Gérard Catroux

Trends in rhizobial inoculant production and use

Rhizobia inoculants have contributed to increase N2 fixation and yield in legumes crops. However, most of the inoculants produced world-wide are of poor or suboptimal quality. We discuss here why some of them are poor products and how to improve their quality and efficacy. Reported data on the inoculation rate effect can be used to design good inoculants. Technologies are now available to produce inoculants with a shelf-life of more than 1 year. Available quality control methods can help to improve the quality of inoculants although they do not take into account the physiological satus of the rhizobia. Unfortunately quality control is not commonly used except in major inoculant companies and the quality of inoculants sold on the market is low. The need for an increase in quality standards is discussed especially for the number of rhizobia delivered per seed and for the presence of contaminants. Some new technologies which able to increase efficacy and reliability of inoculation are discussed.

Effect of some fungicide seed treatments on the survival of Bradyrhizobium japonicum and on the nodulation and yield of soybean [Glycine max. (L) Merr.]

Several commercial fungicide seed treatments were evaluated for their possible effect on the survival of Bradyrhizobium japonicum on seeds and on the nodulation and yield of soybeans in a greenhouse and a field experiment. quinolate Pro (carbendazim and oxine copper), Vitavax 200FF (carboxin and thiram), and Monceren (pencycuron) had a small effect or no effect on the survival of B. japonicum and on the nodulation and yield of soybeans. They can thus be considered compatible with soybean seed inoculation. Germipro UFB (carbendazim and iprodione), Apron 35J (metalaxyl), and Tachigaren (hymexazol) decreased B. japonicum survival and the nodulation and yield of soybeans and thus cannot be considered compatible with soybean seed inoculation.

Changes in the physiological and agricultural characteristics of peat-based Bradyrhizobium japonicum inoculants after long-term storage.

Abstract Commercial soybean inoculants processed with sterilised peat and stored at 20 °C for 1–8 years were used as experimental materials to assess the changes in the physiological activity of Bradyrhizobium japonicum after storage. Viable counts decreased and physiological characteristics of the bacterium changed during storage, with an increase in the time taken for colony appearance on a medium without yeast extract, an increase in the lag time for nodule appearance on soybean grown in glass tubes and a decrease in survival on seeds. All the inoculants produced a significant increase in grain yield in a field experiment. The percentage of efficient cells in the field (relative to the plate counts) decreased as the length of storage increased. These results suggest that the physiological activity of B. japonicum cells changes after storage. Practical implications for inoculant quality control are discussed.

Decomposition de corps microbiens dans des sols fumiges au chloroforme: Effets du type de sol et de microorganisme

Five microbial species (Aspergillus flavus, Trichoderma viride, Streptomyces sp., Arthrobacter sp., Achromobacter liquefaciens) were cultivated in liquid media containing 14C-labelled glucose. The decomposition of these microorganisms was recorded in four different soils after chloroform fumigation by a technique related to that proposed by Jenkinson and Powlson, to determine the mineralization rate of microbial organic matter (Kc coefficient). Three treatments were used: untreated soil, fumigated soil alone and fumigated soil supplied with 14C-labelled cells. Total evolved CO2 and 14CO2 were measured after 7 and 14 days at 28°C. The labelled microorganisms enabled the calculation of mineralization rate Kc (Kc = mineralized microbial carbon/supplied microbial carbon). The extent of mineralization of labelled microbial carbon depended on the type of soil and on the microbial species. Statistical analysis of results at 7 days showed that 58% of the variance is taken in account by the soil effect and 32% by the microorganism effect. Between 35 and 49% of the supplied microbial C was mineralized in 7 days according to the soil type and the species of microorganism. Our results confirmed that the average value for Kc = 0.41 is acceptable, but Kc variability according to soil type must be considered. The priming effect on organic C and native microbial biomass mineralization, due to microbial carbon addition was obtained by comparison between the amount of non-labelled CO2-C produced by fumigated soils with or without added labelled microorganisms: this priming effect was generally negligible. These results indicate that the major portion of the error of microbial biomass measurement comes from the Kc estimation.

Survival and change in physiological state of Bradyrhizobium japonicum in soybean (Glycine max L. Merril) liquid inoculants after long-term storage

Commercial liquid inoculants for soybean, stored at 20 °C for 1–8 years in 400 ml bottles or in 5000 ml containers, were assessed for their efficacy and changes in the physiological activity of Bradyrhizobium japonicum. A decrease in viable counts and in bacterial survival on seeds was observed in inoculants stored for several years. The number of nodules produced per plant in a growth chamber decreased and was correlated to the number of bacteria surviving on the seeds. Changes in physiological properties were assessed using biochemical, physiological and microscopic methods. The cell total sugars content decreased with increased storage of the inoculants. High calculated ratios of suspended solid dry matter/carbon/nitrogen/proteins weight per c.f.u. strongly suggested the presence of dead or viable but non-culturable (VBNC) cells in the inoculants. This was confirmed in a study of bacterial respiratory activity, using p-iodonitrotetrazolium reduction. The time of colony appearance on plates increased in the old inoculants stored for a long time, especially on yeast-free culture medium. The heterogeneity in colony size also increased with storage length. Inoculants stored for more than 2 years could be differentiated from the others by using nalidixic acid against cellular division. Nucleic acid staining of cells showed that the percentage of membrane-compromised bacteria in all the inoculants increased with increased storage length, whatever the type of packaging used for the inoculants. These results demonstrated that the physiological activity of B. japonicum cells in commercial liquid inoculants changes after storage. To complete c.f.u. determination, three methods were proposed to assess the fitness of stored bradyrhizobia, but they remain to be checked for reliability on a variety of commercial inoculants.

Effect of inoculant density, formulation, dispersion and soil nutrient amendment on the removal of carbofuran residues from contaminated soil

Abstract Our study was aimed at investigating some factors affecting the use of a carbofuran-degrading bacterium as an inoculant to remove carbofuran residues from contaminated soil. Soil samples inoculated with a liquid cell suspension of strain C28 to give an initial density ranging from 10 3 to 10 7 cfu g −1 (dry weight of soil) significantly improved pesticide removal. When the soil was inoculated with a granular formulation of strain C28 to give theoretical cell densities of 5 × 10 3 , 5 × 10 4 and 5 × 10 5 cfu g −1 (dry wt of soil), a significant increase in the rate of carbofuran degradation, as compared to uninoculated samples, was seen only for the highest inoculum size. The rate of carbofuran degradation in non-inoculated samples was significantly reduced by chloroform fumigation, but inoculation of natural and chloroform-fumigated soil with equivalent cell densities of strain C28 in both liquid and granular formulation led to a significant increase in the rate of carbofuran breakdown. The effect of soil glucose amendment on pesticide degradation by strain C28, in a liquid or solid formulation, depended on the concentration of the sugar and on inoculant density. A low concentration of glucose significantly reduced the lag phase before pesticide degradation by a low density inoculant. A high concentration, however, led to a significant reduction of the degradation rate for all inoculum sizes. Mixing the siol after inoculation was essential to obtain rapid degradation of the pesticide. When theoretically-equivalent cell densities were introduced in 0.1 and 1 g of microgranules, no significant difference in the rate of pesticide degradation was observed in these experiments.

Evaluation of commercial soybean inoculants from Argentina

Eighteen samples of soybean inoculants representative of the major manufacturing companies in Argentina were purchased from the market and evaluated using plate counts, most probable number (MPN) of Bradyrhizobium japonicum on plants and time of nodule appearance. One or two B. japonicum isolates per product were isolated and typed by analysis of their DNA patterns. The Log10 numbers of B. japonicum obtained were in the range of 0 to 6/soybean seed, with only two products above 1 × 106 bacteria/seed. Of 18 products, 17 were contaminated, and of these 14 contained more contaminants than B. japonicum. The time of nodule appearance varied between 8 and 16 days, indicating a great difference in microbial activity between products. The strains were found to be similar to USDA 138 (five isolates), E45-INTA Argentina (two isolates), USDA 142 (four isolates) and E4-INTA (one isolate). Thus, even if most of the typed strains are considered as good N2-fixing strains, the average quality of the analysed samples was low, and could not support efficient inoculation of soybean.

Use of central composite experimental design for the development ofBradyrhizobium japonicum liquid inoculants

Central composite experimental designs realized with flask cultures and completed by fermentor studies were used to developB. japonicum liquid inoculants. Two different media are described, using glycerol or glucose, in the presence of only 1 g/l yeast extract. They allow the production of cultures containing more than 1010 viable cell/ml, able to nodulate soybean efficiently.

Improved enumeration of Bradyrhizobium japonicum in commercial soybean inoculants using selective media

Two selective media proposed for the enumeration of Bradyrhizobium japonicum were tested using six strains of different origin and eight different commercial soybean inoculants. These media contained tetracycline, rifampicin and chloramphenicol to control bacterial contaminants, and cycloheximide and pimafucin to control fungal contaminants. They were compared with previously described selective media and plant infection technique counts. The proposed media provided better control of contaminants than previously described media, gave counts of B. japonicum similar to those obtained by the plant infection technique, and so may be used for quality control of commercial inoculants.

Compatibility of a soybean peat inoculant with some seed applied fungicides and microgranular insecticides

Soybean cropping is being developed in Italy and France, and the practices of disinfection of seeds with fungicides and the treatment of soil with microgranular insecticides are increasing. Since farmers intend to combine seed and soil treatments with inoculation using Bradyrhizobium japonicum, several questions concerning the compatibility of soybean inoculants with these practices have arisen. We describe the procedure we are using to assess the possible effects of both fungicides and micro‐granular insecticides on survival of Bradyrhizobium japonicum during both inoculation and early nodulation of the soybean in the glasshouse. Reliability and possible conclusions are discussed.