Phaseolus vulgaris (L) is poorly nodulated in all the regions of Tunis
ia where this crop is grown. Selection of effective strains needs to b
e developed. An effectiveness test trial was carried out on Royalnel a
nd Coco cultivars in hydroponic aseptic conditions for seven reference
strains and three local isolates of common bean rhizobia. A significa
nt interaction strain x cultivar was observed and, unlike local isolat
es, most reference strains were very effective. The ELISA technique us
ing polyclonal antibody was used to evaluate nodule occupancy of inocu
lated common bean rhizobia in Tunisian soils. Rabbit antisera were rai
sed against one local isolate (HB) and five reference strains (CIAT899
, USDA2667, USDA2669, USDA2676 and IRATYH112). Analysis of relatedness
between 44 strains and isolates showed that 77% of the local isolates
did not react with any of the sera used, whereas the remaining isolat
es were distributed into three serogroups. An inoculation experiment w
ith the selected reference strains CIAT899, USDA2667 and IRATYH112, wh
ich are serologically distinguishable from native rhizobia, was perfor
med in a green house for Royalnel and Coco cultivars grown in soil cor
es from two sites (Boucharrai and Chrifet). Nodule occupancy of the in
oculated strains was assessed and concomitant effect of inoculation on
nodule number and shoot nitrogen content was analysed. A highly signi
ficant treatment x cultivar x site interaction was observed for the pa
rameters studied. While in Boucharrai soil inoculation leads to a sign
ificant improvement of shoot nitrogen content for Royalnel cultivar, i
noculation of the same cultivar in Chrifet soil leads to an increase i
n nodulation by native rhizobia without improvement of plant growth. F
or the two cultivars, N-fertilization leads for the two cultivars to a
decrease in nodule number, particularly in Chrifet soil. Thus, to imp
rove establishment of symbiotic process, the use of N-fertilizer shoul
d be rationalized.