The metabolic abilities of microbial communities extracted from maize rhizo
sphere and non-rhizosphere compartments were compared by using the Biolog(R
) system. A double sampling protocol (from bulk soil to root surface and al
ong the root) was used for testing the hypothesis that nature and quantity
of rhizodeposits could be key factors governing microbial community structu
re. Rhizoplane microbes are closer to rhizodeposition sites than microbes l
iving in the adhering soil. This proximity generates distinct pools of rhiz
odeposits between adhering soil and rhizoplane; the carbon pool available f
or adhering soil microbes is qualitatively and quantitatively different fro
m the original one because of rhizoplane microbial activity. The first samp
ling was designed to extract microbes from these two compartments. Moreover
the different parts of a root are known to release distinct pools of rhizo
deposits; the second sampling along the root was designed to extract microb
es from the main morphological parts of a root (ramification zone, root hai
r-elongation zone and apex). Bulk soil without plant (control soil) was als
o investigated to obtain microbes deprived from rhizodeposits. The comparat
ive metabolic profiling revealed clear shifts between the microbial populat
ions of the rhizoplane and the control soil ones. The main discriminating s
ubstrates were carbohydrates, amino acids and amides. Microbial populations
from the ramification zone and the control soil were clearly separated fro
m those of the root hair-elongation zone. This discrimination was mainly ba
sed on some carboxylic acids, carbohydrates, amino acids and amides. The sa
mpling procedure and the community structure differences revealed in this s
tudy confirm a possible structural effect of maize rhizodeposits on telluri
c microbial populations. (C) 2001 Editions scientifiques et medicales Elsev
ier SAS.