Sampling and analysis of soil

Joseph H. Massey

Mississippi State University, Starkville, MS, USA

Scott H. Jackson and Manasi Saha

BASF Corporation, Research Triangle Park, NC, USA

Eberhard Zietz

Institut Fresenius, Taunusstein, Germany

Introduction

Field studies designed to investigate agrochemical behavior in soil are conducted for
a variety of reasons. For example, these studies may be performed to understand
soil, climatic, and use-pattern effects on the biological efficacy of soil-applied
agrochemicals. This information is used to establish application rates as a function
of soil texture and organic matter content, or to establish safe rotation intervals for
succeeding crops. However, these studies are most often conducted to determine the
fate of an agrochemical in the soil environment.

Soil represents a significant environmental ‘sink’ for agrochemicals when they are

applied, for example, as pre-emergence herbicide applications, fungicide coatings
on seeds, and ‘drench’ insecticide treatments around the foundations of homes and
buildings. Incidental agrochemical depositions onto soil result from the over-spray
and wash-off of foliar pesticide applications and, to a minor extent, deposition of
atmospheric residues. As a result of direct applications and indirect depositions on
to soil, information is needed on the persistence and mobility of agrochemicals and
their degradates in soil.

Field dissipation studies conducted on bare soil are the primary means by which

soil dissipation rates of agrochemicals, and rates of formation and decline of their
associated degradates, are determined under actual conditions. These studies are used
to validate results of laboratory studies (e.g., metabolism and rate of degradation
in soil and sediment, aqueous and soil photolysis, hydrolysis, sorption, and column
leaching) that are used in formulating conceptual environmental dissipation models
for agrochemicals. In addition, the propensity of an agrochemical to leach below the
root zone can be assessed when these studies are properly conducted. More defini-
tive studies, such as small-scale prospective groundwater studies and/or lysimeter

Handbook of Residue Analytical Methods for Agrochemicals.

2003 John Wiley & Sons Ltd.

Sampling and analysis of soil

841

studies, are necessary to establish fully the leaching behavior of a potentially mobile
compound. Ultimately, the dissipation and mobility parameters derived from these
field soil dissipation studies are used by regulatory and environmental scientists to
predict environmental concentrations and fates of agrochemicals in different soil and
climatic environments using computer simulation models. The purpose of this article
is to present field and laboratory best practices that have proven useful in the success-
ful conduct of field soil dissipation studies. Although focusing specifically on field
soil dissipation studies, the soil sampling, sample processing, analytical, and curve-
fitting techniques described herein pertain to other more general situations requiring
the careful sampling and analysis of soils for agrochemicals.

The four main phases involved in a field soil dissipation study are (I) planning and

design phase, (II) field-conduct phase, (III) sample processing/analysis phase, and
(IV) data handling/reporting phase. Each phase is vitally linked to the next and each
is critical to study success. Results from an otherwise perfectly executed study may
be made useless by uneven test substance application or improper sampling, sample
handling, and/or analytical techniques. Each of these phases is discussed below.

Phase I: field study design and logistics

Before the first soil sample is collected, several important, inter-related factors must be
considered to ensure study success. These factors include the specific physicochem-
ical properties of the agrochemical, its use pattern, the anticipated environmental
conditions during use, and several practical considerations related to the analysis of
soil (Figure 1). They determine the type(s) of soil to be investigated, agrochemical
application method and rate, depth and frequency of sampling, and total amount of
soil to be collected. Additional information, such as how the cores will be shipped,
stored, sectioned and processed, and whether supplemental irrigation, if any, should
be applied during the study must also be determined as part of the study design phase.
These are all important details that should not be left to chance or addressed as an
afterthought once the study is under way. Because each phase of study conduct is sig-
nificantly impacted by the chemical and physical properties of the agrochemical, we
begin with a review of key physicochemical properties affecting field study design.

2.1

Physicochemical properties

2.1.1

Anticipated persistence of parent molecule

Among the first issues to address in study design is how often soil samples should be
collected and how long the study should be conducted. The frequency and duration of
soil sample collection depend on the anticipated soil persistence of the agrochemical
based on results from laboratory and/or other field studies. Because dissipation data
are subjected to regression analysis and mathematical modeling in order to calculate
a dissipation half-life (T

) or DT

value [unlike the T

value that is based on

first-order kinetics, DT values make no assumptions as to the appropriateness of first-
order kinetics in fitting dissipation data; see Phase IV (Section 5) for more details],
the sampling regime must result in decline data that bracket the anticipated T

842

Best practices in the generation and analyses of residues in environmental samples

Use Pattern

• Formulation type

(liquid vs. granular)

• Application type (PRE, PPI, POST,

banded, chemigation)

• Application rate

• Application timing

(spring vs. fall)

• Application frequency

(single vs. multiple applications)

• Target crop(s) and cultural system

Physico-Chemical

Properties

• Anticipated soil persistence and

mobility of parent and degradates

• Water solubility

• Soil sorption (K

, K

)

• Speciation (pKa)

• Vapor pressure

Environmental

• Texture, organic matter content, oxygen

status and pH values of soil in use area(s)

• Amounts and timing of precipitation

• Temperature and solar irradiance in use area(s)

(temperate vs. tropical vs. nordic)

Analytical

•

Anticipated soil concentration

• LOD for parent and degradate(s)

• Sample size requirements

• Stability of analyte(s)

• Sample transportation logistics

• Freezer storage & sample processing

capabilities

• Analytical cost per sample

Figure 1

Inter-related factors affecting the design of terrestrial field soil dissipation studies

value. The sampling points should occur at regular, evenly spaced intervals with

four to five sampling points prior to the T

or DT

value. If the data are heavily

skewed towards the ends of a regression line, with few sampling points in between,
erroneous estimates of the dissipation rate will occur.

For example, Figure 2 shows how the number and spacing of sampling points can

affect T

determination. In Figure 2(a), a total of 15 sampling periods were fitted

using a first-order model, resulting in an estimated T

of 10 days and a coefficient of

determination (R

) of 0.90. Here, six evenly spaced sampling periods occurred before

the half-life and four immediately after the half-life. In Figure 2(b), two sampling
periods prior to the half-life and five intermediate sampling periods were omitted
from the calculations, resulting in a T

of 28 days. This value is almost three times

longer than that estimated using the complete data set, and demonstrates how sampling
frequency and spacing affect dissipation rate estimation.

When an environmental fate profile of an agrochemical and its degradates is not

established, it is prudent to collect soil cores at frequent intervals, recognizing that it
may not be necessary to analyze all of the contingency samples. A standard practice in
field protocol design is to build 20% more sampling periods into the sampling regime
than is thought necessary to adequately characterize the dissipation profiles. If only
laboratory-derived soil persistence data are available, a general guide is to assume a
3–5 times greater degradation rate in the field than was observed in the laboratory.
This empirical factor reflects that dissipation rates are often greater in the field as
compared with laboratory data.

Sampling and analysis of soil

843

100

150

200

250

300

350

-6

-5

-4

-3

-2

-1

ln of Concentration (mg kg

−

)

Days After Last Application

P1 = -0.21898
Alpha = 5.09774
Beta = 0.00515
DT

= 28.3

= 60.7

=0.95

100

150

200

250

300

350

-6

-5

-4

-3

-2

-1

Bare Soil

Days After Last Application

ln of Concentration (mg kg

−

)

P1 = -0.21898
Alpha = 1.43747

Beta = 0.06216

= 10.0

= 26.1

=0.90

(a)

(b)

Figure 2

Influence of sampling frequency on first-order model parameters

Faster degradation in the field results from increased opportunities for volatilization,

leaching, photolysis, and ‘aging’. Moreover, microbial activities and biomass in soil
may be significantly affected by techniques used to store, dry and fortify soils used in
laboratory studies.

–

Thus, if a compound has a laboratory T

of 100 days, assume

a T

of 20–30 days when designing the field sampling protocol. If a wide range

of persistence is exhibited under controlled conditions, it is prudent to design the
protocol to accommodate a range in dissipation rates (e.g., frequent early sampling
coupled with prolonged sampling).

Vapor pressure (VP), water solubility (S

), and soil sorption coefficients

) are key properties that govern volatilization of agrochemicals from soil.

Volatile compounds such as S-ethyl dipropylthiocarbamate (EPTC) (VP

≈ 4.5 Pa,

844

Best practices in the generation and analyses of residues in environmental samples

= 375 mg L

−1

, K

≈ 223 mL g

−1

) and vernolate (VP

≈ 1.4 Pa, S

= 108 mg

−1

, K

≈ 260 mL g

−1

) are readily lost from moist soil. In fact, their volatility

makes it imperative to soil-incorporate these herbicides immediately after applica-
tion to reduce volatilization. Frequent initial sampling is required when investigat-
ing volatile compounds such as EPTC and vernolate. This is especially true for the
first 24 h when volatilization losses are often the greatest from moist soils. Inter-
mediate volatilization losses have been observed for compounds such as trifluralin
(VP

≈ 0.015 Pa, S

= 0.3 mg L

−1

, K

≈ 7200 mL g

−1

) and fonofos (VP

≈ 0.045 Pa,

= 17 mg L

−1

, K

≈ 1920 mL g

−1

Volatility is not a major consideration in

sampling protocols designed for compounds such as atrazine (VP

≈ 9 × 10

−4

Pa,

= 33 mg L

−1

, K

≈ 147 mL g

−1

) and prometon (VP

≈ 0.011 Pa, S

= 720 mg

−1

, K

≈ 95 mL g

−1

Designing a sampling protocol to describe adequately the formation and decline

of degradates requires prior knowledge of a compound’s dissipation behavior. Degra-
dates included in analytical protocols for field dissipation studies must first be identi-
fied in laboratory hydrolysis, soil/aqueous photolysis and soil/sediment metabolism
studies. They usually must comprise

≥10% of the applied dose in at least one labo-

ratory study to warrant inclusion in a field dissipation study. However, degradates of
known toxicological concern must be included, even when present at

<10% of the

applied dose. In some cases, minor transformation products are included to better de-
scribe the degradation pathway or to avoid ‘parent only’ dissipation profiles. Because
of the increased analytical sensitivity they afford, radiolabeled test materials are some-
times used to track the formation and decline of degradates under field conditions.

The appearance of degradates in field soil can be sporadic and transient in nature,

making it difficult to obtain useful kinetic information under field situations. As a
result, sampling protocol design becomes more complicated when the parent molecule
and degradation product(s) exhibit widely different soil persistence. For example,
although chlorsulfuron has a T

of 18 days, samples were collected for 540 days

after last application so that the dissipation profiles of three primary degradates could
be discerned.

Even with frequent and prolonged sampling, additional laboratory

studies may be required to fully establish the dissipation kinetics of degradates whose
dissipation profiles are not adequately captured under field conditions.

2.1.2

Anticipated mobility in soil profile

The anticipated mobility of an agrochemical affects protocol design by determin-
ing maximum depth of sample collection. Properties governing soil mobility in-
clude S

, acid dissociation constant (pK

), soil sorption coefficients (K

, K

)

and VP.

The soil sorption of ionizable compounds is affected by pH and organic

matter and clay content.

Weak organic acids and bases are less retained as soil

pH increases due to increasing (weak acids) or decreasing (weak bases) ionization.
Depending on their persistence, compounds with S

> 30 mg L

−1

, K

< 5 and/or

< 300 mL g

−1

are generally more likely to move below the root zone than

sparingly soluble, strongly bound compounds such as trifluralin (S

< 1 mg L

−1

> 5000 mL g

−1

, logK

> 3).

Thus, deeper soil coring is required for poten-

tially mobile compounds compared with those which are less mobile. Ultimately, this

Sampling and analysis of soil

845

information is used to assess the maximum depth of leaching of the agrochemical
under field conditions.

When the potential mobility of an agrochemical and/or its degradates is not known,

it is prudent to collect sub-surface cores to a depth of 90–120 cm, recognizing that
only certain sections of the cores may ultimately be processed and analyzed. For
studies conducted for regulatory purposes, cores must be collected such that at least
one core section is found to be free of quantifiable residues. As a result, full-length
cores (90–120 cm) are almost always collected for studies conducted for ultimate
submission to regulatory agencies, regardless of the anticipated mobility of the agro-
chemical. Additional information on the physicochemical properties of agrochemicals
can be found elsewhere.

,11,12

2.2

Use-pattern considerations

A guiding principle in the design of a field soil dissipation study is that study conduct
should closely follow agricultural practices associated with the particular use pattern
of the agrochemical being investigated. This requires knowledge of regional agricul-
tural production practices for the targeted crop and an appreciation of how physical
formulation, method of application, and soil and climatic factors affect agrochemical
dissipation in soil. As discussed below, the ideal of closely following realistic agri-
cultural practices has to be carefully balanced with practical considerations and the
ultimate purpose of the study.

2.2.1

Bare-soil versus cropped studies

Determining whether to study the dissipation of an agrochemical in the presence or
absence of a crop represents an important consideration in field study design. Bare-
soil studies are useful in establishing soil dissipation kinetics for agrochemicals at
their labeled rates but will omit potential rhizosphere, leaf photolysis, canopy shading,
crop uptake and transpiration effects on dissipation and mobility. Cropped studies are
useful in determining maximum plateau concentrations under actual-use conditions
but cannot always be relied upon to yield results with the resolution necessary for
accurate determination of soil dissipation kinetics. Moreover, the presence of an over-
hanging canopy, rocks, roots, and/or hardpans common to certain cropping situations
(e.g., orchards) may impede deep soil sampling necessary for maximum depth of
leaching determinations.

An increasingly important agronomic use pattern is reduced or no-till conservation

practices where crop residues are allowed to accumulate on soil surfaces to reduce soil
erosion and improve soil tilth. The effects that accumulated crop residues and con-
comitant changes in soil properties associated with conservation tillage have had on
herbicide dissipation, for example, have been mixed,

but represent another impor-

tant use-pattern consideration for certain agrochemicals. Because soil characteristics
such as organic carbon content, pH, and microbial biomass require several years to be
affected by conservation tillage practices, field sites must be carefully selected when
the effect of crop residues on agrochemical dissipation is to be determined. Clearly,
tillage would not be appropriate for no-till investigations.

846

Best practices in the generation and analyses of residues in environmental samples

Wheat Canopy

3:1 A:B Ratio

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

19-May

(App. 1)

04-June
(App. 2)

19-Jul

20-July

(App. 3)

30-Jul

19-Aug

Days After Last Application

Average Soil Concentration

(mg kg

–

)

100

120

140

160

Cumulative Precipitation (mm)

Compound A

Compound B

Precipitation

--Soil Not Sampled--

Standard error bar shown about mean of three replications representing a composite of 10 soil samples/rep./sample period.

Apple Canopy

1:1 A:B Ratio

0.000

0.050

0.100

0.150

0.200

0.250

0.300

15-July

(App. 1)

29-July

(App. 2)

12-Aug
(App. 3)

26-Aug

(App. 4)

9-Sep

(App. 5)

22-Sep

23-Sep
(App. 6)

2-Oct

7-Oct

Days After Last Application

Average Soil Concentration

(mg kg

–

)

100

120

140

Cumulative Precipitation (mm)

Compound A

Compound B

Precipitation

-------------Soil Not Sampled-------------

Standard error bar shown about mean of three replications representing a composite of 10 soil samples/rep./sampling period.

(b)

(a)

Figure 3

Soil deposition of two agrochemicals after application to (a) wheat and (b) apple canopies

Figure 3 illustrates the difficulties that can arise when investigating agrochemi-

cal dissipation in soil following application to foliage. The figure shows the initial
soil deposition of two agrochemicals that were applied as a mixture to field-grown
wheat or apple trees. Three sequential applications of a formulation with a 3 : 1 ratio of

Sampling and analysis of soil

847

Compound A to Compound B were made to a mature wheat canopy while six sequen-
tial applications of a formulation with a 1 : 1 ratio of Compound A to Compound B
were applied to mature apple trees. These studies closely mimicked actual agrochemi-
cal use patterns in terms of application rate and timing, application volume, irrigation,
and other key agricultural practices.

When applied to wheat, Compound A was readily detected in soil samples collected

beneath the treated canopy. Compound B was detected only sporadically in soil. Due
in part to a greater application volume (1000 vs 300 L ha

−1

water) and a higher

application rate for Compound B, both compounds were more consistently detected
in soil following application to apple foliage. It is often difficult to establish dissipation
kinetics under these conditions because as residues in the soil dissipate, additional
compound may continue to be deposited on the soil, resulting in a complex, variable
dissipation pattern. As a result, it is not always practical or advisable to study soil
dissipation in the presence of a crop.

A compromise to the ‘bare-soil vs cropped studies’ dilemma is to establish a bare-

soil study in close proximity to the target crop. Using the above scenarios, a bare-soil
study could be established by removing the aboveground portion of the crop wheat
or the vegetation existing between the rows of trees in the apple orchard. The bare
surface allows direct application of the agrochemical to soil, eliminating the effects
of delayed and variable agrochemical deposition on soil commonly associated with
foliar applications while exposing the agrochemical to edaphic and hydrogeologi-
cal conditions that approximate those of soils underlying agronomic or horticultural
crops.

For compounds applied to annual crops, another approach is to apply the compound

to bare soil prior to crop emergence and follow the soil dissipation of the compound
as the crop emerges and grows throughout its normal growing season.

–

This is

the appropriate use pattern for pre-emergence compounds and represents another
approach that may used to study the soil dissipation of foliar-applied compounds.

In terms of leaching potential, mobility is often considered ‘worst case’ under

bare-soil conditions because the absence of crops eliminates transpiration losses, a
major avenue of water loss from cropped soils.

Elimination of transpiration losses

results in less upward movement of agrochemicals in the soil profile that, in turn,
increases the potential for agrochemicals to leach over time. Leaching may also be
greater under reduced-tillage conservation systems, largely due to increased numbers
of intact soil macropores as compared with tilled soils (Ref. 13 and references cited
therein). Moreover, use patterns where the compound is injected via sub-surface drip
irrigation increase opportunities for the compound to leach and require additional
design and sampling considerations.

Ultimately, the primary purpose of the study

and the use pattern of the agrochemical dictate whether to use a bare-soil or cropped
study design. Many of the technical considerations and techniques presented in this
article pertain to both study types.

2.2.2

Region of use

Soil, climatic, and hydrogeological conditions vary widely between geographical
regions. This variation occurs on a variety of scales and significantly affects agro-
chemical dissipation rates and fates in the environment. As a result, differences in

848

Best practices in the generation and analyses of residues in environmental samples

soil properties and the range of climates where a compound is used must be carefully
considered. These factors also determine the number of study locations required to
establish the full range of agrochemical dissipation behavior in field soil. A geo-
graphic information system (GIS) has been developed to identify comparable field
study areas in the USA and Canada.

Soil properties affecting dissipation include texture (sand, silt, and clay contents),

clay type, organic carbon content and type, iron and aluminum oxide contents,
porosity, overall fertility, moisture and oxygen status, temperature, pH, salinity, and
microbial community structure and activity. Climatic factors for consideration in-
clude seasonal timing and amounts of precipitation, solar irradiance, and maximum/
minimum temperatures that affect crop production. Field sites that accentuate leaching
(low soil sorption capacity, high rainfall, high pH for weak organic acids) and/or per-
sistence (high soil sorption capacity, low soil moisture, high or low soil temperature,
high or low pH) help to define ‘worst case’ behavior for agrochemicals.

2.2.3

Supplemental irrigation

Crops may be grown under rainfall-fed or irrigated production systems. For a study
designed to mimic an upland use pattern, the ability to irrigate test plots is required
in the event that weather conditions turn drier than normal. Typically, 110% of the
long-term monthly mean precipitation is applied to ensure that leaching opportunities
for the agrochemical exist under study conditions. Water inputs for irrigated crops
are determined using a soil-water budget method. The soil-water budget method is
described in more detail in Section 3.3.9. In addition, irrigation is sometimes required
to facilitate soil sampling in dense, hard-packed soils.

Whatever the case, the ability to irrigate test plots is an important consideration

during field site selection. Sprinkler irrigation is preferred. Flood and furrow irrigation
should be avoided since they may disturb surface residues, resulting in uneven residue
distribution and/or inadvertent agrochemical loss from the study plots. Recommended
irrigation practices are discussed in more detail in Section 3.3.8.

2.2.4

Soil incorporation of agrochemical residues

Certain soil-applied agrochemicals are incorporated into the soil after application to
facilitate better contact with target organisms and/or to reduce losses due to photolysis
and volatilization. Soil incorporation depths typically range from 2 to 10 cm. Incor-
poration is accomplished using a power rotary tiller, rolling cultivator, rotary hoe,
disk harrow, or similar implement.

,21,22

To maintain realistic study conditions, agro-

chemicals that are typically incorporated during use should be incorporated during
field soil dissipation investigations.

Soil incorporation must be done with care to avoid the introduction of significant

variability in agrochemical residues in soil. Studies have found that as much as a 50-
fold variation in agrochemical residues may arise when incorporation is incomplete
or uneven.

,23

A single incorporation pass is not sufficient to mix agrochemicals

thoroughly and, in practice, two or more passes are often necessary.

,24

Provisions

for soil incorporation must be made prior to study initiation and should follow good
agricultural practices.

Sampling and analysis of soil

849

2.2.5

Application timing

The time of year in which a pesticide is applied significantly affects its dissipation
rate due to temperature, moisture, and solar-irradiance effects on abiotic and biotic
dissipation processes. For example, dissipation rates for agrochemical applications
made in the springtime are normally greater than those observed for fall (autumn)
applications.

,15,21

Thus, the timing of agrochemical applications made in field soil

dissipation studies should closely match those occurring under actual-use conditions.

2.2.6

Method of application

Agrochemicals are applied using a number of techniques. The method of application
depends upon formulation type and the particular setting in which the chemicals
are used. Granular formulations may be applied aerially or applied broadcast or
banded using ground equipment.

Liquid solutions may be applied aerially or applied

broadcast using ground equipment or by air-blast or chemigation.

,18

Banded and

chemigation applications require specialized application equipment and additional
study design considerations to ensure that soil residues are sampled in a representative
manner. Use patterns that involve agrochemical applications under plastic mulch affect
dissipation and require access to specialized equipment used in plot establishment
and test-substance application.

The most common application method used in field

soil dissipation is broadcast application using a hand-held or vehicle-mounted spray
boom. Proper broadcast application techniques are discussed in Section 3.

Another application-related factor affecting study design is the quantity of test

material that is available for study. When applied as a commercial formulation,
test-substance availability is generally not an issue, and relatively large, replicated
areas can be treated with commercial application equipment. In contrast, radio-
labeled test materials or agrochemicals being investigated early in the discovery
process are applied in small amounts using small-plot techniques such as described by
van Wesenbeeck et al.

and Zabik et al.

These small-quantity materials are usually

prepared in formulation blanks to approximate the physicochemical properties of
commercial products.

2.2.7

Application rate and frequency

Application rate is generally dictated by the labeled, or anticipated, application rate
relevant to the particular use pattern being investigated. To improve analytical detec-
tion or to compensate for potentially low zero-time application recoveries, application
rates are sometimes increased to 110% of the labeled application rate. An application
rate greater than this level would be subject to regulatory scrutiny and may affect
the dissipation rates of certain agrochemicals owing to potential short-term effects on
sensitive soil microflora.

For low-use rate compounds applied on a grams per hectare basis, it has sometimes

been necessary to apply the cumulative seasonal rate in a single application in order to
improve analytical detection. Advances in analytical chemistry have greatly improved
the trace-level detection of agrochemicals in soil but it is still prudent to verify that
sufficient analytical sensitivity exists to detect agrochemicals at their anticipated soil

850

Best practices in the generation and analyses of residues in environmental samples

concentrations and, thus, allow dissipation rate determinations over time. Calculation
of anticipated soil concentration is discussed in Section 2.3.1.

The proper frequency of applications made during a field study is a source of con-

tention among environmental scientists. For example, some believe that the number
should be determined strictly by the agrochemical’s use pattern. For example, if the
compound is typically applied four times over a growing season at 2-week inter-
vals at a rate of 0.1 kg a.i. ha

−1

per application (a.i.

= active ingredient), it should be

applied in this manner for a soil dissipation study. However, other scientists have
found that dissipation data resulting from sequential agrochemical applications can
be difficult to interpret, especially for degradates.

They would argue that the above

compound should be applied in a single application of 0.40 kg a.i. ha

−1

. For studies

conducted for regulatory purposes, it is recommended that the application rate and fre-
quency represent the ‘worst-case’ scenario in terms of agrochemical persistence and
mobility.

A single application made at the maximum-labeled rate is often viewed

as worst case in this regard. This latter approach also reduces costs associated with
multiple applications, travel to field site, and application verification.

2.2.8

Application volume

The volume of spray solution in which an agrochemical is applied to a given area is
relevant to nongranular formulations. Broadcast application volumes typically range
from 200 to 600 L ha

−1

. In an attempt to improve zero-time recoveries of agrochem-

icals, some practitioners diverge from actual-use conditions in terms of application
volume and/or the number of passes made over the test plots during test substance
application. This practice is based on the assumption that improved coverage of
the soil surface results in more uniform agrochemical residues in soil samples. These
practices deviate from standard agricultural practices but may be necessary to ensure
a uniform application.

2.3

Analytical considerations

2.3.1

Anticipated soil concentration and analytical sensitivity

As mentioned previously, one must ensure that sufficient analytical sensitivity exists
to analyze the agrochemical at its anticipated soil concentration. In order to make
this determination, one calculates the nominal zero-time soil concentration and com-
pares this value to the limit of quantitation (LOQ) determined for the soil analysis
method. Because the dissipation of agrochemicals in soil typically follows a biphasic
or ‘hockey stick-shaped’ pattern (biphasic or hockey stick-shaped dissipation curves
are characterized by initial rapid dissipation rates followed by substantially slower
decline rates, resulting in soil residues that persist at low levels for a period of time

regulatory agencies often require that the time required for an agrochemical to dissi-
pate to 25% or 10% of the initial soil concentration (e.g., DT

or DT

value) be de-

termined in addition to the DT

value. Hence analytical methods are often developed

to quantify residues equivalent to

≤5% of the initial applied mass so that DT

values

may be readily obtained. For relatively high use-rate compounds applied at kilograms
per hectare rates, this generally does not pose a problem. However, for low use-rate

Sampling and analysis of soil

851

compounds applied at gram per hectare rates, the analytical sensitivity necessary for
these low levels cannot be assumed and must be verified prior to study initiation.

For example, the expected zero-time soil concentration (C

) of a compound ap-

plied at a rate of 2.2 kg a.i. ha

−1

would be calculated by dividing the application

rate (mg a.i. ha

−1

) by the total weight of a 15-cm depth of soil. Assuming a soil

bulk density of 1500 kg m

−3

, the total weight of a 15-cm layer of soil is 2

.24 ×

kg ha

−1

.2 × 10

mg a

.i. ha

−1

)

/(2.24 × 10

kg soil)

≈ 1.0 mg a.i. kg

−1

(1)

Similarly, the expected C

of a 0.168 kg a.i. ha

−1

application rate would be:

.68 ×10

mg a

.i. ha

−1

)

/(2.24 × 10

kg soil)

≈ 0.08 mg a.i. kg

−1

(2)

The LOQ value necessary to follow residue decline to 5% of the initial value, as
is typically needed for DT

determination, would be 0.05 mg kg

−1

for the higher

application rate and 0.004 mg a.i. kg

−1

for the lower rate. As a result, an LOQ of

0.01 mg a.i. kg

−1

would be sufficient for the 2.2 kg a.i. ha

−1

application rate but not

for a rate of 0.168 kg a.i. ha

−1

There are several approaches that might be taken to address the issue of insufficient

analytical sensitivity indicated by the above calculations. For example, the LOQ might
be lowered by increasing the total amount of soil extracted, reducing the volume
of the final extract solution, improving method cleanup procedures to reduce the
effects of interferences, and/or switching to a more sensitive method of detection.
A brief overview of analytical techniques used for soils, with an emphasis on liquid
chromatography/mass spectrometry (LC/MS) techniques, is given in Section 4.

Another approach to improving agrochemical detection is to apply more of the

active ingredient to increase the initial soil concentration. As mentioned previously,
however, one must be careful not to exceed greatly the labeled application rate of
the compound as questions may arise as to concentration effects on the observed
dissipation. A more common and acceptable approach is to section the upper soil
core into smaller depth increments, yielding increased residue concentrations as the
total amount of soil mixed with the residues decreases in each processed sample
(Table 1).

Table 1

Anticipated zero-time concentrations (mg kg

−1

) as a function of soil core length

Core section length (cm)

0–15

0–10

0–7.5

0–5

0–2.5

Total soil weight (kg)

2.24

× 10

1.49

× 10

1.12

× 10

7.47

× 10

3.73

× 10

Concentration estimate

0.08

0.11

0.15

0.23

0.45

(mg kg

−1

)

Total soil weight per given depth per hectare; assumes a bulk density of 1500 kg soil m

−3

Calculations are based on a nominal application rate of 0.168 kg a.i. ha

−1

. Soil core sectioning

techniques are discussed in Section 3.

852

Best practices in the generation and analyses of residues in environmental samples

For example, if one must estimate a DT

value given an application rate of

0.168 kg a.i. ha

−1

and an LOQ of 0.01 mg a.i. kg

−1

, one could further section a 0–

15-cm upper core into 5-cm lengths, resulting in an increased ability to detect to
0.011 mg kg

−1

as required by the LOQ:

× 0.05 = 0.23 mg a.i. kg

−1

× 0.05 = 0.011 mg kg

−1

≈ LOQ

(3)

Regardless of how the upper core is ultimately sectioned, the 15–120-cm depth cores
are typically sectioned in 10–15-cm lengths for analysis. Techniques used to section
soil cores are presented in Section 3.3.6.

2.3.2

Agrochemical residue variability and sample number requirements

Variability exists in every aspect of study conduct and must be carefully controlled for
meaningful field soil dissipation results. Variability under the investigator’s control
includes that associated with soil surface preparation, agrochemical application, soil
incorporation (if any), sample collection, sample processing, and sample analysis.

Variations in the biological, chemical and physical processes affecting agrochemical
dissipation in soil can be large within a field, and are responsible for the observed
increases in variability with time.

–

The blocking techniques designed to statistically minimize effects of soil hetero-

geneity require prior knowledge of soil texture, fertility, and/or other gradients that
occur across the test site.

Such information is not typically known when field soil

dissipation studies are being established, and the positioning of study plots is usually
based instead on matching plot dimensions with those of the test site. As a result,
the impact of soil heterogeneity may only be partially minimized through careful
visual assessments of soil conditions made during site selection (see Section 3.1) and
collection of ample numbers of soil cores. The greater the variability in soil residue
levels, the greater is the number of samples required to estimate the dissipation rate
of an agrochemical.

The need for additional samples to compensate for soil heterogeneity must be rec-

onciled with labor, storage, transportation, analytical, and other constraints that add
significantly to study costs. Satisfactory results have been obtained from numerous
field studies using three or four treated replications with 5–10 soil cores collected
from each replication per sampling period.

,7,15,20

These replication/repetition num-

bers strike a reasonable balance between the need for samples sufficient in number
to characterize agrochemical dissipation versus financial and logistical constraints
associated with sample collection and analysis.

2.3.3

Sample homogenization

In practice, the number of soil samples that is actually analyzed is reduced by the
preparation of composite samples. Here, multiple samples from a given replication
and sampling period are blended together to yield one composite sample for analysis.
Composite samples are statistically justifiable as they increase the precision with

Sampling and analysis of soil

853

which the mean residue concentration in soil can be estimated while decreasing the
total number of samples analyzed.

,37

Careful attention must be paid to the homogenization of soil samples because in-

complete or careless blending may result in significant variability among agrochemi-
cal residues and may place in jeopardy an otherwise well-executed study. Clayey soils
are generally more difficult to homogenize than sandy or loamy soils and, therefore,
often require additional processing time. Thorough homogenization also becomes
increasingly more difficult with increasing soil moisture. Soil homogenization tech-
niques are discussed in more detail in Section 4.1.

2.4

Basic experimental designs for field soil dissipation studies

At this stage in planning, the essential study design information listed below should
be determined and a written study plan (i.e., protocol) including these key study
details prepared. A formal, pre-approved study plan is required for field soil dissipa-
tion studies conducted under Good Laboratory Practice (GLP). A written study plan
for non-GLP studies is highly recommended since the document serves as valuable
guidance for study personnel.

Study design consideration

Basis for design

Number and locations of test sites and

required soil properties

Region(s) of test substance use and

use pattern

Cropped vs bare-soil surface

Use pattern and study purpose

Small vs large plot

Test material availability

Application type

Relevant use pattern and physical

formulation

Application rate, frequency, and timing;

need for soil incorporation

Relevant use pattern

Sampling frequency, duration, and depth

Anticipated persistence and mobility

of agrochemical and degradates

Number of replications and repetitions

Anticipated variability in soil

residues and cost constraints

Amount of soil to collect and core

sectioning

Depends upon specific analytical

procedures (and associated LOQ)
and available sample storage and
processing capabilities

Supplemental irrigation (sprinkler)

Necessary for most dryland and

irrigated cropping scenarios

Additional information regarding applicator-boom width, spray-tank capacity, and

the wheelbase of any vehicle-mounted soil sampling equipment used during the study
is also required to ensure that the field plot design accommodates size restrictions of
field equipment.

854

Best practices in the generation and analyses of residues in environmental samples

2.4.1

Control plot

Untreated (control) soil is collected to determine the presence of substances that may
interfere with the measurement of target analytes. Control soil is also necessary for
analytical recovery determinations made using laboratory-fortified samples. Thus,
basic field study design divides the test area into one or more treated plots and an
untreated control plot. Unlike the treated plots, the untreated control is typically not
replicated but must be sufficiently large to provide soil for characterization, analytical
method validation, and quality control. To prevent spray drift on to the control area
and other potential forms of contamination, the control area is positioned

≥15 m away

and upwind of the treated plot, relative to prevailing wind patterns

2.4.2

Treated plots

Factors used to determine treated plot size include the (1) available quantity of test
substance, (2) total number of samples to be collected, (3) specific space requirements
of soil sampling equipment, (4) foot and equipment traffic within and between plots
necessary for plot establishment/maintenance and sample collection, and (5) necessity
for minimizing preferential flow to subsoil through sample boreholes. The treated
test area must be large enough to provide the required number of soil core samples
and ensure that human activities do not affect or influence the dissipation of the
agrochemical. The study design should always allow for extra ‘contingency’ samples
beyond the anticipated level of sampling deemed sufficient at the time of study design.
Once the total number of samples to be collected has been determined, the availability
of necessary study supplies (e.g., plastic probe liners, caps, labels, bags, etc.) and
freezer storage capacity should be determined.

2.4.3

Small-plot designs

A wooden or metal containment box surrounding the treated area is commonly used
when a small quantity of test material is to be applied.

The box, typically rectangular

in shape and partially buried beneath the soil surface, serves to isolate the treated area
from surrounding soil and protect against wind and water erosion. A one- to two-
nozzle application boom that moves along guy wires or tracks is often used to ensure
even application. Radiolabeled materials having two or more label positions often
serve as replicates in these studies.

2.4.4

Large-plot designs

The need to collect soil samples repetitively with time while minimizing soil sur-
face disturbance associated with foot and equipment traffic precludes the completely
random collection of soil samples from study plots. One of the most common field
designs, used successfully in numerous field dissipation studies, divides the treated
test area into three or four blocks (i.e., replicates). The blocks are further subdivided
into subplots as shown in Figure 4. The number of subplots is dictated by the number
of sampling dates plus a 20% contingency since soil samples are taken only once
from each subplot using this design.

Sampling and analysis of soil

855

> 3.5 m

40.5

10.5 m

Treated plots

Buffer

zone

Untreated plots

> 71 m

3 m

2 - 2.5 m

1.5 m

0.5m buffer

Sampling area

Sprayed area

Figure 4

Randomized block design using four replications having 20 sub-plots each

856

Best practices in the generation and analyses of residues in environmental samples

The width of an individual treated replicate should not be wider than 3 m to enable

test substance application using a single pass of a conventional plot sprayer. The
application is made in the same direction as the layout of the plot. If multiple boom
widths are used in treating a single plot, it is critical that areas of potential over-spray
and under-spray are avoided during soil sampling. Study designs requiring multiple
application passes within a single treated area are not recommended owing to potential
issues arising from areas of over- or under-spray.

Five soil cores are typically collected from a predetermined subplot within each

replication at each sampling period. As mentioned previously, the number of soil cores
collected increases with increasing residue variability. The order of subplot sampling
is determined using a randomization procedure

or by random-number subroutines

common to many computer spreadsheet programs. The areas between the treated
replicates serve as buffer zones and provide access lanes for study personnel and
vehicles. Within each row, the subplots are separated by a buffer zone of 0.5 m.
An important advantage of the completely randomized block design is that sample
collection is distributed across the entire test plot, helping to capture effects of soil
spatial variability on agrochemical dissipation. The design presented in Figure 4 is
readily adapted to bare-soil and cropped studies.

Additional planning and sample numbers are often required when agrochemicals

are applied as banded rather than broadcast applications. The soil sampling techniques
devised for banded fertilizer applications provide a good basis for the sampling of
agrochemical residues.

,40

For example, the recommended approach for sampling

fields receiving banded nitrogen fertilizer applications involves the collection of 15–
30 composite cores taken between the banded rows and inter-rows of the field.

Sampling at multiple positions perpendicular to the application band provides a mea-
sure of agrochemical distribution throughout the surface soil. Similarly, determining
representative soil sampling locations for agrochemicals applied by chemigation is not
a trivial undertaking and requires increased sample numbers to account for increased
residue variability.

2.4.5

Plot markers

A field soil dissipation study usually lasts between 1 and 2 years; long-term soil
accumulation studies may last for up to 6 years. Hence, it is essential that test plots
are clearly marked to ensure accurate sampling for the duration of the study. Durable,
highly visible markers (stakes) made of plastic, metal, or wood should be located
at the main corners of the treated and control plots. Additional markers indicating
replication and subplot number or line number, as appropriate, must also be installed.
Weather-proof signs must be installed that clearly indicate the Study Director and
contact information, study number, test substance and application rate, and study
initiation and termination dates. This information helps to prevent application and
sampling errors. Plot markers and signs should be checked regularly to ensure that
they are legible and in good physical condition.

Permanent markers outside the study area should also be located and used in the

event that one or more plot markers are inadvertently moved or lost. One option is
to locate a minimum of two permanent reference points outside of the study area

Sampling and analysis of soil

857

50
to
80
cm

Stake marking
individual
sub-plot

Post marking
individual
replicate

Permanent
marker

50
to
80
cm

Sub-soil marker

Figure 5

Techniques used to mark test plots in field soil dissipation trials

that can be used to re-survey the test area by triangulation (Figure 5). The distances
to prominent points such as the ends of sampling plots should be recorded in the
study records and indicated on corresponding plot maps. Another option is the use
of sub-soil markers that are detected by induction (Figure 5). Because these markers
are placed 60–80 cm directly below prominent points in the study area, it is un-
likely that they will be moved during the study. The sub-soil markers are especially
useful in long-term accumulation studies that involve seasonal plowing or cultiva-
tion and when permanent landmarks are not conveniently located near the study
area.

858

Best practices in the generation and analyses of residues in environmental samples

2.5

Additional considerations

2.5.1

Study documentation

Overall study success depends upon the careful documentation of key aspects of
study conduct. As mentioned previously, a formal, written study plan (protocol) is
required for GLP studies and is highly recommended for non-GLP studies. Other key
information to document in the study records includes example calculations involving
the application rate, anticipated zero-time concentration, and those associated with the
analysis of soil. Additional documentation should include the source, purity, test site
location(s), soil textural class, diagrams of test site layout, type and inner diameters of
soil corers, sampling depths, pertinent weather parameters, amount and timing of all
supplemental irrigation, and the names of all personnel involved with study conduct.
The date and time of each application, sample collection, freezer storage, sample
extraction, and analysis should all be carefully recorded. Any events that result in
deviations from the written protocol must be carefully recorded in the study records
and, in the case of GLP studies, the Study Director notified of these events within 24 h
of their occurrence. Photographs taken during test substance application and sampling
and of the equipment related to these activities are useful in reconstructing key aspects
of study conduct. Thorough documentation is as vital for non-GLP research as it is
for studies conducted for regulatory purposes.

2.5.2

Safety

Equipment used to apply agrochemicals and to collect and process soil is inherently
dangerous. The appropriate personal protective equipment must be worn and mini-
mally includes protective eyewear and gloves. Additional protective equipment may
include spray suits, respirators, steel-toed boots, and hearing protection, depending
on the particular materials being investigated and equipment being used. Large phys-
ical force is required to insert a soil probe into the ground; this same force can crush
or amputate human limbs. Hence, workers must be well trained in the operation of
sampling equipment. Fieldwork also requires physical exertion so caution should be
observed when working in high temperature and humidity conditions. Studies involv-
ing the application of radiolabeled materials require prior written permission from the
appropriate regulatory authorities as well as special provisions for the proper removal
and disposal of treated soils and sub-soils.

Phase II: field study conduct

Each of the five main steps in field conduct (site selection, test plot layout, test
substance application, sample collection, and sample storage/handling) is addressed
below.

3.1

Test site selection

Once the targeted study regions, soil textures, space requirements, and other key
aspects of study design have been determined, the search for suitable test sites

Sampling and analysis of soil

859

begins. Test site selection is critical to the success of a field soil dissipation study
as field-related factors have a major influence on the overall outcome of the study.
Even for bare-soil studies, an ‘agriculturally viable’ soil that would be capable
of growing a healthy crop is usually desired. Hence it is important to ascertain
the soil’s recent cropping and management history before choosing a particular
site.

Table 2 lists basic criteria that can be used during field site selection for bare-

soil and cropped studies. Priority among the selection criteria depends upon the
particular goals of the study but certain factors (e.g., slope

>1%, excessive rocks,

flood prone, potential plot disturbance by wildlife) usually serve to exclude cer-
tain sites automatically. If the region of interest is far away, it is best to seek the
assistance of university investigators, extension agents, and consultants who are
familiar with the regional agricultural practices and local soil and climatic con-
ditions.

Table 2

Site-selection criteria for field soil dissipation studies

Selection
criterion

Priority

Basis for selection

Comments

Region

A or B

Site must match the climatic, soil, and

agricultural conditions typical of the
target crop

Some crops are grown only in certain regions

(e.g., rice) while others are common to
many regions (e.g., maize). Thus, selection
of a test region may be restrictive or
relatively flexible

Soil properties

Soil texture (sand, silt, clay), organic

matter/carbon content, and pH

Stones, roots, and hardpans must be

largely absent to allow representative
sampling of soil profile

Soil properties should appear uniform

over test site

Soil texture data should be available at time

of site selection. Soil properties must
match study purpose. This can be ‘realistic
use’ conditions, ‘realistic worst-case’ or
‘worst-case’ in terms of agrochemical
mobility and persistence

Must ensure that the majority of samples can

be taken from the deepest sampling
horizon. Information about sub-soils can be
obtained from soil maps, test coring and
on-site interviews

Site topography

Exclusion

Must have slope

≤1%

Site must not be susceptible to flooding
Shallow water table or tile drains must not

interfere with sampling

These are exclusion criteria that have to be

carefully determined during on-site
inspection

Site must be level to prevent losses of

agrochemical due to surface run-off and
soil erosion

Site must not be susceptible to runoff from

other areas higher than test site

Size of test site

Depends on study design. The minimum

area required for a typical large-plot
design is about 0.25 ha

Test site must allow for test design plus

sufficient buffer zone around perimeter of
field to protect against external disturbance

For bare-soil studies, shady sites should be

avoided

(Continued overleaf )

860

Best practices in the generation and analyses of residues in environmental samples

Table 2 —Continued

Selection
criterion

Priority

Basis for selection

Comments

Cropping history
and previous
pesticide use

Exclusion

The cropping and pesticide history for the

previous 3 years must be well
documented

The test substance must not have been

applied to site within the past 3 years

This information is crucial and evidence of

careful record keeping reflects favorably
upon the future reliability of a field
cooperator

Prior application of agrochemical forming

identical/similar degradation products as
test substance should be considered as
potential analytical interferences

Previous management practices (e.g., soil

amendments, tillage, crop type) should
have been uniformly applied across test site

Irrigation

Exclusion

Site must be equipped with sprinkler

irrigation

Irrigation is necessary to ensure 110% of

historical rainfall for dryland settings or to
follow regional irrigation practices in
irrigated cropping settings

Test site security

Access of unauthorized persons, livestock,

etc., must be restricted

Potential impact of any nearby construction,

utility lines, rights-of-way, etc., must also
be assessed

Plot maintenance

Expertise must be available to maintain

the test site and, if cropped, to take care
of the crop

For bare-soil studies, the soil surface must be

carefully prepared prior to test substance
application and kept weed-free without
disturbing the test areas. If the test is
cropped, the crop should be treated
according to Good Agricultural Practice. In
case of a soil accumulation study, the field
may be cultivated and cropped each season
for up to 6 years

Ownership

Access to test site must be guaranteed for

the duration of study

Owner must agree to grant access to the site

for duration of study plus possible time
extensions. As a result, sub-leasing of the
test site is not preferred. This criterion is
extremely important for long-term studies
such as field soil accumulation studies

Weather station/
weather data
requirements

On-site weather station is preferred and

may be mandatory for certain studies.
Minimally, a station must be located
within 10 km of test site

In certain cases, a weather station located

within 10 km of the test site may be
sufficient. If water balances are to be
determined, an on-site weather station is
necessary to measure, at a minimum,
precipitation, solar radiation, wind speed,
relative humidity, and air temperature

‘Exclusion’ implies that criteria must be fulfilled without compromise since the study may be jeopardized if the criteria are not

met; ‘Priority A’ implies some flexibility after careful consideration; ‘Priority B’ factors offer the greatest flexibility in terms of
site selection.

3.1.1

Collection of control soil

Once test sites have been identified, control soil should be collected and returned to
the laboratory. This soil is used to (1) verify soil texture and related properties, (2)
ensure adequate analytical recovery of target analytes, and (3) determine the presence
of potential background interferences in the soil.

Sampling and analysis of soil

861

3.1.2

Soil surface preparation

Preparation of the soil surface is critical to achieving acceptable results with minimal
variability. Surface roughness due to the presence of crop debris or soil clods makes
representative sampling nearly impossible. This same material also interferes with
sample homogenization. As a result, the importance of proper soil surface prepara-
tion for bare-soil studies cannot be overstated. If vegetation exists on the selected
site, it must be removed for bare-soil study designs. Vegetation can be removed by
application of a nonselective herbicide such as glyphosate, paraquat, or glufosinate
followed by mowing, raking, and harrowing once the vegetation has died.

A combination of techniques is normally required to smooth the soil properly. For

example, disking is usually followed by multiple passes of a rolling-cage cultivator. If
necessary, individual subplots can be hand-raked. Sandy soils are the easiest to prepare
and dry quickly after rainfall. Silt loam to clay loam soils form clods when worked
too wet. Hence timing field preparation around rainfall and soil moisture content is
always a factor in preparing test plots. Heavy clay soils containing

>40% clay pose

real challenges in terms of surface preparation owing to excessive clod formation
and surface cracking and should be avoided. When clayey soils are investigated,
increased numbers of soil samples should be collected to compensate for the additional
variability typically associated with these soils.

In addition to being smooth, it is preferable that the soil surface be firmly packed.

This is because loose soil is not always retained in large-diameter sampling probes.
Firming of the soil surface may be accomplished using a turf roller or equivalent.
Alternatively, the soil surface may be prepared in advance of study initiation to allow
rainfall or irrigation to settle and firm the soil. This latter approach also allows soil
surface depressions to be observed and avoided when laying out the test plots.

3.2

Test substance application

Accurate and even application of test substance is absolutely critical to study success.
If the application is highly variable or deviates significantly from the target application
rate, the study results may be technically unusable and/or unacceptable to regulatory
authorities. Accurate agrochemical application begins with careful calibration of the
spray equipment. Hence Study Directors should be familiar with sprayer calibration
techniques,

,42

even if they will not be personally making the applications.

Braverman et al.

found that factors responsible for inaccurate pesticide applica-

tions made for crop residue trials (i.e., application rates applied at

>10% or <5%

of the target rate) were improper boom height (60% of errors), miscalculation of
application rate (26% of errors), and variations in pass time (14% of errors). Appli-
cation rate calculations must be carefully performed and double-checked, preferably
by a second individual

. Calculations involving products containing more than one

test substance can be particularly confusing and the application rate for each active
ingredient must be clearly stated in the field protocol. Similarly, one must clearly
distinguish application rates based on active ingredient versus acid equivalents for
agrochemicals prepared in various salt formulations. For a given salt formulation, an
application rate based on acid equivalents will always be more than that based on
active ingredient. For more details on calculating application rates, see Anderson.

862

Best practices in the generation and analyses of residues in environmental samples

Spray nozzle type plays an important role in the success of agrochemical applica-

tion. For broadcast applications to soil, flat fan nozzles should be used. Newer spray
tips such as the DG TeeJet, XR TeeJet, Turbo TeeJet and similar nozzles supplied
by Lechler and Hardy have provided acceptable results in a number of studies. For a
given nozzle type, the lower the application pressure, the larger is the spray droplet
size and the less potential for spray drift. Similarly, the closer the boom is positioned
to the soil surface, the less is the potential for spray drift.

Most applications are

made with spray tips having 80

◦

or 110

◦

spray angles and boom heights of about

50 cm above the soil surface.

Wind speed is another important factor affecting applications. Because modern

analytical techniques used in soil analysis are capable of detecting slight differences
in residue concentrations, experienced applicators are cautious with regard to wind
effects on pesticide drift. A hand-held anemometer should be used to measure wind
speed at spray-boom height prior to and during test substance application. Applica-
tions should not occur when wind speeds exceed 3 m s

−1

In regions where exces-

sively windy conditions are the norm, it may be necessary to build wind blocks to
protect the test plots during application. Wooden frames covered in plastic or fiber-
glass sheeting have been successfully used for this purpose.

3.2.1

Application verification

A combination of techniques is typically used to verify the accuracy and precision
of agrochemical applications to soil. For example, the catch-back method or pass-
time method is typically used in conjunction with analytical results from application
verification monitors to confirm proper application. The catch-back method involves
measuring the spray solution volume before and after application to double check that
the desired volume of test solution was actually applied to the test plots. Experienced
applicators are often able to apply within

±2% of the targeted spray volume.

The pass-time method involves measuring the time that it takes the applicator to

pass over a test plot of known length and comparing this time to the speed used in cal-
culation. For example, a typical walking speed for an applicator carrying a hand-held
boom is about 1.3 m s

−1

. At this speed, it would take about 30 s to apply an agro-

chemical along the 40.5-m length of one replication (i.e., block) depicted in Figure 4.
An actual pass time of 31 s would suggest that about 103% of the target application
rate was applied to the test plot. As with spray volume, experienced applicators are
often able to apply within

±2% of the targeted pass time. Field protocols typically

require that the application be within

±5% of the target spray volume or pass-time

value; pass-times or spray volumes greatly exceeding these criteria should be closely
scrutinized and may warrant termination of the study.

Application verification (AV) monitors are devices that are placed within test plots

to measure actual spray deposition that occurred during application. The main function
of AV monitors is to show whether or not the intended amount of test material was
actually deposited on the soil surface. Application monitors consisting of soil-filled
containers, paper disks, polyurethane foam plugs, and glass Petri dishes have all been
used successfully for this purpose. Prior to using a monitor in the field, it is important
to determine that the test substance can indeed be successfully extracted from the
monitor and that the compound will be stable on the monitor under field conditions.

Sampling and analysis of soil

863

Application monitors are positioned in pre-determined locations shortly before test

substance application. Immediately after application, the monitors are collected and
stored in a freezer until they are extracted and analyzed. Soil samples must not be
collected from locations previously covered by the monitors. Hence the monitors
should be placed only in the unsampled buffer zones between the subplots. Alter-
natively, their positions within the test plots can be clearly marked using plastic
flags or stakes and these locations not used for soil collection. The main advan-
tages of using AV monitors rather than zero-time soil cores to verify application
rates are that delays between application and zero-time sampling are greatly reduced
(important for labile materials), and errors often associated with soil sampling are
avoided.

3.3

Soil sampling techniques

Over the years, many soil collection techniques have been developed and tested to
determine their suitability for field dissipation studies. The biggest challenge con-
fronting researchers is to collect representative samples from various depths of soil.
Of particular concern is how to collect samples from the lower soil profile when a
highly concentrated agrochemical residue layer exists at the soil surface immediately
after application. Depending on the environmental fate properties of the compound,
this challenge may exist for some time after application. The following provides an
overview of sampling techniques that have proven useful in addressing these concerns
in field soil dissipation studies.

The artificial downward movement of agrochemical residues caused by soil sam-

pling is commonly referred to as drag down. Several coring techniques using similar
overall approaches have been devised to prevent drag down and cross-contamination
of soil samples. In the method depicted in Figure 6, a probe (e.g., 5

.7 × 15 cm) with an

associated outer sleeve is inserted into the soil. Once the probe and soil core have been
removed, the outer retaining sleeve is left in the ground to ensure that the resulting
borehole does not collapse, thereby preventing the contamination of lower soil depths
by surface residues. Next, a smaller diameter probe (e.g., 3

.8 × 120 cm) is inserted

through the hole kept open by the outer retainer sleeve and forced down to collect soil
lower in the profile. (Note that the lower probe length typically exceeds the length of
the desired core length to offset less-than-full cores that occur commonly under field
conditions, i.e., a 120-cm probe is used to ensure that 100-cm cores are collected.)
Both probes are designed for use with a plastic liner. During sampling, it is the plastic
liner that actually receives the soil so that soil does not touch the steel tubes except
at the cutting tip.

As the plastic liners are removed from the probe, they are capped on both ends,

the appropriate labels affixed, and promptly placed in a freezer (an in-field sectioning
technique used for further partitioning of the 0–15-cm core is described later in this
section). By convention, red plastic caps are placed on top of the core (i.e., the end
that was closest to the soil surface) and black caps are placed on the bottom. Use
of the two-color capping system is important when the cores are sectioned at a later
time. This approach is referred to as zero-contamination sampling and is the industry
standard in field soil dissipation.

864

Best practices in the generation and analyses of residues in environmental samples

Red Cap:
Nearest to
soil surface

Black Cap:
Farthest from
soil surface

Step 1. Soil probe with outer retainer sleeve (A),

inner probe (B), and probe liner (C) is
inserted into soil profile.

Outer
Retainer Sleeve (A)

Inner Probe (B)

Plastic
Liner (C)

Steps 2 and 3.

The inner probe (with liner) is
removed, leaving the outer
retainer sleeve in the soil profile.
While the liner is still in the inner
probe, a red cap is carefully
placed on the top of the liner.
Next, the probe is inverted, the
liner removed from the inner
probe, and the bottom of the liner
carefully fitted with a black cap.

Upper
soil profile
sample
in liner
with end
caps

Outer retainer

sleeve (A)

remains

in soil to

prevent

‘drag down’

of surface

residues

to lower

soil profile

Soil Surface

Narrow Soil
Probe (D)

Plastic
Liner (E)

Red Cap:
Nearest to
soil surface

Black Cap:
Farthest from
soil surface

Step 4. Inner soil probe

(D) with
plastic liner
(E) is
inserted
through
outer
retainer
sleeve (A)
and into the
lower soil
profile.

Steps 5 and 6.

The narrow probe (with liner) is
removed, again leaving the outer
retainer sleeve in the soil profile.
As before, a red cap is first
liner. Next, the probe is inverted,
the liner removed from the inner
probe, and the bottom of the liner
carefully fitted with a black cap.

Outer retainer

sleeve

(A)

Soil Surface

Figure 6

Diagram of zero-contamination soil sampling procedure

Sampling and analysis of soil

865

The insertion of a sampling probe into soil requires considerable force. As a result,

collection of soil cores by human power is generally limited to the top 20–25 cm.
Sampling below this depth requires some form of mechanical assistance. A common
method used to insert and remove the probes is by hydraulic ram mounted on a three-
point hitch of a tractor. A percussion method using an electric rotary hammer can also
be used but is physically more demanding than hydraulic equipment and requires that
the sampling tubes be removed from the ground via a ground jack. Pressing soil tubes
into the ground is the least soil-disruptive technique that is currently used.

When a tractor for soil sampling is not available or it is logistically not feasible to

transport such heavy sampling equipment to distant field sites, the coring approach
depicted in Figure 7 represents another viable option. Using this method, the top 5 cm
of soil is first carefully removed using a hand trowel from within a metal retainer
sleeve and placed into a pre-labeled container. Next, a narrower probe that can be
lengthened by attaching additional sections of pipe is used to collect discrete sub-soil
samples. As before, plastic cartridges within the probe prevent the soil from actually
contacting the metal probe. The probe can be inserted into the soil profile by electric
drill, rotary hammer, or plastic mallet.

3.3.1

Influence of soil core diameter on study results

Assuming proper soil surface preparation (i.e., smooth with no soil clods or crop de-
bris) and test substance application, the diameter of the soil probe does not generally
impact observed pesticide residue concentrations in soil or associated variability.

,30

Nevertheless, a minimum diameter of 5 cm for the upper soil probe is recommended
to improve sampling under less than ideal conditions. Increasingly, researchers are
using probes having diameters

>5 cm with good results under a variety of field

conditions.

3.3.2

Minimizing plot disturbance and cross-contamination

Great care should be taken while moving in and around the plots so that the sampling
areas are not disturbed. The importance of minimizing soil surface disturbance and
drag down during sampling is critical as one tries to assess the potential mobility
of an agrochemical. This is particularly an issue when one attempts to collect many
samples from a relatively small area. In general, the risk of sub-surface contamination
is greatly minimized by using zero contamination sampling techniques.

To avoid cross-contamination of control samples, untreated controls are collected

before the treated samples. Preferably, personnel who handle the upper cores should
be different from those handling the lower depth cores. This further reduces poten-
tial cross-contamination of lower depth cores. Sampler handlers should change their
gloves each time a new subplot is sampled. The use of disposable shoe covers also
lessens the possibility of cross-contamination.

Once the soil cores have been collected, all boreholes must be backfilled with

untreated soil (with frequent tamping) to prevent bypass flow that could transport
residues into the lower soil profile. After backfilling, flags or stakes should be placed
at the boreholes. This serves as an additional check to ensure that sub-plots are not
sampled more than one time during the study. (Note that these boreholes should

866

Best practices in the generation and analyses of residues in environmental samples

0 cm

10 cm

20 cm

40 cm

Top

Soil Layer

0 - 5 cm

Sub-soil Layer

5 - 20 cm

Stainless steel

retainer sleeve

12-cm (ID)

Step 1

Step 2

5 - 20 cm

Step 3

Disassembling of the

corer

and replacing the

cartridge

0 cm

Step 5

10 cm

20 cm

40 cm

Step 4

Reassembling of the corer
and drilling down to 40 cm

60 cm

80 cm

Step 6

Reassembling of the corer

and drilling further down

5 - 20 cm

20 - 40 cm

Figure 7

Alternative zero-contamination sampling method for soil

Sampling and analysis of soil

867

be periodically checked for subsidence over time and backfilled with soil again, if
necessary, to prevent water infiltration.)

3.3.3

Cleaning procedure for soil sampling equipment

All sampling equipment coming in contact with treated soil (e.g., sample probes and
sectioning equipment) must be thoroughly cleaned between compounds and collec-
tion periods. Cleaning is best accomplished by first brushing off any soil adhering to
equipment. The next step is washing with pressurized water or soap and water, and
finally rinsing with a solvent such as acetone or isopropyl alcohol, alone or in combi-
nation with clean water. The use of a solvent will facilitate faster drying of equipment.

3.3.4

Protection of sample integrity

All application verification and soil samples must be individually labeled with unique
sample identification (ID) and other identifying information such as study ID, test
substance name, sample depth, replicate, subplot and date of collection, as appropriate.
Proper study documentation requires that sample lists and labels be created prior
to work commencing in the field. Water- and tear-resistant labels should be used
since standard paper labels may become water-soaked and easily torn during sample
handling. Sample lists should have the same information on them as the labels and
are a convenient place to record plot randomization, initials of the individual who
collected the sample, and date of collection. As such, the sample list is important in
establishing chain of custody from the point of sample collection until its arrival at
the laboratory.

As soon as the sample has been properly labeled and recorded, it should be placed

in a generator-powered chest freezer located directly in the field. A flat-bed trailer can
be used to transport freezers to and from the field site. Insulated boxes filled with dry-
ice can be used as a substitute for freezers. However, chest freezers typically work
better than dry-ice since they allow more cold air circulation around the samples,
facilitating more rapid freezing.

After the samples have been placed in the freezer, it is critical that they remain frozen

until analysis. Electronic temperature data-loggers can be used to monitor conditions
during storage. Simpler techniques, such as inverting plastic tubes partially filled with
ice or placing plastic bags containing ice cubes, can also be used in combination with
a mercury thermometer (any movement of the ice in the inverted tube or melting of
the ice cubes indicates that the soil samples may have been subjected to temperatures
>0

◦

C and, hence, sample integrity potentially compromised). Since electronic data-

loggers are fairly inexpensive, however, continuous monitoring of freezer storage
conditions is strongly recommended.

3.3.5

Zero-time recovery and importance of the soil micro-layer

Proper sample collection and handling are the key to acceptable agrochemical re-
covery at zero time. The zero-time sample interval is defined as the first sample
collected after application. Zero-time soil samples should be collected within 3 h
after application. Zero-time soil core concentrations, such as those given in Table 3,

868

Best

actices

in
the

ener

ation

and

analyses

of
residues

in
en
vir
onmental

samples

Table 3

Summary of zero-time soil concentration and application verification (AV) monitor results for Pyraclostrobin applied at two field sites

Site

Nominal soil

Calculated soil concentration

Maximum observed

Day maximum

Recovery (%) based

location

concentration

based on average

concentration

on application rate

(state)

(mg kg

−1

)

pass time (mg kg

−1

)

on (mg kg

−1

)

observed (DALA)

(0.28 kg a.i. ha

−1

)

(A) Zero-time soil recovery results

CA – bare soil

0.25

0.281

± 0.003

0.236

94 (104)

FL – bare soil

0.25

0.282

± 0.003

0.123

49 (53)

Days after last application.

The number in parentheses denotes procedural correction using a 90% recovery for the CA site and a 93% recovery for the FL site.

AV – fortified samples:

AV – spray samples:

mean concentration (

µg)

total a.i. recovered (

µg)

Expected fortification

Observed

Expected

Observed

Site/application no.

(nominal/assessed)

fortification

AV – spray

Recovery (%)

(B) Application verification (AV) monitor results

CA – App. 1

420.0/423.8

419.3

535

529.3

CA – App. 2

403.4

535

483.2

CA – App. 3

387.0

535

480.3

CA – App. 4

413.1

535

507.4

FL – App. 1

420.0/423.8

365.4

535

476.2

FL – App. 2

349.1

535

482.4

FL – App. 3

385.0

535

501.2

FL – App. 3

372.3

535

482.2

Sampling and analysis of soil

869

are calculated by first subtracting any parent residue present in the core before last
application (e.g.,

−T4) from the parent residue measured immediately after the last

application (e.g., T4). For example, at the CA site, the soil concentration of BAS 500 F
one day after last application (DALA) was 0.769 mg kg

−1

. Prior to application, the soil

concentration was 0.533 mg kg

−1

. By subtraction, a concentration of 0.236 mg kg

−1

was determined for BAS 500 F in the 0–8-cm section. This results in a zero-time soil
recovery of 94% [(0.236 mg kg

−1

)

/(0.25 mg kg

−1

)

× 100]. The parent residue con-

centration used to calculate recovery was the maximum concentration reached at any
time during sampling after the last application. Zero-time core recoveries (corrected)
ranged from 53 to 104% for the FL and CA sites (Table 3). These data show that even
when considerable effort has been expended on proper test substance application (as
evident by the excellent pass-time and AV recovery results) and sampling, zero-time
recoveries are frequently lower and more variable than desired.

Discrepancies between AV monitor and pass-time (or catch-back) results and actual

zero-time soil concentrations are most likely due to residue losses occurring during
sample handling. Similar discrepancies may also arise for very labile compounds
owing to rapid abiotic and/or biotic losses in soil; the presence of degradates in zero-
time samples would indicate that low zero-time recovery was due to degradation
losses. Immediately after application, all residues, with the exception of those com-
pounds that are soil incorporated, are located in the uppermost layer of the soil core.
This thin layer of surface soil is called the soil micro-layer. Loss of soil micro-layer
residues is believed to be the main reason for low and/or highly variable zero-time
recoveries from soil cores. Initial loss of the soil micro-layer is also believed to be the
reason why maximum residue concentrations commonly occur days to weeks after
application rather than at time-zero.

Until these surface residues are redistributed

into the core by capillary action, precipitation, or irrigation, they remain subject to
loss. Careful handling of the soil samples in the field and laboratory remains especially
critical until surface residue redistribution has occurred.

Empirical evidence supporting the role of soil micro-layer losses in zero-time issues

is given by the often-seen rise in post zero-time residue recoveries. The improved
recoveries likely result from the micro-layer residue redistribution that reduces losses
of the highly concentrated surface residues. There has been some speculation that zero-
time core recoveries may be due to volatilization losses not measured by standard
laboratory studies. If this were the case, however, increases in residue concentrations
would not occur over time since volatilized residues would be lost to the atmosphere.

3.3.6

Sectioning of soil cores

The upper soil core can be further sectioned into

≥2.5-cm lengths according to study

needs and purposes. Sectioning of the upper core can be done in the laboratory but is
most efficiently performed immediately after the core has been removed from the soil
profile. In-field sectioning begins by using a metal or plastic ‘punch’ having a wide
circular surface on one end to push the lower portion (i.e., the end furthest from soil
surface) of the core out of the liner to the desired length. Next, a metal cutting tool
(e.g., knife or spatula) is used to slice the soil core at the correct length. As the soil
is being sliced, it is directed into a pre-labeled sample bag. This process is repeated,
working from the lower to upper portion of the core, until all the appropriate sections

870

Best practices in the generation and analyses of residues in environmental samples

have been sliced away. The sample bags should be rotated in and out of the on-site
freezer until all the sectioning depths have been collected from each core within a
subplot. This technique works well for all soil textures.

Once the lower 15–120-cm cores are completely frozen, they can be further sec-

tioned into 10–15-cm lengths using a hacksaw or band saw. As before, red and black
caps are placed on the tops and bottoms of each newly created core section. Each
new section also receives a unique sample ID number and new label containing all
pertinent sample information. Care must be used when cutting frozen cores to prevent
damage to original sample labels. An advantage of the sampling approach shown in
Figure 7 is that the soil cores generally require no additional sectioning.

3.3.7

Field-fortification samples

In order to determine the dissipation rate and assess the potential mobility of an
agrochemical in soil, it is crucial that the residue level measured in a particular sample
reflects the actual concentration present in the soil profile at the time of sampling. If this
basic assumption cannot be assured, the validity of resulting data may be questioned.
Regulatory concerns have arisen over past improper sample-handling practices that
might have artificially accelerated agrochemical dissipation in the soil samples. This
could occur, for example, whenever samples are exposed to elevated temperatures
and/or direct sunlight for extended periods of time prior to freezer storage. As a
result, regulatory authorities have requested that a set of fortified samples having a
known amount of active ingredient be prepared in the field. These field fortification
samples are intended to indicate how well the integrity of the actual field samples was
preserved during sample collection, transportation, and storage. If the field-fortified
residues are found to be stable, the sample handling conditions are deemed sufficient
also to have protected the integrity of the actual field samples. In contrast, if the
recovery from the field fortification samples is low, this implies that sample integrity
was compromised at some point during study conduct.

Although theoretically sound, field fortification samples often generate as many

questions as they answer. This is because accurate and precise fortification of soil is
difficult to accomplish under field conditions except when the field site is very near
the supporting laboratory. For a distant field site, the fortification solution is typically
prepared and assayed in the laboratory prior to overnight shipment. If agrochemical
recovery from the resulting field fortification samples is low, this may be due to
accelerated dissipation, problems associated with the fortification solution itself or
improper technique used by field personnel. Shipping fortifying solutions to the field
is further complicated by the fact that many active ingredients make only suspensions,
not true solutions. Once frozen or left without agitation for extended periods, these
formulations are difficult to re-suspend, as is required for proper soil fortification. As
a result, acceptable recovery from field spikes helps to address the issue of sample
integrity, but poor recovery only results in more questions as to its cause.

A solution to this dilemma is to place soil samples immediately in a freezer lo-

cated in the field, the temperature of which is continuously monitored, as described
previously. Laboratory-prepared storage study samples can then be used to determine
test substance stability under freezer storage conditions that match those used in the
field and during transportation and final storage. If a valid laboratory storage stability

Sampling and analysis of soil

871

study indicates that residues are stable, any observed decline in soil residues can then
be assumed to have occurred in situ. Details on the conduct of a freezer storage study
are given in Section 4.

3.3.8

Test plot maintenance

The guiding principles in test plot maintenance are to (1) minimize soil surface distur-
bance at all times, (2) ensure that control and treated plots are similarly maintained,
(3) avoid applying other agrochemicals that may interfere with sample analysis or
that are otherwise contrary to the purpose of the study, (4) follow the prescribed ir-
rigation policy determined for the study site, and (5) keep bare-soil test plots free of
vegetation, as follows.

For bare-soil studies, vegetation is controlled on an ‘as-needed basis’ by application

of nonselective herbicides (e.g., glyphosate, paraquat, glufosinate) or by careful hand
weeding. Vegetation control may be required on a weekly basis during the growing
season. The use of glyphosate or paraquat is a widely accepted means of controlling
unwanted vegetation in and around test plots, and has the added advantage of control-
ling weeds without physically disturbing soil surfaces. Because physical disturbance
of the soil surface is to be avoided, hoeing or other forms of mechanical removal
should not be used in the actual test plots. Vegetation that is pulled by hand should
remain on the test plots to avoid inadvertent removal of agrochemical residues.

3.3.9

Irrigation

Because soil moisture plays such a critical role in determining agrochemical dissipa-
tion rate and mobility, it is important to devise carefully an irrigation plan that clearly
specifies the timing and amount of irrigation that is to be added at each study site. One
must be able to justify all irrigation applications based upon the relevant agricultural
practices in the study region and actual use pattern of the agrochemical.

For studies conducted in regions of irrigated agriculture, the plots must be irrigated

according to the soil-water budget method. This is determined by calculating the
evapotranspiration rate for the target crop (ET

) and adjusting irrigation amounts to

110% of the ET

= ET

× K

(4)

Irrigation to apply

= ET

× 110%

(5)

where ET

is the actual daily evapotranspiration rate and K

is the specific crop

coefficient based on the targeted crop and appropriate growth stage. Deficiencies
should be reconciled about every 10 days, as required.

In regions of rain-fed agriculture, the test plots must receive 110% of the monthly

historical rainfall. Differences in this total should be reconciled every 10 days. If the
plots do not receive 110% of historical monthly rainfall, the study may be severely
compromised.

Apply the supplemental water inputs via sprinkler irrigation. Do not flood or furrow

irrigate since these practices may disturb soil surface residues. Be aware that even

872

Best practices in the generation and analyses of residues in environmental samples

sprinkler irrigation can cause uneven application of water and, if leaks occur, severe
erosion of the soil surface. Therefore, regularly inspect irrigation equipment and
function. The control and treated plots must be irrigated in a similar manner. Record
the volume and date of all irrigations, the source of irrigation water, and the type
of irrigation system used. If water begins to pool or run off of the soil surface, stop
irrigating immediately. Resume irrigation only after the risk of runoff is over. To avoid
runoff, carefully match the application rate to the soil infiltration rate. Note that, in
cold climates, irrigation equipment is winterized to prevent damage from freezing
and is generally not available for use during the winter months.

Phase III: sample processing and analysis

Once soil samples have been received and properly logged in by the laboratory, there is
a multi-step process required to isolate agrochemical residues from the sample matrix
so that sensitive, reproducible analysis can occur. Residue methods for agrochemicals
in soil involve the basic steps shown in Figure 8.

Cleanup

Derivatization

Analyte

Quantitation*

Cleanup

Extraction

Homogenization

*HPLC-UV, GC-ECD, GC-MS, LC-MS

Figure 8

Schematic of general analytical method for soil analysis

Sampling and analysis of soil

873

A general overview of each of these steps is given below. This is followed by a

specific example involving an increasingly powerful quantitation technique, liquid
chromatography/tandem mass spectrometry (LC/MS/MS).

4.1

Sample homogenization

Soil homogenization is the critical first step in the analysis of soil samples. Improper
homogenization can lead to variable results that seriously confound the interpretation
of soil residue data. Samples are commonly homogenized using equipment called
size-reducing mills. Size-reducing mills can be further categorized as being ‘grinder’,
‘rotary blade’, or ‘hammer’ type mills. Each of these has advantages and disadvantages
but the ability to mix uniformly the anticipated volume of soil and the ease with which
the mill can be cleaned are key considerations when choosing a particular mill. The
design of the mill should also prevent the loss of fine soil particles generated during the
blending process. Other key aspects of sample homogenization are addressed below.

4.1.1

Protecting sample integrity

When processing samples, they should always be milled using dry-ice in amounts suf-
ficient to ensure that the samples remain frozen during homogenization. As discussed
previously, protecting sample integrity is of utmost concern throughout every aspect
of study conduct. The use of adequate dry-ice also helps keep soil from sticking to
the mill. Some mills have been designed to use liquid nitrogen rather than dry-ice for
cooling, and also work well with soils.

4.1.2

Minimizing cross-contamination

To minimize cross-contamination, soil cores are processed beginning with the lowest
depth samples and progressing to the surface samples. It is very important that the
mill be thoroughly cleaned between samples so as to minimize the risk of cross-
contamination. The machinery should be thoroughly cleaned with water followed by
a water–solvent solution such as acetone. Typically, the machine should be cleaned
after running one replicate set of samples from the lowest depth to the surface. If the
samples have coarse fragments in them, it may be necessary to sieve the samples prior
to homogenization. As mentioned previously, soils with a large percentage of clods
or rocks should be excluded during the site selection process since they also interfere
with sample collection in the field.

4.1.3

Ensuring thorough sample homogenization

Before processing actual study samples, and periodically during the course of a study,
it is important to test the thoroughness of the homogenization procedure using soils
having a range of textures. This is typically done by measuring the analytical variance
between sub-samples, and is the only reliable method for determining the effectiveness
of a blending technique. Depending on the soil type and sample size, it may be
necessary to pass the sample through a mill twice to ensure proper homogenization.
For example, experience has shown that when using a rotary-blade type mill, two
passes are normally required for proper homogenization of turf or sod samples. When

874

Best practices in the generation and analyses of residues in environmental samples

Table 4

Tepraloxydim analytical results used to determine efficacy of soil homogenization

procedure

Description

Residue found (mg kg

−1

)

Sample weight

= 10 g of soil

Sample 1

Control

Not detected

Sample 2

Fortified sample, 0.1 mg kg

−1

0.101

Sample 3

Treated sample, replicate 1

0.120

Sample 3, duplicate analysis

Treated sample, replicate 1

0.110

Sample 4

Treated sample, replicate 2

0.050

Sample 4, duplicate analysis

Treated sample, replicate 2

0.057

Sample weight

= 5 g of soil

Sample 5

Control

Not detected

Sample 6

Fortified sample, 0.1 mg kg

−1

0.099

Sample 7

Treated sample, replicate 1

0.110

Sample 7, duplicate analysis

Treated sample, replicate 1

0.180

Sample 8

Treated sample, replicate 2

0.054

Sample 8, duplicate analysis

Treated sample, replicate 2

0.068

Sample weight

= 2 g of soil

Sample 9

Control

Not detected

Sample 10

Fortified sample, 0.1 mg kg

−1

0.102

Sample 11

Treated sample, replicate 1

0.148

Sample 11, duplicate analysis

Treated sample, replicate 1

0.133

Sample 12

Treated sample, replicate 2

0.059

Sample 12, duplicate analysis

Treated sample, replicate 2

0.063

turf samples are being processed, it is also essential that the sod plug be totally frozen
so that the plug will break up as it passes through the mill.

An example of adequate sample homogenization is given in Table 4. The exper-

iment was conducted with two replicate treated soil samples. Each replicate was
analyzed in duplicate. Three different sample aliquots (2, 5 and 10 g) were used
from each replicate. Analyses of controls and fortified samples were also conducted
concurrently with treated samples to evaluate method performance (i.e., extraction
recoveries). These results show that residue values are the same regardless of sample
size. Thus, thorough homogenization of soil samples coupled with rugged analytical
methodology provides for satisfactory residue analysis.

4.2

Sample extraction

An efficient and reproducible extraction procedure is mandatory when analyzing
agrochemicals in soil. An overview of common soil extraction techniques is given
below.

4.2.1

Solvent selection

Soil samples are generally extracted with one or more organic solvents mixed with
up to 10% (v/v) water. A wide variety of solvents is used for extraction, the choice

Sampling and analysis of soil

875

of which depends upon the polarity of the compound to be extracted.

For example,

extraction with methanol and methanol–water usually works well for compounds
with medium to high polarity. Acetonitrile is another common solvent used in soil
extractions. Sometimes pH adjustment is also required for compounds that are acidic
or basic in nature (e.g., ammonium carbonate is often added to improve the extract-
ability of weak organic acids). Starch-encapsulated formulations may benefit from
an enzymatic pretreatment prior to extraction from soil.

Several extraction techniques are used in the analysis of soil. The following are

brief descriptions of some of the most commonly used techniques.

4.2.2

Mechanical shaker

A commonly used extraction technique involves shaking soil with a suitable solvent on
a mechanical shaker at about 300 rpm. After extraction, the soil extracts are collected
by centrifugation followed by decantation or filtration. This technique could be used
for any amount of soil samples (from 10 to

>100 g). Soil samples greater than 100 g

require efficient agitation to achieve acceptable recoveries.

4.2.3

Soxhlet extraction

This technique is used to extract effectively analytes that are polar in nature and
strongly bound to soil. Typically, a solvent mixture containing a water-miscible solvent
and an apolar solvent (e.g. methanol–dichloromethane) is used. A small aliquot of soil
(10–30 g) is dried by mixing with sodium sulfate and refluxed for 8–16 h to extract
the residues.

4.2.4

Sonication

This technique is used mainly for nonpolar compounds. Typically a small aliquot of
soil (10–30 g) is dried by mixing with sodium sulfate prior to extraction. Next, the
sample is extracted with a solvent for 10–20 min using a sonicator probe. The choice
of solvent depends on the polarity of the parent compound. The ultrasonic power
supply converts a 50/60-Hz voltage to high-frequency 20-kHz electric energy that is
ultimately converted into mechanical vibrations. The vibrations are intensified by a
sonic horn (probe) and thereby disrupt the soil matrix. The residues are released from
soil and dissolved in the solvent.

4.2.5

Supercritical fluid extraction (SFE)

SFE is used mainly for nonpolar compounds [e.g. polychlorinated biphenyls (PCBs)].
Typically, small aliquots of soil (0.5–10 g) are used for extraction. The extraction sol-
vent is a supercritical fluid, most commonly carbon dioxide, which has properties
of both a liquid and gas. The supercritical fluid easily penetrates the small pores of
soil and dissolves a variety of nonpolar compounds. Supercritical carbon dioxide ex-
tracts compounds from environmental samples at elevated temperature (100–200

◦

and pressure (5000–10 000 psi). High-quality carbon dioxide is required to minimize

876

Best practices in the generation and analyses of residues in environmental samples

analytical interferences. Compounds with different chemical natures can be selec-
tively extracted by varying the extraction pressure and temperature. The addition of an
organic modifier, such as methanol, may improve the recoveries of polar compounds.

4.2.6

Accelerated solvent extraction (ASE)

This fully automated process developed by Dionex is used for a variety of compounds
having a wide range of polarities.

Typically, a small aliquot of soil (0.5–20 g) is

extracted using a variety of solvents. As with other techniques, the solvent choice
depends upon the polarity of the compound to be extracted. The unit extracts soil at
elevated temperatures (

>60

◦

C) and pressures (

>1000 psi). The increased temperature

accelerates the extraction kinetics while the elevated pressure keeps the solvent(s)
below the boiling point, thus allowing safe and rapid extraction. Both time and solvent
consumption are dramatically reduced compared with mechanical shaking. There are
now several published United States Environmental Protection Agency (USEPA)
methods that use ASE (e.g., USEPA Method 600/4-81-055, ‘Interim Methods for the
Sampling and Analysis of Priority Pollutants in Sediment and Fish Tissue’).

4.2.7

Microwave extraction

This is a relatively new technique that is used for PCBs and other nonpolar, volatile
and semi-volatile organic compounds. Typically, a small aliquot of soil sample
(0.5–20 g) is used for the extraction. Soil samples are extracted with one or more
organic solvents using microwave energy at elevated temperature (100–115

◦

C) and

pressure (50–175 psi). This method uses less solvent and takes significantly less time
than Soxhlet extraction but is limited to thermally stable compounds.

4.3

Sample cleanup

Trace analysis of soil samples often requires post-extraction cleanup to remove co-
extracted matrix interferences. There are several difficulties that may arise during
chromatographic analysis due to interferences present in sample extracts. To avoid
these and other issues, one or more of the following cleanup techniques are often used.

4.3.1

Liquid–liquid partition

This technique provides a convenient method for separating an agrochemical com-
pound from a highly aqueous extraction mixture. The partitioning solvent is usually
a volatile, water-immiscible organic solvent that can be removed by evaporation after
the desired component has been extracted. This technique is based on the principle
that when a substance is soluble to some extent in two immiscible liquids, it can be
transferred from one liquid to another by shaking. The degree of partitioning from one
solvent to the other depends on the agrochemical’s distribution coefficient between
the immiscible liquids. This technique is particularly useful for the cleanup of ioniz-
able compounds, since the pH of the aqueous solution can be adjusted to maximize
partitioning into the organic or water phases, as desired.

Sampling and analysis of soil

877

4.3.2

Solid-phase extraction (SPE)

This technique is based on the same separation mechanisms as found in liquid chro-
matography (LC). In LC, the solubility and the functional group interaction of sample,
sorbent, and solvent are optimized to effect separation. In SPE, these interactions are
optimized to effect retention or elution. Polar stationary phases, such as silica gel,
Florisil and alumina, retain compounds with polar functional group (e.g., phenols,
humic acids, and amines). A nonpolar organic solvent (e.g. hexane, dichloromethane)
is used to remove nonpolar inferences where the target analyte is a polar compound.
Conversely, the same nonpolar solvent may be used to elute a nonpolar analyte, leaving
polar inferences adsorbed on the column.

The most common technique used for agrochemicals is reversed-phase SPE. Here,

the bonded stationary phase is silica gel derivatized with a long-chain hydrocarbon
(e.g. C

–C

) or styrene–divinylbenzene copolymer. This technique operates in the

‘reverse’ of normal-phase chromatography since the mobile phase is polar in nature
(e.g., water or aqueous buffers serve as one of the solvents), while the stationary phase
has nonpolar properties.

Ion-exchange solid-phase extractions are used for ionic compounds. The pH of the

extracts is adjusted to ionize the target analytes so that they are preferentially retained
by the stationary bonded phase. Selection of the bonded phase depends on the pK

of the target analytes. Sample cleanup using ion exchange is highly selective and

can separate polar ionic compounds that are difficult to extract by the liquid–liquid
partition technique.

A variety of solid-phase cartridges are available from a number of different manu-

facturers (e.g. J.T. Baker, Varian). Most cartridges, however, use a similar extraction
procedure that consists of these basic steps:

1. Conditioning the column. This step prepares the column to absorb the analytes and

also pre-washes the column with the solvents that are used for the cleanup.

2. Sample application. The sample extract is dissolved in the weaker solvent and

applied to the top of the column. The analytes of interest are extracted from the
crude sample extract and are adsorbed on the column.

3. Wash. Solvents, weaker than the elution solvents, are used to remove interferences

selectively.

4. Elution. The compound of interest is selectively eluted with a stronger solvent.

4.4

Derivatization techniques

A derivatization technique is commonly applied to an agrochemical with certain re-
active functional groups (e.g., carboxylic acid, amine, phenol) to make the compound
amenable to either gas chromatography (GC) or LC analysis. An in-depth discus-
sion of derivatization reactions used in the analysis of agrochemicals is beyond the
scope of this article. For more information on this topic, the reader is referred to
Knapp.

878

Best practices in the generation and analyses of residues in environmental samples

4.5

Analytical detection and quantitation techniques

The most common final separation techniques used for agrochemicals are GC and LC.
A variety of detection methods are used for GC such as electron capture detection
(ECD), nitrogen–phosphorus detection (NPD), flame photometric detection (FPD)
and mass spectrometry (MS). For LC, typical detection methods are ultraviolet (UV)
detection, fluorescence detection or, increasingly, different types of MS. The excellent
selectivity and sensitivity of LC/MS/MS instruments results in simplified analytical
methodology (e.g., less cleanup, smaller sample weight and smaller aliquots of the
extract). As a result, this state-of-the-art technique is becoming the detection method
of choice in many residue analytical laboratories.

An example of an LC/MS/MS method with an LOQ of 0.01 mg kg

−1

is illustrated

in Figure 9. This method was used to analyze tepraloxydim and its primary metabolite

Soil (25 g)

- Extract with dichloromethane 3 X 50 mL
- Centrifuge

Combined dichloromethane extract

Marc

(discard)

- Evaporate to dryness

Dissolve in acetonitrile-water (80:20, v/v)

- Dilute with:

• Acetonitrile-water (1:1) + 0.1% formic acid

• Methanol-water (1:1) + 0.1% formic acid,

4 mM ammonium formate

LC/MS/MS determination

Analysis for tepraloxydim (m/z 342 to 250)
and DP-6 (m/z 253 to 197) in positive ion mode

Modifications were used for different soil types.

Tepraloxydim

DP-6

Figure 9

Method diagram for the determination of tepraloxydim and its degradate, DP-6, in soil

(LOQ 0.01 mg kg

−1

)

Sampling and analysis of soil

879

Table 5

Recoveries of tepraloxyim and degradates from soil dissipation studies conducted in the USA and Canada

Recovery range (%)

Mean recovery (%)

Compound
fortified

North Dakota

Mississippi

California

North Dakota

Mississippi

California

(A) US sites

Tepraloxydim

78–119

74–106

86–113

± 10

± 7

100

± 9

= 46)

= 44)

= 26)

DP-6

69–116

71–102

77–102

± 11

± 7

= 46)

= 44)

= 26)

Recovery range (%)

Mean recovery (%)

Compound
fortified

Manitoba

Saskatchewan

Alberta

Manitoba

Saskatchewan

Alberta

(B) Canadian sites

Tepraloxydim

77–110

72–121

70–107

± 9

± 8

= 39)

= 44)

= 43)

DP-6

71–116

74–119

72–118

± 10

± 11

± 16

= 39)

= 44)

= 43)

Fortification range for all three sites was 0.01–0.1 mg kg

−1

DP-6 over 3000 soil samples collected from several terrestrial field dissipation studies.
The sample procedural recoveries using this method, conducted concurrently with the
treated samples during soil residue analysis, are summarized in Table 5. This method
was proven to be short, rugged, sensitive, and suitable for measuring residues in soil
and sediment at levels down to 0.01 mg kg

−1

. The reproducibility of the methods also

indicated acceptable method performance and, as a result, thousands of samples were
analyzed using this methodology.

4.6

Freezer storage stability

Most agrochemicals remain stable in frozen soil for many months. However, it is
important to verify this stability by conducting a freezer storage stability study. One
type of study is conducted by fortifying known amounts of test substance and its
major transformation products into control soil collected from a participating field
site. Fortification normally occurs at two levels: replicate soil samples are fortified
at the LOQ and at the highest expected residue concentration for each analyte of
interest. The fortified soil samples are stored under the same conditions as the field
samples and analyzed at different time periods that bracket the storage time of the
actual field samples. The recoveries of the storage samples are compared with those
obtained from day zero analyses to obtain the storage stability. In general, the method
of analysis is the same as used for the soil residue analysis.

A second approach to determining freezer storage stability involves the reanalysis

of incurred residues found in actual samples that are stored over time. Using this
approach, soil from an actual field sample containing residues is periodically analyzed

880

Best practices in the generation and analyses of residues in environmental samples

during the course of the analysis phase of the study. A key advantage of this method
is that the stability of actual field-derived residues is being determined. The main
drawback is that this approach does not work for degradates that do not form in the
field at concentrations at or above their LOQ values.

Phase IV: reporting of results

Once soil samples have been analyzed and it is certain that the corresponding results
reflect the proper depths and time intervals, the selection of a method to calculate
dissipation times may begin. Many equations and approaches have been used to help
describe dissipation kinetics of organic compounds in soil. Selection of the equation
or model is important, but it is equally important to be sure that the selected model is
appropriate for the dataset that is being described. To determine if the selected model
properly described the data, it is necessary to examine the statistical assumptions for
valid regression analysis.

5.1

Goodness of fit testing

There are two statistical assumptions made regarding the valid application of mathe-
matical models used to describe data.

The first assumption is that row and column

effects are additive. The first assumption is met by the nature of the study design,
since the regression is a series of X , Y pairs distributed through time. The second
assumption is that residuals are independent, random variables, and that they are
normally distributed about the mean. Based on the literature, the second assumption
is typically ignored when researchers apply equations to describe data. Rather, the
correlation coefficient (r ) is typically used to determine goodness of fit. However,
this approach is not valid for determining whether the function or model properly
described the data.

In Figure 10, two solutions (models) are shown for the same data set. The first

solution is based on a linear fit (Hamaker equation) that provided a high correla-
tion coefficient of 0.93. The second solution (Gustafson–Holden model) is based on
a nonlinear solution that provided a high correlation coefficient of 0.98. However,
based on an examination of the residuals from both equations, it is evident that the
linear model failed to describe properly the data based on the second assumption for
valid regression analysis (Figure 11). In other words, the residuals were not randomly
distributed; initially they are greater than zero but become increasingly negative as
time progresses. In contrast, the residuals from the nonlinear model are equally neg-
ative and positive throughout time and it appears, therefore, that the nonlinear model
fulfills the second assumption for valid analysis (Figure 12). The second assumption
for valid analysis becomes especially important when kinetics are implied based on
the fit of the model. However, a kinetic model truly cannot be proven by a fit to data
from a field dissipation study.

–

Therefore, the appropriateness of a model should

be determined by its ability to empirically describe the data without implication of
mechanism (order).

Sampling and analysis of soil

881

100

120

140

160

180

200

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

Gustafson--Holden (R=0.98)
Hamaker (r=0.93)

ln of Concentration (mg kg

−

)

Days After Application

Figure 10

Comparison of linear (Hamaker) and nonlinear (Gustafson–Holden) solutions for a

typical soil dissipation data set

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

Hamaker

Residuals

100

120

140

160

180

200

Days After Application

Figure 11

Residuals plot for the linear model

5.2

Models for agrochemical dissipation in soil

Since many equations and analysis procedures have been described in the literature,
we present here just a few of the most commonly used equations. The solutions to
these equations are obtained using a nonlinear curve fitting routine found in many
commercially available statistical programs.

882

Best practices in the generation and analyses of residues in environmental samples

-20

100

120

140

160

180

200

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

Residuals

Days After Application

Gustafson--Holden

Figure 12

Residuals plot for the nonlinear model

5.2.1

Hamaker equation

The equation by Hamaker

is one of the most commonly used methods for describing

dissipation kinetics using a linear fit. The basic computational form of the equation is

= a exp(−bX)

(6)

This equation is satisfactory for data sets that are linear when ln of concentration is
plotted vs time.

5.2.2

Hamaker equation (power rate form)

As mentioned previously, most agrochemicals do not exhibit linear degradation pat-
terns. As a result, Hamaker

proposed another variation of the linear-fit equation that

allows better description of nonlinear data sets:

= a

−n

+ (n − 1)bX

−n

(7)

where n

= 1; n is the rate order and a and b are solved as unknowns 1 and 2. The

disadvantage of this type of approach is that the user is simply choosing a power or
‘order’ that empirically describes the data better than the single exponential form of
the equation.

5.2.3

Timme–Frehse–Laska equation

In a similar approach to Hamaker, Timme et al.

proposed six functions that are

also empirically based. However, they took the additional step of suggesting that the
choice of the equation should be based on the regression correlation coefficient (r ).

Sampling and analysis of soil

883

However, regression coefficients cannot be used to determine the adequacy of a model
choice, as discussed previously.

Order 1:

log X

log b

(8)

Order 1.5:

(

√

− 1)

(9)

Order 2:

− 1)

(10)

Similarly to the Hamaker parameters, a and b are solved as unknowns 1 and 2.

5.2.4

Gustafson–Holden equation

The Gustafson–Holden equation

is a unique approach that allows both linear and

nonlinear datasets to be solved since it is based on a gamma distribution. The equation
is first order and has three unknowns (a, b and c):

= a − b ln(1 + cX)

(11)

This equation requires more data points than the previous equations.

5.2.5

Wolt equation

The Wolt equation

is also a unique approach that is described as being a quasi-first-

order equation. This equation also has three unknowns that are solved (a, b, and c):

= a + b exp(−cX) + e

(12)

The variable e has been described as an error term, but is not used in most applications
of the equation.

5.3

versus T

values

It is important that a clear distinction be made between DT

and T

values. A DT

implies that the value describes the time required for 50% of the starting concentration
to dissipate or degrade. A T

result implies that the number is derived from a rate

constant, which may or may not describe where 50% of the starting concentration has
dissipated or degraded. If a logarithm concentration data set is nonlinear with time,

884

Best practices in the generation and analyses of residues in environmental samples

a rate constant will not accurately describe the data. If the dataset is linear, the rate
constant and the DT

result should be about equal. A rate constant solution describes

a data set with the assumption that an equal change in concentration occurs with an
equal change in time. The Hamaker equation is an example of one of the most widely
used rate constant equations.

5.4

Determining water balance and leaching potential

One of the objectives for a field dissipation study is to determine how the leaching
behavior of an agrochemical is correlated with water inputs occurring at the field
site. In order to answer this question, researchers often overlay water additions on
top of graphs displaying residue movement. However, this method often falls short of
answering the basic question of whether sufficient water was applied to allow leaching
to occur. For example, clay loam soils have on average a 6.4-cm water holding capacity
per 30-cm depth. If the water content of the clay soil is approximately at permanent
wilt point and a 4-cm irrigation event occurs, the 30-cm depth of soil will not reach
field capacity. If the field capacity is never exceeded, no movement of soil solute from
the 0- to 30-cm depth would be expected to occur. (These techniques do not address
preferential or by-pass flow processes where agrochemicals are transported to subsoils
via water following root channels, cracks, etc. Techniques to address preferential flow
are not well established at this time.) If three days later an additional 3.2-cm rainfall
event occurred, the 0- to 30-cm depth of soil would still not have been brought back
to field capacity (assuming 0.7-cm evaporation on the previous two days).

For these reasons, it is desirable to perform a series of simple calculations to

determine if the field capacity for a given depth of soil is ever exceeded, rather than
simply overlaying water inputs over plots of residue data. The following series of
calculations addresses the primary issue of whether sufficient water was applied to
the test system at appropriate intervals to create leaching opportunities:

Surface-layer calculation:

[(P

+ SM + I ) − (Q − ET

)]

(13)

Sub-surface-layer calculation:

(Inf

− RF

)

(ET

θ in an overlying layer = 0)

(14)

where

= time in days

θ = volumetric water content

= precipitation

= snow melt (when snow pack exists and ambient temperature is >0

◦

Sampling and analysis of soil

885

= irrigation

= runoff

= evapotranspiration corrected for the crop (ET

= ET

× K

) or E

soil

= ET

× k)

Inf

= infiltration

= root extraction factor RF = RF × c, c = 1.0

Once performed, these calculation results can be graphed as shown in Figure 13.

This type of information provides more insight into the soil water status at a site than
simply graphing rainfall. This figure also helps determine if soil water movement
occurred out of a given depth of soil. Moreover, it is useful to overlay Figure 13 with
a graph of compound movement by depth to determine if the predicted water flux at
a given depth corresponds to actual residue movement.

100

150

200

250

300

0.0

5.1

10.2

15.3

Accumulating Flux Past 120 cm

Days

0.06

0.09

0.12

0.15

90-120 cm

0.09

0.12

0.15

0.18

0.21

60-90 cm

0.09

0.12

0.15

0.18

0.21

30-60 cm

0.06

0.09

0.12

0.15

0.18

0.21

0-30 cm

Volumetric Water Content

Figure 13

Volumetric water contents for Haw series soil calculated using the Penman equation

886

Best practices in the generation and analyses of residues in environmental samples

100

150

200

250

300

0.0

5.1

10.2

15.3

Accumulating Flux Past 120 cm

Volumetric Water Content

Days

0.06

0.09

0.12

0.15

90-120 cm

0.09

0.12

0.15

0.18

0.21

60-90 cm

0.09

0.12

0.15

0.18

0.21

30-60 cm

0.06

0.09

0.12

0.15

0.18

0.21

0-30 cm

Figure 14

Comparison of actual volumetric water contents (measured by time domain reflectom-

etry) and calculated soil-water flux values (Penman equation) at four soil depths

More sophisticated methods that actually measure volumetric water content can

also be used, such as time domain reflectometry (TDR). In Figure 14, an example of
TDR results is presented. Both the calculated and measured (i.e., TDR) volumetric
water contents provide a similar picture of the profile water status by depth with time.
Proper soil characterization data, such as those shown in Table 6, are necessary for
these calculations and improve understanding of the test system. The determination
of water-holding capacity (WHC) at 0.03 MPa field capacity (FC) and 1.5 MPa

Sampling and analysis of soil

887

Table 6

Soil characterization results used in water balance calculations and data interpretations

Depth increment (cm)

Soil characteristic

0–15

15–30

30–45

45–60

60–75

75–90

90–105

105–120

Sand (%)

Silt (%)

Clay (%)

Organic matter (%)

2.1

0.9

0.3

0.2

0.1

Bulk density (g cm

−3

)

1.33

1.41

1.49

1.46

1.45

1.43

1.45

1.47

6.1

7.2

6.1

5.9

6.2

6.4

6.2

WHC

at 0.33 bar (%)

9.9

6.4

5.4

4.7

5.8

5.1

5.4

WHC at 15 bar (%)

5.1

3.4

2.4

2.0

2.1

2.7

2.1

2.3

CEC

(mequiv. per 100 g soil)

8.1

5.8

6.4

3.5

2.9

4.0

2.9

3.1

Textural classification

Loamy

sand

WHC

= water-holding capacity.

CEC

= cation-exchange capacity.

permanent wilt point (PWP) is important for any type of soil-water calculations or
for field sensor measurements.

In Table 7, a comparison of actual measurements, and also two well-known pedo-

transfer functions, can be found by depth. It is important to note that there is a
large difference in water content between the disturbed soil core samples and the
undisturbed samples. Additionally, the two pedo-transfer functions also exhibit a large
difference in predicted water content. Therefore, when doing calculations or trying

Table 7

Measured and estimated volumetric water contents as a function of depth and matrix potential for a Haw series soil

(Payette Country, Idaho)

Volumetric water content

Intact

Disturbed

Pedo-transfer

Intact

Disturbed

Pedo-transfer

soil core:

function I:

function II:

soil core:

function I:

function II:

measured

estimated

measured

estimated

Soil matrix potential, 0.03 MPa

Soil matrix potential, 1.5 MPa

Depth (cm)

(field capacity)

(permanent wilting point)

0–15

28.60

33.50

36.93

27.18

18.73

23.40

19.14

12.05

15–30

27.85

34.90

31.30

27.18

19.07

26.10

17.72

12.05

30–45

35.88

41.10

27.82

27.18

25.17

26.60

14.89

12.04

45–60

43.68

43.80

20.98

26.24

32.27

28.50

7.59

11.48

60–75

37.55

41.10

21.12

26.71

30.23

25.80

7.62

11.75

75–90

39.83

42.40

20.58

24.84

28.30

23.50

8.14

10.68

90–105

38.23

40.70

24.04

24.84

26.03

23.50

13.12

10.71

105–120

38.10

37.10

27.40

24.84

26.30

20.60

16.70

10.73

Estimated using the method of Rawls et al.

Estimated using the method of Bauer and Black.

888

Best practices in the generation and analyses of residues in environmental samples

to calibrate field sensors, the magnitude of the differences observed in Table 7 must
be considered and a compromise should be struck between precision and accuracy.

5.5

Weather data requirements for water balance
and mobility assessments

If basic calculations such as those presented are to be conducted, it is important to
collect enough weather parameters to calculate reference evapotranspiration (ET

An on-site weather station should be considered a basic requirement: minimum sensor
requirements to calculate a Penman equation would include solar radiation, wind
speed, relative humidity or actual vapor pressure, and air temperature. An on-site rain
gauge is essential but it is also a good idea to have a rain gauge on the weather station
even if it is not directly on-site. The most accurate variations of the Penman equation
calculate ET

on an hourly basis. However, Penman routines using daily summaries

are typically satisfactory for the purpose of calculating soil-water recharge.

Summary and conclusions

The proper conduct of a field soil dissipation study represents a significant commit-
ment of labor, money, and time. As such, there are many important study details that
cannot be left to chance, or addressed as an afterthought, once the study is under-
way. Each of the four main phases of study conduct, (1) planning and design, (2)
field conduct, (3) sample processing and analysis, and (4) data handling and report-
ing, is vitally linked to the next. Each phase is critical to study success. This article
addresses key aspects of study design and conduct necessary for successful study com-
pletion. When properly planned and conducted, these studies provide valuable infor-
mation regarding the environmental persistence and mobility of agrochemicals in field
soils.

Abbreviations

ASE

Accelerated solvent extraction

Application verification

ECD

Electron capture detection

Gas chromatography

Liquid chromatography

LOQ

Limit of quantitation

Mass spectrometry

NPD

Nitrogen–phosphorus detection

SFE

Supercritical fluid extraction

SPE

Solid-phase extraction

TDR

Time domain reflectometry

USEPA

United States Environmental Protection Agency

Ultraviolet

, K

Soil sorption coefficients

Sampling and analysis of soil

889

Acid dissociation constant

Regression correlation coefficient

Regression coefficient of determination

Water solubility

References

1. S.K. Singles, R.F. Dietrich, and R.D. McFetridge, ‘Degradation of pyrithiobac sodium in soil

in the laboratory and field,’ in “Pesticide Environmental Fate, Bridging the Gap Between
Laboratory and Field Studies,” ed. W. Phelps, K. Winton, and W.R. Effland, ACS Sympo-
sium Series No. 813, American Chemical Society, Washington, DC, Chapter 15, pp. 207–221
(2002).

2. B. Stenberg, M. Johansson, M. Pell, K. Sj¨odahl-Svensson, J. Stenstr¨om, and L. Torstensson, Soil

Biol. Biochem., 30, 393 (1998).

3. S.L. Trabue, T.M. Crowe, and J.H. Massey, ‘Changes in soil biomass and microbial community

structure as affected by storage temperature and duration: effect on the degradation of metsul-
furon methyl’ in ‘Pesticide Environmental Fate: Bridging the Gap Between Laboratory and Field
Studies’, ed. W. Phelps, K. Winton, and W.R. Effland, ACS Symposium Series 813, American
Chemical Society, Washington, DC (2002).

4. L.O. Osa-Afiana and M. Alexander, Soil Sci. Soc. Am. J., 46, 285 (1982).
5. J.E. Woodrow, J.S. Seiber, and L.W. Baker, Environ. Sci. Technol., 31, 523 (1997).
6. I.J. van Wesenbeeck, J.M. Zabik, J.D. Wolt, G.A. Bormett, and D.W. Roberts, J. Agric. Food

Chem., 45, 3299 (1997).

7. J.M. Zabik, I.J. van Wesenbeeck, A.L. Peacock, L.M. Kennard, and D.W. Roberts, J. Agric.

Food Chem., 49, 3284 (2001).

8. H.J. Strek, Pestic. Sci., 53, 52 (1998).
9. R.C. Knox, D.A. Sabatini, and L.W. Canter, ‘Subsurface Transport and Fate Processes,’ Lewis

Publishers, Boca Raton, FL, Chapter 6, pp. 268–272 (1993).

10. P.H. Nicholls and A.A. Evans, Pestic. Sci., 33, 319 (1991).
11. S.Z. Cohen, S.M. Creeger, R.F. Carsel, and C.G. Enfield, ‘Potential for pesticide contamination

of ground water from agricultural uses,’ in “Treatment and Disposal of Pesticide Wastes,” ed.
R.F. Krueger and J.N. Seiber, ACS Symposium Series No. 259, American Chemical Society,
Washington, DC, Chapter 18, pp. 318–319 (1984).

12. H.H. Cheng, ‘Pesticides in the Soil Environment: Processes, Impacts and Modeling,’ Soil Science

Society of America, Madison, WI (1990).

13. M.A. Locke and C.T. Bryson, Weed Sci., 45, 307 (1997).
14. A.W. Taylor, H.P. Freeman, and W.M. Edwards, J. Agric. Food Chem., 19, 832 (1971).
15. J. Rouchaud, O. Neus, K. Cools, and R. Bulcke, J. Agric. Food Chem., 47, 3872 (1999).
16. X. Jianming, W.C. Koskinen and H.H. Cheng, Pedosphere, 10, 289 (2000).
17. D. Hillel, ‘Environmental Soil Physics,’ Academic Press, New York, Chapters 18 and 19, pp. 507–

587 (1998).

18. A.S. Felsot, R.G. Evans, and J.R. Ruppert, ‘Field studies of imidacloprid distribution following

application to soil through a drip irrigation system’, in “Terrestrial Field Dissipation Studies: De-
sign, Interpretation and Purpose,” ed. E.L. Arthur, V.E. Clay, and A. Barefoot, ACS Symposium
Series No. 842, American Chemical Society, Washington, DC (2003).

19. W.R. Effland, N.C. Thurman, R. Gangaraju, I. Nicholson, and D. Kroetsch, ‘GIS decision sup-

port system to evaluate U.S. and Canada field study areas for agrochemicals,’ in “Pesticide
Environmental Fate, Bridging the Gap Between Laboratory and Field Studies,” ed. W. Phelps,
K. Winton, and W. R. Effland, ACS Symposium Series No. 813, American Chemical Society,
Washington, DC, Chapter 2, pp. 2–21 (2002).

20. R.K. Trubey, R.A. Bethem, and B. Peterson, J. Agric. Food Chem., 46, 2360 (1998).
21. D.L. Suett, Pestic. Sci., 6, 385 (1975).
22. L.E. Bode and M.R. Gebhardt, Weed Sci., 17, 551 (1969).

890

Best practices in the generation and analyses of residues in environmental samples

23. R.D. Wauchope, J.M. Chandler, and K.E. Savage, Weed Sci., 25, 193 (1977).
24. C.G. McWhorter and O.B. Wooten, Weeds, 9, 42 (1961).
25. C.L. Stiles, C.E. Sams, D.K. Robinson, D.L. Coffey, and T.C. Mueller, J. Agric. Food Chem.,

48, 4681 (2000).

26. FIFRA Scientific Advisory Panel Meeting Report, ‘Review of Proposed Revised Guidelines

for Conducting Terrestrial Field Dissipation Studies’, SAP Report No. 90-01, United States
Environmental Protection Agency, Washington, DC, p. 32 (1998).

27. ‘Pesticide Rejection Rate Analysis – Environmental Fate,’ EPA 738-R-93-010, United States

Environmental Protection Agency, Washington, DC (1993).

28. M. Alexander, ‘Biodegradation and Bioremediation,’ second edition, Academic Press, San

Diego, Chapter 6, pp. 99–100 (1999).

29. J.H. Massey and J.S. LeNoir, ‘Sources and magnitudes of variability in terrestrial field dissipation

of agrochemicals,’ in “Terrestrial Field Dissipation Studies: Design, Interpretation and Purpose,”
ed. E.L. Arthur, V.E. Clay, and A. Barefoot, ACS Symposium Series No. 842, American Chemical
Society, Washington, DC (2003).

30. A. Walker and P.A. Brown, Crop Protect., 2, 17 (1983).
31. P.S.C. Rao and R.J. Wagenet, Weed Sci., 33 (Suppl. 2), 18 (1985).
32. L.S. Wood, H.D. Scott, D.B. Marx, and T.L. Lavy, J. Environ. Qual., 16, 251 (1987).
33. J.M. Novak, T.B. Mooreman, and C.A. Cambardella, J. Environ. Qual., 26, 1271 (1997).
34. C. Vischetti, M. Businelli, M. Marini, E. Capri, M. Travisan, A.A.M. Del Re, L. Donnarumma,

E. Conte, and G. Imbroglini, Pestic. Sci., 50,175 (1997).

35. K.A. Gomez and A.A. Gomez, ‘Statistical Procedures for Agricultural Research,’ second edition

Wiley, New York, Chapter 2, pp. 7–83 (1984).

36. M.G. Cline, Soil Sci., 58, 275 (1944).
37. R.G. Petersen and L.D. Calvin, ‘Sampling,’ in “Methods of Soil Analysis, Part I. Physical and

Mineralogical Methods,” Agronomy Monograph No. 9, second edition, American Society for
Agronomy, Madison, WI, Chapter 2, pp. 33–51 (1986).

38. D.C. Montgomery, ‘Design and Analysis of Experiments,’ second edition, Wiley, New York,

Chaper 3, p. 51 (1984).

39. S.A. Clay, D.E. Clay, W.C. Koskinen, and G.L. Malzer, J. Environ. Sci. Health, B27, 125

(1992).

40. D.E. Clay, C.G. Carlson, K. Brix-Davis, J. Oolman, and B. Berg, J. Prod. Agric., 10, 446

(1997).

41. W.P. Anderson, ‘Weed Science Principles’, second edition, West Publishing, St. Paul, MN,

Chapter 8, pp. 341–364 (1983).

42. TeeJet, ‘Agricultural Spray Products, Catalog 47,’ Spraying Systems, Wheaton, IL, p. 136

(1998).

43. M.P. Braverman,, J.S. Corley, D.C. Thompson, M. Arsenovic, V.R. Starner, K.S. Samoil, F.P.

Salzman, D.L. Kunkel, and J.J. Baron. Proc. Weed Sci. Soc. Am., 41, 17 (2001).

44. A. Nordby and R. Skuterud, Weed Res., 14, 385 (1975).
45. J.H. Massey and S.K. Singles, ‘Photostability of two fungicides on spray application monitors,’

in “Terrestrial Field Dissipation Studies: Design, Interpretation and Purpose,” ed. E.L. Arthur,
V.E. Clay, and A. Barefoot, ACS Symposium Series No. 842, American Chemical Society,
Washington, DC (2003).

46. D.G. Graham, V. Clay, S.H. Jackson, and R. Jones, ‘Field dissipation studies – The measurement

of zero-time residues,’ in “Terrestrial Field Dissipation Studies: Design, Interpretation and Pur-
pose,” ed. E.L. Arthur, V.E. Clay, and A. Barefoot, ACS Symposium Series No. 842, American
Chemical Society, Washington, DC (2003).

47. T.H. Dao, T.L. Lavy, and J. Dragun, Residue Rev., 87, 92 (1983).
48. B.J. Weinhold and T.J. Gish, Weed Sci., 39, 423 (1991).
49. B. Richter, J. Ezzell, and D. Felix, ‘Single Laboratory Method Validation Report: Extraction of

Organo-phosphorus Agrochemicals, Chlorinated Herbicides and Polychlorinated Biphenyls Us-
ing Accelerated Solvent Extraction (ASE) with Analytical Validation by GC/NPD and GC/ECD,’
Document 101124, Dionex Sunnyvale, CA (1994).

50. D.R. Knapp, ‘Handbook of Derivatization Reactions,’ Wiley, New York (1979).

Sampling and analysis of soil

891

51. G.W. Snedecor and W.G. Cochran, ‘Statistical Methods,’ Iowa State University Press, Ames, IA

(1987).

52. J.D. Wolt, ‘Soil Solution Chemistry,’ Wiley, New York (1994).
53. J.D. Wolt, H.P. Nelson, Jr, C.B. Cleveland, and I.J. van Wesenbeeck, Rev. Environ. Contam.

Toxicol., 169, 123 (2001).

54. D.L. Sparks, ‘Kinetics of Soil Chemical Processes,’ Academic Press, San Diego (1989).
55. J.W. Hamaker, ‘Decomposition: quantitative aspects,’ in “Organic Chemicals in the Soil Envi-

ronment”, ed. C.A.I. Goring and J.W. Hamaker, Marcel Dekker, New York, pp. 253–340 (1972).

56. G.H. Timme, H. Freshe, and V. Laska, Pflanzenschutz-Nachr. Bayer, 39, 187 (1986).
57. D.I. Gustafson and L. Holden, Environ. Sci. Technol., 24, 1032 (1990).
58. R.G. Allen, L.S. Pereira, D. Raes, and M. Smith, ‘Crop Evapotranspiration – Guidelines for

Computing Crop Water Requirements,’ FAO Irrigation and Drainage Paper 56, Food and Agri-
culture Organization of the United Nations, Rome (1998). Also available on the World Wide
Web: http://www.fao.org/docrep/x0490e/x0490e00.htm.

59. W.J. Rawls, D.L. Brakensiek, and K.E. Saxton, Trans. ASAE, 81, 2510 (1982).
60. A. Bauer and A.L. Black. Soil Sci. Soc. Am. J., 56, 248 (1992).

Document Outline

Front Matter
Table of Contents
Volume I
Volume II
Recent Advances in Analytical Technology, Immunoassay and Other Nonchromatographic Methods
Best Practices in the Generation and Analyses of Residues in Environmental Samples
Best Practices in the Analysis of Residues in Environmental Samples: Groundwater and Soil-Water Monitoring Procedures
Preparation and Instrumental Analysis of Agrochemical Residues in Water Samples
Sampling and Analysis of Soil
Sampling Sediment and Water in Rice Paddy Fields and Adjacent Water Bodies
Monitoring of Agrochemical Residues in Air
Biological Sampling: Determining Routes of Wildlife Exposure to Pesticides
Best Practices in Conducting Dislodgeable Foliar Residue Studies
Best Practices to Conduct Spray Drift Studies
Field Methods for Performing Farm Worker Exposure and Re-Entry Studies
Electronic Record Keeping in a Regulated Environment
Compound Class
Individual Compounds
Index

Wyszukiwarka

Podobne podstrony:
06C Nakazy i zasady w wych ) 12
06c.Na czym polega promocja i profilaktyka zdrowia, Zdrowie publiczne, W. Leśnikowska - Ścigalska -
13 sub 875 1140775365 06c chirurgia urazowa(2)
13 sub 875 1140775365 06c chirurgia urazowa(2)
06c CZWÓRNIK 20091216, Notatki, Obwody i Sygnały
kurswww 06c php m7rofqufu46zrmh Nieznany
91942 08w
91942 abb
91942 08q
91942 04m
91942 05d
91942 03d
91942 01e
91942 08j
91942 05b
91942 08p
91942 04b
91942 08u
91942 04i

więcej podobnych podstron

91942 06c

Document Outline