Elsevier

Field Crops Research

Volume 283, 1 July 2022, 108537
Field Crops Research

Seed inoculation with Azospirillum brasilense in the U.S. soybean systems

https://doi.org/10.1016/j.fcr.2022.108537Get rights and content

Highlights

  • Strong yield response to co-inoculation was observed in only 2 out of 25 site-year.

  • Relative abundance of ureides was not affected by Azospirillum or co-inoculation.

  • The probability of maximum yield response (0.124 Mg ha−1) to Azospirillum was 5.3%.

  • Seed yield responses were linked to soil phosphorus and were reduced under drought.

  • There is negligible chance of inoculation to increase seed protein concentration.

Abstract

Symbiotic nitrogen (N) fixation (SNF) is critical to satisfying the nutritional need of soybean (Glycine max (L.) Merr.) and maintaining productivity and high seed protein concentration. Due to its low environmental impact, a key factor for increasing the sustainability of soybean systems is to enhance SNF. Seed inoculation with the free-living Azospirillum brasilense alone or with Bradyrhizobium japonicum (herein called co-inoculation) are plausible strategies that have been explored in tropical environments but lack information in temperate regions. Following this rationale, this study aimed to evaluate the impact of seed inoculation with Azospirillum brasilense (herein called Azospirillum) alone or combined with Bradyrhizobium japonicum (herein called Bradyrhizobium) in a range of environments in the United States (US) for: (i) seed yield, (ii) relative abundance of ureides (RAU) as a proxy of SNF, and (iii) seed protein concentration. Twenty-five field studies across the US states with the same experimental design were performed during the 2019 and 2020 growing seasons. The primary outcomes of this research were: (i) yield responses to co-inoculation were considered significant in only 2 out of 25 site-years, (ii) RAU was not increased by Azospirillum inoculation or co-inoculation, and lastly, (iii) seed protein concentration was marginally associated with the inoculation strategies. Although Azospirillum did not impose remarkable gain in any observed plant traits, future studies should focus on mechanistically understanding whether Azospirillum can naturalize in temperate region soils. Still, strategies for enhancing SNF are required for sustainably improving productivity and quality for US soybean systems.

Introduction

Crop systems are entailed to increase food and energy production while providing environmental services (Ladha and Chakraborty, 2016, Lal, 2015, Thomson, 2003). To some extent, soybean systems are more sustainable than other row crops due to the low dependency on nitrogen (N) fertilizers due to the crop’s reliance for obtaining N from the symbiotic N fixation (SNF) relationship with bacteria of the genus Bradyrhizobium. This large-scale biological process provides 16.5–40 Tg N year−1 (Galloway et al., 2008, Herridge et al., 2008). Although the SNF process occurs spontaneously due to soil indigenous Bradyrhizobium population, some regions do not have naturalized populations and farmers commonly apply inoculants (Chibeba et al., 2020, Hungria et al., 2015a, Leggett et al., 2017a). However, the increase in soybean production has been associated with decreased protein concentration during the last decades (de Borja Reis et al., 2020) with a negative relationship between seed yield and seed composition in modern genotypes (Assefa et al., 2018). Thus, leading the thought that N supplied by SNF is insufficient to simultaneously maintain plant N demand and seed protein concentration (Cafaro La Menza et al., 2019), thereby impacting the needs of the feed industry (Wilson et al., 2014), and impairing further yield increases.

A physiological limitation for SNF in a high-N demanding environment could constrain either yield and/or seed protein synthesis (Cafaro La Menza et al., 2017, Ciampitti and Salvagiotti, 2018), but there is no solid evidence that soybean systems would benefit from a substantial N fertilizer input. An extensive database with more than 207 experiments across the U.S. applying from 0 to 560 kg ha−1 (Mourtzinis et al., 2018) revealed a yield increment of only 110 kg ha−1 in N-fertilized over unfertilized treatments. More recently, Cafaro La Menza et al. (2019) with N fertilizer applications up to 870 kg ha−1 reported a yield gain of 600 kg ha−1 compared to only inoculated treatment. These modest yield responses under unrealistic (and uneconomical) rates for management purposes reinforce results reporting the absence of yield and protein response to N fertilization (de Borja Reis et al., 2021, Hungria et al., 2006, Ortez et al., 2018, Tamagno et al., 2018), and support recommendations not to add N fertilizer for most soybean growing conditions. There is conclusive documentation that soybean suppresses SNF in rich N environments, resulting in non-significant additional N uptake (Fabre and Planchon, 2000, Salvagiotti et al., 2008). Whether the inability to increase N supply represents a physiological limitation for yield and protein synthesis is still a central research question that needs to be addressed by focusing on boosting the ability of the SNF to provide more N to the soybean plant.

From the symbiotic context, developing elite Bradyrhizobium strains with increased fixation efficiency has been proposed as one target for increased N fixation (Mendes et al., 2004), although a trade-off between efficiency and environmental acclimatization hamper the use of elite strains in field conditions (Batista et al., 2007, McLoughlin et al., 1990, Muller and Denison, 2018). Another approach is to increase soybean exposure to higher rates of inoculants (Albareda et al., 2009, Brockwell et al., 1989) and re-application during the growing season (Hungria et al., 2015b, Kaschuk et al., 2016). Both techniques have shown erratic responses that appear highly dependent on environmental conditions with null to marginal results in mature soybean systems (Carciochi et al., 2019). More recently, the effect of co-inoculation of Bradyrhizobium with Azospirillum has been summarized in meta-analyses describing positive impacts on yields and seed N concentrations (Barbosa et al., 2021), nodule count and biomass (Zeffa et al., 2020). Azospirillum is a free-living N-fixing bacteria that also promotes plant growth (Bashan and De-Bashan, 2010, Fibach-Paldi et al., 2012) with positive yield results in soybean (Fukami et al., 2018, Hungria et al., 2015b) and other field crops (Pankievicz et al., 2019, Zeffa et al., 2019). Azospirillum also appears to directly stimulate Bradyrhizobium N-fixing activity (Moretti et al., 2020a). However, field results are still sparse, with most research executed in tropical regions (Barbosa et al., 2021). In fact, the effect of Azospirillum seems to be associated with sandy soil, organic levels below 25 mg dm−3, and to environments prone to drought stress (Barbosa et al., 2021). The estimated yield gain of roughly 3% (Barbosa et al., 2021, Zeffa et al., 2020) is similar to some responses reported for N fertilizer (Mourtzinis et al., 2018), although with an arguably lesser environmental impact. However, it is unknown if Azospirillum can benefit soybean production in temperate regions with more fertile soils (e.g., high soil organic matter, less acidic pH).

A lesson learned from previous investigations on co-inoculation is that the experimental framework should comprehend a broad range of environmental conditions while dealing with a rather small expected effect imposed by the treatments. A slight response is unlikely to be considered a rare event and therefore not caused by random chance (Smith et al., 2002) using traditional frequentist analysis. In this scenario, a hierarchical Bayesian framework (Ellison, 2004) would consider the error structure of multiple environment experiments presenting the full posterior distribution of effect probabilities instead of inferring from a strict significance threshold (Berger, 2003). A Bayesian approach permits obtaining the expected value of the response with a probabilistic component (Ellison, 2004). In addition, Bayesian methods can provide an exact quantification of the likelihood for differences between treatments and their levels of uncertainty, quantifying credibility intervals (Ellison, 2004).

We hypothesized that similar benefits of Azospirillum alone or co-inoculation described in the tropical zone would be partially reproduced in U.S. soybean production systems. To understand potential gains on crop performance and N fixation, the objectives of this study were to evaluate the effects of seed inoculation with Azospirillum alone or combined with Bradyrhizobium sp. in a broad range of environmental conditions on: (i) yield, (ii) relative abundance of ureides as a proxy of SNF, and (iii) seed protein concentration.

Section snippets

Treatment and design structure

A total of 25 field experiments were distributed across 10 U.S. states, encompassing latitudes from 36.01ºN to 47.00ºN and longitudes from 84.46ºW to 97.10ºW in the U.S. Midwest and Midsouth regions (Fig. 1). From the 25 total sites, 13 studies were carried out during 2019 and 12 during 2020 growing seasons. Evaluated treatments constituted of four inoculation strategies: (1) absence of inoculation (control); (2) inoculation only with Bradyrhizobium japonicum; (3) inoculation only with

Results

Seed yield observations ranged from 1.8 Mg ha−1 (Lansing, MI 2019) to 6.6 Mg ha−1 (Princeton, KY 2019) and averaged 4.1 Mg ha−1 in 2019 and 4.0 Mg ha−1 in 2020. According to the individual models for the site-year factor, Lafayette IN 2020, Reed ND 2020, and Prosper ND 2019 were the only locations portraying evidence of seed yield changes coming from inoculation strategies greater than the control posterior probability (Fig. 2, a). At Lafayette, IN 2020, the application of Azospirillum alone

Discussion

This is the first study investigating the effect of co-inoculation of Azospirillum and Bradyrhizobium in an extensive network of field studies carried out in several soybean-producing states of the United States. The overall yield response to co-inoculation was erratic with a modest probability of significant gains (maximum response of 0.096 Mg ha−1 in the full model). From the standpoint of the statistical method, one of the main advantages of implementing a Bayesian approach is that it helps

Conclusion

This study is the first extensive network of field experiments reporting the effect of co-inoculation practice in a temperate U.S. soybean production region. The lack of greater impact of co-inoculation and other strategies portrays the complexity of exploring direct avenues for enhancing N fixation for soybean using beneficial microorganisms. Soil and environmental conditions for the U.S. soybean systems are different relative to the responses of this practice encountered in tropical

CRediT authorship contribution statement

Andre F.B. Reis: Methodology, Formal analysis, Visualization, Writing − original draft. Luiz M. Rosso: Data curation, Methodology, Formal analysis, Visualization, Writing − original draft. Eric Adee: Methodology, Data curation, Writing − review & editing. Dan Davidson: Data curation, Writing − review & editing. Péter Kovács: Data curation, Writing − review & editing. Larry C. Purcell: Data curation, Writing − review & editing. Frederick E. Below: Data curation, Writing − review & editing. Shaun

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

United Soybean Board funded this research, project no. 2020–152-0104. This is contribution no. 22-278-j from Kansas Agricultural Extension Station. We appreciate the support of TerraMax Inc. for providing the inoculants. Michigan State trials received support from Michigan Soybean Committee. The North Dakota trials were funded by the North Dakota Agricultural Experiment Station. This work could not have been possible without the effort of Conner Raymond, Hunter Adams, Curtis Bradley, Sloan

References (75)

  • K.J. Thomson

    World agriculture: towards 2015/2030: an FAO perspective

    Land Use Policy

    (2003)
  • E.-J. Wagenmakers et al.

    Bayesian hypothesis testing for psychologists: A tutorial on the Savage–Dickey method

    Cogn. Psychol.

    (2010)
  • J.T. Abatzoglou

    Development of gridded surface meteorological data for ecological applications and modelling

    Int. J. Climatol.

    (2013)
  • Y. Assefa et al.

    Spatial characterization of soybean yield and quality (amino acids, oil, and protein) for United States

    Sci. Rep.

    (2018)
  • J.S.S. Batista et al.

    Variability in Bradyrhizobium japonicum and B. elkanii seven years after introduction of both the exotic microsymbiont and the soybean host in a cerrados soil

    Microb. Ecol.

    (2007)
  • J.O. Berger

    Could Fisher, Jeffreys and Neyman Have Agreed on Testing?

    Stat. Sci.

    (2003)
  • J. Brockwell et al.

    Effects of soil nitrogen status and rate of inoculation on the establishment of populations of Bradyrhizobium japonicum and on the nodulation of soybeans

    Aust. J. Agric. Res.

    (1989)
  • P.-C. Bürkner

    Advanced Bayesian Multilevel Modeling with the R Package brms

    R. J.

    (2018)
  • P.-C. Bürkner

    brms: An R Package for Bayesian Multilevel Models Using Stan

    J. Stat. Softw.

    (2017)
  • W.D. Carciochi et al.

    Soybean yield, biological N2 fixation and seed composition responses to additional inoculation in the United States

    Sci. Rep.

    (2019)
  • N.W. Chaney et al.

    POLARIS Soil Properties: 30‐m Probabilistic Maps of Soil Properties Over the Contiguous United States

    Water Resour. Res.

    (2019)
  • A.M. Chibeba et al.

    Towards sustainable yield improvement: field inoculation of soybean with Bradyrhizobium and co-inoculation with Azospirillum in Mozambique

    Arch. Microbiol.

    (2020)
  • I.A. Ciampitti et al.

    New insights into soybean biological nitrogen fixation

    Agron. J.

    (2018)
  • A.F. de Borja Reis et al.

    Environmental factors associated with nitrogen fixation prediction in soybean

    Front. Plant Sci.

    (2021)
  • A.F. de Borja Reis et al.

    Historical trend on seed amino acid concentration does not follow protein changes in soybeans

    Sci. Rep.

    (2020)
  • J.L. de Bruin et al.

    Probability of Yield Response to Inoculants in Fields with a History of Soybean

    Crop Sci.

    (2010)
  • A.M.L. de Naoe et al.

    Co-inoculation with azospirillum brasilense in soybean cultivars subjected to water deficit

    Rev. Bras. De. Eng. Agric. e Ambient.

    (2020)
  • W.R. Ellis et al.

    Persistence and recovery of rhizobium japonicum inoculum in a field soil 1

    Agron. J.

    (1984)
  • A.M. Ellison

    Bayesian inference in ecology

    Ecol. Lett.

    (2004)
  • W.R. Fehr et al.

    Stage of Development Descriptions for Soybeans, Glycine max (L.) Merrill 1

    Crop Sci.

    (1971)
  • S. Fibach-Paldi et al.

    Key physiological properties contributing to rhizosphere adaptation and plant growth promotion abilities of Azospirillum brasilense

    FEMS Microbiol. Lett.

    (2012)
  • G.M. Fipke et al.

    Osmoprotectant in soybean seeds can increase the inoculation and co-inoculation time in pre-sowing

    Aust. J. Crop Sci.

    (2020)
  • J. Fukami et al.

    Azospirillum: benefits that go far beyond biological nitrogen fixation

    AMB Express

    (2018)
  • F.S. Galindo et al.

    Revista Brasileira de Engenharia Agrícola e Ambiental Modes of application of cobalt, molybdenum and Azospirillum brasilense on soybean yield and profitability

    Rev. Bras. De. Eng. Agríc. e Ambient.

    (2017)
  • F.S. Galindo et al.

    Technical and economic viability of co-inoculation with azospirillum brasilense in soybean cultivars in the Cerrado

    Rev. Bras. De. Eng. Agric. e Ambient.

    (2017)
  • J.N. Galloway et al.

    Transformation of the Nitrogen Cycle: Recent Trends, Questions, and Potential Solutions

    Science

    (2008)
  • P.H. Graham

    Stress tolerance in Rhizobium and Bradyrhizobium, and nodulation under adverse soil conditions

    Can. J. Microbiol.

    (1992)
  • Cited by (8)

    • System of pre-sowing seed inoculation

      2023, Scientific Horizons
    View all citing articles on Scopus
    View full text