Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Detection of nutrient stress in Douglas-fir seedlings using spectroradiometer data Bracher, Grant Allan 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A1 B68.pdf [ 12.51MB ]
Metadata
JSON: 831-1.0100495.json
JSON-LD: 831-1.0100495-ld.json
RDF/XML (Pretty): 831-1.0100495-rdf.xml
RDF/JSON: 831-1.0100495-rdf.json
Turtle: 831-1.0100495-turtle.txt
N-Triples: 831-1.0100495-rdf-ntriples.txt
Original Record: 831-1.0100495-source.json
Full Text
831-1.0100495-fulltext.txt
Citation
831-1.0100495.ris

Full Text

DETECTION OF NUTRIENT STRESS IN DOUGLAS-FIR SEEDLINGS USING SPECTRORADIOMETER DATA BY i GRANT ALLAN BRACHER .Sc.(Renewable Resources), M c G i l l U n i v e r s i t y , 198 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF THE FACULTY OF GRADUATE STUDIES (Department of Forestry/Remote Sensing) We accept t h i s t h e s i s as conforming t o the r e q u i r e d standard THE UNIVERSITY OF BRITISH COLUMBIA September, 1991 fc\ GRANT ALLAN BRACHER DOCTOR OF PHILOSOPHY i n In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada Date op c^O^ft^V VY\ \ DE-6 (2/88) ABSTRACT Narrow-band spectral reflectance measurements in the visible and near-infrared region of the spectrum were investigated for the detection of nutrient deficiencies, and estimating the f o l i a r concentrations of nitrogen, phosphorus, sulphur, and total chlorophyll. One-year-old Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) seedlings were treated with 24 nutrient solutions containing nitrogen, phosphorus and sulphur levels ranging from 1 to 400 mg/L. After one growing season, newly matured needles were harvested, spectral reflectance measured from 400 to 1100 nm, and f o l i a r samples analyzed for nutrient and chlorophyll levels. Several nutrient deficiencies were diagnosed. There were no unique changes in spectral reflectance which could be attributed to a specific nutrient deficiency; rather changes in reflectance were non-specific responses influenced by how varying nutrient levels affected total chlorophyll concentration. A l l deficiencies caused decreases in total chlorophyll, thus demonstrating the value of total chlorophyll as an indicator of nutrient stress. Correlation coefficients were calculated describing the degree of association between f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll, and the following spectral parameters: the wavelength of the red edge, percent spectral reflectance at 554 and 630 nm (red rise), and 15 vegetation indices (Vis). Vis were f i r s t determined using combinations of spectral reflectances at 480, 554, 800 nm and the wavelength of the red well (674 nm; called red well Vis), and then recalculated using red rise (called red rise Vis) rather than red well measurements to see i f greater correlation with f o l i a r constituents could be obtained. Although the use of red rise measurements in the calculation of Vis 1, 2, 3, 12 and 14 resulted in higher correlation coefficients, differences between coefficients were seldom s t a t i s t i c a l l y significant. Red well and red rise VI15 were most correlated with the needle nitrogen content of individual Douglas-fir, red rise VI10 with phosphorus, and spectral reflectance at 554 nm and red well VI15 with total chlorophyll. These parameters proved useful indicators of relative nitrogen, phosphorus and total chlorophyll content. None of the Vis or other spectral parameters followed the same relationship with f o l i a r sulphur for the sulphur, nitrogen and phosphorus treatments; consequently, none were deemed suitable for sulphur estimation. i i i TABLE OP CONTENTS Page i i iv v i i x i i xvi 1. INTRODUCTION 1 2. LITERATURE REVIEW 3 2.1 SPECTRAL PROPERTIES OF VEGETATION 3 2.2 TERMINOLOGY 6 2.3 IMPACT OF NUTRIENT STRESS ON SPECTRAL REFLECTANCE 9 2.4 RED RISE 13 2.5 RED EDGE 14 2.6 VEGETATION INDICES 16 2.7 CHLOROPHYLL A TO B RATIO 21 3. HYPOTHESES 22 4. METHODS 24 4.1 1987 24 4.2 1988 28 5. RESULTS AND DISCUSSION 31 5.1 NITROGEN SERIES OF TREATMENTS 31 5.1.1 Seedling Growth 31 5.1.2 Foliar Analysis ' 33 I ABSTRACT II TABLE OF CONTENTS III LIST OF TABLES IV LIST OF FIGURES V ACKNOWLEDGEMENTS iv Page 5.1.3 Spectral Reflectance Curves 39 5.1.4 Chlorophyll a/Chlorophyll b 44 5.1.5 Green Reflectance Peak 49 5.1.6 Red Rise 53 5.1.7 Red Edge 56 5.1.8 Vegetation Indices 59 5.1.9 Estimation of Foliar Nitrogen 64 5.2 PHOSPHORUS SERIES OF TREATMENTS 72 5.2.1 Seedling Growth 72 5.2.2 Foliar Analysis 72 5.2.3 Spectral Reflectance Curves 81 5.2.4 Chlorophyll a/Chlorophyll b 87 5.2.5 Green Reflectance Peak and Red Rise 91 5.2.6 Red Edge 99 5.2.7 Vegetation Indices 101 5.2.8 Estimation of Foliar Phosphorus 106 5.3 SULPHUR SERIES OF TREATMENTS 114 5.3.1 Seedling Growth 114 5.3.2 Foliar Analysis 116 5.3.3 Spectral Reflectance Curves 123 5.3.4 Chlorophyll a/Chlorophyll b 126 5.3.5 Green Reflectance Peak and Red Rise 130 5.3.6 Red Edge 134 5.3.7 Vegetation Indices 136 . 5.3.8 Estimation of Foliar Sulphur 140 v Page 5.4 CHLOROPHYLL 147 5.4.1 Green Reflectance Peak and Red Rise 147 5.4.2 Red Edge 150 5.4.3 Vegetation Indices 150 5.4.4 Estimation of Total Chlorophyll 152 5.4.5 Total Chlorophyll 164 6. SUMMARY OF RESULTS, CONCLUSIONS AND RECOMMENDATIONS 168 6.1 TESTING OF HYPOTHESES 168 6.2 NEED FOR NARROW BAND SENSORS 177 LITERATURE CITED 183 APPENDICES 198 APPENDIX A. COMMON AND SCIENTIFIC NAMES 199 APPENDIX B. NUTRIENT SOLUTION COMPOSITION 2 02 APPENDIX C. FOLIAR NUTRIENT CONCENTRATIONS 2 27 APPENDIX D. CHLOROPHYLL A, B and TOTAL CHLOROPHYLL 232 CONCENTRATIONS vi LIST OF TABLES Page 1 Examples of vegetation indices. 17 2 New vegetation indices developed from 19 those of Kleman and Fagerlund (1981). 3 Mean nitrogen growth data. 32 4 Composite f o l i a r nitrogen and total 34 chlorophyll concentrations for the 1987 nitrogen series of treatments. 5 Composite fo l i a r nitrogen and total 3 5 chlorophyll concentrations for the 1988 nitrogen series of treatments. 6 Foliar nitrogen and total chlorophyll 36 concentrations for individual Douglas-fir seedlings subjected to the 1988 nitrogen series of treatments. 7 Correlation between total chlorophyll and 4 4 the f o l i a r concentration of nitrogen. 8 Percent spectral reflectance at the green 50 reflectance peak and red rise, and wavelength of the red edge for composite f o l i a r samples of the 1987 nitrogen series of treatments. 9 Percent spectral reflectance at the green 51 reflectance peak and red rise, and wavelength of the red edge for composite fo l i a r samples of the 1988 nitrogen series of treatments. 10 Percent spectral reflectance at the green 52 reflectance peak and red rise, and wavelength of the red edge for individual Douglas-fir ' seedlings subjected to the 1988 nitrogen series of treatments. 11 Correlation between the f o l i a r concentration of 57 nitrogen and green reflectance peak, red rise, and red edge measurements. 12 Correlation between the f o l i a r concentration of 60 nitrogen and vegetation indices calculated using percent spectral reflectance at the red well and red rise. v i i Page 13 Changes in correlation between f o l i a r nitrogen 63 concentration and recalculated vegetation indices using percent spectral reflectance at the red rise. 14 Comparison of f o l i a r nitrogen levels for 1988 65 individual seedling data as determined by chemical analysis and the red well vegetation index 15 model. 15 Comparison of f o l i a r nitrogen levels for 1988 66 individual seedling data as determined by chemical analysis and the red rise vegetation index 15 model. 16 Comparison of f o l i a r nitrogen levels for 68 composite samples as determined by chemical analysis and the red rise vegetation index 13 model. 17 Comparison of fo l i a r nitrogen levels for 69 composite samples as determined by chemical analysis and the red well vegetation index 15 model. 18 DIAGFOLI's interpretation of Douglas-fir f o l i a r 71 nitrogen concentration. 19 Mean phosphorus growth rate. 73 20 Composite f o l i a r phosphorus and total 75 chlorophyll concentrations for the 1987 phosphorus series of treatments. 21 Composite f o l i a r phosphorus and total 76 chlorophyll concentrations for the 1988 phosphorus series of treatments. 22 Foliar phosphorus and total chlorophyll 77 concentrations for individual Douglas-fir seedlings subjected to the 1988 phosphorus series of treatments. 23 Correlation between total chlorophyll and the 85 fo l i a r concentration of phosphorus. v i i i Page 24 Percent spectral reflectance at the green 92 reflectance peak and red rise, and wavelength of the red edge for composite f o l i a r samples of the 1987 phosphorus series of treatments. 25 Percent spectral reflectance at the green 93 reflectance peak and red rise, and wavelength of the red edge for composite f o l i a r samples of the 1988 phosphorus series of treatments. 2 6 Percent spectral reflectance at the green 94 reflectance peak and red rise, and wavelength of the red edge for individual Douglas-fir seedlings subjected to the 1988 phosphorus series of treatments. 27 Correlation between the f o l i a r concentration 98 of phosphorus and green reflectance peak, red rise, and red edge measurements. 28 Correlation between the f o l i a r concentration 103 of phosphorus and vegetation indices calculated using percent spectral reflectance at the red well and red rise. 29 Changes in correlation between f o l i a r 105 phosphorus concentration and recalculated vegetation indices using percent spectral reflectance at the red rise. 30 Comparison of f o l i a r phosphorus levels for 109 1988 individual seedling data as determined by chemical analysis and models developed using red rise vegetation index 10. 31 Comparison of f o l i a r phosphorus levels for 110 composite samples as determined by chemical analysis and models developed using red rise vegetation index 6. 32 Comparison of f o l i a r phosphorus levels for 111 composite samples as determined by chemical analysis and models developed using red rise vegetation index 10. 33 DIAGFOLI's interpretation of Douglas-fir 113 f o l i a r phosphorus concentrations. ix Page 34 Mean sulphur growth rate. 115 3 5 Composite f o l i a r sulphur and total 117 chlorophyll concentrations for the 1988 sulphur series of treatments. 3 6 Foliar sulphur and total chlorophyll 118 concentrations for individual Douglas-fir seedlings subjected to the 1988 sulphur series of treatments. 37 Correlation between total chlorophyll and 126 the f o l i a r concentration of sulphur. 38 Percent spectral reflectance at the green 132 reflectance peak and red rise, and wavelength of the red edge for composite f o l i a r samples of the 1988 sulphur series of treatments. 39 Percent spectral reflectance at the green 133 reflectance peak and red rise, and wavelength of the red edge for individual Douglas-fir seedlings subjected to the 1988 sulphur series of treatments. 40 Correlation between the fo l i a r concentration of 134 sulphur and green reflectance peak, red rise, and red edge measurements. 41 Correlation between the fo l i a r concentration of 138 sulphur and vegetation indices calculated using percent spectral reflectance at the red well and red rise. 42 Changes in correlation between f o l i a r sulphur 139 concentration and recalculated vegetation indices using percent spectral reflectance at the red rise. 43 Comparison of f o l i a r sulphur levels for 1988 141 individual seedling data as determined by chemical analysis and the red well vegetation index 13 model. 44 Comparison of f o l i a r sulphur levels for 1988 142 composite seedling data as determined by chemical analysis and the red rise vegetation index 12 model. x Page 45 Correlation between total chlorophyll 149 concentration and green reflectance peak, red rise, and red edge measurements. 46 Correlation between total chlorophyll 153 concentration and vegetation indices calculated using percent spectral reflectance at the red well. 47 Correlation between total chlorophyll 155 and vegetation indices calculated using percent spectral reflectance at the red rise. 48 Changes in correlation between total 157 chlorophyll and recalculated vegetation indices using percent spectral reflectance at the red rise. 49 Comparison of total chlorophyll levels as 159 determined by chemical analysis and the green reflectance peak model. 50 Comparison of total chlorophyll levels as 160 determined by chemical analysis and the red well vegetation index 15 model. 51 Models for estimating total chlorophyll and 161 associated standard errors of prediction. 52 Spectral bands of the MEIS vegetation stress 179 f i l t e r set. xi LIST OF FIGURES Generalized spectral reflectance curve for a green leaf and dominant controlling leaf reflectance in the visible and near-infrared spectral regions. Generalized relationship between growth and tissue nutrient concentration. First derivative spectrum of Douglas-fir needles. 1987 spectral reflectance curves showing the effects of nitrogen levels on needle reflectance at the wavelength interval 400 to 1100 nm. 1988 spectral reflectance curves showing the effects of nitrogen levels on needle reflectance at the wavelength interval 400 to 1100 nm. Relationship between total chlorophyll and f o l i a r nitrogen concentration for the 1988 individual seedling data set. Relationships between the ratio of chlorophyll a/chlorophyll b, fo l i a r nitrogen and severity of nutrient deficiency for the 1987 nitrogen composite data set. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r nitrogen and severity of nutrient deficiency for the 1988 nitrogen composite data set. Relationships between the ratio of chlorophyll a/chlorophyll b, fo l i a r nitrogen and severity of nutrient deficiency for the 1988 nitrogen individual seedling data set. Relationship between fo l i a r nitrogen concentration and percent spectral reflectance at the green reflectance peak for the 1988 individual seedling data set. Relationship between the logarithm of f o l i a r nitrogen concentration and the logarithm of percent spectral reflectance at the green reflectance peak for the 1988 individual seedling data set. x i i Page 12 Relationship between fo l i a r nitrogen and the 58 wavelength of the red edge for the 1988 individual seedling data set. 13 1987 spectral reflectance curves showing the 82 effects of phosphorus levels on needle reflectance at the wavelength interval 400 to 1100 nm. 14 1988 spectral reflectance curves showing the 83 effects of phosphorus levels on needle reflectance at the wavelength interval 400 to 1100 nm. Plotted data are from the 1988 composite data set. 15 Relationship between total chlorophyll and f o l i a r 86 phosphorus concentration for the 1988 individual seedling data set. 16 Relationships between the ratio of chlorophyll 88 a/chlorophyll b, f o l i a r phosphorus and severity of nutrient deficiency for the 1987 phosphorus composite data set. 17 Relationships between the ratio of chlorophyll 89 a/chlorophyll b, f o l i a r phosphorus and severity of nutrient deficiency for the 1988 phosphorus composite data.set. 18 Relationships between the ratio of chlorophyll 90 a/chlorophyll b, f o l i a r phosphorus and severity of nutrient deficiency for the 1988 phosphorus individual seedling data set. 19 Relationship between f o l i a r phosphorus 95 concentration and percent spectral reflectance at the red rise for the 1988 individual seedling data set. 20 Relationship between logarithm(0.80 - percent 97 f o l i a r phosphorus) versus percent spectral reflectance at the red rise for the 1988 individual seedling data set. 21 Relationship between f o l i a r phosphorus and 100 the wavelength of the red edge for the 1988 individual seedling data set. x i i i 22 Relationship between fo l i a r phosphorus 102 concentration and red well vegetation index 2 (NIR - red)/(NIR + red) for the 1988 individual seedling data set. 23 Regression lines and associated data for the 107 relationship between f o l i a r phosphorus concentration and red rise vegetation index 10 for the 1988 individual seedling data set. 24 1988 spectral reflectance curves showing the 124 effects of sulphur levels on needle reflectance at the wavelength interval 400 to 1100 nm. Plotted data are from the 1988 composite data sets. 25 Relationship between total chlorophyll and 127 f o l i a r sulphur concentration for the 1988 individual seedling data set. 26 Relationships between the ratio of chlorophyll 128 a/chlorophyll b, f o l i a r sulphur and severity of nutrient deficiency for the 1988 sulphur composite data set. 27 Relationships between the ratio of chlorophyll 129 a/chlorophyll b, fo l i a r sulphur and severity of nutrient deficiency for the 1988 sulphur individual seedling data set. 28 Relationship between fo l i a r sulphur 131 concentration and percent spectral reflectance at the green reflectance peak for the 1988 individual seedling data set. 29 Relationship between f o l i a r sulphur and the 13 5 wavelength of the red edge for the 1988 individual seedling data set. 30 Relationship between f o l i a r sulphur concentration 137 and red well vegetation index 13 (NIR/(green + red + NIR)) for the 1988 sulphur individual seedling data set. 31 Relationship between sulphur f o l i a r concentration 144 and red well vegetation index 13 (NIR/(green + red + NIR)) for the 1988 nitrogen individual seedling data set. xiv Page 32 Relationship between f o l i a r sulphur concentration 145 and red well vegetation index 13 (NIR/(green + red + NIR)) for the 1988 phosphorous individual seedling data set. 33 Relationship between total chlorophyll 148 concentration and percent spectral reflectance at the green reflectance peak for the 1988 nitrogen individual seedling data set. 34 Relationship between total chlorophyll and the 151 wavelength of the red edge for the 1988 nitrogen individual seedling data set. 35 Relationships between total chlorophyll, 165 percent spectral reflectance at the red rise and severity of nutrient deficiency for the 1988 nitrogen individual seedling data set. 3 6 Relationships between total chlorophyll, 166 percent spectral reflectance at the red rise and severity of nutrient deficiency for the 1988 phosphorus individual seedling data set. 3 7 Relationships between total chlorophyll, 167 « percent spectral reflectance at the red rise and severity of nutrient deficiency for the 1988 sulphur individual seedling data set. xv ACKNOWLEDGEMENTS Each graduate student owes part of their intellectual evolution to contact with other professionals in the f i e l d . I have been most fortunate in having Dr. Peter Murtha as a research supervisor; not only did I greatly benefit from his guidance and support during my stay at the University of British Columbia, but also, from his insights into the study of vegetation stress and remote sensing. I am grateful to Drs. Tim Ballard, Karel Klinka, John Worrall and Vernon Singhroy, my committee members, for their encouragement and advice at every stage of the study. I sincerely express my gratitude to the staff of the Ontario Centre for Remote Sensing and the Canada Centre for Remote Sensing for the trust they showed when lending me their spectroradiometers. Without these instruments the research could not have taken place. I also wish to thank the staff of Pacific Soil Analysis Incorporated and Can Test Limited for their rapid analysis of f o l i a r samples and explanations of techniques used. Finally, I must acknowledge the tremendous emotional support provided by my wife, Wenda, who on our f i r s t wedding anniversary presented me with a personal computer thus forcing me to learn how to operate this invaluable tool. I am grateful to the patience demonstrated by Jerry Maedel when answering my personal computing questions. Funding was provided by the Canadian Forestry Service (CFS) through their Canadian Forestry Scholarship Program, and by funds from a CFS Human Resources Program block grant to the Faculty of Forestry, University of British Columbia. xvi 1. INTRODUCTION To meet the world's future needs for wood products and to ensure that British Columbia maintains or increases i t s share of this market, methods must be adopted for increasing production while maintaining competitive prices and preserving environmental quality. Increases in forest productivity from a fixed or decreasing land base can be achieved through the use of genetically improved planting stock, converting natural stands of slow-growing species to plantations of more desirable species, or f e r t i l i z i n g existing stands. Of these, only f e r t i l i z a t i o n of existing stands offers the possibility of economic gain within as short a time as five to ten years (Heilman, 1971). Nutrient deficiencies are not uncommon in the forests of British Columbia. Moderate to severe nitrogen deficiencies are widespread, and the most serious nutritional problem (Ballard, 1985; Ballard and Carter, 1986). According to Ballard (1985) sulphur and phosphorus rank next, though substantially lower in importance, among macronutrient deficiencies. Micronutrient deficiencies of iron and copper have been demonstrated in B.C. interior spruce and lodgepole pine stands1 (Ballard, 1985; Ballard and Majid, 1985; Majid and Ballard, 1990), and f o l i a r analyses of coastal Douglas-fir, western hemlock and Pacific f i r stands suggest that boron and zinc deficiencies may be more widespread than commonly thought (Carter et al., 1986; Carter and Klinka, 1986). Scientific names of species mentioned are listed in Appendix A. 2 Successful forest f e r t i l i z a t i o n programs require the development of rapid and cost-effective techniques for identifying those stands in need of f e r t i l i z a t i o n and the determination of the limiting element(s). Laboratory techniques for f o l i a r analysis are time-consuming and expensive. While visual observations are inexpensive and commonly used in forestry and agriculture to estimate nutrient deficiencies qualitatively (Morrison, 1974; Mengel and Kirkby, 1987; Marschner, 1986), not a l l deficiencies exhibit distinctive symptoms; moreover, nutrient deficiencies severe enough to impair growth are not always severe enough to cause visib l e symptoms (Ballard and Carter, 1986) . Spectroscopic assays are the primary analytical method used in laboratory research for the identification and quantification of plant pigments, nitrogen, lignin, cellulose, and other biochemical components of leaves (Curtiss and Ustin, in press). Consequently, i t is not surprising that crop physiologists have approached the challenge of fast and early detection of nutrient deficiencies in agricultural crops by u t i l i z i n g remote sensing techniques which relate leaf spectral reflectance to physiological properties. Despite the potential benefits of such research, similar studies have been less common in forestry. This thesis explores the possibility of using visible and near-infrared spectral reflectance measurements to assess the nutrient status of Douglas-fir seedlings. 3 2. LITERATURE REVIEW 2.1 SPECTRAL PROPERTIES OF VEGETATION As solar radiation comes into contact with a plant, i t is primarily intercepted by leaves, and i t i s largely the spectral reflectance from these structures that i s detected by remote sensing systems. The interaction of incident solar radiation and green leaves throughout the electromagnetic spectrum has been well documented. Gates et al. (1965) discuss the spectral properties of plants, Gates (1970) the physical and physiological properties of vegetation, Woolley (1971) the reflectance and transmittance of light by leaves, and Gausman (1974; 1977) the reflectance of near-infrared light by leaves and leaf components. A generalized spectral reflectance curve for a healthy green leaf is shown in Figure 1. Plant pigments, primarily chlorophyll a and chlorophyll b, are the main factors influencing reflectance and absorption of radiation in the visible region of the spectrum. Although each pigment absorbs energy of different wavelengths, the main absorption features are centred at approximately 480 and 680 nm. These are the result of strong photon absorption involving electronic transitions in the chlorophyll molecule centred around the magnesium component of the photoactive site (Goetz et al., 1983) . In contrast, there is high reflectance of green light by chlorophyll, peaking at about 550 nm, thus causing the familiar green colour of leaves. Throughout this thesis the wavelengths of greatest green spectral reflectance and least red reflectance are 4 I PIGMENTS STRUCTURE 1 LEAF WATER CONTENT DOMINANT FACTOR CONTROLLING LEAF REFLECTANCE 60 50 40 £ 30 — i < oc S> 20 NEAR INFRARED PLATEAU GREEN REFLECTANCE PEAK \ — RED EDGE RED WELL _J_ _1_ _!_ JL ( VISIBLE 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 WAVELENGTH (nm) X 100 NEAR INFRARED SHORT WAVE INFRARED SPECTRAL , REGION Figure 1. Generalized spectral reflectance curve for a green leaf and dominant factors c o n t r o l l i n g leaf reflectance i n the v i s i b l e and near-infrared spectral regions. Adapted from Goetz et al., (1983). 5 referred to as the "green reflectance peak" and "red well", respectively. High reflectance in the near-infrared plateau is primarily due to multiple refractions occurring at the interface of hydrated c e l l walls with intercellular spaces, as a result of differing refraction indices (Gates, 1970; Woolley, 1971; Kumar and Silva, 1973; Gausman, 1974). Gausman (1977) reported that leaf components, including stomata, nuclei, c e l l walls, crystals and chloroplasts, directly contribute a small percentage of the reflectance within the 700 to 1100 nm wavelength range. Spectral characteristics of the shortwave infrared region are dominated by absorption of energy by liquid water within leaves (Gates, 1970; Knipling, 1970; Myers et al., 1970). Strong absorption bands occur by water at 1450 and 1900 nm (Gates et al., 1965; Gausman et al., 1969; Knipling, 1970; Woolley, 1971), and according to Tucker (1980) the 1650 to 2200 nm region can provide an accurate indication of leaf water content. Spectral reflectance properties of vegetation canopies in f i e l d situations are modified, both quantitatively and qualitatively, from those of individual leaves due to variations in leaf orientation, shadows, and background surfaces, such as s o i l . Knipling (1970), Colwell (1974), Jarvis et al. (1976), Curran and Milton (1983), and Goel (1988) have discussed these effects in detail. 6 2.2 TERMINOLOGY Before exploring the potential of remote sensing to detect nutrient stress and deficiencies i t i s necessary to c l a r i f y the meaning of these terms. It is generally accepted that the term "deficiency" means there is a shortage or lack of an element; however, deficiency and sufficiency are not discrete conditions. Everard (1973) and others have shown how the relationship between growth and the level of a particular nutrient can be described by a growth response curve (Figure 2). A central concept used in the interpretation of growth response curves i s the " c r i t i c a l range" (BC in Figure 2) of each nutrient for each plant species (Macy, 1936; Barrows, 1959). Below this range (AB) trees w i l l likely show deficiency symptoms and respond greatly to applications of the deficient element. Above the c r i t i c a l range (CD) there is luxury consumption, where further nutrient additions do not result in increased growth. Segment DE of the growth response curve represents the toxicity range, where further increases in nutrient level result in a reduction in yield. 7 Figure 2. Generalized relationship between growth and tissue nutrient concentration. Adapted from Everard (1973). 8 Use of the term "stress" has caused concern for remote sensing specialists and plant physiologists alike. Sometimes stress has been used to indicate the external force that causes physiological and morphological changes in the plant, whereas in other cases, i t is used to indicate the changes themselves. To resolve this problem the definitions of Levitt (1972) and Murtha (1982) have been adopted, and presented below: Stress: any environmental factor capable of inducing a potentially injurious strain on a plant (Levitt, 1972) ; Strain: any physical or chemical change in a plant produced by stress (Levitt, 1972); Injury: any stress on a plant that causes a detrimental strain that may be noted because of either temporary or permanent syndromes (Murtha, 1982) ; Damage: any loss, either biologic or economic, due to stress (Murtha, 1982). Therefore, according to Levitt's (1972) and Murtha•s (1982) definitions, nutrient concentrations represented by the deficient or toxic ends of the growth response curve indicate potential stresses capable of inducing a strain which could result in injury 9 or damage to the plant. 2.3 IMPACT OF NUTRIENT STRESS ON SPECTRAL REFLECTANCE Once leaves have fu l l y developed, their spectral properties remain f a i r l y stable unless affected by stress or naturally by aging. Several laboratory studies of nutrient stress on single leaves have been performed. Thomas et a l . (1966) found that decreasing concentrations of available nitrogen resulted in an increase of cotton leaf spectral reflectance in the visible (400 to 700 nm) region, and a decrease in both the near-infrared (700 to 1400 nm) and middle-infrared (1400 to 2500 nm) wavelengths. In contrast, Thomas and Oerther (1972) reported that as nitrogen deficiency increased, sweet pepper leaves increased in visible, near-infrared and middle-infrared reflectance. Mexican squash deficient in nitrogen, potassium, sulphur and magnesium were reported by Gausman et al. (1973) to have higher spectral reflectance in the 500 to 650 nm and near-infrared wavelength intervals than the leaves of control squash. Phosphorus-deficient squash leaves had slightly lower reflectance in the visible region but greater near-infrared reflectance compared to control plants, whereas iron-deficient squash had much greater visib l e reflectance and lower near-infrared reflectance than the leaves of controls. 10 Al-Abbas et al. (1974) measured light absorption of maize plants subject to several nutrient deficiencies and found that leaves deficient in nitrogen, phosphorus and potassium exhibited higher visi b l e reflectance than control plants, and that while some nitrogen and phosphorus-deficient leaves had low reflectance in the near-infrared, potassium deficient leaves had high near-infrared reflectance. Calcium, magnesium and sulphur deficiencies resulted in higher visible reflectance than control maize plants; however, near-infrared reflectance decreased with increasing magnitude in the following order for various deficiencies: calcium, magnesium, and sulphur. Research on the effects of varying nitrogen levels on crop canopies under f i e l d conditions have shown consistent results. Reductions in the amount of available nitrogen have caused increases in visible spectral reflectance and reductions in near-infrared for the canopies of maize (Walburg et al. , 1981; 1982), winter wheat (Hinzman et al., 1984; Nilsson, 1987; Nilsson and Linner, 1987), barley (Kleman, 1985; Demetriades-Shah and Court, 1987; Nilsson and Linner, 1987), buffelgrass (Everitt et al., 1985), and a l i c i a grass (Richardson et al., 1983). Densitometric measurements from aerial photographs have proven unsuccessful for the detection of specific nutrient deficiencies and estimation of f o l i a r nutrient levels. Rennie and Buckner (1972) simultaneously exposed Kodak aerial films (Ektachrome 2448, black and white infrared 5424, and colour infrared 2443) at an altitude of 457 meters (neither scale nor camera focal length were reported) over a loblolly pine plantation subjected to twelve treatments, ranging from the application of no f e r t i l i z e r to f e r t i l i z a t i o n with 55 kg nitrogen, 6.5 kg phosphorus and 18 kg potassium/ha. No significant differences in mean optical density were found between treatments or nutrient levels for any combination of film, camera f i l t e r and microdensitometer f i l t e r (Rennie and Buckner, 1972; Rennie and Cress, 1975). Murtha and Ballard (1983) assessed the use of densitometric data from 1:1,200 colour infrared (Kodak Aerochrome 2443) aerial photographs for detecting nutrient deficiencies in a Douglas-fir stand. Three different classes were identified based on growth rates, as expressed by the varying magenta hue of the tree crowns. Foliar analysis showed significant nutritional differences (especially for nitrogen, magnesium, and calcium) between classes; however, no consistent correlation was found between specific nutrient deficiencies and densitometric data. A series of related experiments reported in Card et al. (1988), Peterson et al. (1988) and Wessman et al. (1988a, 1988b, 1989) measured near-infrared reflectance from dried and ground-up leaves, whole leaves, and forest canopies to determine i f reflectance in the 1200 to 2400 nm range i s influenced by biochemical characteristics. High correlation was found between f o l i a r concentrations of lignin and nitrogen, and the f i r s t 12 derivative 2 at several near-infrared wavelengths, thus demonstrating the potential of spectral reflectance in the 1200 to 2400 nm range for estimating lignin and nitrogen. Few studies have investigated the possibility of using spectral reflectance measurements to estimate f o l i a r nutrient and chlorophyll levels. Benedict and Swidler (1961) reported an inverse relationship between visible spectral reflectance and the chlorophyll concentration of soybean and Valencia orange leaves, and demonstrated that reflectance measurements could be used to follow changes in chlorophyll concentration. Thomas and Oerther (1972) conducted a f e a s i b i l i t y study to determine the nitrogen status of sweet pepper leaves. Reflectance in the visi b l e wavelengths was found to be inversely correlated with leaf nitrogen concentration, and regression equations expressing spectral reflectance as a function of leaf nitrogen levels for greenhouse sweet peppers were successfully used to estimate the f o l i a r nitrogen concentration of field-grown peppers. Tsay et al. (1982) and Nelson et al. (1986) made spectral reflectance measurements of loblolly pine seedlings in the 350 to 700 nm range of the electromagnetic spectrum. Both studies found that there was a strong positive correlation between nitrogen and chlorophyll concentration, that needle nitrogen concentration The derivative of a spectral signature is i t s rate of change with respect to wavelength. Therefore, the determination of a f i r s t order derivative spectrum involves the differentiation or calculation of the change in spectral reflectance divided by the change in wavelength plotted against wavelength. 13 was negatively correlated with reflectance, and that predictive equations for estimating nitrogen and chlorophyll as a function of reflectance can be formulated by regression analysis. The chlorophyll and nitrogen estimation technique described by Thomas and Oerther (1972), Tsay et al. (1982), and Nelson et al. (1986) shall be referred to as the green reflectance peak method for the remainder of the thesis. 2.4 RED RISE Although many published spectral reflectance curves of stressed vegetation demonstrate an interband red rise at about 633 nm (see Figure 1), only Ahern's (1988) study of mountain pine beetle stress on lodgepole pine makes specific mention of this phenomenon. The same spectral reflectance shift can be observed in the spectra of ponderosa pine under attack by mountain pine beetle (Heller, 1968), coniferous trees growing over a mineral deposit (Collins and Chiu, 1979), beech trees undergoing natural senescence (Knipling, 1969), and nitrogen deficient loblolly pine seedlings (Tsay et al., 1982; Nelson et a l . , 1986). Nelson et al. (1986) grew loblolly pine at eight different levels of nitrogen, ranging from 400 to 1 mg/L. With each decreasing nitrogen level there was a corresponding increase in the spectral reflection of the red rise region. This clearly indicates the potential of the red rise region as an indicator of nutrient stress and estimator of f o l i a r nitrogen levels. 14 2.5 RED EDGE The "red edge" is the sharp change in leaf reflectance which occurs in the 680 to 750 nm range of the spectrum ((see Figure 1) Horler et al., 1983). It is the result of internal leaf scattering causing high near-infrared reflectance and chlorophyll absorption giving low red reflectance (Barber and Horler,. 1981; Horler et al., 1983). Investigators have quantitatively defined the red edge as the wavelength of maximum rate of change .between 680 and 750 nm (see Figure 3) of the reflectance spectrum (Horler et al., 1983; Demetriades-Shah et al., 1990). Demetriades-Shah et al. (1990) discuss the use of high resolution derivative spectra in remote sensing. Gates et al. (1965) found that as white oak leaves matured over the growing season and chlorophyll concentration increased, the position of the red reflectance edge shifted systematically toward longer wavelengths. This "red shif t " was also reported by Collins (1978) over the growth cycle of crop plants such as wheat, cotton, and sugar beets. Geobotanists have noted a "shift" in the position of the red edge towards the blue end of the spectrum (referred to as a "blue shift") for trees and crops growing in soils enriched with trace metals (Howard et al., 1971; Chang and Collins, 1983; Collins et al., 1983; Horler et al., 1983; Milton et a l . , 1983; Rock and Vogelmann, 1985). These results demonstrate the potential of the red edge as a metal stress indicator. 15 Figure 3. F i r s t derivative spectrum of Douglas-fir needles. The wavelength of the red edge is defined as the wavelength of maximum dreflectance/dwavelength. Curtis and Ustin (1989) reported a decrease in chlorophyll concentration and a corresponding blue shift in the red edge of ponderosa pine foliage exposed to low levels of ozone pollution. However, high levels of ozone exposure produced a shift to longer wavelengths. Curtiss and Ustin (1989) hypothesized that this red shift was due to chloroplast membrane damage. The slope and position of the red edge are directly correlated with leaf chlorophyll concentration (Barber and Horler, 1981; Horler et al., 1983; Curtiss and Ustin, 1989; Cure, in press); consequently, red edge measurements may provide a new method for remotely determining f o l i a r nutrient status. 2.6 VEGETATION INDICES Several spectral vegetation indices have been developed to make inferences about the extent and condition of vegetation. Most vegetation indices are simple ratios or ratios of linear combinations of green, red, and near-infrared (NIR) spectral reflectance ground features. Presumably, more information is provided by two or more bands than any one spectral band. According to Vygodskaya et al. (1989) vegetation indices are useful in that they reduce the dimension of the i n i t i a l multiband information, and minimize the impact of illumination and viewing conditions. Examples of vegetation indices are presented in Table 1. 17 Table 1. Examples of vegetation indices. Symbol used in Formula based on this study reflectance data Reference VII* VI2b VI3C VI4 VI5 VI6 VI7 VI8 VI9 VI10 VIII VI12 NIR/Red (NIR-Red)/(NIR+Red) SQRT(VI2+0.5) Red/Green Green(NIR/Red) Red(NIR/Green) (Green-Red)/(Green+Red) (NIR-Green)/(NIR+Green) (VI7)(NIR) Red/(Blue+Green+Red) Green/(Blue+Green+Red) Blue/(Blue+Green+Red) Jordan, 1969 Pearson and Miller, 1972 Rouse et al., 1973 Vygodskaya et al., 1989 Vygodskaya et a l . , 1989 Vygodskaya et al., 1989 Vygodskaya et a l . , 1989 Vygodskaya et a l . , 1989 Vygodskaya et al., 1989 Kleman and Fagerlund, 1981 Kleman and Fagerlund, 1981 Kleman and Fagerlund, 1981 a Conventional name in literature is Ratio Vegetation Index (RVI) b Conventional name in literature is Normalized Difference Vegetation Index (NDVI) c Conventional name in literature is Transformed Vegetation Index (TVI) NIR = near-infrared SQRT = square root 18 The development of vegetation indices can not be considered complete. Spectral reflectance measurements from different regions of the electromagnetic spectrum can be combined and manipulated in an i n f i n i t e variety of ways to produce new vegetation indices. The substitution of near-infrared spectral reflectance for blue reflectance in the vegetation indices of Kleman and Fagerlund (1981), listed in Table 1, would maximize the difference between the absorption of red light by chlorophyll and high near-infrared spectral reflectance (see Table 2) , and could lead to the development of three new indices useful for the detection of vegetation nutrient stress. Quantitative assessment of vegetation by vegetation indices was f i r s t reported by Jordan (1969) who used the ratio vegetation index (NIR/red) to derive the leaf area index for forest canopies in a tropical forest. Subsequent work, primarily conducted on agricultural and rangeland environments, found significant correlations between the ratio vegetation index and normalized vegetation index ((NIR - red)/(NIR + red)), and such agronomic parameters as biomass, leaf area index, photosynthetic activity and grain yield, for barley (Kleman and Fagerlund, 1981, 1987; Demetriades-Shah and Court, 1987; Nilsson and Linner, 1987), wheat (Ajai et a l . , 1983A; Patel et al., 1983; Hinzman et al., 1984; Das et al., 1985; Kamat et al., 1985; Nilsson, 1987; Nilsson and Linner, 1987), maize (Tucker et al., 1979; Walburg et al., 1981, 1982), soybeans (Tucker et al., 1979), chickpea (Ajai et al.,1983B; Kamat et al. , 1985), and forage crops (Pearson et a l . , 1976; 19 Table 2. New vegetation indices developed from those of Kleman and Fagerlund (1981). Symbol used in Formula based on this study reflectance data VI13 NIR/(Green+Red+NIR) VI14 Red/(Green+Red+NIR) VI15 Green/(Green+Red+NIR) NIR = near-infrared Tucker, 1979; Richardson et al., 1983). In each of these studies the ratio vegetation and normalized vegetation indices showed high values for nitrogen f e r t i l i z e d plots as compared to nitrogen deficient plots. Jackson et al. (1980) noted that with adequate nitrogen, potassium and water, ratio vegetation index values were significantly higher than those measured over a nitrogen-deficient sugarcane plot and another plot deficient in potassium. Although past studies have indicated the potential of using vegetation indices as a means of estimating f o l i a r nitrogen, the only study to explore the possibility has been that of Richardson et al. (1983) who found significant correlations between the ratio vegetation and perpendicular vegetation (PVI = -0.8736 x red + 0.4866 x NIR) indices and the nitrogen concentration of a l i c i a grass. Richardson et al. (1983) used these relationships to develop regression equations to estimate f o l i a r nitrogen. Likewise the potential of using vegetation indices to estimate chlorophyll has been demonstrated. Significant correlations have been reported between the ratio vegetation and normalized vegetation indices and chlorophyll concentration for a number of agricultural crop canopies (Tucker, 1979; Ajai et al. , 1983B; Curran and Milton, 1983; Hinzman et al., 1984), and also for percent crop chlorosis (Tucker et al., 1979). Demetrides-Shah and Court (1987) noted that plant chlorophyll concentration is better predicted from off-nadir viewing reflectance measurements than from nadir viewing measurements, and that the NIR/red, NIR/green, green/red, and especially (NIR-red)/(NIR+red) vegetation indices are strongly related to chlorophyll concentration. Although the relationship between vegetation indices and chlorophyll has been established no one has yet used them to estimate chlorophyll. Clearly there is a need to determine which of the many vegetation indices currently in use are most strongly correlated with chlorophyll and f o l i a r nutrient levels. Predictive models could then be developed for the estimation of f o l i a r constituents and assessment of nutrient stress. 21 2.7 Chlorophyll A to B Ratio The concentration and proportion of photosynthetic pigments vary with species, environment, and leaf age (Wolf, 1956; Kramer and Kozlowski, 1979; Hoober, 1984; Salisbury and Ross, 1985). Chlorophyll a is usually 2 to 3 times as abundant as chlorophyll b (Salisbury and Ross, 1985); however, with stress the losses of chlorophyll a and b often are not proportional (Kramer and Kozlowski, 1979) . Some forms of stress increase the chlorophyll a/b ratio while other forms result in a decrease. Thornber (1975) reported a decrease in the chlorophyll a/b ratio for plants subjected to low light intensity, while Alberte et al. (1977) noted an increase for drought-stressed maize. Horler et al. (1980) showed that the chlorophyll a/b ratio of pea plants decreased under conditions of excess cadmium or copper, but showed l i t t l e effect with lead or zinc. In contrast, Rock et al. (1986; 1988) reported that the chlorophyll a to b ratio was higher in red spruce from sites of high acid rain damage than from low damage sites. Changes in chlorophyll a/b ratios of vegetation subjected to nutrient deficiencies have yet to be investigated. Should a relationship exist between the ratio of chlorophyll a/b and nutrient stress then i t may be possible to monitor changes in chlorophyll ratios, and thus nutrient stress, via remote sensing of strong chlorophyll a and chlorophyll b absorption bands. 22 3. HYPOTHESES Several hypotheses were postulated while reviewing past studies on the impact of nutrient stress on vegetation spectral reflectance. The objective of this study was to test each of the following hypotheses: One: That percent spectral reflectance at the wavelengths of the green reflectance peak and red rise can be used to detect nutrient (nitrogen, phosphorus and sulphur) deficiencies in Douglas-fir seedlings. Two: That spectral reflectance at the green reflectance peak and red rise are significantly correlated with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll levels in Douglas-fir seedlings, and can thus be used to estimate the concentrations of these f o l i a r constituents. Three: That nutrient (nitrogen, phosphorus and sulphur) deficiencies in Douglas-fir seedlings can be detected by "shifts" in the wavelength of the red edge. Four: That the wavelength of the red edge is significantly correlated with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll levels in Douglas-fir seedlings, and can thus be used to estimate the concentrations of these 23 f o l i a r constituents. Five: That changes in the values of vegetation indices 1 to 15 can be used to detect nutrient (nitrogen, phosphorus and sulphur) deficiencies in Douglas-fir seedlings. Six: That vegetation indices 1 to 15 are significantly correlated with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll levels in Douglas-fir seedlings, and can thus be used to estimate the concentrations of these f o l i a r constituents. Seven: That the use of red rise rather than red well measurements in the calculation of vegetation indices requiring a red reflectance measure in their determination w i l l result in indices more strongly correlated with nitrogen, phosphorus, sulphur and total chlorophyll, and thus provide more accurate estimates of these f o l i a r constituents. Eight: That the chlorophyll a/chlorophyll b ratio can be used as an index of nutrient (nitrogen, phosphorus and sulphur) stress in Douglas-fir seedlings. 24 4. METHODS 4.1 1987 One year old Douglas-fir seedlings were transplanted into plastic trays (26 cm x 52.5 cm x 5 cm) f i l l e d with clean s i l i c a sand. Seedling roots were thoroughly washed before being placed in the sand culture. Five plastic pots (9 cm x 8 cm x 5 cm) , each containing one seedling, were placed in each tray. Treatments were applied as sixteen nutrient solutions containing nitrogen and phosphorus levels of 1, 5, 10, 25, 50, 100, 250 and 400 mg/L in a completely randomized design. A l l treatments contained equal amounts of macronutrients (except for nitrogen or phosphorus) and micronutrients. Appendix B provides a l i s t of nutrient levels and sources for each treatment. Nutrient solution composition was based on the work of van den Driessche (1969) and Swan (1972). The pH of the nutrient solutions was adjusted to 5.5 with NaOH or HCl, as this was the acidity recommended for coniferous tree growth by Tinus and MacDonald (1979). Seedlings were grown from June 3 to September 20, 1987, on an exposed, level, area of the University of British Columbia tree nursery. One l i t r e of nutrient solution was applied to each tray every second day. At one-week intervals the sand in each tray was leached with five l i t r e s of d i s t i l l e d water to prevent salt accumulation. Needle-reflectance measurements were made using a Spectron Engineering (Model SE590) spectroradiometer. The scanning head had a field-of-view of 10 degrees, and was positioned 10 cm above the scanned foliage, perpendicular to the target platform. Newly matured needles were clipped from terminal leaders, and a composite sample produced for each treatment by taking equal f o l i a r weights from each seedling. A 2.5 cm x 2.5 cm target was formed by laying needles top-side-up parallel to one another. A second layer was placed at right angles overtop the f i r s t so that none of the target platform was visible. Measurements were made from 400 to 1100 nm at 3 nm intervals in a dark room. Four scans were made of each sample and the average computed. Barium sulphate (ca. 100% reflectance) was used as a reference. One 600-W CGE photographic lamp positioned 2 0 cm from the target at 45 degrees served as the light source. To adjust for variance in light conditions, needle radiance was converted to a percentage of reflectance from the barium sulphate at each sampled wavelength. Nutrient analyses were conducted on needles dried over an o eight hour period at 70 C in a forced-draft oven. Samples were ground to a fine powder and digested by the method of Parkinson and Allen (1975) , using H2S04, Li 2S0 4, H202, and Se. Digest solutions were analyzed for nitrogen (Berthelot reaction) and phosphorus (ascorbic acid reduction of molybdate complex) on a Technicon auto-analyzer; potassium, calcium, magnesium, manganese, iron, zinc and copper were measured by absorption spectrophotometry. Total sulphur was determined with a Leco sulphur analyzer, while boron was measured by the Wolf (1974) azomethine H colorimetric method as modified by Gaines and Mitchell (1979). Foliar samples destined for chlorophyll analysis were immediately frozen after harvest, and later analyzed spectrophotometrically for chlorophylls a, b, and total chlorophyll following procedures outlined by the British Columbia Department of Environment (1976) and the American Public Health Association (1980, 1985). To determine i f any treatments resulted in nutrient deficiencies, data from fo l i a r analyses were entered into a computerized diagnostic program called DIAGFOLI. DIAGFOLI was originally developed for the B.C. Ministry of Forests by Dr. T.M. Ballard of the University of British Columbia, and expanded for use with several coniferous tree species (Ballard and Carter, 1986). The program compares fo l i a r nutrient concentrations and nutrient ratios with the interpretations of several investigators. Although DIAGFOLI reports on the severity of a deficiency, i t does not distinguish between adequate and toxic conditions. Literature citations for interpretive c r i t e r i a and a discussion of diagnostic limitations are presented in Ballard and Carter (1986) . The ratio of chlorophyll a/chlorophyll b was plotted against the f o l i a r concentrations of nitrogen and phosphorus to determine the potential of this ratio as an indicator of nutrient stress in Douglas-fir seedlings. F i r s t derivative spectra were calculated and used to determine the wavelength of the red edge. The wavelength of the red edge was defined as the wavelength of maximum dreflectance/dwavelength within the 680 to 750 nm range (see Figure 3). The wavelength of the green reflectance peak was determined by averaging the wavelengths of maximum green spectral reflectance for a l l treatments, whereas wavelength of the red rise was the average of the wavelengths of maximum reflectance between 628 and 638 nm. The wavelength of the red well was defined as the average wavelength of minimum red spectral reflectance for a l l treatments. Values for vegetation indices 1 to 15 were determined using percent spectral reflectance at the green reflectance peak and red well, and the reflectance of blue and near-infrared light at 480 and 800 nm, respectively. The red and blue chlorophyll absorption centres were selected since when used in combination with the green reflectance peak, the difference between chlorophyll absorption and reflection is maximized. The 800-nm band was selected for calibration and comparison purposes since i t is well within the Spectron spectroradiometer's sensitivity range, and sensitive to discontinuities in the leaf mesophyll structure (Demetriades-Shah et al., 1990). Relationships between spectral parameters (green reflectance peak, red rise, red edge, and vegetation indices) and needle chlorophyll, nitrogen, and phosphorus concentration were examined by correlation and regression analysis. 28 Vegetation indices were recalculated using percent spectral reflectance at the red rise in place of red well measurements to determine i f greater correlation would result with f o l i a r nitrogen, phosphorus, and total chlorophyll. Correlation coefficients of vegetation indices calculated with red well and red rise spectral measurements were tested for significant differences according to procedures outlined by Zar (1984). Regression lines and curves of spectral parameters most correlated with needle nitrogen, phosphorus and total chlorophyll were used as models to estimate the concentrations of these fol i a r constituents. 4.2 1988 Similar methods were used to grow year-old Douglas-fir seedlings as in 1987. The only differences were that ten seedlings were grown in each tray, leaching of salts was done twice per week using ten l i t r e s of de-ionized water per tray, and eight new treatments were added to the experiment. Eight trays of Douglas-fir seedlings were f e r t i l i z e d with sulphur solutions at rates of 1, 5, 10, 25, 50, 100, 250, and 400 mg/L. Appendix B provides information on the sources of sulphur and the concentrations of other elements used in the nutrient solutions. Each tray was f e r t i l i z e d with one l i t r e of solution, every second day, from May 19 to October 2, 1988. Seedling height and stem diameter were measured pre- and post-treatment. Diameter measurement was made at a point where a line extended across the top of the pot intersected the seedling stems. Height measurement was made from the point of diameter measurement to the apical bud. Post-treatment stem and root masses were o determined following a four-day drying period at 70 C. Growth data were tested for significant differences between treatment means using Tukey's honestly significant difference procedure. A Li-Cor 1800 spectroradiometer f i t t e d with an external integrating sphere (Model 1800-12) was used to measure leaf hemispherical spectral reflectance, at 1-nm scan intervals. Newly matured needles were clipped, selected at random from a cup, arranged top-side-up parallel to one another on a piece of black ele c t r i c a l tape, and placed in the port of the integrating sphere. Spectral radiance between 400 and 1100 nm was recorded for a l l seedlings. The sample and barium reflectance plug in a reference port were interchanged and a reference spectrum was obtained. Final sample hemispherical spectral reflectance was calculated from %SHR = (SR - DS)/(RR - DS)RB where %SHR=sample hemispherical spectral reflectance, SR=sample spectral radiance, RR=reference spectral radiance, DS=dark signal, and RB=reflectance of the barium sulphate reference which was assumed to be 100%. 30 The dark signal is the sum of detector offset voltage, which is independent of wavelength, and scattered light which is wavelength dependent (Li-Cor, 1982). It was determined by measuring the spectral radiance of the integrating sphere with the sample port open in a dark room. Spectral reflectance data were analyzed using the same procedures as in 1987. Due to budgetary constraints i t was not possible to analyze f o l i a r nutrient and chlorophyll levels for a l l seedlings. Consequently, composite samples were produced for each treatment by taking two grams of foliage from each seedling. To provide an idea of within treatment var i a b i l i t y three seedlings were randomly selected from each tray for f o l i a r analysis. Composite spectral reflectance curves were produced by averaging the spectral signatures for a l l seedlings within a treatment. 31 5. RESULTS AND DISCUSSION 5.1 NITROGEN SERIES OF TREATMENTS 5.1.1 Seedling Growth In the 1987 nitrogen series of treatments the sole mortality occurred at the 250 mg/L nitrogen level. Mortality in 1988 was most common among low level nitrogen treatments. One seedling died when subjected to 5 mg/L nitrogen, three at 10 mg/L, two at 25 mg/L, and one at 250 mg/L nitrogen. Seedlings receiving 250 and 400 mg/L nitrogen had significantly greater mean height gains and dry stem weights at the end of the 1988 growing season than Douglas-fir receiving lower amounts of nitrogen (Table 3). From 100 to 250 mg/L nitrogen there was a significant increase in f e r t i l i z e r response for these two variables. Stem diameter change and dry root weight tended to increase with each successively concentrated nitrogen treatment; although, changes in stem diameter were not significantly different for treatments 100 to 400 mg/L nitrogen, nor were root weight changes between 25 and 400 mg/L nitrogen. Of a l l the nitrogen f e r t i l i z a t i o n rates tested, the optimum concentration for overall seedling growth was about 250 mg/L nitrogen. This is higher than the optimum nitrogen concentration of 100 mg/L reported by van den Driessche (1969). 32 Table 3. Mean nitrogen growth data. Nitrogen Diameter Height Stem Root applied change change weight weight (mg/L) (mm) (cm) (g) (g) 1 (± 1.9a 0.56) 5.9a (± 1.30) (± 1. 66a 0.60) (± 2.7la 1.16) 5 (± 2. 0a 0.70) 5.8a (± 1.57) (± 1.80ab 0.51) (± 2.49a 1.84) 10 (± 1.8a 0.11) 6.1a (± 1.96) (± 1.81ab 0.54) (± 2.99ab 1.45) 25 (± 3.5b 0.81) 8.1a (± 3.67) (± 2.83abc 0.69) (± 5.32abc 2.30) 50 (± 4.3bc 0.57) 7.9a (± 1.94) (± 3.82bc 0.71) (± 5.61abc 3.36) 100 (± 5. Ocd 1.22) 9. 5a (± 3.56) (± 4.76c 1.42) (± 6.lObc 2.36) 250 (± 5.5d 0.83) 15.7b (+ 6.10) (± 7.25d 2 . 92) (± 8.37c 3 . 02) 400 (± 5.9d 0.66) 15.4b (± 3.75) (± 8.73d 2.11) (± 8 . 30c 2.28) Means within each column with the same letter are not significantly different as judged by Tukey's honestly significant test (P<0.05). Values are means of 9, 7, 8 and 9 observations for treatments 5, 10, 25 and 250 mg/L nitrogen, respectively. A l l other values are means of 10 observations. Numbers in parentheses are standard deviations. Diameter and height changes were those which occurred from May 19 to October 2, 1988, while stem and root weights refer to measurements made on October 2, 1988. 33 5.1.2 Foliar Analysis Tables 4 to 6 present f o l i a r nitrogen and chlorophyll levels, and results of the DIAGFOLI diagnostic program for the nitrogen series of treatments. Chlorophyll and f o l i a r concentrations of a l l elements for a l l treatments are listed in Appendix C and D. Although needle nitrogen and chlorophyll increased with successively concentrated nitrogen treatments, large within-treatment variation occurred for these and other f o l i a r nutrient concentrations. Genetic variation within a species may cause differences in mineral composition of conifer seedlings (Mergen and Worrall, 1965; Goddard and Hollis, 1984; Timmer, 1991). Because of large within-treatment variation and the incorporation of f o l i a r material from the three individual seedlings analyzed per treatment into the 1988 composite samples, composite and individual seedling data were considered separate data sets and analyzed separately. Data for phosphorus and sulphur treatments were treated in a like manner. The sole use of composite data would have eliminated the spread of data points for each treatment, thus providing a misleading impression of tight clustering (Freedman et al., 1978). DIAGFOLI's interpretation of Douglas-fir f o l i a r nutrient concentrations are based on the work of Gessel et al. (1960), Turner (1966), Heilman (1971), Everard (1973), and Heilman and Ekuan (1973), and indicate that seedlings grown at low concentrations (1 to 10 mg/L) of nitrogen were severely nitrogen Table 4. Composite f o l i a r nitrogen and t o t a l chlorophyll concentrations f o r the 1987 nitrogen s e r i e s of treatments. F o l i a r concentrations Nutrient status Nitrogen applied (mg/L) %N CHL (mg/g) Nitrogen Other elements 1 0.86 0.46 very severe de f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Zn - probable d e f i c i e n c y 5 1.01 0.95 very severe de f i c i e n c y 10 1.24 0.64 severe deficiency 25 1.35 0.78 s l i g h t to moderate deficiency Zn - possible d e f i c i e n c y 50 1.80 0.63 adequate 100 2.34 1.20 adequate Zn - possible d e f i c i e n c y 250 1.84 0.95 adequate Cu - possible d e f i c i e n c y 400 3.15 1.90 adequate %N - percent nitrogen CHL - t o t a l chlorophyll Table 5. Composite f o l i a r nitrogen and t o t a l chlorophyll concentrations f o r the 1988 nitrogen ser i e s of treatments. F o l i a r concentrations Nutrient status Nitrogen applied (mg/L) %N CHL (mg/g) Nitrogen Other elements 1 1.13 (+0.08) 0.49 (±0.03) severe deficiency 5 1.09 (+0.12) 0.57 (+0.03) severe deficiency Ca - l i t t l e , i f any def i c i e n c y Cu - s l i g h t p o s s i b i l i t y of de f i c i e n c y 10 1.60 (+0.05) 0.77 (±0.12) adequate Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 25 1.48 (+0.14) 0.68 (+0.04) adequate 50 1.68 (+0.3.0) 1.08 (+0.06) adequate 100 1.76 (+0.48) 1.23 (+0.46) adequate 250 2.55 (+0.47) 2.28 (+0.29) adequate Zn - possible deficiency 400 3.77 (+0.94) 1.87 • (±0.47) adequate %N - percent nitrogen CHL - t o t a l chlorophyll Numbers i n parentheses are standard deviations. Standard deviations f o r f o l i a r concentrations are based on data from three i n d i v i d u a l seedlings, whereas standard deviations for s p e c t r a l measurements are based on a l l surviving seedlings. T a b l e 6. F o l i a r n i t r o g e n and t o t a l c h l o r o p h y l l c o n c e n t r a t i o n s f o r i n d i v i d u a l D o u g l a s - f i r s e e d l i n g s s u b j e c t e d t o t h e 1988 n i t r o g e n s e r i e s o f t r e a t m e n t s . F o l i a r c o n c e n t r a t i o n s N u t r i e n t s t a t u s N i t r o g e n a p p l i e d (mg/L) %N CHL (mg/g) N i t r o g e n O t h e r elements 1 0.92 0.49 v e r y severe d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 1.06 0.55 s e v e r e d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 0.94 0.50 v e r y severe d e f i c i e n c y Ca - p o s s i b l e s l i g h t t o moderate d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 5 1.04 0.43 v e r y severe d e f i c i e n c y Ca l i t t l e , i f any d e f i c i e n c y Zn - p o s s i b l e d e f i c i e n c y Cu — s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 0.96 0.37 v e r y severe d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 0.80 0.38 v e r y severe d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 10 1.05 0.61 v e r y severe d e f i c i e n c y 0.96 0.37 v e r y severe d e f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Zn - p o s s i b l e d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 1.04 0.52 v e r y severe d e f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 25 1.49 0.75 adequate 1.47 0.76 adequate 1.23 0.68 s e v e r e d e f i c i e n c y 50 1.42 0.67 s l i g h t t o moderate d e f i c i e n c y 1.88 0.65 adequate Cu - s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 1.31 0.77 s l i g h t t o moderate d e f i c i e n c y Cu — s l i g h t p o s s i b i l i t y o f d e f i c i e n c y C o n t i n u e d Table 6. Continued. F o l i a r concentrations Nutrient status Nitrogen applied %N CHL Nitrogen Other elements (mg/L) (mg/g) 100 1.36 0.78 s l i g h t to moderate deficiency Ca _ l i t t l e , i f any d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 1.57 0.94 adequate K - possible s l i g h t d e f i c i e n c y Ca - possible s l i g h t to moderate Cu - s l i g h t p o s s i b i l i t y of de f i c i e n c y 2.28 1.65 adequate 250 2.26 1.30 adequate K possible s l i g h t d e f i c i e n c y 2.88 1.86 adequate K - possible s l i g h t d e f i c i e n c y 1.96 1.44 adequate Zn possible d e f i c i e n c y 400 1.86 1.42 adequate K _ possible s l i g h t d e f i c i e n c y Zn - possible d e f i c i e n c y 3.39 2.10 adequate Zn - possible d e f i c i e n c y 3.57 2.32 adequate %N - percent nitrogen CHL - t o t a l chlorophyll 38 deficient. This was visually demonstrated in both 1987 and 1988 data sets by a gradient of foliage greenness starting with the extremely chlorotic needles of the 1 mg/L treated seedlings, increasing in greenness with each successively concentrated nitrogen treatment, and culminating in the dark green foliage of those seedlings grown at 400 mg/L. Chlorosis was evenly distributed over both new and older needles of 1 to 10 mg/L nitrogen treated seedlings, except for a few seedlings which had necrotic needle tips. A l l seedlings grown at these nitrogen concentrations had slender, short, stalks with l i t t l e branching, and stunted needles. Although DIAGFOLI indicated the slight possibility of zinc, copper and calcium deficiencies, nitrogen was overwhelmingly the most severe deficiency. Furthermore, the seedlings' symptoms were characteristic of nitrogen deficiency (Marschner, 1986; Mengel and Kirkby, 1987; Morrison, 1974). Nitrogen f e r t i l i z a t i o n at 100 to 400 mg/L may have induced the possible slight potassium deficiencies noted for the individual seedling data set. Weetman (1968) found that nitrogen f e r t i l i z a t i o n of black spruce may result in slight potassium deficiencies. 39 5.1.3 Spectral Reflectance Curves For a l l 1987 nitrogen treatments (Figure 4) , needle reflectance remained constant from 400 to about 494 nm, then rapidly increased to a maximum in the visual range at 559 nm. From 559 to approximately 670 nm there was a gradual decrease in reflectance, reaching a low at 674 nm, then dramatically increasing over the red edge portion of the electromagnetic spectrum. With the exception of a slight decrease in spectral reflectance at 960 nm, reflectance remained relatively constant between 774 and 1032 nm, followed by a sharp drop over the remainder of the near-infrared region scanned. Aside from the erratic measurements in the extreme blue and near-infrared regions (Figure 5) , the 1988 composite nitrogen reflectance curves were similar to those of 1987. The only minor differences were that the green reflectance peak occurred at 554 rather than 559 nm, and that spectral reflectance in the 800 to 1000 nm infrared region was more uniform in 1988. The erratic spectral reflectance readings or "noise" at the extreme ends of the spectral region scanned can be attributed to the sensitivity limits of the Li-Cor spectroradiometer being approached. The LI-1800 spectroradiometer measures the spectral distribution of radiation by dispersing the radiation with a diffraction grating monochromator, and measuring the energy in the various wavelengths of the resulting spectrum with a silicon detector. Selecting the proper s l i t size for the monochromator 90 400 500 600 700 800 900 1000 1100 WAVELENGTH (nm) 1 mg/L N 5 mg/L N 10 mg/L N 25 mg/L N Figure 4. 1987 spectral reflectance curves showing the effects of nitrogen levels on needle reflectance at the wavelength interval 400 to 1100 nm. 400 500 600 700 800 900 1000 1100 WAVELENGTH (nm) 50 mg/LN 100 mg/LN 250 mg/LN 400 mg/L N Figure 5. 1988 spectral reflectance curves showing the effects of nitrogen levels on needle reflectance at the wavelength interval 400 to 1100 nm. Plotted data are from the composite data set. represents a trade-off between wavelength resolution and signal-to-noise ratio of the resulting data (Li-Cor Inc., 1982). Needle reflectance over the visible range of the spectrum varied with nitrogen treatment and thus f o l i a r nitrogen concentration; however, no relationship was observed between spectral reflectance in the near-infrared and nitrogen level. Treatments low in nitrogen produced needles which had higher visible spectral reflectance than foliage of treatments high in nitrogen. The shape of the spectral reflectance curves also varied between nitrogen treatments (Figures 4 and 5). Treatments which produced foliage high in nitrogen had low spectral reflectance at the red rise, and a deep, broadened, red well region. This shifted the red edge toward the near-infrared region of the spectrum. These changes in spectral reflectance can be attributed to differences in nitrogen f e r t i l i z a t i o n , and thus, chlorophyll concentration. The relationship between f o l i a r nitrogen and total chlorophyll was linear (Figure 6), and the correlation positive and s t a t i s t i c a l l y significant (p<0.01) for a l l three data sets (Table 7) . Similar relations between chlorophyll and nitrogen have been reported for Scotch pine and poplar (Linder, 198 0; Linder and Rook, 1984). Nitrogen is an essential component of the chloroplast proteins, chlorophylls, lipids, and nucleic acids. Lack of this element results in the collapse of chloroplasts and a disturbance in chloroplast development (Thomson and Weir, 1962; Kirk and Tilney-Bassett, 1967; Mengel and Kirkby, 1987). Since chlorophyll 43 2.5 J1 1 — I 1 - T , r -0-5 1 1.5 2 2.5 3 3.5 NITROGEN (%) Figure 6. Relationship between t o t a l c hlorophyll (CHL) and f o l i a r nitrogen concentration (N) for the 1988 i n d i v i d u a l seedling data set. 44 Table 7. Correlation between total chlorophyll and the f o l i a r concentration of nitrogen. df Correlation coefficient 1987 composite 6 samples 1988 composite 6 samples 1988 individual 22 seedlings ** - p<0.01 df - degrees of freedom strongly absorbs energy in the wavelength bands centred at 480 and 680 nm (Rock et al., 1986) i t s destruction causes reflectance in the blue and red to increase. Often red reflectance increases to the point that foliage shifts from a green colour to yellow to orange, and f i n a l l y a deep brown (Fox, 1978). 5 . 4 . 4 Chlorophyll A/Chlorophyll B Plots of chlorophyll a/chlorophyll b versus the f o l i a r concentrations of nitrogen were used to determine the potential of this ratio as an indicator of nutrient stress in Douglas-fir seedlings. The relationship between chlorophyll a to b and the severity of nutrient deficiency for seedlings subjected to the phosphorus and sulphur series of treatments were examined in a like 0.88** 0.84** 0.96** manner. DIAGFOLI diagnosed most seedlings as being deficient in one or more nutrients, therefore seedlings were classified according to their most severe deficiency and grouped into the following categories: 1) very severely nutrient deficient; 2) moderately to severely nutrient deficient; 3) possible minor nutrient deficiency; 4) not nutrient deficient. Chlorophyll a/chlorophyll b ratios of composite f o l i a r samples for the 1987 series of nitrogen treatments were extremely high for seedlings diagnosed as very severely nutrient deficient (Figure 7) . However, ratios dramatically decreased for severely and moderately nutrient deficient samples, and then increased for composite samples with minor or no nutrient deficiencies. In 1988, composite and individual seedling data sets followed similar trends (Figures 8 and 9) . Chlorophyll a to b ratios of very severely to severely nutrient deficient seedlings were in most cases lower than those of seedlings classified as having minor or no nutrient deficiencies. As there was no consistent relationship between the ratio of chlorophyll a/chlorophyll b and the severity of nutrient deficiency for the 1987 and 1988 data sets i t does not appear to be a reliable indicator of nutrient stress. 46 1.5 2 2.5 NITROGEN (%) 3.5 • — VERY SEVERELY NUTRIENT DEFICIENT 0 —- MODERATELY TO SEVERELY NUTRIENT DEFICIENT + -— POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 7. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r nitrogen and severity of nutrient deficiency for the 1987 nitrogen composite data set. 47 2.15 2 2.5 NITROGEN (%) • — VERY SEVERELY NUTRIENT DEFICIENT + ~ POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 8. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r nitrogen and severity of nutrient deficiency for the 1988 nitrogen composite data set. 48 1.7 0 2 2.5 NITROGEN (%) • — VERY SEVERELY NUTRIENT DEFICIENT 0 — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 9. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r nitrogen and severity of nutrient deficiency for the 1988 nitrogen individual seedling data set. To determine whether the green reflectance peak, red rise, red edge, or one of the vegetation indices would be best for detecting nitrogen deficiencies and estimating f o l i a r nitrogen concentrations, i t was necessary to find which of these measurements were most highly correlated with nitrogen. The following sections explore the relationship of these spectral measurements with f o l i a r nitrogen levels. 5.1.5 Green Reflectance Peak The wavelength of the green reflectance peak occurred at 559 nm in 1987, and at 554 nm in 1988. These were the wavelengths of greatest separation of reflectance curves in the visible region of the spectrum, and therefore, according to Tsay et al. (1982) and Nelson et al. (1986), the best wavelengths for discriminating nitrogen and chlorophyll concentrations. The discrepancy between years for the wavelength of the green reflectance peak can largely be attributed to differences in the spectral resolution of spectroradiometers used, and possibly experimental error and phenotypic variation. Reflectance measurements were recorded for the most narrow spectral bandwidths possible; in the case of the Spectron SE590 this was approximately 3 nm, and for the Li-Cor 1800 1 nm. Spectral reflectance at the green reflectance peak decreased as the concentration of f o l i a r nitrogen increased (Tables 8 to 10) . Plots of these two variables demonstrated exponentially decreasing 50 Table 8. Percent spectral reflectance at the green reflectance peak (559 nm) and red rise (630 nm), and wavelength of the red edge for composite f o l i a r samples of the 1987 nitrogen series of treatments. Nitrogen Green applied reflectance Red rise Red edge (mg/L) peak (%) (nm) (%) 1 22.0 13.5 700 5 17.5 13.0 700 10 23.0 14.0 706 25 16.5 10. 5 700 50 12.5 6.5 712 100 13.5 7.0 717 250 11.5 6.0 717 400 10.0 5.5 723 51 Table 9. Percent spectral reflectance at the green reflectance peak (554 nm) and red rise (630 nm), and wavelength of the red edge for composite f o l i a r samples of the 1988 nitrogen series of treatments. Nitrogen Green applied reflectance Red rise Red edge (mg/L) peak (%) (nm) (%) 1 26.22 14.88 694 (±2.25) (±2.24) (±2.36) 5 26.38 14.78 697 (±2.20) (±1.95) (±5.05) 10 26.10 14.57 696 (±2.15) (±2.14) (±4.35) 25 19.99 9.95 704 (±2.58) (±1.26) (±6.12) 50 18.95 10.35 699 (±2.63) (±2.42) (±6.26) 100 15.06 7.76 699 (±2.22) (±2.81) (±12.23) 250 8.60 4.76 715 (±1.12) (±0.28) (±3.95) 400 8.81 4.70 723 (±1.59) (±0.80) (±7.05) Numbers in parentheses are standard deviations. Standard deviations for f o l i a r concentrations are based on data from three individual seedlings, whereas standard deviations for spectral measurements are based on a l l surviving seedlings. 52 Table 10. Percent spectral reflectance at the green reflectance peak (554 nm) and red rise (630 nm), and wavelength of the red edge for individual Douglas-fir seedlings subjected to the 1988 nitrogen series of treatments. Nitrogen Green applied reflectance Red rise Red edge (mg/L) peak (%) (nm) (%) 1 26.97 13.86 692 26.14 14.37 698 23.36 13.43 696 5 28.11 16.04 695 25.92 13.69 702 27.65 16.67 702 10 25.58 14.67 703 27.57 17.11 694 26.03 13.91 689 25 23.30 10.97 692 16.81 7.60 698 22.65 11.16 703 50 15.38 7.28 701 22.41 14.79 705 20.09 10.43 699 100 15.09 6.83 718 19.13 10.08 696 12 . 39 4.98 729 250 9.62 4.44 721 9.21 4.31 724 10.17 4.93 726 400 11.61 5.01 720 10.67 4.56 728 8.76 5.37 722 relationships for both 1987 and 1988 data, an example of which is provided by the 1988 individual seedling data (Figure 10). Logarithmic transformations of nitrogen and green reflectance peak data as outlined in Spain (1982) yielded an inverse, linear, relationship (Figure 11). Insufficient data was available to test for normality and homocedasity. Significant correlation (p<0.01) was found between logarithmically transformed variables for a l l three data sets (Table 11). As f o l i a r nitrogen increased, chlorophyll concentration increased, resulting in more visible light being absorbed and less reflected. This explains the inverse relationship noted between percent spectral reflectance at the green reflectance peak and f o l i a r nitrogen. Thomas and Oerther (1972), Tsay et al. (1982) and Nelson et al. (1986) noted similar relationships. 5 . 1 . 6 Red R i s e Tables 8 to 10 demonstrate changes in spectral reflectance at the wavelength of the red rise for the nitrogen series of treatments. The red rise occurred at 630 nm in both 1987 and 1988. As with green reflectance peak data spectral reflectance at the red rise decreased exponentially with increased f o l i a r nitrogen; logarithmic transformation of both variables resulted in a linear relationship. The correlation between the logarithm of f o l i a r nitrogen concentration and the logarithm of percent spectral reflectance at 630 nm was significant at the 0.01 level 54 Figure 10. Relationship between f o l i a r nitrogen concentration (N) and percent spectral reflectance at the green reflectance peak (GRP; 554 nm) for the 1988 individual seedling data set. 55 0.7 -0.1- -t--0.2-I . - . . , = , 0.9 1 1.1 . 1.2 1.3 1.4 1.5 LOG SPECTRAL R E F L E C T A N C E (%) AT 554 nm Figure 11. Relationship between the logarithm of f o l i a r nitrogen concentration (N) and the logarithm of percent spectral reflectance at the green reflectance peak (GRP; 554 nm) for the 1988 individual seedling data set. 56 for a l l data sets (Table 11). As nitrogen increased, there was a corresponding increase in chlorophyll concentration, resulting in more red light being absorbed and less reflected. Consequently, an inverse relationship existed between percent spectral reflectance at the red rise and needle nitrogen concentration. 5 . 1 . 7 Red Edge There was a blue shift of the red edge with decreasing concentration of nitrogen treatments (Tables 8 to 10) which resulted in a linear relationship between f o l i a r nitrogen and the wavelength of the red edge (Figure 12). Correlation coefficients calculated for the association between f o l i a r nitrogen and the wavelength of the red edge were positive and significant (p<0.01) for 1987 and 1988 data. With a decrease in needle nitrogen concentration, there was a corresponding decrease in chlorophyll, causing less absorption of red light, and a shift of the red edge toward the blue region of the spectrum. 57 Table 11. Correlation between the f o l i a r concentration of nitrogen and green reflectance peak, red rise, and red edge measurements. df GRP RR RE 1987 composite samples 6 -0.87** (0.10) -0.90** (0.09) 0.93** (0.31) 1988 composite samples 6 -0.92** (0.09) -0.91** (0.08) 0.94** (0.32) 1988 individual samples 22 -0.90** (0.08) -0.86** (0.10) 0.79** (0.48) df - degrees of freedom GRP - green reflectance peak RR - red rise RE - wavelength of the red edge Numbers in parentheses represent the standard error of estimate. Green reflectance peak and red rise correlation coefficients were calculated for logarithmically transformed data. 58 Figure 12. Relationship between f o l i a r nitrogen (N) and the wavelength of the red edge (RE) for the 1988 individual seedling data set. 59 5.1.8 Vegetation Indices The wavelength of the red well occurred at 674 nm for both 1987 and 1988. Plots of percent f o l i a r nitrogen versus each vegetation index calculated using percent spectral reflectance at the red well yielded two types of relationships. Vegetation indices 1, 2, 3, 4, 6, 8, 10, 12 and 13 increased exponentially with increased f o l i a r nitrogen, whereas vegetation indices 5, 7, 9, 11, 14 and 15 decreased exponentially with increased needle nitrogen concentration. Linear relationships were obtained through logarithmic transformation of f o l i a r nitrogen and vegetation indice data, and correlation coefficients calculated for the transformed data (Table 12). Several vegetation indices (1, 2, 3, 8, 12, 13 and 15) were significantly correlated with f o l i a r nitrogen at the 0.01 level for a l l three data sets. Of these, vegetation index 15 had the highest correlation coefficient, followed by 13 and 8. Red rise measurements were highly correlated with f o l i a r nitrogen; consequently, percent spectral reflectance at 630 nm was substituted for red well measurements in the calculation of vegetation indices to determine i f higher correlation could be obtained between indices and nitrogen. Throughout the thesis recalculated indices using percent spectral reflectance at the red rise (630 nm) are referred to as "red rise vegetation indices", whereas those calculated using red well (674 nm) measurements are called "red well vegetation indices". 60 Table 12. Correlation between the fo l i a r concentration of nitrogen and vegetation indices calculated using percent spectral reflectance at the red well and red rise. Red well Red rise VI 1987 comp 1988 comp 1988 ind 1987 comp 1988 comp 1988 ind 0.83** (0.11) 0.85** (0.10) 0.69** (0.14) 0.95** (0.06) 0.90** (0.08) 0.86** (0.10) 0.84** (0.11) 0.83** (0.11) 0.64** (0.14) 0.95** (0.06) 0.86** (0.10) 0.83** (0.10) 0.84** (0.11) 0.83** (0.11) 0.65** (0.14) 0.95** (0.06) 0.87** (0.10) 0.84** (0.10) 0.29ns (0.20) 0.85** (0.10) 0.83** (0.10) -0.73* (0.14) -0.22ns (0.19) -0.42* (0.17) -0.33ns (0.19) -0.86** (0.10) -0.81** (0.11) 0.49ns (0.18) -0.09ns (0.19) 0.40* (0.17) 0.12ns (0.20) 0.82* (0.11) 0.82** (0.11) -0.60ns (0.16) -0.59ns (0.16) -0.39ns (0.17) -0.27ns (0.20) -0.84** (0.10) -0.84** (0.10) 0.69ns (0.15) 0.25ns (0.19) 0.36ns (0.18) 0.93** (0.07) 0.88** (0.09) 0.88** (0.09) 0.93** (0.07) 0.88** (0.09) 0.88** (0.09) -0.31ns (0.19) -0.87** (0.10) -0.79** (0.12) 0.61ns (0.16) 0.02ns (0.19) 0.36ns (0.18) 10 -0.03ns (0.20) 0.79* (0.12) 0.68** (0.14) -0.86** (0.10) -0.86** (0.10) -0.72** (0.13) Continued 61 Table 12. Continued Red well Red rise VI 1987 comp 1988 comp 1988 ind 1987 comp 1988 comp 1988 ind 11 -0.73* (0.14) -0.88** (0.09) -0.90** (0.08) -0.04ns (0.20) -0.81* (0.11) -0.68** (0.14) 12 o'.87** (0.10) 0.84** . " (0.11) 0.79** (0.12) 0.91** (0.08) 0.86** (0.10) 0.82** (0.11) 13 0.94** (0.07) 0.90** (0.09) 0.88** (0.09) 0.96** (0.06) 0.89** (0.09) 0.88** (0.09) 14 -0.76* (0.13) -0.68ns (0.14) -0.52** (0.16) -0.95** (0.07) -0.90** (0.08) -0.85** (0.10) 15 -0.94** (0.07) -0.91** (0.08) -0.91** (0.08) -0.94** (0.07) -0.91** (0.08) -0.91** (0.08) df 6 6 22 6 6 22 * ** ns df p<0.05 p<0.01 not significant degrees of freedom VI - vegetation indices 1987 comp - 1987 composite samples 1988 comp - 1988 composite samples 1988 ind - 1988 individual seedlings Numbers in parentheses represent the standard error of estimate. Correlation coefficients were transformed data. calculated for logarithmically While plots of percent f o l i a r nitrogen versus vegetation indices retained their exponential relationships, vegetation indices 4 and 6 changed from increasing exponentially with increased nitrogen to exponentially decreasing, whereas indices 5, 7 and 9 went from decreasing exponentially to increasing exponentially with increased nitrogen concentration. Correlation coefficients for red rise vegetation indices were determined from logarithmically transformed data (Table 12) , and differences between correlation coefficients for red well and red rise vegetation indices tested for s t a t i s t i c a l significance (Table 13) . The correlation between fo l i a r nitrogen and red rise vegetation indices 1, 2, 3, 10, 12 and 14 was significant (p<0.01), and consistently greater (though seldom s t a t i s t i c a l l y different) for a l l data sets than the correlation between nitrogen and red well values for the same vegetation indices. Therefore, percent spectral reflectance at the red rise should be used rather than red well measurements when using vegetation indices 1, 2, 3, 10, 12 and 14 to assess the nitrogen status of Douglas-f i r seedlings. Correlation coefficients were the same for red well and red rise vegetation indices 8 and 15. This outcome was expected for vegetation index 8 since i t does not incorporate a measure of red reflectance in i t s calculation. Correlation coefficients for red well and red rise vegetation index 15 were only the same when rounded to two decimal places. Table 13. Differences in correlation between f o l i a r nitrogen concentration and recalculated vegetation indices using percent spectral reflectance at the red rise. Vegetation 1987 1988 1988 indices composite composite individual samples samples seedlings 1 + + + 2 • + + + 3 + + + 4 + - -* 5 + - -* 6 + - - * 7 + - -* 8 0 0 0 9 + -* -* 10 +* + + 11 - -* 12 + + + 13 + - 0 14 + + +* 15 0 0 0 * - p<0.05 + - increase in correlation - - decrease in correlation 0 - no change in correlation 64 5.1.9 Estimation of Foliar Nitrogen The determination of which spectral measurements (green reflectance peak, red rise, red edge, or vegetation indices) were best for assessing nitrogen status, and estimating f o l i a r nitrogen in Douglas-fir seedlings was based on finding those measurements most correlated With needle nitrogen concentration. If two or more spectral measurements were equally correlated with nitrogen then the spectral measurement with the lowest standard error of estimate was considered best for estimating f o l i a r nitrogen. The same procedure was also used in the selection of spectral measurements for estimating needle phosphorus, sulphur, and chlorophyll concentration. Vegetation index 15 was the measure most highly correlated with nitrogen for the 1988 individual seedling data set. Calculation of the index using either percent spectral reflectance at the red well or red rise resulted in identical correlation coefficients (r=-0.90; p<0.01) and standard error of estimates (0.08 percent nitrogen). Consequently, models for estimating f o l i a r nitrogen were developed from the regression curves of both red well and red rise forms of vegetation index 15. Models were tested by comparing the estimated and actual f o l i a r nitrogen levels of individual phosphorus and sulphur treated Douglas-fir seedlings (Tables 14 and 15). Table 14. Comparison of f o l i a r nitrogen l e v e l s f o r 1988 i n d i v i d u a l seedling data as determined by chemical analysis and the model -1.2393 %N = (0.2132)(RWVI15). Phosphorus treated seedlings Sulphur t r e a t e d seedlings RWVI15 N obs.(%) N est.(%) Abe. d i f f . 0.2436 1.73 1.23 0.50 0.2022 2.55 1.55 1.00 0.2091 2.11 1.48 0.63 0.1526 2.02 2.19 0.17 0.1639 2.07 2.00 0.07 0.1742 1.62 1.86 0.24 0.1749 1.69 1.85 0.16 0.1809 1.59 1.77 0.18 0.2011 1.52 1.56 0.04 0.1675 1.96 1.95 0.01 0.1854 2.03 1.72 0.31 0.1696 2.04 1.92 0.12 0.2031 1.72 1.54 0.18 0.1658 1.61 1.98 0.37 0.2161 1.73 1.42 0.31 0.2614 1.05 1.12 0.07 0.2265 1.33 1.34 0.01 0.2188 1.46 1.40 0.06 0.2316 1.18 1.31 0.13 0.2530 1.29 1.17 0.12 0.2037 1.47 1.53 0.06 0.2390 1.31 1.26 0.05 0.1802 1.47 1.78 0.31 0.2578 1.17 1.14 0.03 Standard e r r o r of pr e d i c t i o n 0.21 RWVI15 N obs.(%) N est.(%) Abs. d i f f . 0.1672 2.04 1.96 0.08 0.1603 1.91 2.06 0.15 0.1559 2.00 2.13 0.13 0.1322 1.82 2.62 0.80 0.1413 1.59 2.41 0.82 0.1706 1.68 1.92 0.23 0.1727 1.63 1.88 0.25 0.1220 1.80 2.89 1.09 0.1498 2.06 2.24 0.18 0.2124 1.98 1.45 0.53 0.2271 1.75 1.34 0.41 0.1918 1.63 1.65 0.02 0.1850 1.44 1.73 0.29 0.1957 1.53 1.61 0.08 0.1812 1.84 1.77 0.07 0.1820 1.54 1.76 0.22 0.1993 1.61 1.57 0.04 0.1699 1.49 1.92 0.43 0.2125 1.70 1.45 0.25 0.1686 1.59 1.94 0.35 0.2117 1.56 1.46 0.10 0.1959 1.38 1.61 0.23 0.2178 1.53 1.41 0.12 0.1909 1.40 1.66 0.26 Standard e r r o r of p r e d i c t i o n 0.30 RWVI15 - red well vegetation index 15 N obs. - observed nitrogen concentration N est. - estimated nitrogen concentration Abs. d i f f . - absolute difference between observed and estimated concentrations Table 15. Comparison of f o l i a r nitrogen l e v e l s f o r 1988 indiseddalng data as determined by chemical analysis and the model -1.3783 %N = (0.1586)(RRVI15). Phosphorus treated seedlings Sulphur treated seedlings RRVI15 N obs.(%) N est.(%) Abs. d i f f . RRVI15 N obs.(%) N est.(%) Abs. d i f f . 0.2301 1.73 1.20 0.53 0.1645 2.04 1.91 0.13 0.1939 2.55 1.52 1.03 0.1567 1.91 2.04 0.13 0.2005 2.11 1.45 0.66 0.1523 2.00 2.12 0.12 0.1511 2.02 2.15 0.13 0.1294 1.82 2.66 0.84 0.1597 2 .'07 1.99 0.08 0.1406 1.59 2.37 0.78 0.1674 1.62 1.86 0.24 0.1642 1.68 1.91 0.23 0.1689 1.69 1.84 0.15 0.1654 1.63 1.89 0.26 0.1774 1.59 1.72 0.13 0.1213 1.80 2.91 1.11 0.1920 1.52 1.54 0.02 0.1478 2.06 2.21 0.15 0.1647 1.96 1.90 0.06 0.2043 1.98 1.42 0.56 0.1804 2.03 1.68 0.35 0.2146 1.75 1.32 0.43 0.1646 2.04 1.91 0.13 0.1831 1.63 1.65 0.02 0.1960 1.72 1.50 0.22 0.1674 1.44 1.86 0.42 0.1622 1.61 1.95 0.34 0.1880 1.53 1.59 0.06 0.2076 1.73 1.38 0.35 0.1739 1.84 1.77 0.07 0.2366 1.05 1.16 0.11 0.1660 1.54 1.88 0.34 0.2119 1.33 1.35 0.02 0.1834 1.61 1.64 0.03 0.2067 1.46 1.39 0.07 0.1610 1.49 1.97 0.48 0.2120 1.18 1.35 0.17 0.2017 1.70 1.44 0.26 0.2325 1.29 1.18 0.11 0.1580 1.59 2.02 0.43 0.1953 1.47 1.51 0.04 0.1987 1.56 1.47 0.09 0.2236 1.31 1.25 0.06 0.1882 1.38 1.59 0.21 0.1751 1.47 1.75 0.28 0.2057 1.53 1.40 0.13 0.2391 1.17 1.14 0.03 0.1759 1.40 1.74 0.34 Standard e r r o r of prediction 0.22 Standard e r r o r of p r e d i c t i o n 0.32 RRVI15 - red r i s e vegetation index 15 N obs. - observed nitrogen concentration N e s t . - estimated nitrogen concentration Abs. d i f f . - absolute d i f f e r e n c e between observed and estimated concentrations 67 Both models yielded nitrogen estimates with essentially the same standard errors of prediction. The standard errors of nitrogen estimates for sulphur treated seedlings were much greater than those for phosphorus treated seedlings; each model tended to overestimate f o l i a r nitrogen for low vegetation index 15 values, and underestimate nitrogen when values for vegetation index 15 were high. Among composite data, red rise vegetation index 13 of the 1987 composite data set was most correlated (r=0.96; p<0.01) with f o l i a r nitrogen. The regression curve describing the relationship between f o l i a r nitrogen and red rise vegetation index 13 was used as a nitrogen estimation model, and tested by comparing estimated and actual needle nitrogen concentrations of composite f o l i a r samples for the 1987 phosphorus treated seedlings, and 1988 composite samples for nitrogen, phosphorus, and sulphur f e r t i l i z e d seedlings (Table 16). Nitrogen estimates of composite samples tended to be less accurate than those of individual seedlings, with standard errors of prediction ranging from 0.28 to 0.52 percent nitrogen. The red well vegetation index 15 model used to estimate the f o l i a r nitrogen concentration of individual seedlings was applied to composite data in order to obtain the most accurate estimates of nitrogen (Table 17) . Estimates of nitrogen concentration for composite samples tended to be most accurate using vegetation index 15. The standard error of prediction decreased for every composite data set, and ranged from 0.11 to 0.34 percent nitrogen. Table 16. Comparison of f o l i a r nitrogen levels for composite samples as determined by chemical analysis and the model 5.3576 %N = (9.6780)(RRVI13). 1987 phosphorus treated seedlings RRVI13 N obs.(%) N est.(%) Abs. d i f f . 0.7563 1.86 2.17 0.31 0.7607 2.14 2.24 0.10 0.8014 2.53 3.00 0.47 0.7410 1.73 1.94 0.21 0.7861 2.10 2.67 0.57 0.7541 1.95 2.13 0.18 0.7771 1.88 2.51 0.63 0.6890 1.73 1.32 0.41 S.E. of pred. 0.36 1988 phosphorus treated seedlings RRVI13 N obs.(%) N est.(%) Abs. d i f f . 0.7128 2.19 1.58 0.61 0.7423 1.82 1.96 0.14 0.7096 1.66 1.54 0.12 0.7237 1.75 1.71 0.04 0.7284 1.82 1.77 0.05 0.6791 1.51 1.22 0.29 0.6450 1.42 0.92 0.50 0.6612 1.57 1.05 0.52 S.E. of pred. 0.28 1988 nitrogen treated seedlings RRVI13 N obs.(%) N est(%) Abs. d i f f . 0.5944 1.13 0.60 0.53 0.6001 1.09 0.63 0.46 0.5942 1.60 0.60 1.00 0.6756 1.48 1.18 0.30 0.6736 1.68 1.17 0.51 0.7271 1.76 1.76 0.00 0.8179 2.55 3.30 0.75 0.8116 3.77 3.16 0.61 S.E. of pred. 0.52 1988 sulphur treated seedlings RRVI13 Noba.(%) N e s t . ( % ) Abs. d i f f . 0.7754 1.90 2.48 0.58 0.7724 2.12 2.43 0.31 0.7648 2.02 2.30 0.28 0.7200 1.89 1.67 0.22 0.7194 1.94 1.66 0.28 0.7152 1.69 1.61 0.08 0.6993 1.73 1.42 0.31 0.6642 1.50 1.08 0.42 S.E. of pred. 0.31 RRVT13 - red r i s e vegetation index 13 N obs. - observed nitrogen concentration N est. - estimated nitrogen concentration S.E. of pred. - standard e r r o r of p r e d i c t i o n Abs. d i f f . - absolute difference between observed and estimated concentrations cn co Table 17. Comparison of f o l i a r nitrogen l e v e l s f o r composite samples as determined by chemical analysis and the model -1.2393 %N = (0.2132)(RWVI15). 1987 phosphorus treated seedling 1988 nitrogen treated seedlings RWVI15 N obs.(%) N est.(%) Abs. d i f f . RWVI15 N obs.(%) N est.(%) Abs. d i f f . 0.1592 1.86 2.08 0.22 0.2836 1.13 ' 1.02 0.11 0.1519 2.14 2.20 0.06 0.2807 1.09 1.03 0.06 0.1367 2.53 2.51 0.02 0.2855 1.60 1.01 0.59 0.1688 1.73 1.93 0.20 0.2301 1.48 1.32 0.16 0.1353 2.10 2.54 0.44 0.2240 1.68 1.36 0.32 0.1638 1.95 2.01 0.06 0.1855 1.76 1.72 0.04 0.1491 1.88 2.26 0.38 0.1187 2.55 2.99 0.44 0.1750 1.73 1.85 0.12 0.1249 3.77 2.81 0.96 S.E. of pred. 0.19 S.E. of pred. 0.34 1988 phosphorus treated seedlings 1988 sulphur t r e a t e d seedlings RWVI15 N obs.(%) N est.(%) Abs. d i f f . RWVI15 N obs.(%) N est.(%) Abs. d i : 0.2003 2.19 1.56 0.63 0.1763 1.82 1.83 0.01 0.2075 1.66 1.50 0.16 0.1878 1.75 1.69 0.06 0.1944 1.82 1.62 0.20 0.2140 1.51 1.44 0.07 0.2321 1.42 1.30 0.12 0.2215 1.57 1.38 0.19 S.E. of pred. 0.18 RWVI15 - red well vegetation index 15 N es t . - estimated nitrogen concentration Abs. d i f f . - absolute d i f f e r e n c e between observed and estimated concentrations 0.1561 1.90 2.13 0.23 0.1572 2.12 2.11 0.01 0.1634 2.02 2.01 0.01 0.1907 1.89 1.66 0.23 0.1858 1.94 1.72 0.22 0.1820 1.69 1.76 0.07 0.1845 1.73 1.73 0.00 0.1938 1.50 1.63 0.13 S.E. of pred. 0.11 N obs. - observed nitrogen concentration S.E. of pred. - standard e r r o r of p r e d i c t i o n vo The divisions used by DIAGFOLI to categorize the severity of Douglas-fir nitrogen deficiencies are listed in Table 18. Clearly, a •difference of 0.34 percent nitrogen between the actual and estimated f o l i a r concentration can greatly influence the interpretation of plant nutrient status. An error of this magnitude can mean the difference between a Douglas-fir being diagnosed as severely nitrogen deficient or having access to more than adequate levels of nitrogen. Foliar nitrogen predicting models may not currently provide reliable estimates; however, red well vegetation index 15, and other spectral measures highly correlated with nitrogen, may prove useful indices of relative needle nitrogen concentration. As fo l i a r nitrogen decreased vegetation index 15 increased. If several spectral reflectance measurements made from one seedling consistently resulted in high values for vegetation index 15, and low values for another seedling, then in a l l likelihood the f i r s t seedling would have a greater f o l i a r nitrogen concentration than the second. Therefore, i t would be possible to follow changes in nitrogen concentration. 71 Table 18. DIAGFOLI's interpretation of Douglas-fir f o l i a r nitrogen concentrations. Foliar nitrogen Interpretation concentration (%) 0. 00 to 1. 05 very severely deficient 1. 05 to 1. 30 severely deficient 1. 30 to 1. 45 slight moderate deficiency > 1.45 adequate Source: Ballard and Carter (1986) 72 5.2 PHOSPHORUS SERIES OF TREATMENTS 5 .2 .1 S e e d l i n g growth Only one mortality occurred in the 1987 series of phosphorus treatments. One seedling died when subjected to 400 mg/L phosphorus. Two Douglas-fir died in the 1988 phosphorus treatments, one for each of the 50 and 400 mg/L treatments. A l l four growth variables measured demonstrated similar responses over the range of phosphorus treatments (Table 19) . There was a slight increase in growth from 1 to 5 mg/L, between 5 and 100 mg/L the response was essentially unchanged, except for a slight decrease at 25 mg/L, and from 100 mg/L to the most concentrated phosphorus treatments there was a large decrease in growth. The optimum phosphorus concentration for Douglas-fir seedling growth was approximately 10 mg/L. This i s lower than the optimum concentration of 30 mg/L phosphorus determined by van den Driessche (1969) for Douglas-fir seedlings. 5 .2 .2 F o l i a r A n a l y s i s Foliage appearance varied with phosphorus concentration applied, and indicated the occurrence of one or more nutrient deficiencies. Treatments of 1 to 2 5 mg/L phosphorus produced foliage of a dark green, with the exception of 1 to 2 mm of chlorosis at the needle tips. Approximately 5 percent of needles grown at 1 and 5 mg/L phosphorus developed necrotic tips. 73 Table 19. Mean phosphorus growth rate. Phosphorus Diameter Height Stem Root applied change change weight weight (mg/L) (mm) (cm) (g) (g) 1 (± 4.3abc 0.63) 9. lab (± 2.53) (± 3.74ab 1.39) (± 4.35ab 1.97) 5 (± 5.0bc 1.48) 10.8b (± 4.58) (± 5.32ab 2.79) (± 7.04b 3 .51) 10 (± 5.1bc 1.72) 11.2b (± 3.54) (± 5.44b 2.24) (± 7.85b 3.04) 25 (± 4.6abc 1.53) 9.2ab (± 2.53) (± 3.56ab 0.76) (± 5.30ab 2.57) 50 (± 4.9abc 0.39) 10.8b (±3.27) (± 4.67ab 1.77) (± 7.58b 3.38) 100 (± 5. 2c 1.68) 11.7b (± 3.76) (± 4.76ab 1. 60) (± 5.76ab 2 . 01) 250 (± 3. 0a 1.01) 6.3a (± 2.74) (± 2.91a 1.22) (± 3.34a 2.01) 400 (± 3. 2ab 0.76) 5.7a (± 0.46) (± 3.03ab 1.07) (± 3.27a 1.20) Means within each column with the same letter are not significantly different as judged by Tukey's honestly significantly test (P<0.05). Values are means of 9 observations for each of treatments 50 and 400 mg/L phosphorus. A l l other values are means of 10 observations. Numbers in parentheses are standard deviations. Diameter and height changes were those which occurred from May 19 to October 2, 1988, while stem and root weights refer to measurements made on October 2, 1988. Seedlings treated with phosphorus at 5 or 10 mg/L had the darkest green foliage, and were the least chlorotic and necrotic. None of the treatments low in phosphorus demonstrated the purple-tinged foliage characteristic of anthocyanin accumulation in severely phosphorus-deficient plants (Bidwell, 1979; Marschner, 1986). There was a large increase in chlorosis and necrosis between seedlings treated with 25 and 50 mg/L phosphorus. Foliage discoloration increased for each successively concentrated phosphorus treatment, such that, seedlings produced at 400 mg/L phosphorus were approximately 50 percent chlorotic and 25 percent necrotic. Chlorosis started at the needle t i p then spread in a uniform manner towards the leaf stem. Frequently, necrosis developed on chlorotic foliage, beginning with needle tips. Tables 20 to 22 summarize fo l i a r phosphorus and chlorophyll concentrations, and results of the DIAGFOLI diagnostic program for the phosphorus series of treatments. In 1987, f o l i a r phosphorus levels increased with each increasingly concentrated phosphorus treatment. While composite 1988 data followed the same trend for treatments 1 to 100 mg/L, f o l i a r phosphorus levels decreased from 0.81 to 0.70 percent between treatments 100 and 250 mg/L, and then increased to 0.83 percent for seedlings treated with 400 mg/L phosphorus. Foliar phosphorus concentrations of seedlings grown in 1987 and 1988 were essentially the same for treatments 1 to 50 mg/L phosphorus; however, for treatments 100 to 400 mg/L, 1987 f o l i a r phosphorus levels were much greater than those of 1988. Foliage of Douglas-fir treated with 400 mg/L phosphorus in 1987 had more than Table 20. Composite f o l i a r phosphorus and t o t a l chlorophyll concentrations f o r the 1987 phosphorus s e r i e s of treatments. F o l i a r concentrat ions Nutrient status Phosphorus applied (mg/L) %P CHL (mg/g) Phosphorus Other elements 1 0.14 1.12 s l i g h t d eficiency Zn - probable d e f i c i e n c y 5 0.17 1.38 adequate ^possible deficiency 10 0.23 1.77 adequate Zn - possible d e f i c i e n c y 25 0.48 0.36* adequate 50 0.70 1.43 adequate 100 1.04 0.99 adequate 250 1.34 1.05 adequate 400 1.93 0.71 adequate %P - percent phosphorus CHL - t o t a l c h l o r o p h y l l + - based on P/N r a t i o * - deleted from data analysis Ul Table 21. Composite f o l i a r phosphorus and t o t a l chlorophyll concentrations f o r the 1988 phosphorus s e r i e s of treatments. F o l i a r concentrations Nutrient status Phosphorus applied (mg/L) %P CHL (mg/g) Phosphorus Other elements 1 0.17 (+0.02) 1.47 (+0.10) adequate ^possible d e f i c i e n c y K - p o s s i b l e s l i g h t d e f i c i e n c y 5 0.20 (+0.02) 1.36 (+0.17) adequate K - po s s i b l e s l i g h t d e f i c i e n c y 10 ,- 0.26 (+0.02) 0.82 (+0.18) adequate K - p o s s i b l e s l i g h t d e f i c i e n c y Zn - p o s s i b l e d e f i c i e n c y 25 0.41 (+0.03) 1.19 (+0.15) adequate 50 0.65 (+0.05) 0.98 (+0.09) adequate 100 0.81 (+0.07) 0.77 (+0.18) adequate 250 0.70 (+0.09) 0.61 (+0.14) adequate N - s l i g h t t o moderate d e f i c i e n c i e s Ca - p o s s i b l e s l i g h t to moderate d e f i c i e n c y 400 0.83 (+0.08) 0.56 (+0.02) adequate %P - percent phosphorus CHL - t o t a l chlorophyll + - based on P/N r a t i o Numbers i n parentheses are standard deviations. Standard deviations f o r f o l i a r concentrations are based on data from three i n d i v i d u a l seedlings, whereas standard deviations f o r s p e c t r a l measurements are based on a l l su r v i v i n g seedlings. Table 22. F o l i a r phosphorus and t o t a l chlorophyll concentrations f o r i n d i v i d u a l Douglas-fir seedlings subjected to the 1988 phosphorus s e r i e s of treatments. F o l i a r concentrations Nutrient Status Phosphorus applied %P (mg/L) CHL (mg/g) Phosphorus Other elements 0.16 0.13 0.15 1.05 1.24 1.14 adequate s l i g h t d e f i c i e n c y s l i g h t d eficiency K - p o s s i b l e s l i g h t d e f i c i e n c y Ca - po s s i b l e s l i g h t to moderate d e f i c i e n c y Zn - pos s i b l e d e f i c i e n c y K - s l i g h t t o moderate d e f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Zn - po s s i b l e d e f i c i e n c y 0.17 0.21 0.17 1.39 1.58 1.24 adequate adequate adequate K - pos s i b l e s l i g h t d e f i c i e n c y 10 0.22 0.26 0.23 1.24 1.06 1.41 adequate adequate adequate K - pos s i b l e s l i g h t d e f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Zn - po s s i b l e d e f i c i e n c y K - p o s s i b l e s l i g h t d e f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Zn - probable d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 25 0.41 0.39 0.45 1.46 1.70 1.43 adequate adequate adequate K - po s s i b l e s l i g h t d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of de f i c i e n c y 50 0.65 0.72 0.62 1.16 0.99 1.07 adequate adequate adequate Continued Table 22. Continued. F o l i a r concentrations Nutrient status Phosphorus applied %P CHL Phosphorus Other elements (mg/L) (mg/g) 100 0.58 0.47 adequate N very severe d e f i c i e n c y 0.69 0.68 adequate N - s l i g h t t o moderate d e f i c i e n c y 0.71 0.83 adequate 250 0.79 0.63 adequate N severe d e f i c i e n c y Ca - l i t t l e , i f any d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.61 0.43 adequate N - severe d e f i c i e n c y Ca - moderate to severe d e f i c i e n c y Mg - p o s s i b l e s l i g h t to moderate d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.74 0.70 adequate Ca - l i t t l e , i f any d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y of d e f i c i e n c y 400 0.75 0.52 adequate N _ s l i g h t t o moderate d e f i c i e n c y • Ca - p o s s i b l e s l i g h t to moderate d e f i c i e n c y Mg - p o s s i b l e s l i g h t to moderate d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.76 0.49 adequate Ca - p o s s i b l e s l i g h t to moderate d e f i c i e n c y Mg - l i t t l e , i f any d e f i c i e n c y 0.61 0.49 adequate N - severe d e f i c i e n c y Ca - severe d e f i c i e n c y Mg - p o s s i b l e s l i g h t to moderate d e f i c i e n c y Zn - p o s s i b l e d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y of d e f i c i e n c y %P - percent phosphorus CHL - t o t a l c h l o r o p h y l l 09 twice the f o l i a r phosphorus of seedlings subjected to the same treatment in 1988. The present experiment does not provide an explanation for the discrepancies between years. According to DIAGFOLI's interpretations, the absolute f o l i a r phosphorus levels and nitrogen/phosphorus ratios suggest that seedlings treated with 1 mg/L phosphorus were supplied with slightly deficient to adequate amounts of phosphorus. The nitrogen/phosphorus ratio for the 1987 composite data also indicated possible phosphorus deficient conditions for seedlings grown at 5 mg/L phosphorus. While DIAGFOLI reported adequate levels of phosphorus for a l l other treatments, phosphorus f e r t i l i z a t i o n at 100 to 400 mg/L resulted in slight to severe nitrogen deficiencies. Several investigators (ex. Beaufils, 1973; Everard, 1973; Morrison, 1974; Ingestad et al. , 1981) have suggested that the proportion of nitrogen, phosphorus, sulphur and other nutrients may have more influence on growth than the absolute levels of these elements. Increasing the supply of only one mineral nutrient stimulates growth, which in turn can induce a deficiency of the other by the dilution effect (Timmer, 1991); consequently, the c r i t i c a l level of phosphorus increases as the nitrogen concentration increases, and vice versa (Marschner, 1986). In the same way that high levels of nitrogen (100 mg/L) relative to phosphorus may have induced phosphorus deficiencies in 1 to 5 mg/L phosphorus treated seedlings, high levels of phosphorus (100 to 400 mg/L) induced nitrogen deficiencies. Heilman and Ekuan (1973) reported increases in f o l i a r phosphorus and decreases in fo l i a r 80 nitrogen levels for Douglas-fir and western hemlock seedlings f e r t i l i z e d with phosphate at rates of 0, 225 and 1120 kg/ha. Data for the 1988 individual seedlings indicated several possible calcium deficiencies which tended to increase in severity with concentrated phosphorus treatments. These deficiencies, in part, can be attributed to some calcium being precipitated out of solution, and the dilution effect discussed earlier. Precipitates formed when phosphorus nutrient solutions were stored without shaking for one week. The uptake of Ca 2 + could have been competitively depressed by the increased presence of other cations (e.g. H+ and NH4+) in concentrated phosphorus nutrient solutions. According to Mengel and Kirkby (1987) the absorption of cations tends to be a non-specific process, depending mainly on the concentration of the cation species in the nutrient medium, and in some cases also on the specific permeability of membranes to particular cation species. Because Ca 2 + i s readily replaced by other cations from i t s binding sites at the exterior surface of the plasma membrane, the Ca 2 + requirement increases with increasing external concentrations of other cations (Wyn Jones and Lunt, 1967; Burstrom, 1968; Marschher, 1986). The same processes restricting the uptake of calcium could have also resulted in the possible potassium deficiencies noted for the 1 to 25 mg/L phosphorus treatments. 81 The high frequency of possible zinc, copper and magnesium deficiencies for a l l phosphorus, nitrogen and sulphur treatments suggest that insufficient amounts were added to the nutrient solutions. These conditions may have been aggravated by inhibition of micronutrient absorption by the high levels of cations added with concentrated phosphorus treatments (Chaudhry and Loneragan, 1972) , and the dilution of zinc, copper and magnesium in plant tissues by promotion of plant growth through phosphorus f e r t i l i z a t i o n (Loneragan et a l . , 1979). Numerous investigators, including Millikan (1963), Watanabe et al. (1965), Boawn and Brown (1968), and Loneragan et al. (1979), have reported that high rates of phosphorus f e r t i l i z a t i o n can induce zinc deficiencies in agricultural crops. 5.2.3 Spectral Reflectance Curves Foliage of Douglas-fir treated with phosphorus demonstrated the same basic spectral reflectance pattern as nitrogen treated seedlings (Figures 13 and 14). However, increasingly concentrated phosphorus applications resulted in a different sequence of spectral changes than nitrogen treated seedlings. Treatments low (1 to 10 mg/L) in phosphorus produced dark green foliage low in visi b l e spectral reflectance, whereas seedlings grown at concentrated (100 to 400 mg/L) phosphorus levels had high visible spectral reflectance. Visible reflectance decreased or remained constant from 1 to 10 mg/L phosphorus, then increased for foliage 82 80 70-400 500 600 700 800 900 1000 1100 WAVELENGTH (nm) 1 mg/L P 5 mg/L P ' 10 mg/L P 25 mg/L P 400 500 600 700 800 900 1000 1 100 WAVELENGTH (nm) 50 mg/L P 100 mg/L P 250 mg/L P " — 400 mg/L P F i g u r e 13. 1987 s p e c t r a l r e f l e c t a n c e curves showing the e f f e c t s of phosphorus l e v e l s on needle r e f l e c t a n c e a t the wavelength i n t e r v a l 400 t o 1100 nm. 83 "0 u 1 1 1 1 1 1 1 4-00 500 600 700 800 900 1000 1100 WAVELENGTH (nm) 1 mg/L P 5 mg/L P 1 0 mg/L P 25 mg/L P 70 400 500 600 700 800 900 1000 1 100 WAVELENGTH (nm) • 50 mg/L P 1 00 mg/L P 250 mg/L P 400 mg/L P Figure 14. 1988 spectral reflectance curves showing the effects of phosphorus levels on needle reflectance at the wavelength interval 400 to 1100 nm. Plotted data are from the composite data set. 84 of seedlings grown at 10 to 400 mg/L. Percent spectral reflectance at the red rise for treatments 10 to 400 mg/L phosphorus increased relative to blue and green reflectance resulting in narrow, shallow, red well regions, and a shift of the red edge toward the blue end of the spectrum. No relationship was found between spectral reflectance in the near-infrared and phosphorus le^vel. The sequence of spectral reflectance changes can be attributed to differences in the chlorophyll concentration associated with the various phosphorus treatments. Phosphorus compounds play several essential structural and development roles in energy transfer and photosynthetic processes. Chloroplast replication and development of photosynthetic systems would not be possible without nucleic acids, in which phosphorus plays a major role (Jagendorf, 1973). Total chlorophyll and f o l i a r phosphorus were significantly correlated for a l l three data sets, with correlation coefficients ranging from -0.70 to -0.81 (Table 23). Plots of total chlorophyll versus f o l i a r phosphorus demonstrated that chlorophyll i n i t i a l l y increased with phosphorus, but then decreased with the high f o l i a r phosphorus levels associated with treatments high in phosphorus (Figure 15). This resulted in a negative relationship, due primarily to the nitrogen and calcium deficiencies induced at high (100 to 400 mg/L) phosphorus f e r t i l i z a t i o n levels, and the subsequent destruction or manufacture of less chlorophyll. Consequently, nitrogen and calcium deficiencies degraded the correlation between chlorophyll and f o l i a r phosphorus resulting in less of a linear relationship between chlorophyll and needle 85 Table 23. Correlation between total chlorophyll and the f o l i a r concentration of phosphorus. df Correlation coefficient 1987 composite 5 -0.77* samples 1988 composite 6 -0.81* samples 1988 individual 22 -0.70** seedlings * - p<0.05 ** - p<0.01 df - degrees of freedom phosphorus levels than that between chlorophyll and fo l i a r nitrogen. The association between f o l i a r phosphorus levels and total chlorophyll could have been further complicated by possible micronutrient and potassium deficiencies, and at 1 to 5 mg/L phosphorus treatments by possible phosphorus deficiencies. Kirk and Tilney-Bassett (1967) claim that phosphorus deficiency in higher plants does not always cause chlorosis. Phosphorus deficiency appears to allow normal chloroplast development followed by degeneration of the internal structure (Thomson and Weier, 1962). Figure 15. Relationship between total chlorophyll (CHL) and f o l i a r phosphorus concentration (P) for the 1988 individual seedling data set. 87 5 . 2 . 4 C h l o r o p h y l l a / C h l o r o p h y l l b A l l chlorophyll a to chlorophyll b ratios for the 1987 composite data set were high (Figure 16) ; none of the f o l i a r samples were diagnosed as very severely or severely nutrient deficient. The ratio for the only severely nutrient deficient f o l i a r sample of the 1988 composite data set was about the same as those composite samples considered not nutrient deficient (Figure 17). However, seedlings diagnosed as very severely and severely nutrient deficient in the 1988 individual seedling data set had lower chlorophyll a/chlorophyll b ratios than those with minor or no nutrient deficiencies (Figure 18) . Consequently, a consistent relationship did not exist between the ratio of chlorophyll a to chlorophyll b and the severity of nutrient deficiency for phosphorus treated seedlings. 88 4.5 1 ' 5 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 PHOSPHORUS(%) + — POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 16. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r phosphorus and severity of nutrient deficiency for the 1987 phosphorus composite data set. 89 0.3 0.4 0.5 0.6 PHOSPHORUS(%) © — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 17. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r phosphorus and severity of nutrient deficiency for the 1988 phosphorus composite data set. 90 0.3 0.4 0.5 0.6 PHOSPHORUS (%) 0.7 0.8 • —VERY SEVERELY NUTRIENT DEFICIENT 0 — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 18. Relationships between the rat i o of chlorophyll a/chlorophyll b, f o l i a r phosphorus and severity of nutrient deficiency for the 1988 phosphorus individual seedling data set. 91 5.2.5 Green Reflectance Peak and Red Rise Tables 24 to 26 present percent spectral reflectance at the green reflectance peak (559 nm in 1987; 554 nm in 1988) and red rise (630 nm) for phosphorus treated seedlings. Scatter diagrams of f o l i a r phosphorus versus spectral reflectance at the green reflectance peak and red rise approximated exponential saturation relationships (Figure 19) . Foliar phosphorus concentrations rapidly increased relative to the green reflectance peak and red rise, then reached a plateau where phosphorus levels changed l i t t l e with increased reflectance at the green reflectance peak. Spectral reflectance at the green reflectance peak and red rise varied with chlorophyll concentration. As f o l i a r phosphorus levels rapidly increased for seedlings treated with 1 to 25 mg/L phosphorus, there was a corresponding increase in total chlorophyll which resulted in a decrease in visible reflectance. While f o l i a r phosphorus levels changed l i t t l e over treatments 25 to 400 mg/L phosphorus nutrient deficiencies, especially nitrogen, resulted in decreased chlorophyll levels, and a subsequent increase in spectral reflectance at the green reflectance peak and red rise. Table 24. Percent spectral reflectance at the green reflectance peak (559 nm) and red rise (630 nm), and wavelength of the red edge for composite fo l i a r samples of the 1987 phosphorus series of treatments. Phosphorus Green applied reflectance Red rise Red edge (mg/L) peak (%) (%) (nm) 1 12.5 7.0 709 5 12.0 7.5 700 10 9.5 4.5 726 25 13.5 8.0 720 50 11.5 7.0 723 100 14.5 8.0 712 250 12.0 6.5 712 400 17.5 15.0 703 93 Table 25. Percent spectral reflectance at the green reflectance peak (554 nm) and red rise (630 nm), and wavelength of the red edge for composite f o l i a r samples of the 1988 phosphorus series of treatments. Phosphorus Green applied reflectance Red rise Red edge (mg/L) peak (%) (nm) (%) 1 16.06 8.01 700 (±1.98) (±1.20) (±6.91) 5 13.90 7.13 704 (±3.72) (±1.99) (±9.50) 10 16.86 7.78 706 (±4.09) (±2.28) (±8.22) 25 15.41 8.10 705 . (±3.00) (±2.78) (±9.88) 50 15.62 7.02 703 (±2.59) (±1.29) (±4.42) 100 17.99 10.61 703 (±3.99) (±5.75) (±7.89) 250 20.08 13.25 695 (±2.66) (±4.31) (±4.74) 400 19.05 11.85 701 (±3.69) (±2.34) (±3.36) Numbers in parentheses are standard deviations. Standard deviations for f o l i a r concentrations are based on data from three individual seedlings, whereas standard deviations for spectral measurements are based on a l l surviving seedlings. 94 Table 26. Percent spectral reflectance at the green reflectance peak (554 nm) and red rise (630 nm), and wavelength of the red edge for individual Douglas-fir seedlings subjected to the 1988 phosphorus series of treatments. Phosphorus Green applied reflectance Red rise Red edge (mg/L) peak (%) (%) (nm) 1 19.14 10.83 703 15.41 7.68 712 17.33 7.89 709 5 11.15 5.16 718 12.83 5.52 707 12.91 6.27 702 10 14.61 6.54 708 13 .73 5.62 711 15.74 7.19 706 25 13.31 6.26 717 15.18 7.19 708 13.16 6.12 709 50 16.17 7.58 706 12.58 5.39 714 17.43 7.00 703 100 26.42 23.76 691 21.78 17.06 700 17.61 8.74 699 250 20.20 14.42 702 22.44 12.84 698 16.12 8.10 703 400 19.81 10.36 703 15.72 15.25 701 24 .18 13.15 698 95 0.8-0.7n 0.6H CO o I 0.4-O I o_ 0.3H 0.2-0.1-r = -0.44*. p<0.05; n = 24 1 10 12 14 16 18 20 22 24 SPECTRAL REFLECTANCE (%) AT 630 nm Figure 19. Relationship between f o l i a r phosphorus concentration and percent spectral reflectance at the red rise (630 nm) for the 1988 individual seedling data set. A transformation for linearizing exponential saturation equations, as outlined by Spain (1982), was conducted so that correlation coefficients could be calculated, and the least squares method used to f i t regression lines. The transformation involved plotting the logarithm of (A - percent f o l i a r phosphorus) versus percent spectral reflectance at the green reflectance peak (or red rise) , where A i s the value of an asymptote. Asymptotes were estimated by adding 0.01 percent to the maximum f o l i a r phosphorus concentration of each data set, thus eliminating the problem of calculating the logarithm of zero or negative numbers. Data transformation did not yield strong linear relationships between transformed variables (Figure 2 0) ; moreover, the only s t a t i s t i c a l l y significant correlation coefficient calculated for the relationship logarithm(A - percent f o l i a r phosphorus) versus percent spectral reflectance at the green reflectance peak was -0.81 (p<0.05) for the 1987 composite data set (Table 27). Correlation coefficients were greater for transformed red rise data, two of which were s t a t i s t i c a l l y significant. 97 Figure 20. Relationship between logarithm(0.80 - percent f o l i a r phosphorus (P)) versus percent spectral reflectance at the red rise (RR; 630 nm) for the 1988 individual seedling data set. 98 Table 27. Correlation between the f o l i a r concentration of phosphorus and green reflectance peak, red rise, and red edge measurements. df GRP RR RE 1987 6 -0.81* -0.92** -0.29ns composite (0.48) (0.32) (0.66) samples 1988 6 -0.68ns -0.70ns -0.36ns composite (0.54) (0.52) (0.28) samples 1988 22 -0.36ns -0.44* -0.48* individual (0.47) (0.45) (0.22) seedlings ns - not significant * - p<0.05 ** - p<0.01 df - degrees of freedom GRP - green reflectance peak RR - red rise RE - wavelength of the red edge Numbers in parentheses represent the standard error of estimate. Green reflectance peak and red rise correlation coefficients were calculated for transformed data. 5.2.6 Red Edge Tables 24 to 26 l i s t the wavelength of the red edge for the phosphorus series of treatments. Plots of f o l i a r phosphorus versus wavelength of the red edge did not demonstrate a close relationship, linear, or otherwise (Figure 21). Correlation coefficients ranged from -0.29 (not significant) for 1987 composite data to -0.48 (p<0.05) for 1988 individual seedling data (Table 27) . Changes in the red edge were more closely related to total chlorophyll concentration. As chlorophyll increased for seedlings treated with 1 to 25 mg/L phosphorus, the wavelength of the red edge shifted to longer wavelengths. In contrast, reductions in total chlorophyll associated with 25 to 400 mg/L phosphorus treatments resulted in shifts to shorter wavelengths. 100 Figure 21. Relationship between f o l i a r phosphorus (P) and the wavelength of the red edge (RE) f o r the 1988 i n d i v i d u a l seedling data set. 101 5.2.7 Vegetation Indices Vegetation indices liste d in Tables 1 and 2 were calculated for the phosphorus series of treatments, f i r s t using percent spectral reflectance at the red well (674 nm) , and then with reflectance at the red rise (63 0 nm) . Red well and red rise vegetation indices 4, 6, 10, 14 and 15 demonstrated exponentially increasing saturation equations with f o l i a r phosphorus like those of the green reflectance peak and red rise. Foliar phosphorus levels rapidly increased over a narrow range of vegetation index values, then reached a plateau where f o l i a r phosphorus changed l i t t l e . Graphs of red well and red rise vegetation indices 1, 2, 3, 5, 7, 8, 9, 11, 12 and 13 versus needle phosphorus concentration formed exponentially decreasing saturation relationships (Figure 22) . Foliar phosphorus concentrations remained relatively unchanged over a wide range of vegetation index values, then reached a point where phosphorus dramatically decreased over a narrow range of index values. Spain's (1982) procedure for linearizing exponential saturation equations was applied to vegetation index and f o l i a r phosphorus data, and the correlation coefficients calculated for transformed data (Table 28). In most cases use of red rise rather than red well measurements in the calculation of vegetation indices yielded higher correlation coefficients (Table 29) , although none of the differences were s t a t i s t i c a l l y significant. 102 0.65 0.7 0.75 0.8 0.85 RED WELL VEGETATION INDEX 2 0.9 Figure 22. Relationship between f o l i a r phosphorus concentration and red well vegetation index 2 (NIR - red)/(NIR + red) for the 1988 individual seedling data set. 103 Table 28. Correlation between the fo l i a r concentration of phosphorus and vegetation indices calculated using percent spectral reflectance at the red well and red rise. Red well Red rise VI 1987 comp 1988 comp 1988 ind 1987 comp 1988 comp 1988 ind 0.69ns (0.60) 0.79* (0.45) 0.43* (0.45) 0.71* (0.58) 0.72* (0.51) 0.52** (0.43) 0.85** (0.44) 0.83** (0.41) 0.37ns (0.46) 0.87** (0.41) 0.71* (0.52) 0.47* (0.44) 0.85** (0.43) 0.83** (0.41) 0.37ns (0.47) 0.88** (0.40) 0.71* (0.52) 0.46* (0.44) -0.74* (0.55) -0.56ns (0.61) -0.28ns (0.48) -0.88** (0.39) -0.67ns (0.55) -0.49* (0.44) 0.39ns (0.76) 0.54ns (0.62) 0.17ns (0.49) 0.65ns (0.63) 0.65ns (0.55) 0.50* (0.43) -0.82* (0.47) -0.54ns (0.62) -0.29ns (0.48) -0.90** (0.35) -0.66ns (0.55) -0.49* (0.44) 0.70ns (0.59) 0.55ns (0.61) 0.25ns (0.48) 0.85** (0.44) 0.66ns (0.55) 0.50* (0.43) 0.65ns (0.62) 0.67ns (0.54) 0.37ns (0.47) 0.65ns (0.62) 0.67ns (0.54) 0.37ns (0.47) 0.53ns (0.70) 0.57ns (0.60) 0.23ns (0.49) 0.83** (0.46) 0.66ns (0.55) 0.50* (0.43) 10 -0.75* (0.54) -0.72* (0.51) -0.35ns (0.47) -0.84** (0.44) -0.70ns (0.52) -0.60** (0.40) Continued 104 Table 28. Continued Red well Red rise VI 1987 1988 1988 1987 1988 1988 comp comp ind comp comp ind 11 0.03ns 0.03ns -0.08ns 0.76* 0.41ns 0.18ns (0.67) (0.73) (0.50) (0.54) (0.67) (0.49) 12 -0.18ns 0.55ns 0.58** -0.02ns 0.56ns 0.59** (0.81) (0.61) (0.41) (0.82) (0.61) (0.40) 13 0.79* 0.76* 0.43* 0.79* 0.71* 0.45* (0.50) (0.48) (0.45) (0.50) (0.52) (0.45) 14 -0.84** -0.82* -0.36ns -0.87** -0.71* -0.48* (0.44) (0.42) (0.47) (0.40) (0.52) (0.44) 15 -0.55ns -0.63ns -0.33ns -0.52ns -0.64ns -0.31ns (0.68) (0.57) (0.47) (0.70) (0.56) (0.47) df 6 6 22 6 6 22 * ** ns df p<0.05 p<0.01 not significant degrees of freedom VI - vegetation indices 1987 comp - 1987 composite samples 1988 comp - 1988 composite samples 1988 ind - 1988 individual seedlings Correlation coefficients were calculated for transformed data. Table 29. Differences in correlation between f o l i a r phosphorus concentration and recalculated vegetation indices using percent spectral reflectance at the red rise. Vegetation 1987 1988 1988 indices composite composite individual samples samples seedlings 1 + - + 2 . + - - + 3 + - + 4 + + + 5 + + + 6 + + • + 7 + + + 8 0 0 0 9 + + + 10 + - + 11 + • + + 12 - + + 13 + - + 14 + - + 15 - + + - increase in correlation - - decrease in correlation 0 - no change in correlation Note: None of the changes were s t a t i s t i c a l l y significant at the 0.05 level. 106 Red rise vegetation indices 4, 5, 6, 7, 9 and 11 were consistently more highly correlated with f o l i a r phosphorus than red well indices for composite and individual seedling data sets. Consequently, percent spectral reflectance at the red rise should be used when assessing needle phosphorus concentration with these vegetation indices. Correlation coefficients for vegetation index 8 did not change since i t s calculation does not incorporate a measure of red light reflectance. 5.2.8 Estimation of Foliar Phosphorus For individual seedling data, red rise vegetation index 10 was the index most highly correlated (r=-0.60; p<0.01) with f o l i a r phosphorus, while the highest correlation coefficient for either composite data set occurred in 1987 for red rise vegetation index 6 (r=-0.90; p<0.01). These measurements were more strongly correlated with needle phosphorus concentration than the green reflectance peak, red rise, red edge or other vegetation indices, and thus used in the development of models for estimating f o l i a r phosphorus levels in Douglas-fir seedlings. The exponentially increasing saturation relationship of f o l i a r phosphorus concentration versus red rise vegetation index 10 was best described by two separate regression lines; one for the region of rapid increase in f o l i a r phosphorus, and another for the region where phosphorus levels increased slowly over a wide range of index values (Figure 23). 107 Figure 23. Regression lines and associated data for the relationship between f o l i a r phosphorus concentration (P) and red rise vegetation index 10 (RRVI10) for the 1988 individual seedling data set. 108 Determination of which regression line to use as a model for f o l i a r phosphorus estimation was based on the x-intercept of the two equations. If the value of red rise vegetation index 10 was less than or equal to the x-intercept (0.3287) then %P = -0.9422 + 4.9532 X RRVI10 was used, whereas needle phosphorus concentration was estimated with %P = 0.6579 + 0.0848 X RRVI10 when the index value was greater than 0.3287. Use of the models to estimate f o l i a r phosphorus of individual Douglas-fir seedlings subjected to the nitrogen and sulphur series of treatments resulted in standard errors of prediction of 0.13 and 0.15 percent phosphorus, respectively (Table 30). The same method was used to develop regression lines to describe the relationship of red rise vegetation index 6 and needle phosphorus concentration for 1987 composite data. Predictive capabilities were tested by comparing estimates and actual f o l i a r phosphorus levels of composite f o l i a r samples for the 1987 nitrogen treated seedlings, and 1988 composite f o l i a r samples of nitrogen, phosphorus and sulphur treated seedlings (Table 31) . Phosphorus estimates of composite samples tended to be less accurate than those of f o l i a r samples from individual seedlings; standard errors of prediction ranged from 0.11 to 0.26 percent phosphorus. Consequently, estimates were recalculated using equations developed from red rise vegetation index 10 and 1988 individual seedling data (Table 32) . The resulting phosphorus estimates were more accurate. Table 30. Comparison of f o l i a r phosphorus l e v e l s f o r 1988 i n d i v i d u a l seedling data as determined by chemical analysis and models developed using red r i s e vegetation index 10. Nitrogen treated seedlings Sulphur t r e a t e d seedlings RRI10 P obs.(%) P est.(%) Abs. d i f f . RRVI10 P obs.(%) P est.(%) Abs. d i f f . 0.2977 0.50 0.53 0.03 0.2394 0.36 0.24 0.12 0.3073 0.64 0.58 0.06 0.2601 0.35 0.35 0.00 0.3272 0.50 0.68 0.18 0.2108 0.34 0.10 0.24 0.3156 0.54 0.62 0.08 0.2539 0.38 0.32 0.06 0.3006 0.57 0.55 0.02 0.2409 0.33 0.25 0.08 0.3519 0.43 0.69 0.26 0.2866 0.40 0.48 0.08 0.3388 0.56 0.69 0.13 0.2761 0.39 0.43 0.04 0.3532 0.53 0.69 0.16 0.2469 0.47 0.28 0.19 0.3033 0.64 0.56 0.08 0.2541 0.46 0.32 0.14 0.2710 0.74 0.40 0.34 0.2581 0.44 0.34 0.10 0.2664 0.58 0.38 0.20 0.2927 0.46 0.51 0.05 0.2945 0.71 0.52 0.19 0.2727 0.37 0.41 0.04 0.2576 0.63 0.33 0.30 0.4765 0.47 0.70 0.23 0.3511 0.71 0.69 0.02 0.2761 0.43 0.42 0.01 0.2853 0.57 0.47 0.10 0.2617 0.43 0.35 0.08 0.2688 0.39 0.39 0.00 0.4637 0.46 0.70 0.24 0.3077 0.47 0.58 0.11 0.4077 0.44 0.69 0.25 0.2312 0.47 0.20 0.27 0.3637 0.33 0.69 0.36 0.2456 0.33 0.27 0.06 0.2879 0.37 0.48 0.11 0.2458 0.45 0.28 0.17 0.4148 0.38 0.69 0.31 0.2477 0.37 0.28 0.09 0.3022 0.44 0.55 0.11 0.2462 0.25 0.28 0.03 0.2865 0.31 0.48 0.17 0.2252 0.38 0.17 0.21 0.2889 0.37 0.49 0.12 0.2779 0.43 0.43 0.00 0.4051 0.32 0.69 0.37 Standard e r r o r of p r e d i c t i o n 0.13 Standard e r r o r o: f p r e d i c t i o n 0.15 RRVI10 - red well vegetation index 10 P obs. - observed phosphorus concentration P e s t . - estimated phosphorus concentration Abs. d i f f . - absolute difference between observed and estimated concentrations The model %P « -0.9422 + 4.9532 x RRVI10 was used when RRVI10 was les s than or equal to 0.3287, whereas %P = 0.6579 + 0.0848 x RRVI10 was used f o r f o l i a r phosphorus estimation when RRVI10 was H greater than 0.3287. o Table 31. Comparison of f o l i a r phosphorus l e v e l s f o r composite samples as determined by chemical analysis and models developed using red r i s e vegetation index 6. 1987 nitrogen treated seedlings 1988 nitrogen treated seedlings RRVI6 P obs.(%) P est.(%) Abs. d i f f . RRVI6 P obs. (%) P est. (%) Abs. d i f f . 40.1932 0.36 0.71 0.35 49.7714 0.66 1.21 0.55 49.9130 0.87 1.21 0.34 49.7273 0.75 0.73 0.02 35.6200 0.93 0.58 0.35 36.0370 0.86 0.59 0.27 28.4348 0.32 0.37 0.05 36.0250 0.55 0.59 0.04 Standard e r r o r of p r e d i c t i o n 0.25 1988 phosphorus treated seedlings 34.1692 0.66 0.54 0.12 34.6208 0.54 0.55 0.01 33.2500 0.70 0.51 0.19 31.0592 0.75 0.45 0.30 33.0324 0.58 0.51 0.07 31.3293 0.41 0.46 0.05 33.2265 0.37 0.51 0.14 31.0530 0.45 0.45 0.00 Standard e r r o r of p r e d i c t i o n 0.11 1988 sulphur treated seedlings RRVI6 P est.(%) P obs.(%) Abs. d i f f . RRVI6 P est.(%) P obs.(%) Abs. d i f ] 29.8004 0.17 0.41 0.24 28.4275 0.38 0.37 0.01 31.0507 0.20 0.45 0.25 29.5973 0.43 0.41 0.02 27.7862 0.26 0.36 0.10 28.9238 0.49 0.39 0.10 32.3638 0.41 0.49 0.08 33.0215 0.44 0.51 0.07 27.2975 0.65 0.34 0.31 35.6530 0.47 0.58 0.11 35.6790 0.81 0.58 0.23 40.4482 0.43 0.72 0.29 39.9416 0.70 0.71 0.01 46.1523 0.41 1.02 0.61 37.5091 0.83 0.64 0.19 50.3672 0.36 1.24 0.88 Standard e r r o r of p r e d i c t i o n 0.18 Standard e r r o r of p r e d i c t i o n 0.26 RRVI6 - red r i s e vegetation index 6 P est. - estimated phosphorus concentration P obs. - observed phosphorus concentration Abs. d i f f . - absolute difference between observed and estimated concentrations The model %P = -0.4451 + 0.0288 x RRVI6 was used when RRVI6 was l e s s than or equal to 40.6626, whereas %P = -1.4166 + 0.0527 x RRVI6 was used for f o l i a r phosphorus estimation when RRVI6 was greater than M 40.6626. H Table 32. Comparison of f o l i a r phosphorus levels f o r composite samples as determined by chemical a n a l y s i s and models developed using red r i s e vegetation index 10. 1987 nitrogen treated seedlings 1988 nitrogen treated seedlings RRVI10 P obs.(%) P est.(%) Abs. d i f f . RRVI10 P obs.(%) P est.(%) Abs. d i f f . 0.3293 0.36 0.69 0.33 0.3220 0.66 0.65 0.01 0.3768 0.66 0.69 0.03 0.3178 0.54 0.63 0.09 0.3182 0.87 0.69 0.18 0.3240 0.70 0.66 0.04 0.3333 0.75 0.69 0.06 0.2893 0.75 0.49 0.26 0.2889 0.93 0.63 0.30 0.3019 0.58 0.55 0.03 0.2800 0.86 0.71 0.15 0.2935 0.41 0.51 0.10 0.2857 0.32 0.49 0.17 0.2696 0.37 0.39 0.02 0.2750 0.55 0.44 0.11 0.2674 0.45 0.38 0.07 Standard err o r of p r e d i c t i o n 0.17 Standard e r r o r of p r e d i c t i o n 0.08 1988 phosphorus treated seedlings 1987 sulphur treated seedlings RRVI10 P obs.(%) P est.(%) Abs. d i f f . RRVI10 P obs.(%) P est.(%) Abs. d i f f . 0.2855 0.17 0.47 0.30 0.2562 0.38 0.33 0.05 0.2829 0.20 0.46 0.26 0.2731 0.43 0.41 0.02 0.2671 0.26 0.38 0.12 0.2717 0.49 0.40 0.09 0.2827 0.41 0.46 0.05 0.2989 0.44 0.54 0.10 0.2690 0.65 0.39 0.26 0.3108 0.47 0.60 0.13 0.3267 0.81 0.68 0.13 0.3325 0.43 0.69 0.26 0.3525 0.70 0.68 0.02 0.3607 0.41 0.69 0.28 0.3309 0.83 0.68 0.15 0.3918 0.36 0.69 0.33 Standard err o r of p r e d i c t i o n 0.16 Standard e r r o r of p r e d i c t i o n 0.16 RRVI10 P obs. P e s t . Abs. d i f f . The model %P whereas %P « greater than 0.3287. - red r i s e vegetation index 10 - observed phosphorus concentration - estimated phosphorus concentration - absolute d i f f e r e n c e between observed and estimated concentrations = -0.9422 + 4.9532 x RRVI10 was used when RRVI10 was l e s s than or equal to 0.3287, 0.6579 + 0.0848 x RRVI10 was used f o r f o l i a r phosphorus estimation when RRVI10 was 112 Standard errors of prediction ranged from 0.08 to 0.17 percent phosphorus, and decreased for every composite data set. Table 33 l i s t s the c r i t e r i a used by Ballard and Carter's (1986) DIAGFOLI program in the interpretation of phosphorus status of Douglas-fir seedlings. A standard error of 0.17 percent phosphorus can have a great influence on nutrient status interpretation, as an error of this magnitude covers the range of interpretations from severely deficient to having adequate levels of phosphorus. While the models developed for estimating f o l i a r phosphorus in Douglas-fir seedlings may not currently provide reliable estimates for assessing nutrient status, the results indicate the potential of red rise vegetation index 10 as a relative index of needle phosphorus concentration. As the value of vegetation index 10 increased, f o l i a r phosphorus levels tended to increase, and thus, could be used to follow changes in phosphorus concentration. Presumably, further study and the incorporation of more data would result in better models. Table 33. DIAGFOLI1s interpretation of Douglas-fir f o l i a r phosphorus concentrations. Foliar phosphorus concentration (%) Interpretation 0.00 to 0.08 0.08 to 0.10 0.10 to 0.15 > 0.15 severely deficient moderately deficient slightly deficient adequate Source: Ballard and Carter (1986) 114 5.3 SULPHUR SERIES OF TREATMENTS 5.3.1 Seedling growth Mortality was low among sulphur treatments. In 1988, one seedling died when subjected to each of the 1 and 400 mg/L treatments. Stem diameter and height change, and dry stem and root weights showed similar responses over the range of 1988 sulphur treatments (Table 34). From 1 to 5 mg/L there was a slight growth increase, reaching a peak in response at 10 or 25 mg/L, and then decreasing with each successively concentrated sulphur solution. The optimum concentration for seedling growth was 10 mg/L. While no other study has determined the optimum sulphur concentration for Douglas-fir seedling growth, 10 mg/L is substantially lower than that determined for some coniferous species. Hocking (1971) reported that f e r t i l i z e r treatments containing 150 mg/L sulphur were optimum for lodgepole pine and white spruce seedling growth; Schomaker (1969) found that optimum eastern white pine seedlings growth occurred at 96 mg/L sulphur. In contrast, Ingestad (1962) found that Scotch pine and Norway spruce seedlings grew best at 20 mg/L sulphur. 115 Table 34. Mean sulphur growth rate. Sulphur applied (mg/L) Diameter change (mm) Height change (cm) Stem weight (g) Root weight (g) 1 (± 4.3ab 1.07) 8. 6ab (± 2.55) (± 4.3 0ab 1.56) (± 6.84cd 3.45) 5 (± 4.7ab 1.51) 9.2abc (± 3.19) (± 4.11ab 1.49) (± 7.32cd 3.32) 10 (± 6. lb 1.33) 13 . 6c (± 4.53) (± 6.69c 2.20) (± 8.33cd 2.66) 25 (± 5.5b 1.51) 12.7bc (± 4.31) (± 6.65c 2.31) (± 9.35d 3.71) 50 (± 5.3b 1.02) 9.9abc (+ 1.90) (± 5.47bc 1.29) (± 5.76bc 2.05) 100 (± 4.6ab 1.16) 8.0a (± 4.08) (± 3.88ab 1.10) (± 5.17abc 1.40) 250 (± 4.7ab 1.64) 6. 6a (± 1.42) (± 3.61ab 1.18) (± 3.01ab 0.84) 400 (± 3.4a 1.53) 6.9a (± 1-51) (± 2.90a 1.35) (± 1.99a 1.17) Means within each column with the same letter are not significantly different as judged by Tukey's honestly significant test (P<0.05). Values are means of 9 observations for each of treatments 1 and 400 mg/L sulphur. A l l other values are means of 10 observations. Numbers in parentheses are standard deviations. Diameter and height changes were those which occurred from May 19 to October 2, 1988, while stem and root weights refer to measurements made on October 2, 1988. 116 5.3.2 Foliar Analysis Douglas-fir seedlings grown at 1 to 5 mg/L sulphur were dark green, and did not demonstrate the pattern of general chlorosis, followed by necrosis, characteristic of severe sulphur deficiency in conifers (Ingestad, 1959; Hacskaylo, 1960; Hacskaylo et al. , 1969) . Sulphur f e r t i l i z a t i o n rates of 10 mg/L produced dark green needles with chlorotic tips. Leader dieback, a symptom of calcium and zinc deficiencies (Smith, 1943; Stoate, 1950; Purnell, 1958) occurred in about 10 percent of seedlings subjected to 5 to 400 mg/L sulphur. With each increasingly concentrated sulphur treatment from 25 to 400 mg/L, foliage became a lighter shade of green, and more chlorotic and necrotic. Needles produced at sulphur applications of 400 mg/L were stunted, with approximately 50 to 80 percent chlorosis and 3 0 to 50 percent necrosis covering their surface. Necrosis f i r s t appeared on chlorotic needle tips, and as with chlorosis, spread uniformly over the leaf, thus suggesting sulphur toxicity (Mudd, 1975; Rennie and Halstead, 1977). The stunted needles and roots of seedlings treated with 400 mg/L sulphur, and the easy manner in which needles were shed for treatments 100 to 400 mg/L are indicative of zinc deficiency (Smith and Bayliss, 1942; Stoate, 1950). Results of the f o l i a r analyses and the DIAGFOLI program are summarized in Tables 35 to 36. Only 22 of the 24 randomly selected Douglas-fir seedlings were analyzed for total sulphur since one seedling in each of the 1 and 50 mg/L sulphur treatments Table 35. 1988 composite f o l i a r sulphur and t o t a l chlorophyll concentrations f o r the sulphur s e r i e s of treatments. F o l i a r concentrations Nutrient status Sulphur applied (mg/L) %S CHL (mg/g) Sulphur Other elements 1 0.18 (+0.03) 1.27 (+0.15) adequate 5 0.51 (+0.06) 1.83 (+0.13) adequate 10 0.33 (+0.04) 1.33 (+0.20) adequate 25 0.45 (+0.11) 0.91 (+0.14) adequate Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 50 0.42 (+0.08) 0.70 (+0.34) adequate Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 100 0.51 (+0.16) 0.72 (+0.09) adequate Cu s l i g h t p o s s i b i l i t y of d e f i c i e n c y 250 0.60 (+0.19) 0.95 (+0.09) adequate Ca Zn Cu p o s s i b l e s l i g h t to moderate d e f i c i e n c y p o s s i b l e d e f i c i e n c y s l i g h t p o s s i b i l i t y of d e f i c i e n c y 400 0.48 (+0.06) 0.68 (+0.09) adequate Ca Mg Zn Cu -moderate to severe d e f i c i e n c y l i t t l e , i f any d e f i c i e n c y p o s s i b l e d e f i c i e n c y s l i g h t p o s s i b i l i t y of d e f i c i e n c y %S - percent sulphur CHL - t o t a l c h l o r o p h y l l Numbers i n parentheses are standard deviations. Standard deviations f o r f o l i a r concentrations are based on data from three i n d i v i d u a l seedlings, whereas standard deviations f o r s p e c t r a l measurements are based on a l l surviving seedlings. H H «0 Table 36. 1988 f o l i a r sulphur and t o t a l chlorophyll concentrations f o r i n d i v i d u a l Douglas-fir seedlings subjected to the sulphur s e r i e s of treatments. F o l i a r concentrations Nutrient status Sulphur applied %S CHL Sulphur Other elements (mg/L) (mg/g) 1 0.17 1.34 adequate Ca _ l i t t l e , i f any d e f i c i e n c y NA 1.52 NA Ca - l i t t l e , i f any d e f i c i e n c y 0.21 1.63 adequate Ca l i t t l e , i f any d e f i c i e n c y 5 0.25 1.16 adequate Ca _ l i t t l e , i f any d e f i c i e n c y 0.37 1.37 adequate Zn - p o s s i b l e d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.33 1.14 adequate 10 0.22 1.32 adequate Cu _ s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.28 0.95 adequate 0.29 1.25 adequate 25 0.56 1.09 adequate 0.41 0.82 adequate Zn - p o s s i b l e d e f i c i e n c y 0.34 0.93 adequate 50 NA 0.52 NA N s l i g h t t o moderate d e f i c i e n c y Ca - p o s s i b l e s l i g h t to moderate d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.52 0.62 adequate Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.40 1.16 adequate Ca - l i t t l e , i f any d e f i c i e n c y Cu s l i g h t p o s s i b i l i t y of d e f i c i e n c y Continued CO Table 36. Continued. F o l i a r concentrations Nutrient status Sulphur applied (rog/L) %S CHL (mg/g) Sulphur Other elements 100 0.61 0.69 adequate Ca possible s l i g h t t o moderate d e f i c i e n c y Zn - possible d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.30 0.59 adequate Ca - possible s l i g h t to moderate d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.49 0.76 adequate Ca - l i t t l e , i f any d e f i c i e n c y Zn - probable d e f i c i e n c y Cu — s l i g h t p o s s i b i l i t y of d e f i c i e n c y 250 0.83 0.83 adequate Ca severe d e f i c i e n c y Mg - l i t t l e , i f any d e f i c i e n c y Zn - possible d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.50 0.81 adequate Ca - moderate t o severe d e f i c i e n c y Zn - probable d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y 0.49 0.67 adequate Ca - moderate to severe d e f i c i e n c y Mg - l i t t l e i f any d e f i c i e n c y Zn - possible d e f i c i e n c y Cu mm s l i g h t p o s s i b i l i t y of d e f i c i e n c y Continued Table 36. Continued. F o l i a r concentrations Nutrient status Sulphur applied %S CHL Sulphur Other elements (mg/L) (mg/g) 400 0.42 0.85 adequate N - s l i g h t t o moderate d e f i c i e n c y K - pos s i b l e s l i g h t d e f i c i e n c y Ca - severe d e f i c i e n c y Mg - moderate t o severe d e f i c i e n c y Zn - probable d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of deficency 0.53 0.69 adequate Ca - severe d e f i c i e n c y Mg - moderate t o severe de f i c i e n c y Zn - possible d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y o f d e f i c i e n c y 0.43 0.71 adequate N - s l i g h t t o moderate d e f i c i e n c y Ca - moderate t o severe d e f i c i e n c y Mg - po s s i b l e s l i g h t to moderate d e f i c i e n c y Zn - probable d e f i c i e n c y Cu - s l i g h t p o s s i b i l i t y of d e f i c i e n c y %S - percent sulphur CHL - t o t a l c h l o r o p h y l l NA - data unavailable; i n s u f f i c i e n t f o l i a r material a v a i l a b l e f o r sulphur a n a l y s i s . ro o 121 possessed insufficient f o l i a r material for analysis. None of the treatments were successful in inducing sulphur deficiencies in Douglas-fir seedlings. This was likely due to sulphur dioxide pollution from o i l refineries, cement plants, and pulp mills in the Greater Vancouver Regional District. From May to October, 1988, precipitation collected at Vancouver International Airport contained mean sulphate concentrations ranging from 0.99 to 2.05 mg/L (B.C. Ministry of Environment, Acid Rain Monitoring Program, pers. comm.). In an unrelated study, Murtha (pers. comm.) identified damage types in small numbers of mature conifers surrounding the University of British Columbia tree nursery suggestive of air pollution. Others have commented on the d i f f i c u l t y inducing sulphur deficiencies in industrialized areas since airborne sulphur can be absorbed directly or dissolved in nutrient medium (Bidwell, 1979). It i s probable that airborne sulphur was ultimately util i z e d by the seedlings. This may explain the low optimum applied sulphur concentration noted for Douglas-fir seedling growth in relation to other conifers. DIAGFOLI*s interpretation of composite and individual seedling f o l i a r data indicated that the sulphur series of treatments induced calcium deficiencies in the Douglas-fir seedlings. The 1988 individual seedling data exhibited a gradient of calcium deficiency starting with possible small calcium deficiencies for seedlings treated with 1 mg/L sulphur and culminating in severe, calcium-deficient conditions for seedlings subjected to 250 and 400 mg/L sulphur. It is likely that these deficiencies increased in severity 122 with increased rates of sulphur f e r t i l i z a t i o n due to sulphur combining with calcium to form precipitates of calcium sulphate (gypsum), and thus being less available for plant growth. White precipitates formed when nutrient solutions were stored for over two days. In addition, the uptake of Ca + 2 could have been competitively depressed by the increased presence of other cations in concentrated sulphur nutrient solutions. This would explain why fo l i a r sulphur concentrations did not increase in proportion to the rate of sulphur f e r t i l i z e r applied. The high incidence of possible zinc, copper and magnesium deficiencies indicate that insufficient amounts were added to the nutrient solutions. Foliar analysis of individual seedlings indicated slight to moderate nitrogen deficiencies for some seedlings treated with 50 and 400 mg/L sulphur. Both nitrogen and sulphur are essential for protein synthesis (Marschner, 1986), and their organic forms have been found to occur in a constant ratio in the foliage of Douglas-f i r (Turner et a l . , 1977). Consequently, application of sulphur without sufficient nitrogen could involve the ut i l i z a t i o n of a l l available nitrogen in the plant. 123 5.3.3 Spectral Reflectance Curves Spectral reflectance plots of sulphur treated Douglas-fir seedlings demonstrated the characteristic pattern of reflected light from green vegetation. Reflectance was low in the blue spectral region, rapidly increasing over the green and reaching a peak at 554 nm, then gradually decreasing over the remainder of the visib l e spectrum (Figure 24). Between 680 and 750 nm spectral reflectance dramatically increased, then remained constant over the near-infrared plateau, with the exception of a slight dip at 969 nm. Needle reflectance varied with each sulphur treatment; the result of differing amounts of chlorophyll and necrosis. Foliar analysis of composite data showed an increase in total chlorophyll concentration from 1 to 5 mg/L sulphur, then a decrease from 5 to 400 mg/L, except for a slight increase in the foliage of the 250 mg/L sulphur treatment. Visible spectral reflectance tended to follow a gradient starting with the low reflectance of foliage produced at 1 mg/L sulphur, and increasing with each successively concentrated treatment. As the sulphur f e r t i l i z a t i o n rate increased, less chlorophyll was formed or maintained, thus the red well region became narrower and less deep, with a corresponding increase at the red rise. The wavelength of the red edge increased from 1 to 10 mg/L sulphur, then kept shifting toward the blue end of the- spectrum for the remaining treatments. 124 u t 1 1 1 , 1 , 400 500 600 700 800 900 1000 1100 WAVELENGTH (nm) 1 mg/L S 5 mg/L S 1 0 mg/L S 25 mg/L S 70 J 1 l 1 1 1 1 1 400 500 600 700 800 900 1000 1100 WAVELENGTH (nm) 50 mg/L S 1 00 mg/L S 250 mg/L S 400 mg/L S Figure 24. 1988 spectral reflectance curves showing the effect of sulphur levels on needle reflectance at the wavelength interval 400 to 1100 nm. Plotted data are from the 1988 composite data set. 125 Treatments with high visible spectral reflectance tended to display high near-infrared reflectance; however, no definite relationship was found between f o l i a r sulphur levels and near-infrared reflectance. The correlation between percent spectral reflectance at 800 nm and f o l i a r sulphur levels for the composite data set was 0.85 and significant at the 0.01 level, whereas that for the individual seedling data was extremely low (-0.004) and not s t a t i s t i c a l l y significant. Although sulphur i s essential for protein synthesis, and thus chlorophyll formation (Marschner, 1986), sulphur was least correlated with total chlorophyll of the three nutrients studied (Table 37) . The correlation coefficients for the composite and individual seedling data sets were -0.22 (not significant) and -0.60 (p<0.01), respectively. As f o l i a r sulphur levels increased the concentration of total chlorophyll tended to decrease (Figure 25) . This i s probably a consequence of the increasing calcium deficiency noted over the range of sulphur treatments. The relationship between chlorophyll and f o l i a r sulphur was likely further degraded by zinc, copper, magnesium, and nitrogen deficiencies. A l l of these conditions have been reported to cause chlorosis in conifers (Morrison, 1974) 126 Table 37. Correlation between total chlorophyll and the f o l i a r concentration of sulphur. df Correlation coefficient 1988 composite 6 -0.22ns samples 1988 individual 20 -0.60** seedlings ** - p<0.01 ns - not significant df - degrees of freedom 5 . 3 . 4 C h l o r o p h y l l a / C h l o r o p h y l l b Neither of the graphs for composite or individual f o l i a r samples of sulphur treated seedlings demonstrated a relationship between chlorophyll a/chlorophyll b ratios and nutrient stress (Figures 26 and 27) . Ratios of f o l i a r samples diagnosed by DIAGFOLI as severely nutrient deficient were in the majority of cases no less or greater than samples classified as having only possible minor deficiencies or none at a l l . As there was no clear and consistent relationship between the ratio of chlorophyll a to b and the severity of nutrient deficiencies for the sulphur, nitrogen and phosphorus treated seedlings i t appears ineffectual as an indicator of nutrient stress. 127 Figure 25. Relationship between total chlorophyll (CHL) and f o l i a r sulphur concentration (S) for the 1988 individual seedling data set. 128 0.2 0.3 0.4 0.5 SULPHUR (%) 0.7 © — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 26. Relationships between the ratio of chlorophyll a/chlorophyll b, f o l i a r sulphur and severity of nutrient deficiency for the 1988 sulphur composite data set. 129 I O _ i I O 0.1 02. o!3 OA 05 06 07 08 0^ 9 SULPHUR (%) 0 — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • _ NOT NUTRIENT DEFICIENT Figure 27. Relationships between the ratio of chlorphyll a/chlorophyll b, f o l i a r sulphur and severity of nutrient deficiency for the 1988 sulphur individual seedling data set. 130 5.3.5 Green Reflectance Peak and Red Rise Both visib l e reflectance and separation between spectral curves for sulphur treatments were greatest at 554 nm; consequently, f o l i a r sulphur levels were plotted against percent spectral reflectance at this wavelength (Figure 28) . The relationship between f o l i a r sulphur and reflectance at the green reflectance peak was best described as linear, reflectance increased with each successively concentrated sulphur treatment (Tables 38 and 39). Significant correlation (r=0.60; p<0.01) existed between f o l i a r sulphur concentration and spectral reflectance at the green reflectance peak for the individual seedling data set (Table 40), but not for composite data (r=0.63). The relationship between needle sulphur concentration and percent spectral reflectance at the red rise was also best described as linear; however, the red rise was less strongly correlated with f o l i a r sulphur (Table 40) . Correlation coefficients for 1988 individual seedling and composite data were 0.46 (p<0.05) and 0.61 (not significant), respectively. As f o l i a r sulphur levels increased there was a corresponding decrease in total chlorophyll due to increasingly severe nutrient deficiencies, especially calcium and nitrogen. This resulted in reduced absorption of visible light, and an increase in spectral reflectance at the red rise and green reflectance peak. 131 Figure 28. Relationship between f o l i a r sulphur concentration (S) and percent spectral reflectance at the green reflectance peak (GRP; 554 nm) for the 1988 individual seedling data set. 132 Table 38. Percent spectral reflectance at the green reflectance peak (554 nm) and red rise (63 0 nm), and wavelength of the red edge for composite f o l i a r samples of the 1988 sulphur series of treatments. Sulphur Green applied reflectance Red rise Red edge (mg/L) peak (%) (nm) (%) 1 11.81 5.59 706 (±3.33) (±2.34) (±8.34) 5 12.25 5.89 716 (±1.93) (±1.15) (±5.21) 10 12.53 6.01 727 (±1.95) (±0.99) (±12.40) 25 15.66 8.37 705 (±2.45) (±2.82) (±1.90) 50 15.13 8.79 700 * (±0.70) (±4.14) (±6.98) 100 * 15.21 9.80 697 (±1.19) (±3.10) (±2.75) 250 15.73 11.47 698 (±3.25) (±2.05) (±5.81) 400 16.85 13.94 695 (±1.93) (±3.12) (±6.22) Numbers in parentheses are standard deviations. Standard deviations for f o l i a r concentrations are based on data from three individual seedlings, whereas standard deviations for spectral measurements are based on a l l surviving seedlings. 133 Table 39. Percent spectral reflectance at the green reflectance peak (554 nm) and red rise (630 nm), and wavelength of the red edge for individual Douglas-fir seedlings subjected to the 1988 sulphur series of treatments. Sulphur Green applied reflectance Red rise Red edge (mg/L) peak (%) (nm) (%) 1 12.00 5.17 703 13.14 5.35 712 11.64 4.58 720 5 10.53 5.03 704 11.59 5.01 718 14.72 7.34 711 10 14.22 6.50 701 9.09 4.38 722 11.89 5.58 711 25 17.10 7.26 703 18.63 9.47 699 16.72 7.91 716 50 17.02 19.92 692 17.42 8.50 700 14.35 7.08 719 100 14.69 14.86 692 17.77 15.76 693 14.69 10.96 714 250 17.81 8.75 697 14.57 13.47 706 18.95 10.36 705 400 16.23 7.60 713 18.81 9.81 698 17.16 14.71 697 134 Table 40. Correlation between the f o l i a r concentration of sulphur and green reflectance peak, red rise, and red edge measurements. df GRP RR RE 1988 composite 6 0.63ns 0.61ns -0.39ns samples 1988 individual 22 0.60** 0.46* -0.40ns seedlings * - p<0.05 ** - p<0.01 ns - not significant df - degrees of freedom GRP - green reflectance peak RR - red rise RE - wavelength of the red edge 5 . 3 . 6 Red Edge Tables 38 and 39 present red edge data for sulphur treatments. As f o l i a r sulphur levels increased, total chlorophyll decreased, resulting in a shift of the red edge towards the blue end of the spectrum. The relation between needle sulphur concentration and the red edge appeared linear in nature (Figure 29), even though the correlation between these two variables was not s t a t i s t i c a l l y significant for either data set (Table 40) . 135 Figure 29. Relationship between f o l i a r sulphur concentration (S) and the wavelength of the red edge (RE) for the 1988 individual seedling data set. 136 5.3.7 Vegetation Indices Vegetation indices 1 to 15, listed in Tables 1 and 2, were calculated for the sulphur series of treatments. Indices were f i r s t determined using percent spectral reflectance at the red well (674 nm) as a measure of red reflectance, then recalculated using reflectance at the red rise (630 nm) . Plots of f o l i a r sulphur concentration versus values of each vegetation index were found to describe third order.polynomial relationships (Figure 30). Data transformations were unsuccessful in linearizing the relationship between f o l i a r sulphur and vegetation indices. Consequently, the nonlinear estimation module of SYSTAT (Wilkinson, 1988), a personal computer based s t a t i s t i c a l package, was used to estimate correlation coefficients (Table 41) and constants of polynomial equations. While the use of red rise rather than red well measurements in the calculation of vegetation indices did not result in any s t a t i s t i c a l l y significant changes in correlation coefficients (Table 42), some consistent trends were evident. Red rise correlation coefficients for vegetation indices 6 and 12 were greater than red well coefficients for both individual seedling and composite data sets, whereas red rise correlation coefficients for vegetation index 9 were lower for both data sets. 137 Figure 30. Relationship between f o l i a r sulphur concentration (S) and red well vegetation index 13 (RWVI13 = NIR/(green + red + NIR)) for the 1988 sulphur individual seedling data set. 138 Table 41 . Correlation between the : f o l i a r concentration of sulphur and vegetation indices calculated using percent spectral reflectance at the red well and red rise. Red well Red rise VI 1988 1988 1988 1988 comp ind comp ind 1 0.79* 0.53* 0.67ns 0.73** 2 0.78* 0.53* 0.68ns 0.66** 3 0.77* 0.53* 0.69ns 0.66** 4 0.71ns 0.32ns 0.71ns 0.49* 5 0.43ns 0.58** 0.69ns 0.43ns 6 0.72ns 0.26ns 0.74ns 0.40ns 7 0.71ns 0.37ns 0.71ns 0.46ns 8 0.62ns 0.65** 0.62ns 0.65** 9 0.70ns 0.51* 0.69ns 0.46ns 10 0.85* 0.46ns 0.81* 0.57** 11 0.14ns 0.47ns 0.77* 0.36ns 12 0.82* 0.55* 0.89** 0.57** 13 0.68ns 0.80** 0.68ns 0.79** 14 0.78* . 0.44ns 0.71ns 0.65** 15 0.61ns 0.61** 0.63ns 0.57** * - p<0 .05 r VI - vegetation indices ** - p<0 .01 1988 comp - 1988 composite samples ns - not significant 1988 ind - 1988 individual seedlings Note: There i s no exact s t a t i s t i c to compare correlation coefficients for testing the significance of non linear correlation. However, as a practical procedure tabulated r (positive square root of the coefficient of determination) values for (m-1) variables and (n-m+1) degrees of freedom, where m is the number of coefficients being estimated and n is the total number of observations, can be used as a measure of comparison (Draper and Smith, 1981). Degrees of freedom for the composite and individual seedling data sets were 7 and 23, respectively. 139 Table 42. Differences in correlation between f o l i a r sulphur concentration and recalculated vegetation indices using percent spectral reflectance at the red rise. Vegetation 1988 1988 indices composite individual samples seedlings 1 - + 2 - + 3 - + 4 0 + 5 + -6 + + 7 0 + 8 0 0 9 - -10 - + 11 ' + -12 + + 13 0 -14 - + 15 + • -+ - increase in correlation - - decrease in correlation 0 - no change in correlation Note: None of the changes were s t a t i s t i c a l l y significant at the 0.05 level. 140 5.3.8 Estimation of Foliar Sulphur For individual seedling data, red well vegetation index 13 was the index most highly correlated (r=0.80; p<0.01) with f o l i a r sulphur, while red rise vegetation index 12 (r=0.89; p<0.01) was most correlated for composite data. These measurements were more highly correlated with needle sulphur concentration than the green reflectance peak, red rise, red edge, or other vegetation indices; consequently, regression curves describing the relationship between fo l i a r sulphur and red well vegetation index 13, and red rise vegetation index 12 were used as models for estimating sulphur levels in Douglas-fir seedlings. Red well vegetation index 13 and red rise vegetation index 12 models were tested by comparing laboratory-determined and estimated f o l i a r sulphur concentrations of nitrogen and phosphorus treated seedlings. Large discrepancies existed between observed and estimated needle sulphur concentrations for both individual seedling and composite data sets (Tables 43 and 44). The standard error of prediction for nitrogen and phosphorus individual seedling data sets were 1.16 and 0.65 percent sulphur, respectively. The nitrogen composite data set had a standard error of prediction of 0.97, while that of the phosphorus composite data set was 0.24 percent sulphur. Both models frequently underestimated needle sulphur concentration to such an extent that negative estimates resulted. A l l treatments followed the same trend with respect to vegetation indices 12 and 13 and nutrient deficiency. Seedlings Table 43. Comparison of f o l i a r sulphur l e v e l s f o r 1988 i n d i v i d u a l seedling data as determined by chemical analysis and the model 3 2 %S = (2024.169 x RWVI13) - (4689.884 x RWVI13) + (3614.776 x RWVI13) - 926.453 Nitrogen treated seedlings Phosphorus treated seedlings RWVI13 S obs.(%) S est.(%) Abs. d i f f . RWVI13 S obs.(%) S est.(%) Abs. d i l 0.6584 0.30 -1.79 2.09 0.6770 0.37 -0.68 1.05 0.6827 0.22 -0.43 0.65 0.7396 0.48 0.54 0.06 0.6258 0.28 -4.92 5.20 0.7390 0.48 0.54 0.06 0.6863 0.31 -0.28 0.59 0.7870 0.76 0.27 0.49 0.6487 0.22 -2.54 2.76 0.7919 0.68 0.25 0.43 0.6443 0.34 -2.93 3.27 0.7814 0.47 0.31 0.16 0.6379 0.31 -3.56 3.87 0.7821 0.37 0.31 0.06 0.6389 0.29 -3.46 3.75 0.7650 0.30 0.43 0.13 0.6866 0.45 -0.28 0.73 0.7547 0.53 0.49 0.04 0.7420 0.34 0.54 0.20 0.7705 0.52 0.39 0.13 0.6952 0.33 0.01 0.32 0.7544 0.43 0.49 0.06 0.7617 0.58 0.45 0.13 0.7817 0.41 0.31 0.10 0.6887 0.40 -0.20 0.60 0.7381 0.47 0.54 0.07 0.7093 0.35 0.33 0.02 0.7853 0.65 0.29 0.36 0.7632 0.34 0.44 0.10 0.7379 0.44 0.54 0.10 0.7317 0.31 0.54 0.23 0.6082 0.34 -7.37 7.71 0.8014 0.42 0.21 0.21 0.6649 0.43 -1.35 1.78 0.8192 0.41 0.24 0.17 0.7308 0.38 0.53 0.15 0.8335 0.45 0.39 0.06 0.6954 0.23 0.01 0.22 0.7953 0.47 0.23 0.24 0.6903 0.25 -0.14 ' , 0.39 0.8070 0.45 0.21 0.24 0.7370 0.29 0.54 0.25 0.8320 0.34 0.37 0.03 0.7048 0.26 0.24 0.02 0.6742 0.26 -0.82 1.08 0.6799 0.21 -0.55 0.76 Standard error of p r e d i c t i o n 1.16 Standard e r r o r of p r e d i c t i o n 0.65 RWVI13 - red well vegetation index 13 S obs. - observed sulphur concentration S e s t . - estimated sulphur concentration Abs. d i f f . - absolute difference between observed and estimated concentrations Table 44. , Comparison of f o l i a r sulphur l e v e l s for 1988 composite seedling data as determined by chemical analysis and the model i %S = (8883.577 x RRVI12) - (4461.411 x RRVI12] | + (732. 371 x RRVI12) - 38.945 . Nitrogen treated seedlings Phosphorus treated seedlings RRVI12 S obs.(%) S est.(%) Abs. d i f f , RRVI12 S obs.(%) S est.(%) Abs. d i f f 0.1104 0.43 -0.51 0.94 0.1423 0.40 0.53 0. 13 0.1151 0.28 -0.21 0.49 0.1654 0.45 0.33 0. 12 0.0958 0.25 -1.92 2.17 0.1542 0.42 0.48 0. 06 0.1297 0.45 0.38 0.07 0.1796 0.43 0.15 0. 28 0.1453 0.44 0.53 0.09 0.1325 0.55 0.43 0. 12 0.1369 0.48 0.50 0.02 0.1193 0.41 0.01 0. 40 0.2434 0.44 3.11 2.67 0.1132 0.27 -0.32 0. 59 0.2312 0.35 1.69 1.34 0.1372 0.27 0.50 0. 23 Standard error of prediction 0.97 Standard > error of p r e d i c t i o n 0. 24 RRVI12 - red r i s e vegetation index 12 S obs. - observed Bulphur concentration S est. - estimated sulphur concentration Abs. d i f f . - absolute d i f f e r e n c e between observed and estimated concentrations 143 with very severe to severe nutrient deficiencies had low vegetation index values, due to low total chlorophyll levels, and high visible spectral reflectance. Possible minor or seedlings without nutrient deficiencies resulted in higher chlorophyll concentrations, less v i s i b l e reflectance, and higher values for vegetation indices 12 and 13. However, large errors in estimation occurred since the relationship between f o l i a r sulphur and vegetation indices 12 and 13 for sulphur treated seedlings was different than that for seedlings of the nitrogen and phosphorus series of treatments. For sulphur treated Douglas-fir seedlings, f o l i a r sulphur levels were low when vegetation index 13 values were also low (Figure 30). Needle sulphur concentration increased as values for vegetation index 13 increased, peaked at the 0.73 vegetation index value, and then decreased as the vegetation index increased. In contrast, f o l i a r sulphur levels of seedlings subjected to the nitrogen and phosphorus series of treatments increased over the whole range of red well vegetation index 13 values (Figures 31 and 32). Similar relationships were noted between f o l i a r sulphur and red rise vegetation index 12. Needle sulphur concentrations were low when values of vegetation indices 12 and 13 were high for the sulphur series of treatments since these f o l i a r samples were from seedlings treated with low concentrations (1 to 50 mg/L) of sulphur. Nitrogen and phosphorus treated seedlings were supplied with a constant concentration (69 mg/L) of sulphur. Foliar sulphur levels of these treatments increased with an increase in vegetation indices 12 and 144 0.65 0.7 0.75 0.8 RED WELL VEGETATION INDEX 13 0.85 Figure 31. Relationship between sulphur concentration and red well vegetation index 13 (NIR/ (green + red + NIR)) for the 1988 nitrogen individual seedling data set. 145 0.6 0.62 0.64 0.66 0.68 0.7 0.72 0.74 0.76 0.78 0.8 RED WELL VEGETATION INDEX 13 Figure 32. Relationship between f o l i a r sulphur concentration and red well vegetation index 13 (NIR/(green + red + NIR)) for the 1988 phosphorus individual seedling data set. 146 13 as concentrated (250 and 400 mg/L) nitrogen treatments and 5 to 50 mg/L phosphorus treatments stimulated seedling growth, resulting in greater u t i l i z a t i o n of available sulphur and high f o l i a r sulphur concentrations. Because the relationships between f o l i a r sulphur and vegetation indices 12 and 13 were not the same for a l l three series of treatments, the indices were considered useless in the estimation of f o l i a r sulphur. None of the other vegetation indices or spectral measurements (green reflectance peak, red rise and red edge) examined in the study demonstrated the same relationship with f o l i a r sulphur for a l l three series of treatments; consequently, none of the spectral parameters investigated in this study were deemed suitable for sulphur estimation. 147 5.4 CHLOROPHYLL 5.4.1 Green Reflectance Peak and Red Rise Percent spectral reflectance at the green reflectance peak (554 nm in 1987; 559 nm in 1988) and red rise (63 0 nm) decreased exponentially with increased total chlorophyll concentration (Figure 33) . These two relationships were observed for a l l data sets of the nitrogen, phosphorus, and sulphur treatments. As total chlorophyll increased, there was more absorption of visible light and less reflectance in the green and red spectral regions. Consequently, inverse relationships existed between chlorophyll level and spectral reflectance at the green reflectance peak and red rise. Thomas and Oerther (1972), Tsay et al. (1982), Nelson et al. (1986) and Cure (in press) noted similar relationships at the green reflectance peak, as did Tucker et al. (1975), Ajai et al. (1983B) and Everitt et a l . (1985) for spectral reflectance in the red spectral region. Linear relationships were obtained through logarithmic transformation of spectral reflectance and chlorophyll data, and correlation coefficients calculated describing the degree of association between transformed variables (Table 45) . A l l correlation coefficients determined for the relationship between the logarithm of total chlorophyll concentration and the logarithm of percent spectral reflectance at the green reflectance peak were s t a t i s t i c a l l y significant at the 0.05 probability level or better, as were a l l but one correlation coefficients calculated for the red 148 E > x o. O CC O _i X o 2.5 2H 1.5 H r— 0.5 \ •+-\ -+--1.2884 CHL = 32.7429(GRP) \ -+• •+• V. -f- \ + 10 15 20 25 S P E C T R A L R E F L E C T A N C E (%) AT 554 nm 30 Figure 33. Relationship between t o t a l chlorophyll concentration (CHL) and percent spectral reflectance at the green reflectance peak (GRP; 554 nm) for the 1988 nitrogen individual seedling data set. 149 Table 45. Correlation between total chlorophyll concentration and green reflectance peak, red rise, and red edge measurements. df GRP RR RE Nitrogen • 1987 composite samples 6 -0.74* (0.14) -0.64ns (0.16) 0.74* (0.33) 1988 composite samples 6 -0.95** (0.08) -0.94** (0.09) 0.85** (0.37) 1988 individual seedlings 22 -0.94** (0.09) -0.92** (0.10) 0.72** (0.31) Phosphorus 1987 composite samples 5 -0.95** (0.04) -0.88** (0.07) 0.66ns (0.29) 1988 composite samples 6 -0.89** (0.08) -0.81* (0.10) 0.31ns (0.35) 1988 individual seedlings 22 -0.77** (0.12) -0.82** (0.11) 0.71** (0.28) Sulphur 1988 composite samples 6 -0.86** (0.23) -0.82* (0.10) 0.77* (0.28) 1988 individual seedlings 22 -0.66** (0.24) -0.86** (0.07) 0.59** (0.26) * - p<0.05 GRP - green reflectance peak ** - p<0.01 RR - red rise ns - not significant RE - wavelength of the red edge df - degrees of freedom Numbers in parentheses represent the standard error of estimate. Green reflectance peak and red rise correlation coefficients were calculated for logarithmically transformed data. 150 rise. In most instances, spectral reflectance at the green reflectance peak was more highly correlated with total chlorophyll than reflectance at the red rise. 5.4.2 Red Edge The association between total chlorophyll and wavelength of the red edge was linear (Figure 34), and significantly correlated in a l l but two cases (Table 45) . As total chlorophyll concentration increased, more red light was absorbed, causing a shift of the red edge toward the red region of the spectrum. The observed relationship between chlorophyll and the wavelength of the red edge agrees with the findings of Horler et al. (1983), Curtiss and Ustin (1989), and Cure (in press). 5.4.3 Vegetation Indices Plots of total chlorophyll concentration versus values of vegetation indices calculated using percent spectral reflectance at the red well (674 nm) described exponential relationships. As values of red well vegetation indices 1, 2, 3, 4, 6, 8, 10, 12 and 13 increased, total chlorophyll concentration increased exponentially, while chlorophyll decreased exponentially with increased values of red well indices 5, 7, 9, 11, 14 and 15. 151 685 690 695 700 705 710 715 720 725 730 WAVELENGTH OF THE RED EDGE (nm) , Figure 34. Relationship between total chlorophyll (CHL) and the wavelength of the red edge (RE) for the 1988 nitrogen individual seedling data set. 152 Linear relationships were obtained through logarithmic transformation of chlorophyll and vegetation index data, and correlation coefficients calculated for transformed data (Table 46) . Correlation coefficients for red rise vegetation indices 8 and 15 were s t a t i s t i c a l l y significant for individual seedling and composite data sets of a l l treatments. Nitrogen data sets tended to have higher correlation coefficients than those of phosphorus and sulphur. Plots of total chlorophyll versus values of red rise vegetation indices also demonstrated exponential relationships. Correlation coefficients describing the degree of association between total chlorophyll and vegetation indices 1, 2, 3, 12 and 14 were consistently greater when indices were calculated with red rise rather than red well measurements (Table 47); however, these changes were not always s t a t i s t i c a l l y significant (Table 48). In general, higher correlation coefficients resulted when percent spectral reflectance at the red rise was used in the calculation of the other vegetation indices. 5.4.4 Estimation of Total Chlorophyll Of a l l correlations examined for individual seedling data sets, the green reflectance peak and red well vegetation index 15 for the nitrogen series of treatments were the most highly correlated with total chlorophyll. Each had a correlation coefficient of -0.94 (p<0.01) and 0.08 mg/g standard error of Table 46. C o r r e l a t i o n between t o t a l c h l o r o p h y l l concentration and vegetation indi c e s c a l c u l a t e d using percent spectral reflectance at the red w e l l . Nitrogen Phosphorus Sulphur VI 1987 1988 1988 1987 1988 1988 1988 1988 comp comp ind comp comp ind comp ind 1 0.37ns 0.78* 0.75** 0.70ns 0.73* 0.64** 0.71* 0.76** (0.19) (0.16) (0.17) (0.10) (0.12) (0.15) (0.12) (0.09) 2 0.38ns 0.78* 0.72** 0.33ns 0.75* 0.60** 0.65ns 0.73** (0.19) (0.16) (0.17) (0.12) (0.11) (0.15) (0.13) (0.10) 3 0.38ns 0.78* 0.72** 0.72ns 0.75* 0.61** 0.65ns 0.73** (0.19) (0.16) (0.17) (0.10) (0.11) (0.15) (0.13) (0.10) 4 0.78* 0.95** 0.81** -0.41ns -0.22ns -0.25ns -0.48ns -0.40ns (0.13) (0.08) (0.15) (0.13) (0.16) (0.19) (0.15) (0.13) 5 -0.73* -0.95** -0.80** 0.11ns 0.22ns 0.23ns 0.46ns 0.37ns (0.14) (0.08) (0.15) (0.14) (0.16) (0.19) (0.15) (0.13) 6 0.47ns 0.95** 0.79** -0.58ns -0.21ns -0.26ns -0.50ns -0.42* (0.18) (0.08) (0.15) (0.11) (0.16) (0.19) (0.15) (0.13) 7 -0.75* -0.95** -0.81** 0.51ns 0.24ns 0.34ns 0.48ns 0.45* (0.13) (0.08) (0.15) (0.12) (0.16) (0.18) (0.15) (0.13) 8 0.77* 0.93** 0.91** 0.91** 0.88** 0.77** 0.87** 0.64** (0.13) (0.09) (0.10) (0.06) (0.08) (0.12) (0.08) (0.1.1) 9 -0.70ns -0.95** -0.79** 0.28ns 0.25ns 0.32ns 0.45ns 0;41* (0.15) (0.08) (0.15) (0.14) (0.16) (0.18) (0.15) (0.13) 10 0.54ns 0.92** 0.66** -0.48ns -0.37ns -0.33ns -0.50ns -0.47 (0.17) (0.10) (0.19) (0.12) (0.16) (0.18) (0.15) (0.13) Continued Table 46. Continued. Nitrogen Phosphorus Sulphur VI 1987 1988 1988 1987 1988 1988 1988 1988 comp comp ind comp comp ind comp ind 11 -0.94** -0.94** -0.87** 0.30ns -0.20ns 0.02ns 0.37ns 0.13ns (0.07) (0.09) (0.12) (0.13) (0.16) (0.19) (0.16) (0.14) 12 0.61ns 0.87** 0.77** 0.20ns 0.57ns 0.44* 0.39ns 0.33ns (0.16) (0.13) (0.16) (0.14) (0.14) (0.17) (0.15) (0.13) 13 0.69ns 0.94** 0.91** 0.85* 0.88** 0.80** 0.81* 0.81** (0.15) (0.09) (0.10) (0.07) (0.08) (0.11) (0.10) (0.08) 14 -0.26ns -0.53ns -0.59** -0.66ns -0.67ns -0.58** -0.69ns -0.71** (0.20) (0.22) (0.20) (0.11) (0.12) (0.16) (0.12) (0.10) 15 -0.79* -0.95** -0.94** -0.90** -0.86** -0.71** -0.88** -0.55** (0.13) (0.08) (0.09) (0.06) (0.09) (0.13) (0.08) (0.12) df 6 6 22 5 6 22 6 22 * - p<0.05 VI - vegetation indices ** - p<0.01 1987 comp - 1987 composite samples ns - not s i g n i f i c a n t 1988 comp - 1988 composite samples df - degrees of freedom 1988 ind - 1988 i n d i v i d u a l seedlings Numbers i n parentheses represent the standard error of estimate. Cor r e l a t i o n c o e f f i c i e n t s were calculated f o r l o g a r i t h m i c a l l y transformed data. Table 47. Correlation between t o t a l chlorophyll and vegetation i n d i c e s c a l c u l a t e d using percent spectral r e f l e c t a n c e at the red r i s e . Nitrogen Phosphorus Sulphur VI 1987 1988 1988 1987 1988 1988 1988 1988 comp comp ind comp comp ind comp ind 1 0.65ns 0.93** 0.92** 0.86* 0.81* 0.82** 0.82* 0.85** (0.16) (0.09) (0.10) (0.07) (0.10) (0.11) (0.10) (0.08) 2 0.63ns 0.91** 0.89** 0.83* 0.82* 0.78** 0.77* 0.82** (0.16) (0.11) (0.11) (0.08) (0.10) (0.12) (0.11) (0.08) 3 0.63ns 0.92** 0.89** 0.84* 0.82* 0.79** 0.78* 0.82** (0.16) (0.10) (0.11) (0.08) (0.10) (0.12) (0.11) (0.08) 4 -0.23ns -0.17ns -0.54** -0.72 -0.65ns -0.70** -0.74* -0.74** (0.20) (0.26) (0.21) (Q.10) (0.13) (0.14) (0.11) (0.10) 5 0.03ns -0.07ns 0.49* 0.51ns 0.65ns 0.68** 0.75* 0.72** (0.21) (0.26) (0.22) (0.12) (0.13) (0.14) (0.11) (0.10) 6 -0.29ns -0.45ns -0.53** -0.78* -0.64ns -0.70** -0.74* -0.74** (0.20) (0.23) (0.21) (0.09) (0.13) (0.14) (0.11) (0.10) 7 0.18ns 0.19ns 0.50* 0.73ns 0.69ns 0.60** 0.67ns +0.55** (0.20) (0.26) (0.22) (0.10) (0.12) (0.15) (0.13) (0.11) 8 0.77* 0.93** 0.91** 0.91** 0.88** 0.77** 0.87** 0.64** (0.13) (0.09) (0.10) (0.06) (0.08) (0.12) (0.08) (0.11) 9 0.08ns 0.02ns 0.48* 0.68ns 0.70ns 0.59** 0.66ns +0.63** (0.20) (0.26) (0.22) (0.10) (0.12) (0.16) (0.13) (0.24) 10 -0.44ns -0.87** -0.79** -0.80* -0.70ns -0.79** -0.75* -0.77** (0.18) (0.13) (0.15) (0.08) (0.12) (0.12) (0.11) (0.09) Table 47. Continued. Nitrogen Phosphorus Sulphur VI 1987 1988 1988 . 1987 1988 1988 1988 1988 comp comp ind comp comp ind comp ind 11 -0.52ns -0.85** -0.59** 0.55ns 0.38ns 0.40ns 0.68ns 0.56** (0.18) (0.14) (0.20) (0.12) (0.16) (0.18) (0.12) (0.12) 12 0.68ns 0.90** 0.81** 0.35ns 0.63ns 0.51* 0.54ns 0.45* (0.15) (0.11) (0.15) (0.13) (0.13) (0.17) (0.14) (0.13) 13 0.72ns 0.94** 0.92** 0.89** 0.86** 0.82** 0.84** 0.83** (0.14) (0.09) (0.10) (0.06) (0.09) (0.11) (0.09) (0.08) 14 -0.62ns -0.93** -0.91** -0.85* -0.79* -0:82** -0.82* -0.84** (0.16) (0.09) (0.10) (0.08) (0.10) (0.11) (0.10) (0.08) 15 -0.79* -0.95** -0.93** -0.88** -0.86** -0.70** -0.88** -0.46* (0.13) (0.08) (0.09) (0.07) (0.09) (0.14) (0.08) (0.13) df 6 6 22 5 6 22 6 22 * - p<0.05 VI - vegetation indices ** - p<0.01 1987 comp - 1987 composite samples ns - not s i g n i f i c a n t 1988 comp - 1988 composite samples df - degrees of freedom 1988 ind - 1988 i n d i v i d u a l seedlings + - df = 20; two observations were deleted f o r each of vegetation i n d i c e s 7 and 9 as negative values were obtained thus preventing logarithmic transformation. Table 48. Differences i n c o r r e l a t i o n between t o t a l c h l o r o p h y l l and r e c a l c u l a t e d vegetation i n d i c e s using percent spectral r e f l e c t a n c e at the red r i s e . Nitrogen Phosphorus Sulphur VI 1987 comp 1988 comp 1988 ind 1987 comp 1988 comp 1988 ind 1988 ind 1988 ind 1 + + + * + + + + • + 2 + + + + + + + + 3 + + + + + + + + 4 - - * - + + + * + + 5 - — * - + + + + + 6 - — * - + + + + + 7 - - * + + + + 8 0 0 0 0 0 0 0 0 9 - - + + + + + 10 - - + + + + * + + 11 - - - * - + + + + + 12 + + + + + + + 13 + 0 + + - + + + 14 + + + * + + + + + 15 0 0 - - 0 - 0 -* VI p<0.05 vegetation indices 1987 comp 1988 comp 1988 ind - 1987 - 1988 - 1988 composite samples composite samples i n d i v i d u a l seedlings 158 estimate. Regression curves of both green reflectance peak and red well vegetation index 15 were used as models to estimate total chlorophyll levels of individual phosphorus and sulphur treated Douglas-fir seedlings (Tables 49 and 50). The model based on the green reflectance peak yielded slightly better chlorophyll estimates, with standard errors of prediction of 0.22 and 0.20 mg/g for the phosphorus and sulphur data sets, respectively. Standard errors for the total chlorophyll estimates of the red well vegetation index 15 model were 0.24 mg/g for the phosphorus data set and 0.25 for the sulphur data set. It was d i f f i c u l t to select a single spectral parameter on which to base a chlorophyll estimation model for composite f o l i a r samples. Among composite data, the green reflectance peak of the 1987 phosphorus data set had the highest correlation coefficient (r=-0.95; p<0.01), lowest standard error of estimate (0.04 mg/g), and less data than other composite data sets. Several spectral measurements (green reflectance peak, red well vegetation indices 4, 5, 6, 7, 9 and 15, and red rise vegetation index 15) of the 1988 nitrogen composite data set also had correlation coefficients of -0.95, but larger standard errors of estimate (0.08 mg/g). Consequently, models were developed using each of these parameters and tested on other composite data sets. Standard errors of prediction are summarized in Table 51. Table 49. Comparison of i n d i v i d u a l Douglas-fir t o t a l chlorophyll l e v e l s as determined by chemical a n a l y s i s and the model -1.2884 Chi = (32.7429)(GRP). Phosphorus treated seedlings Sulphur treated seedlings GRP Chi obs. (mg/g) Chi est. (mg/g) Abs. d i f f . GRP Chi obs. (mg/g) Chi est. (mg/g) Abs. d i f f . 19.14 1.05 0.73 0.32 12.00 1.34 1.33 0.01 5.41 1.24 0.97 0.27 13.14 1.52 1.19 0.33 17.33 1.14 0.83 0.31 11.64 1.63 1.39 0.24 11.15 1.39 1.46 0.07 10.53 1.16 1.58 0.42 12.83 1.58 1.22 0.36 11.59 1.37 1.39 0.02 12.91 1.24 1.21 0.03 14.72 1.14 1.02 0.12 14.61 1.24 1.03 0.21 14.22 1.32 1.07 0.25 13.73 1.06 1.12 0.06 9-.09 0.95 1.91 0.96 15.74 1.41 0.94 0.47 11.89 1.25 1.35 0.10 13.31 1.46 1.17 0.29 17.10 1.09 0.84 0.25 15.18 1.70 0.98 0.72 18.63 0.82 0.76 0.06 13.16 1.43 1.18 0.25 16.72 0.93 0.87 0.06 16.17 1.16 0.91 0.25 17.02 0.52 0.85 0.33 12.58 0.99 1.25 0.26 17.42 0.62 0.82 0.20 17.43 1.07 0.82 0.25 14.35 1.16 1.06 0.10 26.42 0.47 0.48 0.01 14.69 0.69 1.03 0.34 21.78 0.68 0.62 0.06 17.77 • 0.59 0.80 0.21 17.61 0.83 0.81 0.02 14.69 0.76 1.03 0.27 20.20 0.63 0.68 0.05 17.81 0.83 0.80 0.03 22.44 0.43 0.59 0.16 14.57 0.81 1.04 0.23 16.12 0.70 0.91 0.21 18.95 0.67 0.74 0.07 19.81 0.52 0.70 0.18 16.23 0.85 0.90 0.05 15.72 0.49 0.94 0.45 18.81 0.69 0.75 0.06 24.18 0.49 0.54 0.05 17.16 0.71 0.84 0.13 Standard error of p r e d i c t i o n 0.22 Standard e r r o r of p r e d i c t i o n 0.20 GRP - green reflectance peak; percent s p e c t r a l r e f l e c t a n c e measured at 554 nm Chi obs. - observed chlorophyll concentration Chi e s t . - estimated chlorophyll concentration Abs. d i f f . - absolute diff e r e n c e between observed and estimated concentrations Table 50. Comparison of i n d i v i d u a l Douglas-fir t o t a l chlorophyll l e v e l s as determined by chemical analysis and the model -1.6810 Chi = (0.0577)(RWVI15). Phosphorus treated seedlings Sulphur treated.seedlings RWVI15 Chi obs.(mg/g) Chi est.(mg/g) Abs. d i f f . RWVI15 Chi obs.(mg/g) Chi est.(mg/g) Abs. d i f f . 0.2436 1.05 0.62 0.43 0.2022 1.24 0.85 0.39 0.2091 1.14 0.80 0.34 0.1526 1.39 1.36 0.03 0.1639 1.58 1.21 0.37 0.1742 1.24 1.09 0.15 0.1749 1.24 1.08 0.16 0.1809 1.06 1.02 0.04 0.2011 1.41 0.86 0.55 0.1675 1.46 1.16 0.30 0.1854 1.70 0.98 0.72 0.1696 1.43 1.14 0.29 0.2031 1.16 0.84 0.32 0.1658 0.99 1.18 0.19 0.2161 1.07 0.76 0.31 0.2614 0.47 0.55 0.08 0.2265 0.68 0.70 0.02 0.2188 0.83 0.74 0.09 0.2316 0.63 0.67 0.04 0.2530 0.43 0.58 0.15 0.2037 0.70 0.84 0.14 0.2390 0.52 0.64 0.12 0.1802 0.49 1.03 0.54 0.2578 0.49 0.56 0.07 Standard err o r of p r e d i c t i o n 0.24 0.1672 1.34 1.17 0.17 0.1603 1.52 1.25 0.27 0.1559 1.63 1.31 0.32 0.1322 1.16 1.73 0.57 0.1413 1.37 1.55 0.18 0.1706 1.14 1.13 0.01 0.1727 1.32 1.10 0.22 0.1220 0.95 1.98 1.03 0.1498 1.25 1.40 0.15 0.2124 1.09 0.78 0.31 0.2271 0.82 0.70 0.12 0.1918 0.93 0.93 0.00 0.1850 0.52 0.98 0.46 0.1957 0.62 0.90 0.28 0.1812 1.16 1.02 0.14 0.1820 0.69 1.01 0.32 0.1993 0.59 0.87 0.28 0.1699 0.76 1.14 0.38 0.2125 0.83 0.78 0.05 0.1686 ' 0.81 1.15 0.34 0.2117 ' 0.67 0.78 0.11 0.1959 0.85 0.89 0.04 0.2178 0.69 0.75 0.06 0.1909 0.71 0.93 0.22 Standard e r r o r of p r e d i c t i o n 0.25 RWVI15 - red well vegetation index 15 Chi obs. - observed chl o r o p h y l l concentration Chi e s t . - estimated chl o r o p h y l l concentration Abs. d i f f . - absolute diff e r e n c e between observed and estimated concentrations Table 51. Models f o r estimating t o t a l c h l o r o p h y l l concentration of composite f o l i a r samples and associated standard e r r o r s of pr e d i c t i o n . Model and data set from Standard errors of p r e d i c t i o n (mg/g) for composite data sets which derived N 1987 N 1988 P 1987 P 1988 S 1988 Average 1987 P composite data -1.4748 Chl=(48.7199)(GRP) 0.22 0.97 NA 0.22 0.18 0.40 1988 N composite data -1.1363 Chl=(24.8200)(GRP) 1.7938 Chl=(9.2697)(RWVI4) 4 -1.7286 Chl=(1.0245 X 10 )(RWVI5) 1.8495 Chl=(0.0050)(RWVI6) -2.5516 Chl=(0.2021)(RWVI7) 3 -2.4187 Chl=(4.4589 X 10 )(RWVI9) -1.4765 Chl=(0.0946)(RWVI15) -1.6215 Chl=(0.0687)(RRVI15) 0.27 0.65 0.41 1.02 0.63 0.48 0.38 0.37 NA NA NA NA NA NA NA NA 0.22 0.92 0.57 1.37 1.44 0.81 0.33 0.32 0.20 0.34 0.33 0.35 0.33 0.33 0.19 0.20 0.27 0.70 0.64 0.76 0.88 0.77 0.29 0.29 0.24 0.65 0.49 0.88 0.82 0.60 0.30 0.30 1988 N i n d i v i d u a l seedling data -1.2884 Chl=(32.7429)(GRP) -1.6810 Chl=(0.0577)(RWVI15) 0.23 0.22 0.16 0.16 0.10 0.21 0.17 0.18 0.18 0.20 0.17 0.19 RWVI - red well vegetation index RRVI - red r i s e vegetation index NA - not a v a i l a b l e Chi - t o t a l c h l o r o p h y l l N - nitrogen treated seedlings P - phosphorus t r e a t e d seedlings S - sulphur treated seedlings H Ok H 162 Based on a comparison of average standard errors of prediction, the green reflectance peak model developed with 1988 nitrogen data provided the most accurate chlorophyll estimates of any model based oh composite data. Average standard errors varied from 0.24 mg/g for the 1988 nitrogen based model to 0.88 mg/g for the red well vegetation index 6 model. In an effort to obtain more accurate total chlorophyll estimates the green reflectance peak and red well vegetation index 15 models developed for individual seedling f o l i a r samples were applied to composite data sets (Table 51) . These predictive equations yielded better estimates; the average standard error of prediction was reduced to 0.17 and 0.19 mg/g for the green reflectance peak and red well vegetation index 15 models, respectively. Models developed from individual seedling data probably produced more accurate chlorophyll estimates since they were based on a far greater volume of data. Curran and Milton (1983) mention that chlorophyll measurements made spectrophotometrically have an accuracy of approximately +5 percent due to the small quantities involved and the need for several dilutions, blendings, and extractions. Most estimates of total chlorophyll, even those of the best models, had errors greater than +5 percent. Nevertheless current results are encouraging and demonstrate that spectral parameters which are highly correlated with total chlorophyll, such as the green reflectance peak and red well vegetation index 15, can provide useful indices of relative chlorophyll concentration. Undoubtedly, 163 further experimentation and incorporation of further data in model development w i l l result in better chlorophyll estimates. Correlation coefficients describing the association between total chlorophyll and a l l spectral parameters (red edge, red rise, green reflectance and vegetation indices) for the nitrogen series of treatments tended to be greater than those of the phosphorus and sulphur treatments. This may have been due to stress caused by concentrated (50 to 400 mg/L) phosphorus and sulphur f e r t i l i z a t i o n rates, as indicated by chlorotic and necrotic foliage, and resulting changes in plant pigment concentrations besides chlorophyll. Chlorophyll, chlorophyll decomposition products (e.g. phenophytins) and carotenoid concentrations may vary, and additional pigments (e.g. tannins), may build up within leaves in response to stress (Rock et al., 1986), thus confounding the relationship between chlorophyll and spectral reflectance, and influencing chlorophyll estimation. Curran et al. (1991) reported a near-linear relationship between red edge and total chlorophyll concentration for amaranth leaves with low concentration of the plant pigment amaranthin; however, the wavelength of the red edge occurred at longer wavelengths and was independent of chlorophyll concentration for leaves with high amaranthin concentration. Further study on the influence of nonchlorophyll leaf pigments on spectral reflectance i s required. 5.4.5 Total Chlorophyll 164 Although the ratio of chlorophyll a/chlorophyll b did not prove to be a useful index of nutrient stress, the potential of total chlorophyll as an indicator of nutrient stress became evident early in the study; moreover, spectral parameters highly correlated with total chlorophyll could be used to assess nutrient stress. Of a l l spectral parameters investigated, the red rise was best at discriminating the severity of nutrient deficiencies. Figures 35 to 37 show the relationships between total chlorophyll, spectral reflectance at the red rise, and severity of nutrient deficiencies for individual seedlings of the 1988 nitrogen, phosphorus and sulphur treatments. Consistent trends were observed for a l l three series of treatments. Seedlings without nutrient deficiencies or possible minor deficiencies had high levels of total chlorophyll, and low spectral reflectance at the red rise. As the severity of nutrient deficiency increased, total chlorophyll decreased, and the red rise increased. The same trends were also observed for a l l nitrogen, phosphorus and sulphur composite data sets. 165 5 ? CO d >• I 0. O cc o _l X o _l-o 6 8 10 12 14 16 SPECTRAL REFLECTANCE (%) AT 630 nm • — VERY SEVERELY NUTRIENT DEFICIENT © — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 35. Relationships between total chlorophyll, percent spectral reflectance at the red rise (630 nm) and severity of nutrient deficiency for the 1988 nitrogen individual seedling data set. 166 8 10 12 14 16 18 20 SPECTRAL REFLECTANCE (%) AT 630 nm 22 24 * —VERY SEVERELY NUTRIENT DEFICIENT © — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + _ POSSIBLE MINOR NUTRIENT DEFICIENCY • — NOT NUTRIENT DEFICIENT Figure 36. Relationships between total chlorophyll, percent spectral reflectance at the red rise (630 nm) and severity of nutrient deficiency for the 1988 phosphorus individual seedling data set. 167 8 10 12 14 16 18 SPECTRAL REFLECTANCE (%) AT 630 nm © — MODERATELY TO SEVERELY NUTRIENT DEFICIENT + — POSSIBLE MINOR NUTRIENT DEFICIENCY • —NOT NUTRIENT DEFICIENT Figure 37. Relationships between total chlorophyll, percent spectral reflectance at the red rise (630 nm) and severity of nutrient deficiency for the 1988 sulphur individual seedling data set. 168 6. SUMMARY OF RESULTS, CONCLUSIONS AND RECOMMENDATIONS 6.1 TESTING OF HYPOTHESES This thesis explored the use of narrow-band spectral reflectance measurements in the 400 to 1100 nm range of the electromagnetic spectrum to detect nutrient deficiencies, assess their severity, and estimate the f o l i a r concentrations of nitrogen, phosphorus, sulphur and total chlorophyll in Douglas-fir seedlings. Several nutrient deficiencies, including nitrogen, phosphorus, calcium, potassium and zinc, were induced. Decreases in total chlorophyll were associated with a l l deficiencies, thus demonstrating the value of total chlorophyll as an index of nutrient stress. There were no unique changes in visible or near-infrared spectral reflectance which could be attributed to a specific nutrient deficiency; rather, changes in reflectance were non-specific responses influenced by how varying nutrient levels affected chlorophyll concentration. A l l observed changes in spectral reflectance occurred in the visible region; no relationships were observed between near-infrared reflectance and fo l i a r nutrient levels associated with the nitrogen, phosphorus and sulphur series of treatments. Foliar nitrogen was most correlated with total chlorophyll, followed by phosphorus, and then sulphur. Spectral reflectance measurements most correlated with f o l i a r nitrogen, phosphorus and sulphur were also highly correlated with total chlorophyll. As fo l i a r nitrogen and phosphorus increased total chlorophyll 169 increased in a linear manner, as did needle sulphur and total chlorophyll concentrations of Douglas-fir seedlings subjected to the nitrogen and phosphorus series of treatments. However, chlorophyll decreased linearly with increased f o l i a r sulphur for seedlings subjected to the sulphur series of treatments. This discrepancy was the result of differences in response to f e r t i l i z e r treatments. Increasingly severe calcium, nitrogen and other possible nutrient deficiencies caused reductions in chlorophyll levels over the range of sulphur treatments. In contrast, concentrated nitrogen and 5 to 50 mg/L phosphorus treatments stimulated seedling growth, resulting in greater utilization'of available sulphur, and high f o l i a r sulphur and total chlorophyll concentrations. An important outcome was that none of the vegetation indices or other spectral measurements followed the same relationship with f o l i a r sulphur for the sulphur, nitrogen and phosphorus series of treatments; thus none of the models developed were deemed suitable for sulphur estimation. Future research should investigate the relationship between needle sulphur concentration and total chlorophyll of Douglas-fir growing over a wide range of conditions. Sulphur estimation may require the development of two different models, one for trees growing in areas of low to average sulphur availability, and another for those exposed to high sulphur concentrations. 170 Based on these results the hypotheses postulated at the start of the investigation were tested. One: That percent spectral reflectance at the wavelengths of the green reflectance peak and red rise can be used to detect nutrient (nitrogen, phosphorus and sulphur) deficiencies in Douglas-fir seedlings. While specific nutrient deficiencies could not be determined, percent spectral reflectance at the green reflectance peak and red rise proved useful indicators of nutrient stress. A l l nutrient deficiencies caused increases in reflectance at the green reflectance peak and red rise in proportion to the severity of deficiency. Therefore, by measuring spectral reflectance at these wavelengths, having a p r i o r i knowledge that seedlings are not subjected to other forms of stress, i t i s possible to ascertain that Douglas-fir seedlings are nutrient deficient and the severity of deficiency. Two: That spectral reflectance at the green reflectance peak and red rise are significantly correlated with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll levels in Douglas-fir seedlings, and can thus be used to estimate these f o l i a r constituents. 171 Percent spectral reflectance at the green reflectance peak and red rise were significantly and highly correlated with total chlorophyll and f o l i a r nitrogen; consequently, these measurements are useful in following changes in needle nitrogen and total chlorophyll, and show potential in the development of chlorophyll and nitrogen estimation models for Douglas-fir seedlings. Development of models based on a greater volume of data w i l l likely result in reliable estimates. Foliar phosphorus and sulphur were much less correlated with the green reflectance peak and red rise. Correlation coefficients describing these relationships were significant in only 50 percent of the cases tested. Because there were other spectral parameters more highly correlated with phosphorus and sulphur, neither the green reflectance peak nor red rise should be used in the estimation of these nutrients. Three: That nutrient (nitrogen, phosphorus and sulphur) deficiencies in Douglas-fir seedlings can be detected by "shifts" in the wavelength of the red edge. For a l l nutrient deficiencies, the wavelength of the red edge decreased or shifted toward the blue end of the spectrum in relation to the severity of deficiency; thus the red edge can be used to detect nutrient deficiencies and assess their severity. 172 Four: That the wavelength of the red edge is significantly correlated with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll levels in Douglas-fir seedlings, and can thus be used to estimate the concentrations of these f o l i a r constituents. The degree of correlation between the wavelength of the red edge and total chlorophyll and f o l i a r nitrogen was high and s t a t i s t i c a l l y significant, thus demonstrating the usefulness of the red edge in following changes in nitrogen and chlorophyll concentrations, and i t s potential in nitrogen and chlorophyll estimation. The red edge does not appear a useful estimator of f o l i a r phosphorus and sulphur. Correlation coefficients calculated for the relationship between needle phosphorus and the wavelength of the red edge were low and significant for only one third of data sets examined, while the red edge was not s t a t i s t i c a l l y correlated with sulphur in any of the cases tested. Five: That changes in the values of vegetation indices 1 to 15 can be used to detect nutrient (nitrogen, phosphorus and sulphur) deficiencies in Douglas-fir seedlings. 173 Values of vegetation indices increased or decreased, depending on the particular index, in proportion to the severity of nutrient deficiency. This demonstrates the value of vegetation indices as indicators of nutrient stress; naturally, those indices most correlated with total chlorophyll, such as red well and red rise vegetation index 15, are better stress indicators. Six: That the vegetation indices 1 to 15 are significantly correlated with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll levels in Douglas-fir seedlings, and can thus be used to estimate the concentrations of these f o l i a r constituents. Not a l l vegetation indices listed in Table 1 were significantly correlated with total chlorophyll or each of the f o l i a r nutrients examined; consequently, not a l l vegetation indices are useful for following changes in nutrient levels, or estimating needle nitrogen, phosphorus, sulphur, and total chlorophyll. It was d i f f i c u l t to rank vegetation indices according to their degree of correlation with nitrogen, phosphorus, sulphur and chlorophyll since correlation varied with the measure of red spectral reflectance used in index calculation. Vegetation indices with the highest correlation coefficients for individual seedling data sets were found best for estimating f o l i a r nutrient concentrations; red well or red rise vegetation index 15 for f o l i a r nitrogen 174 estimation, red rise vegetation index 10 for phosphorus, and red rise vegetation index 15 for estimating total chlorophyll. Red well vegetation index 13 was most correlated with f o l i a r sulphur. While models developed using vegetation indices did not provide reliable estimates, results are encouraging. The refinement of models based on a greater volume of data w i l l likely result in more accurate estimates. Models developed from individual seedling data had three times the data of composite data sets, and consistently produced better estimates. Seven: That the use of red rise rather than red well measurements in the calculation of vegetation indices requiring a red reflectance measure in their determination w i l l result in indices more strongly correlated with nitrogen, phosphorus, sulphur and total chlorophyll, and thus provide more accurate estimates of these f o l i a r constituents. While the use of red rise measurements in the calculation of vegetation indices did not always result in higher correlation with f o l i a r nitrogen, phosphorus, sulphur and total chlorophyll, some consistencies were evident. Use of spectral reflectance at the red rise in the calculation of vegetation indices 1, 2, 3, 12 and 14 consistently resulted in higher correlation with f o l i a r nitrogen and total chlorophyll, and with f o l i a r phosphorus and sulphur in the majority of cases tested. Therefore, red rise measurements 175 should be used in the determination of these vegetation indices when investigating vegetation stress or estimating needle nitrogen, phosphorus, sulphur, and total chlorophyll. This is considered an important finding since vegetation indices 1 and 2 are the most frequently used indices in remote sensing, and usually calculated using reflectance in the red well region. For other vegetation indices the influence of the red rise varied with each of the nitrogen, phosphorus, sulphur, and chlorophyll f o l i a r constituents. Correlation between fo l i a r nitrogen and vegetation index 10 increased for a l l cases tested when the red rise was used in index calculation, whereas correlation coefficients for vegetation index 11 always decreased, and those for vegetation index 15 remained unchanged. Correlation coefficients for the remaining vegetation indices tended to decrease when percent spectral reflectance at the red rise was used in index calculation rather than reflectance at the red well. Correlation between total chlorophyll and vegetation indices 4, 5, 6, 7, 9, 10, 11 and 13 determined using red rise measurements increased for the vast majority of test cases, thus indicating that the red rise should be used in total chlorophyll estimation when using these indices. Correlation coefficients for vegetation indices 1 to 7 and 9 to 14 were greater when calculated using red rise rather than red well spectral measurements; therefore, reflectance at the red rise should be used for index calculation when estimating f o l i a r phosphorus with these vegetation indices. 176 Vegetation indices 6 and 12 were more highly correlated with f o l i a r sulphur for a l l test cases when indices were determined with red rise instead of red well measurements; in contrast, correlation coefficients for vegetation index 9 consistently decreased with use of red rise measurements. Therefore, spectral reflectance at the red rise i s recommended for vegetation index calculation when investigating the association between f o l i a r sulphur and vegetation indices 6 and 12, but not when using index 9. When red rise measurements were used, correlation coefficients for each of the remaining vegetation indices increased in f i f t y percent of test cases, and either decreased or remained unchanged in the remaining test cases; thus indicating the need for further research to elucidate the relationship between red rise vegetation indices and needle sulphur concentration. Eight: That the chlorophyll a/chlorophyll b ratio can be used as an index of nutrient (nitrogen, phosphorus and sulphur) stress in Douglas-fir seedlings. As there was no clear and consistent relationship between the ratio of chlorophyll a to b and the severity of nutrient deficiencies for nitrogen, phosphorus and sulphur treated seedlings i t appears ineffectual as an index of nutrient stress. 177 6.2 NEED FOR NARROW BAND SENSORS The present study demonstrated the complex interrelationships between nutrient stress, changes in Douglas-fir physiology and the resulting effects on spectral reflectance, and thus the need for greater co-operation between plant physiologists and remote sensing specialists. To interpret vegetation stress from remote sensing data, Murtha (1982) identified four associated subject areas which investigators should be aware of, namely: a) the possible environmental stress capable of inducing injurious strain; b) the possible syndromes indicative of injurious strain, or the manifestation of damage; c) the effect of the strain on the normal spectral reflectance pattern; and d) the resulting effects of spectral reflectance changes on the imagery or aerial photographs. A thorough understanding of each of these subject areas could best be achieved by a group of plant physiologists and remote sensing scientists. The need for basic spectral reflectance studies was also demonstrated for the identification of spectral regions most affected by the strain of nutrient deficiency or other forms of vegetation stress. Such information i s imperative for sensor design and selection as detection of stress is highly depended on the spectral and spatial sensitivity of the instruments. The designers of MEIS (Multi-detector Electro-optical Imaging Scanner), a recently developed airborne sensor equipped with a special series of f i l t e r s for detecting vegetation stress, appear 178 to have overlooked the need for basic spectral reflectance studies. None of the MEIS bands record spectral reflectance at the 630 nm red rise (Table 52) , a spectral region which the current study proved important in the detection of nutrient stress. Furthermore, MEIS bandwidths range from 11.7 to 35.3 nm, making them too coarse for determining the wavelength of the red edge or measuring spectral reflectance at c r i t i c a l wavelengths. While the centre of one band approximates the position of the green reflectance peak and another the red well, the measurements obtained are averaged over entire bandwidths, thereby hindering the discrimination of differences in spectral reflectance and consequently vegetation stress. Instruments capable of simultaneously measuring spectral reflectance in narrow, preferably less than 3 nm, contiguous, spectral bands over the visible and near-infrared regions are required. While the Spectron SE590 and Li-Cor 1800 spectroradiometers used in the current study meet these specifications, they are designed for ground use. Airborne sensors, such as CASI (Compact Airborne Spectrographic Imager), are required for detection of nutrient stress over forest stands. CASI is a high resolution imaging spectrometer capable of obtaining data over the 425 to 950 spectral region at 1.8 nm bandwidths (Borstad et al., 1989). With a 1.2 milliradian resolution CASI provides 2.5 meter size pixels at approximately 1800 meter altitude, and is thus able to image individual tree crowns. 179 Table 52. Spectral bands of the MEIS vegetation stress f i l t e r set. Centre wavelength (nm) Mean bandwidth (nm) Band limits Minimum (nm) Maximum 548.4 31.9 532.45 564.35 596.4 35.3 578.75 614.05 675.0 20.8 664.60 685.40 698.3 13.1 691.75 704.85 711.2 15.7 703.35 719.05 721.4 11.7 715.55 727.25 734.4 17.0 742.90 759.90 747.8 16.7 739.45 756.15 780.2 33.3 763.55 796.85 Data from Innotech Aviation Limited (pers. com.). The present study demonstrated the potential use of leaf spectral reflectance measurements for the detection of nutrient stress, to follow changes in ,nutrient and chlorophyll levels, and estimate needle nitrogen, phosphorus, and total chlorophyll concentrations in Douglas-fir seedlings. This may lead to the development of rapid, non-destructive, ground-based f o l i a r nitrogen, phosphorus and total chlorophyll estimation techniques useful in nursery and f i e l d situations. Development of such methods would be particularly advantageous over traditional laboratory techniques which require several hours and are particularly complex, thereby increasing the chance of experimental error. While significant correlations were noted for the present study between spectral response characteristics and fo l i a r nitrogen, phosphorus, sulphur and total chlorophyll at the 180 individual leaf level, that does not guarantee that the spectral properties of a forest canopy can also be related to nutrient status. Radiance recorded by an airborne remote sensing system is a composite signal from leaves, branches and boles, tree canopies, ground cover and soils (Colwell, 1974; Goel, 1988; Koch et al., 1990; Spanner et al., 1990), and influenced by such factors as shadowing due to height differences and defoliation, sun-target-sensor geometry, topography, leaf orientation, canopy architecture, and atmospheric turbidity (Jackson et a l . , 1983; Holben et al., 1986; Goel, 1988; Brakke and Otterman, 1990). Methods may have to be required to account for these added influences. Huete (1988) and Major et al. (1990) present transformation techniques to minimize background soil-brightness influences, and reduce the effects of solar- and leaf-angle effects. The red edge may provide an alternate approach. While the wavelength of the red edge was not the spectral parameter most correlated with chlorophyll or f o l i a r nutrient concentrations, i t was highly correlated with total chlorophyll and needle nitrogen concentration. Consequently, use of the red edge in nutrient and other vegetation stress studies may be particularly advantageous since i t appears insensitive to variation in vegetation ground area coverage (Collins et al., 1983; Horler et al., 1983). Remotely sensed data from vegetation often require knowledge of the degree of ground cover for their interpretation since the assessment of stress conditions can be confounded with ground cover variation. It is important to distinguish low ground cover from unhealthy 181 foliage. In an investigation of the influence of ground cover variation on red edge measurements, Horler et al. (1983) reported that the red edge remained unchanged with simulated ground cover ranging from 20 to 100 percent. Investigations by Demetriades-Shah et al. (1990) also suggest that c r i t i c a l wavelengths along the first-order derivative, such as the red edge, or wavelengths along higher-order derivatives may eliminate the effects of s o i l background reflectance and canopy architecture. One avenue of future research should concentrate on canopy spectral reflectance to determine i f the added complexities associated with canopies preclude detection of nutrient stress, or nitrogen, phosphorus, and total chlorophyll estimation. This could li k e l y be accomplished through the analysis of canopy spectral reflectance measurements of Douglas-fir growing under a variety of nutrient conditions, measured with a high resolution airborne sensor such as CASI. Such research may contribute toward remote detection of nutrient stress in forest stands, refinement of models for f o l i a r nutrient estimation, and aid in the identification of forest stands which could benefit from forest f e r t i l i z a t i o n projects. Furthermore, estimation of leaf nitrogen and other chemicals via remote sensing could be helpful in describing ecosystems over large areas. Research in forest ecosystems indicate that canopy chemistry, especially total nitrogen concentration, may be important in describing and modelling productivity and nutrient dynamics (Vitousek, 1982; Pastor et al., 1984; Van Cleve et al., 1983). 182 While this study laid the ground work for the remote detection of nutrient stress, and f o l i a r estimation of nitrogen, phosphorus and total chlorophyll i t must be emphasized that the results are specific to Douglas-fir seedlings and spectral reflectance measurements of individual needles made under controlled laboratory conditions. Further research is required before the techniques developed can be applied to operational forest f e r t i l i z a t i o n programs. 183 LITERATURE CITED Ahem, F.J. 1988. The effects of bark beetle stress on the f o l i a r spectral reflectance of lodgepole pine. International Journal of Remote Sensing 9: 1451-1468. Ajai, M.N. Sashikumar, D.S. Kamat, G.S. Chaturvedi and S.K. Sinha. 1983A. Remote detection of nutrient deficiencies in wheat crops. Pages A-3-1 to A-3-9 in Proceedings of the Fourth Asian Conference on Remote Sensing, November 10-15, Colombo, Sri Lanka. Ajai, D.S. Kamat, G.S. Chaturvedi, A.K. Singh and S.K. Sinha. 1983B. Spectral assessment of leaf area index, chlorophyll content, and biomass of chickpea. Photogrammetric Engineering and Remote Sensing 49: 1721-1727. Al-Abbas, A.H., R. Barr, J.D. Hall, F.L. Crane and M.F. Baumgardner. 1974. Spectra of normal and nutrient-deficient maize leaves. Agronomy Journal 66: 16-20. Alberte, R.S., J.P. Thornber and E.L. Fiscus. 1977. Water stress effects on the content and organization of chlorophyll in mesophyll and bundle sheath chloroplasts of maize. Plant Physiology 59: 351-353. American Public Health Association. 1980. Standard methods for the examination of. water and wastewater. 15th edition. American Public Health Association, Washington, D.C. American Public Health Association. 1985. Standard methods for the examination of water and wastewater. 16th edition. American Public Health Association, Washington, D.C. Ballard, T.M. 1985. Foliar analysis for deficiency diagnosis and f e r t i l i z e r prescription in forestry: varia b i l i t y considerations. Pages 142-153 in H.A. Luttmerding, L.E. Lowe and T.M. Ballard (Editors) . The role of s o i l analysis in resource management. B.C. Ministry of Environment, Victoria, B.C., Technical Report 16. Ballard, T.M. and R.E. Carter. 1986. Evaluating forest stand nutrient status. B.C. Ministry of Forests, Victoria, B.C., Land Management Report Number 20. 60 pages. Ballard, T.M. and N. Majid. 1985. Use of pre-treatment increment data in evaluating tree growth response to f e r t i l i z a t i o n . Canadian Journal of Forest Research 15: 18-22. 184 Barber, J. and D.N.H. Horler. 1981. Fundamental relationships between plant spectra and geobotanical stress phenomena: fi n a l report. Final report on N.A.S.A. Contract NAS 5-25738. Department of Pure and Applied Biology, Imperial College of Science and Technology, London, U.K. 63 pages. Barrows, H.L. 1959. Evaluating the micronutrient requirements of trees. Pages 18-31 in Mineral nutrition of trees: a symposium. Duke University School of Forestry. Bulletin 15. Beaufils, E.R. 1973. Diagnosis and recommendation integrated system (DRIS). Soil Science Bulletin No. 1, University of Natal, Pietermaritzburg, South Africa. 132 pages. Benedict, H.M. and R. Swidler. 1961. Nondestructive method for estimating chlorophyll content of leaves. Science 133: 2015-2016. Bidwell, R.G.S. 1979. Plant physiology. Macmillan Publishing Company, New York. 726 pages. Boawn, L.C. and J.C. Brown. 1968. Further evidence for a P-Zn imbalance in plants. Soil Science Society of America Journal 32: 94-97. Borstad, G.A., D.A. H i l l and R.C. Kerr. 1989. Use of the Compact Airborne Spectrographic Imager (CASI): laboratory examples. Pages 2081-2084 in Proceedings of the 1989 International Geoscience and Remote Sensing Symposium and 12th Canadian Symposium on Remote Sensing, Vancouver, B.C. Brakke, T.W. and J. Otterman. 1990. Canopy bidirectional reflectance dependence on leaf orientation. International Journal of Remote Sensing 11: 1023-1032. British Columbia Department of Environment. 1976. A laboratory manual for the chemical analysis of waters, wastewaters, sediments and biological tissues. 2nd edition. Environmental Laboratory, Water Resources Service, British Columbia Department of Environment, Vancouver, B.C. Burstrom, H.G. 1968. Calcium and plant growth. Biological Reviews of the Cambridge Philosophical Society 43: 287-316. Card, D.H., D.L. Peterson, P.A. Matson and J.D. Aber. 1988. Prediction of leaf chemistry by the use of visible and near infrared reflectance spectroscopy. Remote Sensing of Environment 26: 123-147. 185 Carter, R.E. and K. Klinka. 1986. Symptoms and causes of distorted growth in immature forest stands in coastal British Columbia. B.C. Ministry of Forests, Victoria, B.C., Land Management Report Number 39. 61 pages. Carter, R.E., A.M. Scagel and K. Klinka. 1986. Nutritional aspects of distorted growth in immature forest stands of southwestern British Columbia. Canadian Journal of Forest Research 16: 36-41. Chang, S.H. and W. Collins. 1983. Confirmation of airborne biogeophysical mineral exploration techniques using laboratory methods. Economic Geology 78: 723-736. Chaudhry, F.M. and J.F. Loneragan. 1972. Zinc absorption by wheat seedlings: I. Inhibition by macronutrient ions in short-term studies and i t s relevance to long-term zinc nutrition. Soil Science Society of America Journal 36: 323-327. Collins, W. 1978. Remote sensing of crop type and maturity. Photogrammetric Engineering and Remote Sensing 44: 43-55. Collins, W., S.H. Chang, G. Raines, F. Canney and R. Ashley. 1983. Airborne biogeophysical mapping of hidden mineral deposits. Economic Geology 78: 737-749. Collins, W. and H. Chiu. 1979. Signature evaluation of natural targets using high spectral resolution techniques. Pages 567-582 in Proceedings of the 13th International Symposium on Remote Sensing of the Environment, April 23-27, Environmental Research Institute of Michigan, Ann Arbor, Michigan. Colwell, J.E. 1974. Vegetation canopy reflectance. Remote Sensing of Environment 3: 175-183. Cure, W.W. In press. Assaying the chlorophyll content of intact loblolly pine needles with reflectance and transmittance measurements. Remote Sensing of Environment. Curran, P.J., J.L. Dungan, B.A. Macler and S.E. Plummer. 1991. The effect of a red leaf pigment on the relationship between red edge and chlorophyll concentration. Remote Sensing of Environment 35: 69-76. Curran, P.J. and E.J. Milton. 1983. The relationship between the chlorophyll concentration, LAI and reflectance of a simple vegetation canopy. International Journal of Remote Sensing 4: 247-255. 186 Curtiss, B. and S.L. Ustin. 1989. Parameters affecting reflectance of coniferous forests in the region of chlorophyll pigment absorption. Pages 2633-2636 in Proceedings of the 1989 International Geoscience and Remote Sensing Symposium and 12th Canadian Symposium on Remote Sensing, Vancouver, B.C. Curtiss, B. and S.L. Ustin. In press. The remote detection of early stages of air pollution injury in coniferous forests using imaging spectrometry. International Journal of Remote Sensing. Das, D.K., G. Singh and A.K. Sutradhar. 1985. Remote sensing of wheat grown under differential irrigation, row spacing and nitrogen levels. Pages 400-405 in Proceedings of the Sixth Asian Conference on Remote Sensing, November 21-26, Hyderabad, India. Demetriades-Shah, T.H. and M.N. Court. 1987. Oblique view reflectance for assessing nitrogen status of incomplete canopies. International Journal of Remote Sensing 8: 1049-1055. Demetriades-Shah, T.H., M.D. Steven and J.A. Clark. 1990. High resolution derivative spectra in remote sensing. Remote Sensing of Environment 33: 55-64. Draper, N.R. and H. Smith. 1981. Applied regression analysis. John Wiley and Sons Incorporated, New York. 709 pages. Everard, J. 1973. Foliar analysis; sampling methods, interpretation and application of the results. Quarterly Journal of Forestry 67: 51-66. Everitt, J.H., A.J. Richardson and H.W. Gausman. 1985. Leaf reflectance-nitrogen-chlorophyll relations in buffelgrass. Photogrammetric Engineering and Remote Sensing 51: 463-466. Fox, L. 1978. Previsual detection: the elusive dream. Pages 53-64 in On remote sensing for vegetation damage assessment. Symposium held in Seattle, Washington, by American Society for Photogrammetry. Freedman, D. , R. Pisani and R. Purves. 1978. Statistics. W.W. Norton and Company, New York. 506 pages. Gaines, T.P. and G.A. Mitchell. 1979. Boron determination in plant tissue by the azomethine method. Communications in Soil Science and Plant Analysis 10: 1099-1108. Gates, D.M. 1970. Physical and physiological properties of plants. Pages 224-252 in Remote sensing with special reference to agriculture and forestry. National Academy of Sciences, Washington, D.C. 424 pages. 187 Gates, D.M., H.J. Keegan, J.C. Schleter and V.R. Weidner. 1965. Spectral properties of plants. Applied Optics 4: 11-20. Gausman, H.W. 1974. Leaf reflectance of near-infrared. Photogrammetric Engineering 40: 183-191. Gausman, H.W. 1977. Reflectance of leaf components. Remote Sensing of Environment 6: 1-9. Gausman, H.W., W.A. Allen and R. Cardenas. 1969. Reflectance of cotton leaves and their signature. Remote Sensing of Environment 1: 19-22. Gausman, H.W., D.E. Escobar and R.R. Rodriquez. 1973. Discriminating among plant nutrient deficiencies with reflectance measurements. Pages 13-27 in Proceedings of the Fourth Biennial Workshop on Aerial Color Photography in the Plant Sciences, July 10-12, University of Maine, Orono, Maine. Gessel, S.P., K.J. Turnbull and F.T. Tremblay. 1960. How to f e r t i l i z e trees and measure response. National Plant Food Institute, Washington, D.C. 67 pages. Goddard, R.E. and CA. Hollis. 1984. The genetic basis of forest tree nutrition. Pages 237-258 in G.D. Bowen and E.K.S. Nambiar (Editors). Nutrition of plantation forests. Academic Press, Toronto. 516 pages. Goel, N.S. 1988. Models of vegetation canopy reflectance and their use in estimation of biophysical parameters from„reflectance data. Remote Sensing Reviews 4: 1-221. Goetz, A.F.H., B.N. Rock and L.C. Rowan. 1983. Remote sensing for exploration: an overview. Economic Geology 78: 573-590. Hacskaylo, J. 1960. Deficiency symptoms in forest trees. Transactions of the Seventh International Congress of Soil Science 3: 393-405. Hacskaylo, J., R.F. Finn and J.P. Vimmerstedt. 1969. Deficiency symptoms of some forest trees. Ohio Agricultural Research Division Central Research Bulletin 1015. Ohio Agricultural Research and Development Center, Wooster, Ohio. 68 pages. Heilman, P. 1971. Effects of f e r t i l i z a t i o n on Douglas-fir in southwestern Washington. Washington Agricultural Experimental Station, College of Agriculture, Washington State University, Circular 535. 23 pages. Heilman, P. and G. Ekuan. 1973. Response of Douglas-fir and western hemlock seedlings to lime. Forest Science 19: 220-224. 188 Heller, R.C. 1968. Previsual detection of ponderosa pine dying from bark beetle attack. Pages 387-434 in Proceedings of the Fifth International Symposium on Remote Sensing of Environment, April 16-18, Environmental Research Institute of Michigan, Ann Arbor, Michigan. Hinzman, L.D., M.E. Bauer and C.S.T. Daughtry. 1984. Growth and reflectance characteristics of winter wheat canopies. Laboratory for Applications of Remote Sensing, West Lafayette, Indiana. LARS Technical report 111484, NASA-Johnson Space Center Contract NAS9-16528. 18 pages. Hocking, D. 1971. Preparation and use of a nutrient solution for culturing seedlings of lodgepole pine and white spruce, with selected bibliography. Northern Forest Research Centre, Canadian Forestry Service, Department of Environment, Edmonton, Alberta, Information Report NOR-X-1. 14 pages. Holben, B., D.S. Kimes and R.S. Fraser. 1986. Directional reflectance response of AVHRR red and near IR bands for three cover types and varying atmospheric conditions. Remote Sensing of Environment 19: 213-236. Hoober, J.K. 1984. Cellular organelles: chloroplasts. Plenum Press, New York. 280 pages. Horler, D.N.H., J. Barber and A.R. Barringer. 1980. Effects of heavy metals oh the absorbance and reflectance spectra of plants. International Journal of Remote Sensing 1: 121-136. Horler, D.N.H. , M. Dockray and J. Barber. 1983. The red edge of plant leaf reflectance. International Journal of Remote Sensing 4: 273-288. Howard, J.A., R.D. Watson and T.D. Hessin. 1971. Spectral reflectance properties of Pinus ponderosa in relation to the copper content of the s o i l , Malachit mine, Colorado. Pages 285-296 in Proceedings of the 7th Symposium on Remote Sensing of the Environment, Ann Arbor, Michigan. Huete, A.R. 1988. A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment 25: 295-309. Ingestad, T. 1959. Studies on the nutrition of forest tree seedlings - II: mineral nutrition of spruce. Physiologia Plantarum 12: 568-593. Ingestad, T. 1962. Macro element nutrition of pine, spruce, and birch seedlings in nutrient solutions. Medd. Statens Skogsforsknings Institute 51(7): 1-150. 189 Ingestad, T., A. Aronsson and G.I. Agren. 1981. Nutrient flux density model of mineral nutrition in conifer ecosystems. Studia Forestalia Suecica 160: 61-71. Jackson, R.D., C.A. Jones, G. Uehara and L.T. Santo. 1980. Remote detection of nutrient and water deficiencies in sugarcane under variable cloudiness. Remote Sensing of Environment 11: 327-331. Jackson, R.D., P.N. Slater and P.J. Pinter. 1983. Discrimination of growth and water stress in wheat by various vegetation indices through clear and turbid atmospheres. Remote Sensing of Environment 17: 187-208. Jagendorf, A.T. 1973. The role of phosphate in photosynthesis. Pages 381-392 in E.J. G r i f f i t h , A. Beeton, J.M. Spencer and D.T. Mitchell (Editors). Environmental phosphorus handbook. John Wiley and Sons, New York. 718 pages. Jarvis, P.G., G.B. James and J.J. Landsberg. 1976. Coniferous forests. Pages 171-240 in J.L. Monteith (Editor), Vegetation and the atmosphere; Volume 2 - case studies. Academic Press, New York. 439 pages. Jordan, CF. 1969. Derivation of leaf area index from quality of light on the forest floor. Ecology 50: 663-666. Kamat, D.S., A.K.S. Gopalan, Ajai, and M.N. Shashikumar. 1985. Assessment of water-stress effects on crops. International Journal of Remote Sensing 6: 577-589. Kirk, J.T.O. and R.A.E. Tilney-Bassett. 1967. The plastids. W.H. Freeman and Company Limited, San Francisco. 608 pages. Kleman, J. 1985. The spectral reflectance of coniferous tree stands and of barley influenced by stress: an analysis of f i e l d measured spectral data. Ph.D. Thesis, Department of Physical Geography, University of Stockholm, Stockholm, Sweden. 181 pages. Kleman, J. and E. Fagerlund. 1981. Spectral signature measurements of barley crops: The influence of irrigation and f e r t i l i z a t i o n . Pages 417-424 in International Colloquium on Spectral Signatures of Objects in Remote Sensing, September 8-11, Avignon, France, International Society for Photogrammetry and Remote Sensing. Kleman, J. and E. Fagerlund. 1987. Influence of different nitrogen and irrigation in the spectral reflectance of barley. Remote Sensing of Environment 21: 1-14. 190 Knipling, E.B. 1969. Leaf reflectance and image formation on color infrared film. Pages 17-29 in Remote sensing in ecology. University of Georgia Press, Athens, Georgia. Knipling, E.B. 1970. Physical and physiological basis for the reflectance of visible and near-infrared radiation from vegetation. Remote Sensing of Environment 1: 155-159. Koch, B., U. Ammer, T. Schneider and H. Wittmeier. 1990. Spectroradiometer measurements in the laboratory and in the f i e l d to analyse the influence of different damage symptoms on the reflection spectra of forest trees. International Journal of Remote Sensing 11: 1145-1163. Kramer, P.J. and T.T. Kozlowski. 1979. Physiology of woody plants. Academic Press, New York. 811 pages. Kumar, R. and L.F. Silva. 1973. Light ray tracing through leaf cross section. Applied Optics 12: 2950-2954. Levitt, J. 1972. Responses of plants to environmental stresses. Academic Press, New York. 697 pages. Li-Cor Inc. 1982. LI-1800 portable spectroradiometer instruction manual. Lincoln, Nebraska. Publication Number 8210-0030. Linder, S. 1980. Chlorophyll as an indicator of nitrogen status of coniferous seedlings. New Zealand Journal of Forestry Science 10: 166-175. Linder, S. and D.A. Rook. 1984. Effects of mineral nutrition on carbon dioxide exchange and partitioning of carbon in trees. Pages 211-236 in G.D. Bowen and E.K.S. Nambiar (Editors). Nutrition of plantation forests. Academic Press, Toronto. 516 pages. Loneragan, J.F., T.S. Grove, A.D. Robson and K. Snowball. 1979. Phosphorus toxicity as a factor in zinc-phosphate interactions in plants. Soil Science Society of America Journal 43: 966-972. Macy, P. 193 6. The quantitative mineral nutrient requirements of plants. Plant Physiology 11: 749-764. Majid, N.M. and T.M. Ballard. 1990. Effects of f o l i a r application of copper sulphate and urea on the growth of lodgepole pine. Forest Ecology and Management 37: 151-165. Major, D.J., F. Baret and G. Guyot. 1990. A ratio vegetation index adjusted for s o i l brightness. International Journal of Remote Sensing 11: 727-740. 191 Marschner, H. 1986. Mineral nutrition of higher plants. Academic Press, Toronto. 674 pages. Mengel, K. and E.A. Kirkby. 1987. Principles of plant nutrition. International Polish Institute, Worblaufen-Bern, Switzerland. 687 pages. Mergen, F. and J. Worrall. 1965. Effect of environment and seed source on mineral content of jack pine seedlings. Forest Science 11: 393-400. Millikan, CR. 1963. Effects of different levels of zinc and phosphorus on the growth of subterraneum clover (Trifolium subterraneum). Australian Journal of Agriculture Research 14: '180-205. Milton, N.M., W. Collins, S.H. Chang and R.G. Schmidt. 1983. Remote detection of metal anomalies on p i l o t Mountain, Randolph County, North Carolina. Economic Geology 78: 605-617. Morrison, I.K. 1974. Mineral nutrition of conifers with special reference to nutrient status interpretation: a review of literature. Canadian Forestry Service, Ottawa, Ontario, Canadian Forestry Service Publication 1343. 74 pages. Mudd, J.B. 1975. Sulphur dioxide. Pages 9-22 in J.B. Mudd and T.T. Kozlowski (Editors). Responses of plants to air pollution. Academic Press, New York. 3 83 pages. Murtha, P.A. 1982. Detection and analysis of vegetation stress. Pages 141-158 in C.J. Johannsen and J.L. Sanders (Editors). Remote sensing for resource management. Soil Conservation Society of America, Ankeny, Iowa. Murtha, P.A. and T.M. Ballard. 1983. Foliar nutrients and photo interpretation of Douglas-fir stress. Canadian Journal of Remote Sensing 9: 99-110. Myers, V.I., M.D. Heilman, R.J.P. Lyon, L.N. Namken, D. Simonett, R.J. Thomas, C.L. Weigand and J.T. Woolley. 1970. Pages 253-297 in Remote sensing with special reference to agriculture and forestry. National Academy of Science, Washington, D.C Nelson, V.L., D.H. Gjerstad and G.R. Glover. 1986. Determining nitrogen status of young loblolly pine by leaf reflectance. Tree Physiology 1: 333-339. 192 Nilsson, H.-E. 1987. Experiments on hand-held radiometry and IR-thermography of winter wheat in f i e l d plot experiments. Swedish University of Agricultural Sciences, Research Information Centre, Uppsala, Sweden. Vaxtskyddsrapporter Jordbruk 44. 48 pages. Nilsson, H.-E. and H. Linner. 1987. Remote sensing of wheat and barley in a f i e l d plot experiment with different levels of nitrogen f e r t i l i z a t i o n and irrigation. Swedish University of Agricultural Sciences, Research Information Centre, Sweden. Vaxtskyddsrapporter Jordbruk 45. 246 pages. Parkinson, J.A. and S.E. Allen. 1975. A wet oxidation procedure for the determination of nitrogen and mineral nutrients in biological material. Communications in Soil Science and Plant Analysis 6: 1-11. Pastor, J., J.D. Aber, CH. McClaugherty and J.M. Melillo. 1984. Above ground production and N and P cycling along a nitrogen mineralization gradient on Blackhawk Island, Wisconsin. Ecology 65: 256-268. Patel, N.K., T.P. Singh, S. Baldev and B.R. Patel. 1983. Study of water and nutrient stress in wheat crop through remote sensing technique. Pages O-20-1 to 0-20-8 in Proceedings of the Fourth Asian Conference on Remote Sensing, November 10-15, Colombo, Sri Lanka. Pearson, R.L. and L.D. Miller. 1972. Remote mapping of standing crop biomass for estimation of the productivity of the shortgrass prairie, Pawnee National Grasslands, Colorado. Pages 1355-1379 in Proceedings of the 8th International Symposium on Remote Sensing of the Environment, Environmental Research Institute of Michigan, Ann Arbor, Michigan. Pearson, R.L., L.D. Miller and C.J. Tucker. 1976. Hand-held spectral radiometer to estimate gramineous biomass. Applied Optics 15: 416-418. Peterson, D.L., J.D. Aber, P.A. Matson, D.H. Card, N. Swanberg, C. Wessman and M. Spanner. 1988. Remote sensing of forest canopy and leaf biochemical contents. Remote Sensing of Environment 24: 85-108. Purnell, H.M. 1958. Nutritional studies of Finns radiata Don. I symptoms due to deficiency of some major elements. Australian Forestry 22: 82-87. 193 Rennie, J.C. and E.R. Buckner. 1972. Use of remote sensing for the detection of areas of nutrient deficiencies. Pages 163-169 in B.A. Tschantz (Director), Remote sensor u t i l i z a t i o n for environmental system studies. University of Tennessee, Knoxville, Tennessee, Contract No. F19628-69-C-0016, Project No. 7259, Project Themis Final Report (Volume I), Prepared for Air Force Cambridge Laboratories, United States Air Force, Bedford, Massachusetts. Rennie, J.C. and D.H. Cress. 1975. Multispectral imagery for detection of nutrient deficiencies in pine plantations. Pages 345-355 in F. Shahrokhi (Editor), Technical papers selected from the Fourth Remote Sensing of Earth Resources Conference, Volume IV, The University of Tennessee Space Institute, Tullahoma, Tennessee. Rennie, P.J. and R.L. Halstead. 1977. The effects of sulphur on plants in Canada. Pages 69-179 in Sulphur and i t s inorganic derivatives in the Canadian environment. National Research Council of Canada, Publication No. NRCC 15015 of the Environmental Secretariat. Ottawa, Ontario. 426 pages. Richardson, A.J., J.H. Everitt and H.W. Gausman. 1983. Radiometric estimation of biomass and nitrogen contents of a l i c i a grass. Remote Sensing of Environment 13: 179-184. Rock, B.N., T. Hoshizaki and J.R. Miller. 1988. Comparison of in situ and airborne spectral measurements of the blue shift associated with forest decline. Remote Sensing of Environment 24: 109-127. Rock, B.N. and J.E. Vogelmann. 1985. Stress assessment and spatial characterization of suspected acid deposition damage in red spruce (Picea rubens) from Vermont. Pages 860-870 in American Society of Photogrammetry Technical Papers, Volume 2, 51st Annual Meeting, American Society of Photogrammetry and Remote Sensing, Falls Church, Virginia. Rock, B.N., J.E. Vogelmann, D.L. Williams, A.F. Vogelmann and T. Hoshizaki. 1986. Remote detection of forest damage. Bioscience 36: 439-445. Rouse, J.W. (Jr.), R.H. Hass, J.A. Schell and D.W. Deering. 1973. Monitoring vegetation systems in the Great Plains with ERTS. Pages 309-317 in Proceedings of the 3rd ERTS-1 Symposium, NASA SP 351. Salisbury, F.B. and C.W. Ross. 1985. Plant physiology. Wadsworth Publishing Company, Belmont, California. 540 pages. 194 Schomaker, CE. 1969. Growth and f o l i a r nutrition of white pine seedlings as influenced by simultaneous changes in moisture and nutrient supply. Soil Science of America Proceedings 33: 614-618. Smith, M.E. 1943. Micronutrients essential for the growth of Pinus radiata. Australian Forestry 7:22-27. Smith, M.E. and N.S. Bayliss. 1942. The necessity of zinc for Pinus radiata. Plant Physiology 17: 303-310. Spain, J.D. 1982. Basic microcomputer models in biology. Addison-Wesley Publishing Company, Don Mills. 354 pages. Spanner, M.A., L.L. Pierce, D.L. Peterson and S.W. Running. 1990. Remote sensing of temperate coniferous forest leaf area index: the influence of canopy closure, understory vegetation and background reflectance. International Journal of Remote Sensing 11: 95-111. Stoate, T.N. 1950. Nutrition of the pine. Australian Forest Timber Bureau Bulletin 30. 61 pages. Swan, H.S.D. 1972. Foliar nutrient concentrations in lodgepole pine as indicators of tree nutrient status and f e r t i l i z e r requirements. Pulp and Paper Research Institute of Canada. Woodlands Report 42. 19 pages. Thomas, J.R., V.I. Myes, M.D. Heilman and CL. Wiegand. 1966. Factors affecting light reflectance of cotton. Pages 305-312 in Proceedings of the 4th Symposium on Remote Sensing of the Environment, Ann Arbor, Michigan. Thomas, J.R. and G.F. Oerther. 1972. Estimating nitrogen content of sweet pepper leaves by reflectance measurements. Agronomy Journal 64: 11-13. Thomson, W. and T.E. Weier. 1962. The fine structure of chloroplasts from mineral - deficient leaves of Phaseolus vulgaris. American Journal of Botany 49: 1047-1055. Thornber, J.P. 1975. Chlorophyll-proteins: light harvesting and reaction center components of plants. Annual Review of Plant Physiology 26: 127-158. Timmer, V.R. 1991. Interpretation of seedling analysis and visual symptoms. Pages 113-134 in R. van den Driessche (Editor). Mineral nutrition of conifer seedlings. CRC Press, Ann Arbor, Michigan. 274 pages. 195 Tinus, R.W. and S.E. McDonald. 1979. How to grow tree seedlings in containers in greenhouses. Rocky Mountain Forest and Range Experimental Station, U.S. Department of Agriculture Forest Service, Bottineau, North Dakota. General Technical Report RM-60. 256 pages. Tsay, M.L., D.H. Gjerstad and G.R. Glover. 1982. Tree leaf reflectance: a promising technique to rapidly determine nitrogen and chlorophyll content. Canadian Journal of Forest Research 12: 788-792. Tucker, C.J. 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment 8: 127-150. Tucker, C.J. 1980. Remote sensing of leaf water content in the near infrared. Remote Sensing of Environment 10: 23-32. Tucker, C.J., J.H. Elgin and J.E. McMurtrey. 1979. Temporal spectral measurements of corn and soybean crops. Photogrammetric Engineering and Remote Sensing 45: 643-653. Tucker, C.J., L.D. Miller and R.L. Pearson. 1975. Shortgrass prairie spectral measurements. Photogrammetric Engineering and Remote Sensing 41: 1157-1162. Turner, D.O. 1966. Color and growth of Douglas-fir Christmas trees as affected by f e r t i l i z e r application. Soil Science Society of America Proceedings 30: 792-795. Turner, J. , M.J. Lambert and S.P. Gessel. 1977. Use of foliage sulphate concentrations to predict response to urea application by Douglas-fir. Canadian Journal of Forest Research 7: 476-480. Van Cleve, K.L., L. Oliver, R. Schlenter, L.A. Viereck and C.T. Dyrness. 1983. Productivity and nutrient cycling in taiga forest ecosystems. Canadian Journal of Forest Research 13: 747-766. van den Driessche, R. 1969. Tissue nutrient concentrations of Douglas-fir and Sitka spruce in sand cultures, and value of nutrient concentration levels for interpreting f o l i a r analysis. B.C. Forest Service, Department of Lands, Forests, and Water Resources, Victoria, B.C. Research Note 47. 42 pages. Vitousek, P.M. 1982. Nutrient cycling and nutrient use efficiency. * American Naturalist 119: 553-572. 196 Vygodskaya, N.N., I.I. Gorshkova and Ye. V. Fadeyeva. 1989. Theoretical estimates of sensitivity in some vegetation indices to variation in the canopy conditions. International Journal of Remote Sensing 10: 1857-1872. Walburg, G., M.E. Bauer and C.S.T. Daughtry. 1981. Effects of nitrogen on the growth, yield and reflectance characteristics of corn canopies. Laboratory for Applications of Remote Sensing, West Lafayette, Indiana. Technical Report SR-P1-04044, NAS9-15466. 19 pages. Walburg, G., M.E. Bauer, C.S.T. Daughtry and T.L. Housley. 1982. Effects of nitrogen nutrition on the growth, yield, and reflectance characteristics of corn canopies. Agronomy Journal 74: 677-683. Watanabe, F.S., W.L. Lindsay and S.R. Olsen. 1965. Nutrient balance involving phosphorus, iron, and zinc. Soil Science Society of America Proceedings 29: 562-565. Weetman, G.F. 1968. The nitrogen f e r t i l i z a t i o n of three black spruce stands. Pulp and Paper Research Institute of Canada, Pointe Claire, Quebec. Woodlands Papers Number 6. 45 pages. Wessman, C.A., J.D. Aber, D.L. Peterson and J.M. Melillo. 1988a. Foliar analysis using near infrared spectroscopy. Canadian Journal of Forest Research 18: 6-11. Wessman, CA. , J.D. Aber, D.L. Peterson and J.M. Melillo. 1988b. Remote sensing of canopy chemistry and nitrogen cycling in temperate forest ecosystems. Nature 335: 154-156. Wessman, CA., J.D. Aber and D.L. Peterson. 1989. An evaluation of imaging spectroscopy for estimating forest canopy chemistry. International Journal of Remote Sensing 10: 1293-1316. Wilkinson, L. 1988. SYSTAT: the system for st a t i s t i c s . Systat Incorporated, Evanston, I l l i n o i s . 822 pages. Wolf, B. 1974. Improvements in the azomethine H method for the determination of boron. Communication in Soil Science and Plant Analysis 5: 39-44. Wolf, F.T. 1956. Changes in chlorophylls a and b in autumn leaves. American Journal of Botany 43: 714-718. Woolley, J.T. 1971. Reflectance and transmittance of light by leaves. Plant Physiology 47: 656-662. Wyne Jones, R.G. and O.R. Lunt. 1967. The function of calcium in plants. Botanical Review 33: 407-426. 197 Zar, J.H. 1984. Biostatistical analysis. Prentice-Hall Incorporated, Englewood C l i f f s , New Jersey. 718 pages. 1 9 8 APPENDICES APPENDIX A. COMMON AND SCIENTIFIC NAMES > 200 Table Al. Common and sc i e n t i f i c names of species mentioned in thesis. Common name Scientific name a l i c i a grass Cynodon spp. amaranth Amaranthus spp. barley Hordeum vulgare L. beech Fagus spp. black spruce Picea mariana (Mill.) B.S.P. buffelgrass Cenchrus c i l i a r i s chickpea Cicer arietinum L. cotton Gossypium hirsutum L. Douglas-fir Pseudotsuga menziesii (Mirb.) Franco loblolly pine Pinus taeda L. lodgepole pine Pinus contorta Dougl. maize Zea mays L. Mexican squash Curcurbita pepo L. mountain pine beetle Dendrocktonus ponderosae Hopkins Norway spruce Picea abies (L.) Karst Pacific f i r Picea amabilis (Dougl.) Forbes pea Pisum satiyum L. ponderosa pine Pinus ponderosa Laws. poplar Populus spp. red spruce Picea rubens Sarg. Scotch pine Pinus silvestris L. Continued Table Al. Continued Common name Scientific name soybean spruce sugar beet sugar cane sweet pepper Valencia orange western hemlock wheat white oak Glycine max (L.) Merr. Picea spp. Beta spp. Saccharum spp. Capsicum annum L. Citrus sinensis L. Osbeck Tsuga heterophylla (Raf.) Sarg. Triticum aestivum L. Quercus alba L. APPENDIX B. NUTRIENT SOLUTION COMPOSITION 203 Table A2. Concentration and source of elements in standard nutrient solution for the 1 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 1.00 NH4N03 2.86 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgSO4.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl2. 2H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3B03 2.29 Mo 0.03 Na2Mo04. H20 0.071 204 Table A3. Concentration and source of elements in standard nutrient solution for the 5 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 5.00 NH4N03 14.29 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl 2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl2. 2H20 Fe 5.00 FeS04.7H20 . 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.07,9 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 205 Table A4. Concentration and source of elements in standard nutrient solution for the 10 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 10.00 NH4N03 28.57 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 CI 56.33 70.76 KCl CSLC1L2 • 2H20 Fe 5.00 FeS04. 7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04. 7H20 0.22 B 0.40 H3BO3 2.29 Mo 0. 03 Na2Mo04.2H20 0.071 Table A5. Concentration and source of elements in standard nutrient solution for the 25 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 25.00 NH4N03 71.43 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04. 7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl2. 2H20 Fe 5.0 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2. 4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0. 05 ZnS04.7H40 0.22 B 0.40 H3B03 2.29 Mo 0. 03 Na2Mo04.2H20 0.071 207 Table A6. Concentration and source of elements in standard nutrient solution for the 50 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 50.00 NH4N03 142.86 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI . 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl 2. 2H20 Fe 5.00 FeS04.7H40 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0. 02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 208 Table A7. Concentration and source of elements in standard nutrient solution for the 100 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 100.00 NH4N03 285.72 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl2. 2H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 209 Table A8. Concentration and source of elements in standard nutrient solution for the 250 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 250.00 NH4N03 714.30 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl2. 2H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2. 4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04. 2H20 0.071 210 Table A9. Concentration and source of elements in standard nutrient solution for the 400 mg/L nitrogen treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L N 400.00 NH4N03 1142.88 P 30.00 KH2P04 131.81 K 37.87 K 37.87 62.13 KH2P04 KCl 118.47 CI 56.33 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2.2H20 146.72 S 65.98 2.87 MgS04. 7H20 FeS04.7H20 CI 56.33 70.76 KCl CaCl2.2H20 Fe 5. 00 FeS04.7H40 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 Table A10. Concentration and source of elements in standard nutrient solution for the 1 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 1.0 KH2P04 4.39 K 1.26 N 64.63 35.37 NH4N03 KN03 184.66 255.32 K 98.74 K 1.26 98.74 KH2P04 KN03 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.00 FeS04. 7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 212 Table A l l . Concentration and source of elements in standard nutrient solution for the 5 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 5.0 KH2P04 21.97 K 6.31 N 66.44 33.56 NH4N03 KN03 189.83 242.26 K 93.69 K 6.31 93.69 KH2P04 KN03 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 213 Table A12. Concentration and source of elements in standard nutrient solution for the 10 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 10.00 KH2P04 43.94 K 12.63 N 68.70 31.30 NH4N03 KN03 196.29 225.92 K 87.37 K 12.63 87.37 KH2P04 KN03 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2.2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 214 Table A13. Concentration and source of elements in standard nutrient solution for the 25 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 25.00 KH2P04 109.84 K 31.56 N 75.53 24.47 NH4N03 KN03 215.80 176.97 K 68.44 K 31.56 68.44 KH2P04 KNO3 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 215 Table A14. Concentration and source of elements in standard nutrient solution for the 50 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 50.00 KH2P04 219.69 K 63.12 N 86.79 13.21 NH4N03 KN03 247.98 95.36 K 36.88 K 63.12 36.88 KH2P04 K N O 3 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl 2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.0 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2. 4H20 0.72 Cu 0.02 CuS04.5H20 0. 079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.079 216 Table A15. Concentration and source of elements in standard nutrient solution for the 100 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 79.21 20.79 KH2P04 (NH4)2HP04 348.04 88.64 K N 100.00 18.80 N 18.80 81.20 (NH4)2HP04 NH4N03 232.00 K 100.00 KH2P04 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 ' S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 217 Table A16. Concentration and source of elements in standard nutrient solution for the 250 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 79.21 110.57 60.22 KH2P04 (NH4)2HP04 H3P04 348.04 471.42 190.53 K N 100.00 100.00 N 100.00 (NH4)2HP04 K 100.00 KH2P04 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04. 7H20 FeS04.7H20 Fe 5.00 FeS04.7H20 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04. 7H20 0.22 B 0.40 H3B03 2.29 Mo 0. 03 Na2Mo04. 2H20 0.071 218 Table A17. Concentration and source of elements in standard nutrient solution for the 400 mg/L phosphorus treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L p 79.21 110.57 210.22 KH2P04 (NH4)2HP04 H3P04 348.04 471.42 665.13 K N 100.00 100.00 N 100.00 (NH4)2HP04 K 100.00 KH2P04 Mg 50.00 MgS04.7H20 506.72 S 65.98 Ca 40.00 CaCl2. 2H20 146.72 S 65.98 2.87 MgS04.7H20 FeS04.7H20 Fe 5.00 FeS04.7H2 24.90 S 2.87 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B- 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 219 Table A18. Concentration and source of elements in standard nutrient solution for the 1 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L S 1.0 (NH4)2S04 4.12 N 0.87 N 0.87 22.26 76.87 (NH4)2S04 KN03 NH4N03 160.66 219.63 K 62.13 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 KNO3 KH2P04 Mg 50.00 MgCl2.6H20 195.83 Ca 40.00 CaCl2. 2H20 146.72 Fe 5. 00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0. 05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 220 Table A19. Concentration and source of elements in standard nutrient solution for the 5 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L s 5.00 (NH4)2S04 20. 61 N 4.37 N 4.37 22.26 73.37 (NH4)2S04 KN03 NH4N03 160.66 219.63 K 62.13 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 K N O 3 KH2P04 Mg 50.00 MgCl2.6H20 195.83 Ca 40.00 CaCl2. 2H20 146.72 Fe 5.00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 Table A20. Concentration and source of elements in standard nutrient solution for the 10 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L S 10.00 (NH4)2S04 41.21 N 8.74 N 8.74 22.26 69.00 (NH4)2S04 KN03 NH4N03 160.66 197.15 K 62.13 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 KNO3 KH2P04 Mg 50.00 MgCl2.6H20 195.83 Ca 40.00 CaCl 2. 2H20 146.72 Fe 5.00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04.5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3B03 2.29 Mo 0.03 Na2Mo04.2H20 0.071 222 Table A21. Concentration and source of elements in standard nutrient solution for the 25 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L s 25.00 (NH4)2S04 103.03 N 21.84 N 21.84 22.26 55.90 (NH4)2S04 KN03 NH4N03 160.66 159.72 K 62.13 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 . KN03 KH2P04 Mg 50.00 MgCl2.6H20 195.83 Ca 40.00 CaCl 2. 2H20 146.72 Fe 5.00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H3B03 2.29 Mo 0.03 Na2Mo04.2H20 0.071 223 Table A22. Concentration and source of elements in standard nutrient solution for the 50 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L s 50.00 (NH4)2S04 206.06 N 43.68 N 43.68 22.26 34.06 (NH4)2S04 KN03 NH4N03 160.00 97.32 K 62.13 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 KN03 KH2P04 Mg 50.00 MgCl2.6H20 195.83 Ca 40.00 CaCl2. 2H20 146.72 Fe 5.00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0.05 ZnS04.7H20 0.22 B 0.40 H 3 B O 3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 224 Table A23. Concentration and source of elements in standard nutrient solution for the 100 mg/L sulphur treatment. Element Element mg/L Source Source mg/L Associated Element Element mg/L S 74.53 25.47 (NH4)2S04 K2S04 307.15 138.45 N K 65.12 62.13 N 65.12 34.88 (NH4)2S04 NH4N03 99.66 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 K2S04 KH2P04 Mg 50.00 MgCl2. 6H20 195.83 Ca 40.00 CaCl2. 2H20 146.72 Fe 5.00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0. 079 Zn 0.05 ZnS04. 7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 Na2Mo04.2H20 0.071 225 Table A24. Concentration and source of elements in standard nutrient solution for the 250 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L S 114.46 25.47 65.98 16.00 28.09 (NH4)2S04 K2S04 MgS04. 7H20 CaS04. 2H20 H2S04 471.70 138.45 506.72 85.91 85.92 N K Mg Ca 100.00 62.13 50.00 20.00 N 100.00 (NH4)2S04 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 K2S04 KH2P04 Mg 50.00 MgS04.7H20 Ca 20.00 20.00 CaS04. 2H20 CaCl2. 2H20 73.36 Fe 5.0 FeCl3.6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0. 05 ZnS04.7H20 0.22 B 0.40 H3BO3 2.29 Mo 0.03 NaMo04.2H20 0.071 226 Table A25. Concentration and source of elements i n standard nutrient solution for the 400 mg/L sulphur treatment. Element Element Source Source Associated Element mg/L mg/L Element mg/L s 114.46 25.47 65.98 16.00 178.09 (NH4)2S04 K2S0y MgS04.7H20 CaS04. 2H20 H2S04 471.70 138.45 506.72 85.91 544.76 N K Mg Ca 100.00 62.13 50.00 20.00 N 100.00 (NH4)2S04 P 30.00 KH2P04 131.81 K 37.87 K 62.13 37.87 K2S04 KH2P04 Mg 50.00 MgS04.7H2O Ca 20.00 20. 00 CaS04. 2H20 CaCl 2. 2H20 73.36 Fe 5.00 FeCl 3. 6H20 24.20 Mn 0.20 MnCl2.4H20 0.72 Cu 0.02 CuS04. 5H20 0.079 Zn 0. 05 ZnS04.7H20 0.22 B 0.40 H3B03 2.29 Mo 0. 03 Na2Mo04.2H20 0.071 APPENDIX C. FOLIAR NUTRIENT CONCENTRATIONS ' t Table 26. F o l i a r nutrient concentrations f or composite samples of Douglas-fir seedlings subjected to the 1987 nitrogen and phosphorus treatments. Sample N P Ca Mg K Cu Zn Fe Mn B la b e l % ppm N 1 mg/L composite 0. 86 0.36 0. 25 0. 22 1.05 5 12 135 165 77 N 5 mg/L composite 1. 01 0.66 0. 43 0. 38 1.50 6 17 158 285 NA N 10 mg/L composite 1. 24 0.87 0. 29 0. 32 1.50 6 38 150 233 81 N 25 mg/L composite 1. 35 0.75 0. 60 0. 55 1.58 5 14 128 263 151 N 50 mg/L composite 1. 80 0.93 0. 46 0. 39 1.88 6 49 150 248 113 N 100 mg/L composite 2. 34 0.86 0. 55 0. 44 0.97 5 10 140 282 131 N 250 mg/L composite 1. 84 0.32 0. 35 0. 29 0.83 4 16 130 173 42 N 400 mg/L composite 3. 15 0.55 0. 50 0. 35 0.90 5 30 128 255 111 P 1 mg/L composite 1. 86 0.14 0. 27 0. 26 0.93 6 9 120 177 NA P 5 mg/L composite 2. 14 0.17 0. 37 0. 33 1.00 7 93 143 227 NA P 10 mg/L composite 2. 53 0.23 0. 44 0. 79 0.95 11 15 68 268 NA P 25 mg/L composite 1. 73 0.48 0. 46 0. 41 1.43 6 36 120 255 NA P 50 mg/L composite 2. 10 0.70 0. 46 0. 36 1.20 6 16 120 248 NA P 100 mg/L composite 1. 95 1.04 0. 39 0. 37 1.13 5 18 128 225 NA P 250 mg/L composite 1. 88 1.34 0. 32 0. 30 1.28 5 38 113 180 NA P 400 mg/L composite 1. 73 1.93 0. 38 0. 38 1.35 5 53 113 173 NA NA - data unavailable; i n s u f f i c i e n t f o l i a r material a v a i l a b l e f o r a n a l y s i s Table 27. F o l i a r nutrient concentrations for composite samples and i n d i v i d u a l Douglas-fir seedlings subjected to the 1988 nitrogen treatments. Sample la b e l N P Ca % Mg K S Cu Zn Fe ppm Mn B N 1 mg/L composite 1. 13 0. 66 0.30 0.309 1.56 0.43 5 25 97 251 79 N 1 mg/L #4 0. 92 0. 50 0.46 0.246 1.28 NA 4 28 75 156 53 N 1 mg/L #5 1. 06 0. 64 0.32 0.343 1.82 0.30 4 22 90 160 67 N 1 mg/L #6 0. 94 0. 50 0.20 0.185 0.80 0.22 4 18 134 141 51 N 5 mg/L composite 1. 09 0. 54 0.24 0.213 1.13 0.28 4 17 134 213 57 N 5 mg/L #3 1. 04 0. 54 0.24 0.187 1.11 0.28 4 12 90 139 50 N 5 mg/L #6 0. 96 0. 57 0.26 0.241 1.70 0.31 3 17 74 144 71 N 5 mg/L #10 0. 80 0. 43 0.26 0.197 0.92 0.22 4 28 257 153 47 N 10 mg/L composite 1. 60 0. 70 0.48 0.361 1.18 0.25 4 22 96 413 50 N 10 mg/L #2 1. 05 0. 56 0.36 0.303 1.39 0.34 5 25 86 259 80 N 10 mg/L #4 0. 96 0. 53 0.22 0.178 1.12 0.31 4 13 85 156 51 N 10 mg/L #6 1. 04 0. 64 0.24 0.250 0.98 0.29 4 21 101 216 47 N 25 mg/L composite 1. 48 0. 75 0.38 0.375 1.26 0.45 5 32 160 279 102 N 25 mg/L #4 1. 49 0. 74 0.34 0.344 1.31 0.45 5 22 97 286 115 N 25 mg/L #5 1. 47 0. 58 0.40 0.255 1.09 0.34 5 26 69 243 105 N 25 mg/L #9 1. 23 0. 71 0.36 0.331 1.23 0.33 5 26 90 234 73 N 50 mg/L composite 1. 68 0. 58 0.38 0.314 1.00 0.44 5 22 112 228 84 N 50 mg/L #1 1. 42 0. 63 0.30 0.313 1.68 0.58 5 25 85 144 116 N 50 mg/L #6 1. 88 0. 71 0.44 0.421 0.86 0.40 4 28 107 192 72 N 50 mg/L #10 1. 31 0. 57 0.30 0.274 1.07 0.35 4 19 75 246 85 N 100 mg/L composite 1. 76 0. 41 0.40 0.280 0.86 0.48 5 32 96 216 92 N 100 mg/L #1 1. 36 0. 39 0.22 0.223 0.76 0.34 4 16 90 152 59 N 100 mg/L #2 1. 57 0. 47 0.18 0.185 0.72 0.31 4 29 106 195 45 N 100 mg/L #4 2. 28 0. 47 0.40 0.228 0.94 0.42 5 29 107 168 82 N 250 mg/L composite 2. 55 0. 37 0.28 0.211 0.78 0.44 6 15 122 104 80 N 250 mg/L #3 2. 26 0. 33 0.26 0.353 0.81 0.41 6 16 96 102 76 N 250 mg/L #7 2. 88 0. 45 0.35 0.410 0.70 0.45 5 20 110 166 114 N 250 mg/L #8 1. 96 0. 37 0.36 0.182 0.66 NA 5 13 101 124 44 N 400 mg/L composite 3. 77 0. 45 0.41 0.333 1.03 0.35 6 17 103 162 78 N 400 mg/L #5 1. 86 0. 25 0.26 0.184 0.50 0.47 6 15 106 99 66 N 400 mg/L #6 3. 39 0. 38 0.38 0.237 0.82 0.45 8 13 96 136 65 N 400 mg/L #9 3. 57 0. 43 0.34 0.286 0.82 0.34 6 16 107 200 61 NA - data unavailable; i n s u f f i c i e n t f o l i a r material a v a i l a b l e f o r a n a l y s i s . to to Table 28. F o l i a r nutrient concentrations for composite samples and i n d i v i d u a l Douglas-fir seedlings subjected to the 1988 phosphorous treatments. Sample label N P Ca % Mg K S Cu Zn Fe ppm Mn B p 1 mg/L composite 2.19 0.17 0.30 0.241 0.58 0.40 5 21 96 207 85 P 1 mg/L #3 1.73 0.16 0.20 0.201 0.60 0.37 5 15 90 143 74 P 1 mg/L #4 2.55 0.13 0.22 0.177 0.42 0.48 6 21 80 143 62 P 1 mg/L #7 2.11 0.15 0.26 0.211 1.01 0.48 6 14 95 191 74 P 5 mg/L composite 1.82 0.20 0.28 0.195 0.73 0.45 6 21 105 154 79 P 5 mg/L #5 2.02 0.17 0.30 0.214 1.17 0.76 6 21 111 131 78 P 5 mg/L #9 2.07 0.21 0.26 0.189 0.84 0.68 5 21 111 211 97 P 5 mg/L #10 1.62 0.17 0.34 0.255 0.54 0.47 5 22 74 144 83 P 10 mg/L composite 1.66 0.26 0.32 0.233 0.75 0.42 5 15 74 162 70 P 10 mg/L #1 1.69 0.22 0.24 0.221 0.69 0.37 5 13 89 138 62 P 10 mg/L #3 1.59 0.26 0.24 0.230 0.54 0.30 5 21 79 147 70 P 10 mg/L #4 1.52 0.23 0.28 0.217 0.81 0.53 4 10 79 118 62 P 25 mg/L composite 1.75 0.41 0.30 0.254 0.87 0.43 5 22 117 230 67 P 25 mg/L #5 1.96 0.41 0.26 0.274 0.96 0.52 5 19 91 244 128 P 25 mg/L #6 2.03 0.39 0.42 0.295 0.72 0.43 4 41 85 211 88 P 25 mg/L #8 2.04 0.45 0.42 0.264 0.80 0.41 5 25 101 260 59 P 50 mg/L composite 1.82 0.65 0.40 0.303 1.18 0.55 5 31 85 271 88 P 50 mg/L #3 1.72 0.65 0.46 0.337 1.11 0.47 6 38 87 172 99 P 50 mg/L #5 1.61 0.72 0.58 0.312 1.27 0.65 6 25 85 183 121 P 50 mg/L #8 1.73 0.62 0.43 0.337 1.12 0.44 5 39 110 263 113 P 100 mg/L composite 1.51 0.81 0.33 0.240 1.32 0.41 6 41 79 173 66 P 100 mg/L #1 1.05 0.58 0.28 0.223 0.85 0.34 5 25 79 134 44 P 100 mg/L #3 1.33 0.69 0.37 0.188 0.98 0.43 5 39 68 161 52 P 100 mg/L #5 1.46 0.71 0.26 0.209 1.30 0.38 6 67 100 165 70 P 250 mg/L composite 1.42 0.70 0.20 0.153 0.91 0.27 5 29 94 209 54 P 250 mg/L #6 1.18 0.79 0.22 0.159 1.20 0.23 4 25 100 142 56 P 250 mg/L #9 1.29 0.61 0.14 0.082 1.15 0.25 4 19 104 114 24 P 250 mg/L #10 1.47 0.74 0.25 0.182 1.12 0.29 3 38 89 157 41 P 400 mg/L composite 1.57 0.83 0.28 0.163 1.18 0.27 5 35 89 193 51 P 400 mg/L #6 1.31 0.75 0.18 0.096 1.10 0.26 4 18 94 114 64 P 400 mg/L #7 1.47 0.76 0.20 0.110 0.98 0.26 5 35 94 110 26 P 400 mg/L #8 1.17 0.61 0.10 0.095 0.81 0.21 4 13 100 83 52 NA - data unavailable; i n s u f f i c i e n t f o l i a r material a v a i l a b l e f o r an a l y s i s . Table 29. F o l i a r nutrient concentrations for composite samples and i n d i v i d u a l Douglas-fir seedlings subjected to the 1988 sulphur treatments. Sample la b e l N P Ca % Mg K S Cu Zn Fe ppm Mn B S 1 mg/L composite 1.90 0.38 0.26 0.148 0.95 0.18 5 59 116 170 75 S 1 mg/L #1 2.04 0.36 0.25 0.165 1.18 0.17 5 65 141 122 72 S 1 mg/L #3 1.91 0.35 0.24 0.146 0.77 NA 6 51 184 134 90 S 1 mg/L #6 2.00 0.34 0.23 0.159 1.08 0.21 6 76 94 133 83 S 5 mg/L composite 2.12 0.43 0.94 0.257 1.18 0.51 5 28 94 228 113 S 5 mg/L #4 1.82 0.38 0.34 0.239 1.01 0.25 5 17 79 162 119 S 5 mg/L #5 1.59 0.33 0.22 0.145 1.14 0.37 4 12 94 114 85 S 5 mg/L #9 1.68 0.40 0.29 0.239 1.12 0.33 6 19 73 153 84 S 10 mg/L composite 2.02 0.49 0.42 0.233 1.11 0.33 5 24 74 174 101 S 10 mg/L #1 1.63 0.39 0.39 0.197 0.88 0.22 4 27 79 142 89 S 10 mg/L #3 1.80 0.47 0.31 0.202 1.26 0.28 7 21 84 118 70 S 10 mg/L #7 2.06 0.46 0.33 0.194 1.37 0.29 6 27 110 169 107 S 25 mg/L composite 1.89 0.44 0.31 0.200 1.22 0.45 4 21 94 181 85 S 25 mg/L #2 1.98 0.44 0.43 0.331 0.87 0.56 5 32 108 236 126 S 25 mg/L #6 1.75 0.46 0.38 0.252 1.15 0.41 5 12 69 167 100 S 25 mg/L #7 1.63 0.37 0.46 0.264 0.99 0.34 5 17 79 131 84 S 50 mg/L composite 1.94 0.47 0.28 0.194 0.99 0.42 4 17 79 190 84 S 50 mg/L #1 1.44 0.47 0.20 0.129 0.99 NA 4 17 90 87 54 S 50 mg/L #5 1.53 0.43 0.36 0.234 1.33 0.52 4 19 64 207 76 S 50 mg/L #9 1.84 0.43 0.22 0.168 0.95 0.40 4 21 79 170 59 S 100 mg/L composite 1.69 0.43 0.26 0.165 0.99 0.51 4 20 69 230 86 S 100 mg/L #1 1.54 0.46 0.20 0.152 1.32 0.61 3 14 90 184 65 S 100 mg/L #7 1.61 0.44 0.20 0.122 0.86 0.30 4 22 84 193 95 S 100 mg/L #10 1.49 0.33 0.22 0.130 1.08 0.49 3 10 73 122 72 S 250 mg/L composite 1.73 0.41 0.16 0.157 0.98 0.60 3 13 68 165 64 S 250 mg/L #3 1.70 0.37 0.08 0.116 1.14 0.83 4 12 53 128 72 S 250 mg/L #4 1.59 0.38 0.11 0.134 0.88 0.50 3 9 63 122 36 S 250 mg/L #6 1.56 0.44 0.11 0.109 0.79 0.49 3 11 58 107 25 S 400 mg/L composite 1.50 0.36 0.12 0.111 0.87 0.48 3 12 74 126 42 S 400 mg/L #4 1.38 0.31 0.08 0.075 0.67 0.42 4 9 68 83 31 S 400 mg/L #7 1.53 0.37 0.08 0.067 1.04 0.53 4 11 99 117 51 S 400 mg/L #8 1.40 0.32 0.11 0.081 0.88 0.43 3 9 94 90 41 NA - data unavailable; i n s u f f i c i e n t f o l i a r material a v a i l a b l e for an a l y s i s . 232 APPENDIX D. CHLOROPHYLL A , B AND TOTAL CHLOROPHYLL CONCENTRATIONS 233 Table A30. Chlorophyll concentrations for composite samples of Douglas-fir seedlings subjected to the 1987 nitrogen and phosphorus treatments. Sample Chlorophyll Chlorophyll Total label a b chlorophyll (mg/g) (mg/g) (mg/g) N 1 mg/L composite 0.35 0.11 0.46 N 5 mg/L composite 0.72 0.20 0.95 N 10 mg/L composite 0.38 0.16 0.64 N 25 mg/L composite 0.50 0.23 0.78 N 50 mg/L composite 0.42 0.16 0.63 N 100 mg/L composite 0.78 0.30 1.20 N 250 mg/L composite 0.69 0.26 0.95 N 400 mg/L composite 1.46 0.43 1.90 P 1 mg/L composite 0.85 0.27 1.12 P 5 mg/L composite 1.07 0.31 1.38 P 10 mg/L composite 1.36 0.41 1.77 P 25 mg/L composite 0.09 0.27 0.36 P 50 mg/L composite 1.10 0.33 1.43 P 100 mg/L composite 0.75 0.24 0.99 P 250 mg/L composite 0.77 0.25 1.05 P 400 mg/L composite 0.52 0.13 0.71 234 Table A31. Chlorophyll concentrations for composite samples and individual Douglas-fir seedlings subjected to the 1988 nitrogen treatments. Sample Chlorophyll Chlorophyll Total label a b chlorophyll (mg/g) (mg/g) (mg/g) N 1 mg/L composite 0.28 0.15 0.49 N 1 mg/L #4 0.30 0.16 0.49 N 1 mg/L #5 0.33 0.19 0.55 N 1 mg/L #6 0.29 0.16 0.50 N 5 mg/L composite 0.33 0.18 0.57 N 5 mg/L #3 0.24 0.12 0.43 N 5 mg/L #6 0.22 0.11 0.37 N 5 mg/L #10 0.20 0.11 0.38 N 10 mg/L composite 0.45 0.25 0.77 N 10 mg/L #2 0.34 0.19 0.61 N 10 mg/L #4 0.19 0.10 0.37 N 10 mg/L #6 0.26 0.15 0.52 N 25 mg/L composite 0.42 0.22 0.68 N 25 mg/L #4 0.46 0.23 0.75 N 25 mg/L #5 0.48 0.25 0.76 N 25 mg/L #9 0.43 0.22 0.68 N 50 mg/L composite 0.71 0.35 1.08 N 50 mg/L #1 0.45 0.22 0.67 N 50 mg/L #6 0.40 0.20 0. 65 N 50 mg/L #10 0. 49 0.24 0.77 N 100 mg/L composite 0.82 0.39 1.23 N 100 mg/L #1 0.53 0.25 0.78 N 100 mg/L #2 0.62 0.31 0.94 N 100 mg/L #4 1.10 0.55 1.65 N 250 mg/L composite 1.52 0.76 2.28 N 250 mg/L #3 0.87 0.43 1.30 N 250 mg/L #7 1.23 0. 63 1.86 N 250 mg/L #8 0.97 0.47 1.44 N 400 mg/L composite 1.24 0.63 1.87 N 400 mg/L #5 1.01 0.41 1.42 N 400 mg/L #6 1.42 0.68 2.10 N 400 mg/L #9 1.55 0.77 2 . 32 235 Table A32. Chlorophyll concentrations for composite samples and individual Douglas-fir seedlings subjected to the 1988 phosphorus treatments. Sample label Chlorophyll a (mg/g) Chlorophyll b (mg/g) Total chlorophyll (mg/g) p 1 mg/L composite 0.96 0.48 1.47 P 1 mg/L #3 0.66 0.34 1.05 P 1 mg/L #4 0.78 0.41 1.24 P 1 mg/L #7 0.74 0.37 1.14 P 5 mg/L composite 0.92 0.44 1.36 P 5 mg/L #5 0.92 0.47 1.39 P 5 mg/L #9 1.09 0.49 1.58 P 5 mg/L #10 0.84 0.40 1.24 P 10 mg/L composite 0.52 0.25 0.82 P 10 mg/L #1 0.84 0.40 1.24 P 10 mg/L #3 0.68 0.33 1.06 P 10 mg/L #4 0.94 0.44 1.41 P 25 mg/L composite 0.76 0.37 1.19 P 25 mg/L #5 1. 00 0.46 1.46 P 25 mg/L #6 1.15 0.52 1.70 P 25 mg/L #8 0.97 0.46 1.43 P 50 mg/L composite 0. 64 0. 32 0.98 P 50 mg/L #3 0.78 0.38 1.16 P 50 mg/L #5 0.66 0.33 0.99 P 50 mg/L #8 0.70 0.33 1.07 P 100 mg/L composite 0.50 0.24 0.77 P 100 mg/L #1 0.26 0.15 0.47 P 100 mg/L #3 0.42 0.20 0.68 P 100 mg/L #5 0.54 0.26 0.83 P 250 mg/L composite 0.36 0.18 0.61 P 250 mg/L #6 0.39 0.20 0.63 P 250 mg/L #9 0.25 0.13 0.43 P 250 mg/L #10 0.45 0.21 0.70 P 400 mg/L composite 0.35 0.21 0.56 P 400 mg/L #6 0.32 0.16 0.52 P 400 mg/L #7 0.21 0.13 0.49 P 400 mg/L #8 0.30 0.15 0.49 236 Table A33. Chlorophyll concentrations for composite samples and individual Douglas-fir seedlings subjected to the 1988 sulphur treatments. Sample Chlorophyll Chlorophyll Total label a b chlorophyll (mg/g) (mg/g) (mg/g) S 1 mg/L composite 0.86 0.39 1.27 S 1 mg/L #1 0.92 0.42 1.34 S 1 mg/L #3 1.04 0.48 1.52 S 1 mg/L #6 1.12 0.51 1.63 S 5 mg/L composite 1.26 0.57 1.83 S 5 mg/L #4 0.80 0.36 1.16 S 5 mg/L #5 0.92 0.45 1.37 S 5 mg/L #9 0.78 0.36 1.14 S 10 mg/L composite 0.90 0.43 1.33 S 10 mg/L #1 0.90 0.42 1.32 S 10 mg/L #3 0.65 0.30 0.95 S 10 mg/L #7 0.86 0.39 1.25 S 25 mg/L composite 0.62 0.29 0.91 S 25 mg/L #2 0.74 0.33 1.09 S 25 mg/L #6 0.55 0.18 0.82 S 25 mg/L #7 0.64 0.29 0.93 S 50 mg/L composite 0.43 0.21 0.70 S 50 mg/L #1 0.28 0.14 0.52 S 50 mg/L #5 0.18 0.30 0.62 S 50 mg/L #9 0.78 0.36 1.16 S 100 mg/L composite 0.46 0.21 0.72 S 100 mg/L #1 0.47 0.22 0.69 S 100 mg/L #7 0.36 0.17 0.59 S 100 mg/L #10 0.49 0.23 0.76 S 250 mg/L composite 0. 60 0.30 0.95 S 250 mg/L #3 0.52 0.25 0.83 S 250 mg/L #4 0.44 0.25 0.81 S 250 mg/L #6 0.44 0.22 0.67 S 400 mg/L composite 0.41 0.19 0.68 S 400 mg/L #4 0.60 0.24 0.85 S 400 mg/L #7 0.44 0.21 0.69 S 400 mg/L #8 0.44 0.22 0.71 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100495/manifest

Comment

Related Items