What do weeds and gamblers in Las Vegas have in common? They both manage the risk of unpredictable outcomes through a strategy called bet-hedging. Could farmers use this same strategy by planting cover crop mixtures? And if so, what are the odds of winning? Let’s take a look.
Bet hedging is a strategy of reducing risk of uncertain outcomes by spreading your bets on several or many potential outcomes. A sports gambler in Las Vegas faces unpredictable outcomes and so hedges their risk by placing bets on multiple games, even opposing teams. A weed in a field faces unknown soil and climate conditions so it hedges its bets by spreading as much seed as possible over space and time to increase the probability that one or more find conditions that will allow them to reproduce (Venable and Brown, 1988). Both do this with the goal of winning or surviving somewhere to cover your losses elsewhere. It’s the same idea behind diversifying your investment portfolio, or not putting all your eggs in one basket, and you can use it with cover crops.
Cover crop mixtures for bet-hedging
Bet hedging with cover crops spreads bets on a mixture of species expecting that at least one of them will grow in the conditions they will face. Because we don’t know what those conditions will be, planting a mixture may be less risky than trying to pick one crop species to match the unpredictable conditions. The key factors are the variability of the weather and the adaptability of different crops to different weather.
We all know we can’t depend on the weather, in some places less than others. If the weather was always “normal”, if it held to the long-term average, bet hedging wouldn’t be needed. The best species for the unvarying weather could be figured out and used every year. In reality, the varying weather presents a risk, primarily from precipitation and temperatures that deviates from the average. So, the first thing to consider is just how variable is the precip and temperature at your location for your cover crop growing window?
This variability can be evaluated using a statistical calculation called the standard deviation1. Those locations that are more likely to have rainfall close to average will have a low standard deviation (gray curve below). Where rainfall is more variable, the standard deviation will be higher (orange and blue curves).
The standard deviation (SD) not only tells us how far rainfall can be expected to vary from the average, but also shows the probability of such a variance. Rainfalls in the range of the +/- one standard deviation (SD) from the average will occur in about 68% of the years. For instance, we could have an average of 3″ with a standard deviation of +/- 1.2″ giving a range of 2.8 – 5.2″. Over the long-term, 68% of the years should have rainfall amounts in this range.
The other thing to consider is the amount of rainfall you can expect. In dryer regions, the variability of rainfall will be a greater factor than in wet regions: a 1″ deficit will mean more when the average is 6″ than when it is 10″. So, we need to consider the variability and compare it to the average. A simple way to do this is to calculate SD as a percent of the mean (SD/mean) x100. This is called the coefficient of variation, CV, and is also used as an indicator of stability. For the example above, this would be 1.2/3, or 40%. In a wetter climate with a 6″ average rainfall for the period in question, but with the same variability, CV would be 20% (1.2/6) indicating lower risk for your cover crop.
If I had great programming and data skills, I could make a cool animated map of the US showing long-term rainfall CVs for a moving 3 week fall cover crop planting window. Something like this, https://twitter.com/i/status/1255929005628252160
An animated map of normal daily precipitation. Watch what happens to rainfall over the High Plains in the late summer and fall, a common cover crop planting window). From Brian Brettschneider, @Climatologist49, on Twitter. Used with permission.
I can’t do that, so I will leave it to you to consider the rainfall in your region during your cover crop planting window. Ask your local climate expert for the precipitation coefficient of variation for your critical cover crop growth period. The same can be done with temperature if that is your critical weather factor.
Finally, consider that for more extreme deviations from normal rainfall, crop adaptations are less and less relevant because all crops need a certain amount of water and only thrive in a certain range of temperatures. In these situations, mixtures and monocultures would most likely fare equally badly.
Cover crop species differences
This brings us to the other key factor for bet-hedging with mixtures: the differences in crop species’ adaptability to variable weather. To make bet-hedging work with cover crop mixtures, we need the unpredictable factor; here it is precip and temperature. We also need to place our bets on several different outcomes to account for that unpredictability. The different outcomes are the different species in the mixture.
Here are some species adaptations that would be beneficial in a bet-hedging mixture:
Ability to germinate in lower moisture soil
Quick germination and establishment after a rainfall
Ability to survive a short-term drought after establishment
Crop species differ in these adaptations (e.g., Hunter and Erickson 1952 and many afterwards) but I did not find a source that summarized the differences – a good project for a graduate student.
Broadleaf or grass, legume, or Brassica, it does not matter as long as they expand the range of conditions such that at least one species in the mixture will handle what comes better than the others, i.e. produce some soil cover. Whether the mixture is better than a monoculture depends on if the range of conditions the mixture can handle is wider than that of any one crop species. If the mixture’s range is wider, and the weather varies enough across that range, it should establish successfully more often than a monoculture on average. There will be years where a monoculture will do better than the mixture.
In a review of cover crop mixture research (Florence and McGuire, 2020), cereal rye was nearly always the best producing species for the researched growing windows. It is the favored cover crop because it is easy to establish, fast growing, and surpasses all other species in growing in low temperatures. It is also widely adapted:
Since our most available cover crop windows are in the cool season, there are few cases where using cereal rye in a monoculture or a large share of a mixture would not make sense. Choosing other species that expand a mixture’s adapted range of conditions could be challenging for this window in cooler locations, but there may be some good options in warmer climates. I will not analyze all the situations for bet-hedging mixtures, but here is where I think the strategy may or may not work.
Where to use bet-hedging mixture and where not to?
Given the factors we have discussed, bet-hedging mixtures should have higher odds of success in these scenarios:
Generally, in locations where the critical weather (rainfall, temperature, or both) is highly variable.
Specifically, in regions that are normally dry and with variable rainfall, assuming you can find a mixture to handle the variability. Where variability is higher, the probability of picking the best species for monoculture decreases, but only so far. If the swings in rainfall are too drastic, then there may be no crops that grow well.
Longer growing windows that span two seasons. This is where bet-hedging mixtures have their best chance of success as the conditions across the growing window are going to vary much more than those within a single season. For example, a mid-summer cover crop that grows into fall and winter.
Where bet-hedging mixtures have lower odds of success:
Irrigated cropping systems. Irrigation removes the uncertain rainfall factor.
Late fall planting windows. Cool-season grasses will be the only species that can produce significant biomass, with cereal rye or winter triticale being the best.
Where weather variability is small from year-to-year, or in wetter regions, it will be less difficult to pick the best species to grow as a monoculture or a bi-culture mix of a grass with a legume to take advantage of the proven nitrogen benefits.
If rainfall has filled the soil profile before planting, then a large amount of the weather availability has been eliminated and the risk is much reduced.
Shorter growing windows. The shorter the window, the less variable weather the cover crop will face, and the less bet-hedging makes sense.
The inevitable tradeoffs
Note that bet-hedging is not the beneficial interaction of different plant species (symbiosis). It is solely a strategy to reduce the risk of a cover crop failure based on probability. And we know that because it is allowed in Las Vegas, it will never be a win-win in the long-term; you know who pays for all those lights.
Bet hedging always incurs some loss. Some of your teams will lose, some weed seeds will never produce a plant, but the risk of a complete loss is reduced. With cover crop mixtures, the loss comes in the cost of the seed for species that will not grow well, or that will be outcompeted by a species that does very well. If hedging is successful and some species grow well while others do not, the area given to those low producing species is lost, at least partly. The question is then, do benefits of hedging with mixtures justify the extra cost of seed? As discussed, that depends on your location, your weather, and the crops available to you.
Does it work?
Bet hedging presents a trade-off between stability over time and maximum yearly productivity. If it works, biomass production of the cover crop should be more stable (fewer lows and highs) over time. While the research is limited by the length of individual studies – generally not longer than 3 years – it gives some insight. In our review of cover crop mixture research (Florence and McGuire, 2020), the results of 30 comparisons from multiple studies suggested that monoculture biomass was less variable over time than that of mixtures. Another recent study not included in our review found similar results (Elhakeem et al., 2021). However, because these results are looking backwards after the results are found, they assume that the best species can be identified for each year. We have discussed why this is difficult in some locations and therefore why bet-hedging is beneficial in those locations.
So where does this leave us? For the common, late summer to early spring growing window, the odds of bet-hedging being a successful strategy seem low except in those favorable circumstances already listed. This is because the dominant factor will be the cold which dictates that a cool-season cereals such as cereal rye be used. Nevertheless, mixtures containing the best monoculture will likely produce similarly to the best monoculture, so it is worth trying. It comes down to preference, cost, and how you view the tradeoffs between monoculture and mixture. If the tradeoffs are acceptable and you want to grow a mixture, do it. Otherwise, figure out the best monoculture for your growing window and grow it.
1assuming a normal distribution of rainfall
Elhakeem, A., L. Bastiaans, S. Houben, T. Couwenberg, D. Makowski, et al. 2021. Do cover crop mixtures give higher and more stable yields than pure stands? Field Crops Research 270: 108217. doi: 10.1016/j.fcr.2021.108217.
Florence, A.M., and A.M. McGuire. 2020. Do diverse cover crop mixtures perform better than monocultures? A systematic review. Agronomy Journal 112(5). doi: 10.1002/agj2.20340.
Hunter, J.R., and A.E. Erickson. 1952. Relation of Seed Germination to Soil Moisture Tension. Agronomy Journal 44(3): 107–109.
Venable, D.L., and J.S. Brown. 1988. The Selective Interactions of Dispersal, Dormancy, and Seed Size as Adaptations for Reducing Risk in Variable Environments. The American Naturalist 131(3): 360–384. doi: 10.1086/284795.
Note on bet-hedging’s relationship to ecological biodiversity studies
This bet-hedging strategy with cover crop mixtures is not due to any synergism between species but purely a playing the odds strategy. It is similar to the sampling effect in ecological biodiversity studies (deLaplante and Picasso, 2011). The main difference is that the sampling effect occurs by chance in random mixtures of perennials, whereas bet-hedging occurs by choice in mixtures of specific species of highly bred annual crops. See here for more differences between ecological and agricultural research.
deLaplante, K., and V. Picasso. 2011. The Biodiversity-Ecosystem Function Debate in Ecology. In: deLaplante, K., Brown, B., and Peacock, K.A., editors, Philosophy of Ecology. North-Holland, Amsterdam. p. 169–200