A new study on strategies for feeding the world more sustainable with organic agriculture has been published in Nature Communications.
Prof. David Powlson, Lawes Trust Senior Fellow, Department of Sustainable Agriculture Sciences, Rothamsted Research, said:
“This is a serious and thorough piece of work that acts as a good basis for discussing future food systems. The authors are quite correct in saying that the whole food system needs to be evaluated, not just crop or animal yields in different agricultural systems. Two of the issues they consider – decreasing meat in human diets and reducing food waste – are both highly beneficial and policies are needed to push these, irrespective of any consideration of organic farming. Less meat almost certainly means reduced greenhouse gas emissions. The authors assume that, on average, crop yields in organic farming are 25% less than in conventional. This is probably a reasonable assumption, though will vary greatly between regions and systems. But I am not certain that they fully take account of the different crop mix in an organic system. These rely on crop rotations much more than conventional so (say) a 25% yield reduction in wheat in the organic system is inevitably accompanied by less wheat in the whole rotation as it is grown in fewer years. I therefore wonder if they under-estimate the extra land required for converting to organic.
“The authors draw attention to nitrogen supply being “challenging” in organic systems. They state their assumptions about inclusion of legumes (nitrogen-fixing crops) but I wonder if there is also some over-estimation of the nitrogen supplied from these to subsequent crops. First, they state in “Methods” that their modelling has assumed one legume crop every 5 years – this seems a small number. Second, to provide protein (to substitute for decreased meat) various pulses and beans will be grown. Whilst these do fix atmospheric nitrogen, much of the fixed nitrogen is removed in the harvested product leaving rather little in soil to supply subsequent crops. On the other hand, if green manures are grown as more effective means of getting nitrogen into the soil to supply subsequent crops, these will mean a loss of a cropping season so an even greater “yield gap” between organic and conventional. In some situations the green manure may be grown as a “cover crop” between other crops at periods when the soil would otherwise be bare – but I suggest this is a limited situation and, in any case, a very short-term nitrogen-fixing crop, whilst having some benefit for soil organic matter, is likely to provide a limited quantity of nitrogen for the next crop.
“For me the most significant finding is that, even under the most optimistic assumptions, an increased area of land is required if we depend increasingly on organic agriculture, leading to increased deforestation – this is shown starkly in Fig. 4. This means a large emission of CO2 as extra land is cleared, not to mention loss of habitats. Thus there is a need to prioritise the benefits and disbenefits or organic agriculture. Is the extra deforestation a price worth paying for (say) the decreased use of pesticides? This is fertile ground for discussion but I would say “no”. Of course we wish to use less pesticides and there are some problems with them but I would argue that with increased precision and good regulatory structures we can use them safely. Similarly with nitrogen surpluses. Of course we need to decrease them and fortunately there are emerging ways to manage nitrogen fertilizer better
“Regarding grassland, the authors state that they assume in their modelling that global grassland area will remain constant – all extra land being obtained by deforestation. But is it assumed that some of the grassland area that is now permanent be converted to grass within an organic rotation? If so this means (1) a loss of some permanent grassland that is valued for biodiversity and (2) major releases of nitrogen (as nitrate or gases) and CO2 when old grassland is ploughed. Additionally, Fig. 4 shows soil erosion to be increased under organic agriculture. This is clearly bad from many viewpoints – e.g. degradation of soil, sediments moving into water courses. There is considerable evidence that reduced tillage, especially if accompanied by retaining crop residues on the soil surface, can greatly reduce erosion. The approach is sometimes termed “conservation agriculture” (CA). It generally requires the use of herbicides and fertilizers so its adoption may be decreased if there were a major move towards organic farming.
“I think we should be grateful to these authors for a thorough study that draws attention to a wide range of important issues and provides a useful basis for debate. But my conclusion from the results of their modelling is that organic agriculture, whilst delivering some environmental benefits, causes different and probably greater environmental problems – especially increased deforestation and possibly increased soil erosion. It is good that they draw attention to the need to waste less food and eat less meat – but these are important issues in their own right and quite independent of organic farming.”
Dr Geoff Squire, Principal Scientist, Ecological Sciences, James Hutton Institute, said:
“The paper applies food-system models, interpreted through a range of nutritional and environmental indicators, to estimate the land needed to produce adequate food for a future global human population under various scenarios of climatic impact, food wastage and reduction of animal feed production on arable land.
“The paper’s broad, food-system approach to comparison of organic (no pesticide, no mineral fertiliser) and non-organic production, and its exploration of scenarios in which both can contribute, leads the argument away from the often polarised debate on whether organic production can feed the world. The models suggest that certain combinations of organic production area, reduction in food waste, and transfer of feed-producing to food-producing activities on arable land, coupled with greater use of nitrogen-fixing legumes can sustain the world’s 2050 population with no more than existing farmland.
“However, strategies to reduce waste and animal feed, together with greater adoption of legumes would be advisable in any future scenario, not just an organic one. In addition, the benefits to food security of reducing the area of land grown for alcohol and biofuel compared to food should also be taken into account.
“Moreover, and as the authors point out, estimates of production and impact would need to be tuned to specific regions and continents. It is difficult to see, for example, how a region whose soils have minimal phosphorus content can be anywhere near sufficient in food supply without mineral phosphate additions. Other uncertainties in the models (some considered by the authors) include inadequate data on nitrogen-fixation by crops and forages, unpredictable effects of expanding the use of some environmentally damaging methods in organic crop protection and the unsuitability of most existing crop varieties to reduced pesticide and fertiliser management during transition. Local tuning of the approach could limit or remove such uncertainties but would potentially lead to different outcomes than evident in the global scenarios presented here.”
Dr Martine Barons, Director of the Applied Statistics and Risk Unit, University of Warwick, said:
“This paper sets out to examine if a change to organic farming practices (defined as non-use of synthetic fertilisers and pesticides, use of crop rotation and closed nutrient cycles to promote soil fertility) can feed the expected world population. The scenarios considered are different proportions of organic farming practice, with its associated higher land requirement for the same yield, in combination with reduction in food waste and the reassignment of land currently used for fodder crops to human food production. The aim is to make the food system more sustainable. Intuitively, we understand that by ‘sustainable’ we mean that everything that is taken out is replaced, but how can we measure to what degree a system is sustainable? Can we have a measure that tells us we are not yet sustainable, but are more sustainable than last decade? The precise meaning of ‘sustainable’ envisaged is not made explicit by the authors.
“The food system does not exist in isolation and there is an increasing recognition of the food-energy-water nexus, as each of these systems impact the other. This paper assumes that the water and energy cost of these strategies are all similar, and concludes that the rise in land required could be managed provided all land currently used to grow animal feed is switched to human food production and food waste is cut to 50% of its current level.
“One aspect not explicitly considered here is the quantification of uncertainty in the system. The average levels of protein and calorie production for the 2005-2009 period is used but no mention of the variance and how that might affect our ability to feed the world. Even without a change in mean temperature, climate and weather can be highly variable, season to season with the obvious knock-on effect on food production. The possibility that yield might increase more slowly under climate effects is explored in this research, but not the possibility that yield might be reduced. Biodiversity is mentioned in the study, suggesting that a reduction in pesticide use might ameliorate the effects of deforestation to make way for farmland. This is a plausible assumption, but ecosystem services, such as pollination, require not only diversity but also abundance of pollinators. The drivers for these tend to be very local as the home range of these insects and animals is small and their lifespan short. In addition, some insecticides are necessary to control the deadly Varroa parasite within colonies of honeybees which are so important for pollination, so a blanket ban on pesticides is likely to be impractical. Likewise, one of the things modern farming does very well is reduce the volatility of food supply due to epidemics of fungus, insects and disease. Increasingly, smart technology is being used to apply remedies precisely where they are needed, reducing the volume of synthetic substances used overall. It might be argued that a move to organic farming might leave food crops vulnerable to catastrophic events, and so make people more vulnerable to hunger.
“Finally this research explicitly excludes the human element insofar as farmer, consumer and market behaviour and choices is concerned. These elements are part of the reason that the food system is a complex system – a system where the consequences of intervening cannot be fully foreseen, making decision-making difficult. The manner in which any move towards the scenarios suggested is implemented, the incentives for farmers, markets and consumers and the socio-political regimes in which the change takes place will be strong drivers towards success or failure and must be considered. The focus here is on providing the 2005-2009 average quantities of just two nutrients, calories and protein. There is no link to RDAs for these nutrients and whether, on average, current production is meeting the needs of the population for an active and healthy life. Under these strategies, the human dinner plate begins to look significantly different to what it is today, with very little meat of any kind and a huge uplift in legume consumption. The question is, how is this shift to be managed? What are the cultural and social norms which must be re-written, and how are governments to facilitate or implement such change?
“This paper makes a valuable contribution, under some fairly broad assumptions, to imagining a world where organic farming both provides sufficient calories and protein for the world’s population and is ‘more sustainable’. There are significant subtleties within the food system to be considered alongside these broad brush strategies in order to find a realistic way forward.”
Prof Sir Colin Berry, Emeritus Professor of Pathology, Queen Mary, University of London (QMUL), said:
“As for all models, assumptions have to be made and what weight you attach to which item can greatly change outcomes. The assumption that grassland areas will remain constant is a large one. The wastage issue is important but solutions (not addressed here) to post-harvest- pre-market losses will be difficult without fungicides for grains, as an example. Refrigeration has been proposed for this kind of loss but has its own implications in terms of energy consumption – and so on.
“A major point, addressed in other papers, is that there is a benefit conferred to organic farming by the fact that most farming is inorganic; pest burdens are kept lower by adjacent inorganic farm land, but increase significantly if organic areas are not surrounded by conventionally farmed land. The issue of nitrogen and its impact on yield is a complex one, with adequate levels affected by legume rotation and the use of animal, human, and plant waste. The use of animal and human waste can have further implications such as the transmission of disease (either zoonotic or man-to-man), and the modelled base setting of legumes forming 20% of crops feels rather optimistic.
“Additionally, some populations could do with more protein more protein to grow and develop normally, despite the models here requiring less animal protein.”
Prof. Les Firbank, Chair of Sustainable Agriculture, University of Leeds, said:
“Sustainable agriculture is not just about producing food that is high quality and in ways that work with the environment, it’s also about making sure that there’s food being produced to fulfil people’s needs. One of the question marks about organic farming is that it can’t feed the world. Here, Adrian Muller and colleagues have looked in to the question more closely using simulation models under a range of assumptions. They conclude that organic farming does indeed require more land than conventional methods, but it we manage the demand for food by reducing waste and reducing the amount of crops grown as animal feed, organic farming can feed the world.
“Their models can only be viewed as a guide: there are many assumptions in them that may or may not turn out to be true and all these scenario exercises are restricted by limited knowledge. Additionally these models are fairly simplistic compared to real life, but they are realistic enough to help formulate policy. But the core message is very valuable and timely: we as a society need to seriously consider how we manage the global demand for food.”
Dr Jeremy Franks, Senior Lecturer in Farm Management, School of Natural and Environmental Science, Newcastle University, said:
“The authors conclude that a “mix and match” approach, involving switching a proportion of conventionally farmed cropland to (registered) organic production, and reducing food wastage and livestock numbers will be needed to adequately feed the 2050 global population. Various combinations of the “mix and match” are considered, but none recommended.
“The model shows that switching to 100% organic production would involve converting substantial areas of forest into cropland. This it recognises is an unsustainable solution. Feeding a future population of 9 billion people by offsetting lower per hectare (organic system) crop yields by converting large areas of forest to farmland will add massively to atmospheric greenhouse gas concentrations, removing any possibility of hitting agreed climate change targets, and put further pressure on biodiversity and water cycling. Such a strategy can play only a small part in any sustainable global solution.
“There appears ample scope to improve the model upon which the “mix and match” estimates are based. For example, there is no explanation of why grass crops grown in arable rotations are not converted to human feed production in preference to deforestation, and the paper does not mention the stability of organic compared to conventional yields. Perhaps most striking, the paper does not discuss where the focus of research and investment should lay: in reducing the organic-conventional yield gap, developing solutions to reducing food wastage, or incentivising changes in diets?
“Technology already exists, and policy options are available, to reduce wastage and human consumption of livestock products – but waste reduction is inadequately resourced and governments are reluctant to intervene between the voter and their plate (even in countries where obesity is more of a problem than hunger), though they show little reluctance to intervene in food production and marketing systems (employing, for example, trade tariffs and non-tariff barriers, and quotas, area and livestock headage payments to farmers, agri-environment schemes and greening, to mention but a few).
“Perhaps more progress in future crop production systems would be achieved by dropping the binary division between organic and conventional production systems. Future production systems research needs to focus on identifying techniques and technologies which can successfully be transferred between organic and conventional systems, to raise the yields of the former and the environmental sustainability of the latter. This form of “mix and match”, which exchanges the best from both systems, will help policy makers and farm managers select more appropriately designed production systems which incorporate spatial awareness of existing ecosystem service provisions, local and national environmental sensitivities and global planetary boundaries.”
* ‘Strategies for feeding the world more sustainably with organic agriculture’ by Muller et al. published in Nature Communications on Tuesday 14 November.
Prof. David Powlson: “No conflicts of interest.”
Dr Geoff Squire: “My research is funded by the Scottish Government and the EU. I sit on no committees that impinge on the authors’ work’ and have no working contact with the authors. I have been a joint co-author with one of the authors P Smith (Aberdeen).”
Dr Martine Barons: No conflicts of interest.
Prof Sir Colin Berry: “Sir Colin consults for a number of agrochemical and pharmaceutical companies and for the MHRA. He is advises the European Risk Forum.”
Prof. Les Firbank: No conflicts of interest.
Dr Jeremy Franks: No conflicts of interest.
None others received.