact in serious sociological work, yet in politically charged situations (like the present example) partisans pretend that it somehow does not matter -- that “the science speaks for itself.” Of course the science does speak for itself, but what you hear that science say is filtered through your framing.
The classic example of this phenomenon is the often cited work of Kahneman and Tversky (1984). A group of experimental subjects was posed the following problem:
“Imagine that the U.S. is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientific estimates of the consequences of the programs are as follows: If Program A is adopted, 200 people will be saved. If Program B is adopted, there is a one third probability that 600 people will be saved and a two-thirds probability that no people will be saved. Which of the two programs would you favor?”
72% of the subjects chose Program A.
Alternatively a group of experimental subjects was posed the following problem, after the same initial setup:
“If Program C is adopted, 400 people will die. If Program D is adopted, there is a one-third probability that nobody will die and a two thirds probability that 600 people will die.”
22% of the subjects chose Program C (which, of course, is the same as program A).
In other words, neither your concern for people dying from the Asian disease nor your ability to read the relevant “scientific” evidence was the major determinant of the decision. Rather, the precise way in which the problem is framed is almost exclusively responsible. Many other examples of this framing problem can be cited. Indeed sophisticated psychological industries associated with marketing have evolved to take advantage of this particular human foible. Scientists seem particularly vulnerable because of their naïve belief that theory and data are politically neutral.
As an archetypical example the recent summary, Phalan et al., (2011) repeat the errors made by Green et al., in their original Science piece (Green et al., 2005; also see the critique by Vandermeer and Perfecto in the subsequent issue of Science and the response of Green et al. therein). In an attempt to summarize their position, they present the framework presented in figure 1, as if it were an ideologically neutral framing.
Figure 1
If you were concerned with biodiversity conservation and natural habitats, it would be difficult not to conclude that the land-sparing option is best. Ironically, in this same piece, the authors also make a strong case that ideological biases be excluded from the analysis of this important issue. Of course I disagree. Ideology is an inherent part of any academic analysis, from the choice of topic to the method of analysis to the standards of proof. The framework of figure 1 could easily be redrawn from a different ideological point of view, as I have done in figure 2.
Figure 2
Focusing on figure 2 it would seem foolish to conclude that land sparing (promoting expansion of industrial agriculture) was better for biodiversity conservation than wildlife-friendly farming (construction of high quality matrix). It is naïve to suppose that either figure 1 or figure 2 is non-ideological – both are. The relationship between ideology and academic pursuits is not easily cancelable, given the social nature of our species. The honest way forward is to be open and aboveboard about one’s ideological position. I, for example, believe that figure 2 is more consistent with my own ideology, born of the basic tenants of the Enlightenment (e.g. , social justice and participatory democracy), and I do not apologize for that position. However, while ideology certainly constrains the formulation and analysis of a problem, scientific fact within those constraints nevertheless drive (or should drive) normative prescriptions, a point hardly requiring assertion.
The land sparing/sharing framework is based on a set of sometimes explicit, sometimes only implicit, assumptions. First, among the frequently implicit assumptions is the need to produce more food globally, at least in the near future. The problem with this assumption is that the production of calories in the world already can accommodate the future population expected to stabilize at about nine billion sometime in this century. It is true that more than a billion of the current population remain hungry and malnourished, but it is widely acknowledged that such a state is due to the distribution of food, not to how much is available worldwide. Then, the assumption of the need to produce more food should be stated with the qualifier, “assuming the state of access to food remains the same or gets worse.” Focusing on the overall supply of food traps us into ignoring the current reason that one segment of the population goes hungry while another segment of the population arguably eats too much for its own health.
Second, the negative secondary environmental consequences, the collateral damage so to speak, of the industrial agricultural system are tacitly ignored. Just how much more of the world’s oceans are we ready to sacrifice to dead zones? How many more aquifers will we permit to be poisoned with pesticides? How much more soil will we allow to be washed away? While agroecological techniques are also implicated in some negative consequences, these are trivial compared to the massive consequences generated by the industrial system.
Third, the negative secondary social consequences of the industrial system are minimized. As documented in many studies, intensification along industrial lines can lead to further erosion of natural areas. For example, if a development plan promotes easy availability of chemical fertilizers in one area, it is most natural that migrants will thus be encouraged to come to that area and, if only natural areas remain uncultivated, there will be a tendency to claim those areas for agriculture. While such consequences could also be imagined under an agroecological framework, previous evidence suggests that the industrial system is far more likely to engender such patterns. On the other hand, those of us who promote agroecological solutions need also remain mindful of the sometimes technology-neutral secondary negative consequences (excessive nitrogen runoff is excessive whether it comes from the Haber-Bosch process or Rhizobium).
Fourth, increased per area productivity is always to be encouraged. Although this proposition apparently seems obvious to many people, it is not evident when viewed in a larger historical and social perspective. In many areas of agriculture today the most pressing problem is over production, and has been so for many years. Dealing with this problem, frequently under the generalization of supply management, has captured a great deal of analytical thought over the past century. That Brazil, in the early 20th century purchased and subsequently burned almost 50% of its national coffee harvest is only an extreme form. The famous coffee cartel of the Cold War, instituted as a part of the West’s struggle against international communism is another. And subsidy structures for maize production in the United States even today, allows giant grain companies to purchase maize at a price considerably below the cost of production, allowing grain exports to be competitive with even peasant producers in developing countries. While the desire to increase production by an individual farm may be universal, the knowledge that overproduction by all farms is destructive to all those farms is not information lost on farmers, even if some economists may have trouble fully appreciating it.
A somewhat more complicated issue is suggested in recent land sparing literature. The notion of “sustainable intensification” has emerged as a solution to the negative issues involved in standard assumptions of industrial intensification. The idea is that the environmental consequences of pesticides, excessive runoff of nitrogen and the like, can be mitigated through technological advance, yet high productivity can still be pursued. In many ways the notion of sustainable intensification appears almost the same as agroecological production (and thus the invention of a new term perhaps questionable). Yet both frameworks enter the debate with previous baggage. The sustainable intensification side of the debate emerges from the earlier ideas of intensification, which is to say substitute natural pest control with pesticides, natural soil fertility with chemical fertilizer and so forth. Most recently that debate has ben entered by giant seed companies eager to find entrance to reticent markets for their transgenic products. The agroecological side of the debate emerges from ecological ideas associated with natural systems agriculture where the structure of local ecosystems is taken as a system of cues as to how the agrecosystem should be designed and from farmer-originated ideas of agroecosystem structures that would sustain farming systems into future generations. The intensification framework carries the baggage of attempts to enter agricultural production as an extension of generalized investment opportunities, the agroecological framwork as an extension of attendance to natural processes – in short, the intensification argument gains most support from economics the agroecological from ecology. When they come together, as they appear to some workers, is that background baggage lost? Perhaps, but I doubt it.
Finally there is a slightly more complicated problem with the sparing/sharing framework. A tacit linear assumption is made regarding the ecological dynamics of biodiversity. If an area is divided into fractions that are devoted to 1) natural vegetation, 2) intensive agriculture, and 3) wildlife-friendly agriculture, the total biodiversity in that area is assumed to be given by B1 + B2 + B3, and total food production given as P2 + P3 where Bi is the number of species and Pi is the production contained in the ith habitat type. It is then assumed that B1 is greater than B2 or B3 and that P2 is greater than P3. Given these constraints it is a trivial exercise to construct, at least in theory, an “optimality” model that will almost surely demonstrate that reducing the area devoted to wildlife-friendly agriculture is the best strategy. As in the case of the Asian disease, the data don’t really matter much once the framework has been set. The problem with this assumption is that biodiversity does not really work this way in the real world. Determining the biodiversity in each of the three areas and then summing is known, from elementary facts of community ecology, to be misleading at best. Biodiversity does not accumulate in a linear fashion. Yet more important, this assumption fails to recognize the dynamic nature of landscapes with regard to the community structure contained therein. Spatial structure is now recognized as an important component of community dynamics, with metapopulation and metacommunity theory developing rapidly and largely negating the simple linear assumptions.
In the end, I argue that the framework of land sparing versus land sharing is more obfuscating than enlightening. I would urge that we begin with a simpler framework, rooted in the elementary ideas of landscape ecology. There are those areas that are largely untouched by humans and they exist in a matrix of human-dominated areas, the latter of which are largely agricultural. In most areas of the world the untouched areas are highly fragmented so it makes sense to talk of natural fragments in a matrix of agriculture. Local extinction is a natural process that cannot be detained by politics. Extinctions, local ones, will happen and they will happen more frequently in smaller fragments. The focus should not be on stopping local extinction, a fool’s errand, but rather on insuring that migration among fragments is sufficiently large that the equilibrium condition retains the species extant in the landscape. My position is that the untouched areas should largely remain untouched, and the main question is how to make the matrix as high quality as possible, with the definition of “quality” a negotiable issue, highly dependent on local factors. The overall landscape (or countryside) then becomes this simple combination of fragments of natural vegetation within the matrix, and the key guiding question should be “how do we make as high quality matrix as possible?” The combination of corridors of natural-like vegetation and archipelagos or biodiversity trampolines can certainly be part of that matrix, but most important is the promotion of agroecological techniques of food production that do as little harm to biodiversity as possible and that aim at sustainability as a goal that supersedes the productivist mentality.
Once we acknowledge that the quality of the matrix is the major goal in pursuing a biodiversity rich landscape (or countryside), it naturally follows that we ask who constructs that quality. On the one hand, it is clear that certain levels high up in bureaucratic structures determine many of the background characteristics within which the structure will emerge (e.g., “decision makers” will decide on zoning or tax structures or roadway construction), but the ultimate decisions about land use will normally be local, decisions that individual farmers, farming families and farming communities make. How to generate sociopolitical structures that will encourage ecological thinking in reaching those decisions is not completely obvious, but clearly necessary. Should one concentrate on those upper bureaucrats who appear to hold all the cards in decision making, or should one concentrate on the farmers who actually work the land and thus do the job of transforming the matrix? Do we try to influence the architect or the stonemason? It is a question deeply riddled with ideology.
Kahnenan, D., and A. Tversky. 1984. American Psychologist, 39:341-350.
Phalan, B., et al. 2011. Food Policy 36:561.
Green et al., 2005. Science 307:550.