Environmental Pollution 111 (2001) 407±415
www.elsevier.com/locate/envpol
Spatial variability in the degradation rate of isoproturon in soil
A. Walker *, M. Jurado-Exposito, G.D. Bending, V.J.R. Smith
Soil and Environment Sciences Department, Horticulture Research International, Wellesbourne, Warwick CV35 9EF, UK
Received 28 September 1999; accepted 17 February 2000
``Capsule'': Soils showing rapid biodegradation of isoproturon had higher pH values and more microbial biomass.
Abstract
Thirty samples of soil were taken at 50-m intersections on a grid pattern over an area of 250200 m within a single ®eld with
nominally uniform soil characteristics. Incubations of isoproturon (3-(4-isopropylphenyl)-1,1-dimethylurea) under standard conditions (15oC; ÿ33 kPa soil water potential) indicated considerable variation in degradation rate of the herbicide, with the time to 50%
loss (DT50) varying from 6.5 to 30 days. The kinetics of degradation also varied between the sub-samples of soil. In many of them,
there was an exponential decline in isoproturon residues; in others, exponential loss was followed by more rapid rates of decline; in a
few soil samples, rapid rates of loss began shortly after the start of the incubations. In more detailed studies with soils from a smaller
number of sub-sites (20), measurements were again made of isoproturon degradation rate, and the soils were analysed for organic
matter content, pH, and nutrient status (N, P, K). Measurements were also made of isoproturon adsorption by the soils and of soil
microbial biomass. Patterns of microbial metabolism were assessed using 95 substrates in Biolog GN plates. Soils showing rapid
biodegradation were generally of higher pH and contained more available potassium than those showing slower degradation rates.
They also had a larger microbial biomass and greater microbial metabolic diversity as determined by substrate utilisation on Biolog
GN plates. The implications of the results for the ecacy and environmental behaviour of isoproturon are discussed. # 2000
Elsevier Science Ltd. All rights reserved.
Keywords: Soils; Herbicides; Isoproturon; Biodegradation; Persistence
1. Introduction
The main process leading to dissipation of pesticide
residues from the environment is degradation in soil,
and this usually involves the activities of soil microorganisms. The rate of biodegradation is in¯uenced by
the chemical properties of the soil such as organic matter
content, pH and nutrient status, and is also in¯uenced by
environmental conditions that control soil temperature
and soil moisture content. Variability in degradation rate
between di€erent soils is expected because of the variability in soil properties, and numerous studies have provided evidence for ®eld-to-®eld variation in the
degradation rates of herbicides (Walker and Brown,
1983; Allen and Walker, 1987; Pussemier et al., 1997)
insecticides (Gerstl, 1984; Parkin and Shelton, 1992) and
fungicides (Walker, 1987a). In several of these examples
* Corresponding author. Tel.: +44-1789-470382; fax: +44-1789470552.
E-mail address: allan.walker@hri.ac.uk (A. Walker).
(Walker, 1987a; Parkin and Shelton, 1992; Pussemier et
al., 1997), the most rapid rates of degradation were
associated with soils to which the study pesticide had
been applied regularly, indicating that the phenomenon
of enhanced biodegradation had occurred (Racke and
Coats, 1990). Studies with soils from 10 di€erent ®elds
at Horticulture Research International, Wellesbourne,
identi®ed one soil (from Deep Slade ®eld) in which the
herbicide isoproturon (3-(4-isopropylphenyl)-1,1-dimethylurea) degraded unusually rapidly, with total loss of
extractable residues within 10 days at 15 C (Cox et al.,
1996). This compares with previously published times to
50% loss (DT50) in moist soil at 15 C of 15 to 40 days
(Mudd et al., 1983; Blair et al., 1990). It was considered
unlikely that the rapid rate of loss could be related to
previous applications of isoproturon since none had been
applied in the 3 years prior to sampling, although occasional applications had been made to cereal crops grown
before that time. Also, similar soils with a more recent
history of isoproturon application did not show such a
rapid rate of degradation (Cox et al., 1996). Other
0269-7491/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0269-7491(00)00092-0
408
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
degradation studies with isoproturon in soil samples
taken from di€erent areas of Deep Slade ®eld indicated
marked variation in dissipation rate with some samples
showing rapid rates of loss, and others providing results
more consistent with the norm (Edwards, 1997). This
suggested that there might be signi®cant spatial variability in the degrading ability of soil, even within ®elds
where the soil appears to be uniform. The present
experiments were made to examine this possibility in a
systematic way by taking samples of soil on a grid pattern from Deep Slade ®eld, and measuring degradation
rates of isoproturon under standard conditions. Some
chemical and microbiological analyses of the di€erent
soil samples were also made.
2. Materials and methods
2.1. Preliminary experiment
In order to provide a preliminary indication of the
variations in degradation rate of isoproturon in soil
throughout the study area, soil samples (500 g) were
removed from 30 separate subsites within the central
part of Deep Slade ®eld at Horticulture Research International. The soil in this area is mapped as a single soil
series (Wick; Whit®eld, 1974), and in terms of visual
appearance and texture classi®cation, it is a uniform
sandy loam throughout the cultivated layer (0±30 cm).
The whole ®eld was cropped with winter barley and had
been sprayed with isoproturon (3-(4-isopropylphenyl)1,1-dimethylurea; Arelon SC, 50% a.i.; Rhone-Poulenc)
at 1.5 kg a.i. haÿ1 on 20 November 1996. Samples of
soil were taken in January 1997, 8 weeks after herbicide
application. They were removed from the 0±5 cm layer
on a grid pattern, with samples 50 m apart covering an
area of 200250 m. There were ®ve samples in a north/
south direction (labelled A±E), and six samples in an
east/west direction (labelled 1±6). All 30 samples were
taken with individual, sterile plastic scoops, and the
sampling sub-sites were marked with ®breglass canes. The
samples were individually processed by hand, using disposable gloves, to give uniformly mixed samples free from
stones and plant residues. Subsamples (30 g) of each soil
were dried in an oven at 110 C overnight to determine soil
moisture content. A suspension of the commercial formulation of isoproturon in distilled water was added to
400 g fresh weight of each to give a concentration of 15
mg kgÿ1 with soil moisture at 9.8% (soil water potential, ÿ33 kPa; Cox et al., 1996). This concentration
approximates to the maximum recommended dose of
isoproturon (2.5 kg haÿ1) when present in a 1-cm depth
of soil. The herbicide was uniformly mixed into each
sample by hand, once more using disposable gloves. The
soils were transferred to sterile polypropylene containers that were loosely capped and incubated at 15 C.
Moisture contents were maintained by additions of sterile distilled water as necessary (usually once each week).
The soils were sampled at regular intervals (3 or 4 days)
during the subsequent 65 days. Amounts of soil (20 g)
were weighed into conical ¯asks (50 ml) and isoproturon
residues were extracted by shaking with methanol (25 ml)
on a wrist-action shaker for 50 min. After shaking, the
samples were allowed to stand until the soil had settled,
and samples of the clear supernatant were analysed for
herbicide residues by HPLC using Kontron Series 300
equipment. The column used was Lichrosorb RP-18
(25 cm4 mm, internal diameter; Merck) and the
solvent system was acetonitrile/water/orthophosphoric acid
(70:30:0.25 by volume) at a ¯ow rate of 1 ml minÿ1.
Detection was by UV absorbance at 240 nm, with a
lower detection limit of 0.10 mg kgÿ1 dry soil. Recovery
of isoproturon in the range from 1.0 to 10.0 mg kgÿ1
varied from 97.1 to 99.0%, and no corrections were
made to the analytical data for recovery.
2.2. Detailed experiment
2.2.1. Herbicide degradation rate
Based on the results from the preliminary study, a
more detailed experiment was made using a smaller
group of soils chosen to represent samples with a range
of degradation rates. Twenty of the original 30 sub-sites
were sampled again in March 1997, 16 weeks after isoproturon application in the ®eld. About 750 g soil was
removed from each sub-site at a position as close as possible to the original. The sampling procedures and the
preparation of samples for the incubation study with isoproturon were exactly as described earlier. In addition a
number of soil analyses were made.
2.2.2. Chemical analysis of soils
Organic matter content was determined by loss on ignition at 450 C, pH was measured with a glass electrode in a
1:1 suspension of soil/distilled water, and extractable
nitrate±nitrogen, phosphorus and potassium were measured using the United Kingdom Ministry of Agriculture
Fisheries and Food standard method numbers 52, 59 and
63, respectively (Anonymous, 1986).
2.2.3. Soil microbial biomass
Soil microbial biomass was measured using a modi®cation of the chloroform fumigation incubation
method of Jenkinson and Powlson (1976) proposed by
Mele and Carter (1996). Duplicate amounts (20 g) of
moist soil from the 20 selected sites were weighed into
100-ml ¯asks and 2 ml ethanol-free chloroform were then
added (1:10 v/w). The ¯asks were closed with silicone
grease-treated glass stoppers, and sealed with para®lm.
The samples, together with similar controls without
chloroform, were incubated for 8 days at 30 C. The
¯asks were then opened and transferred to a vacuum
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
dessicator, which was evacuated several times until there
was no further trace of chloroform in the soil. The
fumigated and control soils were extracted with 50 ml 2
M potassium chloride by shaking for 1 h on a wristaction shaker. The samples were centrifuged, and the
extract analysed for ninhydrin-reactive nitrogen. A
sample of the potassium chloride extract (1 ml) was
mixed with 0.5 ml ninhydrin reagent in a glass tube (20
ml) and heated in a boiling water bath for 25 min. After
cooling, the residue in each tube was made up to 10 ml
with ethanol/water (50:50 v/v). The absorbance of the
resulting purple solution was measured at 570 nm against
reagent blanks. The ninhydrin-reactive N (fumigated
minus controls) was calculated using l-leucine standards, and the value converted to mg microbial carbon
kgÿ1 dry soil as described by Mele and Carter (1996).
2.2.4. Substrate utilisation assay
The functional diversity of the microbial communities
in the 20 soils was assessed using Biolog GN microplates (Biolog Inc. Hayward, CA.) using the technique
described by Garland and Mills (1991). Soils were preincubated with soil moisture equivalent to ÿ33 kPa at
15 C for 14 days, after which 5 g (fresh weight) of each
soil was suspended separately in sterile sodium chloride
solution in water (0.85% w/w, 50 ml). The mixtures were
shaken in sterile Duran bottles (100 ml) on a wrist-action
shaker for 30 min. The suspensions were centrifuged at
1500 rpm for 2 min to clear the supernatant, which was
diluted with further sterile sodium chloride to obtain a
®nal dilution of 10ÿ2.
The diluted suspension was added directly (150 ml per
well) to all of the wells in duplicate Biolog GN microplates.
These plates allow simultaneous testing of the ability of the
microbial suspensions to utilize 95 carbon sources, the
oxidation of which is determined colorimetrically following tetrazolium dye reduction. The inoculated plates were
wrapped in para®lm to reduce evaporation and were
incubated at 20 C. Absorbance of all wells was measured
immediately after inoculation, and after 72 h using an
automatic optical density (OD) plate reader (Anthos
Labtec HT2, version 1.06) at 600 nm.
Absorbance values for the wells were ®rst corrected for
background by subtracting the absorbencies measured
directly after inoculation. Secondly the absorbance of the
control well (which contained no carbon source) was
subtracted from that of every well containing a C-source.
Average well colour development (AWCD) was calculated from each plate at each time according to Garland
and Mills (1991). The utilisation of substrates by the
Biolog-culturable microbial population was analysed as
described by Zak et al. (1994) to give estimates of substrate richness (the number of substrates utilised) and
substrate evenness (the distribution of utilisation among
substrates). These were used to calculate Shannon's
diversity index, which gives a measure of the metabolic
409
diversity of the soil microbial population. Additionally,
relationships between the metabolic pro®les of microbial
communities were investigated by canonical variate analysis using the statistical program Genstat (Version 5.3,
Lawes Agricultural Trust, Rothamsted Experimental
Station, UK).
2.2.5. Isoproturon adsorption measurements
Adsorption of isoproturon by the soils was measured
using a solution of analytical grade isoproturon in 0.01
M calcium chloride at a concentration of 5 mg lÿ1 . The
solution also contained 14C-ring labelled isoproturon
(74 bq mlÿ1 ). Herbicide solution (10 ml) was added to
duplicate amounts (5 g) of each soil in conical ¯asks (50
ml). The samples were shaken on a wrist action shaker
for 4 h, allowed to stand overnight, and then shaken for
a further period of 4 h. After shaking, the contents of
the ¯asks were centrifuged and duplicate subsamples (1
ml) of the supernatant were transferred to scintillation
vials. Liquid scintillant (Optiphase HiSafe; 10 ml) was
added and the radioactivity was measured by liquid
scintillation counting. Samples of the solution without
soil were counted in a similar way. The amount of herbicide adsorbed was calculated from the di€erence
between initial and ®nal measurements of radioactivity,
assuming that this was directly proportional to herbicide concentration and that no degradation of parent
compound had taken place.
3. Results and discussion
3.1. Preliminary experiment
The degradation patterns of isoproturon in soil from
the 30 sites sampled initially in Deep Slade ®eld are shown
in Fig. 1. The data are plotted as residual concentrations,
expressed as a percentage of the amount recovered initially against time of incubation for each site. Degradation
of pesticides in soils is usually interpreted using ®rst-order
reaction kinetics with the data showing an exponential
decrease in concentration over time (Walker, 1987b). In
the present experiments, 12 of the 30 data sets gave an
approximate ®t to ®rst-order kinetics. In a further nine
soil samples, an initial exponential decline in residue
concentration was followed by more rapid rates of loss,
and in the remaining nine soils, rapid degradation began
shortly after the start of the incubations. This suggests
di€erences in the reactions of the soil micro¯ora to the
presence of isoproturon. A progressive, exponential rate
of degradation represents co-metabolic activity, where
the herbicide is degraded as a consequence of metabolism of other organic substrates, but the pesticide is not
used as an energy source (Torstensson, 1980). Rapid
decline after an initial slow rate of loss, suggests that
components of the micro¯ora may have adapted their
410
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
Fig. 1. Degradation of isoproturon in 30 soil samples taken on a 5050 m grid pattern. (~, sites A; *, sites B; &, sites C; *, sites D; ^, sites E).
metabolism in order to utilise the compound as an
energy source, or that an active pesticide-degrading
micro¯ora has proliferated in response to the presence
of the chemical (BergstroÈm and StenstroÈm, 1998).
Rapid degradation soon after application indicates that
a population of micro-organisms with the ability to
metabolize the compound is present in the soil initially
(Torstensson, 1980). Spatial variability in degradation
rate may therefore be associated with spatial variability
in the size or diversity of the soil microbial populations,
or in the soil conditions that control their activity. With
the complex kinetics of degradation observed in several
of the soil samples, it is dicult to de®ne a single parameter with which to characterise the individual decay
curves, but the parameter that is often used, irrespective
of the kinetics, is the time to 50% disappearance (DT50).
The DT50 was calculated for all of the degradation
curves shown in Fig. 1 with the results shown in Table
1. Where ®rst-order kinetics apply, this value is the halflife derived from the lines of best ®t calculated by linear
regression analysis of the logarithm of concentration
against time of incubation. The remaining values of
DT50 were derived by interpolation between appropriate data points. The latter values will be subject to
error because they e€ectively take account of only three
of the measured data points, and they will be strongly
in¯uenced by the value of the initial concentration.
Despite these limitations, the data in Table 1 do provide
an assessment of the relative rates of loss at the di€erent
sub-sites and demonstrate a variation in DT50 from 6.5
to 30 days.
3.2. Detailed study
The results from the second incubation study with a
smaller number of samples are shown in Fig. 2. These
soils were selected to represent the di€erent kinetic
`types' identi®ed in the preliminary study and were
arbitrarily chosen as the 10 samples with the longest
times to 50% loss, and 10 of the 11 samples with the
shortest times to 50% loss. Soil from site A1 was not
used in the second experiment because soil from this
area had been extensively sampled for other microbiological studies. In general, the results con®rm those
from the preliminary study, particularly with respect to
the kinetics of degradation. In all instances where the
degradation data from the preliminary study conformed
to ®rst-order reaction kinetics, these same kinetics were
appropriate to the data from the more detailed experiment (Fig. 2b,d). In all of the other samples, the kinetics
of degradation were clearly not ®rst-order (Fig. 2a,c),
which again is in agreement with the results from the ®rst
experiment. In six of these samples, isoproturon had
degraded completely within 10±12 days. In the other four
soils in this group, rapid degradation began after a period of between 15 and 25 days, with residues below limits of detection within 35 days. Derivation of DT50
values (either as the times to 50% loss or ®rst-order
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
411
Table 1
Estimated DT50 values (days) in the ®rst and second experiments together with the pH measurement of the soila
Sites
A
B
DT50(1)
1
pH
7.6
2
pH
28.2*
3
pH
28.3*
4
pH
26.6*
5
pH
15.2
6
pH
21.4
a
DT50(2)
ND
6.50
6.14
6.47
6.44
ND
ND
DT50(1)
6.4
16.3*
30.2*
19.6*
26.3*
18.0*
25.2*
15.9
24.1
ND
C
20.0
DT50(2)
7.44
6.31
6.26
ND
ND
ND
D
DT50(1)
7.1
8.0
17.6*
18.9
17.4*
28.2*
ND
24.3
ND
25.9*
ND
29.2*
DT50(2)
7.00
6.70
6.36
ND
ND
6.26
E
DT50(1)
7.1
8.0
19.8
27.3*
17.9*
22.5
ND
25.3*
ND
18.8
19.2*
20.8
DT50(2)
7.00
6.28
ND
6.20
6.26
ND
DT50(1)
7.0
7.7
17.0*
25.7*
ND
26.1*
15.9*
8.2
13.4
8.8
ND
8.6
DT50(2)
7.63
ND
6.17
7.40
7.15
7.19
6.5
ND
15.7*
8.3
15.4
5.6
ND, not determined; *, indicates ®rst-order half-lives; all other values are interpolated DT50 values.
Fig. 2. Degradation of isoproturon in soils taken from the sites showing the shortest (a, c) and the longest (b, d) DT50 values in the preliminary
experiment.
half-lives as before), gave the results shown in Table 1.
These times were occasionally di€erent from those
derived from the ®rst experiment, which may represent
either small-scale spatial di€erences or temporal variability in degrading potential of the soil. The changes in
DT50 were most pronounced in the soil samples that
initially showed the slower rates of loss, suggesting the
possibility that soil micro-organisms in the ®eld may
slowly develop the ability to degrade isoproturon over
time, in response to a normal ®eld application of the
herbicide. There is insucient evidence to make conclusions regarding these possibilities. The main features
of the results are the close relationship between the
DT50 values derived in the two experiments (r=0.892;
412
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
data, where the coecient of variation in soil pH was
7.3%, and in soil organic matter content was 8.9%.
The intercorrelation between the various soil parameters, and their relationship with the times to 50% loss
of isoproturon recorded in the second experiment are
shown in Table 3. They indicate a strong negative relationship between the times to 50% loss and soil pH
(r=0.820; P<0.001) and a similar negative correlation
with soil microbial biomass (r=0.839; P<0.001). The
relationships obtained are illustrated in Fig. 3. Microbial biomass and soil pH were positively correlated one
with the other (r=0.865; P<0.001), which illustrates
the diculty of assessing cause and e€ect from simple
correlation analysis. The suggestion of a link between
P<0.001), and the reproducibility of the kinetics of
degradation in the samples taken from the same subsites in the ®eld on the two occasions.
Results of the soil chemical analyses are shown in
Table 2. They show extreme spatial variability in several
of the data sets, particularly those for extractable nutrients, with coecients of variation of 12, 19 and 71% for
phosphorus, potassium and nitrate±nitrogen, respectively. Such large coecients of variation in extractable
nutrient levels are not unusual (Davis et al., 1995; Delcourt et al., 1996), although factors such as pH and
organic matter content usually show somewhat lower
within-®eld coecients of variation (Beckett and Webster 1971; Singh et al., 1993). This was so in the present
Table 2
Soil properties
Site
Nitrate
(mg kgÿ1)
Potassium
(mg kgÿ1)
Phosphorus
(mg kgÿ1)
pH
Organic
matter (%)
Kd
(l kgÿ1)
Biomass
(mg C kgÿ1)
A2
A3
A4
A5
B1
B2
B3
C1
C2
C3
C6
D1
D2
D4
D5
E1
E3
E4
E5
E6
4
8
8
12
13
4
3
12
9
4
7
6
10
5
3
16
3
8
29
7
107
104
111
115
128
117
94
120
98
101
141
125
121
102
129
120
107
180
144
179
58
51
68
68
62
62
60
53
60
75
48
54
56
51
53
59
64
70
66
53
6.50
6.14
6.47
6.44
7.44
6.31
6.26
7.00
6.70
6.36
6.26
7.00
6.28
6.20
6.26
7.63
6.17
7.40
7.15
7.19
3.01
2.37
2.27
2.69
2.51
2.44
2.50
2.51
2.40
2.25
2.63
2.80
2.72
2.30
2.28
2.58
2.22
2.93
2.60
2.70
1.44
1.18
1.22
1.33
1.11
1.18
1.14
0.99
1.08
1.08
1.32
1.05
1.32
1.37
1.10
1.03
1.14
1.38
1.19
1.06
173
159
140
203
257
124
125
254
191
134
166
237
165
154
135
233
172
267
221
288
Table 3
Correlation coecient matrix between the soil physical, chemical and microbiological propertiesa
DT50
pH
NOÿ
3
K
P
OM
Kd
Biomass
Richness
Evenness
Diversity
Canonical variate 1
a
pH
NOÿ
3
K
P
OM
Kd
Biomass
Richness
Eveness
Diversity
Canonical
variate 1
Canonical
variate 2
ÿ0.820
ÿ0.215
0.569
ÿ0.579
0.578
0.268
0.102
0.156
0.168
ÿ0.032
ÿ0.375
0.440
0.188
0.507
ÿ0.075
0.407
ÿ0.331
ÿ0.102
0.064
0.056
0.408
ÿ0.839
0.865
0.467
0.678
ÿ0.042
0.528
ÿ0.235
0.529
ÿ0.471
ÿ0.025
ÿ0.803
0.098
ÿ0.401
0.012
ÿ0.519
0.119
ÿ0.178
0.049
ÿ0.057
ÿ0.177
ÿ0.267
ÿ0.257
ÿ0.136
ÿ0.118
0.504
ÿ0.500
0.021
ÿ0.679
ÿ0.058
ÿ0.504
ÿ0.189
ÿ0.506
0.704
0.621
ÿ0.019
0.056
0.121
0.000
ÿ0.044
ÿ0.119
ÿ0.238
ÿ0.052
ÿ0.181
0.756
0.407
0.309
ÿ0.489
ÿ0.372
ÿ0.273
ÿ0.191
ÿ0.339
ÿ0.045
ÿ0.346
0.165
0.077
0.191
0.000
Signi®cance of correlation coecients; 0.444 (P<0.05); 0.561 (P<0.01); 0.679 (P<0.001).
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
413
Fig. 3. The relationships between DT50 in the second experiment and (a) soil pH and (b) microbial biomass.
soil pH and isoproturon degradation by soil microorganisms, however, is consistent with the data of Cox et
al. (1996). They demonstrated in a range of soils from
di€erent ®elds that there was more rapid degradation of
isoproturon at higher pH within the range 5.0±7.5.
Results from these earlier experiments also demonstrated
that repeated applications of isoproturon to the soils in
laboratory incubations were more likely to result in
enhanced biodegradation if soil pH was >7.0. The spatial link between soil pH and isoproturon degradation is
further illustrated in Table 1 in which the data for soil
pH are listed with the appropriate DT50 values. There
are two main areas of higher soil pH (from site A1
through to site E1, and from sites E4±E6). The areas of
lower soil pH were generally associated with the higher
DT50 values.
The data in Table 3 also show a signi®cant negative
correlation between soil potassium level and the times
to 50% loss (r=0.579; P<0.01). This appeared to be
related to just two soils that showed rapid biodegradation of isoproturon and which contained more available
potassium than the other 18 samples. There were no signi®cant correlation if these two soils were excluded from
the calculations.
The results from the measurements of substrate utilisation patterns on Biolog GN plates are summarised in
Table 4. There were signi®cant di€erences between soils
in total substrate utilisation (richness), in the evenness of
substrate utilisation pro®les, and in the overall metabolic
diversity of the soil microbial population. Richness of
metabolism was signi®cantly correlated with available
potassium, microbial biomass, pH and DT50, while metabolic diversity was signi®cantly correlated with potassium,
organic matter content, microbial biomass, pH and DT50
(Table 3).
Relationships between the metabolic pro®les of the
microbial communities, as determined by canonical
variate analysis, are shown in Fig. 4. There was no clear
Table 4
Richness, evenness and diversity of microbial community metabolism
of Biolog GN microplate substrates at soil sites
Site
Richness
Evenness
Shannon's
diversity index
A2
A3
A4
A5
B1
B2
B3
C1
C2
C3
C6
D1
D2
D4
D5
E1
E3
E4
E5
E6
93.5
86.0
92.5
90.0
86.5
91.5
89.0
90.5
94.0
91.5
86.0
89.5
89.5
94.0
85.5
86.5
94.0
76.5
91.5
80.5
0.919
0.945
0.923
0.940
0.938
0.944
0.945
0.950
0.946
0.937
0.950
0.932
0.955
0.930
0.953
0.945
0.952
0.941
0.939
0.928
4.17
4.21
4.18
4.23
4.18
4.26
4.24
4.28
4.30
4.23
4.23
4.19
4.29
4.22
4.24
4.21
4.32
4.08
4.24
4.07
distinction between those soils showing unusual kinetics
of degradation (`fast degradation'; Fig. 2a,c), and those
in which herbicide degradation followed ®rst-order
kinetics (`slow degradation'; Fig. 2b,d). However, with
the exception of sample E6, there appeared to be some
degree of clustering among `fast' and `slow' degradation
sites, with most separation along canonical variate axis
2. This axis was shown to be correlated with soil pH
(Table 3), indicating that pH signi®cantly in¯uenced
microbial community structure.
A limitation in the use of the Biolog system to evaluate soil microbial communities is that the populations of
bacteria which are culturable on the Biolog plates have
been shown to be predominantly members of the g-class
414
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
are less likely to contaminate subsoils and groundwater
resources than are more stable compounds (Gustafson,
1989). A full understanding and quanti®cation of the
extent of spatial variability in degradation parameters
within speci®c aquifer recharge zones is therefore of considerable signi®cance in de®ning the potential for groundwater contamination.
Acknowledgements
Fig. 4. Canonical variate analysis of microbial community Biolog
substrate utilisation pro®les at soil sites showing rapid (&) and slow
(*) biodegradation of isoproturon.
proteobacteria, which includes species of Enterobacter,
Pantoea, Salmonella and Pseudomonas. These are fast
growing species adapted to high substrate concentrations
(Smalla et al., 1998). Biolog substrate utilisation patterns,
therefore, provide a measure of the metabolic diversity of
this readily culturable portion of the microbiota rather
than that of the entire community.
4. Conclusions
Overall, these results demonstrate signi®cant di€erences in both the kinetics and rates of isoproturon
degradation on a spatial scale within a single ®eld.
Although the reasons for these di€erences have not been
identi®ed de®nitively, the data strongly suggest that they
are related to the size and activity of the isoproturondegrading component of the total soil microbial population, and these appear to be in¯uenced by soil pH. The
data also strongly suggest that the overall structure of the
microbial population at di€erent microsites was in¯uenced by soil pH, and clearly, further experiments are
required to investigate in more detail the inter-relationships between microbial characteristics, soil pH, and
degradation kinetics in the di€erent soils.
The rate at which a pesticide degrades has an important in¯uence on other aspects of its behaviour. Persistence of a soil-acting herbicide, such as isoproturon, will
control the duration of the period of e€ective weed
control, and spatial variability in degradation rate may,
therefore, contribute to spatial variability in performance in the ®eld. Degradation rate in combination
with sorption is also a signi®cant factor controlling the
potential of a compound to leach in the soil, since for
similar sorptive properties, short persistence compounds
This work was funded in the UK by the Ministry of
Agriculture Fisheries and Food (project PL0526), and
by the Biotechnology and Biological Sciences Research
Council. M.J.-E. thanks the Spanish Ministry of Education and Culture for her FPI fellowship. We are greatly
appreciative of the advice of Mrs. Kathleen Phelps and
Dr. Julie Jones concerning the statistical analysis of the
data.
References
Allen, R., Walker, A., 1987. The in¯uence of soil properties on the
rates of degradation of metamitron, metazachlor and metribuzin.
Pesticicide Science 18, 95±111.
Anonymous, 1986. The Analysis of Agricultural Materials, 2nd Edition. MAFF/ADAS Reference Book No 427, HMSO, London.
Beckett, P.H.T., Webster, R., 1971. Soil variability: a review. Soils and
Fertilisers 34, 1±15.
BergstroÈm, L., StenstroÈm, J., 1998. Environmental fate of chemicals in
soil. Ambio 27, 16±23.
Blair, A.M., Martin, T.D., Walker, A., Welch, S.J., 1990. Measurements and prediction of isoproturon movement and persistence in
three soils. Crop Protection 9, 289±294.
Cox, L., Walker, A., Welch, S.J., 1996. Evidence for the enhanced
biodegradation of isoproturon in soils. Pesticide Science 48, 253±
260.
Davis, J.G., Hossner, L.R., Wilding, L.P., Manu, A., 1995. Variability
of soil chemical properties in two sandy, dunal soils of Niger. Soil
Science 159, 321±330.
Delcourt, H., Darius, P.L., De-Baerdemaeker, J., 1996. The spatial
variability of some aspects of topsoil fertility in two Belgian ®elds.
Computers and Electronics in Agriculture 14, 179±196.
Edwards, J.A., 1997. Aspects of the Adsorption and Degradation of
Isoproturon in Soils. BSc dissertation, University of Reading, UK.
Garland, J.L., Mills, A.L., 1991. Classi®cation and characterization of
heterotrophic microbial communities on the basis of patterns of
community-level sole-carbon-source utilization. Applied and Environmental Microbiology 33, 34±44.
Gerstl, Z., 1984. Adsorption, decomposition and movement of oxamyl
in soil. Pesticicide Science 15, 9±17.
Gustafson, D.I., 1989. Groundwater Ubiquity Score; a simple method
of assessing pesticide leachability. Environmental Contamination
and Toxicology 8, 339±357.
Jenkinson, D.S., Powlson, D.S., 1976. The e€ects of biocidal treatments on metabolism in soil. V. A method for measuring soil biomass. Soil Biology and Biochemistry 8, 209±213.
Mele, P.M., Carter, M.R., 1996. Estimation of microbial biomass by
ninhydrin-reactive N using liquid chloroform. Canadian Journal of
Soil Science 76, 37±40.
Mudd, P.J., Hance, R.J., Wright, S.J.L., 1983. The persistence and
metabolism of isoproturon in soil. Weed Research 23, 239±246.
A. Walker et al. / Environmental Pollution 111 (2001) 407±415
Parkin, T.B., Shelton, D.R., 1992. Spatial variability of carbofuran
degradation in soil. Journal of Environmental Quality 21, 672±678.
Pussemier, L., Goux, S., Vanderheyden, V., Debongnie, P., Tresinie,
I., Foucart, G., 1997. Rapid dissipation of atrazine in soils taken
from various maize ®elds. Weed.Research 37, 171±179.
Racke, K.D., Coats, J.R., 1990. Enhanced Biodegradation of Pesticides in the Environment. American Chemical Society Symposium
Series 426, Washington.
Singh, R., Singh, B., Singh, Y., 1993. Spatial variability of some
properties of typic ustochrept soil in the semiarid region of northwestern India. Arid Soil Research and Rehabilitation 7, 197±201.
Smalla, K., Wachtendorf, U., Heuer, H., Liu, W., Forney, L., 1998.
Analysis of BIOLOG GN substrate utilization patterns by microbial
communities. Applied and Environmental Microbiology 64, 1220±
1225.
415
Torstensson, N.T.L., 1980. Role of microorganisms in decomposition.
In: Hance, R.J. (Ed.), Interactions Between Herbicides and the Soil.
EWRS, Academic Press, London.
Walker, A., 1987a. Further observations on the enhanced degradation
of iprodione and vinclozolin in soil. Pesticicide Science 21, 219±231.
Walker, A., 1987b. Herbicide persistence in soil. Reviews of Weed
Science 3, 1±17.
Walker, A., Brown, P.A., 1983. Spatial variability in herbicide degradation rates and residues in soil. Crop Protection 2, 17±25.
Whit®eld, W.A.D., 1974. The Soils of the National Vegetable
Research Station, Wellesbourne. Report of the National Vegetable
Research Station for 1973, pp. 21±30.
Zak, J.C., Willig, M.R., Moorhead, D.L., Wildman, H.G., 1994.
Functional diversity of microbial communities: a quantitative
approach. Soil Biology and Biochemistry 26, 1101±1108.