. [73 . . : ' ‘ V“— r--_4vu Jpn—9. W‘*_-l .-..s,' a? IHIHWI”MINIMUMI'll”)!HIIUHUNHINHIJIHHI 01565 0637 LIBRARY Michigan State University This is to certify that the dissertation entitled PREDISPOSING FACTORS CONTRIBUTING TO RED PINE POCKET MORTALITY presented by Carolyn Joyce Randall has been accepted towards fulfillment of the requirements for Ph.D. degree in Forestry 7M7 F m Major professor Date May 6, 1997 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN RETURN BOX to roman this checkout from your record. TO AVOID FINES return on or More data duo. DATE DUE DATE DUE DATE DUE MSU I. An Affirmativo Mint/Equal Opportunity instituion W m1 PREDISPOSING FACTORS CONTRIBUTING TO RED PINE POCKET MORTALITY BY Carolyn J. Randall A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1997 rL a: :— s ‘ FL . r1. s .. vfo f V A » N\~ C. re . .. r,» . e F‘ s F~« ABSTRACT PREDISPOSING FACTORS CONTRIBUTING TO RED PINE POCKET MORTALITY By Carolyn Joyce Randall A disease known as "red pine pocket mortality" (RPPM) has been reported relatively recently affecting older plantation-grown red pines in the Lake States. A common symptom of RPPM is a large, often circular, area of dead trees (pocket), surrounded by trees showing reduced diameter and height growth, thinning crowns and browning foliage. The causes of this disease are as yet unknown. This study investigated the potential causes of the disease particularly those that are "abiotic" in origin. The condition of red pines at mortality sites was tracked over a five-year period (1991—1995) and the element contents of soil, needle, and root samples from disease versus non— disease locations were determined. Plots were established at pocket decline sites within the Huron-Manistee National Forest, Newago County, Michigan for comparison with check plots of healthy plantation red pines within and outside of the Huron-Manistee National Forest. Analysis of variance was used to identify h.".".- v‘v"- ‘Ar” ,— v..- v’ 1.. Iii a. QC 3. 3. 3.. Nu Y‘ at :. .,\\ _. a. significant differences at the 0.05 level between element contents from areas within disease plots, including the center of the pocket, the edge of the decline, and the outer region where trees appear healthy; and between healthy check-plot stands. Results of disease progress studies indicated that drought may be a contributing factor to the decline; measures of disease intensity tended to increase during drought years and decrease during non—drought years. Increasing gradients in live fine-root (roots < 3 mm in diameter) were found from the center to the edge, to the outside region of disease plots, and finally to the healthy plantation check plots. These increasing gradients suggested that fine roots are dying first before symptoms are seen in tree crowns. Statistical analyses of nutrient contents indicated that manganese (Mn) followed by magnesium (Mg) in soil, needle, and root samples were the nutrient factors most strongly associated with areas of decline and were significantly lower in disease plots versus non-disease plots. Further, increasing gradients in Mn and Mg and corresponding decreasing gradients of metals, particularly aluminum (Al), from diseased to healthy areas, suggested that the pattern of disease spread is associated with nutrient stress. 2. Wk. 5: ACKNOWLEDGEMENTS The author thanks, Dr. John H. Hart, my advisor, for his support and guidance throughout the course of this study, without which this project could not have been completed. Special thanks go to the other members of my committee: Dr. Boyd Ellis, Dr. Andrew Jarosz, Dr. J. James Kielbaso, Dr. Phu Nguyen and to the memory of Dr. Robert P. Scheffer. Each member of the committee made a unique contribution to the study from their area of expertise. Thanks also to numerous undergraduate and graduate students who have assisted me in field and laboratory work. Finally, thanks to my family and friends, especially my husband Timothy Grotjohn, for your love, patience, and support. iv (I) (I) ' f (i'( TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES . . . . . . . . . . . . . . . . INTRODUCTION Chapter I. II. III. LITERATURE REVIEW Abiotic Factors Acidification of Forest Ecosystems . Ca/Al Ratios as Indicators of Stress Inhibition Mechanisms . . . . . . . . Biotic Factors . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . Bibliography . . . . . . . . . . . . . ROOT BIOMASS VOLUMES IN RED PINE (PINUS RESINOSA) MORTALITY POCKETS VERSUS HEALTHY RED PINE STANDS IN MICHIGAN Abstract . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . Methods . . . . . . . . . . . . . . Study Locations . . . . . . . . . . Pit- Sampling Method . . . . . . . . Core- Sampling Method . . . . . . . . Statistical Analysis . . . . . Identification of Pathogenic Fungi Results . . . . . . . . . . . . Pit- Sampled Roots . . . . . . . . . Core- Sampled Roots . . . . . . Isolation of Pathogenic Fungi . . . Discussion . . . . . . . . . . . . . . . Bibliography . . . . . . . . . . . . . . CONDITION AND DISEASE PROGRESS OF RED PINE MORTALITY STANDS IN MICHIGAN Abstract . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . POCKET viii xii 61 62 65 IV. Study Locations Data Collection Tree Ring Analysis . . . . Evaluation of Disease Progress . . . . Statistical Analyses . . . Durban— —Watson Test Statistic . . Spatial Autocorrelation . . . Results . . . . . . . . . . . . . Site Conditions . Tree Health and Annual Growth Disease Progress . Spatial Analyses of Disease Progress Discussion . . . . . . . . . Bibliography . . . . . . . ANALYSES OF SOIL, NEEDLE, AND ROOT SAMPLES FROM RED PINE (PINUS RESINOSA) MORTALITY POCKETS FOR DETERMINATION OF NUTRIENT STRESS Abstract . . . . . . . . . . . . . . . . . Introduction . Methods . Study Locations Soil Sampling Soil Analyses Needle Sampling . . . . . . . . . . . . Needle Analyses . . . . . . . . . . . . Root Sampling Root Analyses . Statistical Analyses Results Soil Nutrient Status Needle Nutrient Status Root Nutrient Status Discussion . Ca/Al Ratios and Risk of. Forest Stress . Results of Soil, Needle, and Root Analyses in Red Pine Mortality Pockets . . . . Bibliography . . . . . . RELATIONSHIPS BETWEEN DROUGHT INDICES, DISEASE 97 99 102 102 104 104 105 106 106 108 108 113 113 119 127 136 136 138 144 INTENSITY, NUTRIENT STUDIES, AND FINE ROOT MORTALITY IN RED PINE (PINUS RESINOSA) MORTALITY POCKETS IN MICHIGAN. Abstract . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . Methods . . . . . . . . . . . Results . Disease Intensity and the Palmer Drought Severity Index . . Soil and Root Nutrient Contents and Root Vitality . . . . . . . vi 146 148 151 154 154 156 Soil and Root Nutrient Factors by Plot . . 161 Multiple Regression Analyses . . . . . . . 168 Discussion . . . . . . . . . . . . . . . . . . 171 Conclusions of Red Pine Pocket Mortality Studies . . . . . . . . . . . . . . . . . 175 Implications for Further Study . . . . . . 178 Bibliography . . . . . . . . . . . . . . . . . 180 APPENDIX: ORIGINAL DATA, SOIL NEEDLE, AND ROOT ELEMENTAL CONCENTRATIONS . . . . . . . . . . . . . . . . . . . 182 vii .f. .C C ..: C 1: LIST OF TABLES Table 2-1. Average root volumes per 1—nfi pit in the top (0 to 30 cm) and bottom (30 to 60 cm) layers of soil as separated by diameter classes and locations (disease center, n=5; disease edge, n=5; disease outside, n=5; Huron-Manistee check, n=4; and Roscommon check, n=2) . . . . . . . . . . . . . . . . . 41 Table 2-2. Average root volumes per 1m x 1m x 60 cm deep pit and proportions of live (L), dead (D), and symptomatic (S) roots for roots 50.3 cm in diameter by location within Huron—Manistee disease plots (center, edge, and outside) and check plots (Huron-Manistee and Roscommon) . . . . . . . . . . . . . . . . . . . . 45 Table 2—3. Average root volumes per 1m x 1m x 60 cm deep pit and proportions of live (L), dead (D), and symptomatic (S) roots for roots >0.3 and 50.6 cm in diameter by location within Huron-Manistee disease plots (center, edge, and outside) and check plots (Huron—Manistee and Roscommon) . . . . . . . . . . . 48 Table 2—4. Average root volumes per 1m x 1m x 60 cm deep pit and proportions of live (L), dead (D), and symptomatic (S) roots for roots >0.6 and 51.0 cm in diameter by location within Huron—Manistee disease plots (center, edge, and outside) and check plots (Huron-Manistee and Roscommon) . . . . . . . . . . . 49 Table 2-5. Average root volumes per 1m x 1m x 60 cm deep pit and proportions of live (L), dead (D), and symptomatic (S) roots for roots > 1.0 cm in diameter by location within Huron-Manistee disease plots (center, edge, and outside) and check plots (Huron-Manistee and Roscommon) . . . . . . . . . . . . . . . . . . . . . 50 Table 2-6. Average volumes (cmfi) per core (10 cm wide X 30 cm deep) of roots from the edge to outer regions of Minnie Lake disease plots, Huron-Manistee National Forest and in check plots at Roscommon County . . . 54 Table 3—1. Numerical rating scale for evaluating disease progress of red pine pocket mortality . . . . . . . 67 viii Table 3—2. Flora present in the pocket openings of red pine mortality plots . . . . . . . . . . . . . . . . . . 72 Table 3-3. Average diameter-at-breast height (dbh), basal area, and number and percent in each crown class (dominant, codominant, intermediate or suppressed) of red pines from disease plots and healthy plots . . . 75 Table 3—4. Mortality in red pine pockets, 1991-1995. . . 77 Table 3-5. Expected I values and variances for three distance categories at disease plots in 1991 . . . . 88 Table 4-1. Mean pH, C.E.C., percent base saturation and concentrations in mg/kg of Al, Ca, Fe, K, Mg, Mn and P from three different soil horizons at disease plots (center, edge, and outer), healthy plantations (HM and Rosc.), and natural stands (Osmun) . . . . . . . . 115 Table 4-2. Mean molar Ca/Al, Mg/Al, K/Al, and Mn/Fe ratios from three different soil horizons at disease plots (center, edge, and outer), healthy plantations (HM and Rosc.), and natural stands (Osmun) . . . . . . . . 117 Table 4-3. Analysis of variance of element concentrations and ratios by site (Minnie Lake or Thin—A-Gain) and by crown level (top, middle, or bottom) for 1-year-old (n=59) and 2—year-old needles (n=56) . . . . . . . 121 Table 4—4. Analysis of variance of element concentrations and ratios by crown level and by health (healthy or unhealthy) for l-year-old and 2—year-old needles at Minnie Lake and Thin-A-Gain pocket decline sites, Huron—Manistee National Forest . . . . . . . . . . 123 Table 4-5. Number of cases (n) and average concentrations (mg/kg) of Al, Ca, K, Mg, Fe, Mn, and Zn in one- and two-year-old needles of healthy (H) and unhealthy (U) red pine at Minnie Lake and Thin—A-Gain, Huron-Manistee National Forest . . . . . . . . . . . . . . . . . . 124 Table 4—6. Number of cases (n) and average molar Ca/Al, K/Al, Mg/Al, and Mn/Fe ratios in 1— —year- -old and 2-year— old needles of healthy and unhealthy red pine, Huron Manistee National Forest . . . . . . . . . 125 Table 4- 7. Mean element concentrations (mg/kg) for disease plots and healthy plantations for both fine (< 3 mm in diameter) and large (>3 and < 5 mm in diameter) red pine roots . . . . . . . . . . . . . . . . . . . . 128 Table 4-8. Molar ratios of Ca/Al, Mg/Al, K/Al, and Mn/Fe for disease plots and healthy plantations for fine (5 3 ix mm in diameter) and large (>3 and 5 5 mm in diameter) red pine roots . . . . . . . . . . . . . . . . . . 131 Table 4-9. Average element concentrations (mg/kg) and molar ratios of roots 5 3 mm in diameter for disease (MLl, ML2, ML3, TG4, TG5) and non-disease plots (MLC, TGC) within the Huron-Manistee National Forest . . . . . 133 Table 4-10. Analysis of variance of element concentrations and ratios in fine roots samples (5 3 mm in diameter) from Huron-Manistee disease plots (n=72) and Huron- Manistee non-disease (check) plots (n=26) . . . . 135 Table 5-1. Correlations between Average Palmer Drought Severity Index (PDSI), average disease intensity (D.I.) and disease intensity by plot (MLl, ML2, and ML3) (observations=5) . . . . . . . . . . . . . . . . . 155 Table 5-2. Correlations between soil pH, root and soil element concentrations/ratios and proportions of live, dead, and symptomatic roots by location (observations = 4) . . . . . . . . . . . . . . . . . . . . . . . . 157 Table 5-3. Correlations between soil pH, and root (R) and soil (S) element concentrations and ratios by Huron- Manistee disease plots and check plots (observations =7) 0 O O O O O O O O O O O O 0 O O O O O O O O O O 162 Table 5—4. Multiple regression equations for predicting root (R) Mn, soil (S) Mn, and root Mg concentra- tions . . . . . . . . . . . . . . . . . . . . . . . 169 APPENDIX: Table 1. Element concentrations (ppm) of spodzolic soils from plantation stands at Huron-Manistee disease plots: Minnie Lake (MLl, ML2, and ML3) and Thin-A-Gain (TG4 and TG5); Huron-Manistee check plots (MLC and TGC); Roscommon check plots (R1 and R2); and Osmun Lake natural stand check plots (01, O2, and 03); late July to early August 1992, 1993, and 1994 . . . . . . . 182 Table 2. Needle element data (ppm) for healthy (asympto- matic for RPPM) and unhealthy (symptomatic for RPPM) red pines for 1- and 2-year—old needles at Minnie Lake (MLl, ML2 and ML3) and Thin-A-Gain (TG4, TG5, TGC), Huron—Manistee National Forest, July—August 1993. . 188 Table 3. Root element concentrations for fine roots (< 3 mm in diameter) and large roots (3 to 5 mm in diameter) from plantation stands at Huron-Manistee disease plots: Minnie Lake (MLl, ML2, and ML3) and Thin-A-Gain (TG4 and T65); Huron-Manistee check plots (MLC and TGC); X Roscommon check plots (R1, R2, R3); and Osmun Lake natural stand check plots (Ol, 02, O3), August, 1992, 1993, and 1994 . . . . . . . . . . . . . . . 198 xi - k K... A h: ~.L ‘ ‘ “q. xue LIST OF FIGURES Figure 2—1. Location of disease plots (MLl, ML2 and ML3) and one check plot (MLC) at the Minnie Lake site, Huron-Manistee National Forest, Newago County, Michigan . . . . . . . . . . . . . . . . . . . . . . 33 Figure 2-2. Location of disease plots (TG4 and TG5) and one check plot (TGC) at the Thin-A-Gain site, Huron- Manistee National Forest, Newago County, Michigan . . . . . . . . . . . . . . . . . . . . . . 33 Figure 3-1. Average annual growth for disease—plot (MLl, n=2; ML2, n=3; and ML3, n=3) and non~disease trees (MLC, n=3) at Minnie Lake . . . . . . . . . . . . . 80 Figure 3-2. Average annual growth for disease-plot (TG4, n=4; T65, n=5) and non-disease trees (TGC; n=3) at Thin—A—Gain . . . . . . . . . . . . . . . . . . . . 81 Figure 3—3. Disease intensity in terms of the average of the rating scale . . . . . . . . . . . . . . . . . . 83 Figure 3—4. The rate of change in disease intensity (AY/AT) . . . . . . . . . . . . . . . . . . . . . . 83 Figure 3-5. Disease ratings at MLl in a) 1991 versus b) 1995 . . . . . . . . . . . . . . . . . . . . . . . 85 Figure 3-6. Disease ratings at ML2 in a) 1991 versus b) 1995 . . . . . . . . . . . . . . . . . . . . . . 86 Figure 3—7. Disease ratings at ML3 in a) 1991 versus b) 1995 . . . . . . . . . . . . . . . . . . . . . . . . 87 Figure 3—8. Moran's I values for three distance categories for disease plots (MLl, ML2, ML3, and TG5) in 1991 . . . . . . . . . . . . . . . . . . . . . . . . 88 Figure 5—1. Linear regression for percentage of live root volume (LIVE) and concentrations of a. root molar Mg (RMG), b. root molar Mn (RMN) and c. soil molar Mn (SMN) by location: disease edge (DE), disease outer (DO), Roscommon non-disease plantations (ROS) and Huron-Manistee non-disease plantations (HM) . . . . 160 xii I = -v"__. . .. .-_ , 7 , Figure 5-2. Linear regression between root Al (RAL) and root Fe (RFE) molar concentrations . . . . . . . . 164 Figure 5—3. Linear regression of soil pH to molar concentrations of a) root Mn (RMN), b) root Mn/Fe (RMNFE), c) root Mg (RMG), and d) soil K/Al (SKAL) . . . . . . . . . . . . . . . . . . . . . . 165 Figure 5-4. Linear regression between molar concentrations of soil Mg and a) root Mg (RMG) b) root Mn (RMN), c) root Zn (RZN), and d) root K/Al by plot (RKAL). . . 167 xiii INTRODUCTION There are about 600,000 hectares of red pine (Pinus resinosa) forest and plantations in the Lake States, approx— imately one-third of which are in Michigan's northern Lower Peninsula (Smith and Hahn, 1986; Hahn and Smith, 1987; Spencer et al., 1988). A substantial proportion of these stands are in plantations. Traditionally, red pine has been managed for revenue from pulp wood and sawtimber. In the last decade a sustained market for red pine utility poles has increased the value of this resource in the northern Lower Peninsula (Grossman and Potter-Witter, 1990). Red pine exhibits little genetic variation and has relatively few insect and disease problems. It is noted for its good growth characteristics and high productivity (Bassett, 1984; Schone et al., 1984). Because of these attributes, it is a desirable timber species. When pest problems do occur on red pine they are usually associated with younger plantations but some occur in older plantations as well. These include pine root collar weevil (Hylobius radicis), redheaded pine sawfly (Neodiprion lecontei), European pine shoot moth (Rhyacionia buoliana), Saratoga spittlebug (Aphrophora saratogensis), white grubs (Phyllophaga, spp.), diplodia tip blight 2 (Sphaeropsis sapinea), Sirococcus shoot blight (Sirococcus conigenus), Armillaria root rot (Armillaria, spp.), and schleroderris canker (Ascocalyx abietina) (Bassett, 1984; Schone, et al., 1984). Adding to this list, a new disease has been identified affecting older plantation-grown red pines. The disease is known as "red pine pocket mortality" (RPPM) and was first reported in 1975 (Klepzig and Carlson, 1988; Raffa and Hall, 1988). The causes of the disease are unknown and it occurs on a variety of soil types in Wisconsin, Michigan, and Illinois. The disease is restricted to mature plantation- grown red pines on marginal farmland. A common symptom of red pine decline is a large, often circular, area of dead trees (pocket), ringed by trees showing reduced diameter and height growth. Dead trees and greatly increased growth of understory plants can be found in the center of the pocket (Klepzig and Carlson, 1988). The rate at which RPPM is spreading and the degree of crop loss has not yet been assessed in Michigan or in other Lake States. Aerial surveys during the summers of 1988 and 1989 suggested that the amount of mortality had increased significantly. An aerial survey by the U.S. Forest Service of the Huron—Manistee National Forest found 440 pockets of dying red pine (USDA Forest Service, 1989). Many of these pockets were likely caused by bark beetles (Ips pini) rather than RPPM, but it was not possible to distinguish between the two without follow-up ground surveys. Since the drought 3 years of 1988 and 1989, however, the number and size of red pine pockets in Michigan have not increased appreciably. RPPM may be viewed as a disease of complex origin, i.e., one that is caused by both primary and secondary factors. This study considers the possibility that the primary stress-inducing factor is abiotic in origin, i.e., one that relates to site conditions such as the soil nutri- ent status. Therefore, the primary hypothesis of this study is that predisposing, abiotic factors present in some red pine plantations are the primary cause of RPPM. The predominant factor under consideration is soil acidification as it affects soil nutrient content, but other possible causes such as deficiencies/toxicities of certain nutrients have been considered as well. The main focus of this study, however, has been on the effect that aluminum, which becomes more soluble in soils of low pH, is having on the uptake and availability of other nutrients, especially calcium, magnesium, and potassium. This study compares site conditions and nutritional factors at pocket decline sites versus healthy red pine plantations. The purpose is to examine the abiotic factors associated with RPPM. Hopeful— ly, this information will allow plantation managers to establish guidelines and management practices for avoiding or preventing the disease. BIBLIOGRAPHY Basset. J.R. 1984. Red Pine Plantation.Management in the Lake States: A Review. The University of Michigan School of Natural Resources, Intensive Forestry Systems Project, Publication No. 3. 27 pp. Grossman, G.H. and K. Potter-Witter. 1990. Changing red pine markets--changing forest management? Northern J. Appl. For. Hahn, J.T. and B.W. Smith. 1987. Minnesota's forest statistics, 1987: an inventory update. USDA For. Serv. Tech. Rep. NC-118 Klepzig, K.D. and J.C. Carlson. 1988. How to identify red pine pocket decline and mortality. USDA For. Serv. NA— GR-19. Raffa, K.F. and D.J. Hall. 1988. Seasonal occurrence of pine root collar weevil, Hylobius radicis Buchanan (Coleoptera:Curculionidae), adults in red pine stands undergoing decline. Great Lakes Entomol. 21:69—74. Schone, J.R., J.R. Basset, B.A. Montgomery, J.A. Witter. 1984. Red Pine Plantation.Management in the Lake States: A Handbook, University of Michigan School of Natural Resources, Intensive Forestry Systems Project, Publication No. 4, 43 pp. Smith, B.W. and J.T. Hahn. 1986. Michigan's forest statistics, 1987: an inventory update. USDA For. Serv. Resour. Bull. NC-107. Spencer, J.S. Jr. et a1. 1988. Wisconsin's fourth forest inventory, 1983. USDA For. Serv. Resour. Bull. NC—107. CHAPTER I LITERATURE REVIEW Abiotic Factors Acidification of Forest Ecosystems Acidification from anthropogenic inputs has been associated with forest decline in Europe (Godbold et al., 1988a; Ulrich, 1989; Ulrich, 1980). In central Europe, acid precipitation is thought to have contributed to soil acid- ity, which in turn may have adverse effects on plant health and microbial activity (Ulrich et al., 1980). Some of the most severely affected areas occur in the mountains of Czechoslovakia where forests appear devastated. This severe decline or "Waldsterben" as it is known in Germany has been dramatic in selected regions on silver fir (Abies alba), Norway spruce (Picea abies), Scots pine (Pinus sylvestris), and European beech (Fagus sylvatica) (Smith, 1990). In the northeastern United States, the possibility that air pollution is a contributing factor leading to red spruce (Picea rubens) decline is being considered among other possible causes (McLaughlin and Kohut, 1992; Smith, 1990). It has been shown that acid precipitation in contact with the foliage of red spruce trees can cause damage depending upon factors such as sulfate concentration and length/ 6 intensity of exposure (Schier and Jensen, 1992). Another potential source of injury from acidic precipitation results from the increase in soil acidity. With increasing soil acidity, aluminum (Al) becomes more soluble and available for plant uptake. Increased acidity/Al toxicity and calcium (Ca) or magnesium (Mg) deficiency due to Al inhibition of uptake are also being considered as causes of red spruce decline (Joslin and Wolfe, 1988). Investigations of red spruce decline suggest that an imbalance of Al and Ca in the fine root environment leads to reduced growth rates and renders mature trees vulnerable to extant secondary diseases and insect pests (Shortle and Smith, 1988). The results of red spruce and other forest decline studies in northeastern America, however, have been incon— clusive (Hertel et al., 1993; Johnson and Fernandez, 1992; Cronan et al., 1989). In one study, 48 Adirondack soil profiles that had been sampled in 1930-32, were resampled in 1984. The researchers found changes in pH and dilute-acid extractable Ca during the 50-year plus interval; the change differed in soil horizons. Moderately acidic organic horizons (pH>4.0) showed a significant reduction in Ca without a pH reduction. The E horizons appeared to lose extra Ca while the B and C horizons showed no evidence of acidification. A partial Ca budget in a mixed hardwood- softwood forest in the study showed that Ca uptake was approximately equal to the loss of Ca from the soil. This suggested that uptake by vegetation rather than acid deposition may be the major cause of Ca loss. While acid rain has increased hydrogen-ion loading and base—cation leaching in the region, acid consuming processes are apparently balanced by acid additions in the B and C hori— zons so that no acidification was observed. The authors concluded that there was no "compelling" evidence that acid deposition has had a major adverse influence on Adirondack soil chemistry through the mid-1980's (Johnson, et al., 1994). Acid precipitation has the potential to acidify soils through wet deposition of sulfates and nitrates that come from fuel-burning emissions. Dry deposition in the form of gases and fine particles also contributes to soil acidifica— tion and may damage forest canopy layers (Mohnen, 1992). Not all forest declines, however, can be explained by the presence of pollutants. Forest declines occur in areas throughout the world where acid deposition does not appear to be a significant factor (Johnson et al., 1991). Forest declines may also be attributed to natural, internal acidi- fication processes. In western North America, for example, excessive nitrogen fixation by red alder (Alnus oregona) stands increased the rate at which soils underneath them were acidified (Johnson et al., 1991). Harvesting practices have also been shown to increase the rate of base-cation depletion on managed forest lands versus natural forests (Johnson and Taylor, 1989). Brand et al. (1986) found that the pH of soils under white spruce (Picea glauca) and red .i . . ....n .3 .5 I «G _ . C . I. .. L. . t S C. l C 3 . . 1 . . mm 3 . r . . E .7. S... 5. F. .d C C I .3 .l jc‘ mm» as. “C Kw ad. n2 .. E E .c ”w C. C .3 mu 1 S E “w . C S C .i c; . a l . . . a» . . v F». «L .Fu . 8 pine (Pinus resinosa) plantations in central Ontario decreased as a result of afforestation. The pH change could be attributed to uptake and immobilization of exchangeable cations from the soil nutrient pool by incorporation into the stand biomass and accumulating forest floor and by the production of acidic decomposition products under spruce and pine litter. Soil acidification, whether from anthropogenic deposition or through natural processes, mobilizes aluminum which reaches levels toxic to fine roots (Ulrich et al., 1980). Death of fine roots then leads to development of symptoms associated with reduced uptake of water and nutri— ents. In the case of Norway spruce stands in Germany, it was observed that the fine-root system had shifted to the topsoil as a consequence of increased acidity in the sub— soil. This shift led to chronic water stress and eventual crown thinning in the affected Norway spruce stands (Ulrich, 1989). Drought may interact as a predisposing stressor to fine roots magnifying the effects of A1 toxicity (Cronan et al., 1989; Ulrich, 1985). Lawrence et a1. (1995) report that during the past six decades, concentrations of root-available Ca (exchangeable and acid extractable forms) in forest floor soils of the northeastern United States have decreased. The depletion of Ca from the root zone can result in acidification of soil and surface water and possibly growth decline and dieback of red spruce. Their research does not support the notion that an! acid deposition or net forest growth (i.e., uptake by vegetation) is the cause of Ca depletion. Instead they propose that Al mobilized in the soil by acid deposition is transported into the forest floor in a reactive form that competitively reduces storage of Ca and its availability for root uptake. If acid deposition alone were the cause of decreases in adsorbed Ca ions then increases in hydrogen (H) ions should be observed. But no significant correlations between Ca and H ions were identified in the study. Instead, inverse correlations between exchangeable Ca and reactive forms of Al were observed in the forest floor. The researchers propose that Al, mobilized in the mineral soil by acid deposition, is transported into the forest floor where it accumulates in a reactive, mostly organically complexed form. The mechanism by which Al is transported up to the forest floor is through biocycling (including Al transported from roots in the mineral soil to roots in the forest floor) and through water movement (i.e. upward movement of water through evapotranspiration, rising water tables, etc.). Aside from natural and anthropogenic causes of soil acidification, tree species vary in their sensitivity to Al. Aluminum toxicity may be present in almost any inorganic soil with a pH less than 5.0-5.5 depending upon the sensitivity of the plant species and the amount of organic material in the soil (Foy, 1974). Steiner et al. (1984), reported that some clones of hybrid poplars (Populus, spp.) 10 were relatively sensitive to Al in solution. Honeylocust (Gleditzia triacanthos) and Norway spruce (Picea abies) also are considered Al—sensitive species (Sucoff et al., 1990 and Godbold et al., 1988b, respectively). Studies of pines, however, have shown that they are relatively tolerant to Al stress compared to other tree species including other conifers (Cronan et al., 1989; Schaedle et al., 1989; Hutchinson et al., 1986). A study of red spruce (Picea rubens) and loblolly pine (Pinus taeda) revealed that surprisingly low concentrations of A1 were phytotoxic to both species affecting their growth, but loblolly pine was less sensitive to Al than red spruce (Raynal, et al., 1990). Controlled experiments and field data in the study indicated that reductions of Ca and Mg in tissues may occur at A1 concentrations that are well below those causing direct injury. Thus, it was postulated that Al may be involved in reductions in growth of pine and spruce through interference with nutrient uptake and translocation (Raynal, et al., 1990). Many authorities have found that a more reliable indicator of acid stress in forests were the ratios of Al to other nutrients, primarily Ca and Mg and sometimes potassium (K). These ratios have been found to correlate better with symptoms of forest decline, rather than the total concentrations of Al in soil solution (Ulrich, 1989, 1985; Godbold et al., 1988a; Murach and Ulrich, 1988; Shortle and Smith, 1988). 11 Ca/Al Ratios as Indicators of Stress The soil Ca/Al ratio has been proposed as one of the key indicators of nutritional problems associated with acid conditions at a site. It has been suggested that a soil molar Ca/Al ratio below one in forested areas indicates a potentially Al-toxic condition (Johnson and Fernandez, 1992; Shortle and Smith, 1988; Ulrich, 1983). Low soil solution Ca/Al ratios have been associated with the decline of beech- spruce forests in Germany (Cronan et al., 1989; Ulrich, 1989). By contrast, Johnson et al. (1994), in a study of declining red spruce in the Adirondacks, found that soil solution Ca/Al ratios were well above the threshold (1:1 molar ratio) proposed as detrimental to Ca uptake, and soil- solution Al concentrations were well below levels thought to inhibit spruce root growth. The results suggest that Al toxicity was not a factor in the decline of high elevation red spruce in the Adirondacks. In sum, there was no evidence that exchangeable Al, Ca, or Mg played a role in red spruce decline on the area studied. Correlations have been found between soil and solution culture Ca/Al ratios and root growth. Godbold et al., (1988b) found that root elongation in seedlings of Picea abies was severely inhibited by Al in nutrient solutions. Their study revealed that the concentration of Al itself was not critical, but that the molar ratio of Ca/Al was critical. Aluminum disturbed the mineral nutrition of the tree seedlings, displacing Mg and Ca in the root cortex in solution culture. Additions of Mg and Ca ameliorated the effects of Al inhibition of root growth. A German study of declining spruce showed that fine-root biomass was highest in the upper soil layers with higher amounts of organic matter and higher Ca/Al ratios in the soil solution. In the lower soil mineral layers, the decrease in fine-root biomass was correlated with decreasing Ca/Al ratios within the fine roots and the soil solution (Murach and Ulrich, 1988). Ratios of Ca/Al have been cited as key indicators of forest stress in soil and in plant tissues. Dewald et a1. (1990) grew red oak (Quercus rubra) seedlings in soils that had been amended to achieve different levels of Al in the soil solution. Increasing Al levels in soil reduced concentrations of Ca and Mg in foliar tissues. Molar ratios of Ca/Al and Mg/Al were reduced significantly in the foliage even though tissue Al concentrations did not vary significantly across soil-solution A1 treatment levels. Although there was not a strong correlation between soil Al concentration and shoot growth, the study did find that 40 percent of the variation in root growth was related to soil solution A1, with the Al treatments significantly reducing total root weight. Murach and Ulrich (1988) report that fine roots (< 2 mm in diameter) distinctly reflect variations in soil chemistry. There was a strong correlation between the Ca/Al ratio gradients in fine roots on beech-spruce stands in Germany and the gradients in soil solution. Ulrich (1989) reported that the Ca/Al ratios of dead fine roots approaches that of the soil solution. Although correlations have been found between foliar and soil solution Ca/Al ratios, it appears that the correlation between fine roots and soil Ca/Al ratios is a more reliable indicator of Al—stress in forest ecosystems. Other authorities have examined wood chemistry in an attempt to link divalent cation trends in stemwood to acidic deposition (Bondietti, et al., 1990; Bondietti, et al., 1989). There are criticisms of using Ca/Al as indicators of Al-stress in wood; the same cation trends have been identified in unpolluted environments. Concentrations of elements in sapwood are not stable indicators of wood chemistry over time (Cronan, 1995). Thus it appears that the most reliable indications of Al-stress in forests are in soil and fine—root tissues. Cronan and Grigal (1995) reviewed the literature on Ca/Al ratios as indicators of stress in forest ecosystems. They identified four measurement end points to use as ecological indicators for identifying thresholds beyond which the risk of forest damage from Al stress and nutrient imbalances increases. The four measurement end points were as follows: 1) A soil base saturation < 15% of effective CEC; 2) A soil solution molar Ca/Al ratio 5 1.0 (for a 50% risk); 3) A fine root tissue Ca/Al ratio 3 0.2 (for a 50% risk); and 4) A foliar tissue Ca/Al molar ratio 5 12.5 (for a 50% risk). These threshold values are recommended as guidelines to identify sites where Al stress is likely to l4 adversely affect tree growth. The authors recognized that these threshold values must be used with certain precautions. Variability arises, for example, from differences in species tolerances, sampling techniques, and chemical analyses. The authors concluded from the literature reviewed that there is a strong relationship between tree growth and Ca/Al ratios in soil solutions and in plant tissues. Many research studies have substantiated the relationship between Ca/Al ratios and forest health; but the question is "what are the biological and chemical mechanisms whereby Al interferes with nutrient uptake?" Inhibition Mechanisms Many studies have indicated that monomeric inorganic forms of aluminum, mainly A1“, are more toxic to plants than are polymeric and organic forms (Balsberg-Pahlsson, 1990; Nakos, 1989; Ulrich, 1989; Joslin and Wolfe, 1988; Schaedle et al., 1989; Lathwell and Grove, 1986). In organic soils, even with low pH values, Al is complexed with organic matter and is unavailable for uptake. In low pH (below 5.0) mineral soils, however, monomeric forms predominate. As previously discussed, the data show that Al may be directly toxic to plant cells, especially those of sensitive species. There is also an indirect toxic effect in an antagonistic/competitive relationship with other cations (Cronan, 1995; Schaedle et al., 1989). Aluminum, because of 15 its higher bonding strength, displaces Ca within the soil and root zone. Once inside the root, Al displaces Ca primarily in the apoplast; also there is some interference in the symplast (Raynal et al., 1990; Ulrich, 1989; Schaedle et al., 1989; Godbold et al., 1988a; Murach and Ulrich, 1988). Visual symptoms of Al toxicity on roots include decreased root elongation, decreased number and length of lateral roots, browning of roots, increases in diameter, and necrosis (McQuattie and Schier, 1992; Schaedle et al., 1989). At the cellular level, A1 has been shown to inhibit mitosis. Dividing and elongating cells in the region of the root tip are especially affected. Aluminum also causes premature cell maturation and senescence. Vacuolization of root cortical cells closer to the meristem and vacuolization of meristematic cells have been observed after exposure to toxic levels of Al (Schaedle et al., 1989). Aluminum is taken up by roots in a non-metabolic process. Once taken up, Al3+ may bind to apoplast surface adsorption sites and interfere with plant uptake or selectivity for nutrient ions such as Ca2+ and Mg2+ (Cronan and Grigal, 1995; Arp and Ouimet, 1986). Bengtsson et al., (1988) found that Al3+ ions can reduce the concentration of Ca, Mg, and P in plants and increase concentrations of K in both nutrient and soil solution. The effect of A13+ on Ca“ uptake was immediate and primarily of a competitive nature, preventing Ca“‘from being adsorbed. The immediate effect of w . I I :V E .J . J. m... y 1 . c . I I . A: .7. D. S . 3 .3 A C a c him a. .1 m C n 1. C 1....“ . . e a E .4. C I at _ g .. I .3 c l r .. u. I. .. t. A . Q». a .. a .. . C T. a r. PC C Ln p». .Q .l L. .3 .. 1 O l I. 1 Av .d .- i am... he... e 16 Al“ exposure on Ca2+ uptake in roots indicates that most Al3+ effects are primarily extracellular; the cation balance at the root surface is disrupted. AlF'ions were found to compete especially with Ca2+ and Mg” for binding sites in the free space of roots, leading to a shortage of these ions at key sites in the plasma membrane. This is proposed as a reason why Al toxicity sometimes gives symptoms similar to Ca2+ deficiency. It is not clearly understood how Al disrupts plant cells once it breaches the exclusion mechanism and interacts with components of the symplasm. One or more of several possible disturbances may occur. Aluminum may inhibit cell division by binding to nucleotides and nucleic acids, and/or it may interfere with energy transfer by binding to ATP, ADP, or membrane-bound ATP phases. Aluminum may also disrupt enzyme systems such as acid phosphates (Cronan, 1995). Haug et a1. (1994) found that Al ions disrupt elements of signal transduction pathways within the cell. They indicated that Al interfered with the action of guanine nucleotide-binding proteins (G proteins) and with an interrelated Ca“/phosphinositide signalling pathway. Other studies have found the Al may interfere with the activities of calmodulin, a Ca-dependent regulatory protein (Sucoff et al., 1990; Haug and Caldwell, 1985). Aluminum binds to calmodulin, inhibiting calmodulin-stimulated, membrane-bound ATPase activity (Haug and Caldwell, 1985). v . . _ L . .. : 3 F. I. re .5 .... ». Mo. .. L _ . l C e C l a I I C C 3 A _. A: C . v. A: ~ A by a. v .nx. ”I. «y if.“ by AV A .v A v l s Hi. a t r. a 3... CC *1 w L. a C hC C a at inn m r. 17 Aluminum has also interferes with the structure and function of the plasma membrane by disrupting the movement of lipids, thus altering membrane permeability (Haug and Caldwell, 1985). The various pathways of interference by A1 are not completely understood and require further study. The end result, however, is that Al toxicity will stress plants, affecting energy transformations, cell division, membrane transport, nutrient accumulation, and various activities regulated by calmodulin. There is also evidence that soil acidification and aluminum toxicity may be detrimental to mycorrhizae. Holopainen (1989) studied the distribution of mycorrhizae in Scots pine (Pinus sylvestris) growing near a pulp mill. There were few mycorrhizal types and dominance of one mycorrhizal species was observed in this severely polluted, acidic site. The mycorrhizae also had important structural changes, which included accumulation of tannin in the cortical and pericycle cells, intercellular hyphae in the cortical cells, and increase of electron dense vacuolar accumulations in the fungal cells. These observed ecological and ultrastructural changes suggest that severe disturbance in the composition and function of pine mycorrhizae can occur in certain acidic environments. McQuattie and Schier (1992) exposed pitch pine (Pinus rigida) seedlings to various levels of Al and ozone. Ozone alone reduced the percentage of mycorrhizal colonization. Aluminum, by contrast, increased colonization at low F1» .3 E .1 .4 arc by rm 4‘ 2. rs .u .r.‘ C. m3 ca v~ re 18 concentrations but reduced it only at the highest level. However, the combination of ozone and Al on the anatomy of mycorrhizae was greater than either alone; apparently, there was a synergistic effect. Ozone had less effects on root anatomy than did Al, which caused deterioration of the root cortex. Greater colonization by mycorrhizae at low Al concentrations than at high A1 levels may result from more inhibitory effects of Al on root growth than on fungal growth. With less root system to infect, a greater percentage of the seedlings' roots became colonized. Cumming and Weinstein (1990) grew pitch pine seedlings with or without the ectomycorrhizal fungus Pisolithus tinctorius and exposed the roots to low levels of Al in sand culture. The nonmycorrhizal seedlings were affected by A1, resulting in decreased growth, increased transpiration rates, decreased efficiency in use of water, increased levels of foliar Al and Na, and reduced concentrations of P in foliage. By comparison, seedlings with P. tinctorius mycorrhizae, after exposure to Al, were not altered in growth, physiological function, and ionic relations. The fungal symbiont evidently modulated ionic relations in the rhizosphere, reducing A1 uptake and Al—P precipitation reactions. This suggests that mycorrhizae may be able to protect pine roots, at least in part, from the adverse effects of Al. 1 is . . E1. r L r . I. 1 1. V a . 1 , . . I E I _ 1 to C M1.» Cw rt C. 1 t +1 C .. 1 -. . . Z 1 C r 3 E S a .1 a. . 1 C C. ...1 .. 1 1: 1 :1 I a ..,._ C .C 3 E .F. T. 11 O 1,: . . C aw he I V. .1... 5 . _ .n. S C C . . C S a m... r 1... .1 S C C. 1.1 C. c C. .i 8 L. S 3: . . .. . . . . . 19 Biotic Factors Trees stressed by nutrient deficiencies/toxicities, weather extremes, and/or other adverse site factors are more susceptible to insect and disease problems. Toxic effects of Al, whether direct or indirect, may weaken trees predisposing them to attack by insects and fungi. Abebe and Hart (1990) found that the incidence of Cytospora canker on hybrid poplar increases with increases in Al levels in the soil. There are only a few articles written specifically about the role of insects and fungi in red pine pocket mortality. One was a comparison of declining and of healthy red pine plantations in Wisconsin (Klepzig, et al., 1991). There was a significant difference in the abundance of five insect species in declining stands compared to healthy stands of Pinus resinosa. Insects involved were the root collar weevil (Hylobius radicis), the pales weevil (Hylobius pales), the pitch-eating weevil (Pachylobius picivorus), and the turpentine beetle (Dendroctonus valens and Hylastes porculus). These root- and lower stem-infesting insects consistently carried two fungal species--Leptographium terebrantis and Leptographium procerum. The researchers proposed that Leptographium is transmitted to red pine roots by the five insect vectors; root infections lead to expression of above-ground symptoms. Leptographium, spp., which may spread via root graphs or contacts and/or through soil, are proposed to cause below—ground root mortality ll . . .1 ..._ Z“ . _ J :1 I. I. r . 1 C C. 1 1 ...1 .1 1i] .3 (5 f w C» .19.. a TA ‘4 n u to :1. 71‘ AV A\V h «. RAMs rhwv . Illllrirli n.“ a .61. an I C. a . I C 1. C r 1 i v C I e C r» .. .. u. M. .. C. 6 .hr. Tc .1 l T .1 .5 a «L. AL ”N A c 4 C Ci V. AC s 1 £1 a: 20 which precedes the above—ground symptoms of reduced radial growth, thin crown structure, and infestation by the pine engraver (Ips pini). As trees die they become suitable habitats for H. pales and P. picivorus and other members of the root- and lower-stem—feeding insect group. These insects increase in number and introduce additional inoculum; the result is expansion of the site of infection i.e. pocket dieback. Final tree mortality is attributed to Ips pini and its fungal associate Qphiostoma ips (Klepzig, Raffa, and Smalley, 1991). In further analysis of the above scenario, Klepzig et a1. (1995) examined the possibility that D. valens and H. porculus can vector the associated fungi to red pine. In this study, D. valens and H. porculus were caged with wounded red pine roots. D. valens collected from the field transmitted L. terebrantis, L. procerum and O. ips. with 45%, 30% and 5% success, respectively. H. porculus collected from the field transmitted L. terebrantis, L. procerum and O. ips with 55%, 40% and 5% success, respectively. Only two of the control roots, which were mechanically wounded and not exposed to insect vectors, contained Leptographium, spp. The researchers concluded that D. valens and H. porculus can vector Leptographium fungi to red pine trees and that these organisms probably are involved in red pine decline disease. Leptographium terebrantis, vectored by black turpentine beetles (Dendroctonus terebrans), has been 1 u . AU .1‘. law.‘ 1 V . Q» N; . 1b .. 1 ¢ C . . w .1 l. y 1 1 11.11 W141, C. V. .1 C 11.31 m a .. a. 1 _: 1.11 .i. nus i . Qv - . p a FL Q .1 n 1 . a «d N a Q. .ll\ a 9N O Y . s s s 5 5 & a h . , v i‘ $ 5 «\v F» “V. .1 .\ A+V\ QC {.1 associated with mortality of Japanese black pine (Pinus thunbergiana) and Scots pines (P. sylvestris) on Cape Cod, Massachusetts (Highley and Tattar, 1985). At this site, blue stain caused by L. terebrantis developed beneath larval galleries of the beetle in the lower bole and buttress roots; this was followed by foliar symptoms and attack by bark beetles (Ips spp.) and other staining fungi. The pathogenicity of Leptographium spp. and their Qphiostoma teleomorphs are questionable. Alexander, et a1. (1988) reported that L. procerum, the causal agent of procerum root disease of eastern and western white pine (Pinus strobus and P. monticola), affects mainly young trees (< 20 years old) in plantations and is associated with poor site conditions especially poorly drained soils. Harrington (1993) reports that many Ophiostoma/Leptographium species appear to be weak pathogens, perhaps contributing in a minor way to the disease syndrome, whether of abiotic or biotic origin. Harrington also states that the role of Leptographium in beetle-attacked trees remains unclear and that only two species, 0. ulmi and L. wageneri have been studied extensively and found to be causal agents of major plant diseases: Dutch elm disease and black stain root rot on conifers. Summary It is evident from the literature that acidification of forest ecosystems may occur as a result of both natural and .mu 1.: .C AC 22 anthropogenic processes. One of the primary concerns with soil acidification is the increase in soluble/exchangeable forms of Al (mainly Al“) which can be toxic to plants either directly or indirectly. The biochemical pathways and mechanisms of A1 toxicity to plants are not completely understood. The end result, however, is that the plant/tree may become weakened and prone to secondary insect and disease problems. The following chapters of this dissertation will examine whether or not aluminum toxicity is the primary cause of red pine pocket mortality (RPPM). Emphasis will be placed on the indirect toxic effects of aluminum as a deterent to the uptake of other nutrients since this is a more likely scenario with Al—tolerant red pines. Ca/Al ratios and other measurement end points will be the main parameters used to identify unfavorable site conditions in pockets of red pine mortality. 1: 1: EC 1: 1: 1H» 23 BIBLIOGRAPHY Abebe, G. and J.H. Hart. 1990. The relationship of site factors in the incidence of Cytospora and Septoria cankers and poplar and willow borer in hybrid poplar plantations. USDA Forest Service general technical report, NCFES (140), Aspen Symposium, pp. 163-171. Arp, P.A. and R. Ouimet. 1986. Uptake of Al, Ca, P in black spruce seedlings, effect of organic vs. inorganic Al. Water, Air, and Soil Pollution 31: pp. 367-375. Alexander, S.A., W.E. Horner and K.J. Lewis. 1988. Leptographium procerum as a pathogen of pines. In: Leptographium Root Disease on Conifers, T.C. Harrington and F.W. Cobb, Jr, Eds., APS Press, pp. 97—112. Balsberg-Pahlsson, A. 1990. Influence of aluminum on biomass nutrients, soluble carbohydrate and phenols in beech (Fagus sylvatica). Physiol. Plant. 78, pp. 79- 84. Bengtsson B., H. Asp, P. Jensen and D. Berggren. 1988. Influence of aluminum on phosphate and calcium uptake in beech (Fagus sylvatica) grown in nutrient solution and soil solution. Physiol. Plant. 74, pp. 299-305. Bondietti, E.A., C.F. Baes III, and 8.8. McLaughlin. 1989. Radial trends in cation ratios in tree rings as indicators of atmospheric deposition on forests. Can. J. For. Res. 19, 586-594. Bondietti, E.A., N. Momoshima, W.C. Shortle, and K.T. Smith. 1990. A historical perspective on divalent cation trends in red spruce stemwood and the hypothetical relationship to acidic deposition. Can. J. For. Res. 20, pp. 1850-1858. Brand D.G., P. Kehoe and M. Connors. 1986. Coniferous afforestation leads to soil acidification in central Ontario. Can. J. For. Res. 16, pp. 1389-1391. Cronan, C.S., R. April, R.J. Bartlett, P.R. Bloom, C.T. Driscoll, S.A. Gherini, G.S. Henderson, J.D. Joslin, J.M. Kelly, R.M. Newton, R.A. Parnell, H.H. Patterson, '- 4 ‘ 1 . “1 hm » 1 C i r... .. r1 . a u. T rv C “u . 1 r 1 «(0 P0 «Wu «G in «.1 ”H1 J . 1d rs . . .41. £1 H.H.. 24 D.J. Raynal, M. Schaedle, C.L. Schofield, E.I. Sucoff, H.B. Tepper, F.C. Thornton. 1989. Aluminum Toxicity in Forests Exposed to Acidic Deposition: The ALBIOS Results. Water, Air, and Soil Pollution 48: 181-192. Cronan, C.S. and D.F. Grigal. 1995. Use of calcium/ aluminum ratios as indicators of stress in Forest Ecosystems. J. Environ. Qual. 24, pp. 209-226. Cumming, J.R. and L.H. Weinstein. 1990. Aluminum- mycorrhizal interactions in the physiology of pitch pine seedlings. Plant and Soil 125, pp. 7—18 DeWald, L.E., E.I. Sucoff, T. Ohno, and C.A. Buschena. 1990. Response of northern red oak (Quercus rubra) seedlings to soil solution aluminum. Can. J. For. Res. 20, pp. 331-336. Foy, C.D. 1974. Effects of aluminum on plant growth. In: The Plant Root and its Environment, F.W. Carson, ed., Univ. Va. Press, Charlottesville. pp. 602—642. Godbold, D.L., E. Fritz and A. Huttermann. 1988a. Aluminum toxicity and forest decline. Proc. Natl. Acad. Sci. 85: 3888—3892. Godbold, D.L., K. Dictus and A. Hutterman. 1988b. Influence of aluminum and nitrate on root growth and mineral nutrition of Norway spruce (Picea abies) seedlings. Can. J. For. Res. 18, pp. 1167-1171. Harrington, T.C. 1993. Diseases of conifers caused by species of Ophiostoma and Leptographium. In: Ceratocystis and Ophiostoma, Taxonomy, Ecology and Pathogenicity, M.J. Wingfield, K.A. Seifert, and J.F. Webber, eds., APS Press, pp. 161—172. Haug, A., B. Shi, and V. Vitorello. 1994. Aluminum interaction with phosphoinositide-associated signal transduction. Arch Toxicol 68, pp. 1-7. Haug, A.R. and C.R. Caldwell. 1985. Aluminum toxicity in plants: The role of the root plasma membrane and calmodulin. In: Frontiers of.Membrane Research in Agriculture, Beltsville Symposia in Agricultural Research 9, Beltsville Agric. Res. Ctr (BARC), Beltsville, MD, pp. 359-381. Hertel, G.D., C. Eager, S.A. Medlarz and M.W. McFadden. 1993. The effects of acidic deposition and ozone on forest tree species in the eastern United States: Results from the Forest Response Program. In: Forest Decline in the Atlantic and Pacific Regions. R.F. v. l-‘k.’ Lino ‘éu1 (.- 4 ( JC‘J «4‘ Ta v v.“- 25 Huettl, D. Mueller-Dombois, Eds., Springer—Verlag, New York, pp. 53-65. Highley, L. and T.A. Tattar. 1985. Leptographium terebrantis and black turpentine beetles associated with blue stain and mortality of black and Scots pines on Cape Cod, Massachusetts. Plant Disease 69, No. 6, pp. 528-530. Holopainen, T. 1989. Ecological and ultrastructural responses of Scots pine mycorrhizas to industrial pollution. Agriculture, Ecosystems and Environment, 28, pp. 185-189. Hutchinson T.C., L. Bozic and G. Munoz—Vega. 1986. Responses of five species of conifer seedlings to aluminum stress. water, Air and Soil Pollution 31, pp. 283-294. Johnson, A.H., T.N. Schwartzman, J.J. Battles, R. Miller, E.K. Miller, A.J. Friedland, and D.R. Vann. 1994. Acid ran and soils of the Adirondacks. II. Evaluation of calcium and aluminum as causes of red spruce decline at Whiteface Mountain, New York. Can. J. For. Res. 24. pp. 654. Johnson, D.W. and G.E. Taylor. 1989. Role of air pollution in forest decline in eastern North America. Water, Air and Soil Pollution 48, pp. 21-43. Johnson, D.W., M.S. Cresser, S.I. Nilsson, J. Turner, B. Ulrich, D. Binkley and D.W. Cole. 1991. Soil changes in forest ecosystems: evidence for and probable causes. Proceedings of the Royal Society of Edinburgh, 97b, pp. 81-116. Johnson, D.W. and I.J. Fernandez. 1992. Soil-mediated effects of atmospheric deposition on eastern U.S. spruce-fir forests. In: Ecology and Decline of Red Spruce in the Eastern United States. C. Eager and M.B. Adams, Eds., Springer-Verlag, New York, pp. 235-270. Johnson, D.W., A.H. Andersen and T.G. Siccama. 1994. Acid rain and soils of the Adirondacks I. Changes in pH & available calcium 1930—1984. Can J. For. Res. 24. pp. 39-45. Joslin, J.D. and M.H. Wolfe. 1988. Responses of red spruce seedlings to changes in soil aluminum in six amended forest soil horizons. Can. J. For. Res. 18: 1614:1623. v~‘ .-~.¢~ .., I. ”U. a A. V. _.A. C O. I 1.... .1 . J . J U... u...” 1.1 .2.“ 15 ‘I 4 Pa \HC 26 Klepzig, K.D., K.F. Raffa and E.B. Smalley. 1991. Association of an insect-fungal complex with red pine decline in Wisconsin. For. Sci. 37:1119-1139. Klepzig, K.D., E.B. Smalley and K.F. Raffa. 1995. Dendroctonus valens and Hylastes porculus (Coleoptera: Scolytidae): vectors of pathogenic fungi (Ophiostomatales) associated with red pine decline disease. The Great Lakes Entomologist, 28 No. 1, pp. 81-87. Lathwell D.J. and T.L. Grove. 1986. Soil-plant relationships in the tropics. Ann. Rev. Ecol. Syst. 17, pp. 1—16. Lawrence, G.E., David, M.B and W.C. Shortle. 1995. A new mechanism for calcium loss in forest-floor soils. Nature 378, 9 Nov., pp 162-165. McLaughlin 8.8. and R.J. Kohut. 1992. The effects of atmospheric deposition and ozone on carbon allocation and associated physiological processes in red spruce. In: Ecology and Decline of Red Spruce in the Eastern United States. C. Eager and M.B. Adams, Eds., Springer-Verlag, New York, pp. 338-382. McQuattie, C.J. and G.A. Schier. 1992. Effect of ozone and aluminum on pitch pine (Pinus rigida) seedlings: anatomy of mycorrhizae. Can. J. for. Res. 22, pp. 1901—1916. Mohnen, V.A. 1992. Atmospheric deposition and pollutant exposure of eastern U.S. forests. In: Ecology and Decline of Red Spruce in the Eastern United States. C. Eager and M.B. Adams, Eds., Springer—Verlag, New York, pp. 64-124. Murach, D. and B. Ulrich. 1988. Destabilization of forest ecosystems by acid deposition. GeoJournal 172: 253- 260. Nakos, G. 1989. Concentration of exchangeable and soluble aluminium in acid forest soils in Greece. Forest Ecology and Management, 26, pp. 141-149. Raynal, D.J., J.D. Joslin, F.C. Thornton, M. Schaedle and G.S. Henderson. 1990. Sensitivity of tree seedlings to aluminum. III. Red spruce and loblolly pine. Journal of Environmental Quality, Madison, Wisconsin: American Society of Agronomy, Apr/June 1990 v. 19(2). pp. 180—187. Cc m a 27 Schaedle, M., F.C. Thornton, D.J. Raynal and H.B. Tepper. 1989. Response of tree seedlings to aluminum. Tree Physiology 5, pp. 337-356. Schier, G.A. and K.F. Jenson. 1992. Atmospheric deposition effects on foliar injury and foliar leaching in red spruce. In: Ecology and Decline of Red Spruce in the Eastern united States. C. Eager and M.B. Adams, Eds., Springer—Verlag, New York, pp. 271-294. Shortle, W.C. and K.T. Smith. 1988. Aluminum-induced calcium deficiency syndrome in declining red spruce. Science: 240:1017-1018. Smith, W. H. 1990. Forest dieback/decline: A regional response to excessive air pollution exposure. In: Air Pollution and Forests: Interaction Between Air Contaminants and Forest Ecosystems, Springer-Verlag, New York, pp. 501-524. Steiner, K.C., J.R. Barbour and L.H. McCormick. 1984. Response of Populus hybrids to aluminum toxicity. For. Sci. 30: 404-410. Sucoff, E., F.C. Thornton, and J.D. Joslin. 1990. Sensitivity of tree seedlings to aluminum: I. Honeylocust. J. Environ. Qual. 19, pp. 163-171. Thornton, F.C., M. Schaedle, D.J. Raynal and C. Zipperer. 1986. Effects of aluminum on honeylocust (Gleditsia triacanthor L.) seedling in solution culture. J. Expt. Bot. 37, pp. 775-785. Ulrich, B., R. Mayer and P.K. Khanna. 1980. Chemical changes due to acid precipitation in loess-derived soil in central Europe. Soil Sci. 130: 193-199. Ulrich, B. 1983. Soil acidity and its relation to acid deposition. In: Effects of Accumulation of Air Pollutants in Ecosystems, Panketh J. (eds). Reidel. Ulrich, B. 1985. Interaction of indirect and direct effects of air pollutants in forests. In: Air Pollution and Plants, Proc. of the 2nd Euro. Conf. on Chemistry and the Environment, Clement Troyanowsky, Ed., pp. 149-180. Ulrich, B. 1989. Effects of acidic precipitation on forest ecosystems in Europe. In: Acid Precipitation, Volume 2, Biological and Ecological Effects, D.C. Adriano and A.H. Johnson, Eds., Springer-Verlag, New York, pp. 189- -272. CHAPTER I I Root Biomass Volumes in Red Pine (Pinus resinosa) Mortality Pockets Versus Healthy Red Pine Stands in Michigan ABSTRACT Roots were collected from red pine mortality pockets by two different methods: pit-sampling and core-sampling. The pit—sampling method revealed that percentages and actual volumes of live roots for all root—size classes increase starting from the center to the edge of pockets, and finally to the outside "healthy—looking" regions of disease plots. Pocket centers and edges were characterized by relatively low root volumes and high proportions of dead roots. Relatively high percentages of symptomatic root volumes were also found at pocket edges, where trees were declining. Root volumes within disease plots were largest and healthiest in the area outside the pockets where trees appeared healthy. The core-sampling method enabled quantification of very fine roots (< 1 mm) in diameter that were not discernable using the pit-sampling method. However, the core—sampling method failed to reveal the same trends as the pit—sampling method, probably owing to the fact that roots from competing vegetation interfered with sampling only red pine roots. 28 29 Root volumes collected by the pit-sampling method revealed that disease centers and edges, where red pines were dead or dying, had significantly lower volumes of root biomass than did areas of healthy red pines. These healthy— looking regions included areas within the Huron—Manistee disease plots but outside the edge of decline, healthy plantation plots within the Huron-Manistee National forest, and healthy plantation plots at the Houghton Lake State Forest in Roscommon County, Michigan. There were no statistically significant differences between volumes of roots from outer "healthy" regions of disease plots and the check plots at Huron-Manistee and Roscommon. However, smaller volumes of live roots, and more symptomatic and dead roots in the outside healthy—looking regions of disease plots compared to non-disease plantations suggests that the disease is spreading below ground and that trees are dying from the bottom up rather than from the top down. Roots, especially in the fine, 30.3 cm in diameter category, die first, eventually leading to symptoms of red pine pocket mortality (RPPM) in the crown. However, our attempts at isolating a fungal agent which may be responsible for root mortality in red pine pockets were unsuccessful. INTRODUCTION Red Pine Pocket Mortality (RPPM) is a relatively recent disease affecting 30- to 50-year-old plantation—grown red pine in the Lake States. The disease is characterized by QC 30 declining trees surrounding a large, often-circular, area of dead trees (Klepzig and Carlson, 1988; Raffa and Hall, 1988). Symptoms of the declining red pine include thinning crowns, browning foliage, stunted growth in needles, and reduced diameter and height growth. Many insects and fungi have been found associated with the disease in Michigan but none have been consistently present in all pockets. The causes of this disease are uncertain. A study in Wisconsin identified a complex vector and pathogenic fungi association that may lead to the decline (Klepzig, Raffa, and Smalley, 1991). Investigators in the study found a significant difference in the abundance of five insect species in declining stands compared to healthy Pinus resinosa stands. These root- and lower stem—infesting insects consistently carried two fungal species-— Leptographium terebrantis and Leptographium procerum. It was suggested that Leptographium is transmitted to the red pine root system through the five insect vectors resulting in expression of above-ground symptoms. Leptographium, spp., which may spread via root graphs or contacts and/or through soil, are proposed to cause below—ground root mor- tality which precedes the above-ground symptoms of reduced radial growth, thin crown structure, and infestation by the pine engraver beetle (Ips pini). Our investigations of RPPM in Michigan have focused on the possible abiotic factors involved in the decline rather than on the role of biotic agents. Of particular interest 31 has been the possibility that aluminum (Al) toxicity may be the primary cause of fine-root death. This theory of Al— induced root mortality was derived mainly from the literature pertaining to pollution, soil acidification, and forest decline. Soil acidification has been cited as a cause of forest growth decline by creating nutrient deficiencies through leaching of base cations (especially Ca“) and also by increasing the concentration of aluminum (A1“) in solution to toxic levels (Johnson and Taylor, 1989). In the case of RPPM, the soils under red pine may have acidified as a result of pine litter accumulating on the forest floor as the trees mature. Pine needle accumulation over time could eventually affect soil pH as needles decompose and release organic acids into the soil (Johnson and Fernandez, 1992). This time—dependent factor may explain why symptoms of RPPM are not seen until the trees are mature. Whether the causes are biotic or abiotic in origin, it has been presumed that root mortality leads to the expression of above—ground symptoms of RPPM. This study compares volumes of root biomass from declining stands and healthy stands in terms of the proportions of live and dead roots within a plot. METHODS Study Locations Five RPPM sites as identified by the U.S. Forest Service were selected for this study. Research plots were .r » ~ . ., I . . 2. rd .1. ..c In C» i; it FL T» r. «u .p . F.. t .. C. .. AC .C . v v t . I i . & w Mb “a 1;? r... «C 5- v .MV. . M- .. AV .fiu .0“ “M QVO by u... C. “W NJ. Win r u 1 PM} (I... 1 ¢ fic M «flu. aw“ FLA a” L‘ h H. ‘h o. ‘ m c c an‘ . . * buy r» .. . a A ”U FLA established within the area of decline and for an area outside the pocket. Three disease plots (MLl, ML2, and ML3) are located at Minnie Lake at a latitude and longitude of 43°38' and 85°54' (township and range: T 14 N, R 13 W) and two disease plots (TG4 and T65) are located at an area known as Thin—A-Gain at a latitude and longitude of 43°35' and 83041' (township and range: T 14 N, R 12 W) (Figures 2—1 and 2-2). In addition to disease plots there are two plantation check plots of healthy red pine located near the disease plots: one at Minnie Lake (MLC) and one at Thin—A-Gain (TGC). For comparison with plantation-grown red pine located in a forest other than Huron-Manistee, two additional check plots of healthy red pine were established at the Houghton Lake State Forest in Roscommon County, Michigan (R1 and R2) at a latitude and longitude of 44015' and 84040' (township and range: T 21 N, R 3 W). The Roscommon plots are about 115 km northeast of the disease plots. In sum, there were nine plots used in this study: five disease plots; two plantation check plots located in the same forest as the disease plots; and two plantation check plots located in another forest. Plantation check plots are similar to the disease plots in that they consist of mature, even—aged red pines that had been densely planted. no. FL. . \~.\ 3 3 1 """""" 1...; Saw... I Is -~-_.,v ....... I -nmwi . 3': <2 . g; \3 ‘ Jf K i J 5? - a r 22 Em“,MLERD 5 \M4jgyaRDhlemm-mmmmmmnWM- mw«tmm 4.___,_'Y__ _____ > ....................... i ~-»—w- ... -------- 5_ 9:) g, i d: ‘ z . i" 3 1‘3: E E 45 D- : 5 a QC Figure 2—1. Location of disease plots (MLl, ML2 and ML3) and one check plot (MLC) at the Minnie Lake site, Huron-Manistee National Forest, Newago County, Michi- gan. 1 . MONRggno M. "WNW , a O . i n5 0 5 3 0 nu ,' 3 TGC f“ g 52 MILE no E2 MILE no 5" E' ............................... WW E; “>4 E: 2. g a ("in pine plantation Z 21 MILE RD .1‘ MILE RD i i Figure 2-2. check plot Location of disease plots (TGC) at the Thin—A—Gain site, (TG4 and T65) Huron— and one Manistee National Forest, Newago County, Michigan 34 Pit-Sampling Method Roots were collected in August, 1992 by digging 1-m2 pits in disease plots at three locations within the plot: in the center of the pocket where there were many dead standing trees; at the edge of the pocket where trees were declining; and outside of the decline where the trees appeared healthy. The l-nfi pits were divided into four 0.5 x 0.5 m sections and samples were taken from two depths: 0 to 30 cm deep, and 30 to 60 cm deep. Thus eight bags of root samples were collected from one 1-nfi pit (4 sections x 2 depths). In total, fifteen pits were dug (three at each of the five plots) in disease plots. In addition, two l-nfi pits were taken at each of the two check plots at Huron—Manistee. Samples were taken in August, 1994 from the two check plots at Roscommon County. At that time, only 0.5—n8 pits were dug and bagged—samples removed from the two depths (two bags of roots per pit); four 0.5-n9 pits were dug at each of the two check plots so that the area removed was the same size as a l-nfi pit. The pits were dug randomly at locations throughout the check plots because there were no features of center, edge, or outside area of a pocket within these plots. The roots collected at both check plots and disease plots were extracted by sifting the soil through a mesh screen and separating the red pine roots from the roots of other plants growing in the area. Once collected, the roots 35 were washed by rinsing with water for 30 minutes and dipped in a 10% Chlorox solution for 2 minutes. The length and diameter of each root was recorded and they were classified as to live, dead, or symptomatic. Roots were classified as "symptomatic" if they were partially dead or dying, and/or if there was any root staining below the epidermis. Core-Sampling Method ‘Core-sampled roots were collected at three disease plots located in the Minnie Lake region of the Huron- Manistee National Forest and from the two healthy plots at Roscommon in August, 1994 (MLl, ML2, ML3, R1, and R2). In disease plots, the root systems of red pines were sampled along transects. Three trees were sampled per disease plot (9 trees total). The first tree sampled was located at the edge of the pocket. From the edge, the transect extended out into the healthier regions of the disease plots from where the second and third trees were sampled. At Roscommon, because there were no features of center, edge, or outside area of a pocket, the samples were collected randomly from three different trees within each plot for a total of 6 trees. At both Huron—Manistee and Roscommon, five cores, at a distance of 60 cm from the main stem, were taken around each tree. The sampling depth was 30 cm and the cylindrical cores were 10 cm wide. The soil was rinsed from the samples using the hydropneumatic elutriation method (Srivastava, et al., 36 1981). The roots were placed in Whorl—packLm bags with 15- 20% methanol solution and stored at 4Tb The roots were stained blue with 1.0 cc of Malachite-blue dye. Fine-root (< 1.95 mm in diameter) biomass was determined using the robotic camera method (Smucker, 1989). In this method, images of washed roots are stored on standard 1/2 inch VHS format video tapes. Each tape contains images from numerous root samples. Images from the tape were then processed by an image-processing computer to quantify the root length and width. Roots greater than 1.95 mm in diameter were too large to be measured accurately by the image-processing system. These roots had to be pulled out of the sample and measured and recorded by hand. Statistical Analyses Analysis of variance was used to test the hypothesis that root volumes from disease and healthy locations are significantly different from each other. The model used to test hypotheses of root volumes (i.e., V=LnR?) differing by locations and soil depths (Table 2-1) is as follows: Yijk=u+ai+fij+afljj+€ijk where variable Y is the root volume concentration per pit (k) associated with the ith level of factor a (location) and j” level of factor 5 (soil depth), u is the overall mean of all observations, ai is the effect of iLh level of the location factor, Bjis the effect of the j“ level of the soil depth factor, a8” is the effect of the interaction I'll". (K S 37 between the location and soil depth factors and efiflk is the experimental error associated with the kLh experimental unit (l-nf pit) for the levels of location and soil depth factors. Four different root-size classes were tested in separate ANOVA trials. These four variables belong to root- size categories: 3 0.3 cm in diameter; >0.3 and 30.6 cm; >0.6 and 51.0 cm; and, >1.0 cm. There were five levels included in the location factor: the center of Huron- Manistee disease plots; the edge of disease plots; outside the edge of the decline within disease plots; healthy plots at Huron-Manistee; and healthy plots at Roscommon County. Two levels are included in the soil depth factor from which roots were collected: the top 0-30 cm, and bottom 30 to 60 cm. ANOVA was also used to test differences between root volumes by location and condition (i.e. live, dead, or symptomatic) (Tables 2~2 through 2-5). The model used to test the hypothesis that root volumes differ by location and condition is the same except that fljis the effect of the jth condition factor, and up” is the interaction between location and condition (i.e. volumes of live, dead or symptomatic roots) factors. If the soil depth factor or the condition factor was identified as significant, additional ANOVA's were run by selecting the subpopulation of interest. For example, root volumes in the top level of soil depth were compared to each 38 other and in separate ANOVA trials, root volumes in the bottom level were compared. Similarly, for the condition factor, all root volumes in the live category were compared to each other, dead volumes were compared to other dead root volumes, and symptomatic root volumes to other symptomatic root volumes. These ANOVA's follow the model Yik=u+ai+eik where ai is the effect of the location factor. The effect of the location, soil depth and/or condition factors are considered fixed because they are the only levels of interest in this investigation. If the experiment were to be repeated the same treatments would be included and randomness would be inherent in the replications (n) i.e., the number of soil pits sampled. If, however, the ANOVA had been run using plots (i.e. MLl, ML2, ML3, etc.) as the levels of a factor the effect would be considered random because the disease pockets and healthy plantation plots are considered subsamples of the larger population of all potential disease pockets and healthy plantations. Under the random effects model the objective is to make inferences to the larger population (i.e. disease pockets and healthy red pine plantations in other locations). However in this investigation the effects are considered fixed and any inferences made apply only to the disease and healthy areas included in this study. Analysis of core-sampled roots also follow the model, ‘Yik=u+ai+(eik except that a1, the location factor, consists of four levels: three locations along transects within Minnie ’9' 5“. A" r tn“ 1 A. ‘VH AA? .Vv“ {-f ,x 0’ 39 Lake disease plots which start from trees at the edge of pockets, to trees mid-range from the edge to outer region of the plots, to trees located at outer "border" regions of the plots. The fourth location is Roscommon, where trees were sampled randomly within the plots. In this analysis, the sampling unit k, is the 30 cm deep x 10 cm wide cylindrical core rather than the l-nfi pit. The effect of the location factor in this model is also considered fixed because statistical tests were run between samples from within plot locations rather than between individual plots. If the experiment were to be repeated the same levels of the factor would be included. There were no factors of soil depth or condition included in core—sampled roots. All sample distributions for root volumes were transformed by taking the square root in order to normalize the distribution. However, for clarity, means and other descriptive statistics reported are non-transformed values. Means were separated at the 0.05 significance level using the Tukey's highly significant difference (HSD) procedure (Systat for Windows, Version 5, 1992). Identification of Pathogenic Fungi The live or symptomatic roots collected by the pit- sampling method were cultured using selective and non-selec- tive media for Leptographium, spp. Potato dextrose agar (PDA) was used as non—selective media which was acidified with lactic acid. Selective media consisted of PDA with 100 _ I I.C.|J. i. r. r: a. war .(x.\ «it? FIR vi. .2 . . . .. . E r... c x i e Z a C Q -H . . .H C C. u. .1 .1. C . n... .L .1 S a. .l w. my. . c Rm . i t C a. A . P O a 1.. .3 .3 .3 .t. S .g. r. .1 .n . .3 3 a I Q. ... 1. r t G a c t C C. C. c. 40 ppm cycloheximide and acidified with lactic acid (Harrington, 1981). Roughly 1000 roots were cultured; half on selective and half on non-selective media. RESULTS Pit-Sampled Roots Table 2-1 lists the average volume per l—ufi pit in cm3 of roots collected from locations within disease plots (center, edge, and outside) and from healthy areas (Huron- Manistee and Roscommon) from two levels of soil depth (0 to 30 cm, and 30 to 60 cm). In the pit-sampling method, it was difficult to distinguish red pine roots less than 1 mm in diameter. Therefore, the smallest size category represented is between 1 and 3 mm in diameter. Table 2-1 breaks down the volumes of roots collected according to four diameter size categories with the largest size category being greater than 1 cm in diameter. The largest volume of red pine roots was found in the top 30 cm of soil for all locations (disease and non- disease) and for all root-size categories. Probabilities that there were no significant differences between root volumes by soil depth were p<0.001 for the 50.3 cm and >.6 to 51.0 categories; p=0.003 for the >0.3 to 50.6 cm category; and p=0.055 for the >1.0 cm category. The probabilities that there were no significant differences between root volumes by location were p<0.001 for the 50.3 cm category; p=0.003 for the >0.3 to 50.6 cm category; £1) ,,- I GO (bf-f Ix. 9: slam I 0-1 O r.- Ell Table 2-1. top 41 Average root volumes per 1-nfi pit in the (0 to 30 cm) and bottom (30 to 60 cm) layers of soil as separated by diameter classes and locations (disease center, outside, check, n=2). n=5; disease edge, n=5; disease n=5; Huron-Manistee check, n=4; and Roscommon Soil Plot Top Avg. Bottom Avg. dggth Disease volume Std. volume Std. Center cm’ Dev . % cm’ Dev . % % top / %bot 5.3 an 10.0 10.0 3.2 5.3 6.3 3.0 65.5/34.5 .3) 5.6 20.8 22.4 6.6 18.1 17.1 10.1 53.5/46.5 .6) 51.0 41.0 32.7 13.1 25.3 26.6 14.2 61.8/38.2 )1.0 241.8 236.6 77.1 129.6 98.0 72.7 65.1/34.9 Totals 313.6 100 178.3 100 63.8/36.2 Disease Std. Std. Edge cm" Dev . % cm’ Dev . % % tcgl %bot 5.3 an 29.8 13.7 5.7 19.7 11.7 9.3 60.1/39.9 .3) 5,6 60.5 37.0 11.5 36.6 21.0 17.3 62.3/37.7 .6) 51.0 120.2 90.1 22.8 46.8 20.9 22.1 72.0/28.0 )1.0 316.5 172.5 60.0 108.5 98.5 51.3 74.5/25.5 Totals 527.0 100 211.6 100 71.4/28.6 Disease Std. Std. Outside can3 Dev . % cm:I Dev . % % top/ %bot 5.3 cm. 67.5 20.6 6.7 25.2 8.9 5.5 72.8/27.2 .3) 5.6 109.0 53.6 10.9 28.9 21.3 6.3 79.1/20.9 .6) 51.0 133.6 70.7 13.3 36.8 16.8 8.1 78.4/21.6 )1.0 691.0 535.1 69.0 365.7 399.1 80.1 65.4/34.6 Totals 1001.1 100 456.6 100 68.7/31.3 42 Table 2-1. (Cont'd) Soil Plot Top Avg. Bottom Avg. depth HM Check Std. Std. %top/%bot cm3 Dev. % on? Dev. % 5.3 cm 69.1 26.1 8.8 31.7 11.5 7.0 68.6/31.4 .3» 5.6 67.7 18.8 8.6 34.4 6.5 7.6 66.3/33.7 .6) 51.0 122.9 74.7 15.7 51.1 10.5 11.2 70.6/29.4 )1.0 525.4 117.1 66.9 337.4 294.2 74.2 60.9/39.1 Totals .1 785.1 100 454.6 100 63.3/36.7 Rose . I Std . Std . %top/ %bot Check em3 Dev . % can3 Dev . % 5.3 on 132.4 67.7 20.1 39.7 14.0 15.5 76.9/23.1 .3) 5.6 98.2 54.9 14.9 54.4 47.3 20.9 64.4/35.6 .6) 51.0 194.2 24.9 29.5 55.8 11.0 21.7 77.7/22.3 )1.0 234.1 226.6 35.5 108.4 65.2 41.9 68.4/31.6 Totals 658.9 100 258.3 100 71.8/28.2 .75) (W (i '4 (I) Fd 43 p=0.009 for the >0.6 to 51.0 cm category; and p=0.169 for the >1.0 cm category. Thus, the only root—size category which did not differ significantly at the 0.05 level by location or depth was the >1.0 cm category. There were no significant interaction effects of location X depth for all root-size categories at the 0.05 level. In most cases, 60 to 80% of the root volume was in the top level of soil depth regardless of location (Table 2-1). Roscommon plots, in Table 2—1, had the largest proportion of fine-root mass (roots 50.3 cm in diameter) in both levels of soil depth followed by the check plots at Huron-Manistee. Tukey's HSD procedure revealed that fine- root volumes at Roscommon (132.4 cmfi) differed significantly at the 0.05 level from disease centers (10.0 cm?) and edges (29.8 cm?) in the top level (0-30 cm) of soil depth. Also in the top soil depth, fine—root volumes from disease centers (10.0 cm?) differed significantly at the 0.05 level from fine-root volumes at disease outer regions (67.5 cm” and Huron-Manistee healthy locations (69.1 cm?). In the bottom level (30 to 60 cm) of soil depth, fine- root volumes from Roscommon (39.7 cm?) differed signifi— cantly only from the fine-root volumes from disease centers (5.3 cm?) at the 0.05 level. Fine-root volumes from disease centers (5.3 cm?) also differed at the 0.05 level from fine- root volumes at disease edges (19.7 cm?), disease outer regions (25.2 cm?), and Huron—Manistee healthy locations (31.7 cm?) in the bottom soil depth. F1 rm 4. «A K 1, 1'1) '(1 r 44 Fewer significant differences were found within the larger-diameter root categories in Table 2-1 than for the fine-root category using Tukey's HSD procedure. In the >0.3 to 50.6 cm category, significance at 0.05 was found only between volumes between center (20.8 cm?) and outside regions (109.0 cmfi) of disease plots in the top depth. In the >0.6 to 51.0 cm category, a significant difference was found only between volumes from the top depth of disease centers (41.0 cmfi) and Roscommon (194.2 cm?). No signifi- cance at either soil depth was found in the >1.0 cm category. In short, there was no consistent or easily interpreted pattern associated with larger-diameter roots. Table 2-2 lists the average volumes of fine roots, defined as roots 50.3 cm in diameter, per 1 n? X 60 cm deep pit, by location and condition class (i.e. live, dead, or symptomatic). In the fine-root category, ANOVA revealed significant differences between volumes by location, condition class, and for the interaction between location and condition (p<0.001 for all cases). A significant interaction effect indicates that volumes of live, dead, or symptomatic fine roots are partially dependent upon the location from where the roots were sampled (i.e. disease vs. non-disease areas). In disease centers, for example, where there were many tall dead standing trees, there was little expectation that many live fine roots would be recovered compared to healthier—looking locations. These location 45 OOH N.NOH OOH 0.00H OOH O.vm OOH m.mv OOH N.mH nanuoa O.mfi 0.0m 0.0H H.OH m.mm m.Hm O.mm O.ma v.ON H.m m v.mH N.ON O.N O.N 0.0 v.m N.Ov 0.0H 0.0m 0.0 Q 0.00 H.OHH v.mh O.m> 0.0m H.vm O.>N >.mH O.mm m.m A w ”an w nEu w nEu w men m ago "coflufluooo doaaoouom oouudsgz oowuuso ovum uouaoo lacuna "noowueooq ouooudauooz "nsowuoooa ouuouaa oouudonzusousm .Acoaaoomom pom omumflcmz Icousmv muoHd Mooco cam Aopflmuso paw .oopm .uopcmov muoad ommmmflp moumflcmzucousm cflnuflz coflumooa >o nouoamflp Ga So m.Ow muoou mom muoou Amy oHumEoudE»m cam .AOV poop .Aqv m>HH mo mcoflun0doud pcm Dad ammo So OO x EH x EH nod moEdHo> uoou oomuo>< .mum magma I ‘h I M £1. (I) (T (h 1.1"1 u . L). i h 46 factors had a strong influence on the volumes of live and dead roots recovered. Tukey's HSD procedure, in ANOVA trials on the subpopulation of live roots in Table 2-2, revealed that significant differences in fine—root volumes were found between disease centers (3.5 cm?) and edges (13.7 cmfl, compared to the fine-root volumes in outside regions of disease plots (54.1 cm?), Huron—Manistee non-disease plots (79.0 cmi), and Roscommon non-disease plots (117.1 cm?). Thus, diseased areas differed significantly from regions where trees appeared healthy, even the outside regions of disease plots. None of the healthy—looking regions (disease outside, Huron-Manistee check, and Roscommon check) differed significantly from each other. The population of dead fine roots in Table 2—2, did not differ significantly from each other by location (p=0.146). This can be explained by the fact that dead-root volumes were small in healthy areas (disease outside 8.4 cm}, Huron- Manistee 2.7 cmfi, Roscommon 28.2 cm?) and comparable to the dead-root volumes found at the center (8.6 cm?) and edges (19.8 cmfi) of disease plots. However, dead fine-root volumes constituted a relatively small proportion of the total fine-root volumes from healthy-looking areas (disease outside 8.9%, Huron—Manistee 2.7%, Roscommon 16.4%) compared to the disease centers (56.6%) and edges (40.2%). For symptomatic fine-root volumes (i.e., roots that were partially dead or dying, and/or stained below the 47 epidermis) in Table 2-2, significant differences at the 0.05 level were found between disease centers (3.1 cm?) compared to the fine-root volumes at disease edges (15.8 cmfl), outer regions (31.5 cm?), healthy plots at Huron-Manistee (19.1 cm?), and Roscommon (26.9 cm?). This can be explained by the fact that the total volume of fine roots removed from disease centers was much less than at other locations, including the subpopulation of symptomatic roots. Tables 2-3, 2-4, and 2—5 list the average root volumes per 1m x 1m x 60 cm deep pit by location and condition class for the larger root—size categories (i.e. >0.3 and 50.6 cm; >0.6 and 51.0 cm; and >1.0 cm in diameter, respectively). For all three large-root size categories, live-root volumes from disease centers were significantly lower at the 0.05 level compared to live-root volumes from healthy regions: disease outside, Huron—Manistee and Roscommon. These same relationships were found for the live fine-root (50.3 cm) category. Significant differences at the 0.05 level were also found between disease edges and healthy regions for the larger root-size categories. In the >0.3 and 50.6 cm live- root category, root volumes from disease edges (27.7 cmfl were significantly lower than the root volumes from disease outside regions (89.4 cm?), Huron-Manistee (81.8 cm?) and Roscommon (100.2 cm?) (Table 2-3). These same relationships were identified in the live fine-root category. In the >0.6 and 51.0 cm diameter category, disease edges (35.3 cmfi 48 OOH O.H¢H OOH N.NOH OOH 0.0MH OOH H.>m OOH 0.0m anuOB >.H v.m 0.0N v.0m O.vm N.vm H.Ov m.Om 0.0M O.mH m N.>N v.mm 0.0 0.0 v.OH m.vH v.Hm m.om O.Nm ©.ON a H.Hb m.OOH 0.00 O.HO O.vm v.mw m.ON 0.0N v.OH v.m A w «.80 m ":8 w n80 w «:8 w «80 " coaufioaou ooEEoouom oouuwsnz ofifioupo ovum Season noouom "noowuooon ouuoafiaaooz "enowuaooa oudouwn oouawddzuoousm . .AcoSSoomom cam ooumHSmZISOSDmV mDOHQ xoono paw HopHmuso pom .omom .Soucoov mDOHd ommome oopchmZISousm cHnqu SOHDmooH >Q SouoSme SH So 0.0w cam m.OA muoon Sow muoou Amv oHumSoudS>m paw .AOO poop .Aqv o>HH Ho mooHuSOQOSd cam DHQ moop So OO x SH x SH Sod moSdHo> uoou oomuo>¢ .mum oHQmH 49 OOH H.Omm OOH O.vOH OOH «.me OOH 0.00H OOH m.mo nHuuoa H.v m.OH v.OH m.mm N.mm 0.0v o.mm «.mm m.m v.w m v.m H.O O.N m.m H.m O.vH 0.0v H.me v.vm m.Om 9 «.mm O.mmm m.HO O.NVH O.mm 0.00H 0.0m m.mm m.m v.v A m "=8 w «Bo w n30 m 6:8 m «So " soap goon aoEEoonom oouewsdz oowuupo Goon Henson loosen ”udowunoon ounoawolsoz "noowuuooa annouaa oouudduxneousm .AcoSSoomom pom oouchmZISoudmv muoHd xoozo cam Ampflmuso paw .oopo .uoucoov muoHd ommome oouchmzucousm cHnqu coHpmooH HQ SouoSme SH So O.Hw cam O.OA muoou Sow muoou .AHO o>HH mo mcoHuSOQOSQ cam DHQ down So OO x SH x SH Sod moSdHo> DOOS oomuo>4 .vnm oHQmB Amy oHumSoudem paw .HOO poop OOH m.mwm OOH O.NOO OOH OmOH OOH O.mmw OOH O.Hem manuoa 0.0 0.0 m.mm e.vmm 0.0m O.mmm m.mm O.HOH O.vH H.mm m mw O.v O.mH m.H O.HH v.m O.mm m.Ov H.OON O.vm m.va a «.mm 0.0mm N.mm 0.0mm O.mm m.mmm 0.0N N.OHH 0.0 m.m A w «Eu w “EU m "=8 w 650 w «.80 " noduwuoou oofiuoouom wounded: cowuuao anon no» :00 Idousm "uncauuoon omcoawoueoz "aeowuooog ouuoudo denudedzueousm .AcoSSoomom paw oouchmz Icoudmv muoHd xooco pom HopHmpso pom .oopo .uoucoov mpoHd ommome ooumHszndogsm dHcsz doHumooH HQ SouoSme SH So O.H A muoon mom muoou Amy oHumSouQSHm paw .on poop .Aqv o>HH Ho mcoHDSoQoud cam DHQ ammo So Om x SH x SH Sod moSsHo> poou oomgo>¢ .mum oHQmB 51 differed significantly from Roscommon (233.7 cm?) only (Table 2—4); and in the >l.0 cm category, disease edges (117.2 cmfi) differed from the disease outside areas (695.3 cm?) and Huron—Manistee check plots (596.7 cm3)Ibut not from Roscommon (326.8 cm?) (Table 2—5). Unlike the fine-root category, some significant differences were found between dead—root volumes for the larger root-size categories. In the >0.3 and 50.6 diameter category, significant differences were found for dead-root volumes from disease edges (30.5 cm?) and Roscommon healthy plots (38.4 cm?) compared to the healthy plots at Huron— Manistee (0.0 cm?) (Table 2-3) (p=0.023 and p=0.050). In the >0.6 and 51.0 cm category, significant differences were found between disease edges (79.1 cm?) and Huron-Manistee healthy plots (3.5 cm?) (Table 2-4) (p=0.050). In these cases, disease edges tended to have significantly higher volumes of dead roots than non-disease locations, with the exception of Roscommon in Table 2-3. Of healthy regions, Roscommon had the highest proportions of dead fine roots. No significant differences in dead—root volumes were identified for the >1.0 cm category (Table 2—5). For symptomatic root volumes, significant differences by location were not identified in any of the larger root-size categories. 52 Core—Sampled Roots Table 2-6 illustrates the results of core-sampled roots. The cores were taken along transects around trees starting from the edge of the pockets to the outside regions of disease plots where trees appeared healthy and from the non-disease plots at Roscommon. The center regions of the pockets were not sampled since they consisted mainly of dead standing trees and a diverse range of plant species taking advantage of the pocket openings. It would not have been possible to select only red pine roots from other types of vegetation growing in the center areas. It was expected that core—sampling would show increasing volumes of fine-root biomass starting with the lowest volumes at the pocket edges and increasing towards the outside healthy regions of the plot. The edge trees (Tree A from each plot) were always declining trees with obvious symptoms of RPPM. The trees midway between the transect (Tree B from each plot) appeared healthier compared to edge trees but many were starting to show symptoms of decline. The "C" trees, in the outer regions of the disease plots, were asymptomatic, healthy-looking trees. However, unlike pit-sampled roots, the data in Table 2~6 do not reveal any trend in increasing fine-root biomass from the declining to healthy looking trees within a plot. These data are separated according to seven root- diameter categories ranging from .15 mm to > 3.00 mm in diameter. Roots that were 2.5 mm or larger were pulled out 53 of the sample and measured and recorded separately because they were too large to be processed by the robotic-camera method. ANOVA was used to test differences of root volumes within the seven diameter categories by location. There were no significant differences found between root volumes located along transects within disease plots. But some significant differences were found between disease locations and Roscommon. These differences were only in the .15 and .4 mm diameter categories. Roscommon had signifi- cantly higher volumes of fine roots compared to all three disease locations within these categories (P<0.001 for the check compared to Tree A, Tree B, and Tree C locations in the .15 mm category, and Tree A and Tree B locations in the .4 mm category; and P=0.002 for the check compared to Tree C locations in the .4 mm category). There were no other significant differences identified for any of the other root—size categories except for the total volume of roots at each location which was significantly higher at Roscommon compared to the Tree B locations of disease plots (P=0.050). The significant increase in very fine—root volumes (.15 and .4 mm categories) at Roscommon, suggests that these healthy locations may have higher production or perhaps reduced mortality of very fine roots compared to the disease locations. The core-sampling method failed to reveal the same increasing gradients in root volumes within disease plots revealed by the pit-sampling method. In the core- sampling method it was not possible to distinguish red pine 54 roots from roots of other plant species. In collecting core samples we attempted to avoid areas within the plot where competing vegetation might interfere with the results. However, such as grasses, (Poa, SPPJ, raspberries (Rubus, spp.) roots of many plant species other than red pine, and bracken fern (Pteridium aquilinum), especially near the more open canopy of pocket edges, may be increasing the volumes of fine roots collected. under a closed canopy with little interfering vegetation, The Roscommon trees sampled were SO it is believed that the significant differences identified can largely be attributed to increased volumes of fine red pine roots. Table 2-6. .Average volumes (cm9)jper core (10 cm wide X 30 cm deep) of roots from the edge to outer regions of Minnie Lake disease plots, Huron-Manistee National Forest and in check plots at Roscommon County. Minnie Lake Disease Root Diameter (mm) Plots Transact I Total Location n .15 .4 .75 1.25 1.95 2.50 )3.0 Avg. Tree A- edge 15 0.20 1.21 2.35 0.43 0.03 1.54 0.00 5.75 Tree B- middle 15 0.16 1.50 2.56 0.43 0.02 0.81 0.22 5.70 Tree C- outer 15 0.17 1.63 2.75 0.37 0.02 0.48 0.28 5.78 Roscommon 30 0.37 2.31 2.86 0.37 0.02 1.24 0.09 7.39 55 Isolation of Pathogenic Fungi No pathogenic fungi were recovered from root isolates on non-selective media. Of 489 isolates on selective media, only three cultures of Leptographium, spp. were identified. Two of the cultures were Leptographium truncatum and one was Leptographium procerum. All three Leptographium cultures came from symptomatic roots from disease plots (a subpopulation of 42 root cultures). Black staining was visible beneath the epidermis of the roots from which the three cultures were isolated. Two of the Leptographium isolates were from TG4 and one was from MLl. No Leptographium fungi were recovered from roots from non— disease check plots and black-staining was not observed in check plots and rarely in disease plots. The only other potential pathogenic fungi recovered were two isolates of Qphiostoma, spp., also from symptomatic roots in disease plots (TG4). Most of the isolations came from TG4, the only plot in the study with poorly drained soil (Chapter 3). Perhaps greater availability of moisture within the TG4 plot facilitated the growth of these fungi. It was thought that there were not enough isolations done on symptomatic roots from disease plots, so in 1993 the process was repeated; this time 111 symptomatic roots from disease plots were cultured and no Leptographium nor any other noteworthy root pathogens were recovered. Overall it does not appear that Leptographium, spp. are a significant fungal pathogen in these plots. 56 DISCUSSION The pit-sampled roots collected revealed that the highest concentration of roots were located in the top 30 cm of soil versus the bottom 30-60 cm. This was true for all root-size categories (Table 2—1). In our sampling of soils in disease plots, the top layers of soil, particularly the O and E horizons, were also the most acidic, with pH ranging from 4.1 to 4.8 at the Minnie Lake and Thin-A—Gain sites (Chapter 4). These pH's are within the range for considering the possibility of aluminum-toxic effects which may be involved in fine-root mortality at these sites. Red pines at Roscommon appear to be more proficient at producing roots in the fine—root category (50.3 cm in diameter) and/or the fine-root mortality rates are lower versus the red pines at Huron-Manistee healthy locations and diseased areas; although significance was found only in comparison with the diseased areas (Table 2—1). Conversely, red pines from healthy locations at Huron-Manistee appear to be more proficient at producing roots in the largest size category (>1.0 cm in diameter) compared to Roscommon; but again these differences were not statistically significant. These observations suggest that fine—root turnover rates are higher at Roscommon than at Huron-Manistee. Not only were the fine-root volumes per pit the highest at Roscommon (172 cm9), but so were the proportions of dead roots (16.4%) in comparison with the check plots at Huron— Manistee (2.7%) and the outside regions of disease plots -. .19:WNW-mWivWN-Mwhww* *‘ ~ 57 (9.1%) (Table 2-2). Significant differences were few, however, because of the limited number of replications (disease center, n=5; disease edge, n=5; disease outside, n=5; Huron-Manistee check, n=4; and Roscommon check, n=2). More pits would need to be dug in order to confirm the differences between these sites. Tables 2-2 through 2-5 reveal that the same trends in volumes of live, dead, and symptomatic roots in the center, edges, and outside regions of pockets existed for all root- size categories. The greatest percentage of dead roots was located at the center, followed by the edge. Symptomatic roots tended to be found at the pocket edges but also extending into the outside "healthy-looking" regions. Significant differences in fine-root volumes of pit- sampled roots were found mainly between disease centers and edges and regions where red pines appeared healthy. This was not a surprising result since trees were visibly declining in disease regions. One of the questions raised while observing RPPM was whether above—ground symptoms precedes below-ground root mortality or whether root mortality occurs before symptoms are seen in tree crowns? The fact that there was an increasing gradient in fine-root volumes from the center of pockets to the outside regions of disease plots, to the healthy plantations at Huron—Manistee and Roscommon suggests that roots are dying first before symptoms are expressed in tree crowns. The outside area of disease plots also tended to have higher volumes of 58 symptomatic roots than healthy plantations (Roscommon and Huron-Manistee), suggesting that the disease may be spreading beyond the pocket boundary into the healthy— looking tree population. However, the differences between fine-root volumes in the outside regions of disease plots compared to healthy plantations were not statistically significant. Fine—root volumes were the root-size category of primary interest in this study because of the importance of fine roots in absorption and exchange of soil nutrients. If Al—toxicity is a cause of root mortality in red pine pockets, its effect would be primarily on the fine-root system. Joslin and Wolfe (1988) found that the primary effect of Al—toxicity to red spruce seedlings was upon the root system. Their greenhouse-pot study points to toxic levels of Al as the major cause of root and foliar biomass reductions in seedlings. Soil solution levels of inorganic monomeric Al and total Al were superior predictors of root biomass in the study. It may be that aluminum, or more importantly, the ratios of aluminum to other nutrients, especially calcium, are the primary factors controlling fine-root biomass within disease plots. In trying to characterize the disease in terms of biotic agents, (i.e root pathogens), we were not able to corroborate the studies done in Wisconsin. No significant root pathogen populations were identified. It may be that whatever agents are responsible for the decline in Wisconsin 59 are not significant factors to Michigan pockets. The role of biotic agents should be looked into further because the often circular pattern of red pine mortality pockets suggests a strong biotic aspect to the disease. We did observe bark beetle (Ips pini), pine root collar weevil (Hylobius radicis), Armillaria root rot (Armillaria, spp.), and Diplodia tip blight (Sphaeropsis sapinea) activity in pockets. However these agents were not consistently present in all pockets and we do not have quantified information on them. The significant interaction effects for all root-size categories between location and condition indicate that proportions of live, dead, and symptomatic roots will vary dependent upon the location from where the roots were sampled (i.e. areas within disease plots and from non— disease plots). Although there were no statistically significant differences in root volumes between outer, healthy regions of disease plots and non-disease plots, the increasing gradient in live fine-root percentages and corresponding decreasing gradient in dead fine-root percentages from pocket centers to the non-disease stands at Huron—Manistee suggest that fine-root mortality occurs before symptoms are seen in tree crowns. 6O BIBLIOGRAPHY Harrington, T.C. 1981. Cycloheximide sensitivity as a taxonomic character in Ceratocystis. Mycologia, Vol. 73, pp. 1123-1129. Johnson, D.W. and G.E. Taylor. 1989. Role of air pollution in forest decline in eastern North America. Water, Air, and Soil Pollution 48: 21-43. Johnson, D.W. and I.J. Fernandez. 1992. Soil-mediated effects of atmospheric deposition on eastern U.S. spruce—fir forests. In: Ecology and Decline of Red Spruce in the Eastern United States. C. Eager and M.B. Adams, Eds., Springer-Verlag, New York, pp. 235—270. Joslin, J.D. and M.H. Wolfe. 1988. Responses of red spruce seedlings to changes in soil aluminum in six amended forest soil horizons. Can. J. For. Res. 18: 1614:1623 Klepzig, K.D. and J.C. Carlson. 1988. How to identify red pine pocket decline and mortality. USDA For. Serv. NA- GR-19. Klepzig, K.D., K.F. Raffa and E.B. Smalley. 1991. Association of an insect-fungal complex with red pine decline in Wisconsin. For. Sci. 37:1119-1139. Raffa, K.F. and D.J. Hall. 1988. Seasonal occurrence of pine root collar weevil, Hylobius radicis Buchanan (Coleoptera:Curculionidae), adults in red pine stands undergoing decline. Great Lakes Entomol. 21:69—74. Smucker, A.J.M. 1989. Operations manual for the staining, preparation, and video recording of washed root systems by the robotic camera method. Root Image Processing Laboratory, Department of Crop and Soil, Sciences, Michigan State University, unpublished. Srivastava, A.K., Smucker, A.J.M., McBurney, S.L. 1981. An improved mechanical soil-root sampler. Paper presented at the 1981 Summer Meeting of the American Society of Agricultural Engineers, St. Joseph, Michigan: The Society (fiche no. 81-1038). CHAPTER I I I Condition and Disease Progress of Red Pine Pocket Mortality Stands in Michigan ABSTRACT Red Pine Pocket Mortality (RPPM) is a disease of unknown causes. This study tracked the condition of red pines (Pinus resinosa) in pockets in Michigan during the period 1991 through 1995. Disease progress rates indicated that the pattern of disease increase or decrease may coincide with drought and non-drought years. Declining trees within pockets are more likely to die or decline further during drought years than non-drought years. Studies of tree rings indicated that, as a result of the 1988 drought, red pines in all areas exhibited reduced growth, regardless of disease status. After 1988, trees in diseased plots continued to decline, while healthy trees exhibited a rebound in growth. Compound stresses, including drought, may eventually lead to expression of crown symptoms in previously healthy—looking trees and to the expansion of the pocket beyond the boundaries of the declining red pine. Spatial autocorrelation analyses revealed that the pattern of disease progress is not random. Trees within proximity to each other are more likely to show symptoms of the disease than trees spaced further apart. This pattern 61 if“ 62 suggests movement of an infectious agent from one tree to another. Leptographium fungi were found to be associated with mortality pockets in Wisconsin; however, we were not able to verify this association in Michigan (Chapter 2). Temporal analyses of disease spread revealed that the disease spreads slowly, suggesting that whatever agent(s) are involved, they are not very aggressive and/or virulent. We believe the disease to be one of complex origin involving interactions between biotic and environmental factors. Our further studies of RPPM focus on the abiotic factors associated with this disease. INTRODUCTION This study was conducted to assess the condition and disease progress of red pine (Pinus resinosa) mortality pockets in Michigan. Red pine pocket mortality (RPPM) is a relatively new disease that was first detected in 1975 (Klepzig and Carlson, 1988; Raffa and Hall, 1988). It has been reported with increasing frequency throughout the Lake States. The disease affects mature plantation-grown red pine in stands that are about 30- to 50-years old. Symptoms of RPPM include the presence of a large, often circular pocket opening within a stand; dead trees and greatly increased growth of understory plants in the center of the pocket; trees towards the edge of the pocket exhibit reduced diameter and height growth; and browning and stunted growth in needles. Trees further from the margin of decline appear 63 healthy and vigorous in comparison (Klepzig and Carlson, 1988). In this study, five red pine pocket mortality sites identified by the U.S. Forest Service were selected for observation. The plots are all located in the Huron- Manistee National Forest in Newago County, Michigan. The stands were monitored over a five-year period (1991-1995) in order to determine the overall condition of the stands (i.e., site factors and general health of the trees) and rate of disease spread. The causes of RPPM are under investigation. Studies in Wisconsin have identified a complex insect vector and patho- genic fungi association that may lead to the decline. In Michigan, several species of insects and fungi are asso- ciated with the decline but none are consistently present in all pockets (Dr. John Hart, unpublished). In Wisconsin, significant differences in the abundance of five insect species were found in declining stands compared to healthy P. resinosa stands. These root- and lower stem-infesting insects consistently carried two fungal species-- Leptographium terebrantis and L. procerum. These fungal species are thought to be transmitted to the red pine root system through five insect vectors resulting in expression of above—ground symptoms. They may also spread via root grafts or contacts and/or through soil. Leptographium infection can cause below-ground root mortality which pre— cedes the above-ground symptoms of reduced radial growth, 64 thin crown structure, and infestation by the pine engraver (Ips pini) (Klepzig, Raffa, and Smalley, 1991). C‘I‘l I I'M.“ .1‘5‘;QF‘I‘\)IWF|\‘MQH“ u‘ The rate at which RPPM is spreading and the degree of crop loss has not yet been assessed in Michigan or in other .- _ "3- ,- - u U .. y u . Lake States. An aerial survey by the U.S. Forest Service of the Huron-Manistee National Forest found 440 pockets of dying red pine (USDA Forest Service, 1988-1989). However, many of these pockets were most likely caused by bark bee— tles (I. pini) rather than RPPM. Ground surveys would be needed to distinguish between the two pocket types. Since the drought years of 1988-89, the number and size of red pine pockets in Michigan have not increased appreciably. This study attempts to determine the rate of disease spread and the spatial patterns associated with the disease. This information may be useful in assessing how RPPM is af— fecting red pine as a marketable resource. Traditionally the red pine resource has been managed for revenue from pulpwood and sawtimber. In the last decade a sustained market for red pine utility poles has increased the value of this resource in the northern Lower Peninsula (Grossman and Potter-Witter, 1990). Under the right management scenario, increased production of red pine utility poles is projected to pay substantial stumpage premiums (Grossman and Potter— Witter, 1991). Presently there is no information on how RPPM is affecting red pine pulpwood, sawtimber, or utility pole markets. 65 METHODS Study Locations Five plots of RPPM at two different locations in the Huron-Manistee National Forest were identified. These areas have been mapped, and data collected on the diameter and condition of the trees within the area of decline and for an area outside the pocket. These plots are all located within the Huron-Manistee National Forest in Newago County, Michigan. Three disease plots (MLl, ML2, and ML3) are located at an area known as Minnie Lake and two disease plots (TG4 and TGS) are located about 17 km east of Minnie Lake in an area known as Thin-A—Gain (Figures 2-1 and 2-2). The red pine at the disease and healthy sites examined are growing under similar conditions of soil, climate, and topography. The plots in Newago County are located in plant hardiness zone 5a (average annual minimum temperature -26.2 to -28.8°CD. The two healthy plots in Roscommon County are located in plant hardiness zone 4b (average annual minimum temperature —28.9 to -31.6’CD near the boundary of plant hardiness zone 5a (USDA, 1990). The total annual precipita— tion in Newago County is 85.27 cm, the average annual tem— perature is 7.6”Cn and the growing season is 120—150 days (USDA, 1995). The two check plots located in Roscommon county are about 115 km northeast of Newago county and are further inland from Lake Michigan. Therefore, the average annual temperature is slightly cooler and the growing season slightly shorter. The soils under red pine in both counties 66 are sandy Spodosols of low pH. These are characterized by a layer of pine needle litter; an O horizon (2.5 to 5 cm) of a black to dark grey humus and roots layer; an E (albic) hori- zon (2.5 to 10 cm) consisting of a grey leached mineral layer; followed by the spodic B horizon which consists of a yellow-brown sand. In this study sampling depth was 46 cm so no samples were taken below the spodic B horizon. The pH range, in both disease plots and healthy plantations, was 3.8 to 4.7 for the O horizon, 4.2 to 5.1 for the E horizon, and 4.8 to 5.4 for the B horizon. Data Collection During 1991, the first year of study, the disease plots were mapped, trees numbered, and the diameter and condition of each tree within the plot recorded. The condition of each tree was tracked over the 5-year period. Information recorded on each tree included its condition (i.e., its apparent health and/or any visual symptoms of RPPM), its location inside the plot (row number, and distance within row), crown class (dominant, codominant, intermediate or suppressed) and dbh (diameter-at-breast height). Observa— tion of RPPM symptoms and tree decline allowed development of a numerical rating system that facilitated tracking of disease progress through time. The rating system used to evaluate tree condition is listed in Table 3—1. Other observations made in each plot include the flora, soil type and drainage, and the pH of soil horizons. 67 Information collected from healthy red pine plots included the number of trees within the plot, soil type and drainage, soil horizon pH, dbh, crown class, and number of dead stems. The average dbh, crown class, and number of dead stems were estimated from a 10% selected sample of the plot versus the complete inventory done in the disease plots. The sample was selected by first counting the number of trees in the plot, then randomly choosing a section of the plot and recording the information row—by-row until 10% of the total sample had been observed. Table 3—1. Numerical rating scale for evaluating disease progress of red pine pocket mortality. Code Descriptions 0 No apparent symptoms in foliage. l Browning or stunted growth in needles, thinning crowns, and/or poor terminal growth (30% of crown or less). 2 Browning or stunted growth in needles, thinning crown, and/or poor terminal growth (30% to 60% of crown). 3 Browning or stunted growth in needles, thinning crown, and/or poor terminal growth (greater than 60% of crown). 4 Dead standing tree. 68 Tree Ring Analysis Twenty-three stem cross—sectional areas were taken at 30 cm above the ground from declining and healthy trees at Minnie Lake and Thin—A—Gain in late July, 1993. The nar- rowest and widest radii from the top side of each disc was determined and the average radius calculated. Annual ring widths were measured along the average radius back to the pith. The average diameter growth for declining red pine was determined and compared to the average diameter growth for healthy red pine at both locations (Minnie Lake and Thin-A—Gain). Evaluation of Disease Progress To evaluate disease progress, plots were made of annual disease intensity. Intensity is defined as the quantity of disease present (Campbell and Madden, 1990). In this study, intensity was calculated by averaging the rating scale (Table 3-1) for each disease plot. Disease assessment was made once a year in mid to late June. This time interval was frequent enough to observe and record changes in tree health and survival. Intensity, or the average of the rating scale, was calculated with regard to cumulative mor- tality i.e., all dead (Rating 4) trees observed in previous years were maintained in the data set and used to calculate the average. Also plotted, for each disease plot, was the change in disease intensity over time (AY/AT). The AY/AT represents 69 an absolute growth rate of disease (Campbell and Madden, 1991). The location of each tree was determined by measuring the distance between and within rows. This allowed display of the spatial relationships between trees in different rating classes and the spread of the disease over time. The maps were done for the Minnie Lake sites only because of the limited data available at Thin—A—Gain after thinning at this site in 1993. Statistical Analyses Durbin-Watson Test Statistic The Durbin—Watson test statistic (d) was used to deter- mine whether the pattern of disease progress within each of the plots is a factor dependent upon time. The test was ap- plied to the data on disease intensity. The Durbin—Watson statistic measures whether residuals versus time are simi— lar, (i.e., positively serially correlated) or dissimilar, (i.e, negatively serially correlated) (Durbin and Watson, 1951). The formula is: d=2n-1(et‘1_et)2 262. where et denotes the residual at time t, and n the total number of time points. Residuals (6.) were calculated by subtracting the mean disease rating averaged over the five- year period from the disease rating for each year. 70 When there is no serial correlation, the expected value of the Durbin-Watson statistic d is approximately 2.0. In general, values of d less than approximately 1.5 are con— sidered a positive serial correlation and values of d great— er than approximately 2.5 are considered a negative serial correlation (Ott, 1988). Spatial Autocorrelation Spatial autocorrelation analyses were used to detect spatial patterns of disease progress. Moran's I was the spatial autocorrelation method applied to maps showing the spatial distribution of trees in different rating classes as measured in 1991. The first step in calculating Moran's I was to determine distance classes (k) for establishing linear distance criteria for the distance (d) between trees. Distance classes chosen were a) <5 m b) 5 to 10 m; and c) >10 m apart. Trees were spaced in rows less than five meters apart so it was expected that this category would reflect the similarities between trees most accurately. The linear distance (d) between all X; and.Yg coordi— nates on the map was calculated by taking the square root of (X55g)2 + (Yg—Yflz. The Moran's I formula is: where n is the number of locations sampled, x is the average value of x5, k is the distance class, and wWQj is equal to 1 71 if i and j are both within distance class k and equal to 0 if i and j are not (Slatkin and Arter, 1991; Cliff and Ord, 1981). MatLab”’(1992) was used to calculate distances and Moran's I values. To test the null hypothesis that the pattern has been generated in a random manner, it was necessary to compare observed to expected values. The expected value for I was calculated with the equation E(I)=--(n—l)‘1 (Cliff, et al., 1975). Variances of Ikwere also calculated according to procedures detailed by Cliff, et al., 1975. RESULTS Site Conditions Within disease plots, the vegetation present in pocket openings indicated similar site conditions (Table 3-2). Differences in the composition of flora may be partially ex- plained by soil drainage factors. With the possible excep- tions of ML3 and TG4, all of the disease plots had well- drained, dry soils with no evidence of a saturated water table. In ML3, which is located near a marshland, there was some evidence of a high water table. The soil appeared to be saturated at a depth of 1.37 m, yet the yellow color of the sand indicated that the soil was not permanently satu- rated. Water hemlock, Cicuta maculata, was found growing only at ML3, supporting the observation that this site is wetter than most other disease plots (Table 3-2). The soil profiles examined at TG4 indicated that it was a wetter site 72 Table 3-2. Flora present in the pocket Openings of red pine mortality plots. Common'Name Species Plots found in red maple Acer rubrum ML2, ML3, TG5 brown knapweed Centaurea jacea ML3 water hemlock Cicuta maculata ML3 bull thistle Cirsium vulgare TG4, TG5 sweet fern Comptonia ML1 peregrina beech Fagus grandifolia ML1 common St. John's Hypericum ML1, ML3, TG4 wort perforatum red pine Pinus resinosa ML1*, ML2*,IML3*, TG4’, TG5+ grasses Poa, spp. ML1, ML2, ML3, TG4, TG5 pin cherry Prunus ML2, ML3 ‘pensylvanica bracken fern Pteridium ML1, ML2, TG4, aquilinum white oak Quercus alba ML1, ML3, TG5 swamp white oak Quercus bicolor . ML2 red oak Quercus rubra ML2, ML3, T65 black oak Quercus velutina ML1, TGS currant Ribes, spp. TGS raspberry Rubus, spp. ML1, ML2, ML3, TG4, TG5 mullein Verbascum thapsus ML3, T65 *Abundant red pine regeneration +Moderate red pine regeneration “Little red pine regeneration 73 than ML3. A water table close to the surface at 46 cm was identified. The grey sand and thick organic layer indicated that drainage may be a factor in the decline at this site (B. Ellis, personal conversation). Within this same plot on higher ground, however, soils were better drained. The poorly-drained area of TG4 was located in the center of the mortality pocket. Red pine was not regenerating as well within the TG4 pocket as it was in other disease plots (Table 3-2). Oak was frequently found regenerating in all of the pockets with the exception of the wet site, TG4. From the annual growth (number of nodes on the main stem) of regener- ating oak and red pine, it was estimated that the pockets opened up in the early 1980's. Much of the regenerating red pine at Minnie Lake plots ML1 and ML2 when first observed in 1991 were infected with Diplodia tip blight (Sphaeropsis sapinea). In more recent observations in 1994 and 1995, however, the young red pine had apparently grown out of this disease and appeared quite healthy and vigorous. It is not known why conditions at these sites seemed to favor the growth of young red pine while at the same time mature red pine surrounding the pockets were in a state of decline. Tree Health and Annual Growth The red pine stands examined were all even-aged, densely—planted, mature plantation red pine that ranged in age from 40- to 60-years old. The initial planting density 74 appeared to have ranged from 1.5 x 1.5 to 1.8 x 2.4 m. Row— thinnings in each plot (with the exception of R1) have since reduced stand density. Because of the variation in the size of red pine mortality pockets, the plots vary in size. Most of the disease plots were close to 46 x 46 m and all of the check plots were made this size. The exception is ML1, which contains the largest pocket and is 61 x 81 m. The stand density and basal area information in Table 3-3 has been placed on a per hectare basis for comparison. The percent of trees in each crown class category is also listed. Dominant and codominant trees were those with crowns in the upper forest canopy and receiving full light from above. Codominant trees had moderately sized crowns compared to trees in the dominant size class and also tended to have a slightly smaller diameter-at-breast height (dbh). Intermediate trees had crowns in the middle to upper canopy and were shorter and had smaller crowns than in the dominant and codominant categories. Suppressed trees were in the lower canopy and were receiving little or no direct light from above. The dbh's were the smallest for suppressed trees. In August of 1993, plots TG4 and TG5 were thinned in a timber sale in which over 50% of the trees in each plot were removed. The information given in Table 3-3 for TG4 and TG5 lists the stand density and basal area information prior to this thinning. 75 Table 3-3. Average diameter-at-breast height (dbh), basal area, and number and percent in each crown class (dominant, codominant, intermediate or suppressed) of red pines from disease plots and healthy plots. Stand 3 density Avg. Basal I Plot. stems/ dbh Area ‘ ID 5 ha (cm) nf/ha I Crown Position Category (%) Disease Plots DOM CO INT SUP ML1 714 21.3 26.4 56.0 30.1 9.4 4.5 M12 442 24.6 22.0 78.2 17.6 2.5 1.7 ML3 590 23.1 25.5 73.5 22.0 1.5 3.0 T64 1169 18.5 34.4 62.7 13.4 11.9 11.9 T65 1290 17.5 34.4 48.6 19.0 21.0 11.4 Check Plots* I MLC 736 22.9 31.0 68.2 27.3 .5 .0 TGC 1166 22.9 50.0 66.8 26.6 .3 .3 R1 2602 13.2 42.5 14.9 22.2 27.8 35.1 32 1149 25.4 61.8 54.2 33.3 12.5 0.0 *Estimates of basal area, avg. dbh, and proportions of each crown position category come from a 10% subsample of the plot versus complete inventory done in disease plots. 76 From the stand density, average dbh, and basal area information in Table 3—3, it is apparent that stand structure was the most similar for plots located at the same site. Roscommon plots were an exception. The R1 plot had a much lower dbh and much greater stand density than R2. This was an apparently unthinned plot in which trees in the intermediate and suppressed crown position categories domi- nated. By contrast R2 was characterized by a number of vigorously growing, large-diameter trees. Using 1991 as a base-line, the yearly mortality rate was determined by dividing the number of dead trees during the year of observation by the number of trees alive the previous year. The total or cumulative mortality rate for each plot was determined by summing the mortality rates for years 1992 — 1995. ML3 had the highest mortality rate (.190) followed closely by ML1 (.188) from the 1992 though 1995 study period. Mortality at the Thin—A—Gain plots was substantially lower than Minnie Lake in 1992-1993. The thinning operation at Thin—A-Gain plots in late summer 1993, drastically changed their stand structure. The Thin-A-Gain mortality rates for 1994-1995 are based on the number of trees still standing in the plot; TG4 was reduced from 212 trees to 67 trees and TG5 was reduced from 243 trees to 105 trees. It was noted that some trees at Thin—A-Gain were damaged by the logging activities; yet mortality rates are low in 1994—95 compared to Minnie Lake plots. In terms of yearly . o— 7. mt] ~a¥l .du .1. AU Plot ID 77 mortality, 1992 was the most severe; a total of 38 trees died in all five plots (Table 3-4). Table 3—4. Mortality in red pine pockets, 1991-1995. No. Of Trees Dead Plot in Standing No. of Dead Trees 8 ID Plot Trees Yearly Mortality Rate Total (n) 1991 1992 1993 1994 1995 92-95 ML1 352 62 15 20 10 6 51 .052 .073 .039 .024 .188 ML2 119 12 8 0 4 0 12 .075 .000 .040 .000 .115 ML3 132 27 12 1 5 1 19 .114 .011 .054 .011 .190 TG4* 212 34 2 1 0 3 3 (67) .011 .006 .000 .045 .062 TGS* 242 48 1 0 0 0 1 (105) .005 .000 .000 .000 .005 *Mortality rates for 1994 and 1995 were based on 67 (TG4) and 105 (TG5) operation. trees remaining in plot after thinning The only mortality found in check plots, based on a 10% subsample, was at TGC and R1. These were all suppressed trees that were dying due to the effects of natural thinning within the stand. They were found scattered throughout the plot and did not comprise a significant proportion of the stand. R1 had the largest number of dead and suppressed trees because of the high stand density at this site. It was estimated that 7.5% percent of the red pine at R1, and 3.3% at TGC, were dead and suppressed. Trees in the inter— mediate, codominant, and dominant categories appeared heaii ‘. P. I! .l V? V were . r“ f .1; - sic n OD . r .1 .1 Q c .3 far ba ,4 st IECOI. in orig S a .t I «b A.H.. r Ir. 78 healthy in all of the check plots. Mortality data for check plots are not shown in Table 3—4 because they were only evaluated once. Thus, it is not possible to compare mortal— ity increases in check plots to disease plots. Check plots were visited during each year of the study, however, and no significant disease or insect problems were found. Trees were felled at Minnie Lake (11 trees) and Thin—A- Gain (12 trees) in order to ascertain the growth patterns of declining and healthy trees at these sites. The tree ring analysis revealed that the red pine at Minnie Lake are about 10 years older than at ThineA—Gain. Rings were counted as far back in some of the discs to a year of origin of 1938 for Minnie Lake and 1949 for Thin-A—Gain. Compartment records from the USDA Forest Service reveal the year of origin to be 1937 for Minnie Lake, and 1947 for Thin—A—Gain. Some tree rings were missing in the analysis and may have been lost below the height at which the discs were taken. It was also difficult to distinguish some highly compressed rings, especially in declining trees. At Thin-A-Gain and Minnie Lake the average annual ring width was determined for trees within disease plots and compared to nearby healthy trees growing outside of the plot. Some of the trees selected from within the disease plots appeared healthy, while others were clearly declining. Tree ring analysis suggests that non-disease plot red pine at Minnie Lake (MLC) grew more vigorously than the disease— plot trees as far back as 30 years ago (Figure 3-1). At C I O C. .l v. ‘1“? " u-;g-$ i; rACI‘O-“‘ nU C. more we. as 79 Thin-A—Gain, however, red pine in disease plot TG5 grew more vigorously than the non-disease trees examined from 1963 until 1979. After 1979, growth for non-disease trees exceeded the growth of disease-plot trees at Thin—A-Gain (Figure 3-2). At Minnie Lake and Thin-A-Gain, growth for both the non-disease and the disease-plot red pines declined in 1988, a severe drought year. Annual growth for non-disease red pines, however, increased after the 1988 drought period but continued to decline for the red pines located within disease plots (Figures 3—1 and 3~2). A notable growth spike is apparent at Thin-A-Gain during the early 1980's (Figure 3-2). There is evidence from the cutover stumps in these stands that a thinning occurred around this time, releasing the trees from suppression. Disease Progress Figure 3-3 shows disease progress annually in terms of intensity and cumulative mortality for all five disease plots. The plot shows that pockets are expanding each year as more dead and declining trees appear in the plot. One possible exception is T65 which remains nearly unchanged during 1991-1993. To determine the rate of disease increases or decreases each year it is necessary to plot the change in disease intensity between years. Figure 3-4 shows that the disease rate may increase, decrease, or remain constant with time. :55 5.22. 05¢ 80 +MLC +ML1 +ML2 +ML3 Ring Width (mm) ssuwesmmmmn 7273747578777879$818283845$8788$83919293 Year Figure 3-1. Average annual growth for disease-plot (ML1, n=2; ML2, n=3; and ML3, n=3) and non-disease trees (MLC, n=3) at Minnie Lake. +1IT+IT$IIID 4. 3 2 .EE. 52>) uni 81 6.... .. ... .... I- +TGC +TG4 +TGS 5..— a4“ . E '1 5 . g 3 E ih‘ ’>\ I‘ K 2 ' i \‘ l I l , ' \ I J I I I I r II I IIIIIII I IIIIIIIIIIIIIII 0 LIL i $645$6768®7D7172737475767778798381828384586878889m91 9293 Year Average annual growth for disease-plot (TG4, Figure 3—2. and non-disease trees (TGC; n=3) at n=4; TGS, n=5) Thin—A—Gain. +t C. E +L I FL I .0 r 1 I g a mi 1 .i .3 a... e e .. u e .. a I D .b. a C. S r H .II .blu «L . n.u wu; n... N3 1 a. . 1 hp 41 “1.. I E ~8 C 3. .l I .l «j C I . - I 0 r3 . “I; § ‘ :u If‘ F g,» “kg «UJ a CI; 6 7..“ .‘I‘ m .U , . I S r .i I O S a t m . VW' _..H._ 4; \Pl. C . h 82 Thus, although pockets appear to be expanding in Figure 3—3, this does not mean that the disease rate is increasing each year. For example, little change in disease rate occurred between 1991 and 1992 for T65, T64, or ML1 and the disease rate actually decreased, i.e., some trees improved in condi— tion, for most plots between 1992 and 1993 (Figure 3-4). Even though mortality was the highest in 1992 and the lowest in 1995 (Table 3-4), the rate of disease change was the highest between 1994 and 1995. This can be explained by the fact that several trees declined further i.e., entered higher rating classes in 1995. Disease intensity was the highest at ML1 from 1991-1994 (Figure 3-3). This was the largest mortality pocket ob- served. ML3 exceeded the disease intensity of ML1 in 1995 (Figure 3-3) and also had the highest rate of change between 1994-1995 (Figure 3—4). ML2 had the lowest intensity ratings in 1991-1993 (Figure 3-3) but also had a higher rate of change than ML1 between 1994-1995 (Figure 3—4). The Durbin—Watson statistic (d) was used to determine serial correlation for disease intensity in Figure 3—3. For all plots d was < 1.5 indicating positive serial correla- tions (ML1, d=0.84; ML2, d=0.58; ML3, d=0.76; T64, d=1.01; T65, d=1.00). Positive serial correlations suggest that there are specific patterns to disease progress which are time-dependent factors. However, with only three years of data available for Thin-A-Gain plots the d statistic is less reliable than for the Minnie Lake plots. More years of III-Ill 8. 1 6. 1 4'2 Cc 36:95 1 8. 0 C... 0 ... 0 0 9mm 33:35 0 ( e O L No.3 FL 83 2 + 1.8 ML1 -—+—- 1.6 E “"52 a. 14 +— 8 ML3 %% 1.2 - 1 ‘1 1134 19'91 19'92 1393 19'94 1995 Year Figure 3-3. Disease intensity in terms of the average of the rating scale. 0.4 ., 0.35 f m 013 // ——F— g; 0.25 / if: g 02 / } ML3 g 0.15 // 1T»:— " 0.. - // I: 0.05 M/ m 0 A .. 91:92 92193 93:94 94195 Figure 3-4. The rate of change in disease intensity (AY/AT). * u v 80' vi; H.» .1 CO C. Q» a; 9 u )5. 5 A} a S C? CO C. ”A TI .Id » - u five~' 84 study would be needed for both Minnie Lake and Thin—A-Gain plots to increase the reliability of the d statistic (Bowerman and O'Connell, 1987). However, the data currently available for these plots does indicate a trend toward positive serial correlation. Spatial Analyses of Disease Progress Maps were generated of disease ratings for each year of the study. Figures 3-5, 3-6, and 3-7 are maps of disease ratings from Table 3-1 for the Minnie Lake plots between 1991 and 1995. The increase in the size of pockets is not as obvious in maps between subsequent years and Thin—A—Gain maps show no obvious difference in disease progress. Minnie Lake maps, however, do reveal changes especially over the five-year period. Figure 3-8 shows the results of spatial autocorrelation analyses. Of the three distance classes, the < 5 m category had the highest Moran's I value, i.e. the most strongly correlated. As distance increases, from 5 to 10 m and greater than 10 m, spatial autocorrelation values decrease. This suggests that trees nearer to each other are more closely related in terms of disease rating and that the patterns are not random. This result was evident in all plots examined. The variances in Table 3-5 correspond to each correlation coefficient for each plot and distance category in Figure 3-8. These variances indicate that, for the most guitar?- 85 a 1991 80 AAAAA A AA+++ AAA A AA A AAAAA H» u A +0 AAeAA x+ eAAA 70 ‘0‘+ao:xx I-‘+ o AMAA. ARatingo +1-1- - - II I AA eRatIng1 50 alfllgl I I"‘x*;h +Ratin92 504: x xI. I ++ +4., xRating3 g 3% ' I x 3* Us?» IRating4 “ x x _ €40»me .' x I4-x nun-Lift!- x 5‘ 30 I x- 1" A in h' 6+3: 0.. was A 20 «I.» A AAA A AAA I “‘ A“. AA‘ AAIA I A AAA+IAA AA AAAA 10 A A AA AAI AA IA IA A A AA AA 0 10 20 30 40 50 60 meters b) 1995 ARatingo oRating1 +Ratin92 xRating3 g lRating4 8 E 0 10 20 30 40 5O 60 meters Figure 3—5. Disease ratings at ML1 in a) 1991 versus b) 1995. 86 a) 1991 60 1.1-1.---“ 5 - - 8 ~ ~- ~- - M ~3 50 d_ A A I I IA Ae ;, A A no A e. A e I A A A ' 40 h A I A +A : B I“ x x I A ‘ +5 _ 53 30 H)- .‘X ‘ . A ‘0 ‘ ‘ ‘Ratmg 0 E A A): x I IA A AAA .Ratfnw + +‘ ‘ “ +Ratln92 20 ‘” ‘ ‘ ‘ ‘ xRating 3 A A‘AA A A A ‘A AA .Raflng4 A A A A A 10 I A A I ‘A ‘ ‘ A ‘ A A A A AA .» 0 .A a I q A t I 0 10 20 30 40 50 60 meters b) 1995 50 fi_ A A I I H». Al I I A A All I I A s 40 4 x I A e + ' . A I A IA . 2 .4 A . I I ‘ O I: ARatIng 0 3 30 4- I“ ‘ I A ‘e e ‘ eRatlng1 E A Al IIIA A eAA +Ratrn92 20 q_ + +A A A. xRatmg 3 6 HXA A e“. “A “ lRatmg4 x A A ‘ AA ‘ ‘ A 10 -1- A A A A A ‘ A ‘ A A A‘ AA 0 1O 20 30 40 50 60 meters Figure 3-6. Disease ratings at ML2 in a) 1991 versus b) 1995. 87 a) 1991 45 + A I * I A I e 40 ” x + I A I x x A I x 35 4’ A A A A I I x x z . 2 30 ”T x l x A Rating 0 o A A ‘ ’0. x . A Rating 1 g 25 T . .x x . + Rating 2 20--‘A :‘r. I u I °“ xRating3 15 v A A o + I I I I A I Rating 4 10 p ‘A AA A‘ A+ A‘+ AA + A ( 5+t“t‘At AA A A‘A‘AA 0 A A A A A A A A 0 10 20 30 40 50 meters b) 1995 45 x I I I x I + .1 40 I x I +I I II 35 A_ I + I x x o + + + I I I I g . E 30 4% g . I A Rating 0 3 25 3 ‘ . . x i o Ratmg1 g A. ::I-l .- : +1. E+Ratin92 20 «3 .. . I . 4 + ‘ x Rating 3 15* A x I II I I + ;IRating4 10 _4_ ‘A M A‘ .X .X. +L - A 3 511““: A . mu. O _ A A A A A A A A 0 10 20 30 40 50 meters Figure 3-7. Disease ratings at ML3 in a) 1991 versus b) 1995. 00000000 9 9:95.): Cozwgwtoo flgUIe h . 0.8 07 éEE?§\\§\ '§ 3:: Ek\;:::§:::\\ E 0'. \\. E; (is —‘Ei§>:\\ g at \§§§§:\ (J '0 \Q§\ <5 5t61o >10 Distance Category (meters) + ML1—t— ML2 + ML3 45‘- TG5 Figure 3-8. Moran's I values for three distance categories for disease plots (ML1, ML2, ML3, and TGS) in 1991. Table 3-5. Expected I values and variances for three distance categories at disease plots in 1991. _fl i_l ML1 ML2 ML3 T65 n2352 trees n2119 trees n=132 trees n2243 trees Expected I= Expected 1: Expected 1: Expected I: -0.0028 -0.0085 -0.0076 -0.0041 Distance category Vhriance variance variance variance (k) 82 82 $2 82 <5 m 0.00076 0.0032 0.0021 0.00054 5 to 10 m 0.00024 0.0012 0.00077 0.00022 >10 m 0.00029 0.00022 0.00017 0.000053 89 part, observed and expected I values do not overlap and the differences are significant. Moran's I values for the < 5 m and 5 to 10 m categories were statistically higher than the expected I values. However, in Figure 3—8, it can be seen that in the > 10 m category Moran's I values become negative and much lower than the Expected I values in Table 3-5. This indicates that the > 10 m category was too far of a distance to detect spatial patterns and was not significant. DISCUSSION Tree ring analyses suggest that growth of red pines at both Newago County locations declined as a result of the 1988 drought. This drought had an effect on both disease plot and non-disease plot trees. However, the main differ- ence is that the disease-plot trees did not manage to recov- er from the drought but continued to decline further. The largest pocket decline site was ML1. This site had the greatest number of dead standing trees and the highest stand density of Minnie Lake disease plots. Figure 3-3, indicates that ML1 has the highest level of disease intensity between 1991 and 1994, and is exceeded by ML3 in 1995. However, the graph of disease intensity rates (Figure 3-4 indicate that the smaller pockets, ML2 and ML3, are expanding at a faster rate than ML1. Even though disease levels vary between plots, the trend, or the increase or decrease in disease each year, is very similar for plots located at the same site. This sug— t a: e C.. AC a 5 i + .. ‘. a .AU 1 i ly h t *9 7... a. f"‘ V. afa .I .m: T S E E a C. C C, .i 11. a V h Xv. «b 1 i a t I e e t e d E t t e C S w I S U G .7. a I .I O 4 i D U a t x u E b. O a D. .0 F. f E S t. d t m b. 9O gests that some factor is controlling the pattern of disease at a particular site. Thin-A-Gain followed a trend which differs from Minnie Lake. These pockets had apparently stopped expanding to any significant degree during the years studied (1991 to 1993). Soil pH's, on average, were slight— ly higher at Thin-A-Gain than at Minnie Lake (Figure 5-3). It may be that whatever threshold values of factors such as pH, nutrient contents, insect and pathogen populations, adverse climatic factors, etc. are needed to tip the scale in favor of disease, no longer existed at Thin—A—Gain. TG4 was the only plot in which poor soil drainage may have been a factor that initially lead to the decline. It may be that at TG4 the pocket stopped expanding once the red pine in the poorly-drained area had died. The Durbin-Watson test revealed that the pattern of disease progress within a plot is a time-dependent factor. However, there are many approaches to time series data and further years of study are needed to characterize the dis- ease adequately. Linear regression is one approach to the study of time series data that may be useful in analyzing the spread of RPPM. Regression techniques could be used to develop an equation for predicting the spread of the disease that could be applied to the entire population of red pine mortality pockets. In order to do this, the equation would have to based on randomly selected mortality pockets and on data collected over several years. This equation may be in the form of a multiple regression equation which would IPA; ‘Vilrl ‘. 91 include other factors such as weather patterns and pest populations. In our limited study of mortality pockets, linear regression equations could be developed from the time series data but they could only be applied to each individual pocket studied and not generalized to the entire population of mortality pockets. The Minnie Lake maps of disease ratings show that pockets do expand over time. This expansion is not very apparent in year—to-year comparisons as it is in the five- year comparisons because the disease spreads slowly. The pocket margin is often confined to trees that are already declining and there is little expansion into the healthy tree population. In 1992, most of the increase in disease was explained by increased mortality of trees that were already sick in 1991. It was not until 1995 that the disease appeared to spread significantly into the healthy tree population. The high rate of disease intensity in 1995 is better explained by increasing symptoms of RPPM rather than by increased mortality. The maps show that already sick trees tend to decline further but that healthy-looking trees near pocket edges may not show symptoms for several years. The fact that there is a significant spatial autocorrelation of disease ratings in red pine mortality pockets indicates that the disease spread is not random. Individuals closer to one another are more likely to be affected by the disease than those spaced further apart. H VI AA -e S? 3.. Lu \— {Chant A his a w, . .l t .. t .. L r i «G .d p . 1 V ”b .I-V"A wh and ;: howe -' Z REP in T'n and 18kg. 92 This pattern suggests activity by a biotic agent or agents which cause infection from one tree to another. Root grafting is reported to be common in red pine plantations and infection may be spreading via the root system. However, RPPM spreads slowly suggesting that the disease agent(s) are not very aggressive and/or virulent. In our studies or RPPM, we have not been able to identify any single causal agent that has been consistently present in all pockets. Another possibility is that RPPM results from the interactions of environmental stresses and biotic agents. RPPM in Michigan has only been found in even-aged stands that were originally planted on marginal farmland sites. The soils are all sandy Spodosols of low pH. It may be that soil nutrient factors and acidification over time has stressed the trees, predisposing them to attack by insects and fungi. Significant amounts of fine-root mortality have been observed in mortality pockets versus healthy stands (Chapter 2). Whether this mortality is the result of nutrient factors, biotic agents, or a combination of the two is unknown. Climate variables, such as precipitation, may offer a partial explanation for the pattern of disease at Minnie Lake. The Palmer Drought Severity Index or PDSI is a mea— sure of drought severity and is reported on a monthly basis (National Climatic Data Center, NCDC). Positive values of PDSI indicate wet conditions and negative values indicate 93 dry conditions. The PDSI indices for the Newago County region of Michigan reveal that both 1991 and 1993 were non- drought years (i.e. positive monthly values). Disease intensity rates (Figure 3-4) decreased in 1993 because trees were improving in condition and because mortality rates decreased for some plots (Table 3-4). The PDSI indices also reveal that a mild drought oc- curred in 1992 from May through August (PDSI between —1.0 and —2.0). An incipient drought occurred in 1994 in which the PDSI indices were between -0.5 to —l.0 from March though June and September though December. The "incipient" drought in 1994 contributed to even drier conditions in 1995 in which the PDSI was in the mild (-l.0 to -2.0) and sometimes reached the moderate (-2.0 to -3.0) drought severity range from January though September (NCDC). Mortality was highest during the drought year 1992 (Table 3-4). After a decline in disease rates for most plots in 1993, rates increased somewhat in 1994, a slightly dry year. Disease rates increased substantially in 1995, an even drier year (Figure 3—4). Thin-A-Gain stays nearly level from 1991-1993 but it is suggested that the disease is no longer having an adverse effect on these plots. These observations suggests that during drought years, declining trees are likely to be pushed over the mortality threshold or to show greater symptoms of decline. Seemingly healthy trees near the pocket edges may not show visible symptoms of decline but perhaps the disease syndrome is I I .II .. w. ._ .r. r; T; .C .i ... ‘i g. e 3... + c st a .flu e r. e E + c .J .l C. E .3 I e s . t i S C .6 mm 5 C e O S a cl ”.1 I e «I .1. .J .nl. m .U U a S ‘C kt. S .C S T. ; Q q 94 affecting non—visible parts of the trees such as the root system. After years of compound stresses, including drought, symptoms may show up in the crowns of previously healthy-looking trees as was apparent in 1995. Plotting disease progress by looking only at crown symptoms may be misleading because it is not known how the disease is affecting non-visible parts of the tree. Factors such as fine—root mortality may be a better indicator of how the disease progresses. It will require more data points (more years of study) and a more accurate and invasive observation of symptoms to characterize this disease ade- quately. 95 BIBLIOGRAPHY Bowerman, B.L. and R.T. O'Connell. 1987. Time Series Forecasting, Unified Concepts and Computer Implementation, Second Edition, PWS Publishers, Boston, Massachusetts, 540 pp. Campbell, C.L. and L.V. Madden. 1990. Introduction to Plant Disease Epidemiology, John Wiley & Sons. Cliff, A.D., P. Haggett, J.K. Ord, K.A. Bassett, R.B. Davies. 1975. Spatial autocorrelation, In: Elements of Spatial Structure: A Quantitative Approach, Cam— bridge University Press, Cambridge, pp. 145-180. Cliff, A.D. and J.K. 0rd. 1981. Spatial Autocorrelation, Second Edition, Pion, London. Durbin, J. and Watson, G.S. 1951. Testing for serial correlation in least squares regression, II, Biometrika, 38, pp. 159-178. Grossman, G.H. and K. Potter—Witter. 1990. Changing red pine markets--changing forest management? Northern J. Appl. For. Grossman, G.H. and K. Potter—Witter. 1991. Economics of Red Pine Management for Utility Pole Timber. Northern J. of Appl. For., Pg 22-25. Klepzig, K.D. and J.C. Carlson. 1988. How to identify red pine pocket decline and mortality. USDA For. Serv. NA- GR—19. Klepzig, K.D., K.F. Raffa and E.B. Smalley. 1991. Association of an insect-fungal complex with red pine decline in Wisconsin. For. Sci. 37:1119-1139. Klepzig, K.D., E.B. Smalley and K.F. Raffa. 1995. Dendroctonus valens and Hylastes porculus (Coleoptera: Scolytidae): vectors of pathogenic fungi (Ophiostomatales) associated with red pine decline disease. The Great Lakes Entomologist, 28 No. 1, pp. 81—87. 96 Ott, L. 1988. An Introduction to Statistical Methods and Data Analysis, Third Edition, PWS—Kent Publishing, Co, Boston, Massachusetts, 835 pp. Raffa, K.F. and D.J. Hall. 1988. Seasonal occurrence of pine root collar weevil, Hylobius radicis Buchanan (Coleoptera:Curculionidae), adults in red pine stands undergoing decline. Great Lakes Entomol. 21:69—74. Slatkin, M. and H.B. Arter. 1991. Spatial autocorrelation methods in population genetics, Am. Nat. 1991, Vol. 138, pp. 499-517. United States Department of Agriculture. 1990. Plant .Hardiness Zone.Map, Miscellaneous Publication #1475. United States Department of Agriculture. 1995. Soil Survey of Newago County Michigan, USDA Soil Conservation Service, Forest Service. ." V‘_ a L K CHAPTER IV Analyses of Soil, Needle, and Root Samples from Red Pine (Pinus resinosa) Mortality Pockets for Determination of Nutrient Stress ABSTRACT Soil, needle, and root samples from red pine mortality pockets within the Huron-Manistee National Forest, Newago County, Michigan were analyzed for nutrient content and compared to those from healthy red pine plantations at Huron-Manistee; and between soil and root samples from healthy plantations at the Houghton Lake State Forest, Roscommon County, Michigan, and natural stands at Osmun Lake, in Cheboygan County, Michigan. Our initial hypothesis was that aluminum (Al) stress in the acid soils (pH < 5.0) may be a predisposing factor leading to the decline. The parameters of primary interest were the calcium (Ca)/Al ratios of soil, needle, and root samples for evaluating the risk of Al stress to the forest ecosystems. According to the criteria established by Cronan and Grigal (1995), Ca/Al ratios of soil and needle samples were within the threshold values identifying the red pine stands as being at risk for Al stress. However, this was true for both declining and healthy red pine stands. 97 98 Statistical analyses indicated that manganese (Mn), followed by magnesium (Mg) concentrations in soil, root and needle samples were the nutrient factors most strongly associated with areas of decline. Manganese was significantly lower in the needles and roots and Mg significantly lower in the soil and roots of diseased versus healthy plantation plots. Increasing gradients in Mg and Mn concentrations were observed from disease edges to the outer "healthy—looking" regions of disease plots to plantation check plots nearby. Nutrient studies of fine roots (5 3 mm in diameter) from the Huron—Manistee forest alone also revealed significant differences and increasing gradients in ratios of Ca/Al, Mg/Al, potassium (K)/Al, and Mn/Iron (Fe). In addition to Mg and Mn, K concentrations were found to be significantly lower and metals Al, Fe, and Zinc (Zn) were significantly higher in fine roots from Huron—Manistee disease plots. Fine roots may be reflecting unfavorable nutrient conditions within disease plots. Pine litter accumulation in mature even-aged red pine plantations and subsequent soil acidification may have increased leaching and loss of soil nutrients. Manganese, in particular, was consistently lower in both needles and roots of declining red pines. The gradients in element concentrations and ratios identified in soil and root tissues from declining to healthy regions of disease plots suggest that the pattern of disease spread is associated with nutrient problems. 99 INTRODUCTION Soil, needle, and root samples from red pine (Pinus resinosa) mortality pockets were analyzed for nutrient content and compared to the nutrient contents of soil, needle, and root samples from healthy red pine plantations in order to identify potential predisposing stressors contributing to a decline disease known as Red Pine Pocket Mortality (RPPM). RPPM affects mature plantation-grown red pine in the Lake States (Wisconsin, Illinois, and Michigan). Symptoms of the decline include reduced diameter and height growth, thinning crowns, and browning and stunted growth in needles. Red pine mortality pockets are characterized by an open area of dead standing trees ringed by declining trees. Increased growth of understory vegetation can be found in the center of the pocket (Klepzig and Carlson, 1988). The causes of this disease are unknown. This study examines the role of abiotic factors in the decline including the role that Al may have as a predisposing stressor causing mortality of fine roots. The possibility of deficiencies/toxicities of other elements including some micronutrients in soil and plant tissues is also examined. Aluminum is being considered because in acid, inorganic soils below a pH of 5.2, monomeric and trivalent exchange- able forms of Al predominate (Ulrich, 1989; Arp and Ouimet, 1986; Ulrich, 1986; Foy, 1974). In these forms, Al may become water soluble and available for plant uptake. Therefore, in acidic soils, Al3+ ions taken up by plants may as I 4) J] n 4 . . IE 1 C C C l .. i C .. l S .c C .l S a E C F .. .. i t t C r e e e a h .. i S n .1 r r. C t e O r n E a V .I r i e V. t .I B d .l n... A-.. 0 me n.) t .1 CL ti 6 .I .l h .l .l e O r. r 0 u .1 CL x} S E U 0 ¢ « O 100 have a direct toxic effect to the roots, particularly in plant species that are intolerant of aluminum. However, pines are adapted to growing on low pH soils and are considered to be Al—tolerant (Schaedle, 1989; Hutchinson et al., 1986). If red pines are affected by high levels of exchangeable Al, then they are more likely affected by the indirect toxic effects rather than the direct toxic effects. The indirect toxic effects relate to the competitive nature of trivalent Al3+ ions which displace divalent Ca2+ and Mg2+ ions and perhaps other nutrients within the root zone. Thus an "aluminum-induced calcium deficiency", for example, is of primary concern when considering toxic effects of aluminum in acid soils (Shortle and Smith, 1988). For these reasons, the parameters of primary interest when studying the possibility of Al- toxicity in a pine plantation are the ratios of A1 to other nutrients, especially Ca, rather than the actual concentra- tions of A1 in soil and plant tissues. Cronan and Grigal (1995), by reviewing the literature of Ca/Al ratios as indicators of stress in forest eco- systems, ascertained four measurement end points to be used as ecological indicators for identifying threshold values beyond which the risk of forest damage from Al stress and nutrient imbalances increases. The four measurement end points identified were: 1) a soil base saturation < 15% of effective CEC (cation exchange capacity); 2) a soil solution molar Ca/Al ratio 5 1.0 (for a 50% risk of forest stress); need; Wheth aSSOC “76 re 101 3) a fine root tissue Ca/Al molar ratio 3 0.2 (for a 50% risk); and 4) a foliar tissue Ca/Al molar ratio 5 12.5 (for a 50% risk). These threshold values are recommended as guidelines to identify sites where Al stress is likely to affect tree growth adversely. In this study, soil, needle and root samples collected from red pine mortality pockets and also from healthy non-disease plantations are compared to Cronan and Grigal's four threshold values to determine whether there is a significant risk of forest stress at these sites. Concentrations of Al, Ca, Fe, K, Mg, and Mn in soil, ‘ needle, and root samples were also examined to determine whether deficiencies/toxicities of these elements are associated with the decline. Soil P and needle and root Zn were also examined. There is sparse literature available on micronutrients and the growth of pines. Manganese deficiency has been shown to limit the growth of Pinus radiata in South Africa (Grey, 1988). The deficiency was associated with strongly podzolized soils and good drainage; similar to the Spodosols examined in this study. In North America, young slash pines (P. elliotii) growing on coastal plain soils in northeastern Florida were fertilized with micronutrients (Cu, Zn, Mn, Fe, B and Mo). Manganese was the only micronutrient which elicited a significant growth response in the slash pine stands. The Mn stress was described as subacute (i.e. there were no visible symptoms in the foliage) and the stress was 102 found to be associated primarily with Spodosols (Jokela, et al., 1991). Reported symptoms of Mn deficiency in the South African Radiata pine stands are similar to the symptoms observed in red pine mortality pockets. The symptoms included stunted growth, yellow—brown or bronze needle color, and very sparse or short needles (de Ronde, et al., 1988; Lange, 1969). Although "pockets" of P. radiata decline where not reported in the South African studies, the Mn deficiency symptoms were reported to occur in "patches" at various sites in the southern and western Cape (Grey and de Ronde, 1988). Manganese is known to interact with iron oxides forming co-precipitates and application of Fe fertilizers may result in Mn deficiencies (Grey, 1988; Knezek and Greinert, 1971). Therefore, to discern whether a competitive relationship might exist between Fe and Mn, the anFe ratio at healthy and unhealthy red pine sites has also been determined in our study. METHODS Study Locations Five red pine decline sites as identified by the U.S. Forest Service were selected for this study. Research plots were established within the area of decline and for an area outside the pocket. These plots are all located within the Huron-Manistee National Forest in Newago County, Michigan. Three disease plots are located at an area known as Minnie 103 Lake (ML) and two other disease plots are located about 17 km east of Minnie Lake in an area known as Thin-A-Gain (TG) (Figures 2—1 and 2-2). In addition to disease plots there are also two plantation check plots of healthy red pine located near the disease plots: one at Minnie Lake and one at Thin-A—Gain. To allow comparison of plantation-grown red pine to natural stand red pine, three check plots of healthy, non—plantation red pine were established near Osmun Lake at the Pigeon River Country State Forest in Cheboygan County, Michigan. The Osmun Lake stands are about are about 225 km northeast of the disease plots. For comparison with plantation—grown red pine located in a forest other than Huron-Manistee, two additional check plots of healthy red pine were established at the Houghton Lake State Forest in Roscommon County, Michigan. The Roscommon plots are about 115 km northeast of the disease plots. In sum there were twelve plots used in this study: 5 disease plots; 2 plantation check plots located in the same forest; 2 plantation check plots located in another forest; and 3 check plots in natural red pines stands. Similarities between plantation check plots were that they consisted of mature, even-aged red pine that had been initially densely planted. The natural stand check plots, however, were widely spaced and uneven-aged. All stands selected were growing on sandy Spodosol soils of low pH. i 3 CG .r. C ‘3 . . .. Hy .A 4 . r r\\ .. ‘ I E T. .3 C 1 AU :L AW r“ I! C. «a l c l I a :3 I C G . i .L A c no 3 O n c .3 u.” S T . C . i 104 Soil Sampling Soils were sampled in check plots (MLC and TGC) and at the Osmun Lake natural stands (01, 02, and O3) to a depth of 46 cm. Each plot was sampled randomly, 30 times throughout, and soil samples were separated by horizon (i.e. soil color). The 30-composite soil samples consisted of three different horizons. Thus, 15 soil samples were collected at Huron-Manistee check plots and at the Osmun Lake natural stands (5 plots X 3 horizons). Huron—Manistee disease plots (ML1, ML2, ML3, TG4 and TG5) were sampled by the same method except that 15- composite soil samples were taken at three different locations within the plot: at the center, edge, and outside regions of pockets. Thus, 45 additional soil samples were collected at Huron-Manistee disease plots (5 plots x 3 horizons X 3 locations within the plot). The following year, 1994, three lS—composite soil samples were taken randomly at the Roscommon plots for a total of 18 samples (2 plots X 3 horizons x 3 samples). All soil samples collected were dried at 3STIfOr 48 hours prior to analyses. Soil Analyses Soil pH, calcium, magnesium, phosphorus, and potassium content were determined according to the procedures utilized by Michigan State University's Soil and Plant Nutrient Laboratory. Potassium (K), Ca, and Mg were extracted with ammonium acetate (BHEOAC) and pH was determined in distilled 105 water. Phosphorus (P) determination was done using Bray- Kurtz—Pl. Levels of exchangeable aluminum were determined in a separate laboratory facility using 1N KCL as the extractant; five grams of soil were added to 50 ml of 1N KCl, shaken for 30 minutes, and filtered using Whatman no. 5 filter paper. Aluminum levels were determined using plasma emission spectroscopy. In 1995, soils were also analyzed for manganese (Mn) and iron (Fe) content using 0.1 N HCl as the extractant. Complete results for all elements are listed in the Appendix, Table 1. Needle Sampling In late July and early August of 1993, symptomatic and asymptomatic trees were felled in Huron-Manistee disease plots and four branch ends were collected from the top, middle, and bottom sections of the crown for a total of 12 samples per tree. Some symptomatic tree crowns, because of their small size, were divided into only two sections (top and bottom) and eight branch ends were collected. In total 23 trees were felled; 12 trees at the Thin—A-Gain site and 11 trees at the Minnie Lake site. One-year-old needles were separated from two—year-old needles in the laboratory and two samples belonging to the same age class from each crown section were combined. Thus, in most cases, 12 samples came from each tree felled (3 crown sections X 2 combined samples per section X 2 age classes). The needle samples were dried 106 at 42%: for 48 hours then ground. In total 229 needle samples were analyzed for elemental concentration. Needle Analyses Analysis of needle samples was accomplished by using the perchloric acid digest method (MSU Forestry Dept. Procedures, unpublished). Each 150 ml digestion tube contained 0.5 g of dried, ground, plant tissue. Samples were partially digested by placing them in 5.0 ml of nitric acid (HNCu) overnight before starting the acid digestion block. One-hour into the digestion process 2 ml of perchloric acid (HClCu) was added to each tube. Forty tubes were digested at one time. For quality control, random sample replications, blanks, and NBS standards were included in the digests. Element concentrations of Zn, Cd, Mn, B, Fe, Pb, Al, Mg, Cu, Ca, K, Ni, and Cr were determined using plasma emission spectroscopy (compete results in Appendix, Table 2). Elemental standards used for calculating regression equations were 0, .20, .40, .60, .80 and 1.0 of the total standard. Root Sampling Roots were collected in 1992 by digging 1-nfi pits in disease plots at three locations within the plot: in the center of the pocket where there were many dead standing trees; at the edge of the pocket where trees were starting to decline; and outside the edge of the decline where trees appeared healthy. The 1-nfi pits were divided into four 0.5 107 x 0.5 m sections and samples were taken from two levels: 0 to 30 cm deep, and 30 to 60 cm deep. Thus eight bags of root samples were collected from.each.l—nfi pit (4 sections x 2 levels). In total fifteen pits were dug, three at each of the five disease plots. In addition, two 1—nfi pits were excavated at each of the two check plots at Huron-Manistee and one pit from each of the three natural stand plots at Osmun Lake. Three of the disease plots were partially resampled in 1993 by digging 0.5 m x 0.5 m x 60 cm deep square pits (four 0.5qm square pits at ML3, three pits at TG4; and two pits at TG5); only outside and edge samples were collected at this time. Samples were taken from the two check plots at Roscommon County in 1994 using 0 saw pits. Samples were removed from two levels (two bags of roots per pit); four 1- HF square pits were dug at each of the two check plots. Because there were no features of center, edge, or outside area of a pocket within check plots, the pits were dug randomly at locations throughout the plots. The roots collected at both check plots and disease plots were extracted by sifting the soil through a mesh screen and separating the red pine roots from the roots of other plants growing in the area. Once collected the roots were brought back to the laboratory, washed by rinsing with water for 30 minutes, and sterilized by dipping in a 10% Chlorox solution for 2 minutes. The lengths and widths of 108 each root were recorded and they were also classified into live, dead, or symptomatic categories. Roots were called symptomatic if they were partially dead or dying and/or if there was any fungal staining below the epidermis. Live and symptomatic roots were cultured in attempting to identify the pathogenic fungi associated with the disease (Chapter 2). The roots were kept frozen until thawed for drying and grinding. Roots were re-washed by rinsing for 30—minutes after removing from the freezer. Roots were separated by diameter classes (1 to 3 mm and 3 to 5 mm) and oven-dried for 48 hours at 42%3. Only live and symptomatic roots were included in the analysis; dead roots were discarded. Thus, no samples from the center areas of disease plots are included in the analyses, since they consisted mainly of dead roots. Root Analyses Analyses of root elemental content was accomplished by the same perchloric-acid digest method used for needle analyses. The only change was in the standard used to establish the regression equation. Root samples contained higher concentrations of Al and Fe, and lower concentrations of Ca and K than needle samples and the standard was adjusted accordingly. Elemental concentrations of Zn, Cd, Mn, B, Fe, Al, Mg, Cu, Ca, K, Ni, and Cr are listed in the Appendix, Table 3. r. : in. v. vC C CL 6 +C r. a E A A. S f .. . i e C S V S C C C C. C n C e E i O 7 a S e -C a a O X S d .D. t (O f n f a 11 O f :1 l e 11.. f. t 109 Statistical Analyses Analysis of variance (ANOVA) was used to test hypotheses that soil, needle, and root element concentra- tions and ratios from disease and healthy locations are significantly different from each other. Other soil factors tested were pH, CEC (i.e. cation exchange capacity or the sum total of exchangeable cations that a soil can adsorb), and percent base saturation. For soil, the model used to test hypotheses that element concentrations and ratios differ by locations and soil horizons (Table 4-1) is as follows: Yijk=u+ai+fij+upij+€uk where variable Y is the element concentration/ratio or soil factor (pH, CEC, and % base saturation) per sample (k) associated with the i”‘level of factor a (location) and j” level of factor 0 (soil horizon), u=overall mean of all observations, aj is the effect of 1th level of the location factor, Bjis the effect of the jth level of the horizon factor, abij is the effect of the interaction between the location and soil horizon factors and any, is the experimental error associated with the kLh experimental unit (soil sample) for the levels of the location and horizon factors. There were six levels included in the location factor: the center of Huron-Manistee disease plots, the edge of disease plots, outside the edge of the decline within disease plots; healthy plantation plots at Huron—Manistee; 110 healthy plots at Roscommon County, and healthy non— plantation plots at Osmun Lake. Three levels are included in the soil horizon factor: an O horizon of a black to dark grey humus and roots layer; an E (albic) horizon consisting of a grey leached mineral layer; followed by a spodic B horizon which consists of a yellow-brown sand. Needle samples had to be treated somewhat differently from soil samples in order to test the differences between nutrient concentrations and ratios in healthy and unhealthy red pines. The 229 needles samples consisted of two samples in each age category (one-year and two-year), with one missing case, from each crown level position (top, middle, and bottom) of the tree (Appendix, Table 2). The element concentrations and ratios from the two like—samples from each tree were averaged together for a total of 115 cases for statistical analyses. Averaging the two like-samples increased sample independence because only one sample from each crown position of the tree was used in the analysis. Bimodal distributions and analysis of variance of the elements studied revealed that 1-year-old and 2-year-old needles should be treated as separate populations. Therefore separate ANOVA trials were run on samples from each age category. All samples came from the two locations within the Huron-Manistee National Forest (Minnie Lake and Thin-A—Gain) and samples were collected from unhealthy trees inside of disease plots and healthy trees outside of disease plots. The model used to test differences among needles 111 samples was the same as that used to test soil differences except that there are three levels to a., the effect of the ith level of the crown level factor (i.e., top, middle and bottom), and two levels to flj, the effect of the jth level of the location or site factor (i.e., Minnie Lake or Thin—A— Gain); a0ij is the interaction between crown level position and site. The red pines at Thin-A-Gain were ten years younger than the red pines at Minnie Lake and analysis of variance was used to test the differences between nutrient concentrations and ratios of healthy and unhealthy trees at each site. Separate ANOVA trials were run on red pines at Minnie Lake and Thin-A-Gain which included the crown level factor (a1) and two levels of the health factor Bj (healthy or unhealthy), where abij is the interaction between the crown level and health factors. At Minnie Lake there were three levels to the crown level factors: top, middle, and bottom. However, at Thin-A-Gain, only two levels to the crown level factor were included: top and bottom, because there were no unhealthy trees sampled at Thin—A-Gain which included a middle crown level. The analysis of variance model used to test differences between root nutrient concentrations and ratios is Yu=u+¢i+€m where ai is the effect of the location factor. Separate ANOVA trials were run on two different root-size categories: 5 3 mm in diameter, and > 3 to g 5 mm in diameter. In this model, ai the location factor, consists 112 of five levels: Huron—Manistee disease edges, outside the edge of the decline, Huron—Manistee healthy plantation plots, Roscommon healthy plantation plots, and healthy natural stand plots at Osmun Lake. Unlike soil analyses, disease centers were not tested in this model because they consisted mainly of dead roots and only live roots were analyzed for nutrient content. Analysis of variance was also used to test the difference between root nutrient concentrations and ratios of disease versus non-disease plots within the Huron-Manistee forest only. In this model (11 is the effect of the disease status of a plot: i.e., healthy or unhealthy. These models do not include testing of any interaction effects on root samples. The effects of the location and health factors studied in soil, needles, and roots are considered fixed because they are the only levels of interest in this investigation. If the experiment were to be repeated the same treatments would be included and randomness would be inherent in the replications (n) i.e. the number of root, needle, or soil samples collected. Analyses of variance were run using plots (i.e. ML1, ML2, ML3, etc.) as the levels of a factor. This effect is considered random because the disease pockets and healthy plantation plots are considered subsamples of the larger population of all potential disease pockets and healthy plantations. Under the random effects model, unlike the fixed effects model, inferences can be made to the 113 larger population (i.e., disease pockets and healthy red pine plantations in other locations). All sample distributions for both ratios and element concentrations were normalized by placing them on a log scale. Log-transformation of data strengthened statistical testing. However, for clarity, means and other descriptive statistics reported are back-transformed nonlog values. Means were separated at the 0.05 significance level using the Tukey's highly significant difference (HSD) procedure (Systat for Windows, Version 5, 1992). RESULTS Soil Nutrient Status Tables 4—1 and 4-2 show the results of soil analyses as separated by horizon and location within disease plots. Table 4-1 lists the pH, CEC, and % base saturation for locations within disease plots (center, edge, and outer) and for non-disease plots at Huron-Manistee, Roscommon, and Osmun Lake. The mean concentrations in mg/kg of six different soil nutrients: Al, Ca, Fe, K, Mg, Mn, and P are also listed in Table 4-1 and Table 4—2 lists the ratios of Ca/Al, Mg/Al, K/Al, and Mn/Fe. Analysis of variance revealed that significant differences at the 0.05 level could be found for the soil horizon factor for all variables in Tables 4—1 and 4—2 except for Mg (p=0.375). Magnesium concentrations were fairly constant regardless of soil horizon. The soil pH and 114 percent base saturation tended to increase from the O hori- zon to the B horizon (soil sampling depth was 46 cm) and the CEC tended to decrease. The organic material present in the O horizon explains the higher CEC's. The O horizon was also the most acidic and contained the highest concentrations of extractable Al. Soil nutrients tended to decrease from the O horizon to the B horizon except for P, which tended to increase (Table 4—1). Ratios (Table 4-2) also tended to increase from the O horizon to the B horizon. In many cases, however, the grey leached E horizon had the lowest nutrient concentrations and ratios (Tables 4-1 and 4—2). Significant differences at the .05 level were also found by location. The only variable which did not differ significantly by location was the soil Ca concentration (p=0.170). There were also few significant interaction effects between location and horizon except for soil Fe (p<0.001) and K (p=0.022). This indicates that the soil horizon concentrations of Fe or K depend on the location from where the sample was drawn. Iron (along with Al) tended to be highest at the healthy plantations at Roscommon. Potassium concentrations were also higher at healthy locations at Roscommon, Osmun Lake, and Huron- Manistee versus non—healthy locations (disease centers and edges). Tukey's HSD procedure was used to separate means at the 0.05 level in Tables 4-1 and 4-2. Tables 4-1 and 4-2 show the significant differences found within the same soil 115 3.3 6.83 60.3 36 ”SE 00.3 «mm 03 an: AS. $2 no?” m 23m 60.: am}: 6?: fiGm swig 2a.: mmm mmm mmm SS ”.3 33 m caumo no.3 no.2 3.: 63.3 $98.39? 3m Sam at? 33 33 33 N a. $5 Sim $5 $6 13.: 63.3 mos 86m £2. £3 “.8 mm: m amuse 2; £5 £5 aim 63:: 3.3 «we 64:. m8 mmm $3 32 m omen 35 85 SS 85 3:3 2.3 3m $8 £8 £2 £3 £2 m 323 as M on no .53 £3 amom 3.2 aim mmé 2.3V «T: max: mama mom; mama m 03m at? £62 “am: £9. 08.3. 3.2 3.6 $4.. med 48;.“ M.S.”. moo; m 5:50 “ENS £9: has? 2...? $93 £4: £4. 63G Aim 2:...“ 3m; smog. N 2: fig new: 8N2 3.: mm.» £5 34., A.» 3.3 moms mama $.me m gauge 6...; 32 £2 £4: 3a 35 £6 $6 3.3 mmofi men; 2.34. m ovum m? 22 £2 Siam smog: £6 £6 6&5, ”6.: 304.. 3G6 flame m “3:8 3 83333 23m a own an m _ m A o m _ m d o m _ w _ o m _ m — o a "conHou_ "nouwuom "acuwuou "Gouduom "nowuuuon .4558 mpcmum amusumc paw .A.omom paw ZEV mcoflumucmAQ mauammn .Auwuzo paw .oopo .nwucwov muofim wmmwmflp um mcoNHuon HHom quAmMMHU woman Eonw m paw c2 .02 .M tom .mo .Hd mo mx\oa CA wcofiumuucmocoo paw coflumusumm mmmb Damonom ..o.m.o 1mm cme .Huw magma 116 uwuuma mEmm map >9 .musbmooum mocwnwmwfln ucmoHMHcmHm wapmmcon m.>mxde ocflms Hm>wa mocmoHMflcon mo.o mnu um Hmmwflv uoa U36 Umzoafiom quHoo mamm map cflnuflz mammz* mm.vw mA.me mm.m mo.¢ mm.mH mm.vm m onom mném mm.m «dim mmé mm.m “36 m 5580 2.3m: mm.mm mm.mm mm.m mm.om mmdm m Eu mm.Hm mv.mm mm.mm mm.m mw.bfifl mo.vm m. .Houso mm.om mm.mm wH.mm m>.H mm.m Qmo.mm m Gown mo.mHHHmm.mm mm.>m mm.m mm.> an.oN m .Hauaoo m a! m N O m H 0 an Hauaumz "couwuom "nonauou 33A 5830 .AU.HCOUV HIw wanme 117 .mnSUmooum mocmanMHU 3c00303c03m >3ummco: m.>mxsu 0:30: 3m>m3 «0:003w3003m 00.0 «33 um 300036 30c U3U umuuw3 «80m man an Um3033om c833oo mEmm map G3£333 m:«m2« «03.0 «00.0 «030 «00.0 «00.0 «03.0 «03.0 «00.0 «30.0 «00.0 «033.0 «00.0 0 ouom «330.0 «03.0 «03.0 «80.0 «3.0 «3.3.0 «00.0 «33.0 «00.0 «00.3 «30.0 «00.0 m Eggo «30.0 «00.0 «00.0 Q«00.0 «33.0 «03.0 «0.3.0 «33.0 «00.0 «00.0 «00.0 «00.0 m 2H «03.0 «00.0 «00.0 «30.0 «00.0 «03.0 «33.0 «00.0 «00.0 «00.0 «033.0 «00.0 0 H035 «00.0 «03.0 «3.0.0 «00.0 «00.0 «33.0 «03.0 «00.0 «00.0 «00.0 «00.0 «30.0 0 ovum «3.0.0 «3.0 «3m.0 n«03.0 «00.0 «03.0 «30.0 «00.0 «30.0 «30.0 «00.0 «00.0 0 Hauaoo an}: .3? 3432 32.6 m N O m m“ nu m H O m N nu a "nouwuom "nonwuom ”nonwuom "nonwuou "cowuaooa .Acsemov mvcmum 3musumc ncm .3.omom new 230 mc03umucw3a >zu3mw: .33muso Ucm .m0nm «umucmov m303a mmmmm3u um mcoN3uon 330m ugmpm003n «««03 souw «03u«u «m\cz «c« .3¢\x .3<\0z .3¢\«o ««3oe «««2 .m.« «3n«e 118 horizon. In some cases (soil P, Ca/Al, Mg/Al, and Mn/Fe, for example) the table shows no significant differences even though analysis of variance indicated that the location factor was significant. This is because the significant differences at the 0.05 level were not found within the same soil horizon but across horizons. Tukey's HSD revealed that K and Mg concentrations were significantly higher at healthy locations (Roscommon, Osmun Lake, and Huron-Manistee) than diseased areas (centers, edges, and outer regions). Significant differences within the same soil horizon were also identified for metals Al and Fe which were higher at Roscommon than at disease locations but no significant differences were found between metals for locations within the Huron-Manistee National forest. Ratios of Ca, Mg, and K to Al also tended to be the lowest at Roscommon although the differences are not significant. For the most part, it appears that soil elemental concentrations (especially Mg and K, Table 4-1) are better indicators of differences between diseased and non-diseased areas than the soil ratios (Table 4-2). Soil samples were separated in disease plots according to location within the plot in order to determine if there were any gradients in concentrations of soil nutrients from within the Huron-Manistee National forest. There was an increasing gradient in soil concentrations of Mn, K and Mn/Fe ratios from the center of pockets to the healthy plantation plots at Huron-Manistee. However, for Mn and 119 Mn/Fe the differences were not significant (Tables 4-1 and 4-2). The fact that red pines growing at the same site (Huron-Manistee, Roscommon, or Osmun Lake) were the most similar in terms of soil nutrient contents, raises questions about the selection of appropriate control groups for comparison with RPPM—affected red pines. The plots at Osmun Lake were all natural stands. These stands, unlike plantations, were not densely—planted nor even-aged. Thus it seems that natural stands may not be an adequate control group for comparing to plantation stands. Roscommon stands, on the other hand, were similar to the Huron—Manistee stands in terms of being mature, even- aged, plantations. Variability still existed between Huron Manistee and Roscommon, however, in factors such as location, climate, stand density (Table 3-3), age, soil nutrient status, etc. It seems more appropriate to compare trees growing within the same forest that have adapted to similar site conditions over time. These stands would be more likely to respond in comparable ways to changes within the environment. Therefore, Huron-Manistee healthy plantations are considered to be the most appropriate control group for Huron-Manistee disease plots. Although nutrient data from Roscommon and Osmun Lake red pines are included in subsequent tables, the focus will be on the significant differences that occur between disease and non— disease areas within the Huron-Manistee National forest. 120 Needle Nutrient Status Needle samples were collected from disease plots and healthy regions at two locations within the Huron—Manistee National Forest (Minnie Lake and Thin—A-Gain). Unlike soil and root samples, no needle samples were collected from a forest other than Huron-Manistee. Tree crowns were separated into top, middle, and bottom sections and crown symptoms were used to separate "healthy" trees from "unhealthy" trees. The number (n) of needle samples taken from unhealthy trees in the "mid" section of the crown is less than the number collected from healthy trees because of their smaller crowns that were separated into only "top" and "bottom" categories. Table 4-3 shows the results of analysis of variance which tested differences between element concentrations and ratios from the three crown-position categories at two sites (Minnie Lake and Thin-AeGain) for l-year—old and 2-year—old red pine needles. Site (Minnie Lake or Thin-AeGain), apparently, was a strong factor accounting for differences in needle elemental concentrations. For example, in l-year— old needles significant differences were found for Al, Fe, K, Mg, Ca/Al, K/Al, Mg/Al, and Mn/Fe by site. In two year— old needles, only Mg, Zn, and K/Al differed significantly by site (Table 4—3). Table 4—3 also shows significant differences by crown level. In l—year-old needles Mn, Ca/Al, and Mn/Fe differed significantly by crown level. In 2-year—old needles Ca, Mg, 121 Table 4-3. Analysis of variance} of element concentrations and ratios by site (Minnie Lake or Thin-A-Gain) and by crown level (top, middle, or bottom) for l-year-old (n=59) and 2-year-old needles (n=56). ! l-gear-old needles . Z-gear-old-needlea Site Level Site Level Blame P- F- F- F- ent ratio P ratio P ratio P ratio P A]. 8.28“ 0.006 1.44ns 0.245 2.63ns 0.111 0.93ns 0.402 Ca 0.001ns 0.981 2.49nsl0.092 0.36ns 0.551 7.13** 0.002 Fe 5.43* 0.024 0.99ns 0.377 1.09ns 0.301 0.77ns 0.470 R 11.97“ 0.001 0.81ns 0.452 2.96ns 0.092 0.36ns 0.699 Mg 4.41* 0.040 1.84ns(3.169 12.98** 0.001 4.80* 0.012 :Mn 2.31ns 0.135 3.54* 0.036 0.90ns 0.348 5.05** 0.010 Zn 0.96ns 0.331 0.75ns 0.476 7.11** 0.010 2.07ns 0.137 Ca/Al 11.60** 0.001 8.66** 0.001 0.49ns 0.487 13.80** (.000 KIA]. l6.20** <.000 1.89ns 0.161 5.35* 0.025 1.19ns 0.314 ug/Al 4.98* 0.030 2.51nst3.09l 0.27ns 0.605 4.34* 0.018 Mn/Pe 15.10** <.000 4.59* 0.015 3.54ns 0.066 8.03** 0.001 1Significance:*=significant at 0.05;**=significant at 0.01; ns=not significant; there were no significant site X level interactions. 122 Mn, Ca/Al, Mg/Al, and Mn/Fe differed significantly by crown level. Analysis of variance indicated that there were no significant site X crown level interaction effects (results not listed in Table 4-3). The significance of the site factor indicated that separate ANOVA trials should be run comparing needle element concentrations within a site. Table 4—4 lists the results of analysis of variance on the crown level and health factors for each site and age class. The corresponding means for the needle element concentrations and ratios tested are presented in Tables 4-5 and 4-6. Fewer significant differences were found for the crown level factor when sites were tested separately. Only Ca/Al ratios appeared to be strongly influenced by the crown level factor in both l-year-old and 2-year—old needles at a site (Table 4~4). Tables 4-5 and 4-6 show the actual mean element concentrations and ratios for healthy and unhealthy trees by site and crown level position for both age categories. Crown level position differences, for example, are illustrated by Ca/Al ratios which tended to increase from the top level of the crown to the bottom level in both 1- year—old and 2—year—old needles (Table 4-6). Differences between age categories can also be observed in these tables. For example, more physiologically active l—year-old needles have higher K concentrations and K/Al ratios than 2-year-old needles. On the other hand, 2—year-old needles have higher Table 4—4. and ratios by crown level2 and by health (healthy or unhealthy) for 1-year—old and 2-year-old needles at 123 Analysis of variance110f element concentrations Minnie Lake3 and ThinfiA—Gainf pocket decline sites, Huron-Manistee National Forest. _. _ _,——_. —_ __ ____._ Minnie Lake ' l-Xear-old needles Minnie Lake Z-year-old-needles Level Health Level Health Blem- F-ratio p F-ratio P P F-ratio P ent . 111 0.43nsO.654 4.88* 0.038 80nsO.462 6.51* 0.019 Ca 0.74ns 0.490 2.02ns 0.170 59ns0.099 0.56ns 0.461 Fe 0.53ns 0.599 1.02ns 0.324 77ns 0.086 0.06ns 0.636 K 0.42nsO.664 0.21nsO.653 15nsO.862 l.54nsO.228 Mg _ 0.55nsO.585 1.00ns0.328 17ns0.063 2.27nso.l47 Mn . 1.24nsO.31l 9.59**0.005 96ns0.167 4.98* 0.037 Zn - 0.58nsO.568 7.07* 0.015 30nsO.295 6.23* 0.021 Ca/Al 2.31ns 0.124 0.39ns 0.540 23**0.008 0.74ns 0.399 ; 0.58nso.569 2.50ns0.129 72nsO.498 6.52* 0.018 2 0.79ns 0.466 3.89ns0.062 32ns0.056 2.05ns 0.167 l 1.62nsO.222 1.86nsO.187 99ns0.072 3.84ns 0.064 ‘ Thin-A-Gain Thin-A-Gain 1::ear-old needles Z-Xear-old needles . 0.78nsO.389 8.37**0.009 0.07ns 0.802 0.87ns 0.363 i. 1.06ns 0.316 1.52nsO.233 5.86* 0.026 2.87ns0.107 ' 0.46nsO.504 9.91**0.005 0.08ns 0.778 4.03ns0.060 ; 0.48ns 0.496 0.03ns 0.875 0.09ns 0.762 <0.00ns 0.967 ~ 2.49nsO.130 0.02ns 0.892 4.23ns 0.055 1.23ns 0.281 ; 2.00nsO.17311.96**0.002 4.22ns0.055 16.14**0.001 ' 0.00ns 0.987 1.19ns 0.288 0.99ns 0.334 1.43ns 0.248 7.16* 0.015 l.42nsO.24812.44**0.002 l.60nsO.222 g l.58nsO.223 9.27**0.006 0.15ns 0.706 0.49ns 0.493 2.77nsO.112 9.95** 0.005 2.04nsO.170 0.04nsO.836 3.76n30.067 0.58nsO.454 7.21* 0.015 5.29* 0.034 1Significance:*=significant at 0.05;**=significant at 0.01; ns=not significant; there were no significant level X health interactions. Crown level top, middle, or bottom.at Minnie Lake; and top and bottom only at Thin-ArGain 3Minnie Lake: n=27 for l-year-old and 2-year-old needles 4Thin-A-Gain: n=24 for 1-year-old needles; n=22 for 2-year-old needles 124 3 R .3. 003 mm 00 ,0: 3.30 00mm 83 0.83 33 0mm 0mm 3 e eon m3 HN mvm mvm vm mm 000 Hmb mNHm nvmm mth 0mm: NNN NhN v 0 m0.._.. ooauooz odouuoomuw "oven addendrawna mN mv mom mNOH Nm .30 wmm 00.: mMHm bNbN ova mHmN mmN omN m m 98 0N ow mww mam mm 00 wow Nmm «mmN 33va NQVN mNmN va mNm N m OHS MN mN mow va am 03.. mmm mvw mmmN mth 03.3 wvma NmN 00m m m mOH. «036002 fideluoowwu "ouwm @303 owned: om em mew .003 .00. a m eon ON mN 00.3 mmm MN 3» CNN. For 303m 003m mNNH 0mg“ No.3 omH v m m9“. -noauooz_ono|unomwa “ouwm auuwlduoana MN Nm NmN m: :3 0v Now Nmm wav oamv mama mvma moH mHN m m H8 vN om HmN @Nv mm .3 N00 mmm wmmv MNNv PVNH mowfi v03 ®MN N m OHS ON 0N NON mmm 0v 3m 000 Mbh mva mm: 0mm OFNH mwa ObN m m mOB uoaoooz.oaonunomua "093m oxan owes“: o m _ o u p n o u _ o m o n o m A o a 325.3 oxouu w an a: on _ 02 M no .umouom 3chHumz mmum3cmzlcousm .c3mwuer3nB pom oxm3 03ca32 um 0:30 U03 ADV >533mona: paw 3m0 >333mm£ mo mm3omoc U3onumo>uozu paw Iwco CH cm Ugm .a2 .00 .02 .M .«o .3¢.mo 303\0EV mGOHumupcoocoo ommem>m paw 3:0 mmmmo wo nonfisz .010 o3nme 125 Table 4-6. Number of cases (n) and average molar Ca/Al, K/Al, Mg/Al, and Mn/Fe ratios in l—year—old and 2—year- old needles of healthy and unhealthy red pine, Huron Manistee National Forest. Crown Ce/Al Mg/Al K/Al Mn/Fe ““1 n U n U n U n U a U Minnie Lake Site: 1:¥ear-old.Needles TOP 3 8 3.31 3.63 3.36 5.80 12.03 20.17 7.86 5.12 MID 3 2 4.32 5.58 3.85 5.67 11.77 20.07 10.01 9.51 BOT 3 8 4.89 6.26 4.66 7.62 16.28 28.24 11.88 8.01 Thin-ArGein Site: l-geer-old.Needles TOP 8 4 4.44 6.27 4.97 8.27 22.76 36.18 9.47 12.16 BOT 8 4 7.50 6.86;;80 8.72 29é37 37.88 13.33 17.78 Minnie Lake Site: Z-zeer-old.Need1es TOP 3 8 3.39 4.69 2.60 4.16 4.91 8.46 8.96 5.33 MID 3 2 5.35 6.61 3.26 3.66 5.58 8.53 14.34 14.48 BOT 3 8 ‘7.13 7.04 «4h45 4.45 6.85 9.32 14.76 7.71 Thin-A-Gain Site: 2-geer-old.Needles TOP 7 4 4.47 44$“? 3.09 3.93 8.92 11.24 8.02 8.21 BOT 7 4 8.26 6.03 4.41 4.20 9.48 11.01 16.78 8.83 126 concentrations of Ca in reserve. These tables also allow comparison of elemental concentrations and ratios by site (Minnie Lake and Thin—A-Gain). For example, Al concentra- tions tended to be higher in the older red pines at Minnie Lake and K concentrations tended to be higher in the younger red pines at Thin—A-Gain (Table 4-5). Significant differences can be identified for the health factor for elements Al, Fe, Mn, and Zn and for ratios K/Al and Mn/Fe (Table 4—4). Contrary to expectations, metals Al and Zn were significantly higher in healthy 1- and 2—year-old needles at Minnie Lake, and Al and Fe were signi— ficantly higher in healthy l-year-old needles at Thin-A-Gain versus their unhealthy counterparts (Tables 4-4 and 4-5). This may indicate, however, that healthy trees are more efficient at assimilating and storing metals than unhealthy trees. Also contrary to expectations, K/Al ratios were significantly lower in healthy 2-year-old needles at Minnie Lake, and K/Al and Mg/Al ratios were significantly lower in healthy l-year-needles at Thin-A-Gain (Tables 4-4 and 4-6). This may be explained by the higher Al concentrations present in healthy red pines. As expected, Mn/Fe ratios were significantly higher but only in 2-year-old needles of healthy red pines at Thin—A—Gain (Tables 4-4 and 4-6). There were no significant interaction effects between the crown level and health factors (results not shown in Table 4—4). 127 The one nutrient which was consistently lower in unhealthy red pines was Mn. Manganese was significantly lower in the needles of unhealthy red pines at both sites (Minnie Lake and Thin—A—Gain) and for both age categories and also had the highest F values associated with it (Tables 4—4 and 4—5). These data suggest that Mn may play a significant role in the decline. Root Nutrient Status Statistical analysis (ANOVA) of red pine root tissues revealed that significant differences could be found between two root-size categories (5 3 mm in diameter; and > 3 and g 5 mm in diameter); and by plot location i.e. edge and outer regions of Huron-Manistee disease plots, Huron-Manistee healthy plantations, Osmun Lake, and Roscommon plots. The largest F values were for the root-size factor. Thus, these two populations are treated separately. No significant differences could be found by the soil depth from which the root sample was selected (0 to 30 cm; or 30 to 60 cm). Therefore, this factor was ignored in further analyses. Table 4—7 displays the element concentrations of both fine roots and large roots from Huron-Manistee, Osmun Lake, and Roscommon forests. The fine—root category (5 3 mm in diameter) is of primary interest because fine roots are expected to reflect soil chemistry and associated nutri- tional problems more distinctly than large roots (Ulrich, 1989; Murach and Ulrich, 1988). Analysis of variance 128 Table 4—7. Mean element concentrations (mg/kg) for disease plots and healthy plantations for both fine (3 3 mm in diameter) and large (>3 and 5 5 mm in diameter) red pine roots.* Fine Roots Element Concentrations g 3 mm (mg/k9) Location n A1 Ca Mg K Mn Fe Zn Disease Edge 30 1943ac 2662611) 889a l627ab 115a 429ab 56a Disease Outer 42 IL736a 2690ab 1067b 1388a 26lb 33lab 50ab HMCheck 26 109019 2850ab 1273c 1869bc 478C 247ac 38b Osmnn Lake 16 1034b 2313a 958ab 1863bc 296bc 220c' 37b Roscommon 16 2535c 3062b 1087bc 1924c 36lbc 416b 50ab W Large Roots Element Concentrations 3>§5 mm. (mg/k9) _L°t_°“__ Disease Edge 28 506a 1852a 775a 1646a 137a 144ab 40a Disease Outer 35 530a 2038a 981bc 1301a 257ac 175a 38a HMChOCk 23 5676119 2272ab1042c 1584a 45Gb 141a 33ab Osmun Lake .20 544ab 2670b 773a 1475a 272bc 906 26b Roscomn 16 7761: 2084.36 845ab 1696a 3671: 162a 33ab *Means within the same column and size category followed by the same letter did not differ at the 0.05 significance level using Tukey's honestly significant difference procedure. 129 revealed that significant differences could be found for element concentrations by location in both root-size categories with the exception of K concentrations in large roots at the .05 level. Tukey's HSD procedure was used to separate means at the 0.05 level in Table 4-7. Within the Huron—Manistee National Forest, increasing gradients in Mg and Mn for both fine and large roots can be seen from the edge of the disease plots, to the outside region where trees appear healthy, and finally to the healthy check plots nearby. There is also corresponding decreasing gradients of metals Al, Fe, and Zn in fine roots at Huron—Manistee. These observations suggest that nutrient imbalances may be occurring within the fine root system. Tukey's HSD procedure does show a significant increasing gradient at the 0.05 level in Mg and Mn concentrations as well as a significant decreasing gradient in Al concentrations between diseased and healthy regions of Huron-Manistee in the fine- root category. The differences between fine-root Fe concentrations at Huron-Manistee, however, were not significant (Table 4—7). In terms of comparing the Huron-Manistee National forest to the other forests included in the study (Osmun Lake natural stands and Roscommon plantations), root samples from Osmun Lake had comparable concentrations of Ca and Mg to Huron-Manistee disease plots. Mn in fine roots was significantly higher at Osmun Lake and Roscommon compared to Huron—Manistee disease edges. Osmun Lake also had 130 significantly lower concentrations of Al and Fe in fine roots compared to both the edge and outside regions of Huron-Manistee disease plots. Roscommon plots, on the other hand, had the highest root concentrations of Al compared to all other forests even though decline is absent there. However, Roscommon also has the highest fine-root Ca and K concentrations, which may be compensating for the high Al. Table 4-8 lists the Ca/Al, Mg/Al, K/Al, and Mn/Fe ratios identified at Huron-Manistee disease plots (edge and outer regions), Huron—Manistee healthy plantations, Osmun Lake, and Roscommon plots. The table shows increasing gradients within the Huron—Manistee forest in the ratios of Ca/Al, Mg/Al, K/Al, and Mn/Fe starting from the edge of disease plots to the healthy check plots nearby. These increasing gradients suggest that conditions are improving for fine-root growth outside the diseased areas of Huron- Manistee. Tukey's HSD procedure reveals that Ca/Al, Mg/Al, K/Al, and Mn/Fe ratios are significantly different at the 0.05 level between Huron-Manistee disease and check plots for fine roots. The increasing gradient is most apparent within the Huron-Manistee National Forest for Mn/Fe ratios, in which disease edges have a significantly lower ratio than the outer regions of disease plots, which were significantly lower than the plantation check plots nearby (HM check). The Mn/Fe gradient was also apparent for the large-diameter root category at Huron—Manistee. This relationship is not 131 Table 4-8. Molar ratios of Ca/Al, Mg/Al, K/Al, and Mn/Fe for disease plots and healthy plantations for fine (5 3 mm in diameter) and large (>3 and 5 5 mm in diameter) red pine roots. Fine Roots g 3 Mblar Ratios mm. Location :1 Ca/Al Mg/Al K/Al Mn/Fe mm Disease 30 1.10 a 0.62 a 0.69 a 0.44 a Edge Disease 42 1.37 a 0.90 a 0.78 a 1.07 b Outer HM Check 26 2.18 b 1.53 b 1.43 b 2.27 c Osmun Lake 16 2.24 b 1.63 b 1.79 b 2.22 cd Roscommon 16 0.87 a 0.51 a 0.56 a 1.00 bd Large Roots 3>§5 Molar Ratios mm Location 11 Ca/Al Mg/Al K/Al Mn/Fe Disease 28 3.26 ab 2.38 ab 3.19 a 1.50 a Edge Disease 35 3.18 ab 2.50 a 2.38 a 2.50 a Outer EM Check 23 3.06 ab 2.39 a 2.31 a 4.15 b Osmun Lake 20 4.10 b 1.95 ab 2.34 a 3.89 b Roscommon 16 2.09 a 1.39 b 1.72 a 2.89 ab *Means within the same column and size category followed by the same letter did not differ at the 0.05 significance level using Tukey's honestly significant difference procedure. 132 apparent for Ca/Al, Mg/Al, or K/Al ratios in the large- diameter root category. In comparing Huron—Manistee to Osmun Lake and Roscommon, it appears that the Osmun Lake natural stands are in fairly good condition in terms of fine-root ratios. Roscommon, on the other hand, has the lowest Ca/Al, Mg/Al, and K/Al ratios even though the red pines appear healthy. Roscommon stands are significantly lower at the 0.05 level than Huron-Manistee check plots for all four fine—root ratios. Fine root Mn/Fe ratios, however, are significantly lower at disease edges than at Roscommon. Other than location or region from which root samples were taken, significant differences were also found between plots. Table 4-9 shows differences in element concentra— tions and ratios of fine roots (5 3 mm in diameter) from both disease (ML1, ML2, ML3, TG4, and T65) and non-disease plots (MLC and TGC) within the Huron-Manistee National Forest. Significant differences were found between fine root element concentrations and ratios by plot. These differences occurred from within disease plots and between disease plots and healthy plots. However, there were no significant differences between the two healthy plots (MLC and TGC) which tended to have lower concentrations of A1 and higher Mn and elemental ratios than disease plots. Not all disease plots differed significantly from check plots. The disease plot averages were calculated from root samples collected at both the edge and outer region of the 133 .mMSUooond mocmwowmao quOHMHcon maummcoc m.momsa mean: Ho>ma mocmoflmacmflm mo.o map um womwflo Doc pap uwuuoa oEmm ecu >9 ooonHom Cezaoo mamm may cflnuflz mcmwzw cmo.m oom.H “gm.fi noom.H gem scam omwm owmfim noses momwm gasses ma own 884 8am; 834 coed 8&4 33 8N? 33:: 88m: imam £0: 2 on: oa©.H omoa.o nmws.o mmo.H ammq seam “seam amnmmfi amamm mommm hxmrxxrfl ma non omm.o “66mm.o naflo.o who.H «mm nmfioe 060mm onmfimfi mama mflmemn onmmmsfi ma qua onws.o unfim.o oasm.o nnmfis.a mam Assam ommm amommfi naefimfi mamam ammmsa ma mg: gamm.o mnamfi.fi “snmfl.fl DQHH.N amom ammm «mm onfimafl amsfloi mmmom “aammrn mH «a: mom.o mom.o «64.0 mmm.o acme mmmm naaofi mmefia gamma mmosm momqm ma Ham on}: is 33: 228 an on a: u or no 2 a ”.on *.ummuom Hmcoflumz omumacmzucowom map aacufiz loos .quv muoac mammmaaueoc new Amos .eoe .mgz .mgz .quv mmmmmwn wow smumamfln CH 8: m v muoou mo moflumu umMoE cam Amx\oEv mcoflumuucoocoo ucmaoao ommwo>< .miv magma 134 plots. If these regions are calculated separately as in Tables 4—7 and 4-8, the nutritional problems associated with disease plot edges are apparent. The results of analysis of variance using plots as the levels of factor (considered a random effect) allows an inference to be made that element concentrations and ratios will vary between disease plots (i.e., other mortality pockets) within the Huron-Manistee National Forest, with the exception of fine root Ca concen— trations which did not vary at the 0.05 level (Table 4-9). To simplify understanding of differences between disease and non-disease plots, Table 4-10 lists the F-ratios and probabilities of comparisons between the mean element concentrations and ratios of fine roots. The F ratios listed in Table 4-10 test the average of fine—root samples from disease plots (ML1, ML2, ML3, TG4 and TG5) and non- disease plots (MLC and TGC) listed in Table 4-9. Table 4—10 clearly shows that, on average, fine-root samples from non- disease plots have significantly higher concentrations of nutrients K, Mg, and Mn and significantly lower concentrations of metals Al, Fe, and Zn. Fine root Ca concentrations were the only non-significant variable identified. Ca/Al ratios, however, along with Mg/Al, K/Al, and Mn/Fe ratios were significantly higher in non—disease plots than in disease plots. 135 Table 4—10. Analysis of variance1<3f element concentra- tions and ratios in fine roots samples (3 3 mm in diameter) from Huron-Manistee disease plots (n=72) and Huron-Manistee non-disease (check) plots (n=26). Significance Non>Dis., Element E-ratio Probability Non Mg 21.56** < 0.001 > Mn 28.99** < 0.001 > 2n 11.07** 0.001 < .= Significance Non>Dis., Ratio F—ratio Probability Mon Mg/Al 37.12** < 0.001 > x/zu 27.51** < 0.001 > Mn/Ee 29.34** < 0.001 > lSignificance:*=significant at 0.05;**=significant at 0.01; ns=not significant. 136 DISCUSSION Ca/Al Ratios and Risk of Forest Stress Cronan and Grigal (1995) defined two soil-related measurement end points to aid in identification of forests at risk for Al stress. Disease plots at Huron-Manistee were consistent with these two risk rating criteria: i.e., base saturation percentages below 15%, and soil molar Ca/Al ratios below 1.0 (Tables 4-1 and 4-2). However, soils from healthy plantations at Huron-Manistee and Roscommon also fell within these threshold values. Soils at Osmun Lake natural stands, for the most part, were above 15% base saturation and were above the Ca/Al molar ratio of 1.0 in the B horizon only (Tables 4-1 and 4-2). These observations suggest, with the possible exception of Osmun Lake stands, that all forests studied, whether disease or non-disease, are at a 50% risk for A1 stress. These measurement endpoints were defined, however, by reviewing the literature on several different tree species. There are no known studies that identify soil solution Ca/Al threshold values specifically for red pines. Pines, in general, are considered to be Al-tolerant (Schaedle et al., 1989; Ulrich, 1985). Cronan, et a1. (1989) reported a threshold Ca/Al soil solution ratio as low as 0.5 for Loblolly pine. Conversely, McCormick and Steiner, (1978) reported a threshold Ca/Al molar ratio of 1.35 for both Scotch pine and Virginia pine. Cronan and Grigal (1995) estimate that the overall uncertainty of the Ca/Al ratio 137 with a given probability estimate is :50% depending on the A1 sensitivity of the plant species. Thus further studies are needed in order to better characterize the threshold values specific to red pine plantations. Given this uncertainly, there is no strong indication that soil Al or Ca/Al ratios are a significant factor in RPPM. The needle samples analyzed from the Huron-Manistee National forest included both healthy (symptomatic for RPPM) and unhealthy (asymptomatic for RPPM) red pines. Whether classified as healthy or unhealthy, the molar Ca/Al of needle tissues all averaged below Cronan and Grigal's threshold value of 12.5 (Table 4-5). Again this threshold value is based on studies which include Al-sensitive and insensitive tree species. Cronan et al. (1989) identified a threshold value of 2.3 for a foliar Ca/Al molar ratio in Loblolly pine. A more specific foliar threshold value is needed to identify Al tolerances within red pine. In fine root tissues, Cronan and Grigal identified a Ca/Al molar ratio of 0.2 as the measurement end point beyond which the risk of forest stress increases. Fine roots are the primary organs involved in nutrient absorption, and are critical indicators of nutrient stress. The Ca/Al molar ratios of the "fine" roots included in this study were above the critical value of 0.2 (Table 4-8). However a larger diameter size class was used (between 1 mm and 3 mm in diameter) rather than the less than 1 mm diameter category recommended by Cronan and Grigal. Therefore, adequate 138 comparisons could not be made. Overall it appears that the Ca/Al ratios examined in this study are not reliable indicators of forest stress and are not a likely cause of RPPM. Results of Soil, Needle, and Root Analyses in Red Pine Mortality Pockets Soil analyses in Huron—Manistee forests as well as the Roscommon and Osmun Lake forests revealed that red pines were growing on acidic, nutrient-poor soils. For the most part, these soils were similar to each other in terms of pH, and element concentrations and ratios. However, significant differences were found between disease plots and healthy plots in soil K and Mg concentrations (Table 4-1). There were also increasing gradients in Mn and Mn/Fe from pocket edges to the Huron-Manistee check plots, but the differences were not significant. Metals, A1 and Fe were highest at the healthy plantations at Roscommon and their role in the decline may not be as critical as first hypothesized. Low Mg was the nutrient factor which seemed to have the strongest association with disease-plot soils. The needle nutrient status was assessed between healthy and unhealthy trees at each respective site, Minnie Lake and Thin—A-Gain within the Huron—Manistee National Forest. The goal was to determine differences between nutrient contents of symptomatic and asymptomatic red pine by controlling for all factors as much as possible and thereby isolating the variation attributed to the health factor. Some significant 139 differences were found between healthy and unhealthy red pines in needle Al, Mn, Zn, Fe, K/Al, Mg/Al and Mn/Fe. However, with the exception of Mn, these factors were not significantly different in all needle age-categories (1- and 2-year-old) and at both sites. When needle concentrations of metals Al, Fe, and Zn were significantly different, they tended to be higher in healthy versus unhealthy red pines. It was expected that metals, especially A1, would be higher in unhealthy trees. However, the higher concentrations of Al and other metals in the red pines with healthy-looking crowns may merely be an indication of more efficient nutrient assimilation capabilities compared to the unhealthy red pine with their thinning crowns and browning foliage. In addition, mature pines do have the capacity to detoxify and immobilize Al within foliar cells enabling them to tolerate relatively high concentrations (McQuattie and Schier, 1992). Fine root mortality was much higher in disease plots compared to healthy plantations (Table 2-2). Uptake of A13+ ions and other elements to the needles may have been prevented by the death of fine roots in the unhealthy red pines studied. Manganese was the only nutrient which was significantly lower in the needles of red pines symptomatic for RPPM in both needle age categories (1— and 2—year) and at both sites (Minnie Lake and Thin—A—Gain) (Tables 4—4 and 4-6). The browning and stunted growth in needles observed in the unhealthy red pines is consistent with the Mn deficiency 140 symptoms reported in South African Radiata pine plantations (Grey and de Ronde, 1988; de Ronde, et al. 1988; Lange, 1969). However, the critical foliar level for P. radiata Mn deficiency was reported to be below 10 mg/kg, (Grey, 1988; Lange, 1969) much lower than the needle Mn levels observed in the symptomatic red pines at Huron—Manistee which averaged 232 mg/kg in l—year-old needles (n=26) (Table 4-4). Stone (1968) reports an intermediate range in foliar Mn concentrations for red pines to be between 200 and 900 mg/kg. Most of our unhealthy 1-year-old needle samples are at the lower end of this range but still within the normal limits, although eleven (42%) of the cases are below 200 mg/kg (range 116 to 418 mg/kg, n=26). However the red pines sampled in Stone's report were sampled in the fall (October— November) rather than summer (late July) as in our study. Red pine foliar Mn concentrations are reported to fluctuate widely throughout the year (Stone, 1968). It is also not clear whether Stone was referring to young red pines or to more mature red pines, similar to the ones examined in our study. The specific Mn requirements for mature red pines growing on spodzolic soils should be determined for a more specific comparison. Fine roots are the main absorption and exchange sites for soil nutrients and are expected to reflect soil chemistry very distinctly (Murach and Ulrich, 1988). Soil nutrient factors whether related to Al or Mn stress are likely to be indicated more effectively by root elemental 141 concentrations rather than by needle concentrations. Fine roots did reflect increasing gradients in Mg, Mn, Ca/Al, Mg/Al, K/Al, and Mn/Fe within the Huron-Manistee forest from the edge of the disease plots to the healthy plantation check plots which differed significantly at the 0.05 level (Tables 4-7 and 4—8). There was also a corresponding decreasing gradient in Al concentrations (Table 4—7). The Osmun Lake natural stands fell in line with improving fine root ratios of Ca/Al, Mg/Al, K/Al and Mn/Fe. The healthy plantations at Roscommon, however, had the lowest fine-root ratios of all areas studied (Table 4-8). The amount of root mortality was also relatively high in Roscommon plots as it was in disease plots (Table 2—2). However, among the dying roots at Roscommon, unlike the Huron-Manistee disease plots, there was also abundant new root growth. It is suggested that fine root turnover rates are much higher at Roscommon compared to Huron-Manistee disease plots. Soil conditions may be causing root mortality within Roscommon, but these roots are also being replaced more efficiently. Roscommon plots also had the highest fine root Ca and K concentrations which may be compensating for any Al stress present there (Table 4—7). Mn concentrations were also significantly higher in Roscommon fine roots compared to disease edges and outer regions (Table 4-7). Focusing primarily on the Huron—Manistee National forest, disease plots had significantly higher fine-root 142 concentrations of metals Al, Fe, and Zn and significantly lower concentrations of nutrients Mg, K, and Mn than non- disease plots (Table 4—10). Fine—root Ca/Al, Mg/Al, K/Al, and Mn/Fe ratios were also significantly lower in disease plots compared to Huron—Manistee check plots. Fine-root Ca concentrations were not significantly different between disease plots and check plots, so it is possible that the higher A1 concentrations within disease plots are a limiting factor of Ca absorption as well as Mg and K absorption. Manganese may also be inhibited by Fe concentrations, but overall it appears that elemental Mn concentrations were the most indicative factors of diseased areas. Manganese was significantly lower in roots (both fine and large), and needles of red pine in RPPM plots and had the highest F values associated with it (Tables 4—6 and 4—10). Although Mn-deficiency is usually a problem associated with high pH soils (Kielbaso and Ottman, 1976), Ulrich (1989) reports that after a prolonged period of soil acidification, Mn will also be leached from soils. The may explain why RPPM does not appear until trees have matured. The stands were densely planted on highly disturbed, marginal farmland sites. Over time, the concentrated accumulation of pine litter in the even-aged stands may have been a major factor in the soil acidification process. Magnesium concentrations also appear to be strongly associated with the decline. Increasing gradients in soil and root Mg concentrations were apparent from disease plots 143 to the healthy plots at Huron-Manistee (Tables 4-1 and 4-7). Lange (1969) found that Mn-deficiency symptoms in P. radiata could be corrected on a long-term basis with both foliar spray, and soil applications of MnSO4. The treated trees also gave a favorable growth response. In addition, symptomatic P. radiata treated with soil applications of MgSCu responded in a similar manner to the MnSCh-treated trees for about three years, after which growth rates fell and foliar symptoms returned. Lange postulated that the role of Mn in the metabolism of the trees was replaced by Mg, but only temporarily. In the case of RPPM, soil deficiencies in Mn and/or Mg in addition to higher concentrations of metals, especially Al and Fe, in acid soils may tip the scale in favor of disease. The trees respond with decreased diameter and height growth rates, thinning crowns, and browning and stunted growth in the needles. The trees are then predisposed to attack by insect and disease pests, which may explain the often circular-pattern to the disease. The trees eventually die, but spread of the disease is slow (Chapter 3) and probably limited by adequate soil nutrient conditions in other locations. There is some indication that in pocket centers, soil conditions are improving with reduced acid loading by pine litter, allowing the regeneration of young red pines. 144 BIBLIOGRAPHY Arp, P.A. and R. Ouimet. 1986. Uptake of A1, Ca, P in black spruce seedlings, effect of organic vs. inorganic Al. Water, Air, and Soil Pollution 31: pp. 367-375. Cronan, C.S., R. April, R.J. Bartlett, P.R. Bloom, C.T. Driscoll, S.A. Gherini, G.S. Henderson, J.D. Joslin, J.M. Kelly, R.M. Newton, R.A. Parnell, H.H. Patterson, D.J. Raynal, M. Schaedle, C.L. Schofield, E.I. Sucoff, H.B. Tepper, F.C. Thornton. 1989..Aluminum Toxicity in Forests Exposed to Acidic Deposition: The ALBIOS Results. Water, Air, and Soil Pollution 48: 181-192. Cronan, C.S. and D.F. Grigal. 1995. Use of calcium/ aluminum ratios as indicators of stress in Forest Ecosystems. J. Environ. Qual. 24, pp. 209-226. de Ronde, C., D.B. James, N.T. Baylis, P.W. Lange. 1988. The response of Pinus radiata to manganese applications at the Ruitersbos State Forest. South African Forestry Journal, No. 146, September 1988, pp. 26—33. Foy, C.D. 1974. Effects of aluminum on plant growth. In: The Plant Root and its Environment, B.W. Carson, ed., Univ. Va. Press, Charlottesville. pp. 602-642. Grey, D.C. 1988. A review of the role of manganese in pine plantations. South African Forestry JOurnal, No. 145, June 1988, pp. 42-46. Grey, D.C. and C. de Ronde. 1988. History and treatment of manganese deficient Pinus radiata. South African Forestry Journal, No. 146, September 1988, pp. 67-72. Hutchinson T.C., L. Bozic and G. Munoz-Vega. 1986. Responses of five species of conifer seedlings to aluminum stress. Water, Air and Soil Pollution 31, pp. 283-294. Jokela, E.J., W.W. McFee and E.L. Stone, 1991. Micro- nutrient deficiency in slash pine: response and persistence of added manganese. Soil. Sci. Soc. Am. J. 55: 492-496. 145 Kielbaso, J.J. and K. Ottman. 1976. Manganese deficiency-— contributory to maple decline? JOurnal of Arboriculture, Feb. 1976, pp. 27—32. Klepzig, K.D. and J.C. Carlson. 1988. How to identify red pinegpocket decline and mortality. USDA For. Serv. NA— GR—l . Knezek, B.D. and H. Greinert. 1971. Influence of soil Fe and MnEDTA interactions upon the Fe and Mn nutrition of bean plants. Agronomy Journal, Vol 63, July—August 1971, pp. 617-619. Lange, P. 1969. A manganese deficiency in Pinus radiata at Klein Gouna, Knysna. Forestry in South Africa, No. 10, Oct. 1969, pp. 47-59. McCormick, L.H., and K.C. Steiner. 1978. Variation in aluminum tolerance among six genera of trees. For. Sci. 24:565—568. McQuattie, C.J. and G.A. Schier. 1992. Effect of ozone and aluminum on pitch pine (Pinus rigida) seedlings: anatom 2f mycorrhizae. Can. J. for. Res. 22, pp. 1901-1 1 . Murach, D. and B. Ulrich. 1988. Destabilization of forest Eggsystems by acid deposition. GeoJournal 172: 253— Schaedle, M., F.C. Thornton, D.J. Raynal and H.B. Tepper. 1989. Response of tree seedlings to aluminum. Tree Physiology 5, pp. 337—356. Shortle, W.C. and K.T. Smith. 1988. Aluminum—induced calcium deficiency s ndrome in declining red spruce. Science: 240:1017-10 8. Stone, E.L. 1968. Forest fertilization: theory and practice, Symposium on Forest Fertilization, April 1967, Tennessee Valley Authority, National Fertilizer Development Center, Muscle Shoals, Alabama. Ulrich, B. 1985. Interaction of indirect and direct effects of air pollutants in forests. In: Air Pollution and Plants, Proc. of the 2nd Euro. Conf. on Chemistr and the Environment, Clement Troyanowsky, Ed., pp. 1 9-180. Ulrich, B. 1986. Natural and anthropogenic components of ggél7ggidification. Z. Pflanzenernaehr. Bodenk 149, Ulrich, B. 1989. Effects of acidic precipitation on forest ecosystems in Europe. In: Acid Precipitation, Volume 2, Biological and Ecological Effects, D.C. Adriano and Aégé Johnson, Eds., Springer-Verlag, New York, pp. 189- CHAPTER V Relationships Between Drought Indices, Disease Intensity, Nutrient Studies, and Fine Root Mortality in Red Pine (Pinus resinosa) Mortality Pockets in Michigan ABSTRACT To determine the factors associated with a decline of red pines know as Red Pine Pocket Mortality (RPPM) correlations were performed between 1) disease intensity and the Palmer Drought Severity Index (PDSI) at disease plots; 2) soil pH, soil/root molar nutrient concentrations/ratios and volume percentages of live, dead, and symptomatic roots recovered from disease and non-disease locations; and 3) soil pH, soil, and root molar nutrient concentrations between disease plots and non—disease plots. Volumes of live and dead roots correlated well with root manganese (Mn), soil Mn, and root magnesium (Mg) concentrations, suggesting that these nutrients play a significant role in the decline. These nutrients were negatively correlated with dead root volumes which were highest in disease areas, and positively correlated with live root volumes which were highest in non-disease locations. Correlations of nutrient contents by plot suggest that soil and root concentrations of Mn and Mg may be the 146 147 variables to investigate when identifying red pine planta- tions susceptible to decline. For example, soil Mg, was lower in disease plots compared to non-disease plots and was positively correlated with root Mg and root Mn. Low soil and root potassium (K), ratios K/aluminum (A1), and Mn/iron (Fe), and higher concentrations of metals Al, Fe, and Zinc (Zn) were also implicated as being associated with disease plots. Soil pH's which tended to be lower in disease plots at Huron-Manistee were negatively correlated with root Mn, root Mg, root Mn/Fe and soil K/Al. In addition, root concentrations of Al and Fe were positively correlated and were higher in disease plots versus non-disease plots. Linear regressions of these correlations were significant at p g 0.05. These observations were consistent with our prior studies using analysis of variance which indicated that disease plots had significantly lower fine—root (roots 5 3mm in diameter) concentrations of Mn, Mg, and K and signifi— cantly higher concentrations of metals Al, Fe, and Zn compared to non-disease plots (Chapter 4). However, the variable which has been the most consistently associated with disease plots and fine root mortality is Mn concentration. Multiple linear regression was used to develop a prediction equation for root Mn concentrations from soil pH and Al content that may facilitate identification of sites where fine root mortality and symptoms of RPPM are likely to occur. 148 INTRODUCTION Abiotic factors associated with a disease of mature red pine (Pinus resinosa) stands in the Lake States known as Red Pine Pocket Mortality (RPPM) were studied. The disease is characterized by declining trees with thinning crowns, browning foliage, stunted growth in needles, and reduced diameter and height growth; surrounding a large, often circular area of dead trees (Klepzig and Carlson, 1988; Raffa and Hall, 1988). The red pine mortality pockets studied in Michigan were all growing on sandy, low-pH soils. It was proposed that nutrient imbalances associated with acidic soils may be the initial stress-inducing factors leading to decline. Of particular interest was the effect that increased mobility and levels of exchangeable species of aluminum (primarily Al”) in acidic soils (below a pH of 5.2) may have on interfering with the uptake of other essential nutrients especially Ca“, Mg“, and K7 ions (Arp, et al., 1989; Balsberg-Pahlsson, 1990; Dewald et al., 1990; Foy, 1974). Therefore, the variables of primary interest in this study were ratios of extractableAl3+ to Ca“, Mg”’and.K3 in soil, needles, and roots of declining and healthy red pines. In particular, aluminum-induced calcium deficiencies and ratios of Ca/Al, have been studied globally in relation to soil acidification and forest declines resulting from acid rain and/or other soil acidifying processes (Johnson et a1, 1991; Ulrich, 1989; Murach and Ulrich, 1988; Shortle and 149 Smith, 1988; Ulrich, 1985). In the case of red pine pocket mortality, pine litter accumulation through time in densely planted, even-aged plantations and the subsequent release of humic acids through decomposition is proposed as the cause of soil acidification. In addition to studying macronutrients and the possibility of competition with A13+ ions, levels of micro- elements such as Mn, Fe, and Zn were determined in our study for identifying other possible nutrient deficiencies/ toxicities. Manganese/iron ratios are also included because of studies that have shown that high concentrations of Fe may cause Mn—deficiency (Grey, 1988; Knezek and Greinert, 1971). There are few studies of micronutrient contents in pine. Manganese deficiency, however, has been associated with stands of Pinus radiata in South Africa (Grey, 1988; de Ronde, et al., 1988; Lange, 1969). The Mn—deficiency symptoms on South African pines included stunted growth, yellow-brown or bronze needle color, and very sparse or short needles (de Ronde, et al., 1988; Lange, 1969). These symptoms are similar to those observed on red pines at the edge of mortality pockets. In addition, the Mn—deficiency observed in South Africa was associated with strongly podzolized soils and good drainage; similar to the Spodosols examined in this study. The circular pattern to many red pine mortality pockets suggests activity by a biotic agent or agents. Prior studies of biotic factors associated with the disease in 150 Wisconsin revealed a complex association of five root- and lower-stem infesting insects vectoring two species of Leptographium fungi (L. terebrantis and L. procerum) in red pine mortality pockets versus healthy red pine stands (Klepzig, Raffa, and Smalley, 1991). In our investigations, red pine roots from mortality pockets in Michigan were cultured on both selective and non— selective media for Leptographium, spp. but no significant root pathogens were identified (Chapter 2). Our investigation into RPPM did not include a formal study of the insects or pathogens involved beyond attempting to identify potential root pathogens. Our casual observations of biotic agents in Michigan pockets, however, have identified many associated insects and fungi but none that have been consistently present in all pockets. For this reason, we consider the various insects and fungi found within pockets to be secondary expressions of disease and not the primary cause of decline. Our investigations of potential abiotic factors involved in the decline, have indicated that drought, in addition to nutrient stress, may be a contributing factor (Chapter 3). Symptoms of RPPM in the crowns of declining red pines were the principal indicators of increases in disease intensity. Disease intensity, as measured from a rating scale, was correlated with the Palmer Drought Severity Index (PDSI) to determine if disease symptoms increase during dry years. 151 Whether the causes are biotic or abiotic in origin, it has been presumed that root mortality leads to the expression of above-ground symptoms of RPPM. The results of our nutrient studies from red pine mortality pockets were compared with healthy stands to determine relationships between element concentrations and ratios in soil and fine- root samples. Soil and root nutrient levels were also compared to proportions of live, dead, and symptomatic roots removed from healthy and diseased stands. It is anticipated that identifying significant nutrient relationships at these sites will lend evidence to support the probable cause or causes of this disease. Prediction equations can be determined from significant relationships which will assist forest managers in identifying sites that may be subject to the disease. METHODS The Palmer Drought Severity Index (PDSI) was correlated with disease intensity (D.I.) for three disease plots (ML1, ML2, ML3) at the Minnie Lake site in the Huron-Manistee National Forest, Newago County, Michigan. Two disease plots (TG4 and TG5), in an area known as Thin-A—Gain, are not included because over 50% of the trees were removed in a thinning operation in 1993 making it difficult to continue tracking disease progress. The value for the PDSI was determined by averaging the monthly values from January to June for each year of the study (1991 through 1995). The 152 correlations were between PDSI and D.I. on a year-to-year basis for a total of 5 observations. The PDSI values for the Newago County region of Michigan (Zone 5) were obtained from the National Climatic Data Service through the Internet (http://www.ncd.noaa.gov/onlineprod/drought/xmgrg2.html). Disease intensity was determined by averaging the disease rating for each tree within the plot. The trees were evaluated according to a rating scale in mid to late June of each year (1991 through 1995). The rating scale evaluated trees based on their diameter and percent of the crown showing symptoms of RPPM and was based on a 0 to 4 scale of decline with rating 4 being a dead standing tree (Table 3-1). Crown symptoms included browning or stunted growth in needles, thinning crowns, and/or poor terminal growth. A correlation matrix has been used to examine relationships between root and soil element concentrations and ratios and volume percentages of live, dead, and symptomatic roots by location within disease and non-disease plots. The locations within disease plots (ML1, ML2, ML3, TG4 and TG5) at the Huron—Manistee National Forest, include the edge of pockets, consisting of declining trees and the outer "healthy-looking" regions of disease plots (Figure 2- 2). Check plots of healthy red pine stands included in this comparison are plantation stands at the Huron—Manistee National Forest (TGC and MLC) and at the Houghton Lake State Forest in Roscommon County, Michigan (R1 and R2). Thus 153 correlations with live, dead, and symptomatic root volume percentages and element concentrations were derived from the average of all plots at four locations: Huron—Manistee disease edges; Huron—Manistee disease outer regions, Huron— Manistee check plots, and Roscommon County check plots. Roots were removed from each of the locations, and recorded as live, dead, or symptomatic (Chapter 2). Roots were called symptomatic if they were partially dead or dying and/or if there was any staining below the epidermis. Soil and fine-root nutrient levels have been correlated with proportions of dead, live, and symptomatic roots to determine which nutrient factors may be associated with them. Relationships between root and soil samples were examined by correlating molar ratios of Ca/Al, Mg/Al, K/Al, and Mn/Fe and molar element concentrations of Al, Ca, Mg, Mn, K, Fe, and Zn by plots within the Huron-Manistee National Forest. This includes both the disease plots (ML1, ML2, ML3, TG4, and TG5) and the healthy check plots nearby (MLC and TGC). Red pine stands within the Huron-Manistee National Forest were the most similar to one another in terms of site factors (climate, topography, soil type, etc). Thus, these stands are likely to be the best comparison group to one another rather than stands adapted to growing in a different forest (for example, Roscommon County plantations). Correlations between plots located within the Huron—Manistee forest may indicate which nutrient factors 154 are associated with disease versus non-disease stands. Correlations with significant linear regressions were also determined so that prediction equations could be developed. Stepwise multiple linear regression analyses were used to determine the best—fit predictive models for the nutrient factors of interest. The model used assumes the effects of the independent variables are additive (i.e., no interaction effects) and follows the form: y = 00+ [31x1 + 52x2 + ...+ kak+€ All computations were done with Systatmh RESULTS Disease Intensity and the Palmer Drought Severity Index Table 5—1 illustrates the relationship between the average Palmer Drought Severity Index (PDSI) and disease intensity (D.I.), per plot and as an average, measured on an annual basis (1991—1995). Average disease intensity is the mean D.I. value for the three Minnie Lake plots. The 6— month-averaged PDSI values are negatively correlated with disease intensity at plots ML1, ML2, and ML3 and with the average disease intensity (R=-0.738, R=-0.776, R=-O.764, and R=-0.765, respectively). (A negative correlation is expected since the PDSI values become more negative with increasing drought severity and disease intensity (i.e., greater symptoms of RPPM in tree crowns). With only five observations, one for each year of the study, the linear regressions for the correlations 155 between PDSI and D.I. are not significant. Further data collection and observations of disease intensity patterns in | ' .- .5L’ 5"“ 0.“ I .0 mortality pockets over time are needed to investigate the relationship between drought and disease intensity. The . -.--'.:'..I u ar'l' "'I"||l correlations do suggest, however, that symptoms of RPPM increase in tree crowns during drier years. In addition other factors such as pest problems and/or soil nutrients may contribute to changes in disease intensity. Thus, drought alone can not account for the entire disease intensity rating. Table 5-1. Correlations between Average Palmer Drought Severity Index (PDSI), average disease intensity (D.I.) and disease intensity by plot (ML1, ML2, and ML3) (observations=5). Avg. Avg. D.I. D.I. PDSI D.I. ML1 ML2 Avg. D.I. —0.765 D.I. ML1 —0.738 *0.995 D.I. ML2 -0.776 *0.993 *O.976 D.I. ML3 -0.764 *0.999 *0.998 *0.988 *Correlation with significant linear regression (p50.05). Strong positive correlations with significant linear regressions (p50.05) can be found with disease intensity between plots and also with the average disease intensity for all three plots. This indicates similar patterns of 156 disease progress were occurring in all Minnie Lake plots during each year. The plots consisted of mature, even-aged stands of red pine within the same region of the Huron— Manistee National Forest, and were likely being affected in similar ways by environmental factors such as drought, and biotic factors such as the population of insect pests present in the area. Soil and Root Nutrient Contents and Root Vitality Table 5—2 presents the correlations between soil pH and root and soil element concentrations and ratios in terms of proportions of live, dead, and symptomatic roots. The correlations are between the averages for each variable by location: the edge regions of disease plots (DE); the outer healthy-looking regions of disease plots (DO), the healthy plantation check plots at Huron-Manistee (HM), and the healthy plantations at Roscommon (ROS). Thus only four data points are available for linear regression. Strong positive correlations were found between root Mg, root Mn, and soil Mn concentrations and the proportions of live roots (R=0.951, R=O.979, and R=O.988). These correlations had significant linear regressions as illustrated in Figure 5—1. Conversely, strong negative correlations can be found between root Mg, root Mn, and soil Mn concentrations and the proportion of dead roots (R=- 0.916, R=—0.868, R=—0.862, respectively); although, with only four data points it was not possible to reject the null Table 5-2. 157 Correlations between soil pH, root and soil element concentrations/ratios and proportions of live, gead, and symptomatic roots by location (observations = DEAD LIVE SYMP LIVE -0.926 SYMP 0.417 —0.730 pH -0.487 0.415 -0.118 Root A1 0.461 -0.292 -0.132 Root Ca -O.355 0.656 -0.935 Root Ca/Al —O.624 0.526 -0.137 Root Fe 0.818 -0.682 0.160 Root K -0.l43 0.499 -0.942 Root K/Al -0.591 0.544 -0.240 Root Mg -0.916 *0.951 -0.628 Root Mg/Al -0.678 0.582 -0.172 Root Mn -0.868 *0.979 -0.783 Root Mn/Fe -0.837 0.850 -0.528 Soil Al -0.350 0.623 -0.866 Soil Ca -0.l98 0.454 -0.733 Soil Ca/Al 0.635 -0.845 0.883 Soil Fe -0.039 0.252 -0.535 Soil K -0.617 0.854 -0.938 Soil K/AL -0.222 0.217 -0.119 Soil Mg -0.181 0.533 *- .954 Soil Mg/Al 0.148 0.126 -O.572 Soil Mn —0.862 *0.988 -0.817 Soil Mn/Fe —0.821 0.815 —0.476 *Correlation with significant linear regression (p50.05). 158 a hypothesis of Bf=0 at a 0.05 significance level. However, 1% these correlations do suggest that there is a relationship between proportions of dead roots and root Mg and Mn concentrations. Of metals, root Fe was more negatively correlated with the proportion of live roots (R=-0.682) than root Al (R=- «r..w.(tenrmxfifiiMMW-‘i ‘ -; .- 0.292). Root Fe was also more positively correlated with the proportion of dead roots (R=0.818) than root Al (R=0.46l). This suggests that Fe has the stronger associa- tion with root mortality than does Al. Iron has been implicated as an inhibitor of Mn uptake (Grey 1988; Knezek and Greinert, 1971) and may be having a greater effect on Mn nutrition than the proposed effect of Al on Ca uptake and nutrition. However, the linear regressions for the correlations between Fe and Al to live and dead roots were not significant. Of ratios, root and soil Mn/Fe had the strongest correlations with live and dead root volumes. Higher root and soil Mn/Fe ratios were associated with greater proportions of live roots (R=0.850 and R=0.815) and were negatively correlated with the proportion of dead roots (R: -0.837 and R=-O.821). Root ratios of Ca/Al, Mg/Al, and K/Al also correlate positively with live root ratios and negatively with dead root ratios but not as strongly as root Mn/Fe ratios. In addition, soil Ca/Al, Mg/Al, and K/Al, unlike soil Mn/Fe ratios, did not correlate as 159 expected in terms of positive/negative relationships and/or were only weakly correlated with the root proportions. Again, significant linear relationships could not be established for soil and root ratios with only four observations. The only factors which provided correlations with significant linear regressions were root Mg, root Mn, and soil Mn to the proportion of live roots, and soil Mg to the proportion of symptomatic roots (R=—0.956). Therefore, it is suggested, that root and soil element concentrations of Mn and Mg are the most important factors in RPPM compared to other element concentrations or ratios. Figure 5-1 shows the linear regressions for root molar Mg and Mn concentrations, soil molar Mn concentrations, and the percentage of live roots removed from each of the four locations. In each case, the percentage of live root volume increases with increasing Mn or Mg concentrations; and a gradient is apparent starting from the edge of the disease plots where trees are declining (DE), to the healthy-looking outside regions of disease plots (D0), to the Roscommon healthy plantations (ROS), and finally to the Huron—Manistee check plots (HM). Huron-Manistee check plots were the healthiest in terms of percentages of live and dead fine root volumes (Table 2—2) and also had the highest soil Mn, root Mn, and root Mg concentrations (Tables 4-1 and 4-7). 160 a" HM H9} 70 L - 70 80 ” ‘ 50 - “lweso' ‘ xUMeso j an - q ‘0 j 3” " _, u . " 3.05 20 l l 0.02 0.04 0.05 0.03 20 1 . . 4 I l mun 0.04m mun mun RootMn Figure 1a. 18:0 904 Figure lb. F=18.756 Itf0.958 :0 049 F—54.768 p " p=0.021 -_—_ 7 %LIVE 84-6+3~11RMG %LIVE=15.6+7631RMN Figure 1c. IV=0.977 F=83.823 p=0.012 %LIVE=3.431+1406958MN Figure 5—1. Linear regression for percentage of live root volume (LIVE) and concentrations of a. root molar Mg (RMG), b. root molar Mn (RMN) and c. soil molar Mn (SMN) by location: disease edge (DE), disease outer (DO), Roscommon non—disease plantations (ROS) and Huron-Manistee non-disease plantations (HM). 161 Soil and Root Nutrient Factors By Plot Correlations between soil pH, and molar root and soil concentrations and ratios are presented in Table 5-3. Analysis of variance had indicated that significant differences in element concentrations and ratios could be identified according to plot (Table 4—9). Therefore, the correlations presented are between plots at the Huron- Manistee National forest. This includes both disease plots (ML1, ML2, ML3, TG4, and TG5) and non-disease plots (MLC and TGC). With seven observations (7 plots), significance of linear regressions at the 0.05 level was easier to demonstrate than with only the four observations available for Table 5—2. As expected, linear regression equations predicting root Ca/Al (R=-0.863), Mg/Al (R=-0.949), and K/Al (R=-0.967) ratios can be derived from root Al concentrations. They could also be derived from their respective root Ca (R=0.788), root Mg (R=0.797), and root K (R=0.799) concentrations. Soil and root Mn/Fe ratios could also be predicted from their respective Mn concentrations (R=0.982 and R=O.975). Some surprising results are that root Fe concentrations can also be used to predict root Mg/Al (R=—0.831) and K/Al (R=-0.802) ratios (but not root Ca/Al ratios, R=—0.665). This is explained by the fact that root Fe concentrations increase along with root Al concentrations (R=0.893). This relationship is illustrated in Figure 5-2. Root Al and root nu" . 71'1°P-mf+§-i*lili‘m'¢WWW . g I.l ['1‘ \al 162 Table 5-3. Correlations between soil pH, and root (R) and soil (8) element concentrations and ratios by Huron— Manistee disease plots and check plots (observations =7). pH RAL RCA RCAAL RFE RK RKAL RAL -0 358 RCA. -0.123 -0.483 RCAAL 0.257 *-3863 ‘*0.788 RFE -0.585 *0.893 -0.289 -0.665 RK 0.463 -0.723 0.22 0.429 --0.631 RKAL 0.520 *-.967 0.549 *0.820 *-.831 *0.799 RMG *0.791 -0.666 0.414 0.655 -0.570 0.391 0.613 RMGAL 0.608 *-.949 0.614 *0.899 *-.802 0.601 *0.939 RMN *0.966 -0.656 -0.087 0.355 -0.589 0.558 0.624 RMNFE *0.968 ‘-0.646 '-0.117 0.282 -0.650 0.614 0.640 RZN —0.726 0.385 -0.005 —0.147 0.321 -0.293 -0.478 SAL 0.024 -0.580 0.465 0.729 -0.228 0.169 0.548 SCA 0.129 0.003 -0.002 0.243 0.085 '-0.425 -0.209 SCAAL 0.21C) 0.390 -0.690 —0.567 '-0.004 -0.449 -0.514 SFE -0.326 0.583 -0.515 -0.406 0.586 —O.744 -{L743 SK 0.474 -0.672 0.329 0.694 -0.362 0.251 0.634 SKAL *0.855 ‘-0.353 '-0.179 0.204 -0. 32 0.009 0.269 SMG 0.690 -0.722 0.486 0.649 -0.490 0.542 *0.794 SMGAL 0.649 CL(HE5 -0.443 ‘-0.394 -0.241 0.077 -0.043 SMN 0.169 ‘-0.421 0.153 0.524 -0.092 -0.007 0.337 SMNFE 0.263 '-0.545 0.238 0.608 ..0.209 0.132 0.490 *Correlation with significant linear regression (p50.05). 163 Table 5—3. (Cont'd) RMG RMGAL RMN RMNFE RZN SAL SCA RMG RMGAL *0.797 RMN *0.766 0.680 RMNFE 0.699 0.652 *0.975 RZN -0.461 -0.519 —0.736 *-.776 SAL 0.257 0.598 0.236 0.101 -0.188 SCA 0.410 0.066 0.121 -0.062 0.286 0.257 SCAAL —0.123 -0.427 0.022 0.091 0.043 -0.672 0.189 SEE -0.336 -0.599 -0.321 -0.454 0.475 -0.004 0.684 SK 0.579 0.751 0.635 0.508 -0.550 *0.877 0.316 SKAL 0.690 0.477 '*0.831 *0.768 '—0.722 0.208 0.410 SMG *0.773 *0.863 *0.764 0.730 *-.789 0.522 -0.052 SMGAL 0.255 -0.015 0.468 0.595 -0.538 -0.665 -0.231 SMN 0.280 0.450 0.363 0.193 -0.243 *0.917 0.475 SMNFE 0.364 0.582 0.460 0.311 -0.372 *0.942 0.359 SCAAL SFE SK SKAL SMG SMGAL SMN SCAAL SFE 0.395 SK -0.469 -0.092 SKAL 0.294 0.056 0.630 SMG -0.485 -0.596 *0.783 0.605 SMGAL 0.689 -0.258 -0.259 0.502 0.130 SMN -0.420 0.285 *0.902 0.452 0.451 -0.511 SMNFE -0.491 0.104 *0.959 0.487 0.594 -0.450 *0.982 *Correlation with significant linear regression (p50.05). 164 Fe concentrations tend to be higher in disease plots (ML1, ML2, ML3, TG4, and TG5) compared to non-disease plots (MLC and TGC). 0.012 . . r 0.010. 0.000- Rm“ R2=0.797 F=l9.590 0.0004 p=0.007 RFE=-001+0.113RAL 0.004; 0.002 . . . 0.02 0.04 0.06 0.00 0110 FloatAl Figure 5—2. Linear regression between root Al (RAL) and root Fe (RFE) molar concentrations. It was expected that the soil pH would correlate well with root Al and soil Al concentrations. However, the strongest correlations were found between soil pH and root Mn concentrations (R=0.966), root Mn/Fe ratios (R=0.968), root Mg concentrations (R=0.791), and soil K/Al ratios (R=0.855). Figure 5-3 presents the relationship of soil pH to these factors. As soil pH increases, molar concentrations of root Mg, Mn, Mn/Fe, and soil K/Al increase. The check plot at Thin—A—Gain had the highest pH and therefore had the highest concentrations of root Mg and Mn. Soil K/Al ratios were only slightly higher at check plots. These linear « ,‘q ‘ . .3. 1““. . 165 0.015 . I I j , ‘ I l I I 1 g rec 1 S 0.010 3 Root Mn Root 5 Man4: '3 0.005 0 . . . r 0.000 . . . 1 . . .5 .7 . . . . 4.5 4.6 4.7 4.0 4.9 5.0 5.1 ‘5 ‘ " ‘3 ‘9 5° 5‘ p" ”H Figure 3a. Figure 3b. i$=0.933 Ir=o.937 F=69.995 F=74.342 p<0.001 p<0.001 RMN=-0,069+0,016pH RMNFE=-21.747+4.878pH 006 0.20 T r I m r ' ' 1 fl fee. 0.15 r- . 0.05 * ' ”L3 i MLCo WW Root Mg (ML2 0.04? .TGS . “"0 .MU 4 'TG‘ TG4 ' . . . . . "'03 .5 41.5 41.? 4:8 4:9 510 5.1 0'05“ ‘4‘ ‘-7 "° *9 5'“ 5" I)" "H Figure 3c. Figure 3d. R2=0.626 R“=O.730 F=8.357 F=13.550 p=0.034 P=0-014 RMG=-O.117+0.034pH SKAL=-0.356+0.096pH Figure 5-3. Linear regression of soil pH to molar concentrations of a) root Mn (RMN), b) root Mn/Fe (RMNFE), c) root Mg (RMG), and d) soil K/Al (SKAL). 166 regressions suggest that higher soil pH will favor higher root Mn, Mg, Mn/Fe, and soil K/Al. One of the purposes of Table 5-3, was to determine if soil concentrations of nutrients would correlate well with root concentrations. This would make it possible to predict root concentrations of certain nutrients from soil analyses. There were few strong correlations between soil and root concentrations except in the case of soil Mg. Soil Mg correlated strongly with root Mg and with root Mg/Al concentrations. Surprisingly soil Mg also correlated well with root Mn, root Zn, and root K/Al. These relationships are presented in Figure 5-4. In all cases, soil Mg was highest in check plots compared to disease plots as were root Mg, Mn, and K/Al ratios. Root Zn was negatively correlated with soil Mg and decreased in check plots compared to disease plots. It should be pointed out, however, that even though there is a significant linear relationship between soil Mg and the nutrient factors in Figure 5-4, the predictive power of soil Mg for these factors is somewhat limited; the coefficient of determination (R2) ranged from only 0.597 to 0.630 (Figure 5—4). A multiple regression analysis would be needed to improve prediction of root nutrient factors from soil factors. There are many strong correlations, in Table 5-3, between Mn, Mg, and K (or K/Al) concentrations. For example, soil K increases with soil Mn and with soil Mg 0.05 Root Mg 0.03 0.0003 l I 0.000! 0.0005 0.0006 800 Mo Figure 4a. R2=0.597 F=7.408 p=0.042 RMG=0.014+74.0418MG 0.0010 1 I 0.0000 Root Zn 0.0000 . 16c 0.000‘ I I 0.0003 0.0004 0.0005 0.0000 Soil Mg Figure 4c. R2=0 . 62 3 F=8.272 p=0.035 RZN=0.001+-1.252SMG Figure 5-4. of soil Mg and a) root root Zn (RZN), and d) root K/Al by plot 167 M9 0.015 0.010 Root Mn Imo . 4 0.0003 0.0004 0.0005 0.0006 Soil Mg Figure 4b. R2=0. 583 F=6.999 p=0.046 =-0.006+27.6985MG 2.0 1.5 no (It 1 .0 0.5 MU i 0.0004 4.0005 smum 0.0M3 0.0000 Figure 4d. 112:0. 630 F=8.515 p=0.033 RKAL=-0.608+3860.205SMG Linear regression between molar concentrations (RMG) b) root Mn (RMN), c) (RKAL). 168 (R=O.902 and R=0.783, respectively), and root Mn also increases with root Mg (R=0.766). This suggests that these three nutrients are interacting in some way and, for the most part, are lower in disease plots compared to check plots. Multiple Regression Analyses Correlations between soil and root nutrient factors and volume percentages of live and dead fine roots, have indicated that low root and soil Mn and low root Mg concen- trations are associated with disease areas and lower live fine—root volumes (Figure 5-1). Stepwise regression analyses were used to determine which soil or root nutrient factors should be included in multiple regression equations for predicting the concentrations of these nutrients. Soil factors, pH and element concentrations, were tested in each prediction model separate from root nutrient factors. No soil or root elemental ratios were included in the model. Stepwise regression analyses eliminated factors which did not contribute to improving the fit of the prediction equation by increasing the R?. Using this method, the best— fit models for predicting concentrations of root Mn, soil Mn and root Mg were determined and are listed in Table 5-4. These equations allow prediction of root Mn and soil Mn concentrations from soil nutrient factors which are easier to determine in field studies than root nutrient factors. In turn, root Mn and soil Mn were strongly correlated with '09 -H‘ r 'u's'n'i‘U 1 . . ’ """qa ."".'1‘I'IH concentrations of live and/or dead fine roots Therefore, 169 (Table 5—2). the equations could be used to identify sites from soil factors where root mortality and disease are likely to occur. Table 5-4. Multiple regression equations for predicting root (R) Mn, soil (S) Mn, and root Mg concentrations. Dep. Ind. Var. Var. Std. Toler- F- (Y) (X) Err. ance Model R2 ratio P RMN Ifli 0.001 0.999 RMN=-0.072+0.0l6pH 0.979 92.04 <.001 +0.608SAL SAL 0.208 0.999 SMN EMU; 0.023 0.224 SMN=—0.001+0.09OSAL 0.992 127.1 <.001 +0.557SFE+0.67ISK SFE 0.086 0.966 SK 0.131 0.223 RMG IKHX 0.297 0.992 RMG=-.013+0.674RCA 0.820 9.09 0.033 +2.135RMN I004 0.563 0.992 was provided by root nutrient factors concentrations, analysis revealed that, used to predict root Mg concentrations. The best prediction model for root Mg concentrations Table 5—4). (root Ca and Mn However stepwise regression of soil factors, pH could also be No other soil factors were accepted for inclusion in the prediction equation for root Mg concentrations. to the stepwise analysis, Therefore, according the simple linear regression model of RMG=-0.117+0.034pH could be used to predict root Mg 170 concentrations but with an R? of 0.626 (p=0.034) the prediction capability of the equation is somewhat limited. Root nutrient factors could also be used to predict root Mn concentrations. According to the stepwise regression analysis the model would be: RMN=0.021-0.087RAL-0.321RCA+0.262RMG with an R? of 0.905 (p=0.048). In this model, root Al increased with decreasing root Mn concentrations (Table 5- 3). However this relationship was not true for the soil Al variables used to predict root Mn and soil Mn concentrations in Table 5-4. In these models soil Al increased with increasing root and soil Mn concentrations (Table 5-3). This indicates that, in soil, there is no apparent competi- tive effect of Al on Mn levels. If fine roots are examined, however, there does appear to be a competitive relationship between the levels of Al and Mn. The stepwise procedure did not accept any of the root nutrient factors for prediction of soil Mn. But a good fit was provided for soil Mn prediction from the soil nutrient factors in Table 5-4 (Rfi=0.992). Figure 5-4 illustrated that soil Mg was a suitable predictor of root Mn and Mg. However, stepwise regression analysis did not select soil Mg as a variable for inclusion in the root Mn and Mg prediction equations in Table 5-4. This indicates problems with multicollinearity. There are too many variables that are strongly intercorrelated to allow for inclusion of all variables in the prediction equations. 171 DISCUSSION Our observations of red pine pocket mortality have indicated that drought is a contributing stressor to the disease. Symptoms of RPPM increased in tree crowns during drought years compared to non—drought years (Chapter 3). The PDSI was averaged for 6-months (January though June) and correlated with the average disease intensity for that year. Attempts were made to correlate disease intensity with the previous year's average PDSI as averaged from May to November. Studies of red pine have shown that bud formation and shoot elongation are controlled more by the previous year's rainfall than the current year's (Clements, 1970, Motley, 1949). However, our measures of disease intensity did not correlate as well with the previous year's rainfall. Other studies of red pine have found that needle elongation is more responsive to the current year's water supply (Garret and Zahner, 1973). One of the primary symptoms of RPPM observed when evaluating disease stands was browning and stunted growth in needles. Thus, our measures of disease intensity correlated better with the current year's water supply as it affects needle growth than with the previous year's effect on bud formation and shoot elongation. Strong correlations were also found by location between root and soil nutrient concentrations with the volume percentages of live, dead, and symptomatic roots. In particular molar root and soil Mn and root Mg concentrations 172 were positively correlated with the proportion of live roots and negatively correlated with the proportion of dead roots. Disease edges had significantly higher volumes of dead fine roots than did Huron—Manistee check plots (Table 2—2). This suggests that root and soil Mn and root Mg concentrations play a role in determining volumes of live and dead fine roots (i.e. roots 33 mm in diameter). Prediction equations could be developed from the Mn and Mg concentrations to estimate percentages of live and dead roots. Ulrich (1989) reports that mobile cations, K and Mn, are leached from dead roots and that the Al content of dead roots in acid soils approaches that of the soil solution. Beyond these differences, contents of other nutrients should be similar between live and dead roots. Only live and symptomatic roots were analyzed for element content in our investigations, so the difference in Mn concentrations between disease and non-disease areas can not be attributed to dead roots. Symptomatic roots, however, were defined as partially dead or dying and they do have a strong negative correlation with root K concentrations but also to soil Mg, soil K, and root Ca concentrations. Red pine at the edge of the disease pockets were in a state of decline; whereas, the red pines examined in all other areas, even in the outer regions of disease plots, appeared healthy. However root mortality was higher in the outer regions of disease plots compared to healthy plantations even though tree crowns appeared healthy. This .1. " III"“ ‘11 H'Uw' “MIMI! .“ "Wm H I.l ‘.l'l‘" -| '- Wit r» u "I 173 suggests that the roots are dying first, leading to expression of symptoms in the crowns, rather than tree crowns declining and causing root mortality. The average molar concentrations of root Mg, root Mn, and soil Mn at the outer regions of disease plots were 4.4 x 10”, 4.8 x 10*, and 3.5 x 10”, respectively. These are suggested as threshold values below which the risk of decline in red pines plantations of low pH (<5.0) increases. Further correlations between soil and root element contents in disease and non-disease plots at Huron-Manistee suggest a link between low pH and low root Mn and Mg concentrations and Mn/Fe ratios. This may be due to competition from increased solution concentrations of metals Al and Fe, which tend to increase in acid soils. Both root Al and Fe concentrations tended to be higher in disease plots compared to check plots (Figure 5—2). Of soil variables, soil molar Mg concentrations correlated well with root concentrations of Mg, Mn, Zn and K/Al. Disease plots at Huron-Manistee had lower soil molar Mg concentrations than check plots which correlated with lower concentrations of root Mg, Mn, and K/Al and higher concentrations of Zn. It appears that a soil molar Mg concentration below 4X10“‘may indicate conditions where red pines begin to decline (Figure 5-4). This is suggested as another threshold value indicating conditions favorable to development of RPPM and would apply only to red pine 174 plantations growing in a similar soil type (sandy Spodosols) of low pH and in a similar climatic zone (Zone 4 or 5). The variables of primary interest when we initiated this study were the ratios of Ca, Mg, and K to Al concentrations. It was expected that low Ca/Al ratios, in particular, would indicate conditions favoring root mortality and the decline of red pines. Ca/Al along with Mg/Al, and K/Al ratios did correlate negatively with dead root volumes and positively to live root volumes. However, root and soil Mn, and root Mg concentrations had stronger correlations with dead and live root volumes which allowed development of linear regression equations for predicting proportions of live roots from root Mn, root Mg, and soil Mn (Figure 5-1). In addition, stepwise multiple regression provided equations for predicting root Mn and Mg from soil variables which, in turn, could be used to identify sites prone to disease. In analysis of variance studies, Mn was the only nutrient found to be significantly lower, not only in fine root tissues, but in large roots and needles from disease plots compared to non-disease plots and also had the largest F—ratios associated with it (Chapter 4). I propose that Mn deficiency is the primary cause of fine root mortality in red pine mortality pockets. Although Mn deficiency in trees is usually associated with high pH soils (Kielbaso and Ottman, 1976), Ulrich (1989) reports that after a prolonged period of soil acidification even Mn may be leached through 175 acid soils. The Mn—deficiency found on P. radiata stands in South Africa was associated with acid, highly-podzolized soils similar to the soils included in our study. The South African pines also exhibited similar symptoms of browning, and stunted growth in needles observed in RPPM-affected red pines (Grey, 1988; de Ronde, et al., 1988; Lange, 1969). In the proposed etiology of RPPM, localized conditions of low soil Mn (along with Mg and perhaps K concentrations) and high concentrations of metals Al, Fe, and Zn may be enough to tip the scale in favor of fine—root death. Fine- root mortality inhibits absorption and translocation of essential nutrients to tree crowns, inhibiting growth. Trees become weakened and predisposed to attacks from insect and disease pests. The role of biotic agents in the disease may explain the often circular pattern of RPPM pockets. In addition, low nutrient and water absorption by fine roots increases the susceptibility of red pines to drought. The combination of these stresses results in the observed symptoms of RPPM, i.e. thinning crowns, browning and stunted growth in needles, and reduced diameter and height growth; and eventually in tree mortality. Conclusions of Red Pine Pocket Mortality Studies The following ten points summarize our observations and conclusions regarding the nature of red pine pocket mortality and about the primary cause of root mortality. 1. The total volume of fine roots (roots 3 3.0 mm in 176 diameter) were lower and the volume percentages of dead fine roots were higher in RPPM-affected areas versus healthy locations (Table 2-2). Within the Huron-Manistee National Forest, a gradient could be seen in the volumes of live fine roots which increased from the center of mortality pockets where there were many tall, dead-standing trees; to the edge of pockets where trees were declining; to the outer, healthy-looking regions of disease plots; and finally to the healthy plantations nearby (Table 2—2). These observations suggest that fine roots are dying first before symptoms of RPPM are seen in the crown. 2. Red Pine Pocket Mortality does not appear to be spreading quickly-—a once—a-year evaluation was frequent enough to observe disease progress. Evaluation of disease progress indicated that the disease may increase, decrease, or remain constant during any given year (Chapter 3). 3. Spatial autocorrelation revealed that the pattern of disease progress is not random. Red pines near diseased trees are more likely to develop symptoms of the disease than those further away, suggesting movement of an infectious agent from one tree to another (Chapter 3). 4. Drought is likely to be a contributing factor to the disease. Increased mortality and further decline in tree crowns was observed in drought years versus non-drought years (Chapter 3). 5. Soil concentrations of Mg and K were significantly lower in disease areas compared to healthy locations and may be 177 associated with the decline (Table 4-1). 6. Of needle nutrient contents and ratios, Mn was the only factor that was consistently significantly lower in both one- and two-year old needles of red pines symptomatic for RPPM compared to asymptomatic red pines within the Huron- Manistee National Forest (Table 4-4). 7. An increasing gradient in fine root Mn and Mg concentrations was found at the edge of disease plots, to the outer healthy—looking region of disease plots, to the plantation check plots nearby. There was also a corresponding decreasing gradient in fine root Al concentrations. Significant differences in root elemental concentrations were found within these gradients at the 0.05 level (Table 4-7). 8. Analysis of variance revealed that nutrients Mn, Mg, and K and ratios Ca/Al, Mg/Al, K/Al, and Mn/Fe were signifi— cantly lower, and metals Al, Fe, and Zn were significantly higher in fine roots from disease plots versus non—disease plots (Table 4-10). 9. Despite significant differences identified for fine root Ca/Al, Mq/Al, and K/Al ratios, the theory of Al toxicity in red pine mortality pockets is not as strongly supported as Mn-deficiency. Manganese was consistently lower in disease plots, not only in fine roots, but in large diameter roots (3-5 mm in diameter), and needles. Manganese also had the highest F—ratios associated with it in statistical tests (Chapter 4). Root and soil Mn concentrations along with nu“: 178 L ( “"1, ' root Mg were highly correlated with proportions of live and dead roots (Table 5-2, Figure 5—1). 10. Suggested threshold values for identifying susceptible— to—decline red pine plantations are a molar root Mg ..'1 ‘i'..'1_.1 -.v'|-(i011:f.v.t¥h1‘1‘ML“ '. :7. concentration of 4.4X10”, a molar root Mn concentration of 4.8X10”, a soil molar Mn concentration of 3.5X10'4 and a soil molar Mg concentration of 4X10'4 . Levels of root and soil Mn and Mg which fall below these values in mature red pine plantations with sandy, low-pH soils are at risk for decline. Implications for Further Study Further investigations of red pine pocket mortality should focus on establishing cause and effect. Our studies have indicated low soil and fine-root Mn and Mg concentra- tions in acid soils may be the cause of fine-root mortality. It order to establish cause, it would be necessary to g fulfill experimental requirements as outlined by Koch's postulates. One strategy would be to apply lime to mortality stands and to observe whether increasing soil pH results in increased soil and root Mn and Mg concentrations and in improving the health of declining trees. A magnesium—form of limestone would be recommended to compensate for Mg deficiencies in mortality pockets. Mn fertilization alone might also be tried to determine how much effect this nutrient is having on RPPM symptoms. Lange (1969) found that Mn-deficiency symptoms in P. radiata could 179 be corrected on a long-term basis with both foliar spray, and soil applications of MnSO4. Another step would be to acidify soils around healthy mature red pines and observe whether RPPM symptoms develop. Both Mn and Mg are involved in chlorophyll production and there may be a link to deficiencies of these nutrients to the symptoms of browning and stunted growth in needles. Further investigations should also focus on the biotic factors associated with the disease. The often circular pattern of disease pockets suggests a strong biotic link. The only biotic factor examined in our study was in attempting to isolate pathogenic fungi from red pine roots. Unlike the Wisconsin studies, we were not able to isolate any significant pathogenic fungi even when using selective media for Leptographium. Although casual observations were made of bark beetle (Ips pini), pine root collar weevil (Hylobius radicus), Armillaria root rot (Armillaria, spp.), and Diplodia tip blight (Sphaeropsis sapinea) activity in some pockets, we did not quantify any of this information. In addition, the roots removed from disease as well as healthy plantations did not appear to have mycorrhizal infection. The role of mycorrhizae as a factor in the health of red pine stands should be examined further with regard to red pine pocket mortality. 180 B I BL I OGRAPHY Arp, P.A., C. Akerley and K. Mellerowicz. 1989. Distribution of aluminum in black spruce saplings growing on two New Brunswick forest soils with contrasting acid sulfate sorption. Water, Air, and Soil Pollution 48: pp. 277—297. Balsberg-Pahlsson, A. 1990. Influence of aluminum on biomass nutrients, soluble carbohydrate and phenols in beech (Fagus sylvatica). Physiol. Plant. 78, pp. 79- 84. Clements, J.R. 1970. Shoot responses of young red pine to watering applied over two seasons. Can. J. Bot. 48, pp. 75—80. de Ronde, C., D.B. James, N.T. Baylis, P.W. Lange. 1988. The response of Pinus radiata to manganese applications at the Ruitersbos State Forest. South African Forestry Journal, No. 146, September 1988, pp. 26-33. DeWald, L.E., E.I. Sucoff, T. Ohno, and C.A. Buschena. 1990. Response of northern red oak (Quercus rubra) seedlings to soil solution aluminum. Can. J. For. Res. 20, pp. 331-336. Foy, C.D. 1974. Effects of aluminum on plant growth. In: The Plant Root and its Environment, F.W. Carson, ed., Univ. Va. Press, Charlottesville. pp. 602-642. Garret, P.W. and Zahner, R. 1973. Fascicle density and needle growth responses of red pine to water supply over two seasons. Ecology 54, pp. 1328-1334. Grey, D.C. 1988. A review of the role of manganese in pine plantations. South African Forestry JOurnal, No. 145, June 1988, pp. 42-46. Johnson, D.W., M.S. Cresser, S.I. Nilsson, J. Turner, B. Ulrich, D. Binkley and D.W. Cole. 1991. Soil changes in forest ecosystems: evidence for and probable causes. Proceedings of the Royal Society of Edinburgh, 97b, pp. 81-116. 181 Kielbaso, J.J. and K. Ottman. 1976. Manganese deficiency— contributory to maple decline? JOurnal of Arboriculture, Feb. 1976, pp. 27—32. Klepzig, K.D. and J.C. Carlson. 1988. How to identify red pine pocket decline and mortality. USDA For. Serv. NA- GR—19. Klepzig, K.D., K.F. Raffa and E.B. Smalley. 1991. Association of an insect-fungal complex with red pine decline in Wisconsin. For. Sci. 37:1119-1139. Knezek, B.D. and H. Greinert. 1971. Influence of soil Fe and MnEDTA interactions upon the Fe and Mn nutrition of bean plants. Agronomy JOurnal, Vol 63, July-August 1971, pp. 617-619. Lange, P. 1969. A manganese deficiency in Pinus radiata at Klein Gouna, Knysna. Forestry in South Africa, No. 10, Oct. 1969, pp. 47-59. Motley, J.A. 1949. Correlation of elongation in white and red pine with rainfall. Butler Univ. Bot. Stud. 9, pp. 1—8. Murach, D. and B. Ulrich. 1988. Destabilization of forest ecosystems by acid deposition. GeoJournal 172: 253- 260. Raffa, K.F. and D.J. Hall. 1988. Seasonal occurrence of pine root collar weevil, Hylobius radicis Buchanan (Coleoptera:Curculionidae), adults in red pine stands undergoing decline. Great Lakes Entomol. 21:69-74. Shortle, W.C. and K.T. Smith. 1988. Aluminum-induced calcium deficiency syndrome in declining red spruce. Science: 240:1017-1018. Ulrich, B. 1985. Interaction of indirect and direct effects of air pollutants in forests. In: Air Pollution and Plants, Proc. of the 2nd Euro. Conf. on Chemistry and the Environment, Clement Troyanowsky, Ed., pp. 149-180. Ulrich, B. 1989. Effects of acidic precipitation on forest ecosystems in Europe. In: Acid Precipitation, Volume 2, Biological and Ecological Effects, D.C. Adriano and A.H. Johnson, Eds., Springer-Verlag, New York, pp. 189- —272. IL APPENDIX Original Data Soil, Needle, and Root Elemental Concentrations APPENDIX umdosm maumo ou >Hoh opma “lam aam Emu new goal dammumncflce cam Icousm um mocmum coflumucmfim Eouw maaom oHHoNUOQm Mo “Amo pom muoam xoono coEEoomom “Aowe pom 042v .qu .quv Aqu paw .HOV \ NO AEQQV muoad xomzo ocmum Hausumc mxmq anamo pom muoHQ xomco mwumficmz:consm mxmq maccflz ”muoam ommmmflo omumflcmz mcoflumnucmocoo ucoEmHm “Amoe mm.mH m>.mfi ow.vm mw.o Hm.qm em.H om.wm ofi.o .>mQ Dem mo.qm ov.m> m>.HmH oq.m oa.mwfi om.om om.mm mm.w zm m.v o o «we m.>w mm Nwa m Hmfi o.mm m.mw m.v o 0 m4: O.Hm me HHH m mmH 0.5H o.w v.q o 0 m4: o.mm mm mva m mHm O.Hm m.fiw v.v o 0 H42 mo.HH oo.m nn.mm mv.o nm.mH mv.m Hm.mm NH.o .>mQ Gem mm.mm om.Hw mm.omH om.m om.mmfi oo.mH oa.mm mm.w zamz o.mm an ova m mmfi o.mm moH v.v o m woe o.qm aw ova m mma o.mm o.vm m.w o m «we o.vm mo noH m mmfi o.>H o.ov v.v o m mg: m.mH me wma m mvfi o.nH o.o m.w o m qu >.m mm Hm m ova o.nH m.n H.v o m H92 mm.oH OH.HH mo.mm mv.o Hm.vm Hm.m mH.m> mH.o .>mo obm mo.om oo.mw no.nmfi om.m ov.>>a oa.mH om.>m wm.w ZH m.© H.v o u qu o.vm em mm m mmH o.nH o.>H m.v o 0 H42 22 mm Ad 02 do K m mm +aou *GOflu uoam Inn: nnnnnnnnnn nuns: Aemmv nuaoeme(Innlnnnulnnluuunuln [whom Imooq .vmmH paw .mmmH .mmma .H magma 182 APPENDIX OO.OH NO.m Om.ON OO.O OH.OH ON.N ON.OO O0.0 .>mo new OO.OH ON.OO OO.OOH OO.O O0.00 OO.HH O0.00 Om.O zamz 0.0H OO OOH O OO 0.0 OOH 0.0 m o Owe m.mfi Om OOH O OO m.mH OOH 0.0 m o OOH O.Om OO OOH O mas O.mH 0.00 O.O m 0 mg: O.m OO NNH O OOH m.O m.OH m.O m 0 ma: 0.0N mO HOH O OOH O.NH O.mm O.O m o Hag OO.m OO.OH OO.mm OO.O O0.0N OH.N m0.00 NN.O .>mo new ON.O O0.00 Om.HmH O0.0 Om.mOH OO.OH Om.mm Om.O zmo gem ON.O ON.OO mm.ONH O0.0 OO.mOH O0.0H ON.NO NO.O zamz m.OH OO OO O OOH 0.0 OON H.m m o Owe O.O OO ONO O OO O.O O.OH O.O m o Owe O.O an OOH O OOH O.eH m.mO m.O m o ma: O.H mm OHH O me m.O 0.0 O.O m O Na: O.HH OO OOH O OOH m.NH 0.0m m.e m o Hg: 22 nu A4 or «u n m mm +aou «cow» uoam unlulunlnuulllnlull €53 nucofiaam Inulnnlulluunuuulul uwwom ..noon ”v.0coov .H magma 183 APPENDIX mo.m Om.ofi Om.mH mq.o mH.mm Om.o HO.mm OH.o .>mQ Dem Om.m om.oq mv.wm oq.m om.OOH om.m om.fim om.v 24m: O.m mm «O m we o.m ova o.m m o mme O.m mm m» m mHH o.m omH o.m m o «we m.oH om om m mmH o.m m.mm H.m m 0 mg: O.H om NHH m moH m.m m.om m.v m 0 ma: m.v ow Om m me m.m o.mm m.v m 0 HA: mo.H Om.m me.ma mw.o mm.mw Hm.fi wm.om om.o .>mo Dem «O.H ow.Hv mm.om oq.m om.wm on.m om.ow mo.m zmmz O.H om no m mp o.m ooa o.m m m mwe m.H mm mm m mm o.m o.mw o.m m m woe m.m me m« m Hma m.mH m.Om v.m m m qu O.H me no m ms m.m o.mm m.v m m was an mm m mp m.m o.Om m.v m m H42 mm.fi OH.mH Om.mv mq.o oo.om vm.m mm.mmfi wm.o .>ma Dem mm.m oo.Hm mm.fir om.m on.HHH ow.m om.oHH mo.m z0.nm 00.0 00.n0 Hn.0 0v.N HH.0 .>mQ 090 >0.v0 00.0HH N0.H00 00.0H 00.0vN 00.mv 00.0 00.0 deE 0.00 00H 000 HN 00N 0.0a 0.VH H.v O m Nm 0.00 0HH NNm HN 0NN 0.0v 0.0 0.0 O N Nm v.H0 00H mmm HN 000 0.00 0.0 0.0 O H Nm 0.Hm 00H 0mm 0H 0NH 0.00 0.0 0.0 O m Hm 0.0V 00H 0NN 0.0H 00N 0.0V 0.0 0.0 O N Hm N.00 0HH 00N 0.0H 0N0 0.0m 0.HH H.v O H Hm 00.N 00.0 00.0N 00.0 00.0N 00.0 00.v 0N.0 .>ma 090 00.0 00.00 00.HNH 00.NH 00.00H 00.VH 00.0w 0H.0 24m: 0.0 0N N0 NH 0NH 0.VH 0.v> 0.0 m H OOH v.NH 00 NOH NH 00 0.vH 0.N0 0.0 m H OAS 0H.vH 00.0H 00.00 00.0 00.0 0N.> 00.0N 0N.0 .>mo Dem 00.0N 00.00 N0.m0H 00.NH 00.0NH 0P.HN 00.00 00.? zmmz v.0 0N 00 NH 0NH 0.0H 0.00 N.0 m H 009 0.00 00 MNN NH 0NH 0.0N 0.00 0.0 m H 0H2 00.nN 00.0 00.0 00.N 00.0w 0N.» 0>.0H 00.0 .>mm 050 00.00 00.Hh 00.00H 00.0H 00.00H 00.00 0>.N0 00.v z¢m2 0.Hm ND 00H 0H 0NH 0.0m 0.N> h.v O H 005 H.00 Hr 00N NH 00N 0.H0 0.00 0.v O H OH: 22 mm Ad 02 (U M m mm +aou & uoam nunuuuunnuululnnuul Hmnmv nuGQEOHm(Innuuluuunlnlunuuuu :Huom OHQEum :0 . uCOUv .H mHnme 185 APPENDIX OO HO.H OO.OO OO.O OO.OO O0.0 O0.0 OO.O .>mn OOO OO O0.00 OO.OOH OO.HH O0.00H OH.HO OH.O OO.O ZOO: O. OO HOH HH OOO 0.00 O.O O.O o H Oo O. OO OOH HH OOH O.OO O.O H.O o H Oo O. OO OOO OH OOH O.OO 0.0 O.O o H Ho OO OH.O OO.OH OO.O OO.OO OO.H OO.O OO.O .>Oo OOO OO OO.HO OH.OOH OO.OH OO.OOH OO.O OO.OO OO.O ZOO: O. OO OOH OH OHH 0.0 0.00 O.O O O OO O. OO HHH OH OHH 0.0 O.HO H.O O O OO O. HO OOH OH OHH 0.0 0.00 0.0 m H OO O. HO OOH HO HOH 0.0 O.HO 0.0 O O HO O. OO OOH 0.0H OO 0.0 0.00 0.0 O O HO O. OO OOH 0.0H OO 0.0H 0.00 0.0 m H Hm OO.O OO.O O0.00 OO.O O0.00 O0.0 O0.0 OH.O .>OO OOO OH OH.OO OH.HOO OO.OH OO.OOH OO.OH OO.OH OO.O ZOO: O. OO OOO HO HOH O.OO O.OO O.O O O OO O. OO OOO HO OOH O.OH O.HO O.O O O OO O. OO OOO OH OOH 0.0H 0.00 O.O m H OO O.O OOH OOO HO HOH 0.0H O.OH O.O O O HO O. OOH OOH 0.0H OO 0.0H O.OH O.O O O HO O. OO OOO OH OOH 0.0H 0.0H 0.0 m H HO Ilmm an A4 or «u u H mm +aou O uon Innuuuuunuuuunnannu Asmmv nunQEGHM(unuuunnnnunnnnnulun uwuom ngEdm “G.HQOUV .H OHHOO 186 .coOHHoz Hmv oHpomm pcm .Hmv :oNHqu poHm .Hov coOHHon oHcm0Ho n coOHHom+ .muoHd ommmme omuchmzaconsm cflzqu .H v ocHHoop map 00 mono map oonpzo pom .Hmv ocHHooo map mo ompo .Hov umxooa wo Hmucmo u coHumoqu APPENDIX In: nnnnn unlununuuuu HEQQV nucafioam uuununuununnunuluuu Iwuou Canaan _MU.HCOUV OO.O OO.O OO.OO O0.0 O0.0 OO.OH OO.H OO.O .>Oo OOO OO.H O0.00 OO.OO OO.HH OO.OOH OO.OO OO.OO OO.O 2mm: O.H OO OO HH OOH O.OO O.OO O.O m H Oo O.H HO OO HH OOH O.OH 0.00 0.0 m H Oo O.H OO OOH HH OOH O.OO O.OO O.O m H Ho OO.H OO.O OO.OO OO.O OO.OH OO.O OO.O O0.0 .>OO OOO O0.0 O0.00 OO.OOH OO.HH O0.00 OO.OH O0.0 OO.O ZOO: O.O OO OO HH OO 0.00 O.HH O.O m H Oo O.O OO OOH HH OOH 0.00 O.OH O.O O H Oo O.H OO OHH OH OO 0.0H 0.0 H.O m H Ho 22 mm as as «u m H mm O uon .H mHQmB 187 APPENDIX Nm.0 O0.0 NO00 NO0H 00.0 000 00H N.0H N.00 0.00 0NN 00.0 H.N0 HHH Oom NHE HN.0 00.0 0000 00OH 00.0 0O0 00H 0.0H 0.00 0.0N 00N 0O.H 0.00 HHH 000 NH: 00.H N0.0 OON0 000H 0N.0 0OO 00N O0.0 O.H0 0.N0 00H 0O.H 0.0N HHH 009 NH: 00.0 N0.0 00O0 O00 H0.0 000 00N 00.0 0.00 H.00 0NH 0H.0| O.HN HHH 009 NH: H0.0: ON.0 O0N0 00ON 00.N OHHH H00 O0.H N.O0 0.NN H0O 00.0| 0.0N HOH Rom NH: 0H.0u 0N.0 NOON H000 0N.N 00HH H00 O0.0 00H H.HN 00O 00.H| 0.0N HOH 900 NH: OO.H: 00.0n NNON 0ONN 0O.H H00 000 N0.Hu 0.NO O.NN 0N0 O0.0| 0.0N HOH 009 NH: 00.0n O0.0 00OH 00NN 00.N 0O0H 0O0 N0.0 N.N0 H.HN 0H0 00.0: 0.0N HOH 009 NH: 00.0: 00.0 HOHO HHOH N0.0 000 0NN 00.0 0.0N 0.0H O0N 0O.Nn 0.HN H0 Oom NH: 00.0 00.0 HN00 O00H O0.0 N00 00H N.0H 0.H0 0.0H 000 NH.O 0.0N H0 500 NH: H0.0 H0.0 0O0O 00NH O0.0 000 0HH N.0H 0.00 0.0N HON OO.N 0.00 H0 0H: NH: 0N.0 00.0 0HHO O0OH O0.0 ON0 ONH 0.NH 0.00 0.0H 00N 00.N 0.0N H0 0H: NH: H0.0 0H.0 000O 000 H0.0 000 NOH 0.0H 0.00 0.0N 0NN O0.0 0.0N H0 009 NH: 00.0n O0.0 00HO O0O HO.N 000 00N 0.0H: 0.00 0.0H O0H H0.0u 0.0H H0 009 NH: 00.0u O0.0 000O 00HH 00.0 0H0 NON 00.0: 0.HN 0.0 00H 00.Nu N.HN H0 Oom 0H2 O0.N| 00.H 0O00 00NH 00.0 00O OHN 0.HH: 0.0H 0.0 0O0 00.0: 0.H0 H0 Oom 0H2 HN.H| 00.0 0000 0ONH 00.N O00 NOH 00.0: O.HN 0.0 0O0 O0.Hn 0.0N H0 0H2 0H2 0H.Nu OO.N 0000 0O0H 0O.N 0O0 0HN O0.Nn 0.0N 0.0H 00N 0O.Nn 0.0H H0 0H2 0H2 O0.0: H0.0 HNOO 000H O0.N 00O 00N N0.Nn 0.00 0.NH HHO O0.0| N.0N H0 009 0H2 N0.0 No.0 00HO 000 N0.0 0N0 00H 00.0 0.00 0.0N NON 00.H 0.0N H0 009 0H2 00.0 H0.0 OO0O 000 00.0 O00 H0 HN.0 0.0H 0.0H 0HH H0.0 O.NN HN Oom 0H2 N0.0| 00.0: 0O0O 00O H0.0 N0O HO 00.0 N.OH 0.0 0HH 00.0 0.HH HN Oom 0H2 00.H O0.0 NOHO 0N 0O.N O00 H0 00.0 0.0H 0.0H 00 00.Hu 0.0H HN moe 0H2 H0.H ON.O 00OO 0ONH 00.0 0O0 0HN O0.0 0.0N 0.0H 00N N0.0| 0.0N HN moe 0H2 madam can unannondb .noavoonnwnGMIH .ouam oxuq ofidoaz mo Hz M 40 Do 02 Ad mm mm m 2: no ZN * «Hoboq HOHm In:unuuanuuuunauuuunnunnn Hfiflmv unafioHM In:nullnuuunnunulunuauuuuuu ooua clown .000H pmd0d¢ -OHOO .HOOHOO HOcoHHOz mmHmHCOZIcOHsm .HoOO .OOO .OOOV aHOwu¢ucHnO new HOHz new OH: .Hsz ome oHccHz um mmHUmmc UHouHmmzuN paw 1H How moch Umn H2000 How oHumaoumemmv OzuHmmma: paw A2000 How 0Hpmaouma>mmv OguHmmn How H8000 mumU ucmEmHm memmz .N mHQmO 188 APPENDIX 00.0 00.0 0000 0NOH O0.0 000H O0H 0.0H 0.00 0.0H 00O O0.H 0.00 N0 900 HHS O0.N 00.0 ON0O HO0H 00.0 000 0NN 0.HH 0.0N N.0H 000 0N.0 0.00 N 900 HHS 00.0: 00.0 00OO O00H Hm.O 000 NON OO.H: 0.0N H.NH OHO 00.H: 0.00 N 002 HHS 0N.0: 00.H 00O0 HO0H N0.0 O00 N00 HO.0I 0.00 0.0 O00 O0.Nn 0.0N N0 002 HHS 0N.N: 0H.N N0OO 000H 00.0 0H0 OO0 H0.Hn 0.00 0.0 00N OO.H: H.HN N0 009 HHS 00.0: HN.0 OOHO HO0H 00.0 HOO 000 NH.N 0.N0 N.HH 0O0 N0.0| 0.N0 N0 009 HHS H0.0: NO.N 000O HO0H 00.0 0H0 O0N O0.0 0.00 0.0 000 00.H: 0.0N NO 900 HHS 00.0 00.0 00N0 ONON 00.0 OO0H O0N ON.O 0.00 0.0 00O 00.H: 0.00 NO 900 HHS H0.0 00.0 000O 0O0H 0m.O N00 ONN 0.0H 0OH 0.0H O00 N.0H 0.HO NO 002 HHS OH.H O0.N 0NOO O00H ON.0 H00 HON O0.0: 0.00 0.0 0H0 0H.O 0.H0 N 002 HHS N0.H 00.0 N00O HHNH 00.0 0N0 00N 00.0 0.HO 0.0 HON 00.N H.ON N 009 HHS HH.0 0N.0 000O 0O0H 00.0 00O 0ON 0N.0 0.00 0.HH HON 0N.N 0.0N NO 009 HHS madam van Mauauou .uodvoonnuuamnu .ouflm 030A adddflz 00.0 0O.N 00HH 000 00.H 00H 00 00.0 0.HN 00.0 00 00.N 00.0 .>ma 090 H0.0: 0N.0 O0OO 0OHH 00.0 O0O NOH 00.0 0.00 0.0H 00N 00.0: 0.NN 2402 HO.OI 0H.0| HO00 OO0 0O.H O00 00H O0.0 0.N0 0.0 0NH 00.0n 0.0H O0H 900 HHS 00.0: O0.0 OHOO 000 NO.N 000 0OH OH.O| 0.00 0.0 0HH 00.0n 0.HH O0H 900 HHS 0H.0: 0H.H 0OHO 0N0 0H.N 0H0 0OH HH.0 0.00 0.HH 0NH O0.Na 0.0H O0H 009 HHS H0.0 HN.0 0O0O 00N Hm.N 0N0 00H 00.0 0.00 0.0H 0NH 00.0: 0.NH O0H 009 HHS 0O.N 00.N O000 N00 00.0 OOO N0 0.0H 0.N0 0.0H 00H O0.0 0.0H 00N 900 HHS 0N.O 00.0 HO00 NO0 00.0 00O OO 0.0H 0.N0 H.0H 00H ON.0 0.0H 00N 900 HHZ 00.0 O0.0 0000 0O0 00.0 NO0 H0 0.0H N.O0 0.0H NN O0.H 0.0H 00N 009 HHS OO.N 00.0 0HO0 N00 0O.N OOO 00 0.0H 0.00 0.0H 00H 00.0 N.HN 00N 009 HHZ 00.0: 00.H 000O 0HOH ON.0 00O 0HH 00.0 N.0N 0.0H N00 00.0 N.ON HH 900 0H2 ON.0: 00.H 000O HO0H 00.0 0OO 00H 0N.0 0.0H 0.0 H00 00.0: 0.0N HH 900 0H2 00.0: HN.0 0H0O 0OO OO.N ON0 00H 00.0 0.0N 0.0 O0N H0.0: 0.NN HH 009 0H2 H0.0: H0.0 O0OO 00O NO.N 0O0 00 0O.N 0.00 0.0H 0ON N0.0| 0.0H HH 009 0H2 do Hz M do Do 02 A4 mm mm m z: oo Zn * «Haboq uoam Inlnnununlnnuuuuunnnunnun A8000 uaoeme Innun:nuunuuulnuulunnuunluu cows ciouo H0.0:000 .N mwH009 189 HO.Hcoov 0 mHQM9 190 IIllllllllll[III-llillllllrJlllllllmmlllllllllllIlIlllllllllllllllIllIllllllIlIIIIIIIIIIIIIIIIIIIIIIIIII OO.N: 0O.H 0O00 000 ON.0 0OO 00 N0.0 0.0N 0.0H 00H H0.Hn 0.0H 0OH 900 O09 00.0: H0.0 000O 00O 0H.0 0O0 OO 00.N N.0N N.0H O0N 00.0 0.0H 0OH 900 O09 0H.Hs 0O.H H000 000 00.0 HNO 0O O0.0 H.00 N.0H OHN 0H.0: N.0H 0OH 009 O09 00.0 00.0: N00O O00 HO.N 000 00 00.0n 0.00 0.0 00H O0.0: 0.0N 0OH 009 O09 00.0 NO.H 00OO 000 00.0 00O 0NH 00.0 0.0N O.HH: 000 0H.0| 0.0H 00H 900 009 00.0 0O.N 0000 O0NH OH.O O00 HNH O0.0 0.0N H.N 0OH 00.H 0.0N 00H 900 009 H0.0 0O.N 000O 000 0N.0 HNO 0NH 00.0 0.0H 0.0 HOH 00.N 0.0N 00H 009 009 OO.N 00.N 00N0 H00 ON.0 0O0 HHH 0N.0 0.0N 0.0 00H 00.0 0.0H 00H 009 009 00.N: O0.0: 00N0 OOoH HN.O O0O 0NH N0.0 H.0 0.0H O0H 00.0: 0.0H 00 900 009 0O.H| 0O.H NHO0 000 HO.N N0O 0NH 0O.Hu H.0H: N.0H 00H 0.HH: 0.NH 00 900 009 O0.0: H0.H 00O0 OO0 00.0 HO0 O0H O0.0 0.0 N.ON O0H 00.0 0.0H 00 009 009 00.0: O0.0 OOO0 0O0 00.0 O0O 0HH HN.0 0.Hu O.HN 00H 00.0 0.0H 00 009 009 0N.0: 00.H 00O0 00HH N0.N 00O 00 00.H 0.0H N.0H 0ON 0H.N 0.0N O0H 900 009 X 00.0 0O.N 00O0 HO0H O0.0 000 OHH O0.0 H.0m 0.0H 000 00.H 0.0N O0H 900 009 m O0.0: 0N.N 0ON0 O0NH 0N.0 HNO 00 H0.0 0.0N 0.0H 00N 00.H 0.0N O0H 009 009 m 0N.Hn ON.H 0O0O O0NH 0H.0 OOO 00H 00.0 0.00 0.0H 0NN 00.H 0.0N O0H 009 009 m 00:00 vow maudnondn .noauoonuquMIH .ouwm awnwudnnwna HO.N 00.N 00O 000 00.H 0NH 00 O0.0 0.0N OH.O 0NH O0.0 00.0 .>00 090 H0.0 NN.O 0O0O OOOH 00.0 000 OON 0H.0 0.00 O.HH 0HO 00.0 0.0N 2002 00.N N0.0 00N0 HO0H 00.0 ON0 00H 0.0H 0.N0 0.0H 0O0 N0.H 0.HO N0 900 HHS O0.0: 00.H H0OO 000 00.0 000 HNN 0.0H| 0.00 0.HH 000 00.0u 0.0H N0 900 HHS 0H.N: 00.N 0000 HO0H H0.0 00O NON 00.0: 0.00 0.0 000 H0.0: 0.NN N0 0H2 HHE O0.H: HO.N 0000 HNOH OH.O 00O NON 00.H: 0.0N 0.0 000 00.N: 0.0N N0 002 HHS NO.H HH.O 0O0O 000H H0.0 N0O NON 00.0: 0.N0 0.HH 0H0 0H.H| 0.0N N0 009 HHE 00.0 N0.0 N0OO OOmH ON.O 000 NOH 0.0H 0.00 N.HN 0O0 00.0 0.0N N0 009 HHS mo Hz M ‘0 Do u: at mm mm m 2: no 2N * «Hobon HOHA Inunuuuuunnnnnuullnnuuuau 6300. uGQEMHM‘nuuuununuuuunuuuunnuuunuuun cons clown APPENDIX 00.0 0.0H 0ONO OHOH NN.O 00O NON 0.0H 0.H0 O.N0 O00 O0.0 0.00 0OH 009 O09 O0.0 0.HH 00N0 000H 00.0 NNO 00H 00.0 0.00 0.NN 0OO H0.0 0.0N 0OH 009 O09 O0.0: 0H.0 OH0O 0NO 00.N OO0 00 00.0 O.HN N.0H 0NH 0H.0 0.NH O0N. 009 O09 0N.0 00.H 00O0 OOOH 00.0 O0O 00H 00.0 0.H0 0.0N 000 0N.N 0.0H O0N 900 O09 H0.0 O0.H 0N00 O00 00.N HO0 00 O0.H 0.0H 0.0H NOH 00.0 0.HH O0N 0H2 O09 H0.0 0O.H 000O N00 0N.N 000 0O 00.0 0.H0 0.0H OO0 N0.N 0.NH O0N 0H2 O09 O0.H: 0N.0 00NO 00O 00.N O00 OHH OH.H| 0.0 0.0H 0OH N.HH: H.0H O0N 009 O09 N0.H: 0H.0: 000N 000 H0.H 0O0 O0H 0N.0 N.H N.0H H0O 0.0H: N.0 O0N 900 O09 N0.H 00.0 NNOO 000N 00.0 OO0 00H 00.0 0.NO N.OH H00 H0.0 0.0N 0O 900 009 O0.0 O0.0 00NO 00NN 00.0 H00 O0H 00.0 0.00 H.OH 00O 00.0 0.00 0O 900 009 00.N O0.0 O000 O0OH NH.0 000 0OH N0.0 O.H0 N.0H 00O HN.0| 0.0N 0O OH: 009 00.H 00.0 N000 0NON 00.0 0OO 00H 00.N 0.00 0.NH HHO 00.0 0.0N 0O 0H2 009 N0.0: 0H.O N0O0 0O0N 00.N 000 0NO 00.0| 0.00 H.0N 0OO 00.0 0.0H 0O 009 009 N0.0 0.HH O000 NO0H 0N.0 H0O 00N 00.0 0.00 0.0H 0OO 00.0 0.0N 0O 009 009 OH.0: 00.N 0O0O H00 ON.0 00O HHH 00.0 0.00 0.NH 0H 00.0 H.0H 00H 900 009 O0.0: NH.N H0OO O0NH 0O.N 000 ONH 00.H 0.0N 0.0H NNO 0O.N 0.0N 00H 900 009 H0. 1 00.N 0OOO N00 O0.N O0O 0HH 00.0: N.HN 0.0H 000 00.H 0.HN 00H 0H2 009 00.0 O0.0 00OO 000H 00.0 000 00H O0.0 0.00 0.0H OOO 00.0: O.H0 00H 009 009 N0.0: O0.0 0OH0 0HOH HN.0 0O0 OOH 00.0 H.OO O.HH N00 0O.H 0.NO 00H 009 009 00.0: 0O.N O0H0 O00 O0.H 0NO O0H N0.0 0.00 H.OH 000 OH.0: H.0N 00 900 O09 0H.0: 00.0 00H0 00O 00.N 00O 0HH 0O.N 0.0N 0.0N 000 O0.H: 0.0N 00 900 O09 O0.0: H0.0 HOH0 ONO 0H.N 000 O0 0H.N N.ON 0.0H 000 HN.Nu H.0H 00 0H2 O09 00.0u 0H.0 0OH0 0O0 0O.H 00O 00 00.N 0.NN H.OH 00N 00.N 0.HN 00 002 O09 00.0: 00.0 H000 O0O HH.N HO0 0HH 0O.H 0.00 0.0H 00H 0H.0I 0.0N 00 009 O09 00.N 00.0 O0HO 000 0N.N O0O 0NH 0O.H N.O0 H.0N OH0 0H.0: 0.0H 00 009 O09 madam can hauHuom .noduoonnunOMIH .ouwm naumudunwna 00.N 00.0 0OO 0ON N0.0 0O ON 0O.N 0.0H 0.0 NO OH.0 0H.O .>mo 090 NH.Hu OO.H 00N0 N00 00.0 OOO O0H O0.0 0.0N 0.NH NNN O0.0: N.0N 2002 mo Hz M 40 Do 02 Ad. mm mm m 2: no an * aa0>¢q uoam Inniuununnnuunuunuuuuunun H3009udaaoamiIn:unuuunnluuunuununnunnunn AU coma atouo .Hcoov .N mHQm9 191 APPENDIX 00.H 0N.0 0O0N OHOH 00.N 000H OOH 0.HH N.00 0.0N 00N 0H.H 0.0N HN 009 0H2 0O.H 00.N HO0N NOOH 0N.N 0H0 O0N 00.0 H.00 0.0H 0ON 00.HI 0.0H HN 009 0H2 monam_vou anuduonnb .uoduoonuunoMIN .ouum 030A owdnfiz 00.0 00.0 0NOH 000 00.H 0HH 00 0H.0 N.0H N0.0 HOH 00.N 00.0 .>mo 09m 0N.H 0N.0 0N00 000H 00.0 0OO NOH O0.0 H.00 N.0H 00O 00.0 0.0N 2002 00.0 00.N 00O0 000 HN.0 000 O0 H0.0 0.0N 0.NN 0ON 0N.0 O.HN NN 900 009 0N.0 00.0 O0O0 O0O 00.0 000 00 00.0 0.0N 0.0H 00N 00.0: 0.0H NN 900 009 0H.0 N0.0 0OH0 0ONH OH.O 000 00N 00.0 0.00 0.0H 000 00.0: 0.0N NN 0H2 009 00.0 0N.O 00N0 NOOH 00.0 000 O0H 0N.0 0.00 0.0N 00H OO.H: 0.00 NN 0H2 009 O0.0 00.0 000O 0H0 O00 O0 0.0N 0.0H O0N O0.0: N.OH NN 009 009 N0.0 00.0 0OOO 00O 00.0 000 00H O0.0 0.00 0.0N 000 O0.0| 0.0H NN 009 009 O0.0u 00.0 0000 00HN 0H.0 0H0 0OH N0.N 0.0N 0.0H O00 N0.N N.00 NH 900 009 00.H: 00.0 HO00 0O0H 00.0 00O O0H O0.H 0.0N H.0H 000 N0.0 H.H0 NH 900 009 O0.0 00.0 O0OO 00OH HN.0 0HO 00H 0O.N N.0O N.0H 000 OO.H 0.0N NH 0H: 009 00.0: 00.0 000O O00H H0.0 000 00H 00.N N.0O 0.0H O0O 0N.N N.ON NH 0H2 009 00.0 00.0 NO0O 0ONH 00.0 000 HON 00.0 0.00 0.0N 0O0 00.0 0.00 NH 009 009 0H.Hu 0H.0 OOO0 O0OH N0.0 00O NOH 00.0 0.00 0.0H N0O H0.0 0.0N NH 009 009 00.0: 00.0 0000 0HON 00.0 000 O0H 00.0 H.0N H.OH 00O 00.H 0.0N N0 900 009 O0.0 0N.0 000O HNON NO.N OO0 0OH H0.0 H.0O 0.0H 0NO 00.H 0.0N N0 900 009 0H.0 N0.0 0OO0 0HOH 00.0 0N0 O0H O0.H 0.0N 0.0H HNO H0.H 0.00 N0 0H2 009 00.0: 00.0 OO00 HO0H 00.0 000 00H 00.0 0.00 H.0H OHO 00.0: 0.00 N0 OH: 009 00.0: 00.0 O000 0OOH O0.0 000 NOH 00.0 0.00 0.0H 00O O0.0 N.NN N0 009 009 O0.0: 0.NH 00O0 00HH O0.0 O00 HNN 0H.N 0.00 H.HN 0H0 00.0 0.0N N0 009 009 00.0 00.0 HONO 000N 00.0 000 00H 0N.0 O.N0 H.0H 0N0 00.0: 0.0N 0OH 900 O09 00.0| 00.0: 0000 0O0H 00.0 000 O0H 00.0 0.0; 0.HN H00 O0.0 H.NN 0OH 900 O09 00.0a N0.0 000O 0NNN OH.O 0H0 00N 00.0 0.HH 0.0H 00O H0.0 0.0N 0OH 0H2 O09 H0.0 00.0 00N0 O00H 00.0 0N0 00H 0H.0 0.00 0.0H 00O 00.0 0.0N 0OH 0H2 O09 mo Hz M do Do 02 Ad mm mm m 22 no 2N * «Ho>oa 90H0 Insulinulnuuusnnuuaununun H8000 unqaon nuuuunuuIannauuuununuunuuau 0009 ntouo H0.Hcoov .N mHQM9 192 APPENDIX m0.M 00.H HMHN 0NHH 00.N 0M0 MqN N0.v H.Hm M.NH 00M H0.0: 0.0 0MN Hom HHZ HH.¢ 0H.N 0QMN NMH VH.M M00 00H 0.0H m.00 0.0N «PM mo.M 0.0H 0mm Rom HHZ mm.M 00.H MOVN HMOH 00.M 000 00H 0.MH M.>0 N.HN MMM 0N.M 0.0H 0MN moe HHZ m0.M 00.N hmHN MONH 0N.N v00 v0H v.oN 0.0m m.MN NOM M0.N O.HN 0MN mOB HHZ OH.0I N0.0: mme Mv0N 00.N 000 0NN M.NH 0.m> 0.0 000 ON.0: 0.0M HH 90m MHZ 0M.H O0.H 000N 000M 00.M oqu 0HM 0o.m 0.00 M.HN PPOH mm.ou H.Hm HH 80m MHZ NM.m1 0v.o MO0N «NOH H0.N 0H0 HHN NM.> 0.00 0.0 00m H0.0: 0.0H HH woe MHZ 0M.NI 00.H HHvN 00.N 000 0HN M0.0 v.00 N.OH H00 00.H 0.NN HH mOB MHZ M0.M OO.N >00M 000v hm.M HMNH MMN 0.0N MHH 0.0N 0H0 mM.v v.0v HHH Hom NHZ om.M 00.H MO0M 0NM¢ 0N.M nva 0MN 0.0H 0.00 0.0N 0mm 00.H 0.0V HHH Bom NHZ H0.0 mH.H 00Mv Hoom 0N.M 0N0 mnm M.oH 0.00 0.NM NON 0v.o O.HM HHH moe NHZ 0M.OI 0H.0: NOM¢ HOQN 00.M M00 POM 00.H: v.v0 0.HN vbm MM.OI v.MN HHH mOB NHZ 00.0: wm.o: 000N 000M MN.N vONH 0NM 0H.M HOH 0.0N omv 00.H: 0.HM HOH Bom NHZ «0.0: 00.01 mmwm 0NOM 0N.N 0vHH How VM.h NvH M.MH 00¢ 00.v| m.0N HQH Rom NHZ 00.H: om.ol 0O0N mva N0.H N00 00M 00.H n.00 H.HN 0NM M0.NI v.vM HQH moe NHZ 0M.HI 00.0 Nvmm oHNM 00.H OVOH NvM 00.0 M.00 0.0H oHv M0.MI 0.0N HOH mOB NHZ 0M.OI v0.0 om0v 00MN v0.M HMO 00M MH.0 N.00 >.MN Nvm 0v.HI 0.0N H0 90m NHZ 00.0 «H.M 0MOM >0NM M0.M MMHH Mvm v.nH NOH 0.0N 0M0 0H.M «.mm H0 90m NHZ 00.N 00.H 00mm Hn0H 0H.0 000 vMN 0.0H 0.00 0.00 000 N0.H 0.0N H0 QHZ NHZ v0.0 0H.M MO0M NOMN 0H.M HMO 00H M.NN 0.00 0.0M vmm m.M 0.HM H0 QHZ NHZ 00.0 0M.M MOMM vaH 0M.M H00 MHN m.mH M.N0 0.00 HON MN.N M.MN H0 mOB NHZ NH.H Hm.H MO0N MONH 00.N O00 MNN M0.0 0.00 m.>m 00M v0.0 0.0N H0 woe NHZ H0.0 00.N 000N >MON 00.N N00 00N N0.0 0.00 N.NN HOOH hM.N 0.0N HM Rom MHZ 00.N: MN.o VNNN Mm0N vo.N 000 NHM 00.H: >.Nv o.mH PMOH 0M.NI M.0N HM Rom MHZ 00.0: oH.H 000N MNNN M0.H N00 mHM 00.H: 0.00 0.HH >00 00.NI 0.0N HM QHZ MHZ N0.N: M0.0: NOMN vbMM H0.H M00 0HM NM.MI N.0M M.HH 0HOH 00.MI H.0N HM QHZ MHZ 00.0: Nv.H 00mm MO0H Hv.N 00> NHM vm.v 0.N0 v.0H VMF v0.0: 0.0N HM mOB MHZ NH.H MH.v 0NHN 000H 00.N ov0 00m v0.0 o.M0 0.0M mmw 00.0 M.HN HM m0? MHZ 00.H 0M.M OMOM O00H N0.N >00 00H M0.0 0.00 0.0H MOH H0.0 0.0N HN 90m MHZ 0M.MI om.on hva 0vMH 00.N 000 0HH 0m.b N.mv 0.NH HNN 0m.NI 0.HH HN Rom MHZ mu Hz M ‘0 DU 92 Ad mm mm m 22 GO ZN * «H0bon Moan Inuununuunuunnununnuuunuu .693 vaginaluluunuuuunnuuuIIIIIIIIIIIII acne Epouo HU.HCOUV .N mHQme 193 APPENDIX 00.0 N0.N NON0 000N 0H.0 000 00N 0.HN 0.Hn 0.0N 0HvH 00.H v.00 N0 50m HHE 00.0 00.N 000N 0HON 00.N 000H 00N 0.0N H.00 H.0N HO0H 00.H N.00 N0 50m HHZ Hn.0: 00.0: 000N 000N 00.H 000 000 >.oHu 0.00 0.0H 0HNH 00.0: 0.0N N0 DH: HHS 00.N: N0.0: NHNN 000N N0.N 00> 000 00.0: 0.00 0.HH HqHH 00.N: 0.0N N0 OH: HHz N0.H: 00.0 000N 000H 00.N 000 0P0 HN.o 0.00 0.NN 000 00.H: 0.NN N0 009 HHS N0.0: 05.0 0HON 000H 00.N 0H0 000 00.H: v.00 0.0H 000 00.0- 0.0N N0 009 HHZ 00.0 H0.0 000N 0000 0H.N 0HOH NON 0.0H 0.00 N.0N 000H 00.0 N.o0 N0 Rom HHS H0.0 H0.0 0H00 vHNN N0.N oooH 00N 0.0H v.00 0.0H 0H0 N0.0 0.00 N0 eom HHZ 00.H 00.H 0HON 0VNN 00.N 000 0H0 0N.) p.00 0.0N 000 00.0: 0.00 N0 DH: HHZ 00.0 00.H 000N 0NHO 00.N 000H 000 00.0 0.N0 O.HN 0H0 0H.H 0.00 N0 DH: HHZ H0.Nu 0N.O 0NON 000N 00.N 000 NNv N0.0 0.00 N.0H 000 00.N: 0.H0 N0 009 HHz 00.H: N0.0: NNvN HOON 0N.0 000 000 00.0 0.00 >.0H N00 00.H: 0.0N N0 009 HHZ H0.0: 0N.H 000N 0000 00.0 000H 000 00.0 0.00 0.HH H00 0H.H: 0.00 N0 00m HHS HN.o| 00.0 0vHN H000 00.0 0nHH 00N v0.0 «.00 N.NH HN 0N.H: 0.00 N0 90m HHZ 00.0 00.H 0000 0HON 0N.0 HONH 000 0.0N N.00 0.0N 000 N.HH «.00 N0 QHZ HHZ 00.0 00. H H0nN NOHN 00.0 000 00N v0.0 0.N0 0.0H >00 00.0 0.00 N0 DH: HHS 00.H N0. H 000N 000H H0.0 000 000 00.0 0.H0 0.0H 5N0 0.0 0.H0 N0 009 HHS H0.0 00.N 000N 000H 00.0 000 0N0 00.N 0.00 0.0H 00v 00.H 0.N0 NV 009 HHZ «mafia van Mauanon .uoauaaanunohum .ouwm 035A afiaaflz 00.N 00.H 00> 000H >0.o >0H v0 00.r 0.0N v.0 00N 00.N 0.0 .>mQ new 0N.0 00.H N000 NNNN 00.N 0o0 00N o0.r H.00 0.0N 000 0H.0: 0.0N z 0NN NN.0: 0.HH 00H 50m HHS HN.0: 0H.H: NO0N HO0H 00.H Hun 00N N0.0: 0. 00 0.0H H00 O0.0: 0.0H 00H Hem HHz 00.N: 0H.0 >000 0NOH 00.N 000 NNN N0.H: v. 00 0.0H n00 0H.0: 0.0H 00H Noe HHS 00.0 00.H 000N 00> 0H.N 0H0 0HN 00.H H. 00 0.0H 000 00.0: 0.0N 00H Noe HHZ mu Hz a ‘0 Do a: H4 mm mm m z: 00 an * «Hoboq uoam In:nauuunnnnauaaunnnunnuu Hammv uaaawHMwnuununuuuunnnununnunnuununn ooua nzono Hb.u:ouv .N mHQme 194 APPENDIX 0>.H 00.N 000N 0>0N HN.N 000H 000 H0.0 0.00 0.0N 00> 00.H: N.0N 00H 009 009 0N.0 N0.N o0>N 000H 0>.H N00 NON 0>.H 0.0> 0.NN 000 o>.ou 0.0N 0OH 009 009 madam can mauauam .uoauwonuuuoaum .wudm awauudnawaa 00.N 0>.H 000 >00 00.0 00H 00 00.N 0.0N 0.0 00H 00.0 0.0 .>mo 090 00. 00.0: HON0 Ho>H 0H.N H0> 0NN N0.0 H.00 0.0H N00 00.o 0.0H zmmz 00.N: 00.0 000 0>>H 00.N >>0 0NH 00.0 0.00 0.0H 000 0H.H: 0.0H 0>H Rom 009 00.N: H0.0 00N0 00HN 00.N 000 00H 0.0H H.0> 0.0H 00> H0.0: 0.0H 0>H 90m 009 00.N: 00.o 0No0 OOHH 00.N H00 0NH 00.0 N.N> >.> 000 HH.O| 0.0H 0>H 009 009 0>.o HN.o: N000 000H 00.N 00> 0HH 00.0 0.o> >.0 N00 0H.N N.0 0>H 009 009 0>.o 00.0 0000 000H NN.N H0> 00N 00.0 0.00 0.HH: >00 00.H: 0.0N 00H 00m 009 00.0 00.0 0>H0 NO0N 00.N >N> NHO >0.0 N.N0 0.0 000 N>.H 0.0N 00H Rom 009 00.0 00.o N>>0 00>H 00.N 000 000 00.0 0.>0 0.> 0>N 0H.N 0.0H 00H 009 009 0H.H 00.0 0000 000H 00.N >00 HON 0N.0 0.00 H.0H 000 o>.0 0.0H 00H 005 009 0H.0: 0N.0: 0o>0 0HOH 0H.N 000 00N 0N.0 H.>H 0.NN O0N H0.0: 0.NH 00 900 009 H>.Hu 0H.0: 00N0 N0>H 0N.H 00> 00N 00.H 0.00 0.0H NON NN.0| 0.> 00 00m 009 H0.0n 00.N: 00 O0NH NO.N >00 0HN 00.0 0.0N 0.NN HON 00.o 0.0H 00 009 009 00.0: 00.0: 0000 o>0H 00.H 0N0 00N HN.0 H.0 0.0N 00H H0.H: 00.0 00 009 009 H0.H >0.on 0>>N 0NOH H0.H o>0 00N 00.0 0.00 H.0N H>0 0H.N 0.0H 00H Rom 009 0N.0: 00.N: NNON >0HN 00.H 00> 00N No.0 0.00 H.>H >00 NN.N 0.0N 00H 900 009 00.0 00.0: >NNO 0NOH H0.N 00> 00N >0.H N.00 N.0H o>0 N0.N 0.0N 00H woe 00> N0.0: 0H.N: 000N >00H 00.H 000 NON 0H.0 0.N0 0.0N 000 00.0 0.HN 00H 009 0GB madam van mauaownao .uoHUomuIHQOMuN .ouwm uwuouduuwna 00.N 00.H 000 000 0H.H NOH 00 H.OH 0.0H 00.0 0H0 00.0 0.0H .>ma mam NH.H 00.H 000N NO0N 0H.0 o>0 0N0 0.> 0.o> 0.0H 0N0 00.0 0.00 z.N O0OM 0NON 0N.N 0N> M>N M>.N 0.00 N.NN 0N> 0H.O: M.0N N woe owe O0.OI NO.N 0MON HOON 0H.N 0>0 MNM 0N.O 0.0> 0.0N 0O> 00.H N.>M N woe owe HH.O 0O.H 0HOM 0NOM M0.0 >H0 0HN N0.0 0.00 M.ON MOHH HH.H O.HM N eom owe >0.0 0H.H NN>N NNOM 00.H HNO 0NN H>.M M.00 >.0H ONMH 0N.H 0.HM N eom owe 00.o >0.H HOHM HO0M 0O.N 000 0NM 00.N OHH H.HN O00H >0.H 0.00 N QHZ owe HH.O 00.H >0NM 00>H NH.N >>0 MON H0.M 0.H0 >.HN >H> 00.H 0.0N N QHZ owe 00.o: H0.0 N>0M NOON 0H.N OO0 O00 NH.M HOH M.HN >O> 0O.NI 0.0H N woe owe M0.OI OH.0 0H00 000N 00.N 0>> 0HM NN.0 >OH N.0N 0MOH 00.H 0.0N N woe owe 0>.0 0H.0 N>N0 OOO0 M0.N HOHH HMN 0.0H H.00 0.0N 00HH >0.0 M.0M 00H eom 0oe M0.0I 00.0: MMOM 0MOM O0.N 00HH NMN H.0H N.HH M.>N 00HH >H.OI >.0N 00H eom 0oe ON.0 M0.0 000M HO0N NN.M >>0 O0N H.NH 0.00 0.HN 000 ON.OI H.0N 00H QHZ 0oe 00.0 0N.0 000M H000 00.N 00HH 0N0 >.HH N.00 0.0H 0NOH 0>.0n 0.>0 00H QHZ 00B H0.0 N0.0 0HMO MMOH 00.M N0> 0N0 0.0H H.>0 N.0N 00> 00.0 H.0N 00H woe 00e 0>.0 >0.0 NN>M MO0N >0.M 0N0 0M0 0.0H O.H0 >.0N 000 00.N N.0N 00H woe 0we H>.HI O0.MI 000N N0>H 00.o 00> 00H N0.N 0.0N H.ON >00 0>.0I 0M.> >ON eom 0oe >M.0 0>.O 00>N NMON 0>.N 00> 00H H.HH 0.00 0.0H 000 HM.M N.0H >ON eom 0oe 00.0 0H.0 >0OM O0MH 00.N >>0 0NH N0.0 0.00 H.0H N00 00.o 00.0 >ON QHZ 0oe 00.H OO.MI 000N 00OH M>.H O00 00H OH.0 H.00 O.HN H00 H0.N: N0.0 >ON QHZ 0oe 00.H: NM.MI >0NN O00 >0.0 M00 >0H >N.H 0.H0 M.0H HOH 0.HH: 0N.O >ON woe 0we M0.0 0H.Mu 00>N N0> 00.H >00 H>H HN.O| 0.00 0.MN 0HN 0H.0I 00.0 >ON woe 0we O0.H 00.H NH>M 0>0M H0.N >00 >>M O0.H 0.00 0.0N MO0H 00.o H.00 00 eom moe OH.N H>.O HONM >000 00.N 0HOH 000 00.0 0.0> 0.0N 000H 00.0 H.00 00 eom mwe 00.0 HH.0 000M >00N 0O.N H>> H>N N0.N N.00 0.NN >>OH OO.M N.>N 00 QHZ moe >0.H MO.M HNOM OHOM 00.N 0H0 O00 N0.H >.H0 0.0N NONH HM.O 0.0N 00 QHZ moe N.00 O.M0 0000 >O0N 00.N >0> OHM >H.O NON 0.>H >00 00.o 0.>N 00 woe moe >O.HI 0.HH >000 0MOH OH.> OO> 0NN 0O.H: H.00 H.0H >00 N0.OI H.0N 00 woe moe 0H.O N0.H 0HOM >HOH H>.H 0>0 00N >N.N 0.00 M.HN 000 MM.Ou 0.0N 0OH eom 0oe NM.OI 0N.OI HHON 000H 0>.H ON0 NMN 0N.0 O.H0 O.MN 00> 0N.0 0.0H 0OH eom moe 00.0: 00.0: 0MON 00NH 00.H 0N> 00N 0H.N N.O0 H.ON 0O0 HH.N 0.0H 0OH QHZ moe 00.0 N0.0 O00N 000H O00.H MO0 00N M>.N 0.00 0.0H H00 00.H H.0H 0OH QHZ moe m0 H2 M do Do 02 Ad mm mm m 2! no ZN * «HO>OA 00HH ulnnauuuununnuuuulununuuu Hawwv uaqfiaamxInIn:unuulnnuuunlunnuuuunuu cane alone HU.HCOUV .N mHQme . ... __. Avpgd... .H—.u....:... .,.. .UmeEmm mm: czouo mmuu map 00 MHmcumco Eouuon Ucm wow map >Hco mmmmo meow cH .cZOHo mmuu mgu mo UgHauumco A9000 600000 go .AonO mHunfle .Amoev 005 u Hm>mH czou00 X I m 00.> 0.0H 0>0 0>0 00.H 00H 00 0N.0 0.00 N.0 000 0N.0 >.0H .>mQ new mw % 00.N H0.0 >0H0 000N 00.N 000 00N 00.0 >.N> >.HN 000 0N.0: 0.0N 20m: 41 P A NN.0 00.0 0>0N 000H N0.N H0> 00H N0.0 0.00 0.0N 0H0 00.N: >.0H NN eom owe 00.0 00.N 00H0 0HOH 0>.H H00 00H 00.0 0.00 0.0H 0>0 N0.0: 0.0H NN eom owe 00.> 00.0 000N 00HN 0>.N 00> >0N 0.0H 0.00 N.00 000 00.0 0.0N NN QHZ owe 0N.0 0>.0 NN>N 00HN 0>.N H00 >>N 0H.0 H.N0 0.0N H00 00.0: 0.>N NN OH: owe 0H.0 N0.H 00>N 00NH H0.N HHO 00H >0.0 0.>0 0.>H NHO 00.0 0.0H NN woe owe H0.0 00.0 000N HHOH 00.H 0H0 00N 0.0H 0HH N.0H 000 0H.0: 00.0 NN woe owe 00.0: 0N.0 000N 0000 0>.H H00 00N N0.N 0.00 0.NN 00HH 00.0 0.00 NH eom owe N0.0: 00.0 000N 0000 N>.H >00 00N >0.0 0.00 0.0N HONH H>.0 >.00 NH eom owe 00.0 00.H 000N NHON >>.H 00> 00N 0N.0 0.>0 0.0H HO0H 0N.N 0.00 NH QHZ owe 0H.0 0.00 00>N >0>N 00.0 N0> 00N 0.00 H.0H 000 0H.O H.H0 NH QHZ owe mo Hz N do Do 02 Ad mm mm m 22 no 2N * *Hobon uoam lulu:Inuunununuuuuuunnunu Hamwv uaoaaHmUnlunuuununnuuuanuuuuuuuuuuu ooua atouo HU.ucoov .N m HQme APPENDIX HO0H HONN H>.HH >HOH NO0H 000 0NH 0.00 HHZ >00H NHON 00.> 000H 000N 000 00 H.0> HHZ 000N >00H 0N.0 0>0 000H 00H >0H 0.00 MHZ 00.>H 0H.H 00HH 000N 0H.0H 0>> 00> HNH 00.0 0H 00.N: N.00 NHZ 0N.0 H0.H 00NH 0000 00.0N 00HH 000H 0HOH 00.0H 00H H0.0 0.00 0we 00.0 00.N NH>H 0000 H>.>H 000 000N HHO 00.0 00 H0.0: 0.0> MHZ >0.0NI 0H.H >0>H 0>0N >N.0 0>0 H>0 00N >>. NNH HH.0: >NH 0HZ 00.> 0H.0 000H 0NON 0N.0H 000 0>HH HON >0.> 0N 00.0 0.00 HHZ 0H.0: >0.0 0>0 000N 00.0H 000 >HOH 00N 0>.0 00 00.H: 0.00 0oe H0.0: NN.0 NO0N 000N 00.0H 000 00>H >NN >N.0 0> N0.0 0.00 0we 00.0: 00.H 0HOH H0>H H>.0 000 NO0H 00N H0.0 NOH N0.0 H.00 MHZ H0.0Nn 0>.H: 0>0H 0>0N NH.0 >0HH HOHH 00H >0.0 00 N0.0: 0.>N NHZ 0H.0 00.01 000H 00N0 00.0N 000 000N 00N 00.0 00 0>.N >.00 NHZ H>.00: 00.0: 0NOH 0HON 00.0H 00> 0>HH 00N 00.0 00 ON.0: 0.00 0oe >0.0N N0.0 HNO NO0H >0.0 00> HOHH 00H 0>.0H 00H H0.0 0.N> 0HZ 0H.0H 00.0 000 HHHN 00.0 000 000H 00H 00.0H 00H 00.0 0.00 0oe HH.0: 00.0: 000N 0000 0>.0N 0HOH 0HON 00N 0N.> 00 >0.H 0.N0 NHZ 00.HH: NN.0 000H NO0N 00.0N 000H 000H 0>0 0N.NH >0 N0.0 H.00 NHZ 00.0 N0.0 0HOH 000N H0.0H 0H0 0N0 >0H 00.NH 00H HN.0 0.00 0we Nr.0: 00.0: >>00 00>0 H0.00 00NH HO0N HON 00.HH 0>H 0O.N 0.0> 0we 00.0H 00.N 000 000N 00.0H 000 0HON 0ON >0.> 00H 00.H: >.00 0we 00.0 00.H >0H0 0000 0H.0 >>0 >0NN 0H> H0.> 00N 0H.0 0.0> 0oe >0.0HI 00.0 N>> 00HN 0>.> M00 00NN 000 00.0 00H 00.N! 0.00 00e 00.H 0>.H: >00 >0H0 00.0H 000H 000N 000 00.0 HMN 00.0: >.00 0oe .EE 0 v uuoom oadh .aowunooq 00cm ‘uuoam ouuwudn amuuwan2|aousm mo Hz M £0 DU 52 A4 mm m 2: no 2N uoam IIIIIIIIIInullIllulaluunlllullllllu Hewwv nuaaEGHm.IIIIIIIquuullunannullnnnluu .000H 0cm .mmmH .mmmH .pmsosm .000 .00 .Hov mpoHQ xomno vamum HmHSpma mme asamo 0cm «H00 .Nm .Hmv muon xumno coEEoomom «Home 0cm UHZO muon xomno mmuchmchousm «H0we cam 0wev chou¢chne 0am HMHZ cam .NHZ .HHZO wme chcHZ "muon mmmmmHU mmuchmZucousm um mvcmum coHumucme Eoum HumqumHU CH E8 0 on 00 muoou m0umH Ucm AmmumEmHU CH 8E M vv muoou mcHw mow mcoHumMHCmocoo ucmeHm Doom .0 mHQme 198 APPENDIX 00.0H >0.N 0>0 00H0 00.0 000H 0N0 >00 0H.0H NNH 00.0 00H 0H2 N0.N 00.0 HOHH HO0N 0>.0 000 000H 0N0 N.0 00N >>.H: 0.00 HHS 00.H: >0.0 00HN 000H >0.H 00NH 000 0> 00.0 000 0H.N: 0.00 00e 00.HN 00.0: 000H 00H0 N0.00 0H0 000H 0HN 00.> 0H 00.0: N.00 NHZ 00.Hn 00.0: NO0H 0H00 N0.>H H00 000H 00H 00.> 0N 00.0: H.>0 NHZ 00.0 H0.0: 000H >0>H 00.0N 00> 0>0H 00H 0H.NH 00N 00.0 H.00 009 HH.0: H>.0: HO0H 00NN 00.0 00> 0HON 000 00.0H NNH >N.0 >.00 HHS 00.0 0H.0: HHOH 0NON 00.0 NoHH 000H 0>N 00.N 000 00.0 0.00 00B 0N.0H: 00.0: 000H >00N 00.0 0HOH 000N 0>0 00.N 000 >>.0s H.00 0H2 0H.> Hr.H: 0HNH 00NN 00.0 000H >000 0N> 00.> >00 N>.0n 0.>0 009 HH.H 00HH 000N 00.HH 0HOH 000H 000 H0.HH 00N 0N.H 0.00 00e 00.0: 0NOH 0NNN 00.N 000H NOHH 00N 00.0 HHO 00.0: 0.>N 0H2 N.0 000H 000N N0.0 H>0 000 >0H >0.> HH >0.0: 0.0N NHZ 00.N: 0>NH 00NN >0.0 N00 0000 000 0N.N 0HH 0.HN 0we >00H NO0N 00.0H 000H 000N 0H0 >00 NNN 0H2 000H N000 N0.0 0NNH HNO 00N 00H 0.H0 NHZ N>0 000H 0>.0H 000 000H 00H NHO N.>0 0we >0.> 0H.0 00NH >00H 00.0N 000 000H 0HN >0.0H 0H0 00.H 0.00 0oe BE 0 v «000% «can .aoaunoon onwuuno .uuoam ounmuwn amuuwauz nousm 00.0H 00.H 0N0 000 00.> N>H >00 000 00.N H0 N0.0 0.0H .o.m 0H.H: >H.H >NOH NO0N H0.NH 000 000H 0N0 00.0 0HH 0>.0 H.00 2002 00> NONN 00.0 00> N>0N >0N 00N 0.N0 0H2 HON 0NOH 00.N 00> 00N0 00NH 00 >.00 HHZ 000H 0N00 0N.0H 0NOH 000H 00H 00H >.00 NHZ 0>0N 000 00.HH 000H 0HOH 000 00 0.00 NHZ 00NH 0>0H 0H.N 000 NNNN 000 00H 0.00 0we N00 000H >0.0 0H0 0000 00HH 0N 0.00 HHZ mo Hz 3 «0 Do 08 an mu m 2: no 2N uoam ununuuuuaInn:uuuuununnnnnunnnuunnnI Hewwv unawadeu nun:nunuuununuuununuuuuununn ”n.0couv .0 mHQme 199 APPENDIX 00.HH 00.H 000 050 00.5 000 005 000 05.0 050 50.0 0.00 .0.0 50.0- H5.o 0000 0000 00.00 5000 0050 000 00.0 H00 50.0- 0.00 2002 0000 00H0 00.H 05HH 0000 050 000 5.00 000 000H 0HON 00.0 0500 0000 000 000 0.00 000 05HH 5H00 00.H 0000 50H0 000 00H 0.00 H02 050H 0000 00.5 000 000 0HH 00 0.00 002 0000 0050 00.00 500 000 000 00 5.00 002 0000 0000 50.5 5NHH 500 000 00 0.00 002 000 0500 H5.HH 050 000 000 000 0.00 000 0050 0000 00.00 000 0050 000 000 0.00 000 000H H000 00.0 000H 0050 00m 000 0.00 000 00.0 00.0 0000 5000 00.0 000H H500 000 00.00 000 00.0 0.00 002 50.0 00.0 0000 0000 HH.00 000 0000 000 00.0 000 00.0. 5.00 000 00.00. 0H.0- 000 0000 00.00 000H 0000 000 50.0 000 05.H- 0.00 002 00.00. H0.H 0000 0000 00.5 055 0000 000 50.5 000 00.0- 0.00 H02 00.H 00.0 000H 0000 50.00 05H0 0000 500 50.5 000 50.0 0.00 000 50.0- 00.H 000 0000 00.5 005 0000 050 00.00 00 00.0 0.00 002 00.0 00.H 000 0050 00.0 0000 5000 050 00.0 H0 55.0 0.00 002 00.0- 00.0 H00 0000 00.0 000 0000 000 50.0 000 0.0- H.50 002 00.0. 05.0 00HH 0000 H0.0H 000H 0000 000 50.0 000 50.0- 0.00 002 00.0 00.H 0000 5500 H0.0 5H0H 0000 000 00.0 000 00.0. 0.00 002 00.0 00.0 0000 0500 00.00 000H 0000 500 00.0 000 00.0. 0.00 002 00.0 00.0 050 0000 00.0 000 0000 000 55.5 00 00.0. 0.00 002 00.0 00.0 000 0500 00.0 500 5000 000 50.0 000 50.0 0.00 002 00.H H0.0 0000 0000 00.00 000H 0000 000 50.0 000 00.0 0.00 002 00.00- 00.0 0000 0000 05.0 00HH 000H 0HN 00.00 000 05.H 0.00 000 m0 H! K (U DU 62 A4 Nb m 22 00 Zn uoam ----------------------------------- 05000 .uaaaaHm -----------u---------------- 00.00000 .0 00000 200 APPENDIX swam vam mm.m 0mma whaa mmm mow m.vm owe mmma mmbm mm.m mmm mmm mva aoa m.vm 0a: mm.m: mm.o: awmm Nmmm mm.m mmva vmm Nma aN.m 0mm vv.al m.mm owe bh.a| am.m nmam hmmm mo.ma moma vama mmm mv.m mnm mo.o N.0m DUB ho.m Na.a mmma momm mN.NN mmma mava mow vo.Na 0mm hm.a| a.mm OAS mv.a| mm.o amom omvm va.m mum mav Naa mm.aa nmm nm.a a.mv 042 3. : mm.o amma mmmm «om mmoa mmoa mow 3.0 maa mm.a mam oaz am.oa| mn.m wwma mmom mm.vm omva mhaa mma mo.m mnv wm.ml m.mm DUB mm.m| ma.o maom mamm hm.>m mvma mafia mam mv.m who mm.mt >.mm owe mm.>n ma.o: NNaa mmmm mm.m mmoa mmma vmw mm.b bvm vm.a| m.mv OAS mm.mu ba.o bmaa amnm mm.aa omOa oov mma hm.m «ma >>.m N.Oq Uaz N©.ma| mm.ol mmaN Oamv a>.oa mmma mhm Nva mv.v mvm mo.ml m.ma DUB mm.om mm.m mmom ovnm mw.ma mNNa ovm mma mm.mm 0mm bw.m v.mm oaz mw.mu NP. ammm mmom NN.M mvma Nvm ama vm.ma nmn mm.ml m.mm 009 hm.m1 mv.m mmmm mama mm.oa vhma Maaa mom om.oa amm hm.m N.0v owe Na.a| mm.a| mvma 00>N vm.m moaa Noma mmm aa.oa Nam >N.0 m.om 0a: ao.m vm.au mama Nham Na.mm mmma mmm mom mv.m amv mm.a a.mm DOB Na.ml no.a| mmaa ommm wo.mm nmm mmma how vw.aa mmm mm. N.mm oaz va.m ao.o mmwa mmhm mm.ma mmaa mmva vmm mm.ma mmn m>.N m.om 0a: mo.oa am.m wbma mmma mm.oa abOa omm oma mN.Na 0mm aN.N m.mm DUB am.a ana «mom mm.v mmoa mmn mmm mo.m maa >0.N >.mm 0a: ma.a mmom mmmm mm.m mmva mmma hmm mm.Na who mm.a m.nq uaz mm.m mmam omvm mm.aa vaa maaa mam mv.0a mmn mo.a m.mw oaz mm.o amom mmbm mv.aa mmma vmaa mmm mm.m man o.am DUB no.0: amam ammo wv.m Mava vaa mmm mv.oa mom mm.mu v.mm owe mm.ma mmma Nmam mv.ma amma mova vvm om.oa mum a>.01 m.ow 009 E: n v nuoom mafia Juoam 0.00:0 oauuwcgncouum MU H2 M (U DO 08 Ad mm m 22 00 Zn “GHQ nununununnunununnunulnunuunnnnuuunn .8003 nuaaaHm ..qululnuuunnuuuuunuuunuuunu AU.uCOUV .m manme 201 APPENDIX om.a aama mnom N0.0 mvm mmma «ma mm.oa mma mm.m m.mm mo mm.a maam vama m0.aa mmm mam mma mo.0 mom 5m.m >.mm mo mm.m omma vmma ha.m mam «mm amm mm.0a «mm mm.m a.mm a0 0:: m v uuoom onwh \uuoam xomnu oxan c5530 on.a mm.o mmm mum mm.> mna mow mma mm.a mma ma.m m.mm .Q.m mn.m mv.a «mma mmom ma.aa 5moa mmmm mav na.m amm mn.a v.0m z¢mz amma 00mm vo.m mnma omvm vmv Nmn 0.0m am 00mm mmmm mm.m mmma mmmm mom avm m.mq mm momm mqam mm.m mmma momm avm mow a.mm mm mmqa ommm am.oa omoa vmmm mav mav m.m0 am omom 00mm ha.m moma Mamm Nmm mma m.mq mm mmam mnmm mm.» 0000 ammv «mm omm m.mm am amma mqvm mm.m mmoa mnma mma 0mm m.mm am mmmm omow nm.>m maaa mvmm mmw oma baa mm mmmm roam «>.ma Ohm nvwa 0mm pm 0.00 mm mmma mmmm ah.m mom mmmm mmv mmm m.vv am mama moaw mm.ma 00m mmom mmm mvm v.om am mn.o mmma mmom mm.m mmva nmma mqv mm.m mmm ma.v n.mv mm mm.m vama ammm ma.ma amaa moma vmm 0m.aa «mm vm.v w.mm am mm.m no.0 mmma mmmm «0.> vqm «mom omm an.> wma mm.an 0.0m mm mm.v mm.a avma 00>N >m.aa mum oaam mmv vm.m mna 0m.a m.qm am nv.a mm.a moma aamm mv.m mom mvnm mmm wm.m mom om.o >.Nm mm EB m v uuoom «can .uuoam guano aoaaoouom an. mm.a >mv aqm Nm.m 0mm omv ooa mm.m vow mv.m N.m .Q.m ma.mn mm.o mmma ommm mm.ma mnma omoa nvm mm.oa mnv 00.o o.mm 24m: mo Hz K ‘0 Do as. Ad. uh m 22 no as uoam Inunnunlluunnnlnuunnunununlunuunnnn aaflmv nuaqaaau sinus:uuuunnununnnunlluuunun 0U.ucoov .m magma 202 APPENDIX om.v ma.om omma «mam ha.ma mmn mum oma >m.m 0m m0.o N.OM maz mp.0a am.ou mmm omma mm.m mmm amm Nma wv.aa mma m0.n m.mq ma: ma.mu mm.o- mmvm mmma ma.ma mvm mmva mom mm.m 0mm «m.m m.oq «we mm.m mm.mn mmvm mqmm mm.ma mph mow am ma.m mma mm.o a.mq «we mm.mu om.mu mmma mmna mm.o mam mom am mo.m mma ww.on m.am may mm.ma mm.ou omvm mvaa 0v.ma mmm mmm mOa mm.v mm mm.mn 5.Nm v09 ma.va mm.o: mmmm mmma mm.> mom 0mm mm «m.m oma sq.mn o.mv woe rm.o: am.m ammm coma hm.v «vb mmm mmm ao.oa 00a v0.0 m.mm «we mw.au mo.ma 0mm moma ow.m «mm oqm mvm mm.a aom qm.0: m.0a woe mo.oa 0vm mmma mw.a wmv mom voa 0m.m mm 00.0 m.vm woe om.a: anoa mmba am.on vmm mmm moa mm.m mma mm.a o.mm mop BE mum uuoom Gouda .nowuuooq 006m .uuoam oauouwa oouuwauxlcouau am.ma mm.a mmm mam mo.m boa 00m mma av.a mm mm.m m.n .Q.m am.an mm.o mmma mamm 00.> mmm wmoa 0mm mm.m mmm ov.o m.mm z¢mz moma 00am mv.m amm 00> om mma 0.Nq mo oama wmmm mm.ma amp mmm maa oom m.mw a0 ammm wvmm mo.ma mum mmvm mw5 Nov N.mv mo aomm mmwa mm.ma mom mmm ava 0mm m.mm mo vnma ammm «0.NN 000a maom omm aam m.mq ao ommm nmmm a5.v mboa mama mom oav v.mm mo mmma moom mm.0 mQOa Nooa vm 0mm m.mm a0 mn.m mn.o omma omom mw.a Maoa Nva mv mm.> qvm 00.o: >.mm a0 mm.oau mm.ou mmna mwom om.a waaa moaa mma up. mmv mv.mu q.vm ao mo.ma mm.m mmva mmma mm.m mmaa oom vm ma.0a mmm mm.v m.vm ao Na.m ma.ou mmna mmma mm.a mum aam vma mm.m qmv am.mn «.mm mo om.nn am.o «mma mmvm mv.a mmn mava >>N mm.m qua oa.on m.nm mo ao.mmn «m.aa momm amvm mm.a 0mm omv mm mm.m 0mm mm.o >.aq mo mu Hz & do DU 68 A4 an m 22 no as uoam Innuuuuuuuuuuuunuuuuuuuuuuuuuuuuuuu AEEQV nuaaEUHu IlulunnluluuunInuuuunnuuuuuu “U.ucouv .m magma 203 APPENDIX mo.aa hm.a whma mmma mw.m mmVa omm mmm no.a haw m>.ml v.am ma: mm.>a| mm.o maoa aama ho.m mam va baa ao.m mma mm.a| a.mm woe hm.>al mm.al hama moma ha.va om» mom oma ma.m ha mo.o: v.vm Na: am.o| amwa mmha mm.a wMaa mam mma mm.m wmm mo.m m.am woe ma.a amma mmoa mm.v Pom Nmm ama ma.m om am.m m.om ma: ov.o amva amma vv.v whoa mmm mmm No.0 Nvm No.0: m.om ma: om.mma am.a mvaa ahma vm.mm mmm hmm omm an.b am o>.ml m.om Na: 85 mum uuoom wound ~aowunoon ovwnupo .uuoam ooaauwo oauowad2ucouam mm.Ma No.ma oam ohm om.v mma vom oma on. «m vo.v m.ma .O.m mm.aa na.m mvma Nmma aa.w mph mom vva ow.h nma mm.on m.mm Z «Na Nv.o m.>m maz vm.m mm.a mbma whoa mv.oa amm mow mmv vv.m ava mm.ol v.vm «we oo.n am.a mama amom ov.m amoa our man vm.m qaa ov.o| a.am maz mm.v mm.m mmva amom ao.m mom ova mm «v.m ava mm.ol m.om maz hm.oa| Vb. mooa amma mv.m mmw mmm mm mm.v mm nm.o| a.mm awe mm.m om.a ooma mmva vm.v mmm amm moa mm.m om no.ml m.mm maz wo.m mo.at own papa mo.m mmOa mom on mm.v: «mm am.Mat a.v ma: Ma.ma om.ou mbma mama bo.m our ama mm hm.m moa «>.o >.om Na: No.m vb.mt omma amaa mm.a amw aam ow om.o am mm.NaI N.am aaE vv.mml mo.aa ammm oomm mm.m mmm mmm om ao.m ava va.o m.mm maz ma.mal vm.mu Nmom mmMa mw.a NoOa «Ma mm oa.oa ooa aa.mt m.mm aaz mm.qmu mp. nova mama ma.> hmm mow mva mo.aa vaa om.on m.vv mwe mm.mmn mm.m| mmmm aama am.h mom mam mma aa.oa ama >N.o1 N.av maz aa.N| nn.an oom moam mm.v mmm maaa mma mm.m mmm oo.on m.mq ma: mv.ma mm.m amn vmma mv.m mom Pom mm ao.aa «ma mh.o v.mq woe mo.aa mm.o nama mmvm ao.m vvn ohm mm mw.va mom hv.m a.mm woe vo.m mm.a> mvva ommm mv.ma hmm nmoa aMa ao.m mm mm.o m.mm maz an H! K ‘0 D0 02 Ad an m 22 no 2N HOHH uuuununuuunuuuunuuunnnnuuunuunnuunn aammv nuaqavmm lulu:nunnauuununnnnnuunnuunn ”C.DCOUV .m magma 204 APPENDIX mvaa amna mm.a omm mmm moa amm n.mm moe hm.aaI wn.a mmma mmmm mm.oa o>ma amm vma mm.m mom mo.o o.nm ma: mm.a ma.o mmoa mmna ma.m ovm mmm om om.> mmm ma.a N.mm mwe vo.m mv.a mama nmoa wm.mm mmm omv mm mm.m «mm mv.aI m.mv vwe mm.>aI va.aI omoa «mam a>.0a voma ohm ova hm.m oma mm.VI m.mm ma: mp.omI mm.o amaa nmvm am.m vaaa mmm ova no.m mam om.aI N.mv aaz mm.m am.a own mmam mm.ma omm mom omm mm.v baa mo.o N.mm aaz ha.mI mm.o mam mmmm om.¢ mmm amm moa mo.m maa mm. I m.vw aaz Nh.o ma.m wvm aomm mm.v omn Mmm aom mm.m «ma ma.a v.m aaz «m. I mm.oI «maa ammm m>.> mom avm vma mo.o ooa mm. I m.ma ma: mo.a aaaa ovam ao.m oama amv mMa «m.m omm av.0I m.mm ma: ow.m 0mm 0mmm mm.m mmma maaa vq vm.m mmv vm.oI m.>m ma: mm.o amm 0amm >N.m mom mom mma vm.m mma mm.aI m.am aaz vo.oI ao.a amma mvma no.m who 5am vm mo.o mow mo.a a.vm mwe ha.o om.a omm mqam NN.m omaa mom amp >N.h Mam mm.m m.mm ma: mm.maI mp.OI omma mmmm mm.> mmm mma on >m.v mm vm.vI v.am maz mm.q om.m woma nama om.m mam mmm oma ma.m omm ov.OI m.mm mwe va.mmI mm.aI maam mmmm ma.v wwm nq hm nm.oa vb vm.oI a.mm was Na.om m>.m avm amvm va.m moma mmv mom va.ma mmv mo.w a.mm ma: mm.>a mo.m 0mm mwmm mm.m mom omm mmm mm.oa an oa.v m.mm maz mo.m mm.o mmwm Nmm mm.m amm mam mv vm.v mmm mo.mI >.bm v09 «m.o mv.ma mmm mnma am.a orb aqv rma Nm.> mmm am.aI N.Nm aaz rm.m mh.NI omma voma mo.aI 0mm ava mm ma.m wvm ma.mI o.mw may mv.v mo.o mmma woma ma.Oa mmm omm am mm.m mma mm.m v.vm v09 mm.ma mo.oI coma mmmm mv.am ohm own mmm ao.m oa mm.aI a.NN ma: mm.m am.a mmma moma mm.m own mmm ooa ao.ma mmm a>.m >.mw wee mo.a qm.o mnoa omma ma.m vmm «0v vna ma.aa mma mm.a N.mq aaz mm.mI mm.a mmma Nmma ho.b amm arm am ao.m mmv mm.m >.om mwe mu H! M do Do 32 Ad mm m 22 no 2N uoam IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII gammy nuGQECHm IIIIIIIIIIIIIIIIIIIIIIIIIIII AU.ucoov .m man5 205 APPENDIX mn.mau mm.o ommm «mmm oa.> mmma mav am mm.m won am.oI a.mm owe aw.mI mo.o nama mmva >0.m mmma aov no one mm.o v.00 owe a0.oI mm.m mama 00am no.0m maaa amm nma mv.m mum >q.o m.aq uaz ma.m mm.m ovaa mmma ow.m mmm mam mma mo.o 00m mo.oI o.am oaz mo.a mm.o 50m oamm mm.m mom aam maa >m.m 0ma om.o 0.0m 002 0>.aI om.o 5mma mmmm ma.m omm mmm oma am.> ova va.a m.mw 002 mm.ou oa.a 0mm mnmv mm.» mmm mmv «ma mo.o an Na.a o.ma 002 mm.aaI mm.o omma omma wm.0a m5m mwm «ma oa.v mmw mm.mI v.0m owa vm.oaI mo.aI nmna momm ma.ma amaa mmm ama wo.m mmm mm.mI >.am owe mm.ma 0a.ou omoa Nmmm a>.m 0mm amm mom 0m.m amm «N.NI N.Nm oaz oo.mI mo.o «moa mamm vm.v mmaa mow mma 0m.m ovv mo.au m.mm oaz 0a.amI mn.oI mmma vamm mm.m 0mma mvm «m m0.m anv mm.mI v.mm owe av.ma mm.w oaam nmma mo.ma mum vmv mma mm.aa mmv mm.m m.aq uaz mh.aaI >0.N «mmm wwma mm.m mnm amm >0 0m.m amv vm.mI m.am owe ow.vaI NP. mmmm mmaa ma.0 mmm mnv mm mv.oa mmn mo.on m.vm owe mo.mn mm.o mmma movm ao.m mooa omq ova aa.oa ona om.a N.0m oaz mv.mm mm.m vwma mmna mm.ma wmaa mom maa am.» mom mo.o v.mm owe va.a m0.aI acaa nmom ma.aa mam 0mm «mm mv.m 5am «N.0 N.NN 002 ao.m «m.OI mama mmma ma.m pom «m0 «ma mo.ma mmw ao.w o.m¢ oaz am.a mm.a mmma nqma mv.m 000a mom «a mm.ma moo mv.m o.mm woe aa.m woma mmow mo.m mmma mmma arm mm.m Non vm.m m.mm was mm.o mmma mmom mm.v 0mm mvm ooa 0.0a mow nm.m 0.00 was 00.0I 0mma mmmm aw.> mmma 0mm mwa >m.m mqm mm.vI m.om owe mum nuoom Gonna .nuoam guano oouadauxIcouan 0a.mm mm.m mom mom mm.m amm mvm ama mm.m mma om.m 0.0a .o.m om.m mm.a aoma mmom Nv.m amm omm mna om.> 0mm om.OI m.0m z¢mz mo Hz K do Do as an an m 22 no 2N voam IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII Aeamv nuaoadflm IIIIIIIIIIIIIIIIIIIIIIIIIIII “n.0coov .m magma 206 APPENDIX mm.a omwa maam v.m amo arm mm mo.o 50m am.a 0.0a ao am.a mmma vamm mv.m moo mav am mo.o mm mm.m 0.0m mo mo.o mmma omma mm.o mow «cm om mm.m mam am.a m.oa ao .EE mIm uuoom manna ‘auoam 300:0 «Jug Guano mm.a mo.o ohm mav 00.N mna mmm om mm.a qma ao.m w.m .Q.m vo.v 0m.a omma vmom mo.o mom 0>> moa mw.m 0m nn.a m.mm zmmz anva mamm ao.m wmaa woo voa woo m.ov am mmmm mama mm.m mama mam mna own m.m¢ mm «mam owma mo.m on» mmna mmm mom a.wm mm mmma omna om.oa now vmm vo mow 0.N0 am amma mwmm va.v mmoa mom moa mvm m.mm mm mmma aama mo.o mm» «mm mm «mm o.am am 00am mmva N>.m mmn «mm mm qvo m.om am omma 00mm am.a mnm mmn aha 0am m.mv mm mmma omwa mm.m awm qmm mvm 0mm v.ma mm mnma Nwoa mm.m man moma mma 0mm m.0m am mmma mmnm m0.m mmo mom moa mmv m.mm am >a.m mvna 500m mm.a mmoa 00m mma mm.0 onv vm.m >.om mm mo.a mvma ammm ao.m awn aam mom mo.o «mm mo.o m.>v am mn.m mo.m mama Nmmm mo.o omm mom mmm o.m mma vo.aI m.mm mm ow.o mn.m mmaa womm oo.m amm «om mma «0.0a 00a mo.aI o.aa am mm.m mm.o mmna vmwa m>.m mmm man ona no.m 0mm aa.a m.mm mm .85 mIm uuoom Gouda ~nuoam Moono aoafibouom o.aa mm.m mom mmn mo.o mma vow mm mv.m 0ma a>.m 0.0 .o.m mm.mI vm.a «mma NN>N ma.m mvoa now ava mv.m mow av.o >.mm z¢mz mu Hz M do Do 02 4‘ an m 22 no as uoam IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII asamv uuaaaaam IIIIIIIIIIIIIIIIIIIIIIIIIIII 0U.ucooo .m magma 207 APPENDIX mm.ma mm.a omm vmm ao.m on mom vv No.m boa mo.o m.m .Q.m mm.mI mo.ol mnwa onmm vm.m mow «om om mm.> mom mm.OI m.mm zmmz mo.o mo.a mama mavm mo.o aam mam oaa w.m wom mo.v m.mm mo mm.oa mo.o mama mama mv.a mmm nma hm oa.> ova Nvm.aI m.mm a0 wm.mmI mo. I qvma mamm ma.a mom who mm mo.o omm mm.mI m.0m mo mo.aaI mm.o woma awom ma.m hon mmm mm wm.v mmm mm.MI m.ma ao ma.mI hm.o mooa mmam hv.m mmm vow ah ma.v boa No.OI h.ha a0 ma.m mm.mI mmm omma mm.OI amp mmv on mm.mI qmv wm.maI a.ma mo om.m am.a: omma mmma mm.OI mow mom mm na.> mom wm.>I m.mm mo mm.o mo.aI mmma moam mo.o amm vmm on om.m oma mv.aI m.om mo ow.mmI «v. I moom Nmov bv.o aom mom ow aa.m mmm ov.a m.om mo ov.val om.NI mhom vmmm mo.m aam mow oaa vm.m mom am.a N.mm mo oa.mI mm.oI mmma mmmm mm.a our mmm vma mo.o mmv mo.a q.mm mo mv.mmI mm.NI whom mmov mo.oI mam hmm on mo.oa mmm vo.o b.mv NO mo.o mo.oI Nova avma ma.v nmm omma vwa mm.m «av ma.MI N.mm mo on.maI hv.NI amma mmmm aa.N mmh own mma mm.m mwa mm.m m.>m mo vo.o: mmaa onmm mv.m Mam mmoa ova m.m mma ao.m m.va NO Nm. I nhma amom mm.v mow mow om hm.oa mom mv.a N.>m ao ov.OI «ama mmha mo.o ooh own Nma mo.o vmm mo.o m.>a a0 «0 H! M ‘0 DU 92 Ad ”h m 22 no ZN HOHm IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII Afimmv unamaafim IIIIIIIIIIIIIIIIIIIIIIIIIIII AU.uGOUV .m maQMB 208 .l J.- "‘IIIIIIIIIIIIIIIII