Mariusz Fotyma – Redaktor naczelny – Executive Editor Janusz Igras, Przemysław Tkaczyk
Monography
Soil testing methods and fertilizer recommendations in Central–Eastern European countries
Methoden der Bodenuntersuchung und Düngeempfehlung der mittel- und osteuropäischen Staaten
Edited by Mariusz Fotyma, Eike Stefan Dobers – Poland Erstellt durch Mariusz Fotyma und Eike Stefan Dobers (Polen)
Co-authors, Koautoren *
Gerhard Breitschuh – Germany, Valli Loide – Estonia, Regina Timbare – Latvia, Gediminas Staugaitis – Lithuania, Heide Spiegel – Austria, Dorota Pikuła – Poland,
Frantisek Kotvas – Slovak Republic, Barbara Ceh – Slovenia, Pavel Cermak – Czech Republic, Jakab Loch – Hungary
* full names and the addresses of the affiliation units are given in Appendix 8
*die Namen und Adresse von alle zusamenarbeitene Anstalten sind im Appendix 8 hergestellt
Zeszyt wydany w ramach realizacji programu wieloletniego IUNG-PIB Zadanie 1.7
Copyright by Polish Fertilizer Society – CIEC
ISSN 1509 – 8095
Adres Redakcji – Adress Executive Editor Zakład Żywienia Roslin i Nawozenia IUNG-PIB
Czartoryskich 8, 24-100 Pulawy fot@iung.pulawy.pl
Druk: IUNG-PIB zam.24/F/08 nakł. 200 egz. B-5
Table of Contents
1. COMPARISON OF METHODS FOR SOIL ANALYSIS ... 7
1.1 DESCRIPTION OF THE METHODS ... 7
1.2 QUANTITATIVE ANALYSIS OF SOIL DATE. ... 8
1.3 CALIBRATION OF SOILS TESTS RESULTS FOR AVAILABLE P, K, MG 11 1.4 QUALITATIVE ANALYSIS OF SOIL DATA ... 13
1.5 SOIL TEXTURE ... 16
2. COMPARISONS OF RECOMMENDED FERTILIZER RATES... 17
2.1 RECOMMENDED RATES OF NUTRIENTS FOR GRAIN CROPS ... 17
2.1.1 Phosphorus ... 17
2.1.2 Potassium K ... 19
2.1.3 Magnesium Mg ... 21
2.2 RECOMMENDED RATES OF NUTRIENTS FOR POTATO AND SILAGE MAIZE ... 23
3. ALGORITHMS OF FERTILIZER RECOMMENDATIONS ... 26
3.1 GENERAL REMARKS ... 26
3.2 ALGORITHMS OF FERTILIZER RECOMMENDATIONS IN COLLABORATING COUNTRIES ... 26
3.2.1 Estonia EE ... 26
3.2.2 Latvia LV ... 27
3.2.3 Lithuania LT ... 27
3.2.4 Austria AT ... 28
3.2.5 Poland PL ... 28
3.2.6 Slovenia SL... 29
3.2.7 Czech Republic CZ ... 29
3.2.8 Hungary HU ... 29
4. DISCUSSION AND CONCLUSIONS. ... 30
4.1 SOIL TESTS ... 30
4.1.1 Comparison of extraction methods ... 30
4.1.2 Comparison of calibration schemes. ...33
4.2 FERTILIZER RECOMMENDATIONS. ...35
4.2.1 Recommendations schemes (algorithms) ...35 Fertilizers and Fertilization
Nr 30/2008
4.3 CONCLUSIONS ... 38
REFERENCES. ... 39
APPENDICES ... 41
1. VERGLEICH DER METHODEN DER BODENUNTERSUCHUNG...50
1.1 BeschreiBung der Methoden ... 50
1.2 Quantitative analyse der untersuchungsergeBnisse ... 51
1.3 einstufung der ergeBnisse der Bodenuntersuchungen für verfügBares P, K und Mg .54 1.4.Qualitative analyse der Bodendaten ... 56
1.5 Bodenarten-ansPrache ... 59
2. VERGLEICH DER DüNGUNGSEMPFEHLUNGEN... 60
2.1 düngungseMPfehlung für Körnerfrüchte ... 60
2.1.1 Phosphor ... 60
2.1.2 Kalium ... 63
2.1.3 Magnesium ... 65
2.2 düngungseMPfehlung für Kartoffel und siloMais ... 67
3. ALGORITHMEN FüR DIE ERMITTLUNG DES DüNGEBEDARFS.... 70
3.1 allgeMeine BeMerKungen ... 70
3.2 BeschreiBung der algorithMen für die jeweiligen staaten ... 70
3.2.1 Estland (EE) ... 70
3.2.2 Lettland (LV) ... 71
3.2.3 Litauen (LT) ... 71
3.2.4 Österreich (AT) ... 72
3.2.5 Polen (PL) ... 73
3.2.6 Slowenien (SL) ... 73
3.2.7 Tschechische Republik (CZ) ... 73
3.2.8 Ungarn (HU) ... 74
4. DISkUSSION UND SCHLUSSFOLGERUNGEN... 75
4.1 Bodenuntersuchung ... 75
4.1.1 Vergleich der Extraktionsmethoden ... 75
4.1.2 Ermittlung der Nährstoffversorgung ... 78
4.2 dünge-eMPfehlungen... 80
4.2.1 Algorithmen zur Ableitung der Dünge-Empfehlungen ... 80
4.3 schlussfolgerungen ... 83
5. LITERATUR... 85
6. ANHANG ... 86
1. SUPPLEMENT ... 94
Pikuła D., Tkaczyk P. Comparison of aerometric and laser diffraction methods for soil texture analysis... 94
1.1 Methods ... 95
1.2 Results ... 96
1.3 Discussion ... 99
1.4 Conclusions ... 99
1.5 Literature ...99
1.6 Annexes ...101
1. ERGÄNZUNG ...102
Pikuła,D. Dobers, S. Vergleich der Sedimentations- und der Laser-Difraktions- methodik für die Bestimmung der Bodentextur...102
1.1 Methoden ...104
1.2 Ergebnisse ...104
1.3 Diskussion ...107
1.4 Schlussfolgerungen ...107
1.5 Literatur ...108
1.6 Anhang ...109
In course of the 9th meeting of the MOEL group (MOEL working group of agrochemical services in Central-Eastern Europe – Arbeitsgruppe der agrochemischer Untersuchunsdienste der mittel- und osteuropeischen Lander) in 2006 at Piran – Slovenia, it has been decided to organize the inter-laboratory exchange of soil samples with the following fertilizer recommendations between 10 collaborating countries. In each country two representative soil samples were collected, analysed for the basic properties (texture, humus content) and distributed among all participants with the task for estimating the soil’s pH and the content of available phosphorus, potassium and magnesium, using methods officially accepted in the country. Further, six crops typical for most countries e.g. winter wheat, winter rape, winter rye, grain maize, potato and silage maize have been selected with presumed yields 7 t grain, 4.5 t seeds, 6 t grain, 10 t seed, 30 t tubers and 10 t dry matter – respectively. For potato 30 t FYM and for silage maize 30 m3 slurry were to be applied. All the crops were to be grown on both “sites” where soil samples have been collected. On the base of soil indices and crop characteristics, fertilizer recommendations were launched supported by the system officially used in the country. The data concerning soil properties and recommended rates of fertilizers were collected in Excel sheets and processed by means of analysis of variance and regression using the statistical package Statgraphic 5a.
Von den Redakteuren
Während des 9. Treffens der MOEL-Gruppe (Vereinheitlichung der landwirtschaftlich-chemischen Untersuchungsdienste der mittel- und osteuropäischen Länder) in Piran (Slowenien) im Jahre 2006 wurde beschlossen, einen Ringversuch zwischen den 10 beteiligten Staaten zu organisiseren. Dieser Ringversuch hatte zum Ziel, Düngungsempfehlungen abzuleiten.
In jedem teilnehmenden Staat wurden zwei repräsentative Bodenproben gezogen, für diese die Bodenart und der Humusgehalt ermittelt, und dann an alle Teilnehmer des Ringversuchs versandt. Die Teilnehmer des Ringversuchs hatten die Aufgabe, an den Bodenproben unter Verwendung der jeweils landestypischen Methoden den pH- Wert und den Gehalt an pflanzenverfügbarem Phosphor, Kalium und Magnesium zu bestimmen. Man einigte sich ebenfalls auf die folgenden, für die beteiligten Staaten typische Fruchtarten und Ertragsziele: Winterweizen (7 t*ha-1 Korn), Winterraps (4.5 t*ha-1 Samen), Winterroggen (6 t*ha-1 Korn), Körnermais (10 t*ha-1 Korn), Kartoffel (30 t*ha-1 Knollen) und Silomais (10 t*ha-1 Trockenmasse). Für den Anbau von Kartoffel wurde die Verwendung von 30 t*ha-1 Stallmist, und für den Silomais-Anbau die Ausbringung von 30 m³ Gülle. Die Düngebedarfsermittlung fand unter der Annahme statt, dass alle Fruchtarten auf Standorten angebaut würden, die den gezogenen Bodenproben des jeweiligen Staates entsprachen. Auf der Grundlage der Analyseergebnisse und Eigenschaften der angebauten Pflanzen wurden Düngebedarfsermittlungen durchgeführt auf der Basis des jeweils landestypischen Verfahrens.
Die Ergebnisse der Bodenuntersuchung und die empfohlenen Düngermengen wurden in Excel-Tabellenblättern gesammelt und unter Verwendung des Statistik- Programms Statgraphic 5a der Varianz- und Regressionsanalyse unterzogen.
.
1. Comparison of methods for soil analysis 1.1 Description of the methods
In the participating 11 collaborating countries a total of two methods for measurement of pH, four methods for P and K, and five methods for Mg are officially used (table 1).
Table 1. Listing of methods used in participating countries
Country and laboratory Code* pH P and K Mg
Germany, Thuringen,
Deutschland DE TH CaCl2 CAL CaCl2
Germany, Hessien,
Deutschland DE HE CaCl2 CAL CaCl2
Estonia, Estland EE KCl ME-3** ME-3**
Latvia, Lettland LV KCl DL DL
Lithuania, Lithauen LT KCl AL. AL.
Austria, Österreich AT CaCl2 CAL CaCl2
Poland, Polen PL KCl DL CaCl2
Slovak Republic, Slovak SK CaCl2 ME-3 ME-3
Slovenia, Slovenien SL KCl AL CaCl2
Czech Republic, Tschechien CZ CaCl2 ME-3 ME-3
Hungary, Ungarn HU KCl AL KCl
* these codes will be used in the whole report
**until 2002, DL method for P, K, and AL method for magnesium has been in use
The concentration of hydrogen ions (pH) is measured in either “weak” CaCl2 or, as so-called exchangeable acidity in “stronger” KCl solvent. Phosphorus and potassium are measured in a combined (group) extract of buffered calcium lactate (CAL, DL, AL) or in Mehlich solvent. The last method is used to measure the content of available magnesium as well. However, the most often used solvent for Mg measurement is the weak CaCl2 solution. Few countries only are using AL, KCl or DL methods for magnesium estimation. A brief description of chemical details for the different methods for soil analysis is presented in table 2.
Fertilizers and Fertilization Nr 30/2008
Table 2. Description of soil analysis methods
pH P and K Mg*
Code Description Code Description Code Description CaCl2 0.01 mol*dm-3
CaCl2 DL Egner-DL, calcium
lactate, HCl, pH 3.5 CaCl2 Schachtschabel, 0.0125 mol*dm-3 CaCl2 KCl 1 mol*dm-3
KCl AL Egner-Riehm-Domingo, ammonium lactate, acetic acid, pH 3.5
CAL
CAL (Schüller, 1969):
calcium acetate, calcium lactate, acetic acid, pH
KCl 1 mol*dm-3 KCl
ME-3 Mehlich 3, 0.2 mol/l CH3COOH, 0.015 mol/l NH4F, 0.013 mol/l HNO3, 0.25 mol/l NH4NO3, 0.001 mol/l EDTA ,pH
1.2 Quantitative analysis of soil date
Analysis of the variance was performed for each measured nutrient using classical MANOVA with two factors (1) country providing the sample and (2) laboratory for analysis (in fact in the same countries). Individual laboratories have used different methods for analysis (see table 1), and therefore the laboratory is practically the synonym for the method. For almost all nutrients, the interaction between the two factors was insignificant, and hence the differences between levels of the first (countries) and second (laboratories, i.e. methods) factor are interpreted only. With regard to the project’s objectives, the differences between methods are of paramount importance, because differences between the countries were to be expected. Table 3 shows the average values of the soil characteristics for different methods of analysis performed by the laboratories, with the number of laboratories using the same method included in brackets.
Table 3. Average values of the soils characteristics
pH P mg*kg-1 soil K mg*kg-1 soil Mg mg*kg-1 soil method average method average method average method average
CaCl2
(5)* 6.50 a CAL (3) 66.5 a** CAL
(3) 121 a CaCl2 (4) 121 a KCl (6) 6.33 b DL ( 2) 68.2 a DL (2) 136 b KCl ( 1 ) 123 a ME-3 (3) 88.3 b AL (3) 167 c ME-3 ( 3) 213 b AL (3) 105 c ME-3 (3) 183 d DL (1) 271 c AL (2) 576 d
* number of laboratories employing the same method
**averages marked by the same letter do not differ significantly
The average value of soil pH measured in KCl is slightly lower than in CaCl2. However, the relation between the two measures is approximated best by a double reciprocal regression function of the form:
pHCaCl2 = 1/(0.0375 + 0.735/pHKCl), R = 98 (1)
This means that in acid soils pHKCl values are by 0.5 – 0.6 units lower than pHCaCl2, for neutral soils, both measures are similar, and for alkaline soils values determined with pHKCl are slightly higher as those with pHCaCl2 (Fig 1).
However such relation concerns only the date from this research and should not be generalized. Potassium ion from KCl has higher exchange power then calcium ion from CaCl2 and replace more hydrogen ions from soil solid phase. Therefore the pH measured in KCl solution is regularly lower then in CaCl2 though this relation is not necessarily proportional.
Fig. 1. Relation between pH in CaCl2 and in KCl
The average values for phosphorus and potassium extracted with CAL and DL methods are practically the same, which means that both methods can be used likewise. The AL and ME-3 methods belong to a distinctly different group, because higher quantities of phosphorus and potassium are extracted than using CAL and DL methods. The differences between average values determined with AL and ME- 3 methods are statistically significant, but they are not too big and anyhow much lower than the differences between the two groups of methods (CAL & DL versus AL & ME-3). The relation between values determined using methods from these two groups can be approximated by the linear functions:
for phosphorus: PME-3/AL= 29.79 + 0.96CAL/DL, R2 = 54 (2) for potassium: KME-3/AL= 30.17 + 1.12CAL/DL, R2 = 70 (3)
The slopes of both functions are close to one, and intercepts are around 30 mg P and/or K per 1 kg of soil. From these relations a conclusion can be drawn that the results obtained by more “aggressive” solvents as e.g. ME-3 and AL differ by constant values from those obtained by using CAL and/or DL methods. However, one must take into consideration that the determination coefficients for both nutrients, particularly for phosphorus, are not too high, and these relations are strongly sample- dependent (variable).
The lowest amounts of magnesium are extracted by unbuffered salts e.g. CaCl2 and KCl. The ME-3 method results in magnesium values that are nearly 100% higher.
However, the advantage of this method is the possibility to extract simultaneously phosphorus and potassium as well. The AL method, employed by two laboratories, seems to be unsuitable for magnesium estimation due to a very high extraction power exceeding five-fold the extraction power of CaCl2. Besides the only one significant interaction found between laboratories and countries was in the case of magnesium.
A more detailed analysis of the interaction showed that the content of magnesium in soil samples from AT estimated by means of the AL method was much higher in comparison to other methods. With regards to a comparison of methods, the relation between CaCl2 and ME-3 methods has been examined. This relation was best approximated by the following equation:
MgME-3 = 30.49 + 1.467 MgCaCl2, R2 = 58 (results from Slovenia were excluded) (4) The slope of this function is much higher than one, which casts doubts on the possibility of its generalization. Besides, the determination coefficient is not too high.
The main aim of this project is to compare the approaches to soil testing and deriving the recommendations for fertilization. However, it can be interpreted as a ring test for laboratories of modelling the intra- and inter-laboratory variation, as well. Obviously, such an evaluation can be performed only for the same method used by different laboratories. In table 4, the statistical evaluation of the results produced by at least three laboratories is shown.
From table 4, the conclusion can be drawn that laboratories that use the same method of analysis, produce statistically uniform results. This proves the reliability of both the methods and laboratories. Only the data for potassium and magnesium from the SL laboratory differ significantly from the average for laboratories performing the AL method.
Table 4. Statistical evaluation of the inter – lab variation Labora-
tory average
values pH average values
P mg*kg-1 soil average values
K mg*kg-1 soil average values Mg mg*kg-1 soil CaCl2 KCl CAL DL ME-3 AL CAL DL ME-3 AL CaCl2 ME-
3 AL
DE TH 6.51 - 62.2 - - - 124 - - - 118 -
DE HE 6.54 - 80.2 - - - 114 - - - 118 -
EE - 6.28 - - 92.2 - - - 179 - - 242
LV - 6.28 - 64.4 - - - 146 - - - -
LT - 6.31 - - - 96.1 - - - 177 - - 653
AT 6.55 - 57.0 - - - 127 - - - 136 -
PL - 6.40 - 72.0 - - - 127 - - 113 -
SK 6.39 - - - 83.1 - - - 189 - - 172
SL - 6.40 - - - 110 - - - 139* - - 499
CZ 6.49 - - - 89.6 - - - 179 - - 225
HU - 6.32 - - - 110 - - - 184 - -
Mean 6.50 6.33 66.5 68.2 88.3 105 121 136 183 167 121 213 576
* LSD test shows significant difference
1.3 Calibration of soils tests results for available P, k, Mg
The usability of soil test results for fertilization recommendations depends on its calibration against plant indices or against another, already well-calibrated measure.
Theoretically the calibration should be as simple as possible but as complicated as necessary. In the most simple and simultaneously most universal approach the content of the nutrient is taken into consideration only (marked further as Q). In more complicated approaches, besides the nutrient content other soil parameters (pH, texture) are included as well (marked further as QS). The most complex calibration systems take into consideration the nutrient content, soil properties and the type of crop grown (marked further as QSC). A very rough classification of different calibration systems in the collaborating countries is presented in table 5.
For all nutrients and countries, the content of nutrient forms the base for the calibration. In all countries, except Lithuania, five levels of available nutrient are used. For phosphorus in most countries, the content of this nutrient only suffices to classify the soil in five groups. In a few countries, other soil characteristics (textural groups, humus content, CaCO3 content) are accompanying factors. For potassium and magnesium in almost all countries, soil textural classes are taken into consideration as interacting soil properties. The most complicated calibration system is used in Hungary. In this country, in addition to soil properties crop demands for phosphorus and potassium are recognised as well, which results in an amount of calibration classes between 50 and 210. A more detailed description of the calibration systems in the collaborating countries is included in the appendices 1-3. The philosophy of soil test
calibration system influences strongly the philosophy of fertilizer recommendations, which will be discussed further.
Table 5. Classification of calibration systems for soil test results for mineral arable soils*
Country Phosphorus Potassium Magnesium
type explanations type explanations type explanations DE TH Q P content only split into 5
classes QS K content (5 levels) and soil textural groups (5), altogether
25 classes
QS Mg content (5 levels) and soil textural groups
(5), altogether 25 classes
DE HE same as DE TH
EE QS P content (5 levels), humus content (2 levels),
altogether 10 classes
QS K content (5 levels), soil textural groups(5),
altogether 25 classes
QS Mg content (5 levels), soil textural groups(5), altogether 25 classes LV QS P content (5 levels),
soil textural groups(4), altogether 20 classes
QS K content (5 levels), soil textural groups(4),
altogether 20 classes
QS Mg content (3 levels), soil textural groups(4), altogether 12 classes LT Q P content only split into 6
classes Q K content only split
into 6 classes QC Mg content(5 levels), group of crops (3) altogether 15 classes AT Q P content only split into 5
classes QS K content(5 levels), soil textural groups (3)
altogether 15 classes
QS Mg content(5 levels), soil textural groups (3)
altogether 15 classes PL Q P content only split into 5
classes K content (5 levels), soil textural groups(4),
altogether 20 classes
Mg content (5 levels), soil textural groups(4), altogether 20 classes SK, CZ QS P content (5 levels),
soil textural groups(3), altogether 15 classes
QS K content (5 levels), soil textural groups(3),
altogether 15 classes
QS Mg content (5 levels), soil textural groups(3), altogether 15 classes SL Q P content only split into 5
classes QS K content (5
levels), soil textural groups(2), altogether
10 classes
K content (5 levels), soil textural groups(2),
altogether 10 classes MEMHU ** QS P content (5 levels) soil
types (6) CaCO3 content (2 levels), altogether 60
classes
QS K content (5 levels) soil types (6) , soil textural groups (2) , altogether 60 classes
QS Mg content(3 levels), soil textural groups(3) altogether 9 classes MTAHU *** QSC P content(5 levels), soil
textural groups (5), pH(2 levels), CaCO3 content (5 levels), crop demands (2 levels), altogether 210
classes
QSC K content (5 levels), soils textural groups
(5 levels), crop demands (2 levels), altogether 50 classes
QS Mg content(3 levels), soil textural groups(3) altogether 9 classes
* Q – (quantity) – only nutrient content, QC- (quantity-crop demand)- nutrient content and crop demand for a nutrient, QS –( quantity-soil properties)- nutrient content and selected soil properties, QSC – (quantity, soil properties, crop demand) – nutrient content, soil properties, crop demand for a given nutrient
** official recommendations,
*** recommendations of the Institute of Soil Science Hungarian Academy of Science (pilot scale)
1.4 Qualitative analysis of soil data
With such differences in calibration systems, it was unwise to presume that the same soil sample would be classified by all laboratories into the same class of the soil’s nutrient supply. However, the results of comparative analysis are even worse than expected. To perform statistical calculations the five main classes of nutrient content were coded into the numbers 1 (very low) to 5 (very high). The two-factorial analysis of variance was then performed with country (delivering the samples) as the first and laboratory as the second factor. In all analyses, the interaction of the first and second factors proved to be not significant and hence the results may be interpreted for countries and laboratories separately. The results in the form of analysis of means for laboratories are presented in Fig. 2 – 4. This plot constructs a chart similar to a standard control chart, where each sample mean is plotted together with a centreline (CL) and upper (UDL) and lower decision limits (LDL). The centreline is located at the grand average of all observations Y. Any means that fall outside the decision limits indicate that the corresponding sample differs significantly from that overall mean.
For phosphorus, nine laboratories remained in the range of 95% confidence limit, 1 below and 2 above this limit (Fig 2). However, these nine laboratories classified the test soil samples in availability classes of 2.5 (between low and medium) to 4 (high) which cast doubts on the reliability of soil phosphorus analysis data across the nine neighbouring countries. Two laboratories HU MEM and HU MTA obviously overestimated the phosphorus content in comparison to the mean value, and the AT laboratory slightly underestimated it. It is worth noting that such discrepancies in soil analysis concerned the element for which the calibration system is the least complicated and in most countries includes the phosphorus content only.
Fig. 2. Analysis of means for phosphorus (1 – very low P content ….. 5 – very high P content)
From the Annex 1, the average contents of phosphorus in the “medium” class, irrespective of other soil parameters (if any), is calculated and shown in table 6.
The figures for laboratories using ME-3 and AL methods have been recalculated to values for CAL and DL method using equation (2). In relation to the mean value for 11 collaborating laboratories, the content of phosphorus in the “medium” class seems to be slightly underestimated by EE, LV, and LT laboratories.
Table 6. Average contents of phosphorus in the “medium “class [expressed in CAL/
DL units]
Average content of phosphorus in “medium” class in mg P*kg-1soil extrapolated to CAL/
DL units in the country
DE TH DE HE EE LV LT A T* PL SK SL CZ HU Mean
49-72 49-72 37-68 29-55 40-51 47-111 45-65 62-83 44-75 60-79 54-79 47-73
*In the latest edition of the “Austrian guidelines for appropriate fertilisation” (BMLFUW, 2006) the fact of the broad range and the high upper limit of the “medium” class was met with a reduction of fertiliser recommendation up to 50% in the “high” medium class of phosphorus and potassium (see also Spiegel et al., 2006).
For potassium, the situation is worse, because six laboratories only remained in the range of 95 % confidence limit (Fig. 3). These laboratories classified the average soil sample in availability classes from 2.5 (between low and medium) to 3.5 (between medium and high), i.e. in the range of one available class only. However AT, SK, and SL laboratories underestimated the potassium content in comparison to mean value and LT, HU MEM and HU RIS laboratories overestimated it considerably.
Fig. 3. Analysis of means for potassium (1 – very low K content …. 5 – very high K content)
Like phosphorus, the average contents of potassium in the “medium” class are calculated irrespective of other soil parameters and shown in table 7. The figures for laboratories using ME-3 and AL methods have been transformed to figures for CAL and DL method using equation (3). Compared to the mean value for all laboratories the content of potassium in the “medium” class is strongly overestimated in AT, SK, SL and CZ laboratories and underestimated in LV and LT laboratories.
Table 7. Average contents of potassium in “medium “class (expressed in CAL/ DL units)
Average content of potassium in “medium” class in mg K*kg-1soil extrapolated to CAL/
DL units in the country DE
TH
DE HE EE LV LT A T* PL SK SL CZ HU Mean
93- 148
93-148 92-132 69-137 71-97 113-212 94-150 147- 207
125- 182
143- 215
122- 172
105-163
* see explanation under table 6
For magnesium, 8 laboratories were in the range of 95 % confidence limit and this range was rather narrow from about 3 (medium content) to 4 (high content), i.e.
including one availability class (Fig 4).
Fig. 4. Analysis of means for magnesium (1 – very low Mg content …5 – very high Mg content)
The average contents of magnesium in the “medium” class; irrespective of other soil parameters, were calculated, and are given in table 8. The data for laboratories using ME-3 method have been transformed to CaCl2 units using the equation (4). In comparison to the mean value for all laboratories, the content of magnesium in the
“medium” class is strongly overestimated by SK and CZ laboratories and slightly underestimated by PL laboratory.
Table 8. Average contents of magnesium in “medium “class [expressed in CaCl2 units]
Average content of magnesium in “medium” class in mg Mg*kg-1soil extrapolated to CaCl2 units in the country
DE TH DE
HE EE LV* LT** A T PL SK SL** CZ HU Mean***
– 8962 62 – 89 73
– 87 126 – 210
234 – 350
– 10556 41 – 65
110 – 145
– 15076 108 – 145 68 -
120 73 - 105
* single laboratory using DL method
** laboratories using AL method
*** mean from eight laboratories
From the data in this chapter, the general conclusion can be drawn that the classification of soil test measurements into qualitative nutrient supply classes, even by laboratories using the same methods, differs to a much greater extent than the quantitative data, i.e. the amounts of extracted nutrients. These differences between laboratories can be partly explained by different criteria for soil texture estimation.
1.5 Soil texture
Soil texture as an accompanying factor is taken into consideration in all calibration systems for potassium and magnesium. However, the approaches to select the soil texture classes are different in the collaborating countries (Ann. 4). In four countries, DE THU, DE HE, EE, and HU, respectively, five soil textural classes are used.
Two countries, LV and PL, respectively, employ a four-class system and all others – a three-class one. Besides the differences in number and denomination of classes, different criteria are used for their selection. In DE THU, DE HU and EE the content of clay is a differentiating factor, in PL the content of silt (particles <0.02 mm) is considered, and in HU the so-called Arany values (App. 4) are used. With such different approaches, it is no wonder that soil classification to calibration classes for potassium and magnesium (when soil texture is taken into account) by several laboratories can be quite different. In fact, the heterogeneity of soil classification with respect to texture is the main obstacle in preparing a more uniform approach to soil testing in collaborating countries.
2. Comparison of recommended fertilizer rates
In two-factorial analysis of variance, countries (providing two soil samples each) were recognised as the first and laboratories (performing analysis and giving fertilizer recommendations) as the second factor. Both main factors proved to be significant while interaction was not and therefore the differences of the first order were interpreted only. Fertilizer recommendations were derived for the following crops: winter wheat (WW), winter rape (WRP), winter rye (WR), maize for grain (MAG), potatoes (POT), and maize for silage (MAS). The recommended fertilizer rates for all crops are presented in tables 9 – 14. However, due to high discrepancies between four crops grown for grain and without manure and two crops grown for fodder on manure the results for these two groups are discussed separately.
2.1 Recommended rates of nutrients for grain crops 2.1.1 Phosphorus
The average recommended rates of phosphorus for selected crops are presented in table 9 and 10. In table 9, the average values of the soil test for phosphorus are included as the decisive factor that all laboratories took into consideration calculating phosphorus rates.
Table 9. Average recommended phosphorus rates for countries kg P*ha-1 Country Average
value of soil P test*
Recommended P rate for crop WW**
.n=242 WRP
.n = 242 WR
n = 242 MAG
n = 176 POT
n=242 MAS
n = 176 DE TH 3.00 b*** 36.0 cd 49.2 bc 32.0 b 41.2 c 2.3 bc 8.2 cd DE HE 4.00 a 20.4 ab 31.0 a 19.0 a 24.6 ab -5.5 ab 0.6 ab EE 4.23 a 18.8 a 27.0 a 17.8 a 26.3 ab -10.5 a 0.7 ab LV 2.27 c 47.6 ef 61.9 de 44.0 cd 61.5 de 13.1 d 17.7 ef LT 1.73 c 57.5 f 72.2 e 53.0 d 71.7 e 17.4 e 24.2 f AT 3.36 b 29.5 bc 44.3 b 29.0 b 38.1 bc 9.3 c 13.8 de PL 4.19 a 16.5 a 26.2 a 17.8 a 21.4 a -8.1 a -0.5 a SK 3.00 b 40.1 de 54.0 bcd 36.4 bc 49.4 cd 4.4 c 11.0 cde SL 4.10 a 17.6 a 26.8 a 18.7 a 24.8 ab -11.2 a -0.6 a CZ 2.32 c 47.3 ef 63.2 cde 43.6 c 58.5 de 13.2 d 16.3 e HU 2.95 b 34.6 cd 45.5 b 33.8 b 44.6 c 1.0 bc 6.7 bc
Average 3.19 33.3 45.6 31.4 42.0 2.8 8.9
* explanations see Fig. 2 and text
** this destination will be used in all tables and drawings
*** values marked by the same letter do not differ significantly
Table 10. Average recommended phosphorus rates by laboratories kg P*ha-1 Laboratory Recommended P rate for crop
.n=242WW WRP
.n = 242 WR
n = 242 MAG
n = 176 POT
n=242 MAS
n = 176
DE TH 40.9 d 49.9 de 37.5 d 36.6 ab -20.4 b 15.9 ef
DE HE 38.2 cd 39.1 bcd 34.0 cd 45.3 c -31.1 a 7.9 cd EE 24.3 ab 46.4 cde 22.0 ab not given 26.1 f 20.2 f
LV 41.0 d 63.2 e 36.9 d not given 51.1 g 0.0 b
LT 19.1 a 22.4 a 17.0 a not given 21.3 f 18.4 ef
AT 26.9 ab 36.5 bc 26.7 bc 38.7 ab -10.0 c 13.2 de
PL 29.0 abc 36.0 bc 25.4 abc 39.6 ab 18.6 f 5.6 bc
SK 26.5 ab 34.3 abc 26.0 abc 37.0 ab 3.7 de 18.6 ef
SL 63.5 e 84.1 f 55.2 e 72.2 d -32.5 a 2.3 bc
CZ 23.7 ab 31.0 ab 22.8 ab 33.2 a -4.2 cd -7.6 a
HU 32.9 bcd 58.9 de 41.6 d 33.5 a 7.5 e 3.8 bc
Average 33.3 45.6 31.4 42.0 2.8 8.9
Already from a visual analysis of the data in table 9, the conclusion can be drawn that the differences in the rate of phosphorus recommended for individual countries are mainly attributed to the differences in soil phosphorus content. The poorer the soil in phosphorus (lower number for P test), the higher the average recommended phosphorus rate by most of the laboratories. However the results in table 10 show that laboratories from LV and SL recommend regularly (for all countries) higher rates of phosphorus than the average one. Indirectly, it can be therefore concluded that these laboratories seem to overestimate the recommended P rate beyond the rate justified by the phosphorus content in the soil. After eliminating these two laboratories, the rates of phosphorus recommended by the remaining nine laboratories are very similar, and mostly do not differ significantly. The average rates of P for five classes of P content in the soil recommended by 9 laboratories (excluding LV and SL are presented in Fig. 5.
Fig. 5 Average rate of phosphorus for different classes of P content in the soils recommended by nine laboratories.
The most interesting are phosphorus rates recommended for medium P content in the soil. These rates (rounded) recommended by eight laboratories are 31, 40, 28 and 38 for winter wheat, winter rape, winter rye and maize for grain, respectively. In most systems of fertilizer recommendations, the rate of phosphorus in medium soil class is focused on the potential (expected) uptake of this element by the crop. It is worth to remind that the assumed crop yields were 7 ton of winter wheat, 4 ton of winter rape, 6 ton of winter rye and 8 ton of maize.
Phosphorus taken up with these yield (grain + straw) according to PL norms (Appendix 5) equalled (rounded) 33 kg P, 40 kg P, 31 kg P and 50 kg P for WW, WRP, WR and MAG respectively. The P uptake with exception of maize thus corresponds very closely to the recommended rates of this element. It can be therefore concluded that nine laboratories among the eleven, are giving recommendations in pursuance of the principle “input equals output” for medium class of nutrient content. These laboratories are applying also a very similar approach to the increasing and/or decreasing P rates in dependence on the content of available phosphorus in the soil.
The average for all crops recommended P rates are 160 % for very low, 131 % for low, 60 % for high and 14 % for very high of the rate for medium phosphorus content.
For explaining the reasons of outlying the P rates given by LVand SL laboratories from the average ones it is necessary to examine the algorithms of fertilizer recommendations in individual countries (Chapter 4).
2.1.2 Potassium k
Similar procedure of data processing was applied for potassium. The average recommended rates of potassium for selected crops are presented in table 11 and 12.
Again, in table 11 the values of the soil test for potassium are included.
The rates of potassium, as those for phosphorus depend on the content of available potassium in the soil (Tab.11). This rule holds for all laboratories, although laboratories from DE, HE, SK, and SL, are generally recommending higher K rates as an average (Tab. 12). The average rates of potassium for four crops grown for grain depending on the content of potassium in the soil are presented on Fig.6.
Potassium rates recommended for medium content of potassium in the soil are (rounded) 94, 158, 98 and 161 kg K*ha-1 for winter wheat, winter rape, winter rye and maize for grain respectively. The uptake of potassium with the final yield of these crops (grain + straw), according to PL norms (Appendix 5) is however 109, 204, 115 and 276 kg K*ha-1 for WW, WRP, WR, and MAG.
Table 11. Average recommended potassium rates for countries kg K*ha-1 Country Average
value of soil K test
Recommended K rate for crop .n=242WW WRP
.n = 242 WR
n = 242 MAG
n = 176 POT
n=242 MAS
n= 176 DE TH 4.00 e 52.5 a 98.3 a 60.0 a 104 a 2.0 a -8.2 a
DE HE 3.91 e 55.7 a 101 a 60.5 a 118 a 4.7 a -8.3 a
EE 2.86 bc 104 cd 166 d 107 de 187 cd 72.7 d 39.7 d
LV 2.73 b 105 cd 170 d 109 de 193 d 65.8 cd 42.5 d
LT 1.59 a 140 e 217 e 140 f 244 e 115 e 61.5 e
AT 3.27 d 82.8 b 130 b 77.7 b 154 b 44.4 bc 38.0 cd
PL 2.73 b 107 d 172 d 112 e 189 cd 65.3 cd 38.0 cd
SK 3.32 d 79.9 b 138 bc 87.0 bc 166 bcd 41.7 b 23.7 bc SL 3.18 cd 89.7 bc 149 bcd 96.0 cd 169 bcd 44.9 bc 20.3 b CZ 3.04 bcd 91.7 bcd 150 bcd 95.7 cd 167 bcd 50.3 bc 23.0 bc HU 2.95 bcd 102 cd 161 cd 109 de 160 bc 41.0 b 30.5
Average 3.05 91.8 150 95.9 169 49.8 27.5bcd
explanation under table 9
Table 12. Average recommended potassium rates by laboratories kg K*ha-1
Laboratory Recommended K rate for crop
.n=242WW WRP
.n = 242 WR
n = 242 MAG
n = 176 POT
n=242 MAS
n = 176
DE TH 101 e 171 cd 124 d 127 a -133 a -21.6 b
DE HE 144 f 177 cd 146 e 235 c 17.4 b 14.5 cd
EE 77.9 bc 124 b 59.2 b not given 54.5 d 64.5 f
LV 62.3 b 103 b 53.1 ab not given 99.5 e not given
LT 54.7 a 57.1 a 41.1 a not given 52.2 d 44.7 e
AT 70.7 bc 192 d 67.7 b 175 b 46.4 cd 99 g
PL 85.8 cde 125 b 104 c 140 a 148 f 24.1 d
SK 95.9 de 168 c 103 c 194 b 82.1 e 127 h
SL 159 f 259 e 170 f 222 c 29.1 bc 14.9 cd
CZ 86.7 cde 155 c 92.9 c 124 a 48.5 cd -56.8 a
HU 82.0 cd 122 b 93.3 c 131 a 106 e not given
Average 91.8 150 95.9 169 49.8 27.5
Fig. 6 Average rate of potassium for different classes of K content in the soils recommended by 11 laboratories.
Therefore, for cereals, the average recommended rates of potassium correspond quite closely to the uptake of this element, while for winter rape and grain maize recommended K rates are lower or much lower than the removal of potassium with the assumed yield of both crops. Even laboratories from DF HE, SK and SL recommending higher than average rates of potassium for winter rape and maize for grain seem to abstain from the principle “ input equals output” for these crops. All laboratories are applying a similar approach to the increasing and/or decreasing K rates in dependence on the content of available potassium in the soil. The average for all crops recommended K rates are 163 % for very low, 136 % for low, 59 % for high and 36 % for very high of the rate for medium potassium content. With exception of very high K content these increases and/or decreases of potassium rates correspond astonishingly closely to the similar figures for phosphorus.
2.1.3 Magnesium Mg
Not all laboratories are recommending magnesium rates at all, and therefore the results for this element are based on the smaller set of date than the data for phosphorus and potassium. Moreover, the magnesium rates recommended by SL laboratory (except those for AT country) seem to be unrealistically high and therefore they have been discarded from further analysis. Finally, the two-factorial analysis of variance was performed for nine laboratories. The results by countries are presented in table 13 and by laboratories in table 14.
Table 13. Average recommended magnesium rates for countries kg Mg*ha-1 Country
Average value of soil Mg
test*
Recommended Mg rate for crop .n=220WW WRP
.n = 220 WR
n = 220 MAG
n = 176 POT
n=220 MAS
n = 176 DE TH 3.83 b 8.1 bc 10.1 bc 8.5 cdef 21.4
bcde -12.1 a 2.1 abc DE HE 2.89 a 16.2 ef 17.2 de 13.5 fg 31.1 ef -3.6 def 10.6 e EE 2.67 a 18.8 f 21.6 e 16.9 g 35.3 f -1.9 f 8.4 bcde LV 4.05 b 2.3 a 2.7 a 1.9 a 5.1 a -7.5 bcde -1.1 a LT 4.17 b 5.5 ab 6.2 ab 4.8 ab 12.8 ab -8.1 abc 3.3 abcd AT 3.89 b 8.9 bcd 9.0 b 7.7 cde 19.8 bcd -6.8 cde 1.9 ab PL 3.11 a 14.1 def 15.9 de 12.8 efg 27.7 def -3.7 cdef 9.9 de SK 4.28 b 5.7 ab 5.6 ab 3.6 ab 11.3 ab -7.8 abcd 2.9 abc SL 3.78 b 8.1 bc 8.9 b 7.3 cd 18.8 bcd -7.0 cde 5.8 bcde CZ 3.05 a 12.8 cde 15.5 cd 11.0 def 25.3 cdef -3.1 ef 8.9 cde HU 4.05 b 7.3 abc 8.6 b 6.6 abc 15.3 abc -11.6 ab 4.1 abcd
Average 3.62 9.8 11.0 8.6 20.4 -6.6 5.2
* explanation under table 9
Table 14. Average recommended magnesium rates by laboratories kg Mg*ha-1 Laboratory Recommended Mg rate for crop
.n=220WW WRP
.n = 220 WR
n = 220 MAG
n = 176 POT
n=220 MAS
n = 176 DE TH 14.2 de 16.5 de 13.4 d 19.0 bc -25.2 ab 10.3 de DE HE 14.4 de 11.4 bce 11.1 cd 34.7 d -29.5 a 0.4 bc
EE 6.4 ab 7.7 bc 4.5 ab not given 3.1 cd 4.5 cd
LV not given not given not given not given not given not given LT 3.7 a 4.9 a 2.4 a not given not given not given
AT 7.9 abc 7.1 bc 7.1 bc 7.1 a -29.1 a -7.4 a
PL 10.4 bcd 12.0 bce 8.3 bc 22.6 c 8.1 e 4.9 cd
SK 15.7 e 18.0 e 13.8 d 34.7 d 18.2 f 14.4 e
SL 59.1 f 60.4 d 59.5 d 62.7 e 5.0 de 32.7 f
CZ 11.6 cde 13.4 cde 11.2 cd 14.1 abc -24.5 b -5.1 ab
HU 4.1 a 8.2 bce 5.4 ab 10.2 ab 0.6 cd 2.0 c
Average 9.8 11.0 8.6 20.4 -6.6 5.2
In spite of a smaller amount of data, the general regularities are the same as those already presented for phosphorus and potassium. The recommended rates of Mg depend strongly on the content of available magnesium in the soils (Fig. 7).
The only difference is in similar recommendations for very low and low content of magnesium. All laboratories are applying a similar approach to the increasing and/or decreasing Mg rates, dependent on the content of available magnesium in the soil.
The average for all crops recommended Mg rates are 152 both for very low and low, 38 % for high and 15 % for very high of the rate for medium magnesium content.
These figures are somewhat different from those for phosphorus and potassium.
Fig. 7 Average rate of magnesium for different classes of Mg content in the soils recommended by nine laboratories.
Magnesium rates recommended for medium content of magnesium in the soil are (rounded) 14, 17, 13 and 29 kg Mg*ha-1 for winter wheat, winter rape, winter rye and maize for grain respectively. The uptake of magnesium with the final yield (grain + straw) of these crops, according to PL norms (Appendix 5) is however 16, 23, 14 and 46 kg Mg*ha-1 for WW, WP, WR and MAG respectively. Therefore, the recommended rate of magnesium, except winter rye is lower than the uptake of this element by crops.
2.2 Recommended rates of nutrients for potato and silage maize
The main difference between four grain crops (winter wheat, winter rape winter rye and grain maize) and two other crops (potato, silage maize) was that the last ones were grown on manure (30 t*ha-1 FYM for potato, 30 m3*ha-1 cattle slurry for maize).
The content of nutrients in FYM was not uniformly accepted, while for slurry it was assumed that it contains 0.74 kg P, 5.3 kg K, and 0.50 kg Mg per m3. Therefore, with this amount of slurry 22.2 kg P, 159 kg K, and 15 kg of Mg were incorporated per
ha. According to Polish norms (Appendix 5) FYM contains 0.13 % P, 0.58 % K and 0.12 % Mg and consequently with the amount of 30 t FYM *ha-1 39 kg P, 174 kg K and 36 kg Mg were incorporated under potato crop. The approach to the amount of nutrients in manure and its utilisation by crop is different in collaborating countries and it is no wonder that the rates of mineral fertilizers recommended by laboratories differ considerably (Tables 9 - 14).
The average rates of phosphorus for both crops are very low, but difered substantially between laboratories (Tab. 10). Some laboratories recommend “minus”
phosphorus rates, which practically means that the farmer should abstain from application of P fertilizers for potato and/or maize and even can deduct the excess of this element from the rates for succeeding crop. The rates of phosphorus, particularly for potato proved to be more or less independent of the content of available P in the soil (Fig. 8). The significant differences were found between very low (in case of maize also low) content and all other classes of P tests only. The average recommended rates of phosphorus for medium class of P content were 3.3 for potato and 10.5 kg P*ha- for maize. According to Polish standards with the assumed yield of 30 t tubers and 10 t dry matter maize the removal of phosphorus from the soil is about 19 kg P and 32 kg P, respectively. Therefore, the potato was supplied in some surplus with phosphorus from FYM, while the requirements of maize were satisfied with the amount of phosphorus from both sources only.
Fig. 8 Average rate of phosphorus for different classes of P content in the soils recommended by laboratories
There is a large scatter of recommended potassium rates (Table 12). For potato the lowest recommended rate is -133 kg K*ha-1 (DE TH laboratory) and the highest 148 kg/*ha (PL laboratory) e.g. the range is about 280 kg of potassium. The corresponding figures for maize are -56.8 kg K*ha-1 (CZ laboratory) and 127 kg K*ha-1 (SK laboratory) with the range of about 180 kg K*ha-1. Phosphorus alike, the dependence of potassium rates on the K content in the soil is much lower than in case of grain crops (Fig.9).
Fig. 9 Average rate of potassium for different classes of K content in the soils recommended by laboratories
Potassium rates in the range of K availability classes from medium (for potato even from low) to very high content are practically the same. Therefore, in most recommendations for systems of potato and silage maize crops grown on manure, the content of this element in the soil is not taken into proper consideration. The average recommended rates of potassium for medium class of K content were 51 kg K for potato and 17 kg K*ha- for maize . According to Polish standards with the assumed yield of 30 t tubers and 10 t dry matter maize the removal of potassium is 170 kg K and 200 kg K, respectively. For potato, this amount of potassium is supplied in manure, while for maize the total amount of potassium from slurry and mineral fertilizers is slightly below the uptake of this element.
With a few exceptions, the laboratories do not practically recommend magnesium fertilization for potato and silage maize crops (Table 14). According to Polish standards the uptake of magnesium with assumed yield of tubers is about 10 kg Mg*ha-1 and with yield of maize about 20 kg Mg*ha-1. The amount of magnesium in FYM is in the excess to Mg uptake by potato crop while the amount of magnesium provided with slurry does not meet the requirements of maize. Phosphorus and potassium alike, the recommended magnesium rates, with the exception of very low magnesium content in the soil were practically independent of the results of soil tests for this element (Fig.10).
Fig. 10 Average rate of magnesium for different classes of Mg content in the soils recommended by laboratories
3. Algorithms of fertilizer recommendations 3.1 General remarks
Fundamentally, most of the recommendation systems are based on algorithms e.g. the rates of nutrients are calculated using external input data, some internal parameters, and the set of relations between them. In MOEL project the input date (crops, yield, manure for potato and silage maize) have been fixed in advance and the content of available nutrients in the soil was estimated by each laboratory. The internal parameters and algorithms itself are briefly described by most collaborating parties and presented in this chapter. Unfortunately, not all countries revealed these algorithms. The most important internal, quantitative parameters are unit uptake of nutrients by the crops (App. 5), amount of nutrients in manure (App. 6) and correction factors for the content of available nutrients in the soils (App. 7).
3.2 Algorithms of fertilizer recommendations in collaborating countries 3.2.1 Estonia EE
Step 1. Crop nutrient requirements CNR are calculated for so-called standard yield (4t/ha WW and WR, 2.5 t/ha WRP and 23 t/ha POT) and increased or decreased for each tonne of yield, if different from the standard one.
Step 2. CNR for N, P, K is corrected (multiplied by correction factors) depending on soil texture, CNR for N is further corrected for humus content (in multiplicative way) and preceding crop (in additive way)
Step 3. After steps 1.2 the CNR for P, K is corrected depending on the content of these nutrients in the soil (App.7)
Step 4. For crops grown on manure CNR is diminished by the amount of nutrients in manure (App. 6) multiplied by utilisation factor
3.2.2 Latvia LV
Step 1 Calculation crop nutrients requirements CNR (standard demand) Step 2. Correction for yield
Step 3. Correction for soil texture Step 4. Correction for soil pH (P only)
Step 5. Correction for the content of P and K in soil Step 6. Correction for previous crop
Step 7. Correction for manure application, taking into account nutrient utilization from manure
3.2.3 Lithuania LT
Step 1. N, P, K fertiliser rates are calculated for the planned yield. In order to simplify the calculation, four crop productivity groups are set: low, average, high and very high.
Step 2. Correction for preceding crop. Depending on the type of preceding crop P rates would be corrected by maximum 20 % and K rates by maximum 25 %.
Step 3. Correction for after-effect of fertilizers applied for preceding crop. Correction coefficients are in the range 0.8-1.2 for P and 0.9-1.1 for K.
Step 4. Correction for soil texture. Correction coefficients: 0.9-1.2 for P and 0.8-1.1 for K.
Step 5. Correction for soil pH. Correction coefficients: 0.9-1.2 for P and 0.8-1.1 for K.
Step 5. Correction for the content of available P and K content in the soils (App.7) Step 6. Correction for the manure application. Manure application is based on amount of applied nitrogen – the aim is to apply 170 kg/ha of total nitrogen (N). Single manure application brings into soil 170 kg/ha N, 100 kg/ha P2O5 and 200 kg/ha K2O (N: P2O5: K2O = 1:0.5:1.18). The amount of nutrients is multiplied by utilisation coefficients (Table App.6).
Step 7. Step Magnesium rates depend on the class of Mg content in the soil and crop requirements for magnesium:
Mg content in the soil Mg kg*ha-1 for the crop with magnesium requirements
High Medium Low
A very low 15 12 9
B low 12 9 6
C medium 9 6 3
D high 6 3 0
E very high 0 0 0
3.2.4 Austria AT
Step 1. Standard rates of P, K for “average” yield (24 kg P, 66 kg K for WW and WR, 33 kg P, 166 kg K for WRP, 37 kg P, 166 kg K for MAG, 26 kg P, 150 kg K for POT, 39 kg P, 187 kg K for MAS),(“average” yields are: 3.5-5.5 t WW, 2.0-3.0 t WRP, 3.5 – 5.5 t WR, 6.0 – 10.0 t MAG, 25 – 35 t POT , 13.3 – 16.7 t MAS). All rates for medium content (class 3) of P and K in the soil, but if the P and K content is in the upper half of the content in this class then the standard rates are decreased by 50 % for grain crops (for K and root crops by 25%).
Step 2. Correction of the rates from step 1 depending on the yield level, -10 % for the “low” yield and +15 % for the “high” yield. Correction of K rates depending on K/Mg ratio in the soil.
Step 3. Correction of the rates depending on the content of P, K in the soils (App.7).
Step 4. Correction of the rates (in additive way) depending on the standard amount of P, K left in the soil after different preceding crops:
Preceding crop and left by-
product Kg P*ha-1 left
Kg K*ha-1 left depending on preceding crop yield
“lower” “average” “higher“
Cereals (straw), potato
(haulm) 4.4 33 41 58
Grain maize, rape (straw) 8.8 75 99 124
Step 5. Mg rates depend only on the content of Mg in the soil and ratio K/Mg in the soil:
Mg content in the soil K/Mg <5:1 K/Mg >5:1
Very low 42 54
Low 30 42
Medium 18 30
3.2.5 Poland PL
Step 1..Calculation of crop nutrients requirement CNR as the product of yield and unit nutrient uptake (App 5)
Step 2. Correction of CNR depending on the content of available nutrients in the soil (App.7)
Step 3. If manure is applied then correction is made depending on the content of nutrients in manure and utilisation coefficient (App. 6). In Poland, the term nutrient replacement value NRV is used instead of utilisation coefficient.
3.2.6 Slovenia SL
Step 1. Calculation of crop nutrient requirement CNR (according to the uptake) Step 2. Correction of CNR depending on the content of available nutrients in the soil (App.7)
Step 3. The rates of Mg depend on the magnesium content in the soil (class of Mg content) and the relation of K/Mg:
Class of content
Fertilization norm Mg kg*ha-1
A 36-48*
B 30-36*
C 0-24*
D 0
E 0
* if the ratio between K (by AL) and Mg (by CaCl2) is wider than 6:1, then higher norm is advised, otherwise the lower
3.2.7 Czech Republic CZ
Step 1. Calculation of crop nutrients requirement CNR as the product of yield and unit nutrient uptake (App. 5)
Step 2. Correction of CNR depending on the content of available nutrients in the soil (App. 7)
Step 3. If manure is applied then correction is made depending on the content of nutrients in manure and utilisation coefficient (App. 6)
3.2.8 Hungary HU Step 1. Affiliation of soil site on the base of soil properties.
Step 2. Forecasting the yield level on the base of past 5 years yields
Step 3. Quantification of P, K soil supplying capacity on the base of available nutrients content, taking into consideration soil type and other soil properties (pH.
CaCO3 content)
Step 4. Calculation of plant nutrient requirements using the formulae:
Nutrient requirement = Q*f Where : Q – forecasted yield
f – standard coefficient including crop, soil nutrient supply, soil type Step 5. Correction of nutrient requirement (step 4) taking into consideration the preceding crop, manure, soil management etc.
Note: These steps characterise the official system of fertilizer recommendations, which has been in use since 1979 under the Ministry of Agriculture (MEM). At present, the new, more sustainable system is being implemented under the supervision of Institute of Soil Science (MTA). The simplified comparison of both systems is presented below:
MÉM-NAK (1979) MTA-TAKI (pilot scale)
Intensive fertilization, obtaining the highest
possible yields Sustainable fertilization, economically justified yield
Fertilization of soil Fertilization of crop Building and maintenance of high or very
high content of available nutrients in the soil Building and maintenance of medium content of available nutrients in the soil Bringing the soil to optimal P, K level in short
time Prolonged period of soil bringing to
optimal P, K level
Every year fertilization Fertilization of selected crops in rotation
Fertilization of soils showing high content of
P, K Fertilization of soil showing low and
medium content of P, K
Uniform threshold P, K content values Differentiated threshold P, K, content values
4. Discussion and conclusions 4.1 Soil tests
4.1.1 Comparison of extraction methods
In the eleven collaborating countries, fertilizer recommendations for P, K, and Mg are based on chemical soil tests. The whole procedure of soil testing includes several steps: soil sampling, extraction of nutrients, measurement of nutrient content, and classification into one of availability classes. Laboratories in eleven countries are using two extraction procedures for pH, four procedures for P and K, and five procedures for Mg (Table 1).
In half of the laboratories, soil pH is measured in CaCl2 solution and by half in KCl. The inter-laboratory variation for both methods is very low indeed (Tab.4), which is no wonder because soil pH is the most common soil analysis perfectly well mastered by all agrochemical laboratories. The average value of soil pH measured in KCl is slightly lower than in CaCl2. The relation between the two measures is approximated best by a double reciprocal regression function and not by linear function (Fig. 1). From this function it might be concluded that in acid soils pHKCl
values are by 0.5 – 0.6 units lower than pHCaCl2, for neutral soils, both measures are the same, and for alkaline soils values determined with pHKCl are slightly higher as those with pHCaCl2 (Fig 1). However such relation concerns only the date from this research and should not be generalized. Potassium ion from KCl has higher exchange power then calcium ion from CaCl2 and replace more hydrogen ions from soil solid phase. Therefore the pH measured in KCl solution is regularly lower then in CaCl2 though this relation is not necessarily proportional.
Practically both methods can be used therefore alterably, but due to the lack of linearity it would be better to agree upon one uniform extracting solution. Though most countries are using KCl, other considerations are in favour of CaCl2 solution.
The ionic composition and strength of CaCl2 resembles more the composition of soil solution in soils kept in good culture. Besides, in the same extract the content of available magnesium could be measured after slight modification of original Schachtschabel method.
All laboratories are using one solution for P and K extraction, though the chemical composition of these solutions is different. Approximately half of the laboratories are using solutions containing Ca2+ in acid medium (acidified with HCl or acetic acid) e.g. DL or CAL methods. Practically the same quantities of P and K are being extracted by both solutions and they can be used likewise. The problem arises with alkaline (calcium carbonate) soils where most of the phosphorus is in the calcium phosphate forms freely soluble in the acid medium of DL and/or CAL solutions.
To overcame this problem PL laboratory is using AL method for alkaline soils and laboratories from DE TH and DE HE are using the empirical formulae to modify the extracted (measured) amount of phosphorus, taking into consideration the pH of CAL solutions measured after extraction of P and K (final pH).
P,K adjusted= P,K measured*(1+0,83*(pHCAL)final-4,1)
The remaining laboratories are using either AL method or Mehlich – 3 method.
The amounts of P and K extracted by these methods are significantly higher in comparison to DL and/or CAL methods (Table 3). The advantage of Mehlich method is in the possibility of simultaneous extraction of Mg (as well as micronutrients) and this method seems to gain importance in Central – Eastern European countries.
Some years ago, the SK and CZ laboratories switched over to this method and in 2004, the EE laboratory changed extraction solution from DL into Mehlich-3. This possibility is also being considered in PL. There is a straight relation between the amounts of P and/or K measured by DL/CAL methods and AL/Mehlich-3 methods with the slope of regression line close to one. It gives the possibility to recalculate the values from one method to another as has been presented in Tables 6, 7. However, the determination coefficients for both nutrients, particularly for phosphorus, are not too high and these relations are strongly sample-dependent (variable).
Laboratories, which use the same method of analysis, produce statistically uniform results (Table 4). This proves the reliability of both the methods and laboratories.
