i
New Tools for Interpreting Foliar Nutrient Status
Robert P. Brockley
Contract Research Report to Forest Practices Branch, Ministry of Forests, Lands
and Natural Resource Operations
Contract OT12FHQ299
March 2012
i
i
TABLE OF CONTENTS
Introduction 1
Normalization of Foliar Nutrient Data 3
Inter-laboratory Comparisons 3
Laboratory Methods 3
Pacific Soil Analysis 3
Ministry of Environment 4
Data Analysis 4
Normalization Rationale 5
Total Nitrogen 5
Total Sulphur 6
Sulphate-sulphur 6
Phosphorus, Potassium, Calcium, Magnesium 7
Boron 8
Copper, Zinc, Iron, Manganese 8
Normalization Spreadsheet 9
Precautions 9
Revised Foliar Nutrient Interpretative Criteria 10
Data Sources 10
Macronutrients (N, P, K, Ca, Mg, S) 12
Sulphate-sulphur 12
Boron 13
Other Micronutrients 13
Nutrient Ratios 14
Use of Interpretative Tables 15
Precautions 16
Reference 17
TABLES
1 Interpretation of macronutrient concentrations in current year’s foliage of lodgepole pine, interior spruce, and Douglas-fir 25
2 Interpretation of sulphate-sulphur concentrations in current year’s foliage of lodgepole pine, interior spruce, and Douglas-fir 26
3 Interpretation of nutrient ratios in current year’s foliage of lodgepole pine, interior spruce, and Douglas-fir 27
FIGURES
1 Relationship between foliar N (PSA) and foliar N (MoE) 28
2 Relationship between foliar S (combustion) (PSA) and foliar S (combustion) (MoE) 28
3 Relationship between foliar S (combustion) (MoE) and foliar S (ICP) (MoE) 29
4 Relationship between foliar S (combustion) (PSA) and foliar S (ICP) (MoE) 29
5 Relationship between foliar sulphate-S (PSA) and foliar sulphate-S (MoE) 30
6 Relationship between foliar K (MoE) and foliar K (PSA) 30
7 Relationship between foliar Ca (MoE) and foliar Ca (PSA) 31
8 Relationship between foliar Mg (MoE) and foliar Mg (PSA) 31
9 Relationship between foliar P (MoE) and foliar P (PSA) 32
10 Relationship between foliar B (PSA) and foliar B (MoE) 32
11 Relationship between foliar Cu (PSA) and foliar Cu (MoE) 33
12 Relationship between foliar Fe (PSA) and foliar Fe (MoE) 33
13 Relationship between foliar Zn (PSA) and foliar Zn (MoE) 34
14 Relationship between foliar Mn (PSA) and foliar Mn (MoE) 34
15 Foliar data “normalization” spreadsheet 35
i
Introduction
Forests in British Columbia are commonly nutrient deficient, and foliar analysis is widely used by forest practitioners to evaluate their nutrient status as part of the stand selection process for large-scale fertilizer operations. Foliage is collected using standardized foliar sampling methodology (Ballard and Carter 1986; Brockley 2001). Following laboratory nutrient analysis, foliar analytical results are compared with published interpretative criteria to confirm nitrogen (N) deficiencies, and to infer whether other nutrients will either limit growth response or become growth-limiting after N is added. Analytical results are often used to develop appropriate fertilizer formulations to correct inferred nutrient deficiencies.
The interpretation of foliar nutrient data is not straightforward. Foliar nutrient interpretations are subject to serious shortcomings when foliage is collected using non-standardized methods, and when foliar data are reviewed without knowledge or consideration of site ecological characteristics. Also, to arrive at the correct diagnosis of stand nutrient status, appropriate weight must be assigned to several different components of foliar nutrition: 1) absolute levels of individual foliar nutrients, 2) balance of foliar levels of one nutrient to another, and 3) levels of inorganic fractions of specific nutrients (e.g., SO4-S). Finally, foliar analytical results may differ depending on the methodology used for laboratory extraction and determination. For example, recovery of N and S from plant tissue is typically lower with wet (i.e., acid) digestion methods than with dry (i.e., combustion) methods (Randall and Spencer 1980; Guthrie and Lowe 1984; Simonne et al. 1994; Horneck and Miller 1998). In some cases, differences may be large enough to affect diagnoses of nutrient sufficiency or deficiency based on available interpretative criteria. Published nutrient interpretative criteria do not typically account for differences in laboratory analytical methodology. However, known differences in laboratory analytical results for different nutrients can be used to “normalize” foliar nutrient data prior to interpretation. By removing the effect of differences in laboratory analytical methodology on results, “normalization” can improve both interpretative reliability and the development of appropriate fertilizer prescriptions. No inferences about the quality or integrity of results provided by any specific laboratory should be drawn from the “normalization” requirement. It is simply a tool to facilitate reliable interpretation of stand nutrient status.
Reliable interpretation of foliar nutrient data is also dependent on the availability of appropriate foliar nutrient interpretative criteria for the tree species of interest. Ballard and Carter (1986) suggested some interpretations of foliar nutrient levels in several conifer species occurring in British Columbia, based largely on a review of earlier published research undertaken elsewhere. These interpretative criteria were slightly revised by Carter (1992). A large amount of forest nutrition and fertilization research with lodgepole pine (Pinus contorta Dougl. var. latifolia Engelm.), interior spruce (Picea glauca (Moench) Voss and Picea engelmannii Parry, or naturally-occurring hybrids of these species), and Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) was undertaken in the BC interior in the 1980’s and 1990’s by the B.C. Ministry of Forests. Published and unpublished results from this work were used by Brockley (2001) to update lodgepole pine foliar nutrient interpretative criteria. These data, and additional foliar data collected from repeated fertilization experiments with interior spruce and lodgepole pine during the past decade, provide an excellent opportunity to develop revised interpretative criteria for interior spruce and Douglas-fir and to further refine the lodgepole pine criteria.
This report summarizes contract work undertaken in 2011/12 to develop a user-friendly tool to “normalize” foliar nutrient data and to develop revised nutrient interpretative criteria for lodgepole pine, interior spruce, and Douglas-fir. When used in conjunction with each other, the “normalization” tool and the revised interpretative criteria should result in more reliable interpretation of foliar nutrient data and improved fertilizer prescriptions.
“Normalization” of Foliar Nutrient Data
Inter-laboratory Comparisons
The first step in the “normalization” process is to obtain a reliable set of comparative foliar data from different analytical laboratories. In November 2011, 48 previously analyzed lodgepole pine foliage samples were selected for use in an inter-laboratory comparative study between the Ministry of Environment (MoE) laboratory (Victoria, BC) and the Pacific Soil Analysis Inc. (PSAI) laboratory (Richmond, BC). These two laboratories undertake the vast majority of nutrient analyses on conifer foliage samples collected by government and industrial forestry clients in British Columbia. The foliage samples for the inter-laboratory study were specifically selected to cover a broad range of foliar levels for most nutrients. All samples had been previously collected from several forest nutrition studies (EP 886.13, EP 1185) in the BC interior using standard foliar sampling protocol. Each foliage sample was thoroughly mixed and split into two sub-samples, one of which was subsequently sent to each laboratory. The following analyses were completed by each laboratory: N, phosphorus (P),potassium (K), calcium (Ca), magnesium (Mg), total sulphur (S), inorganic sulphate-S (SO4), copper (Cu), zinc (Zn), iron (Fe), manganese (Mn), and boron (B). Each laboratory used their standard procedures for extraction and determination.
Laboratory Methods
Pacific Soil Analysis
Foliage was digested using a variation of the sulphuric acid – hydrogen peroxide procedure described by Parkinson and Allen (1975). The digests were analyzed colorimetrically for N on a Technicon Autoanalyzer using a phenol–hypochlorite reaction (Weatherburn 1967). A spectrophotometer was used to measure P, using a procedure based on the reduction of the ammonium molybdiphosphate complex by ascorbic acid (Watanabe and Olson 1965). Total K, Ca, and Mg were determined using a Perkin-Elmer atomic absorption spectrophotometer.
Separate subsamples were dry-ashed and Cu, Zn, Fe, and Mn were determined by atomic absorption spectrophotometery. After dry-ashing, B was determined colorimetrically using the azomethine-H method described by Gaines and Mitchell (1979).
Total S was determined by combustion using a LECO SC-132 sulphur analyzer.
Inorganic SO4 was extracted with dilute, boiling HCl and determined colorimetrically on a hydriodic acid – bismuth reducible distillate (Johnson and Nishita 1952).
Ministry of Environment
Analysis of total N was by combustion using a Fisons NA-1500 elemental analyzer. All other macro- and micro-nutrients were wet-ashed with concentrated nitric acid – HCl and hydrogen peroxide, using a closed-vessel microwave digestion system (Kalra and Maynard 1991). The digest solutions were diluted with HCl and individual nutrients were determined by inductively coupled plasma (ICP) optical emission spectrophotometer.
Total S was determined using two different methodologies – by combustion using a Fisons NA-1500 elemental analyzer and by wet-ashing (as above) followed by determination by ICP.
Sulphate-S was extracted with dilute HCl and determined by ion chromatography (Waters IC system).
Data Analysis
For any given nutrient, the results provided from the PSAI laboratory are not functionally dependent on the results from the MoE laboratory, or vice versa. Lacking a truly dependent relationship between two variables, regression analysis is arguably an inappropriate statistical technique. However, simple regression offers a convenient way to quantify the relationship between the sample results from each laboratory.
For each individual nutrient, both linear and polynomial models were tested. Across the data ranges tested, both models gave similar fits. For simplicity, and to possibly provide more conservative estimates when extrapolating outside the data range of the model, only linear models were used in the subsequent normalization process. For each nutrient, the “normalization rationale” discussed below was used to determine which laboratory would serve as the dependent variable in the regression analysis. Regression lines were not forced through the origin.
Normalization Rationale
Total Nitrogen
Published foliar N interpretative criteria for most conifer species, both in BC and elsewhere, have used analytical results obtained from wet digestion (i.e., modified Kjeldahl) procedures (Ballard and Carter 1986). Because the MoE laboratory uses combustion methodology for N analysis (which, as noted previously, typically produces higher N values), the MoE inter-laboratory results were “normalized” (i.e., adjusted downward) to facilitate interpretation. The regression equation defining the relationship between PSAI (dependent variable) and MoE (independent variable) foliar N results (r2=0.78) was used to convert the “raw” N data provided by the MoE laboratory to a lower “normalized” value (Figure 1). As stated previously, this does not mean than N analyses undertaken by the MoE are inferior to PSAI data. The “normalization” of the MoE data simply recognizes known differences in analytical results based on different methodology, so that the adjusted N values are consistent with wet digestion N values from which N interpretative criteria are typically developed.
Total Sulphur
Published foliar S interpretative criteria have typically used analytical results obtained from combustion procedures (Ballard and Carter 1986). As noted previously, dry combustion typically yields higher S values than wet digestion analytical methods. The PSAI laboratory uses dry combustion to determine total S. The MoE laboratory reports total S from both wet (S-ICAP) and dry (S-comb) analytical methodologies. The results from the inter-laboratory comparison indicated that the relationship between the S-comb (MoE) method and the PSAI dry combustion method is relatively weak (r2=0.53) (Figure 2), as is the relationship between the MoE wet (S-ICAP) and dry (S-comb) methods (r2=0.46) (Figure 3). Conversely, the relationship between S-ICAP (MoE) wet digestion method and the PSAI dry combustion method is much stronger (r2=0.76) (Figure 4). Therefore, the regression equation defining the relationship between PSAI total S (dependent variable) and MoE (S-ICAP) wet digestion total S (independent variable) was used to convert the “raw” S-ICAP (i.e., wet digestion) data provided by the MoE laboratory to a higher “normalized” value for interpretative purposes (Figure 4). NOTE: The total S-comb (i.e., combustion) values reported by MoE should not be used in the “normalization” spreadsheet.
Sulphate-Sulphur
The use of foliar SO4 as a diagnostic tool for the evaluation of tree S status was pioneered by research with radiata pine (Pinus radiata D. Don) and Douglas-fir (Kelly and Lambert 1972; Turner et al. 1977, 1979). Kelly and Lambert’s (1972) original procedure involved boiling acid extraction of SO4 and precipitation with barium. The SO4 concentration was calculated indirectly after determination of the barium concentration by atomic absorption (AA) spectrophotometry. The laborious and time-consuming method was later modified so that SO4 was quantified in the extracting solution by ion chromatography (IC) (M. Lambert, per. commun.). The modified IC method was later adopted by the MoE laboratory for SO4 determination. The PSAI laboratory uses a similar method to extract SO4 from foliage, but determination is by hydriodic acid (HI) reduction of the extract and bismuth colorimetry using the procedure of Johnson and Nishita (1952). In the inter-laboratory comparison, the IC and HI methods gave similar results and the relationship between the two methods was very strong (r2=0.91). Both methods have been used in British Columbia, but most research with SO4 as a diagnostic indicator of foliar S status has been based on the PSAI method (Brockley 2000a,b). Therefore, the regression equation defining the relationship between PSAI SO4 (dependent variable) and MoE SO4 (independent variable) was used to convert the “raw” SO4 data provided by the MoE laboratory to a slightly lower “normalized” value for interpretative purposes (Figure 5).
Phosphorus, Potassium, Calcium, Magnesium
The MoE and PSAI laboratories extract P, K, Ca, and Mg with different acid digestion methods. At the MoE lab, determination is by ICP, whereas determination at PSAI is by AA spectrophotometry. In the inter-laboratory comparison, the PSAI results for all four nutrients were typically higher than the MoE results within the ranges tested. For K, Ca, and Mg, the inter-laboratory comparison indicated that the relationships between the results from the two laboratories were relatively strong (r2=0.74–0.79) (Figures 6–8). For P, the relationship was somewhat weaker (r2=0.61) (Figure 9).
Few fertilization research experiments have been undertaken to verify inferred deficiency levels for P, K, Ca, and Mg in BC forests. For Douglas-fir, laboratory analysis has typically used methodology consistent with that used by PSAI (Carter and Klinka 1988; Carter et al. 1998). Both analytical methods have been used for lodgepole pine and interior spruce foliage samples collected from research trials involving fertilization with these nutrients. However, development of revised interpretative criteria has mostly been based on MoE methodology (Brockley 2007a, 2010a,b). In the “normalization” spreadsheet, PSAI “raw” data for P, K, Ca, and Mg are converted to slightly lower “normalized” values for interpretative purposes (Figures 6–9).