The Functional Role of Diversity in Agroecosystems

Diversification is the Key to sustainability, according to most agroecologists.

Diversity in cropping systems:

Monoculture:

• Continuous
• Crop Rotation-short rotations vs. long rotations

Polyculture:

• Intercropping
• Agroforestry
• Home-garden systems

Key Questions:

Why are monocultures the dominant system in US and European agriculture while intercropping systems are common in much of the rest of the world?

• Labor efficiency
• Management skills required
• Machinery investment
• Demand for different crops

How do ecological processes differ in simple and complex systems? E.g.

• Resource use
• Resource-use efficiency
• Nutrient cycling
• Competition--Intraspecific vs. Interspecific
• Herbivory

Gliessman (1998) and other agroecologists suggest that diversification results in the enhancement of emergent properties (such as stability, improved pest regulation, soil fertility increases, etc.)

Theoretical Considerations

I. Diversity and Stability

Ecologists of the 1950s and 1960s described apparent causal relationships between ecosystem diversity and ecosystem stability. Many agroecologists today assume (somewhat uncritically) that such a relationship exists and is critical to agroecosystem function.

It has been difficult, however, to test theories relating diversity to stability because the terms themselves are ambiguous; i.e., they have different meanings in different ecological contexts.

You have to be able to measure both diversity and stability before you can study their relationship; to measure them, you have to first define them.

Diversity has been defined as:

• Richness-number of species
• Equitability-number and relative abundance
• Connectance or complexity-usually as food-web complexity.

Stability--Most definitions involve notions of: "Degree to which system doesn't change, or, given some sort of disturbance, the degree to which the system recovers to its original state.":

• Resilience--Rate of recovery after a disturbance.
• Resistance--The amount of change caused by a disturbance
• Persistence--The tendency of system properties to stay more or less constant through time
• Variability--The variance in system properties over time.

Persistence and variability are of most concern to agroecology. Persistence is a major component of sustainability.

II. Diversity and Ecosystem Function

In recent years the apparent increase in the rate of extinction of species has led to many investigations of the roles of particular species and species diversity per se in the functioning of ecosystems.

Ecosystem function is usually defined in terms of:

• energy capture (i.e., productivity-yield in agriculture)
• nutrient cycling
• population regulation (including food web structure)
• stability

Several widely-cited empirical studies appeared to show that more diverse ecosystems function "better", that is, have higher productivity, lower nutrient losses, or increased stability in their productivity.

• Naeem et al. (1994) constructed model communities in large growth chambers (the "ecotron") and found that higher diversity communites had higher productivity.
• Tilman et al. (1996) constructed grassland communities differing in species number and found increased productivity and decreased nitrate leaching with higher species richness.
• Several more recent studies found that richness in "functional groups" was more important than richness per se.

Functional groups are considered to be groups of species that capture carbon or cycle nutrients in fundamentally different ways. For example,

• nitrogen fixers (usually legumes) vs. non-nitrogen fixers
• Early season species vs. late season species
• C3 grasses (cereals) vs. C4 grasses (cereals)
• Deeply rooted species vs. shallowly rooted species

The strongest response to increasing species or functional group richness in experiments is usually found at low species numbers (1-5 species), typical of the range of species that might be found in diversified agroecosystems.

Theoretical reasons why increased richness should be associated with enhanced ecosystem function in experimental studies:

• The "sampling effect"--larger assemblages are more likely to include more productive species.
• Compensatory growth--decreased growth of one species due to abiotic or biotic stresses can be compensated by growth of other species in a mixture.
• Resource-use complementarity--different species in a mixture can acquire resources from different parts of the ecosystem or at different times. The mixtures of species would thus be expected to capture more resources than monocultures, resulting in higher productivity and lower resource loss.

The last mechanism is probably the key one with regard to agroecosystems. If the niches of crops in a mixture differ in any way that affects the capture of resources, then a mixture should have a higher ecological resource-use efficiency than a monoculture.

Niche differences that might affect resource capture include: phenology, root architecture, shoot architecture, and symbiotic relationships.

Crop Rotation

Prior to development of agrichemicals, rotations were the standard practice to control pests and diseases and maintain soil fertility. Development of pesticides and herbicides made continuous monoculture possible. Thus continuous monoculture is a relatively recent agricultural practice.

Short rotations vs Long (Extended) Rotations:

Short rotation:

• Usually just 2 years
• Objective is typically pest control
• Corn-soybean is the commonest crop system in the US-both crops have a high demand

Long (extended) rotations:

• 3 years or longer
• Objectives are pest control, maintain soil organic matter, reduce agrichemical inputs
• Usually includes hay, pasture, or "green manure" to improve soil fertility.

Example of a long rotation: Soil fertility in England during 18th-19th centuries maintained by the Norfolk four-course rotation:

• turnips (fodder crop)
• wheat or barley
• clover (for soil fertility)
• wheat or barley

Rotation Effect. This term refers generally to the higher yields of most crops when grown in rotation, and more specifically to the yield increases that cannot be compensated for by input substitutions. Most crops produce higher yields in rotation than in continuous cultivation, usually 10-15% higher in maize (Singer & Cox, 1998).

That is, if rotations simply worked by provided additional nutrients, or better insect and weed control, then fertilizers, pesticides, or herbicides should result in crop yields of continuous monocultures similar to yields found in rotations; but they usually cannot. Thus the exact causes of the rotation effect remain unknown.

• Pest control
• Disease control
• Weed control
• Enhanced fertility. Microbial diversity is higher under wheat-legume rotations than in wheat monocultures (Lupwayi et al., 1998)
• Autotoxicity (a form of allelopathy). For example, Chou (1999) suggests that toxins released from decaying rice residues reduce yields of subsequent rice crops.

Legume-based Rotations

Legume-based rotations are most common. Incentives for including legumes in a rotation include (1) reduced nitrate leaching and (2) reduced fertilizer costs.

How much nitrogen is fixed by legumes grown in rotations? Do following crops require N fertilization? Methods of measuring BNF include:

• Acetylene reduction -- an instantaneous measure.
• Fertilizer replacement-- confounds N effects and rotation effects.
• Mass balance (N budget): N(bnf) = N(crop) + N(losses) - N(fert) -dN(soil); [dN(soil) = change in soil N].
• ¹⁵N enrichment. Useful for following the fate of fixed N.

Distinguishing nitrogen effects from rotation effects:

Honeycutt (1997) evaluated a method for quantitatively distinguishing N and non-N rotation effects; data supports the assumption that N and non-N effects are additive and non-interacting. For potatoes non-N effects were found to be ± constant over fertilizer levels; N-effects decrease predictably with increasing N fertilization of potato crop. The method involves:

• N-uptake vs. yield determined in fertilizer trials;
• N uptake estimated after rotation using an N budget;
• non-N effects = actual yield - predicted yield (based on N uptake).

Maloney et al. (1999) compared maize rotated with nodulating vs. non-nodulating soybean, both with and without N fertilizers. The yields of maize in continuous monoculture were lower than those of maize rotated with either form of the soybean. But there was no difference in yields in either rotation at any N fertilizer level (0, 80, 160 lb N/acre)-suggesting that N-fixation is not responsible for the rotation effect in the maize-soybean rotation.

Why don't farmers use crop rotations more often?

1. Chemical inputs. Fertilizers and pesticides--have substituted for many of the benefits of rotations.

• Reduced effectiveness or reduced availability (higher cost) could change grower practices.
• Need to stop or reverse soil erosion would also affect practices.

2. Economics.

• Limited markets for some rotation components.
• Commodity price support programs (base-acreage and cross-compliance restrictions).

3. Management. Producing more than one crop requires more expertise, machinery, and marketing skills.

Intercropping

Intercropping involves growing two crops in the same field at the same time. The following are different ways of intercropping, in order of increasing degree of association between crop components:

• Relay-intercropping-planting a second crop before harvesting the first crop.
• Strip-intercropping-growing 2 or more crops in alternating strips. Smith & Carter (1997) found that maize grown in a strip intercrop with alfalfa produced yields 6% higher in 40-ft wide strips, 11% higher in 20-ft wide strips, and 17% higher in 10-ft wide strips. May be due to extra light in border rows of maize.
• Between-row intercropping -growing 2 or more crops in alternating rows.
• Within-row intercropping -growing 2 or more crops in the same rows.

Between-row and within-row intercrops may be either additive or replacement designs.

Intercropping Concepts.

Additive vs. replacement intercrops. In an additive intercrop both species are planted at the same density as in their respective monoculture; in a replacement intercrop a row of one crop "replaces" a row of the second crop in forming the intercrop. Additive intercrops double the density, and therefore may use resources more completely.

Duration refers to the temporal overlap of the intercrop components:

• Differing duration-usually combines a short season crop and a long season crop. Intercrops of differing duration are usually additive.
• Similar duration-competition more intense because both components are using resources at the same time. Intercrops of similar duration tend to be replacement types.

Dominant vs. subordinate components. Typically, one crop component of the intercrop is more competitive and hence dominates the mixture in terms of growth and yield. Dominance may be due to:

• Rapid initial growth
• Height
• Photosynthetic pathway (C4 crops tend to be dominant when grown with C3 crops)
• Legumes are usually subordinate

Measuring Intercrop Performance

The performance of intercrops relative to monocultures of the component crops is usually measured as Land-equivalent ratios (LER) or relative yield totals (RYT):

• Relative Yield (RY) = Yield in intercrop/Yield in monoculture
• LER = RYT = Y(i)/Y(m) = RY(1) + RY(2) + RY(3) + ....

When LER or RYT > 1, the intercrop is said to show overyielding. That is, the intercrops are more productive than the monocultures of the components crops.

The RYs of dominant components are often close to 1.0; efforts to increase intercrop performance often center on increasing the RY of the subordinate component.

Cereal-legume intercrops are particularly common.

• Cereal component dominant--why? [Fast growth of C4 cereals, slow growth of N-fixing legumes, taller cereals.]
• Fertilization increases absolute yield of both monocultures and intercrops but decreases RYT of the intercrop.

Ecological Theory and Intercrop Performance

Why intercrop? (Instead of having many separate plots of different crops.)

• Higher Yields. Most likely to be achieved with additive intercrops.
• Lower Risk. Most likely to be achieved with replacement intercrops.

Ecology of intercropping--what are the mechanisms for overyielding?:

1. Intercrops could capture more resources. That is, ecological resource-use efficiency could be increased through higher LAI, longer LAD, greater root-surface area, etc.

2. Intercrops could increase physiological resource-use efficiency. Dominants often modify the habitat of the subordinates, e.g. lower temperature and higher humidity below canopy of the dominant might permit greater water-use efficiency of the subordinant. Reduction in herbivore populations in intercrops would also have the effect of increasing physiological RUE.

One or more component crops could have a higher harvest index when grown in intercrops. Increased competition reduces excess vegetative growth in crop specieds with lower HI (suboptimal assimilate partitioning).

3. Reduced weed populations in intercrops would effectively increase resource supply.

Competition vs. Facilitation. John Vandermeer has suggested that mechanisms for overyielding fall into two general groups-those involving competitive interactions and those involving facilitative interactions:

• Competition--An ecological interaction between two plants that has a negative effect (usually reduced growth) on at least one component). (Defined this way would include allelopathy.)
• Facilitation--An ecological interaction between two plants (or species) that has a positive effect on at least one component.

The "Competition Production Principle" depends on niche differentiation of the component crops. If crop species growing together harvest resources at different times or from different parts of the agroecosystem, then the combination of crops can acquire more resources than either crop growing in monoculture, resulting in overyielding.

Among animal species, competition often leads to increased specialization--allocation of energy or effort to capture resources in that portion of the environment or habitat where competition is least severe. For example, seed eating rodents, when occupying the same habitat, will tend to gather seeds of a particular size. Where competitors are absent, each species forages over a wider range of seed sizes. Can competing plant species do anything similar?

• The competitive production principle predicts that overyielding will only occur when one or more resources are limiting (almost always the case in agroecosystems in the developing world).
• If none of the crops are resource-limited, i.e., when limiting resources are in excess supply, then there is no opportunity for increasing total resource capture through niche differences.
• The potential for increased resource-utilization efficiency is greater when resources are scarce. The increased competition for scarce resources may accentuate niche differences. For example, increased BNF when soil mineral N is low.

The absolute yields of crops in the intercrop are lower than their respective yields in monoculture due to competition, but the RYT of the intercrop will be greater than 1.0 if the intercrop captures more limiting resources that the individual monocrops. Greater resource capture is possible if the niches of the crops differ with respect to resource capture.

Facilitation mechanisms include:

Nitrogen transfer, legumes to cereals/other non-N-fixers. N-transfer can be (1) direct (via VAM); (2) indirect (mineralization of legume residues during the current growing season), or (3) residual (mineralization after the current growing season). For example, indirect transfer can occur when an early season legume is intercropped with longer duration non-legume (such as soybean-cassava). In temperate zones, most N transfer is residual.

Weed suppression, via increased shading or allelopathy.

Reduction of insect pests. E.g., Trenbath (1993) cites a study which found that cotton grown without insecticides in an intercrop with sorghum yielded 25% higher than sole-crop cotton with insecticides. Insect pest reduction can occur due to two general mechanisms:

• Enemies Mechanism--greater biological control in intercrops.
• Resource Concentration Mechanism(s)--host plants harder to find (less apparent) in intercrops.

Discussion Question-Why not intercrop more in the US?

• Crop management in developed countries seeks to eliminate resource limitations rather increase resource-use efficiency.
• Overyielding by competitive production mechanisms is reduced by inexpensive fertilizers.
• Overyielding by facilitative production mechanisms is reduced or eliminated by effective insecticides and herbicides.
• Mechanization is often cited as an obstacle to intercropping.

Genetic Diversity in Crops

Diversification within a monoculture (i.e., planting a field to more than a single variety of a crop) may also be advantageous with respect to:

• Disease resistance
• Yield

For example, Zhu et al. (2000) found that mixtures of rice cultivars showed greater yield of disease-susceptible varieties and less disease incidence than in monocultures. Glutinous varieties of rice are both the most valuable economically and the most susceptible to rice blast. Their study was conducted on 1000s of farms; mixtures consisted of one row of glutinous rice for every four rows of hybrid rice. Rice blast on glutinous varieties was 94% less severe and yields (per row of the glutinous variety) were 89% higher in the mixtures.

URL: http://ag.arizona.edu/~spmcl/lecturenotes/diversity.htm