Reducing Herbicide Use While Maintaining Crop Yields: Insights from a Crop Rotation Experiment

Weed communities in three cropping systems suitable for the Midwestern USA were studied from 2017 through 2020 to examine how crop diversification and the intensity of herbicide use affected weed stand density and aboveground mass. A baseline 2-year cropping system with corn and soybean grown in alternate years was diversified with cool-season crops, namely oat and red clover in a 3-year system, and with oat and alfalfa in a 4-year system. Herbicides were not used in the cool-season crops. In the cool-season crops, non-chemical weed suppression, was achieved via early crop planting (Picture 1) that gave crops a head start against weeds, crop harvest (Pictures 2 and 3), post-harvest stubble clipping as necessary, and fall tillage. This study was conducted to address the current gaps of information concerning how the density and biomass of weeds respond to different crop environments (the environment that a crop provides) and weed management programs.

Integrating chemical and cultural weed management tools resulted in an overall reduction in the amount of herbicide applied (Tables 1 and 2). In all the studied rotations, the corn phases under a low herbicide regime received banded herbicide application and interrow cultivation (Picture 4); the soybean phases received broadcast herbicide, and the oat and alfalfa phases (3-year and 4-year rotations) did not receive herbicide or cultivation. The reduction in herbicide use was associated with increases in weed density and aboveground mass. In the cool-season crop phases (oat intercropped with red clover, and oat intercropped with alfalfa) of the 3-year and 4-year rotations, the density of emerged weeds increased, but weed biomass did not increase, as compared with the warm-season crops (corn and soybean). Though there were more weeds (Table 3) in the lower-herbicide regime (banded herbicide and interrow cultivation) (Table 1), corn and soybean yields in the 3-year and 4-year rotations were as high or higher than in the conventionally managed 2-year rotation (Table 4).

Table 1: Weed management program from 2017 through 2020 field seasons. The baseline was a 2-year rotation of corn and soybean under conventional herbicide weed management. In the 2-year rotation under the low herbicide weed management regime, corn received banded herbicide on top of crop rows and interrow cultivation, while soybean received broadcast herbicide. The same conventional versus low herbicide designation was also used for the corn and soybean phases of the 3-year and 4-year rotations. Oat, red clover, and alfalfa received no herbicide or cultivation, regardless of the herbicide regime applied on corn and soybean. 

 “: identical to previous year

Table 2: Reduction in the amount of herbicide active ingredients applied in the more diverse 3-year and 4-year cropping systems as compared to a conventionally managed 2-year corn and soybean system, averaged over all crop phases of each rotation system from 2017 through 2020. See Table 1 for the description of other weed management practices. The minus sign in front of the percentages indicates the reduction in herbicide amount in the corresponding system as compared to the baseline.

Table 3: Weed abundance in three cropping systems. The baseline was a 2-year rotation of corn and soybean under conventional herbicide management, i.e., broadcast application. See Table 1 for description of other weed management practices. Zeroes were due to rounding. Means (± standard error, SE) were obtained from two linear models describing the effects of crop identity (crop species and rotation in which they occurred) and weed management regimes on weed density and biomass, respectively. The numbers in the table were converted from g/m² and plants/m² to oz/yd² and plants/yd².

¹: Soybean received broadcast herbicide, regardless of the preceding corn crop’s herbicide regime, but soybean grain yields were determined separately by the overall system herbicide regime.

²: Red clover intercropped with oat and alfalfa intercropped with oat were not harvested.

³: Oat grain and alfalfa hay were pooled at harvest, over two weed management regimes applied to their preceding corn phases.

Table 4: Mean crop yields in three cropping systems from 2017 to 2020. See Table 1 for a description weed management regimes. Corn, soybean, and oat yields are reported in bushels/acre (bu/ac); alfalfa yield is in tons/acre (ton/ac). County-specific alfalfa hay yields were not available for 2019 and 2020. Therefore, the Boone County alfalfa yield shown below was averaged over data for 2017 and 2018.

As shown in Table 5, giant and yellow foxtail were the dominant weeds in corn, common waterhemp dominated in soybean, common lambsquarters dominated in oat, and dandelion dominated in alfalfa. This observation suggested that alfalfa has the potential to shift weed community composition to weed species that are less challenging for corn and soybean phases of rotations. Knowing the challenging weed species in a field and documenting the weed pressure in response to a weed management program would be useful for adjusting management strategies to avoid outbreaks. Weed seedbank density could be used as a sustainability indicator (Storkey and Neve, 2018; Liebman et al., 2021), and having a record of weed seedbank composition over years could provide additional information for making long-term decisions about effective and sustainable weed management (Davis et al., 2005; Forcella et al., 1992; Forcella, 2003; Menalled et al., 2001).

Table 5: Relative abundance (% biomass of single species divided by the total biomass) of the most abundant weed species at the experiment site from 2017 through 2020. Means (± standard error, SE) were obtained from a linear model describing the effects of crop identity (crop species and rotation in which they occurred) and weed management regimes on weed relative abundance over four years. Zeroes without asterisks were due to rounding. Zeroes with asterisks were due to the nondetection of the species. See Table 1 for descriptions of herbicide management regimes.

The corresponding publication can be found at: https://www.frontiersin.org/articles/10.3389/fagro.2022.848548

The data can be found at: https://doi.org/10.25380/iastate.19111376

The code for data analysis can be found at: https://doi.org/10.5281/zenodo.5980943

References:

Davis, A. S., Renner, K. A., and Gross, K. L. (2005). Weed seedbank and community shifts in a long-term cropping systems experiment. Weed Science, 53(3), 296–306. https://doi.org/fkd3dj

Forcella, F. (2003). Debiting the seedbank: Priorities and predictions. Aspects of Applied Biology, 69, 151–162.

Forcella, F., Wilson, R. G., Renner, K. A., Dekker, J., Harvey, R. G., Alm, D. A., Buhler, D. D., and Cardina, J. (1992). Weed seedbanks of the U.S. Corn Belt: Magnitude, variation, emergence, and application. Weed Science, 40(4), 636–644.

Liebman, M., Nguyen, H. T. X., Woods, M. M., Hunt, N. D., and Hill, J. D. (2021). Weed seedbank diversity and sustainability indicators for simple and more diverse cropping systems. Weed Res, 61(3), 164–177. https://doi.org/ghz5bj

Menalled, F. D., Gross, K. L., and Hammond, M. (2001). Weed aboveground and seedbank community responses to agricultural management systems. Ecological Applications, 11(6), 1586–1601. https://doi.org/dpj63j

Storkey, J., and Neve, P. (2018). What good is weed diversity? Weed Res, 58(4), 239–243.https://doi.org/gdwv5r

Acknowledgment

Huong T. X. Nguyen*¹ and Matt Liebman²

^{*, 1} Corresponding author, former graduate student. Current address: School of Integrative Plant Science, Cornell University. hn337@cornell.edu

² Emeritus Professor. Department of Agronomy

The authors thank the following individuals and groups for their assistance with field work and laboratory activities: Matt Woods, Mike Fiscus, and the Iowa State University's Agronomy Research Farm crew for field management; Wendy Borja-Diaz, Lydia English, Jessica Juarez-Morales, Samantha Kanselaar, Jessica Nelson, Elizabeth Oys, Ana Poznanski, Andrew Riehl, Angela Soto-Saenz, Mickala Stallman, David Weisberger, and Wyatt Westfall for field and laboratory assistance.

The authors thank the following individuals and groups for their assistance with the corresponding publication: Katherine Goode, Audrey McCombs, Philip Dixon, and Iowa State's statistical consulting group for data analysis assistance; Russ Lenth and other Stackoverflow community members for answering HN's coding questions; and Micheal Owen for reviewing the manuscript.