Agriculture Reference
In-Depth Information
5.4 DISCUSSION
Both DGGE and PhyloChip analysis revealed that the community com-
position of bacteria at various taxonomic levels differed significantly be-
tween the arable soils and the adjacent scrubland soils. The methods used
to analyze the effects of the transition in land use were complementary but
have clear differences which prevented the use of similar statistical analy-
sis methods to analyze the datasets. In contrast to the DGGE fingerprints
that are assumed to reflect the relative abundance of the dominant bacte-
rial populations, the fluorescent signals detected after PhyloChip hybrid-
izations do not necessarily reflect the relative abundance of OTUs. The
strength of the PhyloChip approach is that the fluorescence signals of each
OTU can be compared horizontally between treatments and thus allows
to identify taxa with a high proportion of OTUs with treatment dependent
changes in abundance. However, both methods used in this study to ana-
lyze PCR-amplified 16S rRNA gene fragments lead to the same conclu-
sions: A strong and significant effect of land use and a higher variability of
the community structure was observed for samples from scrubland (likely
due to the absence of mixing). Furthermore, compared to the effect of land
use, the effect of the site was less pronounced.
Tillage is probably one of the major forces driving shifts in the soil
microbial community structure. In addition, irrigation, fertilization, and/
or application of agrochemicals have been shown to affect the bacterial
community structure [14], [39]-[42]. The minor differences found be-
tween the arable soils from site 1 and site 2 may be linked to differences
in their previous cropping history and agricultural management, which
were most likely not identical. In general, the conversion of scrubland into
agricultural land is associated with the replacement of a diverse, indig-
enous, highly adapted vegetation of the semi-arid desert with a few crop
plants. Plant effects on soil microbial communities have been frequently
observed in the rhizosphere, which refers to the soil directly infl uenced
by root exudates [30], [43]. The infl uence of plant species varied between
different studies, in some of them it was regarded to be the major factor
shaping the microbial community structure [44]-[49], while in others only
a minor infl uence of plant species and vegetation composition on the soil
bacterial community composition was observed [50]. However, only a few
