Whether organic agriculture can feed the world is a controversial topic and it is the subject of much debate in recent literature. A number of studies have sought to answer this question by quantifying yield gaps between organic and conventional agriculture, with the recent estimates of reductions in yield for organic systems ranging from 9% to 25%. Diverse meta-analytical approaches have been employed to arrive at these values, in some cases presenting global averages and in others by separating the analysis by crop type, geographical region, or other moderating variables. Recent meta-analyses are reviewed here in Section 1
, with a brief discussion of the individual categories of cereals, legumes, oil crops, and tubers.
The meta-analytical design is useful in understanding the average value and range of yield gaps, and in many cases meta-analyses consider moderating variables, such as climate, fertilization rates, or rotational diversity, which allow for a more detailed discussion of how yield gaps might vary under certain conditions. However, even statistically rigorous and nuanced meta-analyses rely on the assumption that organic and conventional yields can be compared directly. Section 2
describes why conventional agriculture is not always a suitable benchmark by which to measure organic agriculture and then argues for an alternative approach to viewing the complex differences between conventional and organic systems.
Modeling approaches complement meta-analytical studies and allow for further exploration of how moderating variables affect yield. Multiple models of agricultural systems have been proposed, from input/output models to more complicated equations (Section 3
). We present a novel model describing cropping systems as processes that transform natural resources and inputs into yield (Section 3
). Each cropping system is viewed as a unique case in which inputs and their relative importance differ in comparison to alternative cropping systems. This model can be used to explain some of the observed variation in yield gaps between organic and conventional agriculture among crop types and environments (Section 3
). In addition, the model provides important insights into how to direct organic agriculture research priorities in the future.
3. Reframing the Yield Gap Debate
Meta-analytical studies are a valuable technique to summarize yield comparison studies, but we suggest that it is time to reconsider the question that they address and the benchmarks they use.
Meta-analytical approaches comparing conventional and organic yields worldwide often seek to contribute to the debate of whether organic agriculture can feed the world. However, that is the wrong question, or perhaps it is the right question at the wrong time. Today, when organic agriculture accounts for 1.2% of worldwide agricultural land [31
], it does not make sense to question whether it can feed the world. Perhaps in thirty or forty years, in a world where organic agriculture accounts for 40–50% of arable land, this question may gain renewed meaning. However, at present, the focus must be on questions relevant to the current state of affairs: how, and how much, can organic methods contribute to feeding the world? The “Can organic feed the world?” debate has thus far led to lively controversy, but few satisfying answers; reframing the question can move the debate toward concrete examples, as presented in Section 3
Using conventional agriculture as the benchmark against which organic agriculture must be compared falsely assumes that the systems have the same goals and values. When comparing a conventional wheat farmer’s yields in Germany of 7–10 Mg/ha, achieved using all available synthetic fertilizers and plant protection agents, with the 3.5–6.5 Mg/ha achieved by his neighbor with organic practices assumes that the two systems are essentially the same, except that organic agriculture uses non-synthetic inputs. In reality, however, the farmers are operating under distinct paradigms. Conventional and organic agriculture have different values, even when low-input practices are employed in conventional systems and the conventional-organic comparison becomes less of a dichotomy than a continuum. The conventional approach assumes that the production of food, fiber, and fuel must be maximized to satisfy the demands of a growing human population. Organic agriculture seeks to balance yield with other values, such as biodiversity and conservation of natural resources, as, for instance, required by Reg. (EC) 834/2007. The maximum yield achievable in the organic paradigm must necessarily lie somewhere below the conventional level if it is to leave room for other creatures to exist and to avoid exploitation of the natural environment. The abundance of many insect and plant species is negatively correlated with yield [11
] and a more recent study even suggested pesticide usage, increased application of fertilizers, and year-round tillage in intensive farming as plausible reasons for a 75% decline of flying insect biomass over 27 years [32
]. Organic agriculture must be judged not by the production-driven values system of conventional agriculture, but instead by standards that are consistent with its own values. The model that is presented in Section 3
shows how the distinct values of organic and conventional agriculture cause a divergence in inputs that accounts for a large proportion of yield gaps, as calculated by previous methods.
Furthermore, the geographical bias of the conventional benchmark used in many meta-analyses distorts the reader’s perspective on the debate over feeding the world. While some meta-analyses discriminate between geographical regions, others fail to do so, and thus often use the artificially high benchmark of intensive agriculture in the developed world. In Central Europe, yields that were achieved under the optimized intensive cultivation approach the theoretical maximum that was established by climatic conditions. Nonetheless, their direct contribution to eradicating world hunger is small, as many of the crops are commodities produced for the global commodity markets. Discussion about feeding the world based on these systems is misleading. Feeding the world primarily requires raising yields in subsistence agriculture, not incremental gains in the production of low-value commodities, and the conversion to organic agriculture in developing regions is predicted to make a greater contribution to global food security than conversion in Europe and North America [33
]. Farmers in Central Europe who produce commodities for the world market desire optimizing yields by all available means and will do so as long as it is affordable. Farmers in developing countries produce food to fill the needs of the local community, not global markets, and must do so on soil that is often more vulnerable than that in the global North. Section 3
describes how focusing on the transformation of natural resources rather than transformation of high inputs can greatly contribute to fighting world hunger by addressing yield gaps where they are most critical for food security.
5. Conclusions and Future Directions
Here, we have called for a reframing of the yield gap debate, changing the question under consideration from “Can organic agriculture feed the world?” to “How can organic agriculture contribute to feeding the world?” The model that is outlined in Figure 1
represents a novel approach in that it seeks not to quantify yield gaps with absolute values, but to explain and predict their magnitude under diverse starting conditions. This model does not conclusively establish the future role of organic agriculture, but it rather provides an indication as to how its research priorities should be directed in the future.
First, we need a new benchmark (Figure 2
). For any meaningful discussion of yield gaps between organic and conventional farming to take place, it needs to be clear where the upper boundary lies without violating the values of organic farming. It is clear that there are limits that we cannot exceed if the other goals of organic agriculture are to be pursued. However, the threshold of an ecologically sustainable yield may differ depending on the respective agroecosystem, contemporary technological capabilities, and the presiding social values. For instance, in a highly sensitive agroecosystem, a sustainable balance of the tradeoffs between crop yield and ecological impacts will presumably result in lower yields than in an agroecosystem with high environmental buffering capacity, where crop production may be managed more intensively without excessive negative ecological impacts. Alternatively, if society places a higher value on clean groundwater, for example, maximizing yields becomes less important than minimizing soluble nitrate emissions and the yield benchmark may be lowered. Currently, societal values allow for production to exceed the ecologically sustainable limits in the quest to maximize yields. However, those high yields must not be mistaken for an appropriate benchmark for yield comparisons of conventional and organic farming, as the latter system already integrates societal values, such as contributing to biological diversity and minimizing the use of non-renewable resources.
New ecologically sustainable benchmarks are not a fixed target, and as such they cannot be defined by an absolute value, but rather they must be established through consensus over the balance of tradeoffs that are considered to be acceptable by the respective society. Debate over this approach and the alternative paradigms has begun, for example, with the concept of ecological intensification, in which ecosystem services partially replace reliance on anthropogenic inputs as a source of crop productivity [69
]. Recent metrics that were developed for sustainable intensification are particularly useful when they go beyond the original efficiency-focused framework and emphasize various aspects of sustainability. Other useful frameworks include measuring environmental impacts on a yield-scaled basis [70
] and assorted “eco-efficiency metrics” [71
] that integrate multiple criteria that are related to sustainability and productivity [72
]. However, the strengthening of this debate must continue. Where is the acceptable upper limit, or phrased alternatively, how large of a potential yield gap are we willing to accept in order to avoid the negative environmental tradeoffs that are associated with high conventional yields?
Second, we need to set new priorities for developing agriculture to focus on raising the lowest yields rather than the higher ones, especially in organic systems. Best-practices organic agriculture is already highly refined. Seufert et al. [3
] found that the yield gap is lower when the comparison is between the organic and conventional systems that both use the best respective management practices. Ponisio et al. [4
] likewise found that multi-cropping and crop rotation in organic systems reduced the yield gap to 9% and 8%, respectively, as compared to 14% when the organic systems did not use these best-practice techniques. In organic systems that use best-practice methods, yields might already approach the current ecologically sustainable maximum. Rather than investing resources here, where they would bring only incremental gains, the organic branch should prioritize cases in which the yield gap is largest, which is an example of sustainable intensification [74
]. Technologies developed and adopted in organic farming systems are already highly attractive in developing countries: they require little capital investment or technological know-how, conserve resources, such as soil and groundwater where they are especially vulnerable, and reduce risks as compared to less-diverse systems [33
] (Figure 3
). The transfer of best-practice knowledge can help to raise yields under these conditions and in underperforming organic systems in developed countries as well.
To do this, we must make a third change: redirecting the agricultural research focus from maximizing yield due to transformation of inputs towards maximizing yield due to the transformation of natural resources. Breeding for rhizosphere traits can play an important role here [75
]. Root system architecture and interactions with beneficial rhizosphere microorganisms strongly influence nutrient uptake, and the breeding for these traits can create cultivars that are able to make full use of inherent soil fertility [76
]. Breeding for high nutrient use efficiency, encompassing utilization efficiency as well as acquisition efficiency and translocation efficiency, can help in transforming a greater proportion of this natural fertility into yield, especially if agroecosystem-specific characteristics are taken into account [77
]. At present, today’s conventional breeding is often focused on developing cultivars that transform synthetic inputs into yield under intensive management practices. Hildermann et al. [21
] showed that conventionally bred wheat cultivars out-yielded organically bred cultivars under conventional management, but that there was no yield difference under organic conditions. Organic plant breeding can present an attractive alternative to the conventional model by developing cultivars that are suited to low-input conditions for use in developing countries, and developing cultivars that maximize rhizosphere interactions that transform the natural capital of the soil into yield should be a focus for conventional and organic systems alike.
The Green Revolution worked best where industrial techniques could be implemented, such as in parts of India and China, but it was less successful in parts of the African continent where these methods were impractical [33
]. Pretty and Hine [78
] point out that expanding sustainable agriculture in areas with low food security will do more to combat world hunger than in attempting to increase total food supply through industrialization of agriculture. Organic methods can thus substantially contribute to feeding the world, as they can increase yields where those increases lead to food security and self-sufficiency for the farmers and local communities. As organic agriculture seeks to set its research priorities for the future, however, the focus should also be on raising below-average yields in developed countries, by addressing the factors that most limit yields. Nitrogen availability in cereals and tubers, weeds in grain legumes, and insect pests in oilseed rape are a few examples of research needs that will help in substantially increasing organic yields. By establishing appropriate benchmarks, re-prioritizing research needs, and focusing on transforming natural resources rather than inputs, organic systems can raise yields and thus play an ever-greater role in global sustainable agriculture and food production in the future.