Agriculture Reference
In-Depth Information
5.3.3 CORRELATION OF LAND USE RESPONSIVE TAXA WITH
LAND USE DEPENDENT SOIL PARAMETERS
Two-way ANOVA of soil physicochemical parameters revealed that the
transition of scrubland into arable land resulted in a significantly increased
concentration of sulphate, sodium, salinity, and the ratio of nitrate to total-
N. The phosphate concentration and organic matter content were signifi-
cantly lower in agricultural soils (Table 1).
Redundancy analysis was performed to fi nd the correlation between
discriminative physicochemical characteristics and variation in the com-
munity composition of phyla identifi ed by PhyloChips with signifi cant
response to land use. Sulphate and phosphate (phosphate positively col-
linear with organic matter content) were the main factors that jointly in-
fl uenced the bacterial community composition (Figure 2). In total 33% of
the variation of the bacterial community could be signifi cantly explained
by sulphate and phosphate concentrations. Each factor (sulphate, phos-
phate and organic matter) independently could explain 17% to 21% of the
variation. Sulphate and phosphate also explained a considerable amount of
the variation within sub-communities of
Proteobacteria
(33%), to which
most detected OTUs were affi liated (data not shown) (Figure S6). Dis-
criminative soil parameters also explained a large amount of the varia-
tion within
Firmicutes
(19% variation explained by sulphate; Figure S7),
Actinobacteria
(40% explained by phosphate and sulphate; Figure S8),
Acidobacteria
(30% explained by nitrate; Figure S9),
Bacteroidetes
(16%
explained by nitrate; Figure S10),
Chlorofl exi
(30% explained by nitrate;
Figure S11),
Verrucomicrobia
(24% explained by phosphate; Figure S12)
and
Gemmatimonadetes
(49% explained by sulphate and sodium; Figure
S13). None of these discriminative soil parameters explained the variation
within
Spirochaetes
, which appeared to be linked to the pH values of the
soils (explaining 32% variation; Figure S14). In general, a considerable
amount of variation in the community composition of phyla responding
to land use changes could be signifi cantly explained by discriminative soil
parameters, suggesting that these soil parameters and the bacterial com-
munity structure are strongly connected.
