“Straight up” beats “cocktails” for cover crop ecosystem services
Cover crop mixtures, known as “cocktails” by some, are being promoted as having benefits over cover crops planted as monocultures. As I described in Part I, I reviewed recent research results to get at the answer to the question, “are monocultures or polycultures better when it comes to cover crops?” I found that, for biomass production at least, monocultures were actually best. Now, let’s look at other services provided by cover crops and compare polycultures and monocultures. (See an explanation of monocultures, polycultures, overyielding and transgressive overyielding in my post Ecological Theories, Meta-Analysis, and the Benefits of Monocultures.)
Cover crops provide benefits even in monocultures
Weed suppression by a cover crop is often desired, even if (because herbicides are used in the main crop, for example) it is not the primary goal. Smith et al. (2014) found no evidence that mixtures (five species, in their case) enhanced weed suppression over monocultures. They noted that weed suppression is related to the competitiveness of a crop, where biomass production is a major factor. If you have a mixture that includes a less aggressive, less productive species, like many legumes, this will decrease weed control. Similarly, a team from Penn State found that there were no significant differences in spring weed suppression between the most productive monocultures and the mixtures. As their webinar states: “You do not need a mixture to get good weed suppression.” The legume monocultures did not suppress weeds well, but mixtures with legumes were similar to the best monocultures, probably because the aggressive species in the mixes, here oat and rye, dominated the legumes in the mixes. Think about this: if the dominant species is suppressing weeds, would it not also be suppressing your less dominant planted species? After all, to a rye plant, a clover plant is as much a weed as lambsquarters. Thus, getting weed suppression is limiting your opportunity to supply nitrogen from the legume – there is a tradeoff; mixtures are not win-win. We will return to this.
A Miyazawa et al. (2014) study looked at nutrient recycling by cover crops (often referred to as the “improved soil fertility” ecosystem service, although no nutrients are actually added to the system). This would be important for reducing the leakiness of some agricultural systems, especially where nitrates are leached in the winter. They found no differences, over four years, in the nitrogen, potassium or phosphorus uptake between monocultures and mixtures. Similarly, the Penn State study found that monoculture rye reduced nitrate leaching as well or better than mixtures. Do you see a trend here?
Although the Penn State team was the only one of these recent cover crop studies that looked at pest control benefits of mixtures – they counted the beneficial insects – what they found fits the pattern; the presence and density of flowers in the crop (mono- or polyculture) was more important than species diversity. In their study, the monoculture canola crop had the most flowers per acre, and so attracted the most beneficial insects.
The exception to this recurrent theme of monoculture out-performing polyculture, is when you grow legumes in low nitrogen soils (for legume crops in rotation, see Smith et al. 2008). In this situation, it is the powerful nitrogen effect (an “interspecific synergy” according to Hayden et al. 2014), as has been seen in so many studies over so many years. In Hayden et al., as opposed to Wortman, monoculture yields of legumes were equal or greater than those of rye. Make no mistake, this is a nitrogen effect, even when it is labeled as synergy or other such terms. The effect of nitrogen on plants, whether it comes from legumes or from nitrogen fertilizer is well known and reported in many studies each year, so it should not surprise anyone to find that when mixtures including legumes are compared to monocultures without legumes on low-nitrogen soils, that the mixtures produce more biomass, and so suppress more weeds, and provide more soil cover. In soils with higher levels of nitrogen, however, the effect disappears and monocultures beat the best mixtures, even those including legumes.
Hayden et al. did not find any synergistic effect of mixtures on N fixation by the legumes, or on weed suppression. What they found were tradeoffs; the more rye in a mix, the better the weed suppression; the more legume, the higher the amount of nitrogen fixed, but the worse the weed suppression.
Polycultures require tradeoffs
This is the general conclusion of current research, that there are tradeoffs in obtaining ecosystem services from cover crop mixtures. One crop grows quickly, produces a lot of biomass and suppresses weeds well. A legume grows more slowly but fixes nitrogen. When you mix them, the productivity of the first crop is reduced as is the nitrogen fixation of the legume – there is a tradeoff. This is contrary to the win-win hypothesized in ecology, where complimentary effects between species overcome competition between them. Other than the nitrogen effect from legumes, however, evidence for this is rare. Damour et al. (2014) evaluated 28 species for cover crop use in banana production. They found that most species were good at providing some services, but not at others. No one species provided all the services they desired, and while mixtures provided more services, the level of those services was reduced (diluted). Tradeoffs, rather than complementarity, seem to be the rule, at least with annual, short-season crops, with one species diluting the effects of the other and vice-versa. And the more species in the mix, the more diluted each species’ unique services become. The services are, in a twist of ecological terminology “density dependent”; the fewer plants of a species in a field, the less of that species’ services per acre.
These results challenge popular thought on the topic, including statements made by some in NRCS (here and here, although more often they are spoken and not in print), to the point that some researchers have conducted studies specifically to test these proclamations. One such claim was that cover crop mixtures use less water than monocultures, a critical factor in use of cover crops in regions like the Great Plains. Nielsen et al. (2015) grew 10 monoculture cover crops and compared their use of water to a 10 species mixture. Contrary to the claims, over two years the research found no significant differences overall in soil water contents or growing season crop water use between the monocultures and mixture. The cover crops all used more water than would have just evaporated (fallow) showing there is a cost for farmers growing any cover crop in that region. The researchers surmised that when mixtures do not use as much water as monocultures it is not some synergy in the mixture, but rather a reflection of the lower yields in the mixture compared to the best monoculture species, and since water use is related to yield, less water is used in the mixture. Again, there is a tradeoff here between yield and water use, not synergy between species.
The figure above (from Tilman et al., 2014), is similar to one I shared in Part I of this post. It displays results of modeled ecological theory: an increase in the number of species will result in higher total community biomass. While it may hold true in natural systems, it is not supported by recent cover crop studies, where increasing the number of species did not enhance their performance (in Wortman, going from 2 to 8 species, in Miyazawa, going from two to three species, did not increase biomass or stability). Perhaps if we were growing perennial species over several years, then a mixture would be an advantage, to counter soilborne diseases (Schnitzer et al. 2011) if nothing else. However, in annual cropping agriculture, and especially for short season cover crops, there appears to be no advantage to mixtures with the exception of combining grasses and legumes on low-nitrogen soils.
We can’t have our cake and eat it, too
“Diversity may be there for a reason…but that does not mean that diversity is there for a purpose.” This is a favorite quote from R. Ford Denison’s book, Darwinian Agriculture. In short, it means that we should not assume that everything in nature is optimal for agriculture, and so we should not mimic nature haphazardly. For crops, and for cover crops, we should use the best monocultures, rotated in time to provide diversity. Don’t dilute the strength of a species by putting it in a mixture with lesser (for a specific goal) species.
I believe this leaves you with the choice: 1) choose one primary goal for your cover crop and then plant a monoculture of the species that will best achieve that goal, or 2) choose a few goals and plant mixtures of species that best achieve those goals, recognizing that each of them will provide less of their service in a mixture than they would as a monoculture. However, if one of your goals is weed control, the second option is not a very good one. Achieving some degree of other goals at the expense of weed control means that some weeds will not be controlled; they will go to seed and potentially increase future problems.
Finally, instead of continuing to try to show that polycultures are better than monocultures, let’s improve the rotations of monocultures, including cover crops, and reap their known benefits. And let’s recognize that although ecological theory may work in nature, agriculture is not nature, and works differently. Figuring out how agriculture works best is a big enough job for both agronomists and agroecologists.
- Cardinale, B. J., Matulich, K. L., Hooper, D. U., Byrnes, J. E., Duffy, E., Gamfeldt, L., … Gonzalez, A. (2011). The functional role of producer diversity in ecosystems. American Journal of Botany, 98(3), 572–592. http://doi.org/10.3732/ajb.1000364
- Damour, G., Dorel, M., Quoc, H. T., Meynard, C., & Risède, J.-M. (2014). A trait-based characterization of cover plants to assess their potential to provide a set of ecological services in banana cropping systems. European Journal of Agronomy, 52, Part B, 218–228. http://doi.org/10.1016/j.eja.2013.09.004
- Hayden, Z. D., Ngouajio, M., & Brainard, D. C. (2014). Rye–Vetch Mixture Proportion Tradeoffs: Cover Crop Productivity, Nitrogen Accumulation, and Weed Suppression. Agronomy Journal, 106(3), 904. http://doi.org/10.2134/agronj2013.0467
- Miyazawa, K., Takeda, M., Murakami, T., & Murayama, T. (2014). Dual and Triple Intercropping: Potential Benefits for Annual Green Manure Production. Plant Production Science, 17(2), 194–201. http://doi.org/10.1626/pps.17.194
- Nielsen, D. C., Lyon, D. J., Hergert, G. W., Higgins, R. K., Calderón, F. J., & Vigil, M. F. (2015). Cover Crop Mixtures Do Not Use Water Differently than Single-Species Plantings. Agronomy Journal, 0(0), 0. http://doi.org/10.2134/agronj14.0504
- Penn State cover crop project: webinar, slides, publication, website.
- Tilman, D., Isbell, F., & Cowles, J. M. (2014). Biodiversity and Ecosystem Functioning. Annual Review of Ecology, Evolution, and Systematics, 45(1), 471.
- Schnitzer, S. A., Klironomos, J. N., HilleRisLambers, J., Kinkel, L. L., Reich, P. B., Xiao, K., … Scheffer, M. (2010). Soil microbes drive the classic plant diversity–productivity pattern. Ecology, 92(2), 296–303. http://doi.org/10.1890/10-0773.1
- Smith, R. G., Gross, K. L., & Robertson, G. P. (2008). Effects of Crop Diversity on Agroecosystem Function: Crop Yield Response. Ecosystems, 11(3), 355–366. http://doi.org/10.1007/s10021-008-9124-5
- Smith, R. G., Atwood, L. W., & Warren, N. D. (2014). Increased Productivity of a Cover Crop Mixture Is Not Associated with Enhanced Agroecosystem Services. PLoS ONE, 9(5), e97351. http://doi.org/10.1371/journal.pone.0097351
- Wortman, S. E., Francis, C. A., & Lindquist, J. L. (2012). Cover Crop Mixtures for the Western Corn Belt: Opportunities for Increased Productivity and Stability. Agronomy Journal, 104(3), 699. http://doi.org/10.2134/agronj2011.0422
Very interesting analysis. Even if mixtures aren’t better on average, are there bet-hedging benefits, where they reduce the risk of drastically low cover?
By the way, as I note in Darwinian Agriculture, Tilman’s monoculture plots had very low plant cover, in contrast to agricultural monocultures and also to many natural ones. My suggestion that this low cover was due to accidental removal of the “resident” species during annual weeding of the “nonresident” ones may be wrong, but I don’t know what conclusions we can draw from Tilman’s plots without a solid explanation of why the monoculture plots have such low plant cover.
Maybe I should blog here rather than trying to maintain my own blog, now that my university has discontinued support for our platform.
Yes, you are right about the bet-hedging; where there is risk of not having ideal conditions in which the best monoculture could thrive, it would be less risky to go with a mixture. However, in dry regions, like the High Plains, where this might apply, the water use by a cover crop in dry years, even if a species included in the mix did well under dry conditions, would add to the risk for the following cash crop. Perhaps a mixture could reduce risk for varying temperatures in some situations?
Regarding Tilman’s plots, I wonder if, with perennial plants, the risk of building up a soilborne disease in monocultures would be higher than in a polyculture, which could also explain the deteriorating stand in the monoculture after a few years.
I am currently a masters student with Dr. Wortman working on cover crop mixtures and how they affect certain parts of the broad concept of “soil health” in small organic farming systems. In total I work with 18 different species in 5-way mixtures. We work with six species at a time in seasonally appropriate plantings, so we have 6 species in spring, 6 in summer, and 6 in fall. Each of these species sets typically contains 2 brassicas, 2 grasses, and 2 legumes.
One thing I wanted to mention after reading these two articles is the challenge of crop establishment with vastly different seed sizes. Our plots are small, so we hand-broadcast our mixtures and then either rake them in by hand or cultipack them with a small tractor and roll-behind. We have noticed that the bigger seeds, especially those like faba bean and cowpea, are often not buried deep enough for good establishment. Our other species typically germinate well and get established relatively quickly by either raking or cultipacking, but these seeds are smaller and respond better to shallow planting. Legumes like cowpea need to be sown around 1-2 inches deep for good establishment, which our methods are not fully capable of providing. We think that this could be a factor in why our legumes are typically dominated by mustards and grasses. Maybe legumes are partially at a disadvantage because they are not incorporated well enough into the soil in mixture, and thus get a slower start than the other species? We are always looking for more effective ways of incorporating our mixtures, since our seeds range from the very tiny (clover, mustards, turnips, radishes) to the very large (peas, faba beans, cowpeas, even vetch!).
Comments are closed.