Advertisement
Research Article

Increasing Cropping System Diversity Balances Productivity, Profitability and Environmental Health

  • Adam S. Davis mail,

    adam.davis@ars.usda.gov

    Affiliation: United States Department of Agriculture/Agricultural Research Service, Global Change and Photosynthesis Research Unit, Urbana, Illinois, United States of America

    X
  • Jason D. Hill,

    Affiliation: Department of Bioproducts and Biosystems Engineering, University of Minnesota, St. Paul, Minnesota, United States of America

    X
  • Craig A. Chase,

    Affiliation: Leopold Center for Sustainable Agriculture, Iowa State University, Ames, Iowa, United States of America

    X
  • Ann M. Johanns,

    Affiliation: Department of Economics, Iowa State University Extension and Outreach, Osage, Iowa, United States of America

    X
  • Matt Liebman

    Affiliation: Department of Agronomy, Iowa State University, Ames, Iowa, United States of America

    X
  • Published: October 10, 2012
  • DOI: 10.1371/journal.pone.0047149

Reader Comments (6)

Post a new comment on this article

Non-specialists need to be careful with the yield units.

Posted by daedalus2u on 20 Oct 2012 at 05:17 GMT

These yield figures can be easy to misinterpret to those not well versed in how to calculate yields over time. There is an article in the NYT about this that does misinterprets the yield figures, but I am unable to leave a comment there for some unknown reason (but have tried to contact the author).

The authors do not account for land-use efficiency. The “productivity” they are measuring is single year productivity, not the average over a long time. They explicitly state that this is what they are doing, but non-experts may not appreciate that even though the per-year yields per hectare are comparable, the total yields over the period per unit of land are not.

In the 2 year cycle, they get a maize and soybean crop every 2 years. In the 3 year cycle they get maize and soybeans every 3 years and in the 4 year cycle it is every 4 years.

If you have 120 hectares, then with the 2 year cycle you could alternate maize and soybeans on 60 hectares each. With the 3 year cycle you could alternate maize, soybeans and small grains on 40 hectares each. With the 3 year cycle you could alternate maize, soybeans, small grains and alfalfa on 30 hectares each.

Using the productivity values from the paper (read by eye from the graph), over the 9 years of the study, I estimate 6630, 4560 and 3471 Mg maize production for farming 120 hectares using the 2, 3 and 4 year cycles.

Doing the same thing for soy beans I get 1836, 1316 and 1020 Mg soy production for the 2, 3 and 4 year cycles.

These yields are not equivalent as some non-specialist journals have reported.

No competing interests declared.

RE: Non-specialists need to be careful with the yield units.

daniel_hicks replied to daedalus2u on 21 Oct 2012 at 20:28 GMT

After reading the above comment and reading the paper and supporting documents with this in mind, it seems to me the paper is ambiguous on exactly how mean yield is being calculated.

Figure S1 indicates that the overall test plot was divided into 4 large blocks, and each large block divided into 9 small blocks, with 1 small blocks used for each stage of the 3 different cropping systems. Consequently, the 4-year system takes up twice as much land as the 2-year system. Each small block was 18m x 85m, or 1530 m2, or 0.153 ha.

In such an arrangement, I see two different ways of calculating yield per hectare (Mg ha-1) for a given cropping system, crop, and year. Specifically, hectares can be measured as either (1) the total amount of land devoted to the cropping system "whatever crop is planted on it" or (2) the total amount of land devoted to the cropping system "planted with the given crop in the given year".

The paper reports a mean productivity per year for maize of 12.3 Mg ha-1 for the 2-year cycle and 12.9 Mg ha-1 for the 4-year cycle. If this is measured using measure (1), then in the average large block the 2-year cycle produced about 3.8 Mg of maize per year on 0.306 ha, while the 4-year cycle produced about 7.9 Mg of maize per year on 0.612 ha. This last is equivalent to about 3.95 Mg of maize per year on 0.306 ha. So the 4-year cycle produces about 4% more maize than the 2-year cycle, using measure (1).

If instead measure (2) was used, then in the average large block the 2-year cycle produced about 1.88 Mg of maize per year on 0.153 ha of 0.306 ha devoted to this system, while the 4-year cycle produced about 1.97 Mg of maize per year on 0.153 ha of 0.612 ha devoted to this system. This last is equivalent to about 0.985 Mg of maize per year if a total of 0.306 ha were devoted to the 4-year system. So the 4-year cycle would produce about 48% less maize than the 2-year cycle, using measure (2).

The above comment seems to assume that the authors of the paper used measure (2), and the paper is not explicit about which measure was used. The above comment seems to think that this is somehow related to short-term vs. long-term production, but I do not see how. As a non-specialist in agronomy, I'm not sure which of the two measures might be standard in the field. I would appreciate it if the authors would clarify their methods.

No competing interests declared.

RE: RE: Non-specialists need to be careful with the yield units.

asdavis1 replied to daniel_hicks on 22 Oct 2012 at 01:59 GMT

Dear Commenters 1 and 2,

First, thank you for your interest in our work. We are glad to have feedback from readers as well as reviewers.

Commenter #1 acknowledges that we state that corn and soybean are grown less frequently in the 3-yr and 4-yr rotations. Therefore I'm not clear about the source of his concerns.

Commenter #2: We report mean yields for crops on the land areas that they are grown on, not the ones they're not grown on. In the case of biological productivity, we present data concerning the total, rotation-level mass of harvested crop products, which would seem to me to be appropriate for addressing concerns about low productivity in low input systems.

For the lay reader, the most important point is not that corn or soybean are grown less frequently in the diversified systems, as commenter 1 suggests. Rather it is that all crops in the rotations have a food and market value, not just corn and soybean. Consequently, we make comparisons among rotations on the basis of net returns to a hectare of land split among two, three, or four crop phases.

There is a widespread belief that only corn and soybean have only food or market value. That is patently untrue, especially from the standpoint of global food security, considering that only a small fraction of U.S. corn and soybean exports make it to food insecure countries (Olmstead 2011; http://www.iatp.org/docum... )

Finally, crop rotations operate as systems. It is for that reason that we put the last paragraph into the manuscript and emphasized throughout the system level comparisons.

Regards,

Adam Davis

No competing interests declared.