Crop Diversity is Not the Cause

I previously covered reasons why the results of many biodiversity studies do not apply to agriculture. Here, I want to emphasize a related issue: how ecological biodiversity research has distracted us from figuring out the actual causes of benefits in crop mixtures.

Title of journal article "Diversity increases yield but reduces harvest index in crop mixtures"

Consider the paper above (Chen et al., 2021). Although the title says “Diversity increases yield… in crop mixtures,” this is not true. Diversity didn’t increase the yield. Crop diversity may result in, be linked with, or be related to these benefits, but diversity itself is not the cause. This is a clear case of correlation, not causation. Plant diversity is a feature of species mixtures, not a driver of function. The actual cause — the specific physical and chemical interactions between species in the mixtures—is not determined, not in this study nor in a multitude of other plant diversity studies. In fact, it is rare for such studies to even attempt to figure out a mechanism (Barry et al., 2019; Cardinale et al., 2011; Letourneau et al., 2011). Here is the problem: this same ecological take on plant diversity research has been brought to research on crop diversity. However, in agriculture the results are less than helpful: without knowing the actual mechanism, mixtures are just a guess and benefits are uncertain. So why don’t biodiversity studies determine causes or mechanisms?

“Diversity is the outcome of ecological processes and not an ecological process in itself… Alas, diversity is not good or bad, it simply is.” Shade (2017)

Diversity is easy, mechanisms are not

First, because figuring out the mechanism is difficult. It’s relatively easy to link function to diversity. You count the species, you measure function. Finding the mechanism, however, requires much more time and effort. It can involve genetics, biochemistry, or even worse, digging, sorting, and measuring roots. It may be beyond our current ability to determine. And the results are often unclear.

Mixture of plants and grasses
Like a gambler with a hand of cards, a farmer with a potential crop mixture would like to know what to throw away and what to keep (HT Kenny Rogers). Knowing the mechanism that provides the benefits would allow this. Photo: A. McGuire.

Given all the intercropping and plant diversity research done over the years, we have found two primary mechanisms. First there is the long-known nitrogen effect (mostly) of mixing legumes with non-legumes on low nitrogen soils. The other, more recently identified (van Ruijven et al., 2020), is not due to anything happening in mixtures but rather, a disadvantage of monocultures. In monocultures of perennials, or continuous cropping of annuals, soilborne disease builds up and eventually leads to reduced yields. Mixtures avoid this fate by dispersing host plants of specific soilborne pathogens. That’s it for reliable mechanisms. Other than these two1, potential mechanisms are more often assumed than measured (Barry et al., 2020; Cardinale et al., 2011; Turnbull, 2014; Wuest et al., 2021).

“increasing biodiversity effects over time could therefore result from declining monocultures instead of reflecting increases in the functioning of mixtures” (Marquard et al., 2013)

While this difficulty in finding actual causes and the scarcity of identified mechanisms may dissuade such research, the real damper on finding mechanisms is that ecology doesn’t need to do this challenging work to accomplish its purposes.

Ecology and agriculture have different goals

“discourses on biodiversity have become a rhetorical veneer hiding academic strategies to optimize funding and an opiate sustaining the comfortable belief that biodiversity knowledge unquestionably contributes to the mitigation of biodiversity loss.” (Devictor and Meinard, 2020)

The aims of biodiversity research are not those of agriculture. Ecologists conduct biodiversity research as a response to the “biodiversity crisis” (Maier, 2012). Their answer to the question, “How do we best go about conserving biodiversity?” is “We measure the benefits of having higher biodiversity over lower and make decisions based on those benefits.” To justify conservation of biodiversity, all ecology needs to show is that higher biodiversity is linked to higher function. The actual mechanisms are not relevant. Instead, ecology studies most often point to the “complementary effect,” which is statistically determined and only hints at the actual mechanisms. Or they name potential mechanisms in general terms: better resource use, complementarity, facilitation, etc. (Fridley, 2001; Vanelslander et al., 2009). While sufficient for ecology’s purposes, this does not help agriculture manage crop mixtures.

Ecological biodiversity research methods do not advance agriculture

“[Soil] delivers multiple ecosystem services, which are provided by soil processes and functions performed by soil biodiversity” (Bertola et al., 2021)

Finally, as shown in the paper mentioned at the beginning (Chen et al., 2021), agricultural research of plant biodiversity has borrowed the same research methods used in ecology even though the goals are different. The results are not surprising. Instead of agriculture being laser focused on ag-relevant metrics and determining mechanisms, it often gives the same results as ecological research: biodiversity is linked to improved function. It even appears that many doing this work in agriculture have come to presume that biodiversity is the actual cause and so showing the link between plant diversity and biomass production is enough. This is an easy mistake to make given all the ecological papers that use causal language to describe the link. And if one believes that biodiversity is the cause, then why look further?

Agriculture needs mechanisms to manage

“so far the underlying mechanisms remain poorly understood” Wuest et al., (2021), Ecological and evolutionary approaches to improving crop variety mixtures.

In agriculture, we don’t want to guess which mixtures will perform best. We don’t need more observed links between crop diversity and function. We need to know the real cause of benefits coming from the better performing mixtures. We can manage a cause, we can’t manage a correlation. To achieve this, we must shift our focus from biodiversity itself to determining the specific interactions between specific species in specific environments. There are good examples of papers that do look for, and sometimes find, the actual mechanisms, including benefits of legume+non-legumes mixtures that are not related to nitrogen (Hu et al., 2021). Hard work? Yes, but necessary to improve reliability of all forms of crop diversity.

1I don’t count bet-hedging (here for a post on that topic) because it is not a physical mechanism but a method of improving the odds of matching a species to the varying weather.


  • Barry, K.E., L. Mommer, J. van Ruijven, C. Wirth, A.J. Wright, et al. 2019. The Future of Complementarity: Disentangling Causes from Consequences. Trends in Ecology & Evolution 34(2): 167–180. doi: 10.1016/j.tree.2018.10.013.
  • Barry, K.E., J. van Ruijven, L. Mommer, Y. Bai, C. Beierkuhnlein, et al. 2020. Limited evidence for spatial resource partitioning across temperate grassland biodiversity experiments. Ecology 101(1): e02905. doi: 10.1002/ecy.2905.
  • Bertola, M., A. Ferrarini, and G. Visioli. 2021. Improvement of Soil Microbial Diversity through Sustainable Agricultural Practices and Its Evaluation by -Omics Approaches: A Perspective for the Environment, Food Quality and Human Safety. Microorganisms 9(7): 1400. doi: 10.3390/microorganisms9071400.
  • Cardinale, B.J., K.L. Matulich, D.U. Hooper, J.E. Byrnes, E. Duffy, et al. 2011. The functional role of producer diversity in ecosystems. Am. J. Bot. 98(3): 572–592. doi: 10.3732/ajb.1000364.
  • Chen, J., N. Engbersen, L. Stefan, B. Schmid, H. Sun, et al. 2021. Diversity increases yield but reduces harvest index in crop mixtures. Nat. Plants: 1–6. doi: 10.1038/s41477-021-00948-4.
  • Devictor, V., and Y. Meinard. 2020. Empowering biodiversity knowledge. Conservation Biology 34(2): 527–529. doi: 10.1111/cobi.13367.
  • Fridley, J.D. 2001. The influence of species diversity on ecosystem productivity: how, where, and why? Oikos 93(3): 514–526. doi: 10.1034/j.1600-0706.2001.930318.x.
  • Hu, H.-Y., H. Li, M.-M. Hao, Y.-N. Ren, M.-K. Zhang, et al. 2021. Nitrogen fixation and crop productivity enhancements codriven by intercrop root exudates and key rhizosphere bacteria. Journal of Applied Ecology. doi: 10.1111/1365-2664.13964.
  • Letourneau, D.K., I. Armbrecht, B.S. Rivera, J.M. Lerma, E.J. Carmona, et al. 2011. Does plant diversity benefit agroecosystems? A synthetic review. Ecological Applications 21(1): 9–21.
  • Maier, D.S. 2012. What’s So Good About Biodiversity?: A Call for Better Reasoning About Nature’s Value. Springer Netherlands.
  • Marquard, E., B. Schmid, C. Roscher, E. De Luca, K. Nadrowski, et al. 2013. Changes in the Abundance of Grassland Species in Monocultures versus Mixtures and Their Relation to Biodiversity Effects (J. Moen, editor). PLoS ONE 8(9): e75599. doi: 10.1371/journal.pone.0075599.
  • van Ruijven, J., E. Ampt, D. Francioli, and L. Mommer. 2020. Do soil-borne fungal pathogens mediate plant diversity–productivity relationships? Evidence and future opportunities. Journal of Ecology.
  • Shade, A. 2017. Diversity is the question, not the answer. ISME J 11(1): 1–6. doi: 10.1038/ismej.2016.118.
  • Turnbull, L.A. 2014. Ecology’s dark matter: The elusive and enigmatic niche. Basic and applied ecology 15(2): 93–100.
  • Vanelslander, B., A.D. Wever, N.V. Oostende, P. Kaewnuratchadasorn, P. Vanormelingen, et al. 2009. Complementarity effects drive positive diversity effects on biomass production in experimental benthic diatom biofilms. Journal of Ecology 97(5): 1075–1082. doi: 10.1111/j.1365-2745.2009.01535.x.
  • van Ruijven, J., E. Ampt, D. Francioli, and L. Mommer. 2020. Do soil-borne fungal pathogens mediate plant diversity–productivity relationships? Evidence and future opportunities. Journal of Ecology.
  • Wuest, S.E., R. Peter, and P.A. Niklaus. 2021. Ecological and evolutionary approaches to improving crop variety mixtures. Nat Ecol Evol: 1–10. doi: 10.1038/s41559-021-01497-x.


11 comments on "Crop Diversity is Not the Cause"
  1. To dismiss ecological investigations as focused on only one question is insulting. Is the only question in agriculture “how do we improve yield”? Obviously, we need to consider what our actions do to soils and the biotic commununity they contain. Will that increased yield be sustainable or are we just pissing away the benefits we have gained through previous complex ecological systems?

    1. Gary, thanks for the comment. While my example study was focused on yield, diversity studies address many other questions. In all of them, diversity is not the cause of measured benefits, which is my main point.

  2. Functionality as a goal and understanding mechanisms (and potential pitfalls e.g. pathogens, varying performance dependent on weather etc) are key to useful agricultural applications or advice. Your distinctions are logical, valuable, and (unfortunately) brave given certain sensitivities and prevailing dogmas. Great article thank you.

  3. Andrew
    As a cherry orchardist and ecologist, I understand how complex biological and physical interrelationships can be. Attempting to understand the causal mechanisms is critical and extremely difficult in either field. I admire any attempt to separate out these mechanisms. Often this can’t be accomplished with the information we have today.
    However there are numerous studies that show the removal of certain pieces of any ecosystem can reduce it’s functionality and persistence. That may only be a correlation, but not one to be taken lightly. Trying to reduce any system to it’s individual parts can and has lead to significant mistakes.

    Thanks, Gary

  4. Thanks for this excellent and thought provoking article. Also the link to your earlier blog. As always your articles are challenging and clear sighted. This makes them immensely valuable to all – farmers and citizens as well as researchers. I have been involved in reserach projects in what we termed “functional biodiversity” since the 1980s; including work on mixtures, hrabal leys, population breeding and inter-cropping. The points you make about appropriateness of methods and metrics such as Land Equivelent Ratios (in the earlier post) are things we have struggled with over the years. Unfortunately, researchers and some influential farmers seem more unwilling to challenge “buzzwords” and fashionable ideas today than they were in the past. Because of this I fear we are venturing down a myth and cliched driven mnarrative cul-de-sac that helps no-one. Thank you for your reason and balance.

  5. Why Ecological Biodiversity Research Results Seldom Apply to Agriculture | CSANR | Washington State University says:

    […] The questions that ecology needs to answer for its conservation efforts are not relevant to agriculture. Rather, agriculture wants to know which mixtures would provide greater productivity, resource-use efficiency, or other benefits above those found in monocultures. Even more important, it wants to know the mechanisms behind such effects. At present, the only reliable mixture is that of legumes with non-legumes in a low-nitrogen soil. It is reliable because we know the mechanism that causes the benefits. Autecology (Walter and Hengeveld, 2014) is a branch of ecology that could help here. It focuses not on population and community dynamics, but on the specific physiochemical interactions between species and between species and their environments. More on this here. […]

  6. Crop diversification is need of time especially for farmers in poor economies. This climate change impact is wreaking havoc of farmers’ livelihoods.

  7. […] It is not the cause of soil function, and therefore not so helpful in determining management. Read more here and […]

  8. […] (Jones et al. 2023). Why? Because the focus on diversity from the holistic/emergence viewpoint obscures real mechanisms. Reductionism works because it gives us real, understandable mechanisms to […]

Leave a Reply

Your email address will not be published. Required fields are marked *