Moving from monoculture
A significant development in modern, conventional agriculture is the scale of monoculture of the major crops. Intensive systems rely on regular agrochemical input including widespread use of fungicides on crops that are vulnerable to disease attack. Today the global market for fungicides has reached US$5.7 billion representing a global pesticide market value of US$29.2 billion.
In high input countries of Western Europe, heavy fungicide use occurs, particularly on cereals. The situation in France, Europe’s biggest pesticide market, is typical of many European Union countries where the strobilurin sector in particular (including fungicides such as azoxystrobin and epoxiconazole) has grown rapidly. The advantages seem clear: the crop can be treated as a single commodity from seed production, through planting, pesticide application, harvesting, processing and marketing. However, such systems are entirely dependent on continuous inputs of synthetic chemicals, which leads to both direct costs (such as the fungicide azoxystrobin where resistance is becoming a problem, PN 51, p21) and indirect costs. The indirect costs include for example, checking for and removal of fertiliser and pesticide residues in water and health and other external
costs(1).
At the other extreme, in the tropics, there exist systems of highly diversified polyculture, typified by, for example, the forest gardens of Java. Here, a wide range of perennials and annuals are inter-cropped in such a way that synthetic inputs are not needed. Costs, direct or indirect, are minimal: virtually the only form of human intervention is the year-round harvest process. Such systems are highly energy efficient and encourage a wide range of natural biodiversity. Unfortunately, despite their inherent sustainability, they are not readily adaptable to larger scale farming systems in more temperate climates.
A central question for future development, therefore, is how far we might move away from the current monoculture approach towards polyculture so as to attain a practical balance between the positive and negative aspects of these two approaches. To answer this we need to know more about how polycultural systems work so effectively.
Plant biomass increases with diversity(2). One mechanism of this correlation is ‘complementation’, which is the degree to which different species or varieties grown together are able to use different parts of the available environment so as to produce more as a whole than each component could produce if grown alone in that environment. For example, shade-tolerant plants can grow beneath taller, light-seeking plants, and so on. Because of this more comprehensive exploitation of the environment, the polyculture can also be more competitive against weeds. And, because polycultures encourage a wide range of insect and animal species, it is also likely that competitive interactions among these organisms will help to restrict development of insect pests. However, one of the most common, reliable and predictable factors in polycultures is the restriction of disease spread.
Restricting disease with crop mixtures
Experimentally, it is now clear that mixtures of species or varieties restrict disease development through different
mechanisms(3,4). The first and most important is the spatial or dilution effect. If we take two species, one resistant and one susceptible to a particular disease, then, in a mixture of the species, the susceptible plants are separated by greater distances than they would be in a monoculture. Thus, spores of a pathogen produced on one plant to which they are adapted are less likely to be blown on to an identical plant that they can infect – so disease spreads more slowly. Second, the space between identical susceptible plants is occupied by plants that may be resistant to spores coming from their neighbours – so they provide a physical barrier to spores being blown from one susceptible plant to another, further reducing the probability of successful infections. A third mechanism is a kind of vaccination process. Spores that are not able to infect a particular plant induce a resistance reaction. If a spore which would normally infect that plant lands by chance in the area of the resistance reaction, it may die or develop only to a limited extent.
Disease develops slowly in the mixture, but, inevitably, one of the varieties in the mixture will be more infected than the remainder. Because of the infection, this variety may use less space and fewer resources (light, water, nutrients) than its neighbours – which can then use those surplus resources. So, the reduced yield of the more susceptible plants is compensated by the increased yield of the more resistant plants. Such compensation is less likely in a monoculture where all plants are equally susceptible. This is part of the reason why mixtures are more stable than monocultures in terms of yielding ability in different environments.
The net effect of these mixture processes has been shown to be highly significant in many trials over many years both in restricting disease development and in increasing
yield(5). Just as one example, in our organic wheat trials in Suffolk in 2000, six different mixtures, each of three wheat varieties reduced powdery mildew by 65%, leaf rust by 35% and
Septoria tritici by 17%, while increasing yield by 11%, relative to the means of the component varieties grown as pure stands (unpublished). An important feature of these observations is that different diseases were restricted simultaneously in each variety mixture.
Foxing the pathogen
Experience with large-scale use of crop varieties in monoculture and of single fungicides has shown that, time and again, the pathogen produces a response in terms of new races that are able to overcome the resistant varieties or fungicides. The absence of competition found in a massive area of a single variety, or of fungicide-treated crop, provides an ideal opportunity for previously rare forms of the pathogen to multiply rapidly. However, in a mixture of varieties grown on a large scale, selection on the pathogen is not uniform. There may be competition among individual pathogen genotypes that are well adapted to specific varieties in the mixture, and those that thrive on different combinations of varieties but are less specialised. From field experience, it is not clear which of the competitors will succeed, nor how quickly they will do so.
To slow the rate of pathogen response still further requires a simple stratagem, which is to change one or more components of the variety mixture, so that the selection process in the pathogen population is diverted from time to time. Some years ago, we tried an additional
stratagem(6), which was to treat the seed of one of three barley varieties (usually the one most susceptible to disease) with a fungicide, and then to mix the fungicide-treated seed with the untreated seed of the other varieties. The end result was, as expected, that the highest disease control and yield were obtained from the mixture with all components treated with fungicide, but the differences between fungicide treated and untreated mixtures were very small. It was simpler, cheaper and more environmentally acceptable, to continue with variety mixtures without adding fungicide.
Scaling-up dramatic results in China
It has often been observed with field trials, that the effectiveness of mixtures in small plots can be affected by neighbouring plots. If the nearest neighbour happens to be a plot of a highly susceptible sole variety, then the mixture may exhibit only a limited degree of disease restriction because of the continuous bombardment of spores from the neighbour. The converse of this problem is that, with an increasing field area, the effectiveness of the mixture should increase.
Chinese researchers have investigated this point over two years for rice blast in trials increasing from 800 to more than 3,000
hectares(7). Would the damping effect of their rice mixture multiply across neighbouring mixture fields? First, the authors had to persuade all the rice farmers in a large area that they should grow a particular mixture of rice varieties. The effect was dramatic, with disease reductions up to 97% recorded. This led to a dramatic response from the farmers, thousands more of whom participated in the experiment, so that it was relatively simple to increase its size still further in subsequent years. The level of rice blast was hugely decreased across the whole target area, and the farmers stopped using fungicides.
Mixture effects in crops and pathogens
Because of the fundamental nature of the mechanisms by which mixtures restrict disease development, we should expect mixtures to be effective for a wide range of pathogen and crop species. For pathogens, it was expected that variety mixtures would be most effective against those fungi that are spread by airborne spores among relatively small
plants(8). However, research has demonstrated that useful effects can be obtained even for soil-borne
pathogens(9). At the very least, one might expect that plants that are resistant to a soil-borne pathogen would compensate for the lesser growth of susceptible neighbours in a mixed stand.
The range of crops for which mixtures are effective in restricting at least airborne pathogens now includes such widely different hosts as apple, willow, rape, coffee, wheat, barley, oats, triticale, maize, lettuce, peppers and soya beans. Elm Farm’s current work includes variety mixture trials with potato, to try to restrict late blight, in a large-scale project on blight funded across seven countries by the EU. Earlier pilot trials showed that mixtures of three varieties can have a useful effect, although the scale of restriction may not be as dramatic as with pathogens of cereal crops, which develop more slowly than does the potato blight pathogen.
Marketing needs to adjust
Despite the clearly demonstrated benefits of variety mixtures, farmers generally have been slow to make use of this approach. A major reason appears to have been the perceived difficulties of marketing the mixed product. Where mixtures are used for on-farm feed or for industrial crop use, there should, of course, be no difficulty. The main bottleneck appears to have been the reluctance of, for example, millers and maltsters to buy mixtures even though they often use mixtures themselves. Their argument has had two main points. First, the processors reason that they prefer to mix specific components that may be different from those that would be useful on the farm. Second, they are concerned that less scrupulous farmers might try to conceal grain of low quality in a declared mixture.
These arguments may well be missing a major opportunity. Where mixtures have been, or are being, used on a large scale, there has been no problem in terms of product quality. In the former German Democratic Republic, for example, where the entire spring barley crop was grown as variety mixtures, collaboration between maltsters, breeders and pathologists ensured that the mixtures recommended to farmers gave high yields, major protection against powdery mildew and leaf rust, and uniformly high quality
malt(10). In the west and north-west of the US, where there are now about half a million hectares of wheat
mixtures(11), grain is sold on the basis of type (e.g. hard red spring, soft white winter etc.) rather than by variety, so that a mixture of hard red spring varieties is just as acceptable as a single variety of the same type. Interestingly, European buyers of American milling wheat are importing mixtures for use in their processing.
For the longer term, it is logical to reason that a mixture made up of varieties grown in the field that have different but complementary quality characteristics could well be selected to produce an overall better quality than any one of the varieties grown alone. Indeed, there is now
evidence(12) to support this.
Future developments
Variety and species mixtures can provide a simple and satisfactory means of restricting plant diseases and stabilising crop yields without external inputs. Their performance so far in experimental situations merits their wider uptake, particularly in organic farming.
Mixtures of species provide another layer of crop diversity, with many potential advantages, waiting to be exploited in contemporary
approaches(13,14). It is widely recognized, for example, that high-yielding mixtures of grains and legumes (grass plus clover, maize plus beans, and many other combinations) can restrict the spread of diseases, pests and
weeds(15). At the same time, such mixtures can provide near-complete nutrition for animals and humans alike, without recourse to expensive and uncertain forays into genetic engineering.
Working out the best polycultural combinations for different agricultural systems requires integration of farmer experience and needs on the one hand, and of the needs of the end-user on the other. Ideally, there is also a need for breeding varieties and species specifically for use in mixtures. Such programmes will be long-term and expensive, but the returns on such investments will be considerable in terms of agricultural sustainability. In the meantime, those concerned with variety recommendations, and, indeed, farmers themselves, can carry out simple trials to determine useful mixtures of varieties or species for particular areas or purposes based on currently available plant material.
References
1. J Pretty, High costs of intensive agriculture, Pesticides News No. 50, p18.
2. D Tilman, CL Lehman, and KT Thomson, Plant diversity and ecosystem productivity: Theoretical considerations. Proceedings of the National Academy of Sciences, US, 1997, 94:1857-1861.
3. MS Wolfe, The current status and prospects of multiline cultivars and variety mixtures for disease resistance, Annual Review of Phytopathology, 1985, 23: 251-273.
4. KA Garrett, CC Mundt, Epidemiology in mixed host populations, Phytopathology, 1999, 89: 984-990.
5. MR Finckh, ES Gacek, H Goyeau, C Lannou, U Merz, CC Mundt, L Munk, J Nadziak, AC Newton, C de Vallavieille-Pope, and MS Wolfe, Cereal variety and species mixtures in practice, with emphasis on disease resistance, Agronomie, 2000, 20: 813-837.
6. MS Wolfe, Integrated use of fungicides and host resistance for stable disease control, Philosophical Trans. of the Royal Society, B 295, 1981, 175-184.
7. Y Zhu, H Chen, J Fan, Y Wang Y Li, J Chen J Fan S Yang L Hu, H Leung TW Mew, PS Teng, Z Wang and CC Mundt, Genetic diversity and disease control in rice, Nature, 2000, 406: 718-722.
8. Op. cit. 4.
9. V Vilich-Meller, Pseudocercosporella herpotrichoides, Fusarium spp. and Rhizoctonia cerealis stem rot in pure stands and inter-specific mixtures of cereals, Crop Protection, 1992 11:45-50.
10. MS Wolfe, H Hartleb, Sachs, and H Zimmermann, Sortenmischungen von Braugerste sind gesuender, Pflanzenschutz-Praxis, 1991, 2: 33-35.
11. Op. cit. 4.
12. AC Newton, JS Swanston, Annual Report 1998/99 55-59, Scottish Crop Research Institute, 1999.
13. Op. cit. 2.
14. SL Pimm, In search of perennial solutions, Nature, 1997, 389:126-127.
15. LE Jackson, (ed.) Ecology in Agriculture, Academic, San Diego, US, 1997.
Martin Wolfe is Research Director at Elm Farm Research Centre (UK’s main organic research station), Hamstead Marshall, Newbury, Berks. RG20 0HR, UK.
[This article first appeared in
Pesticides News No. 52, June 2001, pages 10-11]