Can Organic Agriculture Feed The World?
- K.W.T. Goulding (1) and A.J. Trewavas (2), AgBioView, June 23, 2009 http://www.agbioworld.org
(1 Department of Soil Science, Rothamsted Research, Harpenden; Herts AL5 2JQ; 2 Institute of Molecular Plant Science, University of Edinburgh, Edinburgh EH9 3 JH., Scotland.)
Abstract: In a recent publication, Badgley et al. (2007) claimed that organic farming, if used worldwide, would provide sufficient food for a growing world population. This claim was based on a literature survey of two kinds: (1) A comparison of organic and conventional yields, assembled, so far as one can judge, from a mixture of largely research experiments of rather variable quality and sometimes unpublished material. (2) An assessment of nitrogen (N) fixed by legumes from published literature. The two were then combined to calculate, incorrectly in our view, potential food production.
We have examined the literature basis of these claims particularly on wheat. There are many omitted references that indicate organic yields are substantially lower than Badgely et al. (2007) indicated. There are calculation errors in some of the references used by Badgely et al., (2007). Also Badgely et al., (2007) are equating organic procedures only with the use of either manure or cover crops and are ignoring certified organic procedures that prohibit synthetic pesticide use. We have also examined the claims by these authors that there is sufficient N fixed to provide for fertiliser and have found that mineralisation levels are wrongly equated with the N appearing in seed yield. We agree with Badgely et al., (2007) that maintenance of organic material in soil is important but consider that this is not a specific organic procedure. There would be insufficient food for the world population provided by global organic farming.
INTRODUCTION: In a recent publication, Badgley et al. (2007) claimed that organic farming, if used worldwide, would provide sufficient food for a growing world population. This claim was based on a literature survey of two kinds: (1) A comparison of organic and conventional yields, assembled, so far as one can judge, from a mixture of largely research experiments of rather variable quality and sometimes unpublished material. (2) An assessment of nitrogen (N) fixed by legumes from published literature. The two were then combined to calculate potential food production. We have four criticisms of the paper: (1) A very inadequate literature search-we have not instigated a rigorous search but easily found many scientific references that were not included by Badgley et al. (2007) and change the perspective on organic yields. (2) A failure to establish the credibility of published material. (3) An incorrect equating of N released by the mineralisation of organic material in soil with that taken up into grain. (4) No consideration of the economic base that underpins farm practice.
In addition, Badgley et al., (2007) stated that they were evaluating the potential contribution of organic agriculture to the global food supply but then admit that they were not referring to any particular organic certification criteria. It is clear from their paper that they have arbitrarily identified organic farming with just the application of manure or the use of cover crops. Neither of these procedures is particularly organic; they are used by many “conventional” farmers in the UK, for example, particularly those on mixed farms. On the basis of the criteria used by Badgley et al. (2007) for organic farming, most UK farms of all types would currently qualify as 'organic'. Also, from a UK perspective, it is difficult to identify any particular farm with what is claimed by Badgley et al. (2007) to be conventional farming. UK farmers practice what can be called Integrated Farm Management (IFM) to varying extents, with short or long rotations, no-till or min-till (conservation tillage) sometimes combined with ploughing, livestock fed on grass, grass/clover or other forms of silage, and returning variable amounts of manure, slurry and bought-in composts and other “wastes” to the land. The actual practices result from long experience by the farmer and with some experimentation and are usually site specific. Managerial skill is crucial; most UK farmers do take a holistic view of their farm, but the over-riding criterion is the market price for produce, set against the cost of production (Trewavas, 2004). The primary goal of any successful farmer is to sustain both income and the quality of their land.
If arguments are to be made about feeding the world from organic farming then the primary concern must be the yield ratios of organic/conventional for the major cereal crops. Whether organic cabbage, tomatoes or even oats or apples for example can match conventional yields and used by Badgley et al., (2007), is of relatively little importance, so we have not considered these at all. Wheat is grown on 220 million ha worldwide, a substantially greater area than that of corn (maize) or rice, is tolerant of arid climates and, containing more protein than corn or rice, is one of the primary food staples. Therefore, for brevity and simplicity, our critical assessment of the claims made by Badgley et al. (2007) is limited to wheat. However the criticisms and serious omissions that we describe are likely applicable to all the data provided by these authors.
The importance of establishing likely organic yield is highlighted by the strong support for greatly increasing organic farming world-wide by the International Assessment for Agricultural Science and Technology in Development (IAASTD). The claims by Badgley et al., (2007) may have underpinned this recommendation (Riverra-Ferre, 2008) given that the senior author on Badgely et al., (2007) was also a member of IAASTD.
THE PROBLEMS WITH USING EXPERIMENTAL YIELD MEASUREMENTS.
Badgley et al., (2007) have relied heavily on data from experiments and, while this is better than anecdotal information, which they have also used, there are at least three important caveats:
(1) Experiments lack the economic drivers that determine what farmers actually do; experimental station staff income is unaltered and they don't lose their jobs if the experiments fail.
(2) While yields from experiments are useful indicators of potential yield they are obtained under closely-controlled and near-ideal conditions with scientific expertise that does not exist on practising farms. The difficulties of transferring experimental results to actual farming practice have long been recognised (Davidson and Martin, 1968).
(3) Badgley et al., (2007) made no attempt to rate the credibility of the ratios of organic to conventional yield they presented. Measurements made in just one year are presented as having the same credibility as those made over 10-25 years or more.
Comparisons between farms
The correct way to compare how organic performs as against other forms of farming is to assess the averages from many farms over large areas and several years. Although this information is available in the UK it does not seem to be available elsewhere. Figures from Elm Farm Research Centre, the primary organic research institute in the UK, indicate that the average UK organic winter wheat yield (assessed from data on many farms) is c. 4 tonnes/ha but averages c. 8 tonnes/ha for “conventional” agriculture; a ratio of 0.5 compared to Badgley et al.'s ratio [what we refer to throughout as the 'BR ratio'] of 0.93 (Lampkin and Measures, 2001; Welsh et al., 2002; Wolfe, 2001). Record conventional wheat yields are c.14-15 tonnes/ha and, occasionally organic yields have reached 7 tonnes/ha, again suggesting an organic/conventional ratio of 0.5. UK organic associations make lower yields a virtue, claiming they are the result of working with the grain of nature as opposed to conventional production that dominates nature.
Individual farms vary enormously in yield due to soil type, land history use, crop variety, and topography as well as farmer skill. Thus when small numbers of organic farms are compared to other larger groups of conventional farms, anomalous farms in the small group can skew the comparison. Bender (2001) compared traditional Amish and conventional farms and clearly showed they could be more clearly separated by their geological history than by farm type.
To illustrate in extremis the potential flaw in comparison of a few farms with a local average, if one half of the yield of an area chosen as apparent control results from one very large, poorly-run farm then every other farm in the area is better regardless of farming method. Lockeretz et al. (1981) [BR 0.57] made observations from the same farming area as Roberts et al. (1979) [BR 1] but quoted conventional yields 50% higher. But Lockeretz et al. (1981) at least made the effort to match conventional farms with organic farms on the basis of size and soil quality, although they provided no statistics on the extent of variation. However both papers were mainly concerned with the economics of conversion to organic farming. Collating yields was incidental and the authors only quoted a few years of data, which was not intended for a direct comparison of efficiency in production.
Statistical methods were developed in the 1930's by Fisher to enable researchers to measure differences in yields between farms and fields; their use should be mandatory but seem to find little or no place in the organic/conventional lists provided by Badgley et al., (2007). Comparisons would be greatly helped by the selection of small groups of farms at random and indicating statistical variations in yield. Entz et al., (2001) surveyed 13 organic farms and found a six-fold variation in wheat yields. Kitchen et al., (2003) reported four-fold variations in conventional yields and even higher variation for organic yields.
Stanhill (1990) published 10 years of data from the 25-year organic Haughley experiment (organised by Eve Balfour, founder of the Soil Association), which ended when there was a clear continual decline in organic yields in this closed cycle experiment, resulting in much lower organic yields than local 'conventional' county averages. Oddly enough this data finds no place in Badgley et al., (2007) although they use other papers, e.g. Roberts et al. (1979), Koepf (1976, 1981), that compare organic farms with local averages. Furthermore Badgely et al., (2007) do use other older references from Stanhill (1990) and reference Stanhill (1990) as well.
Annual variations in wheat or bean yields were measured at Haughley of up to 4-fold in adjacent years. Single year measurements can be selected from the data which give an organic/conventional ratio of either 1.25 for wheat or 0.2 for beans; or merely taking two years in succession, a ratio of 1.15 or 0.36 (years 1957-1959). But crucially the ten year average of organic/conventional ratios are 0.81 and 0.61 for wheat and beans, respectively. Selecting organic/conventional ratios made on one or two years data lack not only scientific credibility but can clearly mislead as to the truer state of affairs. Even averages of long-term measurements can disguise trends such as the decline in the closed cycle at Haughley. This decline was caused possibly by a cumulative mineral deficiency. Produce sold off-farm contains minerals that must be replaced or the soil is mined.
A recent careful two year study comparing organic and 'conventional' produce found no significant difference in mineral content or mineral retention in feeding experiments (Kristensen et al., 2008).
A MORE DETAILED INSPECTION OF THE BADGLEY ET AL. (2007) LISTS OF WHEAT YIELDS
In this section, comment is made on the individual papers used by Badgley et al., (2007) and also lists those not included.
Leu (2004) is a political polemic by the President of an organic association. Only a one-year wheat yield is reported by Leu, with a ratio of organic to conventional of 1.45. However no information is presented about the farm, the work is unpublished; is it one of a series; is it reproducible? Leu omitted the data from Elm Farm Research Centre, listed above as did Badgley et al., (2007).
Also omitted by Badgley et al. (2007) were:
o the 10-year Boarded Barns Farms study, with certified organic and conventional matched fields, with 7 annual wheat yields that gave organic/conventional ratios of 0.61, 0.75, 0.64, 0.65, 0.7, 0.61, 0.51 and an average conventional yield of 8.3 t/ha compared with an average organic yield of 5.1 t/ha (Higginbotham et al., 2000; Trewavas, 2004);
o the measurements on 9 spring wheat cultivars under conventional and certified organic management over a three-year period in which grain yields fell by 47 to 56% when managed organically, giving again a ratio of about 0.5 (Poutala et al., 1993);
o the experiments on 27 varieties of spring wheat (bred over 113 years of wheat breeding) and grown under organic and conventional conditions in which organic to conventional ratios ranged from 0.46 to 0.69 with a mean of 0.61 (Mason et al., 2007).
o the investigations on matched organic and conventional farms with ratios of 0.83, 0.37, 0.53 and 0.16 by Ryan et al. (2004); those of Entz et al. (2001) who observed an average ratio of 0.8 comparing 13 organic farms with a large number of conventional farms; Kitchen et al. (2003) who reported that grain yield was variable but significantly lower in 11 out of 14 organic farms, with an average ratio 0.75; Grimm (1988) who compared 40 'alternative' farms with over 800 conventional farms to give an average ratio 0.69; Bochenhoff (1986) who found ratios of 0.7 for winter wheat on 145 farms and 0.72 for spring wheat on 52 farms; Wookey (1987), who found a ratio of 0.6; Stoppler (1989) who compared 23 varieties of wheat in organic and conventional agriculture and found ratios of 0.75, 0.95 and 0.73, with an average of 0.83; Dlouhy (1981) who averaged yields over a range of N applied and constructed equations from which yields could be deduced, obtaining a ratio of 0.58; Vereijken (1986), who examined the Nagele experiment containing 3 years of data comparing conventional with biodynamic and integrated farming giving ratios of 0.75 and 0.67, respectively.
Finally, also missing was the review by organic researchers with references that concluded ”In Europe, arable crop yields in organic systems (including wheat) are 60-80% those of conventional systems” (Stockdale et al., 2002).
Surprisingly none of the above figures found their way into Badgley et al. (2007) and yet there are more ratios in total in the above paragraph than provided by these authors. We have for example omitted the individual 27 ratios in Mason et al., (2007). Our search has not been rigorous and we noted many references to other investigations in the papers we found. The omission by the authors of these references in peer reviewed agricultural journals, which in general give ratios that are much lower than those they quote, diminishes substantially any claim to have provided balanced information.
Many of Badgley et al. (2007) elderly references often using biodynamic agriculture are taken directly from a review by Stanhill (1990). How much credence should be placed on results in that review when they stem from Germany in the 1930's, as quoted in Koepf et al. (1978) is moot. Strong ideological support for biodynamic farming was given by the German government of that time. Furthermore many of the comparisons relate to farming methods that would now be regarded as completely out-dated. Convenional yields of wheat have doubled in Europe since the 1930s.
The Broadbalk data indicate important conclusions that can be drawn about manure and wheat yields.
Leu (2004) referred to claims by George Monbiot that wheat grown on the 165-year old Broadbalk Winter Wheat Experiment at Rothamsted with livestock manure has produced consistently higher yields that that given fertilisers. This comparison relates to two experimental treatments on Broadbalk, one given only moderate amounts of fertiliser N and the other 35 tonnes/ha cattle manure. Badgley et al. (2007) refer to these experiments twice in their article under different references (e.g. Jenkinson et al. 1994) and used them as a supposed comparison of organic to conventional. The BR ratio quoted was 1.15.
On the contrary the yields from manure have matched those from moderate amounts of fertilisers within statistical variation over 160 years (Goulding et al., 2008) and are currently about 7 tonnes/ha; near the recent UK conventional average. However, this level of manure application is about 3 times the amount (maximum 1.2-1.4 livestock /ha) permitted by organic regulations. These regulations are imposed to minimise external impacts of the farm on N run-off, for example, and to eliminate dependence on external sources of feed. Also (1) the manure-only experiment is not organic because substituting manure for fertiliser is not certified organic practice, besides which pesticides are used on Broadbalk, and (2) the fertilizers applied to the conventional treatment are not optimal. The best yields obtained on Broadbalk with fertilizers are c. 11 tonnes/ha, giving a “supposed organic”/conventional ratio of c. 0.7. Eleven tonnes/ha can also be achieved with manure, but only if extra nitrogen (N) fertilizer is applied in spring, giving very poor N use efficiency and very high losses of nitrate. If the permitted manure input is based on 1.2 livestock/ha this is equivalent to about 72 kg N/ha and about 12 tonnes/ha manure weight. From the known effects of N upon wheat yield this would provide about 4 tonnes of wheat/ha; approximately the yield currently obtained on well-managed organic farms (Trewavas, 2004).
However, the Broadbalk experiment makes a very important point. Animal manure contains free nitrate and ammonia from partly degraded organic N in plant material. Supplying an amount of manure containing soluble N equivalent to that of N fertiliser will likely give an identical yield as found in Broadbalk. This important observation is entirely absent in Badgley et a., (2007) who have usually failed to identify amounts of manure used. Also, although Badgley et al. , (2007) and Monbiot focus on the manure-only part of the experiment, the fertiliser treatment is equally important because it has continued for over 160 years with identical yields to the manured treatment and with levels of organic matter in the soil less than one-third that of the manure-only treatment. These results contradict a common claim that 'conventional' farming is somehow unsustainable and suggest instead that poor management, not technology, causes most of the recognised difficulties with conventional agriculture. But the Broadbalk experiments, started in 1843, have used winter wheat continuously; whereas conventional farming practice is to rotate crops, a procedure known to reduce pests and disease. Crop plants like other plants interrogate the soil in which they grow. If the untreated soil was regarded as 'unhealthy' then equivalent yields would not be obtained.
Another crucial point to note from the Broadbalk and other long-term experiments at Rothamsted, is that they were started in the mid-19th century because there was insufficient manure, even then, to provide nutrients for crop growth. To provide 35 tonnes of manure/ha requires 3.5 adult cattle/ha. There are about 4 million adult cattle in the UK and 4 million ha of arable farmland. (Respective figures in the USA are 100 and 200 million). Thus to provide sufficient manure for high wheat and other crop yields would necessitate increasing the numbers of cattle about 3.5 fold (7 fold in the USA). Their winter consumption of corn and wheat would increase by the same amount and the land devoted to food crops would have to decline. Methane production, a potent greenhouse gas, would also increase. The much larger numbers of cattle would see prices of both meat and milk drop substantially thus increasing the likelihood of increased animal fat in the diet.
The references listed by Badgley et al. (2007) on wheat.
Granstedt and Kjellenberg (1996) [BR 0.8] state “ The fertiliser application rates for the various treatments were adjusted to bring about comparable yields”. Raupp and Konig (1996) [BR 1.01] “In order to achieve comparable yields with organic and mineral fertiliser, composted manure had to be used in relatively high amounts” sometimes up to 60 tonnes/ha. In view of these statements, yield ratios are clearly obviated. The actual aim of these experiments was to compare food quality, not yield.
Mader et al. (2002) [BR 0.87]. A long-running Swiss experiment applied manure at an organic regulation level of 1.4 animals/ha (and yielded about 4 tonnes/ha) but the fertiliser application of 125 kg N/ha was well below the conventional optimum. The conventional Swiss average over 20 years according to the FAO is 6.5 tonnes/ha which would give a ratio of about 0.6. Mader et al. (2002) state that “cereal yields under organic management are typically 60-70% of those under conventional management”.
McGuire et al. (1998) [BR ratios 0.98, 0.93, 0.83, 0.81]. Investigations examined cover crops on dryland sites over winter but pesticides were used to control weeds and thus organic procedures were not followed. The elimination of weeds by pesticides is known to increase yields substantially.
Smolik et al. (1995) [BR ratio 1.09]. Badgely et al., (2007) have calculated the ratio incorrectly. Smolik et al. (1995) quote figures for the conventional wheat yield of 6403 bushels/acre (4.3 tonnes/ha) and the 'organic' system as 5606 bushels (3.7 tonnes/ha) giving a ratio of 0.87. Although the experiment lasted for 7 years no statistics are supplied.
Stonehouse (1991) [BR ratio 1.14]. Stonehouse (1996) has described his actual research. The primary concern was the economics of converting to organic farming using 9 farms. Two yield ratios were provided; organic/ conventional and organic/ reduced input (i.e. integrated) 0.81 and 0.71, respectively. Stonehouse (1996) states that “no attempt is made to draw any definitive conclusions” by comparing organic with other farm systems indicating awareness of how limited farm numbers can seriously mislead.
Lockeretz et al. (1981) [BR 0.57] matched conventional with 14 organic farms managed by organic farmers without ideological motivation. However, this was again primarily concerned with economic aspects of organic farming rather than yield. No yield statistics were provided.
Nguyen and Haynes (1995) [BR 0.68] examined three pairs of farms and provided one estimate of a ratio of 0.68 for biodynamic/conventional and a separate one of 0.68 for organic/conventional. The paper comments that soil nutrient reserves are being depleted rapidly at one biodynamic and one organic farm.
National Research Council (1989) [BR 1.02] reviewed 'alternative' agriculture. The aims of 'alternative' systems, as described on page 27, would be subscribed to by many farms in the UK (see for example www.leafuk.org) which are not organic. In addition, the organic/conventional ratio refers to the yields of two different rotations both of which received insecticide and are not therefore organic.
Berardi (1978) [BR 0.83] published the results of a very limited 2-year study focusing primarily on economics and energy. Only 10 organic farms could be found in the locality and conventional farms were chosen from a project list. The value of this work is limited by the few farms and the lack of a comparative selection.
Smith et al. (2007) reported results from the Kellog Experimental Station [BR 0.56, 0.55 ] of particular crop rotations repeated every four years , started in 1992. Compared to 3 other management systems, conventional, no-till and low input, yields were lower in the organic system in the winter wheat phase of the rotation.
Levi (1979) [BR 1.09] reported data from a single 20-ha farm owned by a convinced organic farmer who was prepared to accept the loss of a crop rather than compromise organic principles. However the wheat yields reported by Levi at 3.85 tonnes/ha lie within the present day range for Israeli conventional yields (from c.2-9 tonnes/ha), but whether the organic/conventional comparison would have relevance to the present day is difficult to ascertain.
Liebhardt (2001) [BR 0.97] is no longer available but may refer to some of the figures used by Badgley et a., (2007).
Conclusions on the Badgley et al. (2007) lists of wheat yields. The data from these authors was poorly researched and heavily biased towards the conclusion that organic and conventional yields were about the same. The extra references we have included do not support that view.
If sufficient but an unsustainable level of manure is added, similar yields will probably be obtained. Entz et al. (2001) report that the 13 organic soils investigated contained anywhere from 34-246 kg N/ha; the average 92 kg N/ha was higher than the provincial average. At 246 kg N/ha no doubt high wheat yields will have been obtained but is unsustainable because the N leached from these farms would have been large and unacceptable. Entz et al. (2001) also reported that in the oldest organic farms (70 and 30 years old), phosphate levels were seriously deficient.
LEGUME NITROGEN FIXATION.
In a second part of Badgley et al. (2007), it is claimed that legume N fixation would be sufficient to replace the current use of N fertiliser. Legume N fixation is a biological process subject to numerous uncontrollable climate variables such as rainfall and soil temperature, also site-specific variables such as soil micronutrient status.
Reliable representative values of annual fixation rates for different countries still cannot be set and vary over an order of magnitude or more (Smil, 2001). Using two crops (one to fix N and one to provide food yield) introduces variation in inputs and so likely greater variation of the final cereal or vegetable yields. Current predictions that climate change will result in more variable weather in the future thus increases risks. The data used by Badgley et al., (2007) are largely experimental and subject to the same caveats indicated above for crop yield.
How much N is fixed in the course of ordinary and, in many cases, traditional farming is the crucial issue.
Badgley et al. (2007) reference Kramer et al. (2002) as indicating that cover crops can provide the same yield as mineral fertiliser. However Kramer et al. (2002) state that this was achieved using a mixture of a vetch cover crop with 330 kg N/ha from turkey manure, totalling 435 kg N/ha; compared to a conventional fertiliser application of about half that value. The turkey manure is equivalent to about 47 tonnes of cattle manure; an amount that cannot be supplied sustainably. Badaruddin and Meyer (1994) report that legumes produced N equivalent to 75g N/ha, but the amount will depend on the level of N fixation in the cover crop and this can be extremely variable with experimental station results only a loose guide.
Although Badgley et al., (2007) use figures for N inputs apparently only for cover crops, the standard organic procedure in the UK and, so far as is known in Northern Europe and many other countries, is two years of fertility-building ley (usually grass:clover) followed by three years of arable before resumption of ley. On an area x time basis, organic yields will therefore be substantially lower whatever the actual yield/ha compared to farms that use fertiliser to support continuous arable.
The N fixation figures quoted by Badgley et al., (2007) indicate large standard deviations of N fixation of 33%-95% of the mean which is close to 100 kg N/ha. How much of this N is incorporated into grain? The authors reference a single paper (Hoyt and Hargrove, 1986) and give a figure of 66%. This estimate is based on statements in Hoyt and Hargrove (1986) in their summary, that a selection of green manures (crimson clover, hairy vetch and Austrian winter peas) accumulated more than 150 kg organic N/ha and subsequently released 100 kg inorganic N/ha for the subsequent summer crop. This estimate, in turn, is based on data from US experiments of the early 1980s, most of which are referenced in Hoyt and Hargrove (1986) as 'unpublished'.
But it is not what is available to the plant that is important but what gets incorporated into grain. World-wide, 'conventional' N-use efficiency is 33%, a figure easily calculated from the fertiliser N used in agriculture and the N that ends up in seed (Raun and Johnson, 1999). What is the actual N-use efficiency from green manure? A number of papers place the figure at 20% or less, not 66% (Kramer et al., 2002; Ladd and Amato, 1986; Harris et al., 1994). Others report that only 9-33% of the legume N incorporated into the soil the previous year, is actually taken up into the crop, let alone translocated into the grain (Muller 1987, 1988a,b; Ladd et al., 1983). Berry et al., (2002) used a well-established relationship between soluble soil N and the protein content of wheat seed and showed that, in an organic soil containing a potential 300 kg N/ha, the wheat plant only appeared to be able to access 50-60 kg N/ha. It can be argued that continued use of legumes will build up soil N and thus increase the yield long-term. Measurements using manure described by Bochenhoff et al. (1986) and Dabbert (1990) suggest that the increase in N availability as a result of this treatment is only about 15% in total and plateaus after about ten years.
The oft-quoted reason why N-use efficiency is so low for organic wheat (and other annual crops) is the lack of synchrony of mineralization of organic nitrogen with crop requirements. Wheat, for example, needs to take up a lot of N very quickly in spring to form the leaf canopy. Spring applications of soluble fertiliser easily provide that requirement. Organic mineralisation certainly occurs in spring but not usually quickly enough to meet crop demand. Mineralisation continues throughout the season and beyond the growth period in late summer and autumn, ensuring N loss during winter rains. Leaching of N from organic and non-organic farms is now known to be similar in magnitude (Trewavas, 2004). The risks of eutrophication and global warming potential from organic farms are now thought to be higher than conventional farms (Williams et al., 2006).
The average figure quoted for cover crop N fixation is about 100 kg N/ha. About 6 tonnes of wheat and 10.5 tonnes of corn contain 100 kg N (Raun and Johnson, 1999). Even if the unrealistic figure of 66% of N-use efficiency of Badgley et al., (2007) is accepted this could provide enough N for only 4 tonnes of wheat and between 6-7 tonnes corn. But the lower figure of 20% N-use efficiency would yield only 1.3 tonnes of wheat or 2 tonnes of corn, and no more than 60-80 kg protein/ha; enough to support about 4-5 people /ha on a very simple vegetarian diet; a substantially lower density than S. China around the beginning the 20th century (Smil, 2001). In addition, estimates suggest that spoilage from fungal disease and waste accounts currently for 10-30% of world yields.
Organic agriculture cannot provide the necessary N to drive such a system to feed the projected world population of 9 billion, even if all farm land is under food crops and using all of the 1-1.5 billion ha of current arable land available. More arable land could be obtained by cutting down rain forest (the only suitable soil left) but the effects on global warming would prohibit such an approach. There may be improvements in N fixation by genetic manipulation (currently banned by organic associations) or conventional breeding. And of course experimental stations such as Rodale, highlighted by Badgley et al., (2007) can no doubt produce greater levels of N fixation, but these are not representative of most farms and the huge variations in N fixation currently reported give no confidence that they will ever be. The variability in N fixation is of crucial concern to farmers since it is the key factor in organic rotations that determines final crop production and income. What is missing from the Badgley et al., (2007) Appendix 2 are the figures indicating the lower variability of yields using fertilisers, and the reason that fertilisers were first used, in addition to legumes, to feed a growing world population. Once fertiliser is applied the farmer is at least assured that, depending on rain, soluble N will be available for crop growth. Reliability is important in present world agriculture. Good N fixation on one farm is of little help to another where it fails; nodular N cannot be transported between farms!
The insistence that the mineralisation of soil organic matter and crop residues is the only way to provide nutrients to crops, misses the obvious potential of using organic matter, augmented by fertilisers and manures where available. Soluble fertiliser can much more accurately provide nutrients when the growing crop needs it (synchronisation), whilst organic material improves soil structure, water holding capacity, microbial numbers and root development. Claims that fertilisers damage the soil (reiterated by Badgley et al. 2007) were first made by Steiner without evidence and can be rejected (Pfeiffer, 1940). As the Broadbalk Experiment clearly shows, soluble fertilisers do not damage the soil but merely provide a readily available form of the same ions that plants would take up from mineralised organic matter.
The article by Badgley et al. (2007) was, in our view, misleading. It was also confused about what it was trying to demonstrate: was it organic, IFM, alternative or no-till? The data were poorly researched and presented. If the question about organic versus conventional yields is worth asking, what is needed is a rigorous comparison of yield ratios and of the land area needed for organic production: while land is building fertility, it cannot grow wheat. Much of the literature in this area is weak and of little scientific value. It should not have been used without critical assessment!
Current world agricultural production has seen the price of major food-stuffs halved in the last 50 years (up to 2007), which benefited the poorest most. Organic food is more expensive, often greatly so, the higher price being compensation for lower yields and greater variation and greater risk of crop loss. Any attempt to convert world agriculture to organic would increase world food prices enormously, and those most at risk would be the poorest nations that are unable to provide sufficient produce of their own. There is plenty of potential to increase “conventional” production based on scientific understanding that does not seem to exist in organic farming, with its arbitrary, often ideological, regulations that can take many years to change. Current public support for organic production seems based on lack of knowledge of the toxicology of natural food chemicals, incorrect perceptions of current farming procedures and an unsustainable belief that nature knows best.
However, there is always room for improvement in all kinds of farming and we have no difficulty in agreeing with Badgley et al. (2007) about the need to improve soil quality by adding organic material, reducing unnecessary use of fertilisers and agricultural chemicals, and optimising rotations to reduce losses to pests and diseases. This is just common sense, since it will save farmers money as well as improve the farmed environment; it is a 'Win-Win'situation. In areas of the world where soils and income are poor, benefit will follow from these prescriptions.
Badaruddin, M., Meyer, D.W. 1994.Grain legume effects on soil-nitrogen,grain-yield and nitrogen nutrition of wheat. Crop Science 34, 1304-1309.
Badgley, C., Moghtader J., Qunitero, E., Zakern, E., Chapell, J., Avile´s-Va´zquez, K., Samulon, A., Perfecto, I. 2007. Organic agriculture and the global food supply. Renewable Agric. Food Sys. 22, 86-108.
Bender M.H. 2001. An economic comparison of traditional and conventional agricultural systems at a county level. American J. Alternative Agric. 16, 2-15.
Berardi, G.M. 1978. Organic and conventional wheat production - examination of energy and economics. Agro-Ecosys. 4, 367-376.
Berry, P.M., Sylvester-Bradley, R., Philips, L., hatch, D.J, Cuttle, S.P., Rayns, F.W., Gosling, P. 2002. Is the productivity of organic farms restricted by the supply of available nitrogen? Soil Use and Management. (supplement)18, 248-256.
Bochenhoff, E. 1986. Analyse der Beitrebs und Produktionsstrukturen sowie der Naturalertrage Imalternativen Landbau. Berichte uber Landwirtschaft 64: 1-39.
Dabbert, S. 1990. Zur Optimalen Organisation Altenativer Landwirtschaftlicher betreibe Agrarwirtschaft Sonderheft 124, Frankfurt, Verlag Alfred Strothe.
Davidson, B.R., Martin, B.R. 1968. Experimental research and farm production. Ag-Econ Res. Report No 7. University of West Australia Press Nedlands, West Australia.
Dlouhy, J. 1981. Alternative forms of Agriculture- Quality of Plant Products from Conventional and Biodynamic Growing. Report 91. Department of Plant Husbandry, Swedish University of Agricultural Sciences. Uppsala. 143 pp.
Entz, M.H. Guildford, R., Gulden, R. 2001. Crop yield and soil nutrient status on 14 organic farms in the eastern portion of the northern Great Plains. Canadian J. Plant Sci. 81, 351-354.
Goulding, K.W.T., Jarvis, S.C.J. ,Whitmore, A.P. 2008. Optimising nutrient management for farm systems. Phil. Trans. Roy. Soc. London, Series B, in press.
Gransted, A.G. , Kjellenberg, L. 1996. Quality investigations with the K trial, Jarna and other Scandanavian fertilisation experiments. In Raupp, J. (Ed). Quality of Plant Products Grown with Manure Fertilisation. Volume 9. Institute for Biodymanic Research. Darmstadt. pp.3-12.
Grimm, M. 1988. Zum Agrarbericht 1988. Lebendige Erde 3, 166-169.
Harris, G.H., Hesterman, O.B., Paul, E.A., Peters, S.E., Janke, R.R. 1994. Fate of legume and fertiliser 15N in a long term cropping systems experiments. Agronomy. Journal. 86, 910-915.
Higginbotham, S., Leake, A.R., Jordan, V.W.L. , Ogilvy, S.E. 2000. Environmental and Ecological Aspects of Integrated, Organic and Conventional Farming Systems. Aspects Appl. Biol. 62,15-20.
Hoyt, G.D. & Hargrove W.L. 1986. Legume cover crops for improving crop and soil-management in the southern united-states. Hortscience 21, 397-402.
Jenkinson, D.S., Bradbury, N.J., Coleman, K. 1994. In: Leigh, R.A. ,
Johnston A.E.(Eds). Long-term Experiments in Agricultural and Ecological
Sciences. CAB International, Wallingford, UK. pp. 117-138.
Kitchen, J.L., McDonald, G.K., Shepherd, K.W., Lorimer, M.F. , Graham, R.D. 2003. Comparing wheat grown in South Australian Organic and Conventional wheat farming systems. 1. Growth and Grain Yield. Australian J. Agric.Res. 54, 889-901.
Koepf, H.H. 1981. The Principles and Practice of Biodynamic Agriculture. In B.Stonehouse (ed). Biological Husbandry. London, Butterworths. pp.237-250.
Koepf , H.H., Peetersson, B.D. and Schaumann, W. 1976. Biodynamic Agriculture The Anthroposophic Press. Spring Valley, New York. 429 pp.
Kramer A.W., Doane, T.A., Horwath, W.R., and van Kessel, C. 2002. Combining fertiliser and organic inputs to synchronise N supply in alternative cropping systems in California. Agric. Ecosys. Environ. 91, 233-243.
Kristensen, M., Ostergaard, L.F., Halekoh, U., Joprgensen, H., Laurdisen, C., Brandt, K., and Bugel, S. (2008). Effect of plant cultivation methods on content of major and trace elements in foodstuffs and retention in rats. J. Sci. Food Agric. 88, 2161-2172.
Ladd, J.N., and Amato, M. 1986. The fate of nitrogen from legume and fertiliser sources in soils successively cropped with wheat under field conditions. Soil Biol. Biochem. 18, 417-425.
Ladd J.N., Amato, M., Jackson, R.B., Butler J.H.A. 1983. Utilisation by wheat crops of nitrogen from legume residues decomposing in soils in the field. Soil Biol. Biochem.15, 231-238.
Lampkin, N., Measures, M. 2001. Organic Farm Management Handbook. WIRS, Aberystwyth and EFRC Newbury.
Leu, A.F. 2004. Organic Agriculture Can Feed the World. Acres, USA, Vol. 34, No1.
Levi, M. 1979. Principles of Bio-Organic Agriculture. The Biosphere 8, 30-35.
Lockeretz, W., Shearer, G., Kohl, D.H. 1981. Organic farming in the corn belt. Science 211, 540-547.
Mader, P., Fliessbach, A., Dubois, D., Fried, P., Niggli, U. 2002. Soil fertility and biodiversity in organic farming. Science 296, 1694-1697.
Mason, H.E., Navabi, A, Frick, B.L. O'Donovan, J.T., Spaner D.M. 2007. The Weed-Competitive ability of Canada western red spring wheat cultivars grown under organic management. Crop Sci. 47, 1167-1175.
Muller, M.M. 1987. Leaching of subterranean clover derived N from a loam soil. Plant Soil 102, 185-191.
Muller, M.M. 1988a. The fate of clover derived 15N nitrogen during decomposition under field conditions: effects of soil type. Plant Soil 105, 141-147.
Muller M.M. 1998b. The fate of clover derived nitrogen 15N during decomposition under field conditions. Effects of liming and fertilisation. Plant Soil 111,121-126.
National Research Council. 1989. Alternative Agriculture. Washington, D.C. National Academy Press.
Nguyen, M.L., Haines, R.J. 1995. Energy and labour efficiency for three pairs of conventional and alternative mixed cropping farms in Canerbury, New Zealand. Agric. Ecosys. Environ. 52,163-172.
Pfeiffer, E. 1940. Biodynamic farming and gardening: soil fertility renewal and preservation. New York/London, Anthroposophic Press/Rudolf Steiner.
Poutala, R.T., Korva,J., Varis, E. 1993. Spring wheat cultivar performance in ecological and conventional cropping systems. J. Sustainable Agric. 3, 63-83.
Raun, W.R., Johnson, G.V. 1999. Improving nitrogen use efficiency for cereal production. Agronomy J. 91, 357-363.
Raupp, J. , Konig, U.J. 1996. Biodynamic preparations cause opposite yield effects depending upon yield levels. Biol. Agric. Hort. 13, 175-188.
Riverra-Ferre, M.G. 2008. The Future of Agriculture. EMBO Reports 9, 1061-1066.
Roberts, K.J., Warnken, F. ,Schneeberger, K.C. 1979. The economics of organic crop production in the Western Corn belt. Agricultural Economics Paper No. 1979-6. Department of Agricultural Economics, University of Missouri, Columbia, MO. pp. 63-101.
Robertson, G.P. Paul, E.A.,Harwood, R.R. 2000. Greenhouse gases in intensive agriculture. Science 289, 1922-1925.
Ryan, M. H., Derrick, J. W. , Dann, P. R. 2004. Grain mineral concentrations and yield of wheat grown under organic and conventional management. J. Sci. Food Agric. 84, 207-216.
Smil, V. 2001. Enriching the Earth. Cambridge, MIT Press.
Smith, R.G., Menalled, F.D., Robertson, G.P. (2007) Temporal yield variability under conventional and alternative management systems. Agronomy J. 99, 1629-1634.
Smolik, J.D., Dobbs, T.L., Rickerl, D.H. 1995. The relative sustainability of alternative, conventional, and reduced-till farming systems. American J. Alternative Agric. 10, 25-35.
Stanhill, G. 1990. The comparative productivity of organic agriculture. Agric. Ecosys. Environ. 30, 1-26.
Stockdale, E.A., Shepherd, M.A., Fortune, S., Cuttle, S.P. 2002. Soil fertility in organic farming systems - fundamentally different? Soil Use & Management 18 (Supplement), 301-308.
Stonehouse, D.P. 1991.The economics of tillage for large-scale mechanised farms. Soil Tillage Res. 20, 333-351.
Stoppler, H Kolsch, E, Vogtmann H. 1989. Impact of wheat breeding in an agricultural low input system. J. Agron. Crop Sci. 162, 325-332.
Trewavas, A. 2004. A critical assessment of organic farming-and-food assertions with particular respect to the UK and the potential environmental benefits of no-till agriculture. Crop Protection 23,757-781.
Vereijken, P. 1986. From conventional to integrated agriculture. Netherlands J. Agric. Sci. 34, 387-393.
Walker, D., Smith, A. 1992. Evaluation of spring wheat, oat and hulless oat and barley cultivars under an organic production protocol. Adaptation Res. Reports. 14: 257-260.
Welsh, J.P., Phillips, L, Cormack, W.F. 2002. The long term agronomic performance of stockless rotations. In. Powell, J (Ed) UK Organic Research 2002. Proceedings of the COR conference. Aberystwyth pp 47-50.
Williams, A.G., Audsley, E., Sandars, D.L. 2006. Energy and environmental burdens of organic and non-organic horticulture and agriculture. Aspects Appl. Biol. 79,19-23.
Wolfe, M. 2001. Recognising and realising the potential of organic agriculture. Global AG 2020 Conference, John Innes Centre.
Wookey, B. 1987. Rushall - the story of an organic farm. Oxford, Blackwell.