Which nitrogen fertilization techniques and crop traits to promote biological weed regulation by competition?
Résumé
Background and Aims
The necessary reduction of herbicide use may result in a residual weed flora in arable fields. Moreover, the use
of mineral fertilizer must be reduced for environmental reasons. Thus, crop-weed competition for nitrogen might
increase.
Nitrogen competition among plants is driven by soil nitrogen availability and plants traits related to nitrogen nutrition.
Selecting species for their nitrogen nutrition traits and adapting nitrogen fertilization (rate, date, number of
nitrogen applications, straw burial or export) could change competitive relationships between plants. Weeds with
high nitrogen nutrition requirements are often hard to manage and could be hampered by such changes. This
study aimed to identify which traits and which nitrogen fertilization options can reduce weed harmfulness while
maintaining crop production in the long term.
Material and Methods
A maize monoculture from Aquitaine (France) was simulated over 15 years and 5 weather repetitions using FLORSYS
(Gardarin et al. (2012), Munier-Jolain et al. (2013), Colbach et al. (2014)). This process-based model simulates
crop growth and weeds dynamic over the years with a daily time step from cropping system, pedoclimate and
species traits. We simulated numerous combinations of nitrogen fertilization (nitrogen input date and rate, straw
burial vs export) and initial soil nitrogen content with 3 maize varieties (1 actual, 2 virtual) differing by their trait
values related to nitrogen nutrition. The effects of fertilization, soil nitrogen supply and maize variety on weed
disservice and service indicators were studied (Figure 1B).
Key results
The simulations showed that service indicators were often positively correlated to disservice indicators (Figure 1A).
The sensitivity analysis revealed that maize variety and initial soil nitrogen content were the variables that affected
the most weed impact indicators. The interactions between maize variety and nitrogen fertilization techniques
were strong.
Compared to the actual maize variety, one virtual variety improved weed (dis)services except nitrate leaching
(Figure 1B). The other virtual variety had the opposite effect. The effect of nitrogen application date was different
for all 3 varieties, except for bee food which decreased with the lateness of application, regardless of the variety.
For the actual variety, increasing nitrogen rate decreased weed harmfulness over the long term, and increased
weed services (except bee food). The 2 virtual varieties showed the opposite for weed services and a similar behavior
as actual variety on disservices.
Increasing the initial soil nitrogen content decreased current but increased future weed harmfulness (Figure 1B).
On the long term (beyond 15 years), a high initial soil nitrogen content might increase the weed soil seed bank and
thus decrease grain yield loss.
Further analyses and simulations will be carried out to (1) disentangle the effects of fertilization techniques from
maize variety and initial soil nitrogen content and (2) identify which weed traits drive weed (dis)services in these
maize-based cropping systems.
Conclusions
The preliminary results of these simulations of maize cropping systems showed that choosing varieties based on
their nitrogen nutrition traits can reduce weed disservices. However, it was almost impossible to conciliate reduction
of weed harmfulness and increase of weed services in these systems.