9.

Soil Testing/Critical Level Determination

SOIL 5813
Soil-Plant Nutrient Cycling and Environmental
Quality
Department of Plant and Soil Sciences
Oklahoma State University
Stillwater, OK 74078
email: [email protected]
Tel: (405) 744-6414
•Assess the relative adequacy of available nutrients (or lime requirements)
•To provide guidance on amounts of fertilizers (or lime) required to obtain
optimum growth conditions for plants (McLean, 1977)
•Diagnose nutrient limitations before a crop is planted so that corrective
measures can be taken.
*Must be fast, reliable and reproducible
Philosophical Use/Interpretation of Soil Testing
1. Base Cation Saturation Ratio
2. Nutrient Maintenance
Disregarding the soil test level, a quantity of nutrient should be added to
replace the amount expected to be removed by the crop. All required
nutrients- not feasible.
3. Nutrient Sufficiency
No yield response to nutrients above a certain soil test level
a. response assured very low
b. response likely low
c. response possible medium
d. response unlikely high
Continuing Problems in Soil Testing

Depth of Sampling
1. 0-6, 0-8, 0-12, inclusion of subsoil (micronutrients)
 
Critical Levels
1. Cate Nelson
2. Mitscherlich
3. Quadratic
4. Square Root
5. Linear-plateau
6. Quadratic-plateau
Economic and Agronomic Impacts of Varied Philosophies of Soil Testing
(Olson et al., 1982)

Field experiments (1973-1980)

5 locations
Irrigated Corn (Zea mays L.)
5 soil testing laboratories
Lab E (University of Nebraska)

No differences in yield
No agronomic basis for ‘base-cation-saturation-ratio‘ or 'maintenance' concepts
Maintenance: whatever the soil test level, a quantity of nutrient should be added
to replace the amount expected to be removed by the crop.
K, S, Zn, Mn, Cu, B, Mg, Fe

This experiment tests other issues:

Private versus Public Sector Research
Need for Watchdog Agencies
Value of Long-Term Research
Need for subsoil data
University of Minnesota
Kansas State University

“I think it is awfully important that dealers and producers use the lab that bases
its recommendations on research from the state in which the dealers operate.”

George Rehm, Univ. of Minnesota

“Part of the problem with university soil testing programs is that they are not
cost competitive with commercial labs…. University labs typically undercut the
commercial operations” Randy Hemb, Minnesota Valley Testing Lab
Fertilizer “Dealers have an obligation to provide a fertilizer recommendation that
will keep that farmer in business. If the farmer goes out of business, the dealer
goes out of business, and 99 percent of the dealers today recognize that.”
George Rehm, Univ. of Minnesota
Cate and Nelson (1965)
% yield versus soil test level
Two Groups:
1. probability of response to added fertilizer is small
2. probability of response to added fertilizer is large
A.Percent yield values obtained for a wide range in locations (fertilizer rate
studies)
Percent yield = yield at 0 level of a nutrient / yield where all factors are adequate
B. Soil test values obtained (Check Plot)
Will generate a single % yield and one soil test value for each location
C. Scatter diagram, % yield (Y axis) versus soil test level (x axis)
Range in Y = 0 to 100%
D. Overlay
-overlay moved to the point where data in the +/+ quadrants are at a maximum
-point where vertical line crosses the x = critical soil test level
 120

100
ld

80
g ie
eY

60
rcenta

40
Pe

20
Critical Level
0
0 20 40 60 80 100 120 140 160
Soil Analysis, ppm P
Critcal level depends on the extraction method used and crop being grown.
Cate-Nelson: Maximizes the computed chi-square value representing the test of
the null hypothesis that the # of observations in each of the four cells (quadrants
is equal).
2. Mitscherlich
3. Quadratic
4. Square Root
5.Linear Plateau: obtaining the smallest pooled residuals over two linear
regressions.
 
Equation MR MER (dy/dx = PR)
________________________________________________________________________________
2. Mitscherlich Log(A-Y) = Log A - C1(x+b) x=log((2.3*A*c)/PR)/c-b
3. Quadratic y = b0 + b1(x) - b2(x2) x=0.5 b1/b2 x=(PR-b1)/(2*b2)
4. Square Root y = bo + b1(x) + b2(sqrt(x)) x=0.25(b2/b1)2 x=(b2/ 2*(PR-b1))2
5. Linear Plateau y = bo + b1(x) when x < joint
y = bo + b1(joint) when x > joint
________________________________________________________________________________
Use of Price Ratios
PR = (price per unit fertilizer) / (price per unit yield)
Optimum rate of fertilizer capable of generating the maximum economic yield is
dependent upon the price of fertilizer, the value of the crop and magnitude of fixed
production costs. The value of a crop defined as a function of yield and rate of
fertilizer can be expressed as:
V = Y * Py = F(x) * Py
where yield (Y) for each fertilizer rate is multiplied by the crop price (Py) per unit of
yield. A line describing fertilizer costs per unit area cultivated can be expressed as
a function of fixed costs (F) and fertilizer price (Px) times the amount of fertilizer
(X)
T = F + Px * X
where total cost (T) is a linear function of fertilizer amount, the slope of the line is
given by the price of fertilizer and the intercept by the amount of fixed costs
involved (F).
A plot of the value and cost functions illustrates the areas where use of fertilizer is profitable.
Net profit can only be generated by use of a fertilizer amount equal or greater than 0-x1.
Fertilizer should not be used if the value curve is lower throughout than the total cost curve for
fertilizer plus fixed costs (F). With fixed costs involved, the amount of fertilizer that can be
used profitably is greater than zero or an amount equal to or greater than 0-x1. For fertilizer
input greater than 0-x1, crop value exceeds costs and net profit is generated. Profit from
fertilizer application can be increased until input reaches the value of 0-x2. This is the level
which maximizes profit. At 0-x2 the difference between value and cost is at a maximum.
For each production function the amount of fertilizer which maximizes profit can be found by
obtaining the first derivative and setting it equal to the price ratio (PR).
PR = Price per unit of fertilizer / Price per unit of yield (from Barreto and Westerman, 1985)

Crop Value

Profit
Cost

0 x1 x2
Soil Testing for Different Nutrients
 
Total Nitrogen in Soils:
Surface soils: 0.05 to 0.10%
precision 0.01% = +/- 200 lb/ac

Why would we run total N on soils if the precision is so low?

 
        long term experiments (differences greater than 200 lb N/ac)
        C:N relationships at the same level of precision
A. Kjeldahl 1883 (organic + inorganic N)
-digestion to convert organic N to NH4
-determination of NH4 in the digest
(N pool consists of NO3-, NH4+, NO2-, organic N)
devardas: reducing agent, that is a finely powdered mixture of metals that act
as a source of donor electrons to reduce NO3- and NO2- to ammonium
devardas
N pool + K2SO4, CuSO4, Se, H2SO4 -----> (NH4)2SO4
Digest
(NH4)2SO4 + NaOH ----> NH3 + NaSO4 (catch in boric acid)
titrate
K2SO4 is used to raise the temperature of the digest (increases speed and
completeness of the conversion of organic N to NH4)
Se, Cu are used as catalysts to promote the oxidation of the organic matter
NO3 and NO2 are not included in the total N analysis from dry combustion, but
it does not matter since there will be less than 20 lb N /ac as NO3 and the total
N procedure detects to only +/- 200 lbs N/ac
e.g.
0.01 +/- 200 lbs/ac 20 lbs N/ac as NO3 is lost between 0.01
and 0.02 %total N
0.02 +/- 400 lbs/ac because its small value exceeded the
detection limits.
On a KCl extract: (have both NH4 and NO3 in the extract)
-distill over once (to collect NH4)
-add devardas alloy (distill over again to collect NO3 and NO2)
devardas alloy: acts as a source of donor electrons to reduce NO2 and NO3 to
NH4
-problems: N-N and N-O compounds
Dry Combustion (Dumas 1831)
-Sample heated with CuO at high temp (above 600 °C) in a stream of CO2
-Gasses lost are passed over hot Cu to reduce nitrogen to N2
-Next over CuO to convert CO to CO2.
-N2-CO2 mixture is collected in a nitrometer containing concentrated alkali
which absorbs CO2 and the volume of N2 gas is measured.
2NH4Cl + 4CuO -----> N2 + 4H2O + 2CuCl + 2 Cu (CO2)
problems: heterocyclic compounds (pyridine) are difficult to burn
NA-1500
Sample weighed in a tin (Sn) container
Combustion reactor enriched with pure oxygen (sample oxidation) 1020 °C in
combustion tube
Reaches 1700 °C during flash combustion (complete oxidation)
Flash combustion converts all organic and inorganic substances into
elemental gases (stable compounds combusted)
Combustion products carried by He pass through an oxidation catalyst of
chromium oxide
 Combustion Reactor Reduction Reactor

Cr2O3 Cu

CO + 1/2O2 = CO2 (Cr2O3 is accepting electrons)

Cr2O3 ensures complete combustion (oxidation) of all organic materials

NOx N2 (Cu is donating electrons)

Combustion products (CO, N, NO) and water pass through a reduction

reactor (metallic Cu)
Excess O2 is removed in the reduction reactor (Cu at 650 C)
N oxides from the combustion tube are reduced to elemental N2
Taking CO, N, NOx and converting them to CO2, N2
Gases are separated in a chromatographic column and detected using a
thermal conductivity detector (TCD) which gives an output signal proportional
Inorganic Nitrogen
NO3-N
Inorganic N may represent only a small fraction < 2% of the total N in
soils (Bremner, 1965)
Nitrate testing does not work in Illinois. Why?
high OM
high mineralization potential
consideration of NH4
R-NH2 groups from N cycle
rapid changes (biological transformations) affect inorganic N analysis
NO3-N and NO2-N
1. Phenoldisulfonic acid or chromotrophic acid
interference of organic matter, Cl and Fe have affected these colorimetric
procedures
2. Selective ion electrodes
interference of Cl
(NH4)2SO4, AgSO4 extracting solution: Ag used to precipitate Cl
3. Cadmium reduction
2 M KCl extract (colorimetric procedure) - samples are stable for several
months if stored at low temperatures
not subject to interference, extremely sensitive making dilution possible.
NO3 reduced to NO2 by passing through a column of copperized Cd
NO2 reacts well with the diazotizing reagent (sulfanilamide) and NO3 does
not, thus explaining the need for reducing NO3 to NO2 for analysis using the
Griess-Ilosvay method
4. Steam distillation with Devardas alloy (reductant) reduce NO2
and NO3 to NH4
NH4-N
Bremner (1959) Soils contain a large amount of fixed (non-exchangeable NH 4). Defined
as the NH4 that cannot be replaced by a neutral K salt solution present as NH 4 ions in
interlayer positions of 2:1 type clay minerals
Air drying can lead to small but significant changes in NH 4-N
-Steam distillation with MgO (alkaline reagent) color: indophenol blue
-2 M KCl (indophenol blue) phenol and NH3 react to form an intense blue color
-Ammonia gas sensing electrodes
Problems in N analysis:
-accuracy is measured by the least precise measurement.
-weight of the soil is the largest error (propagates through to +/- 0.01%N
0.01% N = +/- 100 ppm (0.01* 10000)
total N in soils 0.10 = 1000 ppm +/- 100 ppm
inorganic N in soils 0.002 = 20 ppm +/- 1 ppm
Total N Inorganic N Organic N?
1000 ppm 20 ppm 980 ppm
-Inorganic N not determined on percent basis, done on an aliquot basis
-Cannot subtract 20 from 1000 to get organic N (determined on a different basis)
-Unrealistic because of the incompatibility of error terms
-Organic-N is difficult to determine (by subtraction, we have an extremely poor estimate)
Organic N
Procedures exist, but are unreliable and are not reproducible

Mineralizable N
1.Leach with CaCl2 - dissolves all the soluble N (NO3 and NO2)
2.Incubate the soil - over time - to determine the amount of NO3 that has been
mineralized (set period of time under set conditions)
3.Leach with CaCl2 again (sample now has NO3)
4.Determine concentration
Phosphorus Soil Index Procedures

Bray and Kurtz P-1

0.025 N HCl and 0.03N NH4F (pH = 3.15)
 
Removes easily acid soluble forms of P, (Ca-P & Fe, Al-P)
NH4F dissolves Al and Fe-P by its complex ion formation with these metal ions in
acid solution. This method has proved to be very successful in acid soils.
 
In view of the high efficiency of the fluoride ion in dissolving phosphate, Bray (1945)
recommended the use of this reagent together with HCl as an extractant
(effectively removed sorbed phosphate)
 
Al reacts with F and inactivates Al leaving P in solution. Use of NH4F will increase
extractable P, or stabilize P (restricting Al from precipitating with P because of the
solubility constants)
Mehlich II
0.20 NH4Cl, 0.2N CH3COOH, 0.015N NH4F and 0.012N HCl
(pH = 2.5)
The concentrations of HCl and NH4F used in Mehlich are half that used in Bray
and Kurtz P-1. However this extracting solution also contains NH4Cl and
acetic acid which probably buffer the solution (i.e., keeps its acidic strength for
a longer period of time). Therefore, it can dissolve more of the P in apatite.
Mehlich III (more acid than Bray) short shaking time (filtering time is
problematic) 11%CCE will neutralize Mehlich – making the extracting
solution water
0.2N CH3COOH, 0.015N NH4F, 0.25N NH4NO3, 0.13N HNO3, 0.001M EDTA
(pH = 2.4)
Designed to be applicable across a wide range of soil properties ranging in
reaction from acid to basic. Can also be used for exchangeable cations (Ca
and Mg). Because this extractant is so acid, there is some concern that the
soil can be dissolved, increasing exchangeable amounts.
Olsen
0.5N NaHCO3 (pH = 8.5)
This extracting solution is used to extract phosphorus in calcareous soils. It
will theoretically extract the phosphorus available to plants in high pH soils.
This extractant decreases the concentration of Ca in solution by causing
precipitation of Ca as CaCO3; as a result, the concentration of P in solution
increases.
CaHPO4 = Ca+2 + HPO4=
HCO3-
CaCO3
Essentially, increase the activity of CO3 in solution which reacts with Ca, and
CaCO3 precipitates.
Nelson et al. (1953) (Mehlich I and or "Double Acid")
0.05N HCl and 0.025N H2SO4 (pH<2.0)
Found to be effective in high P-fixing soils of North Carolina. H2SO4 was
found to be more effective than HCl in dissolving Fe phosphates but that both
were equal regarding Al phosphates.
Extractable P discussion:

The pH of the extracting solution is an indicator of what forms of P will be

extracted. However, this should be used with caution as the shaking time is
important in terms of reaching an equilibrium.
Susuki et al. (1963) noted that 0.1N HCl extractable P was positively
correlated with Ca-P.
NaHCO3 was negatively correlated with Ca-P on 17 Michigan soils (pH 4.8-
7.8)
What would happen if Bray P-1 was used on a calcareous soil?
The lime in the calcareous soil would neutralize the acidity in the extracting
solution thus decreasing its ability to extract the Fe and Al-P forms which
would be available at that soils pH.
Calibrations for the Bray-Kurtz P-1, Mehlich III and Olsen soil tests (Tisdale, Nelson, Beaton
and Havlin, 1993)
____________________________________________________________________________

P sufficiency level Bray-Kurtz P-1 Mehlich III Olsen Fertilizer P

Recommendation
____________________________________________________________________________
lb P2O5/ac kg P/ha

Very low <5 <7 <3 50 25

Low 6-12 8-14 4-7 30 15
Medium 13-25 15-28 8-11 15 8
High >25 >28 >12 0 0
____________________________________________________________________________
Total P ?
Analysis for total P in soils abandoned in the early 1900's
Scientists recognized that total P was not correlated with plant
availability.
Various strengths of extracting solutions were evaluated for specific
soils at selected soil pH that mirrored what the plant would find in soil
solution.
All of these are indices that determine orthophosphate
concentrations (from the dissolution of precipitated forms).
Attempts to correlate extractable P (x - procedure) with total P will
result in meaningless information.
Total P (strong acid digest) will in essence dissolve P forms that will
not be available at that soils specific pH.
Nutrient Interactions

Bray and Nye:

K applications on soils with high K by mass action displace Al+++ which
complexes with P inducing a net P deficiency (pH < 6.0)

P and Zinc
Zinc deficiencies attributed to the immobilization of zinc owing to the
increase in the concentrations of P in the roots above the threshold values.

Depression of zinc concentrations in plant tissue by P (interaction

occurred in the plant and not in the soil).
Source of N by P
NO3- uptake (increase pH)
NH4+ uptake (decrease pH)

1.0
Fraction of H3PO4 species

0.8
NO3-

0.6
HPO4=
H2PO4-
0.4 NH4+

0.2

0.0
4 5 6 7 8 9 10 pH
Optimum pH for species uptake
Increased uptake
Effect on rhizosphere pH
Spectroscopy

Light is considered to be a stream of particles. The discrete particles or units of

energy are called photons or quanta. A photon of blue light contains much
more energy than a photon of red light.
Interaction of light with matter
1nm = 1mu (millimicron) = 10A (angstrom) = 10-7 cm
The interaction of radiation with matter may result in the absorption of incident
radiation, emission of fluorescence or phosphorescence, scattering into new
directions, rotation of the plane of polarization, or other changes. Each of these
interactions can provide useful information about the nature of the same in
which they occur (Tinoco et al. 1978).
Color is characteristic of the spectrum (in the visible region) of light transmitted
by the substance when white light (or sunlight) shines through it, or when light
is reflected from it.
<0.01 Gamma (non particulate photons)
0.01-10 X-Ray (photons)
10-380 Ultraviolet

Wavelength absorbed, nm Absorbed Color Transmitted Color (Complement)

380-450 Violet Yellow-green

450-495 Blue Yellow
495-570 Green Violet
570-590 Yellow Blue
590-620 Orange Green-blue
620-750 Red Blue-green
________________________________________________________________

750-1x106 Infrared
1x106-1x1011 Micro and short radio waves
>1x1011 Radio, FM TV
________________________________________________________________
Wavelength: distance of one complete cycle
Frequency: the number of cycles passing a fixed point per unit time
Wavelength

= c/v
= wavelength in cm
v = frequency in sec-1 or hertz (Hz)
c = velocity of light in a vacuum (3x1010 cm/sec)
Electromagnetic radiation possesses a certain amount of energy. The energy of a unit
of radiation, called the photon is related to the frequency by E = hv = hc/
where E is the energy of the photon in ergs
h is Planck’s constant 6.62 x 10-27 erg-sec
The shorter the wavelength or the greater the frequency, the greater the energy. Energy
of a single photon (E) is proportional to its frequency (v) or inversely proportional to its
wavelength.
If a molecule absorbs radiation, it is raised to a higher energy level, with the
increase in energy being equal to the energy of the absorbed radiation (h
The relative energy levels of the three transition processes are in the order
electronic > vibrational > rotational
If the electromagnetic force results in a change in the arrangement of the
electrons in a molecule, we say that a transition to a new electronic state has
occurred. The absorbed photon results in the excitation of the molecule from
its normal or ground state, G, to a higher energy or excited electronic state, E.
The excited electronic state has a rearranged electron distribution.
When considering absorbing substances that are either liquids, solids or
gases, each will have a characteristic transmission of light. Suppose that light
of intensity Io is incident from the left, propagates along the x direction and
exits from the right with decreased intensity It. At any point x within the
sample, it has intensity I, which will decrease smoothly from left to right.
 dx

I0 I I-dI It

x
l

If the sample is homogeneous, the fractional decrease in the light intensity

is the same across a small interval dx, regardless of the value of x. The
decrease for a solution depends linearly on the concentration of the
absorbing substance.

Not all molecules can absorb in the infrared region

The wavelength of absorption is a measure of the energy required for the
transition
Each molecule will have a complete absorption spectrum unique to that
molecule, so a 'fingerprint' of the molecule is obtained
 Absorption of Visible Light SPAD 501, 502
by Photopigments (430, 750)

Sunlight reaching
earth Phycoerythrin

Chlorophyll b
Absorption

Phycocyanin

B-Carotene

Chlorophyll a

300 400 500 600 700 800

Wavelength, nm Lehninger, Nelson and Co

 Short wavelength Long wavelength
High energy Low energy

Phycoerythrin
Chlorophyll b
Phycocyanin

B-Carotene

Chlorophyll a
Ultraviolet

Infrared
X-Rays

0.01 10 380 450 495 570 590 620 750

wavelength, nm
 White Light

Interference Filter
Photodiode
Phycocyanin

Chlorophyll b

B-Carotene
Phycoerythrin
Chlorophyll a

380 450 495 570 590 620 750

wavelength, nm
 Near-Infrared Absorption
Major Amino and Methyl Analytical Bands
and Peak Positions

CH3 CH3 CH3 CH3 CH3 CH3

RNH2 RNH2 RNH2 RNH2

| | | | | | | | | | | | | | | |

700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200

Wavelength, nm
Mean surface soil test P and fertilizer P recommendations,
bemudagrass pasture, Burneyville, OK

70 30
Mehlich III
60

P FERTILIZER, kg/ha
P Fertilizer 25
MEHLICH III P, mg/kg

50
20
40
15
30
10
20

10 5

0 0
0 2 4 6 8 10 12 14 16 18 20 22
DISTANCE, m
160 acre field ($10/soil sample)
Sensor Based 9 ft2
$7,744,000 100
Map Based
$77,440
900 ft2 x 4840
48.4

1 Acre 43560 ft2

Grid Based
$1600
Soil Testing versus Non-destructive Sensor Based VRT
Soil Testing Sensor Based VRT
low resolution high resolution
Chemistry-Site specific Site specific
Reliable and tested untested
Years of correlation/calibration new technology
Economical high potential of being economical
Crop specific untested
Variety specific untested
Management specific untested
row spacing/tillage
Nutrient interactions untested
NA weed recognition
NA time of day
NA shadow/clouds
NA direction of travel
 Bray’s Mobility Concept
Mobile Nutrients Immobile Nutrients
•Depends on Growth Stage •Independent of Growth Stage

•Indicators •Indicators
•Biomass •Biomass
•Color •Color
•Concentration

Sensor
NO3- H2PO4- HPO4=
Soil Test

Root System Sorption Zone Root Surface Sorption Zone

• Responsedepends on reservoir • Response does not depend on reservoir

• Dependent on environment • Independent of environment

Root system and root sorption zones as related to mobile and immobile nutrients
and the use of soil and sensor tests.
 Evolution of a soil test

Extraction Correlation Calibration On-farm Sufficiency

Procedure with field of rates at validation Index
response a given soil
test level

Evolution of a spectral test

Wavelength Correlation Calibration On-farm Sufficiency

/Index/band with field of rates at validation Index
response a given (index)
Prediction of YIELD POTENTIAL
Experimental Design/Soil Testing and Field Variability
Replication gradients: Do slopes (up and down or side to side) in fields
adequately represent which direction a particular nutrient will increase or
decrease? Are Blocks actually needed?

Number of Replications: If plot size remains large and greater than the field
element size, increasing the number of replications will unlikely lead to
increased power for detecting differences between treatments

Plot Size: Because field variability has been demonstrated to be somewhere

around 9 square feet, field experiments as we now know them must change.
Common plot sizes are between 250 to 1000 square feet. Plant breeders have
generally employed much smaller plot sizes and because of this, CV's from their
work are generally smaller than that found in fertility/weed type trials.
50 m

30 m
60 feet

150 feet