Nodulation and nitrogen fixation of field grown common bean (Phaseolus vulgaris) as influenced by fungicide seed treatment

A field experiment was conducted at Bel Air station, in Dakar using N isotope dilution technique and the non nodulating soybean (Glycine max) variety m129 as reference plant to test the compatibility of Dichlorofenthion-thiram (DCT) fungicide to the inoculation of common bean (Phaseolus vulgaris) Paulista variety with both Rhizobium etli ISRA 353 and R. tropici strain ISRA 554. Nodulation was not induced with R. etli ISRA 353 and nitrogen fixation did not occur. With R. tropici ISRA 554, a decrease in nodulation was observed, but nitrogen fixation was not significantly different compared to that of the non DCT-treated common bean.


INTRODUCTION
In Senegal, common bean (Phaseolus vulgaris) needs to be inoculated with elite Rhizobium strains in the growing area called Niayes zone (Diouf et al., 1999). Usually, seeds of common bean supplied to farmers are often treated with fungicide to prevent losses due to seedborne pathogens.
Since compatibility between rhizobial inoculants and fungicides has not been so far studied in Senegal, our objective was to examine effect of inoculation of fungicide-treated seeds of a common bean variety, Paulista which is one of the most commonly grown varieties in Senegal, in relationship to growth and grain yield, nodulation and nitrogen fixation.

MATERIALS AND METHODS
A field experiment was carried out in Dakar, at Bel-Air experimental station. The soil was a sandy type (94% of sand) containing approximately 100 native Rhizobium/g, counted by infection test method (Brockwell, 1982 ;Vincent, 1970) using common bean seedlings. The pH of the soil was 7.0 with 0.025% nitrogen (Bremner, 1965), and 40 ppm available phosphorus (Olsen et al., 1954). The seeds of common bean variety, Paulista, and that of non nodulating soybean variety m129 used as non-fixing control plant were hand sown in a randomised completed bloc design with four replicates. The size of each plot was 1.35 x 2.55 m, with 15 cm and 45 cm within and between rows respectively. There were four treatments: (i) seeds of common bean treated with Dichlorofenthion-thiram(DCT) fungicide and inoculated with R. etli  (Diouf et al., 2000); (ii) seeds of common bean non treated with DCT and inoculated with R. etli strain ISRA 353; (iii) seeds of common bean treated with DCT and inoculated with R. tropici strain ISRA 554 (Diouf et al., 2000); (iv) seeds of common bean non-treated with DCT and inoculated with R. tropici strain ISRA 554. Rhizobial inoculants were applied as peat slurry containing 10 7 Rhizobium/g. The rhizobial strains ISRA 353 and ISRA 554 were selected at MIRCEN rhizobial culture collection for their high nitrogen fixing potential in association with common bean (Guene, 2002). Within each plot, a 0.60 X 0.45 m microplot was delimitated for the application of 15 N-labelled fertilizer solution, ( 15 NH4)2SO4 containing 5 atom % 15 N excess to supply 20 kg N ha -1 . Unlabelled (NH4)2SO4 was applied at the same rate to the remaining plots. To all plots a basal fertilizer was added and consisted of 60 kg P ha -1 as a triple superphosphate and 120 kg K ha -1 as KCl.
At 60 days after sowing, all plants were harvested from the microplots. The harvested plants were separated into different parts. Nitrogen content (%N) and atom % 15 N excess (% 15 Nae) for individual plant part were determined at the laboratory of soil biochemistry in ISRA-IRD centre of Dakar, Bel Air. Nitrogen fixation (%Ndfa) was estimated using the isotope dilution equation (Fried and Middelboe, 1977

Use of Rhizobium strain ISRA 353
That the seeds were treated or not with the DCT, there were a good vegetative development of the plants and a good production of pods. There was however no significant difference on shoot dry weight and pod yield between the plants from DCT-treated seeds and that from non DCT-treated seeds. Averaged shoot dry weight and pod yield were 1.3 Mg ha -1 and 1.2 Mg ha -1 respectively (Table 1). On the other hand, no nodule was found on the roots of plants derived from DCT-treated seeds. Consequently nitrogen fixation did not occur in these plants, whereas average of 31.5 mg plant -1 of dry nodules recorded on the plants derived from non DCTtreated seeds (Table 1) resulted in 18 and 35% as the proportion of nitrogen derived from fixation (%Ndfa) in the shoot and in the pods, respectively. Corresponding amounts of nitrogen derived from fixation (Ndfa) were 6 kg N ha -1 and 15.4 kg N ha -1 ( Table 2).

Use of Rhizobium strain ISRA 554
The plants were also well developed. No significant difference was observed on the shoot dry weight between plants from treated and non DCT-treated seeds with an average of 2.45 Mg/ha. The pod yield was however decreased (-19.7%) by treating the seeds with DCT (Table 1). Although nodulation of plants was not inhibited by the DCT, nodules dry weight was decreased in the plants from DCT-treated seeds in comparison to that of plants from non DCT-treated seeds, 59.2 mg plant -1 vs 86.2 mg plant -1 (Table 1). No difference was however observed in both the %Ndfa and Ndfa between the two treatments: 38.2% and 18.5 kg N ha -1 respectively in the shoot, 54% and 48.9 kg N ha -1 in the pods (Table 2).

DISCUSSION
Organic fungicides are usually used in agriculture in order to protect seeds to diseases caused by fungi. The main known fungicides are however toxic to rhizobia (Diatloff, Afr. J. Biotechnol.  In each column, for each Rhizobium strain and for each plant organ, values followed by the same letter do not differ significant at p = 0.05 1970; Hofer, 1958). In most cases, the rhizobia remain viable, but are not able to nodulate the host plants or their ability to fix nitrogen is reduced (Fisher, 1976;Staphorst and Strijdom, 1976). DCT is derived from thiram (tetramethyl-thiram-disulphide), one of the less toxic fungicide to rhizobia. We have tested the compatibility of DCT to the inoculation of Paulista variety with Rhizobium strains ISRA 353 and ISRA 554. In the first case there is incompatibility: no nodulation was induced and nitrogen fixation did not occur. In the second case, there is compatibility with less nodulation, and equivalent fixed nitrogen compared to that of plants from non DCT-treated seeds. In addition to the results of Hashem et al. (1997) indicating difference in compatibility with fungicides between peanut (Arachis hypogaea) and Bradyrhizobium inoculants, our results reflect and reactualize the discussion on inoculation of fungicides-treated seeds because the discrepancy between ISRA 353 and ISRA 554 rhizobial strains may be due to the difference of Rhizobium species, that is, R etli and R. tropici. Considerable differences in tolerance among species and strains of rhizobia to different fungicides have been reported by Tesfai and Mallik (1986).
The effect of DCT on the nodulation of common bean Paulista variety inoculated with Rhizobium strain ISRA 353 was similar to that observed with Cicer arietinum (Thomas and Vyas, 1984;Welty et al., 1988), Glycine max (Tesfai and Mallik, 1986) and Pigeon pea (Rennie et al., 1985) treated with different fungicides. Although effect of fungicide application on the viability of Rhizobium strains has been already reported by several authors, the difference in nodulation of the Paulista variety by the Rhizobium strains ISRA 353 and ISRA 554 as influenced by DCT-treated seeds justify the necessity to study the variability of effect of fungicides on the legume-rhizobia symbiosis. Apron fungicide reduces the viability of Bradyrhizobium japonicum (-61%) on soybean seeds after 1 h incubation (Revellin et al. 1993). Similarly, Captan and pentachloronitrobenzene fungicides reduce the viable B. japonicum, -18% and -78% respectively, after 1 h exposure (Curley and Burton, 1975). Graham et al. (1980) have observed on Captan-treated seeds that only 10% of Rhizobium phaseoli had survived after 24 h contact with fungicide whereas 90% of the strains had survived on the non treated seeds after the same time contact. Nevertheless developing fungicide-resistant rhizobial strains remains one approach to overcome this current constraint for delivering inoculants (Tesfai and Mallik, 1986;Hashem et al., 1997).