I... 1.. MICHI HINGA STAT as :1 ‘ 7 HQ” muIII/mumIII///IIII/Iill/IIIII/‘II/‘Ii/7/I 57566 LIBRARY Michigan State University This is to certify that the dissertation entitled Genetic Variation in Two Contrasting Habitats of Eastern Cottonwood: Responses to Different Water Status and Nitrogen Levels presented by Eko Bhakti Hardiyanto has been accepted towards fulfillment of the requirements for ' Ph.D degree in Forest Genetics Jam/(”M Major profe Date 7/”?(9/y7 / / MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 I PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE .inN 1 1992 v.3 ‘J n v “N 0 5 199? WWW MSU Is An Affirmative Action/Equal Opportunity Institution GENETIC VARIATION IN TWO CONTRASTING HABITATS OF EASTERN COTTONWOOD: RESPONSES TO DIFFERENT WATER STATUS AND NITROGEN LEVELS BY Eko Bhakti Hardiyanto A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1989 553%6 I no ABSTRACT GENETIC VARIATION IN TWO CONTRASTING HABITATS OF EASTERN COTTONWOOD: RESPONSES TO DIFFERENT WATER STATUS AND NITROGEN LEVELS BY Eko Bhakti Hardiyanto Selection for clones of eastern cottonwood (Populus deltoides Bartr.) that grow well on marginal lands is very important for plantation establishment. The objective of the study was to assess the genetic variation of trees from two contrasting habitats of eastern cottonwood in their response to water stress and nitrogen fertilization. Four clones from a sand dune (dry site) and another four clones from a floodplain (wet site) were used to establish field plot and greenhouse studies. Both studies used a split-split plot arranged in a randomized complete block design. Three different water regimes were applied: severe, moderate, and no-water stress. More severe water stress treatments were applied in the greenhouse study. Nitrogen levels used in the field plot study were 0, 200, and 400 kg N/ha/yr, while those in the greenhouse were 0, 2.25, and 4.50 gr N per plant. The following measurements were made for the field plot study: height, diameter, leaf area, specific leaf weight, leaf water potential, stomatal conductance, transpiration, and photosynthesis. Similar measurements were taken for the greenhouse study with the following additional measurements: shoot biomass, root biomass, and root-shoot ratio. Water stress reduced all characteristics, except photosynthesis in the field plot study. Nitrogen fertilization had no significant effects on any characteristics. The effects of water status were independent of nitrogen level in all characteristics. Clones between sites (habitats) differed. significantly' only for height and diameter. No interactions involving clones with other treatments were detectable. Clones from the wet site grew faster than those from the dry site. For height and diameter the contribution of clones between sites to the total variation was large. In the greenhouse study, the effects of water status usually depended upon nitrogen level. The effect of nitrogen was more profound under well-watered conditions. Under (well-watered conditions, nitrogen fertilization resulted in increased height, diameter, shoot biomass, leaf area, specific leaf weight, stomatal conductance, and transpiration, but. decreased. root. biomass and. root-shoot ratio. Leaf water potential and photosynthesis were not affected significantly by nitrogen fertilization. For height, diameter, and shoot biomass, clones between sites differed significantly under well-water conditions. For other characteristics, the site of origin of clones was not significantly different. Except for height, diameter, and shoot biomass, interactions involving clones between sites and water status were of little importance. None of the interaction involving clones between sites and nitrogen level was significant. The contribution of clones between sites to the total variation was small. However, clones from the wet site were more plastic to changes in water status. Results of these studies suggested that clones from the dry site did not grow better than those from the wet site under water stress. Selection for clones as source material for plantation establishment or breeding programs for marginal lands needs to be undertaken cautiously. ACKNOWLEDGEMENTS I wish to extend my sincere gratitude to my major professor, Dr. Daniel E. Keathley, whose guidance and assistance were essential to the successful of my graduate program. I would like to thank to the other members of my guidance committee: Drs. Donald I. Dickmann, James F. Hancock, and Kurt S. Pregitzer for their essential contributions. I would also like to thank to Luis Sadina for his help in designing the irrigation system, and to Dr. Phu v. Nguyen for his assistance in using ADC infrared gas analyzer. I would like further to express my appreciation to : Mike Stine, Randy Klevickas, Roy Prentis, R. Wasito, M. Charomaini, Omar Essady, David Freville, and others who are not mentioned here for their valuable aid in the completion of this study. I owe a special debt of appreciation to Dr. Soekotjo, Rector of University of Bengkulu who gave me an opportunity to pursue a graduate program. I wish also to extend my appreciation to the Government of Indonesia and Western Universities Agricultural Education Project for providing the leave of absence and the financial support throughout my graduate study. Finally, I wish to express my gratitude to my parents for their support and encouragement throughout my academic career . ii To my parents and grandmother and to the memory of my mother (1933-1963) iii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION LITERATURE REVIEW Genetic Variation in Eastern Cottonwood Genotype-Environment (GE) Interactions in Eastern Cottonwood FIELD PLOT STUDY Experiment One MATERIALS AND METHODS Establishment Data Collection and Analyses RESULTS Soil Analysis and Rainfall Height Diameter Variance Component Estimation Experiment Two MATERIALS AND METHODS Establishment Data Collection and Analyses RESULTS Rainfall and Temperature Height iv Page vii ix 12 15 15 15 21 24 26 26 26 26 28 28 28 Diameter Leaf area Specific Leaf Weight Leaf Water Potential Stomatal Conductance Transpiration Photosynthesis Correlations Between Characteristics Variance Component Estimation GREENHOUSE STUDY MATERIALS AND METHODS Establishment Data Collection and Analyses RESULTS Height Diameter Leaf Area Specific Leaf Weight Shoot Biomass Root Biomass Root-Shoot Ratio Leaf Water Potential Stomatal Conductance Transpiration Photosynthesis Correlations Between Characteristics Variance Component Estimation V 32 35 35 40 40 44 44 44 50 53 53 53 54 58 58 63 67 71 74 78 82 85 88 88 92 92 97 Stability Assessment DISCUSSION LIST OF REFERENCES vi 97 101 121 LIST OF TABLES Table 1. Expected mean square estimation used to calculate variance component 2. Physical and chemical properties of soil in the filed plot study 3. Rainfall and temperatures in experiment one of the field plot study 4. Analysis of covariance for height in experiment one of the field plot study 5. Analysis of covariance for diameter in experiment one of the field plot study 6. Variance component estimates for height and diameter in experiment one of the field plot study 7. Rainfall and temperatures recorded in experiment two of the field plot study 8. Analysis of variance for height in experiment two of the field plot study 9. Analysis of variance for diameter in experiment two of the field plot study 10. Analysis of variance for leaf area in experiment two of the field plot study 11. Analysis of variance for SLW in experiment one the field plot study 12. Analysis of variance for leaf water potential in experiment two of the field plot study 13. Analysis of variance for stomatal conductance in experiment two of the field plot study 14. Analysis of variance for transpiration rate in experiment two of the field plot study 15. Analysis of variance for photosynthetic rate in experiment two of the field plot study vii Page 14 16 18 19 22 25 29 30 33 36 38 41 43 46 48 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. “Correlations between the physiological characteristics measured in experiment two of the field plot study Variance component estimation in experiment two of the field plot study Physical and chemical properties of soil media in the greenhouse study Average temperatures in the greenhouse study Analysis of variance for height in the greenhouse study Analysis of variance for diameter in the _greenhouse study Analysis of variance for leaf area in the greenhouse study Analysis of variance for SLW in the greenhouse study Analysis of variance for shoot biomass in the greenhouse study Analysis of variance for root biomass in the greenhouse study ' Analysis of variance for root-shoot ratio in the greenhouse study Analysis of variance for leaf water potential in the greenhouse study Analysis of variance for stomatal conductance in the greenhouse study ' Analysis of variance for transpiration rate in the greenhouse study Analysis of variance for photosynthetic rate in the greenhouse study Correlations between physiological characteristics in the greenhouse study Variance component estimation for characteristics measured in the greenhouse study Stability parameters of two populations of eastern cottonwood across three soil water status viii 49 51 59 60 61 65 69 72 75 79 83 86 89 91 94 96 98 99 LIST OF FIGURES Figure 1. 2. The location of clone collection (circles) Moisture retention curves for soil in the The effect of water status on height after one growing season in experiment one of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; NWS= moderate water stress; NWS= no water stress. The effect of water status on diameter after one growing season in experiment one of the field plot study. Any means with the same letter are not significantly different by the DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on height after one growing season in experiment two of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5%. SWS= severe water stress; NWS= moderate water stress; NWS= no water stress. The effect of water status on diameter after one growing season in experiment two of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress NWS= no water stress. ‘0 The effect of water status on leaf area at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; NWS= moderate water stress;NWS= no water stress. The effect of water status on SLW at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. ‘ ix Page 10 17 20 23 31 34 37 39 10. 11. 12. 13. 14. 15. 16. The effect of water status on leaf water potential at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on stomatal conductance at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on transpiration rate at the end of July, 1988 in experiment two of the field plot study. Any means with same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. Soil moisture retention curve for soil media of the greenhouse study. The effect of water status on height after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of site origin of clones on height after 2.5 months, as affected by water status in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on diameter after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe-water stress; NWS= moderate water stress; NWS= no water stress. The effect of site origin of clones on diameter after 2.5 months, as affected by water status in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 42 45 47 6O 62 66 68 17. 18. 19. 20. 21. 22. 23. 24. The effect of water status on leaf area after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= no water stress; NWS= no water stress. The effect of water status on SLW after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on shoot biomass after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of site origin of clones on shoot biomass after 2.5 months, as affected by water status in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on root biomass after 2.5 months in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of nitrogen levels on root biomass after 2.5 months in the greenhouse study. The effect of water status on root-shoot ratio after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on leaf water potential after 2.5 months in the greenhouse study. 70 73 76 77 80 81 84 87 Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. xi 25. 26. 27. The effect of water status on stomatal 90 conductance after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on transpiration rate 93 after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. The effect of water status on photosynthetic 95 rate after 2.5 months in the greenhouse study. Any means with the same letter are not significantly different by DMRT at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. xii INTRODUCTION Eastern cottonwood (Populus deltoides Bartr.) is one of the largest and most widespread tree species in the eastern United States. Due to its rapid growth rate and ease of vegetative propagation, this species has become commercially important, particularly in the southern United States along the Mississippi River Valley. Eastern cottonwood provides an excellent material for pulp, sawlogs and veneer (Drew and Bazzaz, 1978). Eastern cottonwood is also an excellent species for genetic improvement work since intra- and interspecific hybridization can be done relatively easily (Dickmann and Stuart, 1983). In. the North. Central Region. of the ‘United States, eastern cottonwood and hybrid poplars perform well under conventional and short rotation intensive culture (SRIC) . With improved varieties and cultural methods, eastern cottonwood could be planted more extensively in Michigan and elsewhere in this region (Kelly gt al., 1978). Although eastern cottonwood can grow in a wide variety of site conditions, it achieves optimal growth. on rich bottomland soil. In the future, however, it seems that sites designated for forest plantations will be on more marginal sites due to the use of forest land for other purposes. Furthermore, the forested acreage is shrinking due to conversion to agriculture. Consequently, eastern 2 cottonwood will be grown on sites that are less than optimal for this species. On such sites the availability of water and nutrients may be critical for plantation establishment. Selection for genotypes of eastern cottonwood that perform well under water and nutrient- poor conditions is a logical measure. Effective selection for plant characteristics can be achieved only if genetic variation exists. With these factors in mind, the objectives of this study were to: 1. determine if there is genetic variation between cottonwood populations found in two contrasting habitats, namely sand dune and floodplain populations for growth and physiological characteristics; 2. estimate the magnitude of genotype-environment interaction and genotype stability of these populations; 3. observe the effect of water deficit and nitrogen fertilization upon growth and physiological characteristics of these two contrasting populations of eastern cottonwood. LITERATURE REVIEW Genetic Variation in Eastern Cottonwood J . Eastern. cottonwood occurs .naturally in. most of the eastern half of the United States and southern Canada. It is usually associated with bottomland, alluvial, and riparian areas. Fertile, well-drained, fine sandy-loams are the most satisfactory site, but eastern cottonwood will grow almost anywhere and is relatively resistant to drought (Dickmann and Stuart, 1983)°v Since eastern cottonwood. grows in. a ‘wide, range of environments, the existence of natural genetic variation is expected. Provenance studies are the first step for assessing genetic variation in tree breeding programs. Mohn and Pauley (1969) vreported from a provenance study in Minnesota that high growth rate of eastern cottonwood was associated with southern latitude, but seedlings from the more southern latitude had poor survival in Minnesota due to winter injury. A somewhat similar result was reported by Ying and Bagley (1976)ifrom a provenance study in Nebraska representing a major part of the natural range of eastern cottonwood. variation of growth, morphological, and phenological characteristics followed a clinal pattern from north and west to south and east. Genetic variation at stand and family levels has also been reported in eastern cottonwood. Farmer and Wilcox 4 (1966) progeny tested open-pollinated cottonwood families and measured variation in a number of traits at the end of the second growing season. Variation among families for height, diameter, specific gravity, and fiber length was significant. Farmer (1970 a)Vconducted a similar study and found that variation among families was also significant for various characteristics such as height, diameter, specific gravity, and incidence of leaf rust. Nelson and Tauer (1987)v reported from an open- pollinated progeny test that differences among stands were significant for height, diameter, incidence of leaf rust, and number of branches at two years of age. Differences among family within stands were also significant for those characteristics. Several studies have been carried out to document clonal variation in eastern cottonwood. Farmer and Wilcox (1968)flvfor example, reported that clonal variation in the test population was great for several characteristics such as height, diameter, volume, specific gravity, fiber length, and leaf rust. This indicates that improvement through clonal selection is possible. Randall (1973) \Jcollected clones from different populations in western Kentucky, western-central Mississippi, north-eastern Texas, southern Arkansas, and east-central Illinois. The clones were then planted in Illinois. Data were collected when the trees were one, two, three, and five years of age. The clones from the 5 southern populations had larger diameter and faster height growth rate than those from local populations at all ages. There were also differences among clones within populations for the traits measured. Randall and Cooper (1973)Lreported in a test study that clones differed significantly for height, diameter, and specific gravity at five years of age. Genetic variation in physiological characteristics has also been studied in eastern. cottonwood. For example, Kelliher and Tauer (1980) Vcompared stomatal resistance between clones collected from a dry site and from a wet site in. northern Oklahoma after subjecting them. to different water regimes. The result indicated that the dry-site plants had lower stomatal resistance values than the wet- site plants, even under well-watered conditions. Genotype-Environment (GE) Interactions in Eastern Cottonwood A number of studies have indicated that eastern cottonwood genotypes vary in their response to environmental differences. Randall and Mohn (1969)" found substantial clone-site interactions for height and diameter at ages one to four among 79 clones grown on two sites. Mohn and Randall (1973)’obtained a similar result from a different study for height, diameter, and number of first year branches at three years old. Clone-site interactions seemed to be more important than clone-planting year interactions. 6 Bridgewater (1972) selected clones from natural stands and tested them at two locations in Oklahoma. Significant clone-site interactions for growth rate and yield were observed. Only three clones performed well at both locations. Randall (1973) gollected clones from several localities (Kentucky, Mississippi, Texas, Arkansas and Illinois) and planted them in Illinois. He found that there were significant interactions for clone-site, clone within population-site, and population-site. Randall and Cooper (1973)jreported that GE component of variance was large. Among 32 cottonwood clones tested at three locations, the GE component of variance for height growth was as large as the genotype component at one year of age. At later ages, it was approximately half as large as the genotype component. In a recent study, Nelson and Tauer (1987)\) progeny tested 159 open-pollinated families representing 40 natural stands. At two years of age, significant stand-location interactions were detected for height, leaf rust, and number of’ branches, while family-location. interactions ‘were significant for date of leaf fall only. Differential responses of cottonwood genotypes upon application of different water regimes and fertilizer levels I have also been reported. Curlin (1967) found strong clone- fertilizer interactions for height, diameter, and volume, but not for specific gravity at two years old in the field. 7 Broadfoot and Farmer (1969);applied two different water regimes on 30 clones of cottonwood. Significant clone- moisture interactions were not detected. On the other hand, Farmer (1970 bf” reported in a similar study that clone- moisture interactions were significant for growth, shoot- root ratio, and wood properties, although the interaction component of variance for these characteristics was relatively small. In tree breeding programs, genotypes that have high productivity and perform well across different environmental conditions are desirable (Shelbourne, 1972;l Zobel and Talbert, 1984). Genotypes that show little GE interaction have high stability or low plasticity. There are many ways to assess GE interaction and genotype stability. The earliest approach involves analysis of variance (Spraque and Federer, 1951; Plasteid and Peterson, 1959; Comstock and Moll, 1963). Another approach for analyzing GE interaction is joint linear regression. This method was popularized by Finlay and Wilkinson (1963) and used among others by Freeman and Perkins (1971), Eberhart and Russell (1966) and Tai (1971). Multivariate methods have also been employed to analyze GE interaction and genotype stability in recent years. These methods include, cluster’ analysis and. principal component analysis. Cluster analysis was used, among others, by Abou- El-Fittouh gt gt. (1969), Mungomery gt gt. (1974), Lin and Thompson (1975), Lin (1982) and Gadhery gt gt. (1982), while 8 principal component analysis was used, among others, by Kempton (1984), Wescott (1987), Crossa (1988) and Crossa gt a1. (1988). There is no best method for analyzing GE interaction. Every method has its advantages and disadvantages. This matter has been discussed quite extensively in ”_many scientific journals (Freeman and Perkins, 1971; Wescott, M 1986, 1987). FIELD PLOT STUDY Experiment One MATERIALS AND METHODS Establishment Eight clones were used in this study. Four dry-site clones were collected from sand dunes at the Saugatuck State Park, while another four clones, representing a wet site, were collected from the floodplain of Kalamazoo River about 10 to 15 kilometers from the sand dunes. These two habitats are located at the township of Iaketown, Allegan County, Michigan (Figure 2). These clones were collected in March, 1987. ‘ The plants were grown from hardwood cuttings, 1 - 2 cm in diameter, 20 cm in length and having at least two buds. Cuttings were soaked in tap water for 72 hours and then dipped in 4 ppm Indole-3— Butyric Acid (IBA) diluted in a 1 : 1 ratio of distilled water and 75 95 alcohol. Cuttings were then grown in paper pots (7.61 cm diameter, 27.94 cm high) using media containing a 3 : 2 ratio of peat moss and vermiculite. Plants were grown in a greenhouse under ambient light conditions and given supplemental fluorescent lighting to maintain a 16 hour photoperiod. Temperatures were i 27 ° C during the day to :I: 18 ° C during the night. The plants were transferred to a shadehouse for two weeks before outplanting. The experiment was established on May 10 - . . n... o . FIIL . . \- Illllo 1i .\ T . . r4 . .- _ _ .I--. it. u... r ilm. . . ...- . .. T: I-’ I'Ild . r‘nl-|ld - a . 1 x . Tuitfiiti. 'II I." 5-- -I‘O'---- 5-- 1 .41 .q. I. u w . . . . 32.-.... . 1.1.1 . n . .u . . 1ItJ ‘- . 1--J'--Cd - a--- . . . . . r-..!...--J . a}; L -8, '.I-- u c 10"---“ VIII-I'd..-‘ * a . - . u a - . rile-I‘QH'IIFU‘ r‘-c\& . _ . . . - . . . r- -gl'. . d ”Hull-'I‘ Hui-"II" _ les) irc Figure 1. The location of clone collection (c 11 27 - 29, 1987 at the Tree Research Center, Department of Forestry, Michigan State University, East Lansing, Michigan. The experiment used a split-split plot arranged in a completely randomized block design. Water regimes were the main plot and nitrogen levels were the subplot, while clones were used as the sub-subplot. Each experimental unit was represented by two ramets. The spacing was 1.5 x 1.5 m. The following water treatments were applied: 1. no water stress (NWS): soil water potential was kept at -0.01 MPa or less; 2. moderate water stress (MWS): soil water potential was brought back to -0.01 MPA whenever it reached -0.03 MPa; 3. severe water stress (SWS): no water supply except natural rainfall. The main plots were separated from one another by a plastic barrier 0.5 m deep in the soil. For the water treatment, microsprinklers were installed 30 cm from the plants. Every plant had two microsprinklers. There were also three nitrogen levels used: 1. no nitrogen application (N1); 2. 200 kg N per hectare per year equivalent (N2); 3. 400 kg N per hectare per year equivalent(N3). The nitrogen treatments were applied using ammonium nitrate (NH4NO3) in the months of June, July, August and September. 12 The experiment was kept free from weeds throughout the year by spraying with glyphosate a week before and about a month after planting. Data Collection and Analyses Soil 'was analyzed and a soil moisture retention curve was developed using a pressure plate apparatus according to the method developed by Richard (1965). Rainfall data were also collected from the weather station at the Tree Research Center. Height and diameter growth were measured at planting time. At the end of the first growing season, height and root-collar diameter data were collected. No physiological characteristics were measured. The data were analyzed using analysis of covariance with the following model: Yijkl = (1+ Ri + wj + Rwij + Nk + ijk 4 RNKijk + C1 + wcjl + chl + wncjkl + b(xijkl - i) + eijkl where Yijkl = height or diameter in replication i, water status j, nitrogen level k and clone 1; H = grand mean: Ri = the effect of the ith replicate; Wj = the effect of the jth water status; Rwij = the experimental error for the main plot: Nk = the effect of the kth nitrogen level; ijk = the effect of interaction between the jth 13 water status and kth nitrogen level; - the experimental error for the sub-plot; C1 = the effect of the 1th clone; Wle = the effect of interaction between the jth water status and 1th clone; ch1 = the effect of interaction between the kth nitrogen level and 1th clone; WNCjk1= the effect of interaction between the jth water status, kth nitrogen level and 1th clone; 0‘ II regression coefficient between Y and X; Xijkl = initial height or diameter in the replication i water status j, nitrogen level k and clone 1; i = the mean value of X; Eijkl = the experimental error for the sub-sub- plot. Analysis of covariance was used because of heterogeneity of the initial height and diameter. In the model, water status and nitrogen level were considered as fixed effects, while clone and site (population) were considered ,as random effects. Clones were nested within sites. Plot means were used as data entries. To ascertain the amount of genetic variation that can be attributed_to clones between sites, Clones within sites, and second and third order interactions involving clones, components of variance were calculated from the expected 'mean squares (Table 1). 14 Table 1. Expected mean square estimation used to calculate variance component Source of Mean Square variation Expected Mean Square Clones Between sites MSBS Within sites MSWS W X C MSWC N X C MSNC W x N x C MSWNC Error MSE 02E 02E rwnozws + rwncOZBs rwnOZWS rnOZWC erZNC 'rOZWNC W, N and C are water status, nitrogen level and clones, respectively. r,w,n and c are the number of replicate, water status, nitrogen level and clones within sites, respectively. RESULTS Soil Analysis and Rainfall The physical and chemical properties of the soil in the field experiment are shown in Table 2. A soil moisture retention curve and the amount of rainfall occurring during the study are presented in Figure 2 and Table 3, respectively. Height A test of homogeneity of variance indicated that the variance among clones across two sites was homogeneous. The measurement data in height growth after one growing season were then analyzed using analysis of covariance as shown in Table 4. Differences between water status in height growth were significant. The adjusted mean height growths were 181, 183, and 153 cm for NWS, MWS, and SWS, respectively. The Duncan’s multiple range test in Figure 3 indicated that the adjusted mean height growth between NWS and MWS was not significantly different. SWS had a significant effect in reducing height growth. Although differences between nitrogen levels in height were statistically not significant, nitrogen application did tend to increase the height growth. The adjusted mean heights due to nitrogen application were 167, 174, and 177 cm for N1, N2, and N3, respectively. Significant 15 16 Table 2. Physical and chemical properties of soil in the field plot study Texture (%)* Concentrations (ppm) . pH Sand Silt Clay N P K Ca Mg Na 67.8 18.7 13.4 555.8 281.4 36.1 315.3 71.0 20.2 6.5 * sandy loam l7 -012 -010- If 3 -0.08 - E 5 -0.06 - o O. é’ -o.04 — . E 2 -002- I l l l l O 6 10 16 20 25 30 35 Mass water content (1.) Figure 2. Moisture retention curves for soil in the field plot study. 18 Table 3. Rainfall and temperatures in experiment one of the field plot study Amount of rainfall (inches) Date May June July Aug. Sept. Oct. 1 - - 0.01 0.02 - 0.15 2 - - - 0.07 0.15 0.10 3 0.18 0.44 - - - - 4 - - - 0.25 - - 5 — - _ — .— _. 6 - 0.24 0.15 - - 0.08 7 - 0.21 - - - 0.02 8 - - - - 0.09 - 9 0.05 0.10 - 0.56 0.24 - 10 - - 1.00 0.04 - - 11 - - 0.29 - 1.20 - 12 0.33 0.28 - - - - 13 - - - - 0.23 - 14 - - 0.03 - - - 15 - - - 0.30 0.48 - 16 - - 0.11 - - - 17 - - - 0.40 0.13 0.10 18 - - - - 0.12 0.08 19 0.11 - - 0.10 - - 20 0.18 - - - - 0.52 21 0.40 - 0.03 - 0.15 0.13 22 - 1.14 - 1.99 0.24 0.07 23 - - - - - 0.30 24 - - - - - 0.13 25 - - 0.54 - - 0.21 26 - - 0.35 - - 27 - - - 1.41 - 0.36 28 - - - 0.12 - 29 - - - 0.07 0.58 0.04 30 - 0.16 - - 0.14 31 0.22 - - 0.21 - Total 1.47 2.57 2.51 5.52 3.75 2.30 Normal*) 2.57 3.50 2.78 3.04 2.54 2.13 Average temperature (0C) Max. 23 28 28 26 23 11.5 Min. 9.5 18 17.5 15 11 1.5 *)Source: United States, Dept. of Commerce. 1 l9 .Acmuuuucmum a0: um: .m~m>aaumammL .m_w:w~ \. L mzm N m mfiu um uzmomdwcmsm u xx x .m~w>nuumammt .Commmm UCAEOLm mo Uco m£u Ucm mesa mc_me~Q mfiu um Umtnmmme ufimwwfi u > .x ..x mm._c unoc m; ”n mv.rm “Ago we ”a mo.mn ufimo ma .N we. n~ unuv :L N mv.m_ HALO 3U N m_.mw "wmc Du m: on .: amn.mm¢ omo.mmmm1 mm vmnmainm oxzx: m: we. : Iwm.m1m 1mm.n14¢ v1 nmnmzmum ox: u: mm. a m¢_.oom ovo.mvom vs nmrmz_vm ox: x. .. .m mmm.m~¢m mmm.m1m2 m nmnmcenm mmnflm c_rnu: xx 25.1m 3mm.mwmmm cam. menu I :mnmzanm mmnlm cmmznmm r rm .v ccm. nmfim 0mm. mmmmw. m Dmamafinm u mm.fimvmcd mmfi ONm.mmva~ Mum.mmm mmm.mmm~ cm“ LOLLw+uxzzz Nnm.¢mmmm mm" mmo.mmnmm vm~.mam amm.mcm_ ova LOLLw + or: mmm.monmm mm“ m~m.m0¢mm 0mm.mvm www.mmow ova LOLLw + 0x: me.:mwmod mm" DNN.wOmG~_ www.mmom mm_.mD_N cm“ LDLLm + m.: Dom.mfiamv_ wmfl ONm.vmmmw_ www.mfimw www.mm_m NN~ LOLLw +.m.m an.t.bnm_ Nma Dmm.mm_mm_ -v.mm~m mmm.?mmN mm" LDLLm + q mmm.mmh LNm.@w~vm mm— .Omm.wmmvm m_o.m®m ww~.¢hm_ flm_ new LOLLu nw~.OVNN~ mfim.mm mmwtmmm mm 0 x z 7 : mom.ommv .mmH.QDN m_N.mm vfi U x z wm~.mm~m mmm.mmN www.mmm v“ 0 x : va.mv¢m— va. mum“ fimm.vm¢ m mmumm Cains: www.mwvfim arm. mmfiw www.mwm a mmaflm cmmsamm mmm.m~mmm wmm.momh mm.omm m ADV mwco_u m... SN. :5. 3mm vmfinmoo— poor... 33%. z A z u: mm.“ m: .mwg mwmdmrm umnm 1mm 2 CD Ch CD O D LI'.‘ 1 ("\J N v-‘U'lLDNV' H moo.momvm 666.com omn.mm1 on Lotto + : Hoo.msso_ 1 14m.ommmu mom.mo~ u mmm.mm 41 tortm + z mvm.1m11 4mm.¢vvo~ 1 616.1mvm1 mmo.mm mm6.mm m1 any Lorre mmo.omom~ mmo.mmm mmo.mm v z x : mmm.mmmm mmm.mm~ I mmo.m m Azo m_m:m~ :mmotn_z x mm.m_ mmn.ommm1 mm1.1cmam N .vwrmzavm : mk1.m¢_wv m mm1.mmmmv cco.mm1 . mmm.ov_ m Lorre + : mic.msm1 ovo.mwmm m omm. mama .86. mm_ lim.mm 4 Awe Lorre mam.mkmkm vom.~mm I m n.1o1 m ago msnmnm imam: nno.mmoam _mm.moim "mm. om_ m mmbmoinama an: cmm to 3.} >.x x.x counur.m> w2~m> u . I . CD 10 x Lou nmLm:_um > seawata mmoro lo 5:. mu-.:em Dwaum coda tfimdu mLa no mco acmemtmaxm Cm uzmmmfi L04 wucmmLm> 00 do wmmmfimcc .v ofinmk 20 200 b b 180 L 160 _ a 140 — 120 r II 0 SWS MWS NWS Water status Hebht(cm) Figure 3. The effect of water status on height after one growing season in experiment one of the field plot study. Any means with the same letter are not significantly different by Duncan's multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 21 interactions between water and nitrogen treatments were not detected, indicating that the effect of the water stress treatments on height growth was independent of the rates of nitrogen. Differences among clones between sites, as well as differences among clones within sites were significant. The adjusted mean height growths were 155, and 190 cm for the dry site and the wet site, respectively. None of the two-way or three-way interactions involving water status, nitrogen levels and clones were significant, indicating that each factor acted independently in affecting height growth. Diameter As with height growth, a test of homogeneity of variance showed that the variance among clones across the two sites was homogeneous“ The result of analysis of covariance is presented in Table 5. Water treatment had a significant effect on diameter after one growing season. The adjusted mean diameters were 24, 21, and 19 mm for NWS, MWS, and SWS, respectively. As can be seen in Figure 4, the adjusted mean diameter at NWS was significantly different from that of MWS and SWS, iruiicating that both MWS and SWS reduced diameter growth. As with height growth, nitrogen application did not lurve a statistically significant effect upon. diameter growth. The adjusted mean diameters as affected by nitrogen application were 20, 22, and 21 mm for N1, N2, and N3, 22 .LcmuwuLcmLm Lo: u m: umLm>LaumammL .mLm>mL x L Ucm m m£u um acmoLmLcmLm H .mLm>L»umammL .Commmm mCLSOLm no Uta OLE vcw wELu mcmuiwLQ 05L um DwLDmmme quwemLfl H} .K L Xx .x . x 4m.LoL Loose .. mo.mLL u Amowa ”x Lm.Lm Lmoma .x oL.vL u Locpo .x no.6L u Lasso x Lo.mm Loose m: vm.o mmmo.m mem.mm am earmafivm a 2x: m: em.o om-m.m nmmm.mn 4L . umnmsmnm ox: mc mo.L mmmL.m vme.nm 4L nmnmaLum ox: m: em.L mmww.m wmmm.?aL o nmnmqflvm mmLLm :LrLLz x LL.m kLm?.wcL an .moL L umnannm mmLLm cmmznmm x mm.m swam.mm LmLm.mmm a nonmaaum u mewm.momL mmL mmmm.meL mme.mm ommm.om me Lottw. uxzx: momk.LomL mmL mamm.LomL mmmm.mm mme.mv ovL Lotto. an: womm.m4mL mmL mme.eLmL mmmv.mm mmmo.mm osL Lotto + ox: mevv.mmmL LmL oemm.m:mL ammo.mm emm¢.vw mmL Lorin + .m.: momm.mmmL me omom.ommL oovm.¢mL oomw.mm mm Lottm + .m.m mme.mmmL mmL mmmL.LLmL mamm.LmL mev.mw mmL tortm + u mmmw.m mmom.LuLL mmL LmLm.mmmL ammo.cm mmvm.mv emL . Luv forth wmwm.mmL mmmL.om mem.oL mm o x z x : wsem.vu mmmm.o omvm.m 4L u x z mLLm.mm mumm.¢Ln mmov.m 4L 6 x : mmmm.om vmmm.m I mmvm.m m maLLm :LLLL: mmmm.vom mem.mm ksoo.vL L mLLm cmmsbwm mamm.vmm mamm.mm mmvm.mm n Luv mcoLU m: La.L mnom.mm mmmm.vm v umnmaan :3: m: mm.m vnmm.mm qum.mm m nmnm:_1m z mLmo.m¢u mL mmLa.mmm momm.th kmkm.m 6L . Lorin + 2x: :nmv.mmm mL mmmm.mmm oomm.le momm.m .wL tattm + z mnom.mL mem.LmL LL mkum.mmL owem.mL- man.m ..NL Lao Lotto mmmm.Lm momm.v mem.o a -z x z amwv.mm oeum.o mmam.o N Lzo mLm>mL :mqotn_z x vm.mL mamm.oqv ommm.ouc N e . vmnmsnnm : . mmmv.me m Nmkm.meL vme.mmu omwm.m m Lotta + z mme.mu mvmm.on m ham.koL momm.n n nmLm.L q Lmo retro ommr.mmoL mme.Lm- mmmm.m w Lzo mnnmnm ting: mem.kvm mmmL.mv mme.mL m mmnmoLLa.a cm: cmm to >> >x xx coLLmLtma mun-LAN) u QD nth. v“ L0.“— fimfiwmj—ifim up. ”mongol—Q MWOLJ “*0 630. WULerm mvznm LoLa meLa mrn to mco quELLmuxw CL qumemLD Lou mOCmLLm>OU do mmemec .m mLEML 23 24 ‘u 22 - a 20 a 18 16 14 12 0 SWS I I I Diameter (mm) r I MWS NWS Water statue Figure 4. The effect of water status on diameter after one growing season in experiment one of the field plot study. Any' means ‘with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 24 respectively. Interaction between water and nitrogen treatments was also negligible. There were significant differences among clones between sites, as well as among clones within sites. The adjusted mean diameter for clones originated from the dry site was 20 mm, while the adjusted mean for clones from the wet site was 22 mm. As with height growth, no significant effect for two- way or three-way interactions involving clones, water status and nitrogen levels was detected. Variance Component Estimation The estimation of variance components for height and diameter growth is presented in Table 6. Of the total genetic variation (clones), clones between sites were the major source of variation for both height and diameter. The contribution of clones between sites to the total variation was higher for height growth. 25 Table 6. Variance component estimates for height and diameter in experiment one of the field plot study Component estimates (%) Source of variation Height Diameter Clones Between sites 36.2 8.2 Within sites 4.8 3.2 W X C 0 0.3 N x C O 0 W X N X C 0 0 Error 59.0 88.4 Experiment Two MATERIALS AND METHODS Establishment The plantation of experiment one was coppiced to about 20 cm above ground in March, 1988. The stumps were allowed to sprout the following spring, and thinned to one vigorous sprout per stump. This material was the basis of experiment two. In this experiment water and nitrogen treatments were similar to those applied in experiment one. Soil moisture sensors (Soiltest, Inc.) were installed in every SWS treatment in addition to tensiometers, to record soil water potential below - 0.08 MPa. In May, 1988 glyphosate herbicide was sprayed onto the experiment for weed control. Data Collection and Analyses On July 29 - 31, 1988 a number of physiological traits were measured. Stomatal conductance, photosynthetic and transpiration rates were measured using an ADC open system infrared gas analyzer. The measurement was taken on the youngest fully-expanded leaf of each tree. A leaf of the tallest tree within plot was used for the physiological measurements. A preliminary study indicated that the results of measurement conducted at the youngest fully-expanded leaf were strongly correlated with those taken from whole crown (an average of several measurements taken at several height positions). For example, the coefficients of correlation 26 27 between the youngest fully-expanded leaf and the whole crown were 0.8 and 0.79 for stomatal conductance and photosynthetic rate, respectively. These correlations were significant at the 1 % level based on 24 samples which were taken 5 days before the actual measurement. Stomatal conductance, photosynthesis and transpiration were measured from 10.00 am to 2.30 pm. To reduce error due to measurement time, each block was measured at approximately the same time. At the time of measurement photosynthetically active radiation (PAR) was above 2000 QmOl m—ls—l. Leaf water potential was determined on the same leaf as other physiological traits soon after the measurements were finished using a PMS pressure chamber (PMS Instrument Co., Corvallis, Oregon) The leaf was then taken to the laboratory for specific leaf weight (leaf area/leaf dry weight) determination. At the end of the growing season, height and root collar-diameter were measured. Additional rainfall and daily temperature data were collected from the weather station nearby. The data were analyzed using analysis of variance. The variables and assumptions regarding the variables were similar to those used in experiment one. Plot means were used as data entries. RESULTS Rainfall and Temperature The data of rainfall and temperatures during the experiment are presented in Table 7. During the first three months of ‘the experiment, the rainfall was considerably below normal, resulting in a severe drought. Then, from August onward the average amount of rain fell. Height Water status had a highly significant effect upon height after one growing season (Table 8). The mean heights as affected by water status were 280, 324, and 332 cm for severe water stress (SWS), moderate water stress (MWS) and no water stress (NWS), respectively (Figure 5). Water stress reduced height growth 15.5 % and 2.5 % for SWS and MWS, respectively. Nitrogen level as well as interaction between water status and nitrogen level had no statistically detectable effect on height growth. Nevertheless, nitrogen fertilization enhanced height growth. The mean heights as influenced by nitrogen rate were 306, 316, and 314 cm for no nitrogen (N1), 200 kg/ha/yr (N2), and 400 kg/ha/yr (N3), respectively. There were highly significant differences among clones between sites as well as among clones within sites. However, no second or third order interactions involving 28 29 Table 7. Rainfall and temperatures recorded in experiment two of the field plot study Rainfall (inches) Date May June July August Sept. Oct. 1 L L L L L L 2 - 0.14 - - - 0.75 3 - 0.02 - - 0.30 - 4 L L L L L L 5 - - - - 0.20 0.11 6 - - - 0.80 - - 7 L L L L L L 8 L L L L L L 9 0.04 0.04 - 0.03 - 0.49 10 0.21 - - 0.06 - 0.05 11 - - 0.05 - - - 12 - - - - - 0.04 13 - - - 0.10 0.46 - l4 - - - - - - 15 - - - 0.59 - - 16 0.31 - - - - 0.06 17 0.04 - 1.43 0.10 - 0.50 18 - - - 1.50 - 0.83 19 - - 0.20 0.14 0.29 0.13 20 - - - - 1.16 - 21 - - - - 0.11 0.04 22 - - - - - - 23 - - - 0.64 1.70 - 24 - — - 0.10 - 0.53 25 - - 0.33 - - 0 25 26 - - 0.04 - - - 27 - - - - - - 28 - - - - - 0.04 29 - 0.06 - - - - 30 - - - - - - 31 - - - - - - Total 0.60 0.26 2.34 4.08 4.22 3.82 Normal*) 2.57 3.50 2.78 3.04 2.54 2.13 Average temperature (UC) Max. 22.5 29 31 29 22.5 12 Min. 7.5 12 16 17 10 2.5 *) Source: Unites States, Dept. of Commerce. 30 Table 8. Analysis of variance for height in experiment two of the field plot study Source of Variation Df. Mean Square F value Replicates 2 -------- Water status(W) 2 55091.040 22.20 ** Error (a) 4 2481.629 Nitrogen levels (N) 2 2084.145 1.14 ns W X N 4 1415.079 0.77 ns Error (b) 12 1833.238 Clones (C) 7 21829.570 28.73 ** Between sites 1 102643.100 12.28 * Within sites 6 8360.650 11.02 ** W X C 14 460.932 0.61 ns N X C 14 594.789 0.78 ns W X N X C 28 594.663 0.78 ns Error (c) 126 759.936 CV(a)= 15.96 %; CV(b)= 13.72 %; CV(c)= 8.83 %. *,**=significant at the 5 and 1 % levels, respectively. ns= not significant. 31 340 b b 320- 300~ " a E 280- E E? 260- o I 240. ' 220- .I SW8 MWS NWS Water Statue Figure 5. The effect of water status on height after one growing season in experiment two of the field plot study. Any' :means with the same letter are not significantly different by Duncan's multiple range test at the 5% level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 32 water status, nitrogen level and clone were detected. This indicates that the site where clones were collected had an independent effect of water status and nitrogen rates on height growth. Plants originated from the dry site grew more slowly than those from the wet site regardless of water status and nitrogen levels. The average height growth across all treatments for plants from the dry site was 290 cm, whereas the corresponding figure for plants from the wet site was 334 cm. Clones from the wet site grew 13 % faster in height than those from the dry site. Diameter As with height, water status influenced diameter growth significantly after one growing season (Table 9). The mean diameters were 27, 33, and 36 mm for SWS, MWS, and NWS, respectively, and all three differed significantly (Figure 6). Water deficit reduced diameter growth by 26 % and 8 % for SWS and MWS, respectively. Nitrogen level had a minor effect on diameter growth. The interaction between water status and nitrogen rate was also negligible. Differences among clones between sites as well as among clones within sites were highly significant. As with height, no interaction involving water status, nitrogen levels and clones was statistically significant. The average diameter of clones from the dry site was always smaller than that of the wet site clones, irrespective of water status 33 Table 9. Analysis of variance for diameter season in experiment two of the field plot study Source of Variation Df. Mean Square F value Replicates 2 ------- Water status (W) 2 1679.832 60.38 ** Error (a) 4 27.822 Nitrogen levels(N) 2 57.469 0.78 ns W X N 4 84.660 1.15 ns Error (b) 12 73.622 Clones (C) 7 322.629 14.32 ** Between sites 1 1410.002 9.97 * Within sites 6 141.399 6.28 ** W X C 14 3.523 0.16 ns N X C 14 16.096 0.71 ns W X N X C 28 12.568 0.56 ns Error (c) 126 22.523 CV(a)= 16.40 %; CV(b)= 26.68 %; CV(c)= 14.76 %. *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 34 40 c as .. A b E 5 E 30 - O 5 5 a . 25 - o sws Mws NWS Water status Figure 6. The effect of water status on diameter after one growing season in experiment two of the field plot study. Any' means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 35 or nitrogen level. The mean diameters were 30 and 35 mm for clones from the dry site and clones from the wet site, respectively. Clones from the wet site grew 15 % faster in diameter than clones from the dry site. Leaf area Water status was the only variable that showed a significant effect upon leaf area (Table 10), and all three water status treatments differed significantly (Figure 7). All interaction effects were negligible. Even though there were no significant differences among the nitrogen treatments, nitrogen application tended to increase leaf area. The mean leaf areas as affected by nitrogen level were 134, 138, and 140 cm2 for N1, N2, and N3, respectively. Clones between sites did not differ significantly, despite the fact that clones from the wet site had greater leaf area than those from the dry site. The mean leaf areas due to site were 113 and 161 cm2 for clones from the dry site and wet site, respectively. Specific Leaf Weight Water regimes influenced specific leaf weight (SLW) significantly (Table 11). Water deficit appeared to increase SLW. The mean SLWs were 8.9, 8.2, and 7.7 for SWS, MWS, and NWS, respectively (Figure 8). All three water status treatments differed significantly in SLW. 36 Table 10. .Analysis of ‘variance for leaf area in experiment two of the field plot study~ Source of variation Df. Mean square F value Replicates 2 ------- Water status (W) 2 7533.884 21.54 ** Error (a) 4 349.766 Nitrogen levels (N) 2 756.479 1.23 ns W X N 4 49.097 0.08 ns Error (b) 12 697.746 Clones (C) 7 56476.310 75.84 ** Between sites 1 122136.600 2.68 ns Within sites 6 45532.933 61.14 ** W x C 14 1067.196 1.43 ns N X C 14 1261.636 1.69 ns W X N X C 28 745.227 1.00 ns Error (c) 126 744.672 CV(a)= 13.63 %; CV(b)= 18.05 %; CV(c)= 19.89 %. ** = significant at the 1 % level. ns = not significant. 37 160 c 140 ~ b &~ 130 ~ ‘5 a a 2 9 120 L ”a e .4 110 P I 0 SWS MWS NWS Water status Figure 7. The effect of water status on leaf area at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 38 Table 11 . Analysis of variance for SLW in experiment two of the field plot study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 23.7125 29.60 ** Error (a) 4 0.8012 Nitrogen levels (N) 2 2.5127 1.16 ns W x N 4 0.4983 0.23 ns Error (b) 12 2.1617 Clones (C) 7 4.0800 6.63 ** Between sites 1 0.7609 0.16 ns Within sites 6 4.6332 7.53 ** W X C 14 0.3640 0.59 ns N X C 14 1.0315 1.68 ns W x N x C 18 0.5719 0.93 ns Error (c) 126 0.6149 CV(a)= 10.85 %: CV(b)= 17.30 %; CV(c)= 9.50 %. ** - significant at the 1 % level. ns not significant. 39 9.0 . a 8.6 _- b 5 80 _- U) C 7.6 _ 4. 0.0 sws MWS Nws Water statue Figure 8. The effect of water status on SLW at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by Duncan’s multiple range test at 5 %. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 40 The effects of nitrogen levels and clones between sites on SLW’ were not significant- None of the interaction effects were statistically significant. There were, however, large differences among clones within sites that were statistically significant. Leaf Water Potential Soil water status affected leaf water potential significantly (Table 12). As the water deficit increased in the soil, leaf water potential decreased. The mean leaf water potentials were -1.06, -1.01, and -1.01 MPa for SWS, MWS, and NWS, respectively. Only the SWS treatment was significantly different from the other water regimes for leaf water potential (Figure 9). Nitrogen treatments did not have a significant effect upon leaf water potential. Differences among clones between sites, as well as all interaction effects, were not significant. Clones within sites, on the other hand, differed significantly. Stomatal Conductance Among the many variables involved in the analysis, only water status had a significant effect on stomatal conductance (Table 13). No interaction terms were significant. The mean stomatal conductances, as affected by water status, were 0.46, 0.61, and 0.72 mol m-zs-l for sws, MWS, and NWS, respectively. SWS, MWS and NWS treatments were 41 Table 12. Analysis of variance for leaf water potential in experiment two of the field plot study Source of variation Df. Mean square F value Replicates 2 ------- Water status (W) 2 6.1817 7.29 * Error (a) 4 0.8484 Nitrogen levels (N) 2 3.4456 2.23 ns W X N 4 0.9682 0.63 ns Error (b) 12 1.5464 Clones (C) 7 3.3953 5.39 ** Between sites 1 4.5938 1.44 ns Within sites 6 3.1956 5.08 * W X C 14 0.4211 0.67 ns N X C 14 1.0144 1.61 ns W X N X C 28 0.6659 1.06 ns Error (0) 126 0.6296 CV(a)= 8.95 %; CV(b)= 12.09 %; CV(c)= 7.71 % *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 42 ~108 a l~ -L06- n a. 3 -1.05- E g -1.04 — o a 6 -1.03 - ‘6 3 :5 -L02- 3 b -1.01 — b 000 lli SWS MWS NWS Water statue Figure 9. The effect of water status on leaf water potential at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 43 Table 13. Analysis of variance for stomatal conductance in experiment two of the field plot study Source of Variation Df. Mean Square F value Replicates 2 -------- Water status (W) 2 1.2101 37.12 ** Error (a) 4 0.0326 Nitrogen level (N) 2 0.0061 1.22 ns W X N 4 0.0127 2.54 ns Error (b) 12 0.0050 Clones (C) 7 0.0008 0.07 ns Between sites 1 0.0006 0.06 ns Within Sites 6 0.0095 0.82 ns W X C 14 0.0099 0.85 ns N X C 14 0.0125 0.89 ns W X N X C 28 0.0452 1.25 ns Error (C) 126 0.0116 CV(a)= 30.29 %; CV(b)= 11.86 %; CV(c)= 18.07 %. **= significant at the 1 % level. ns= not significant. 44 significantly different from one another in stomatal conductance (Figure 10). SWS and MWS reduced stomatal conductance about 57 % and 18 %, respectively. Transpiration As with stomatal conductance, only water status had a significant effect upon transpiration. None of the other main effects or interaction terms were statistically significant (Table 14). The average transpiration rates were 9.5, 10.5, and 10.3 mol m-zs-1 for the SWS, MWS, and NWS treatments, respectively. The SWS treatment reduced transpiration rate significantly (Figure 11). The ZMWS and. NWS, treatments, however, did not differ significantly in transpiration rate. Photosynthesis Unlike stomatal conductance and transpiration, no main factors had a significant effect on the photosynthetic rate (Table 15), although water (deficit tended to lessen photosynthesis. The mean photosynthetic rates, as affected by water status were 15.9, 17.9, and 17.3)lmol COZm-Zs-l for the SWS, MWS, and NWS treatments, respectively. Correlations Between Characteristics A correlation analysis was carried out on the physiological characteristics measured in the experiment (Table 16). Most of the correlation coefficients were low. 45 10.8 c I.“ b E 10.6 . o 5 g a o i: n h- ‘5 10.4 3 'c I: o o 2 10.2 .- n E 2 m b 0.0 SWS MWS NWS Water status Figure 10. The effect of water status on stomatal conductance at the end of July, 1988 in experiment two of the field plot study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 46 Table 14. Analysis of variance for transpiration rate in experiment two of the field plot study Source of variation Df. Mean square F value Replicates 2 --------- Water status (W) 2 18.0080 9.48 * Error (a) 4 1.9001 Nitrogen level (N) 2 3.4548 3.59 ns W x N 4 1.3952 1.45 ns Error (b) 12 0.9631 Clones (C) 7 0.3245 0.39 ns Between sites 1 0.0937 2.43 ns Within sites 6 0.0385 0.05 ns W X C 14 0.9715 0.39 ns N X C 14 0.6255 0.76 ns W X N X C 28 0.7660 0.93 ns Error (c) 126 0.8249 CV(a)= 13.68 %; CV(b)= 9.74 %; CV(c)= 9.01 % * = significant at the 5 % level. ns = not significant. 47 110 b a- "15 - 5 b E '0 10.0 — 5 e '5 9.6 - a ' C 2 1'! 9.0 ~ a C 2 F 135 ~ F 00 EflNS NHNS hflNS Water status Figure 11. The effect of water status on transpiration rate at the end of July, 1988 in experiment two of the field plot study. Any means with same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 48 Table 15. Analysis of variance for photosynthetic rate in experiment two of the field plot study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 72.9058 1.70 ns Error (a) 4 42.9984 Nitrogen levels (N) 2 22.1677 1.67 ns W X N 4 17.2935 1.30 ns Error (b) 12 13.2593 Clones (C) 7 2.1968 2.42 ns Between sites 1 1.6119 0.70 ns Within sites 6 2.2943 0.45 ns W X C 14 9.0633 1.76 ns N X C 14 6.3590 1.23 ns W X N X C 28 6.6598 1.29 ns Error 126 5.1496 CV(a)= 38.54 %; CV(b)= 21.40 %; CV(c)= 13.34 %. ns = not significant. 49 Table 16. Correlations between the physiological characteristics measured in experiment two of the field plot study Leaf Water Stomat. Transp. Photosyn. Potential Cond. Rate Rate Leaf Water -0.15* -0.07ns -0.15* potential Stomatal 0.53** 0.45** conductance Transpiration 0.23** rate *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 50 However, there are several points worth mentioning. Leaf water potential was negatively correlated with stomatal conductance, transpiration, and photosynthetic rates. In this regard the coefficients of’ correlation. were -0.15, -0.07, and -0.15 for stomatal conductance, transpiration, and photosynthetic rates, respectively. Only the correlation between leaf water potential and transpiration rate was not significant. Stomatal conductance was positively correlated with transpiration and photosynthetic rates. Variance Component Estimation The amount of variation for the characteristics measured that can be attributed to clones between sites, clones within sites, interaction involving clones, as well as residual error was estimated using components of variance (Table 17). The contribution of clones between sites to the total variation was quite large for characteristics such as height and diameter." The amount of variation that accounted for by clones between sites were 46 and 30 % for height and diameter, respectivery. The contribution of clones within sites to the total variation was relatively small for height and diameter. For leaf area, clones within. sites were the major source of variation (52 8;), but the variation that was attributable: to the clones between. sites was relatively large (22 %). Table 17. 51 two of the field plot study Variance component estimation in experiment Component of variance (% of total) Characteristics 02 BS ozws ozwc 01ch 02WNC 02E Height 45.6 14.7 0 0 0 39.7 Diameter 30.4 11.4 0 0 0 58.2 Leaf area 22.1 51.7 1.1 1.8 0. 23.2 SLW 0 2.2 0 6.9 0 90.9 Leaf water 1.5 10.7 0 4.8 12.3 70.8 potential Stomatal 0.1 0 0 0.5 6.9 92.5 conductance Transpiration 0 0 1.9 0 0 98.1 rate Photosynthetic 0 0 7.2 2.1 8.0 82.7 rate 52 Unlike the growth parameters, the major source of variation for SLW and physiological characteristics was residual variance. Both clones between sites and clones within sites made little or no contribution to the total variation. GREENHOUSE STUDY MATERIALS AND METHODS Establishment As mentioned in the field plot study, the plants in experiment one were coppiced in March, 1988. The shoots removed were used to establish a greenhouse study. These plant materials were kept in cold storage until further use. Hardwood cuttings (25 cm in length, 1 - 2 cm in diameter and having at least two buds) were soaked in tap water for 72 hours. The cuttings were then planted in polyethylene containers containing a 2 : 1 ratio of sand and sandy-loam soil. The containers were 15.2 cm in diameter and 61 cm in height. The spacings were 35 and 25 cm between and within plots, respectively. Temperatures varied from i; 36.5 °C‘during the day to i 17 °C during the night. Plants were grown under ambient light condition. The experiment was established on April 17, 1988 in a greenhouse of the Tree Research Center, Department of Forestry, Michigan State University. Treatment combinations and design of the experiment were similar to those of the field plot study. However, in the greenhouse study the plants were subjected to more severe water stress treatments. The following water treatments were applied: 53 54 1. severe water stress (SWS): plants were watered when soil water potential reached -0.2 MPa; 2. moderate water stress (MWS): plants were watered when soil water potential reached -0.1 MPa and 3. no water stress (NWS): plants were kept at soil water potential -0.001 MPa or less. Soil water potentials at the NWS treatment were monitored using tensiometers, while those at the SWS and MWS treatments were monitored with soil moisture sensors (Soiltest, Inc.). For nitrogen fertilization the following rates were used: 1. no nitrogen fertilizer (N1); 2. 2.25 gr N equivalent per plant (N2) and 3. 4.50 gr N equivalent per plant (N3). Water treatment and nitrogen fertilization were applied 1.5 months after planting. Ammonium nitrate (NH4NO3) was used as the nitrogen source. Data Collection and Analyses In August, 1988 stomatal conductance, transpiration, and photosynthetic rates were measured on the youngest fully-expanded leaf using an ADC open system infrared gas 1 1 when the analyzer. PARs were above 1000 umol m- 5- measurements were carried out. Correlations between values for the youngest fully-expanded leaf and measurements taken over the whole plant (an average of several measurements 55 taken at several height positions) were high. For example, the coefficient of correlation for stomatal conductance was 0.97, while that for photosynthetic rate was 0.79. These correlations were significant at the 1 % level. The correlation analyses were based on 24 samples taken 9 days before the actual measurement. Leaf area and specific leaf weight (SLW) were determined. on ‘the leaf‘ ‘where stomatal conductance, transpiration, and photosynthetic rates were measured. Leaf water potential was measured using a PMS pressure chamber (PMS Instrument Co., Corvallis, Oregon) at several height positions. The reading was then determined as an average of these measurements. Height and root-collar diameter measurements were made at the end of the experiment (2.5 months after the water treatments and nitrogen fertilization were started). The shoots were harvested and dried in an oven. The roots were extracted by soaking in tap water and then dried in an oven. The dried shoot and root were then taken to the laboratory for weighting. Data were analyzed using analyses of variance with the model and assumptions being similar to those of the field plot study. Water status and nitrogen level were considered as fixed effects, while clone and site (population) were considered as random effects. Clones were nested within sites. Plot means were used as data entries. 56 The contribution of population to the total variation was estimated based upon variance component estimations. Stability or plasticity of the populations to changes in ' water status were examined by the method of Eberhart and Russell (1966). The parameters are defined with the following model: Yij = “i + {3in + 61]- th where Yi' is the population mean of the i population J at the jth water status (i=1,2 ...v, j= 1, 2 ... n); th “i is the mean of the i population over all environments; Bi is the deviation from regression coefficient that measures the response of the 1th population to varying water status; is the deviation from regression of the 1th 613- population at the jth water status and Io is the environmental index obtained as the 3 mean of all populations at the jth water status minus the grand mean. The first stability parameter is, L... .2. b 2] Y1] I/ 23 I 3 and the second stability parameter is Sdzi = [fjozij/n-z] - Sez/r where Sez/r is the estimate of pooled error and .2..= .2....2. .. .2. 236 13 [ZJY 13 Y l./n] [:3y131312/231 3 57 This method defines a stable population or lack of plasticity as one with a regression slope (b) of unity and a small residual mean square (Sdzi). A population exhibiting a high b value is defined as a population that is more responsive to an environment of high productivity, while a low b value is associated with a population that does not respond to favorable environments. RESULTS The physical and chemical properties of the soil media used in the greenhouse study are given in Table 18, while the soil moisture retention curve is presented in Figure 12. Average temperatures recorded during the study are shown in Table 19. Height There were significant differences in height between water status treatments after 2.5 months (Table 20). There were also significant height growth differences between nitrogen levels. The analysis of variance showed highly significant linear and quadratic effects on height for nitrogen fertilization. The effect of nitrogen, however, was dependent upon water status, since the interaction between water status and nitrogen levels was highly significant. Both linear and quadratic responses of this interaction were significant. The nature of this interaction effect can be seen further in Figure 13. The added nitrogen dramatically increased height growth if the plants were not under water stress. Nitrogen fertilization had little effect, or even tended to decrease height. growth, when. the plants ‘were subjected. to water stress. 58 59 Table 18. Physical and chemical properties of soil media in the greenhouse study Texture (%)* Concentration (ppm) pH Sand Silt Clay N P K Ca Mg Na 90.3 4.4 5.4 354.8 177.1 36.7 450 48.3 20.4 7.5 * sand 60 -012 I -0J0 -008 I -006 Matrlo potential (MPa) I -004 -002 I l L l l L 0 6 10 15 20 25 30 Mass water content (9») Figure 12. Soil moisture retention curve for soil media of the greenhouse study. Table 19. Average temperatures in the greenhouse study Average temperature (0C) May June July August Maximum 36.0 36.5 36.5 35.5 Minimum 17.0 18.0 20.0 20.0 Table. 20. Analysis greenhouse study 61 of variance for height in the Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 101151.300 731.99 ** Error (a) 4 138.187 Nitrogen levels (N) 2 11650.140 24.16 ** Linear 1 16673.266 34.58 ** Quadratic 1 6627.004 13.75 ** W X N 4 11594.170 24.05 ** Linear 2 19156.083 39.73 ** Quadratic 2 4032.257 8.36 ** Error (b) 12 482.119 Clones (C) 7 2230.928 21.45 ** Between sites 1 5033.921 2.85 ns Within sites 6 1763.763 16.97 ** W X C 14 308.246 2.97 ** N X C 14 128.019 1.23 ns W X N X C 28 114.723 1.10 ns Error (c) 126 103.945 CV(a)= 16.87 2; CV(b)= 31.52 2; CV(c)= 14.63 2. ns not significant. significant at the 1 % level. 62 160 140- . a NW8 1% 100‘ 3 ~ 80 - in b be 0 be ““8 C l —+ I 60 L if ‘ women”): 6 4o- h_I—h_‘—““*—*—~—~————~—LL:“W3 Y-48.07-1.90X 20' 0 ' ‘ l 0 2.25 4.50 Nitrogen level (gr/plant) Figure 13. The effect of water status on height after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 2 level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 63 The effect of site origin of clones upon height growth varied according to water status. This effect was significant only when the plants were not under water stress (Figure 14). The nature of the interaction was apparently not a difference in the direction, but rather a difference in the response magnitude. Plants from the dry site grew more slowly than those from the wet site under well-watered conditions. Diameter The nature of responses for diameter after 2.5 months was similar to that of height (Table 21). The interaction between water status and nitrogen levels was highly significant, indicating that. the effect of 'water status varied with the rate of nitrogen applied. Both linear and quadratic effects of the interaction were significant (Figure 15). As with height growth, nitrogen fertilization showed little impact on diameter growth when plants were under water deficit. In fact, the added nitrogen had a negative effect in the SWS treatment. Nitrogen increased diameter growth dramatically in the NWS treatment, but an increase of nitrogen rate above 2.25 gr produced little additional response. The growth of clones collected from different sites varied according to water status, which was shown by a significant interaction effect between water status and 64 140 - Dry sue a 120 _ Wet Site 100- E 80- \ s 3 60- §§ I \ a 40 — ‘ A§ o . t\ SWS MWS NWS Water status Figure 14. The effect of site origin of clones on height after 2.5 months, as affected by water status in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 2 level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 65 Table 21. Analysis of variance for diameter in the greenhouse study Source of variation Df. Means square F value Replicates 2 --------- Water status (W) 2 193.8419 405.03 ** Error (a) 4 0.4786 Nitrogen levels (N) 2 16.6126 11.59 ** Linear 1 17.6680 12.33 ** Quadratic 1 15.5572 10.86 ** W X N 4 17.3625 12.12 ** Linear 2 29.1562 20.35 ** Quadratic 2 5.5688 3.89 * Error (b) ' 12 1.4329 . Clones (C) 7 5.6829 18.47 ** Between sites 1 7.4370 1.38 ns Within sites 6 5.3905 17.52 ** W X C 14 0.6753 2.19 * N X C 14 0.4089 1.33 ns W x N x C 28 0.4869 1.58 ns Error (c) 126 0.3077 CV(a)= 10.02 2; CV(b)= 17.34 2; CV(c)= 8.03 2. *,**=significant at the 5 and 1 % levels, respectively. ns= not significant. 66 II ' a _*. NW3 1o _ v-7.06o1.68x-2.44x 9 L E h 5 a _ g 3 4 ? uws Y'U.‘75‘0.0IBX -§'7 .— E 3 Li sws O Y'B.715‘0.0IX 6 L 5 L Ii 0 I L 1 0 225 450 Nitrogen level (ppm) Figure 15. The effect of water status on diameter after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’multiple range test at the 5 96 level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 67 clones (Figure 16) . Again, the nature of the interaction among clones between sites and among water regimes was in magnitude of response instead of direction of response. Clones from the dry site had smaller diameter than those from the wet site in all three water status treatments. Only those in the NWS treatment were significantly different. Leaf Area The analysis of variance for leaf area was conducted on transformed data due to the heterogeneity of variance among the water treatments. Natural log-transformations were employed, since the standard deviations of the treatments were more or less proportional to their means. The effect of water status on leaf area was dependent upon the rate of nitrogen applied, since the interaction between water status and nitrogen was highly significant (Table 22). Both linear and quadratic responses of this interaction were highly significant. To further elucidate the joint effect of water status and nitrogen level, an additional analysis was carried out (Figure 17). As with height and diameter, nitrogen fertilization had a significant effect under the condition where plants were not lacking water (NWS) . Leaf area was reduced significantly when nitrogen was not added in the NWS treatment. Increasing the nitrogen level to more than 2.25 gr had little effect on increasing leaf area. By contrast, 68 10 - Dry Site 5’5 “Nivunsuo Diameter (mm) a: I SWS MWS NWS Water statue Figure 16. The effect of site origin of clones on diameter after 2.5 months, as affected by water status in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 69 Table 22. Analysis of variance for leaf area in the greenhouse study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 8.8019 20.57 ** Error (a) 4 0.4279 Nitrogen levels (N) 2 9.8364 49.88 ** Linear 1 14.6564 74.32 ** Quadratic 1 5.6951 28.88 ** W X N 4 5.0879 25.80 ** Linear 2 8.0856 41.00 ** Quadratic 2 2.0902 10.60 ** Error (b) 12 0.1972 Clones (C) 7 0.5000 9.63 ** Between sites 1 0.4606 0.91 ns Within sites 6 0.5066 9.77 ** W X C 14 5.7450 1.11 ns N X C 14 0.0639 1.23 ns W x N x C 28 0.0798 1.53 ns Error (c) 126 0.0519 CV(a)= 15.01 2; CV(b)= 10.19 2; CV(c)= 5.22 2. ** - significant at the 5 2 level. ns not significant. 70 250 I 200- NW9 2 {E V-80.70081.706X-10.266X ~g 150r e 2 “ 6 6 ‘5 100 ~ - : uws .3 v-11.07.17.409x-2.6‘aax 2 ed ed _ 1_ L_.8ws 60 :7 v-60.67oa.927x-0.616x o l l l 0 225 450 Nitrogen level (gr/plant) Figure 17. The effect of water status on leaf area after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. II 71 adding nitrogen above 2.25 gr tended to reduce leaf area when the plants were under water stress. Differences among clones between sites in leaf area were not significant, but clones within sites differed significantly. No interaction involving clones was detected, indicating that the response of clones was independent of water status and nitrogen levels. Despite the fact that differences among clones between sites were not significant, clones from the dry site had smaller leaf areas than those from the wet site. The weighted mean leaf areas were 74.7 and 81.9 cm2 for clones from the dry and wet sites, respectively. Specific Leaf Weight Water status had no significant effect on leaf specific weight (SLW) (Table 23). The effect of nitrogen, on the other hand, was highly significant and its effects varied according to water status. The interaction effect between water status and nitrogen level was primarily due to the linear component. The added nitrogen reduced SLW in a dramatic fashion in the NWS treatment (Figure 18). In every water status, the addition of nitrogen at levels greater than 2.25 gr tended to reduce SLW. As with leaf area, differences among clones between sites were not significant. Clones within sites, however, differed significantly. The effect of the second order interaction was also significant. Analysis of the Table 23. Analysis greenhouse study 72 of variance for SLW in the Source of variation Df. Mean square F value Replicates 2 ------- Water status (W) 2 4.8380 1.90 ns Error (a) 4 2.5524 Nitrogen levels(N) 2 12.7808 6.99 ** Linear 1 25.1004 13.72 ** Quadratic 1 0.4612 0.25 ns W X N 4 2.3770 8.28 ** Linear 2 24.3272 13.30 ** Quadratic 2 5.9680 3.26 ns Error (b) 12 1.8295 Clones (C) 7 4.1550 6.01 ** Between sites 1 3.8346 0.91 ns Within sites 6 4.2084 6.08 * W X C 14 1.0459 1.51 ns N X C 14 1.1841 1.71 ns W X N X C 28 1.2671 1.83 ** Error (c) 126 0.6918 CV(a)= 24.12 2 CV(b)= 20.42 2; CV(c)= 12.55 2. *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 73 8.5 8.0 - Y-8.012-O.527X 7.0 " be ”c - v-o.60oo.oox 6.6 ’- he Y-O.818-0.127X 5° 6.0 — SLW U be “W8 5.5 _ nws 0.0 l l l 0 2.25 4.50 Nitrogen level (gr/plant) Figure 18. The effect of water status on SLW after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 74 interaction involving site origin of clones and other treatments revealed no significant effect. shoot Biomass Data transformations were conducted prior to performing the analysis of variance for shoot biomass. Log transformation was used, since the standard deviations of the water treatment were proportional to their means. The effects of water stress and nitrogen treatments were highly significant for shoot biomass after 2.5 months (Table 24). The interaction between water status and nitrogen level was also significant. The interaction effect was primarily linear. (Figure 19). Again, the added nitrogen fertilizer had its greatest effect when the trees were not under water stress. The addition of nitrogen fertilizer at levels greater than 2.25 gr resulted in little increase in shoot biomass in NWS. Conversely, when the plants were under water stress, nitrogen fertilization had a minor or negative impact on shoot biomass. As with the previous characteristics, shoot biomass of clones between sites did not differ significantly. Clones within sites, however, showed significant differences. The interaction between site origin of clones and water status was also significant (Figure 20). Differences in shoot biomass between clones from the dry site and wet site were apparent only when plants were not under water deficit. Under water stress conditions (SWS Table 24. the greenhouse study 75 Analysis of variance for shoot biomass in Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 45.2280 436.77 ** Error (a) 4 0.1036 Nitrogen levels (N) 2 2.8553 8.85 ** Linear 1 1.6394 5.08 * Quadratic 1 4.0707 12.61 ** W x N 2 3.2431 10.05 ** Linear 1 5.8900 18.25 ** Quadratic 1 0.5959 1.85 ns Error (b) 12 0.4220 Clones (C) 7 1.4769 12.37 ** Between sites 1 3.2593 2.76 ns Within sites 6 1.1797 9.88 ** W x C 14 0.2890 2.42 ** N X C 14 0.1471 1.23 ns W x N x C 28 0.1083 0.91 ns Error (c) 126 0.1194 CV(a)= 13.25 %; CV(b)= 23.39 %; CV(c)= 14.23 %. *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 50 4O - } uws E- 3 30* u wan-13.4.9271: E .2 .n o- 20 r- / o o 2 a: b 9 eliuwe 10 L “i— ' v-1o.aesoo.1e7x r L c ‘ ifiaws Y-0.887-0.483X o J l L 0 2.25 4.50 Nitrogen level (gr/plant) Figure 19. The effect of water status on shoot biomass after 2.5 months, greenhouse study. the5% stress; 76 as affected by nitrogen levels in the Any means with the same letter are not significantly different by Duncan’s multiple range test at level. SWS= severe water stress; NWS= no water stress. MWS= moderate water 77 35 - Dry 8m 30‘ Wel8lte ‘2 25— 8 8 a 20- E .2 n 15.. ‘6 2 an 10F 5:______ii§§ o. NHNS DHNS Water statue Figure 20. The effect of site origin of clones on shoot biomass after 2.5 months, as affected by water status in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 78 and. MWS) these differences were negligible even though clones from the dry site had lower shoot biomass than those from the wet site. Root Biomass Log transformations were also employed for root biomass prior to performing an analysis of variance. water status treatments had a highly significant effect upon root biomass (Table 25). The effect of water status on root biomass was independent of the rates of nitrogen applied. Figure 21 shows that all three water regimes differed from one another significantly in root biomass. The weighted mean root biomass after 2.5 months as affected by the water stress treatment were 2.2, 3.5, and 6.4 gr for SWS, MWS, and NWS, respectively. Nitrogen fertilization also had a significant impact on root biomass, but unlike shoot biomass the interaction between ‘water status and nitrogen rates was of little significance. Nitrogen application appeared to reduce root biomass. The major effect of nitrogen on root biomass was linear (Figure 22). The effect of nitrogen was also independent of clones. There were no significant differences among clones between sites, but clones within sites differed significantly. The effect of clones seemed to vary according 11) water status. However, the interaction involving water status and site origin of clones is of 79 Table 25. Analysis of variance for root biomass in the greenhouse study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 19.9920 36.98 ** Error (a) 4 0.5406 Nitrogen levels (N) 2 1.2051 9.58 ** Linear 1 1.8732 14.90 ** Quadratic 1 0.5370 4.27 ns W X N 4 0.0822 0.70 ns Error (b) 12 0.1311 Clones (C) 7 0.9566 13.65 ** Between sites 1 1.0895 1.17 ns Within sites 6 0.9344 13.33 ** W x C 14 0.2456 3.50 ** N x C 14 0.0861 1.23 ns W x N x C 28 0.0864 1.23 ns Error (c) 126 0.0701 CV(a)= 56. 58 %; CV(b)= 27.28 %; CV(c)= 20.37 = significant at the 1 % level. ns = not significant. r1 80 7 C 6” 1: 5’ 3 s 4- E b O 3 3- 8 a m 2_ 1t ’ 0 SW8 MWS NWS Water statue Figure 21. The effect of water status on root biomass after 2.5 months in the greenhouse study. Any means with the same letter are not significantly different by Duncan's multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 81 5 4 .. 1; Y-4.002-0.18X : a — G E .2 .0 ‘6 2 L O a: 1 l— o l l l 0 2.25 4.50 Nltrogen level (gr/plant) Figure 22. The effect of nitrogen levels on root biomass after 2.5 months in the greenhouse study. 82 interest, even though it was not significant. Clones from the dry site and clones from the wet sites did not differ significantly under the different levels of water stress. Root-Shoot Ratio As with shoot and root biomass, log transformations were needed prior to performing the analysis of variance on data for root-shoot ratio, due to the heterogeneity of variance. Water regimes and nitrogen levels had highly significant effects (n1 root-shoot ratios after 2.5 months (Table 26). In addition, the effect of water status was dependent upon nitrogen levels. Both linear and quadratic responses of interaction were significant. Figure 23 shows that water status had a great impact upon root-shoot ratios. Plants under water deficit tended to have greater root-shoot ratios than those under well-watered conditions. Under’ all three ‘water status treatments, the root-shoot ratio decreased when the rate of nitrogen was increased. However, a more dramatic effect was observed in the NWS treatment, where the additional nitrogen resulted in a significant reduction in the root-shoot ratio. The average response of the clones was not significantly different between the two sites, but the performance of Cflones within sites differed significantly. The first order’ and second. order interactions involving clones with other treatments were highly significant. Analyses of the interaction involving site origin of clones 83 Table 26. Analysis of variance for root-shoot ratio in the greenhouse study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 3.7676 21.72 ** Error (a) 4 0.1735 Nitrogen levels (N) 2 3.5334 32.59 ** Linear 1 6.7664 62.42 ** Quadratic 1 0.3005 2.77 ns w x N 4 1.2534 11.56 ** Linear 2 1.9946 18.40 ** Quadratic 1 0.5122 4.72 * Error (b) 12 0.1084 Clones (C) 7 0.1084 6.64 ** Between sites 1 0.8981 1.92 ns Within sites 6 0.4667 5.86 * W X C 14 0.2215 2.78 ** N X C 14 0.0512 0.69 ** W X N X C 28 0.1322 1.65 * Error (c) 126 0.0796 CV(a)= 11.87 %; CV(b)=9.39; CV(c)= 8.04 %. *,**=significant at the 5 and 1 % levels, respectively. ns= not significant. 84 05 I _ u' v-o.423-o.01ax 0 4 - -b+ "’ . sws 3 v-o.402-o.otex uws 2 0.3 - ‘6 o .2 O J- (lzr 8 2 NW3 m Y-0.46-0.163X-0.021X 0.1 r 0.0 L 4 I 0 2.26 4.50 Nitrogen level (gr/plant) Figure 23. The effect of water status on root-shoot ratio after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 85 and other treatments showed no significant effect. Although the root-shoot ratio of clones between sites was not significantly different, clones from the dry site tended to have higher root-shoot ratios in every water status than those from the wet site. Leaf Water Potential The differences in leaf water potential between water status were highly significant (Table 27). Nitrogen rates, on the other hand, had a negligible effect. The interaction effect between water status and nitrogen levels was also very small. The leaf water potentials in all three water status treatments were significantly different (Figure 24). Water stress caused the leaf water potential to be more negative. There were no significant differences among clones between sites for leaf water potential. Clones within sites, however, differed. significantly' in leaf ‘wate ’jpotential. The interaction between water status and clones was also significant. Again, the interaction involving site origin of clones and water status was analyzed further. Its result indicated that the interaction was not significant in this regard. Nonetheless, the leaf water potential for clones from the dry site tended to be higher than those from the wet site. 86 Table 27. Analysis of variance for leaf water potential in the greenhouse study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 1157.6860 78.14 ** Error (a) 4 14.8154 Nitrogen levels (N) 2 3.6213 1.25 ns W X N 4 7.0364 2.44 ns Error (b) 12 2.8789 Clones (C) 7 16.9418 9.69 ** Between sites 1 44.6901 3.67 ns Within sites 6 12.1702 7.05 ** W X C 14 5.7876 3.31 ** N x C 14 1.8302 1.05 ns W X N X C 28 2.0747 1.19 ns Error (c) 126 1.7480 CV(a)= 30.50 %; CV(b)= 13.45 %; CV(c)= 10.48 %. ** = significant at the 1 % level. ns = not significant. 87 -2.0 A a if 3 -1.5 P b .7! E .2 a -1 0 - 1.: O 2' E-os— _l 0.0 SW8 MWS NWS Water Statue Figure 24. The effect of water status on leaf water potential after 2.5 months in the greenhouse study. Any means with the same letter are not significantly different by Duncan’ 5 multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 88 Stomatal Conductance Because of the heterogeneity of 'variance, log transformations for stomatal conductance data were required before conducting an analysis of variance. Log transformation was used. There were highly significant differences in stomatal conductance between the water stress treatments, as well as between nitrogen levels (Table 28). However, the effects of water status and nitrogen levels on stomatal conductance was inter-dependent. The interaction effect was primarily linear (Figure 25). Added nitrogen had little influence upon stomatal conductance when plants were under water stress. In contrast, nitrogen increased stomatal conductance under well-watered conditions. Neither ‘the clones Ibetween sites. nor' clones ‘within sites terms contributed significantly to the observed variance in stomatal conductance. However, plants from the dry site had a slightly higher stomatal conductance than those from the wet site. The weighted mean stomatal conductances for plants from the dry site and wet site were 0.21 and 0.20 mol m—Zs-l, respectively. No interaction involving clones was detected. Transpiration The analysis of variance for transpiration rate was carried out on log-transformed data. The analysis of variance in Table 29 shows that water status and nitrogen levels influenced transpiration rate significantly. However, 89 Table 28 . Analysis of variance for stomatal conductance in the greenhouse study Source of variation Df. Mean square F value Replicates 2 --------- Water status (W) 2 64.8919 109.82 ** Error (a) 4 0.5910 Nitrogen levels (N) 2 0.7236 5.80 ** Linear . 1 1.3843 11.10 ** Quadratic 1 0.0629 0.50 ns W X N 4 0.4073 3.27 * Linear 1 0.6316 5.06 * Quadratic 1 0.1831 1.47 ns Error (b) 12 0.1247 Clones (C) 7 0.2240 1.50 ns Between sites 1 0.1017 0.68 ns Within sites 6 0.2444 1.64 ns W x C 14 0.1957 1.31 ns N x C 14 0.1591 1.07 ns w x N x C ' 28 0.1463 0.98 ns Error (c) 126 0.1489 CV(a)= 25.59 %; CV(b)= 11.76 %; CV(c)= 12.85 %. *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 9O 08 .- NW8 I 05 Y-O.4O1OO.OOX Stomatal conductance (mol/m’le) 10 ch 0.2 _ v-o.1:ooo.001x i E i wwe I 1 —4-ews ;' ed :0 v-o.1ooo.oozx o o J l l 0 2.25 4.50 Nitrogen level (gr/plant) Figure 25. The effect of water status on stomatal conductance after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % leNel. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 91 Table 29. Analysis of variance for transpiration rate in the greenhouse study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 16.4883 139.73 ** Error (a) 4 0.1180 Nitrogen levels (N) 2 0.0864 4.67 * Linear 1 0.1436 7.76 * Quadratic 1 0.0292 1.58 ns W X N 4 0.0793 4.29 * Linear 2 0.1041 5.63 * Quadratic 2 0.0546 2.95 ns Error (b) 12 0.0185 Clones (C) 7 0.0020 0.05 ns Between sites 1 0.0501 1.28 ns Within sites 6 0.1412 3.61 ns W X C 14 0.0236 0.60 ns N X C 14 0.0127 0.32 ns W X N x C 28 0.0148 0.38 ns Error (c) 126 0.0391 CV(a)= 20.56 %; CV(b)= 8.22 %; CV(c)= 11.83 %. *,**=significant at the 5 and 1 % levels, respectively. ns = not significant. 92 the effect of water status was not independent of the effect of nitrogen. This interaction was mainly due to the linear component, since the quadratic response was not significant. Figure 26 indicates that nitrogen levels did not have a great impact upon transpiration rate when the plants were under water deficit. The addition of nitrogen, on the other hand, resulted in increased transpiration rates when plants were well-watered. As with stomatal conductance, no significant effect on transpiration was observed among clones between sites, nor among clones within sites. The interaction involving clones was also of little importance. Photosynthesis The photosynthetic rate data were also log-transformed due to the heterogeneity of variance. Among the many factors involved in the analysis, apparently only water status influenced photosynthetic rate significantly (Table 30). All water status treatments differed significantly from one another (Figure 27), with the SWS treatment showing no net photosynthesis and the MWS treatment showing negligible rates. Correlations Between Characteristics A correlation analysis between physiological traits was conducted (Table 31) . All the characteristics analyzed were strongly intercorrelated. Leaf water potential was 93 12 a 3 1o _ - NW3 a E b >_ - Y-8.47200.343X o .. .§ 8 .2 2 6r- : v-ueoonax O c :1 c 5 ._tuws E _ WI 1 _ E 4 c c E ewe g Y-3.89200.002X I: 2 — o l ‘ l l 0 225 450 Nitrogen level (gr/plant) Figure 26. The effect of water status on transpiration rate after 2.5 months, as affected by nitrogen levels in the greenhouse study. Any means with the same letter are not significantly different by Duncan’s multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 94 Table 30. Analysis of variance for photosynthetic rate in the greenhouse study Source of variation Df. Mean square F value Replicates 2 -------- Water status (W) 2 628.5542 58.32 ** Error (a) 4 10.7782 Nitrogen levels (N) 2 6.7379 2.41 ns W X N 4 4.7412 2.80 ns Error (b) 12 2.7980 Clones (C) 7 3.8863 0.60 ns Between sites 1 0.3057 0.05 ns Within sites 6 4.4830 0.69 ns W X C 14 8.1016 1.25 ns N x C 14 6.6128 1.02 ns W X N X C 28 7.2619 1.12 ns Error (C) 126 6.4789 CV(a)= 39.72 %; CV(b)= 20.25 %; CV(c)= 30.80 %. ** = significant at the 1 % level. ns = not significant. 95 ... 10 O “\ E h c o 8 ‘- (J 3 S a 5” IE: 2 1'; 4L 2 3 5 s. 2- O .2 .2 a b SWS MWS NWS Water statue ' Figure 27. The effect of water status on photosynthetic rate after 2.5 months in the greenhouse study. Any means with the same letter are not significantly different by Duncan's multiple range test at the 5 % level. SWS= severe water stress; MWS= moderate water stress; NWS= no water stress. 96 Table 31. Correlations between physiological characteristics in the greenhouse study ** Leaf water Stomatal Trans. Photosyn. potential conduct. rate Leaf water -.74 -0.81 -0.79 potential Stomatal 0.87 0.87 conductance Transpiration 0.90 ** = all correlations are significant at the 1 % level. 97 negatively correlated .with stomatal conductance, transpiration, and photosynthesis. As leaf water potential became more negative, stomatal conductance, transpiration, and photosynthetic rates decreased. Stomatal conductance, on the other hand, correlated positively with transpiration and photosynthesis. As stomatal conductance increased, transpiration and photosynthetic rates were enhanced. Variance Component Estimation The variance component attributable to cdones between sites was smaller in comparison to the variance component accounted for by clones within sites for height, diameter, shoot biomass, root biomass, and root-shoot ratio (Table 32) . Despite the fact that clones between sites were not the major source of variation, their contribution to the total variation was high for height and shoot biomass. For physiological traits, the contribution of clones between sites, as well as clones within sites seemed to be of less significance. The major source of variation for these characteristics was residual error. Stability Assessment The stability or plasticity parameters were assessed only for the characteristics which showed site origin of clones-treatment interactions. Since the only treatment that interacted with clones was water status, the stability parameters referred to this treatment. Clones from the dry Table 32. 98 Variance component estimation for characteristics measured in the greenhouse study Variance component ( % of total) Characteristics 0283 ozws ozwc 02NC GZWNC 023 Height 13.5 27.4 10.1 1.2 1.6 46.3 Diameter 3.0 30.15 6.5 1.8 9.5 49.1 Leaf area 0 21.0 0.8 1.6 11.6 65.0 SLW 0 11.8 3.6 4.9 17.3 62.5 Shoot biomass 9.7 19.7 9.4 1.6 0 59.7 Root biomass 1.1 24.6 15.0 1.4 4.2 53.8 Root-shoot 3.1 10.9 12.0 0 13.4 60.7 ratio Leaf water 10.0 12.3 15.0 0.3 3.6 58.2 potential Stomatal 0 2.2 3.3 0.7 0 93.8 conductance Transpiration 1.2 8.6 0 0 0 90.3 Photosynthesis 0 0 2.6 0.2 3.8 93.4 99 Table 33. Stability parameters of two populations of eastern cottonwood across three soil water levels Characteristics Population b Sd2 Height Dry site 0.93 0.1146 Wet site 1.07 0.1949 Diameter Dry site 0.90 79.0937 Wet site 1.10 113.9021 Shoot biomass Dry site 0.90 0.0470 Wet site 1.10 0.0706 y 100 site had smaller b and 8&2 values than those from the wet site for all characteristics examined. Clones from the wet site were more responsive to favorable water conditions than those from the dry site, which is indicated by the higher b values. In addition, clones from the wet site were more plastic in response to changes in water status in comparison with those from the dry site as shown by higher values of 862 . DISCUSSION Water stress has a profound effect upon plant growth. Cell enlargement and differentiation are all reduced by water deficit with the obvious result being a reduction in plant size (Kramer, 1983; Kramer and Kozlowski, 1979). Despite the fact that the severe water stress (SWS) treatment in experiment one only occurred during the first three months after its establishment (Table 2), the resulting water deficit reduced height and diameter growth of eastern cottonwood profoundly. During the first three months after planting, the highest recorded soil water potential in the SWS treatment was -0.07 MPa. Thereafter plants in all water treatments were essentially in well- watered conditions due to the adequate amount of rainfall (Table 2). Supplemental irrigation was added quite rarely for the moderate water stress (MWS) and no water stress (NWS) treatments from this period of time until the end of the growing season. MWS had apparently no significant effect on reducing height and diameter growth. The results of experiment two were similar to those of experiment one. Height and diameter growth were affected by water deficit in a significant way. In this experiment, plants in the SWS treatment experienced quite severe water deficit until the end of July (Table 7). As a matter of fact the drought this year was the worst drought to occur in Michigan in recent years (United States, Department of 101 102 Commerce, 1988). The lowest recorded soil water potential in the SWS treatment was —0.1 MPa. Again, from August until the end of the experiment, all water treatments were in well-watered conditions. Additional irrigation for the MWS and NWS treatments was only rarely needed. Water deficit also had significant effects upon leaf area and SLW. Leaf enlargement is more severely affected by water deficit than other physiological processes such as photosynthesis and transpiration (Boyer, 1976). This field plot experiment clearly showed that water deficit had a dramatic impact on leaf area of eastern cottonwood. Water stress reduced leaf area even with moderate water stress. SLW was also very sensitive to changes of water status. It has long been documented that soil water status affects leaf water potential, stomatal conductance, transpiration, and photosynthesis (Hsiao, 1973; Boyer, 1976; Farquar and Sharkey, 1982; Schulze, 1986). The results of this field plot study substantiated the previous findings, except for photosynthesis. leaf water potential, stomatal conductance, and. transpiration. 'were all reduced. by decreasing water availability. Stomatal conductance has long been recognized as a key variable influencing leaf‘ gas exchange: through. its regulation of water vapor and C02 diffusion. Researchers have attempted to correlate stomatal conductance to leaf water potential. However, there is accumulating evidence that under field conditions (mild water stress), the 103 decrease in stomatal conductance is not associated with a change in leaf water potential. For example, Osonubi (1985) found that substantial decreases in stomatal conductance of COWpea (yigna unguiculata L.) were independent of leaf water potential. Blackman and Davies (1985) also found an independence of stomatal conductance with leaf water potential in maize plants subjected to soil drying. Cock gt a1. (1985) observed decreased stomatal conductance in unirrigated cassava (Manihot esculenta Crantz.) plants even though the leaf water potential was slightly higher than that of well-irrigated plants. I In the field plot study the correlation between leaf water potential and stomatal conductance was very low (- 0.15), indicating that stomatal conductance was independent of leaf water potential, even though both stomatal conductance and leaf water potential decreased as water deficit increased. A number of researchers have suggested that stomatal conductance decreases only after a threshold of leaf water potential is attained (Hsiao and Acevedo, 1974; Turner, 1974, Baldocchi g; ,al., 1985; Teskey and Hinckley, 1986). The correlation between stomatal conductance and photosynthesis found in the field plot study was moderate (0.45). water status also had little effect on photosynthesis in this regard. Farquhar and Sharkey (1982) discussed this matter extensively, and suggested that stomata generally function to minimize water loss, while 104 only marginally limiting photosynthesis. In addition, stomatal closure is not the only mechanism by which water deficit influences photosynthesis (Boyer, 1971; Hsiao, 1973; Jones, 1985; Nicolodi gt gt, 1988; Teskey gt gt, 1986). The correlation between leaf water potential and photosynthesis was also low (-0.15). Some researchers have observed this phenomenon and proposed that photosynthesis responds to soil water depletion independent of leaf water status alterations, presumably through a still unknown signal coming from the roots (Passioura, 1980; Bates and Hall, 1982; Turner gt ,g;., 1985; Schulze, 1986). In conifers Grieu gt g;. (1988) found that increasing soil drought affected stomatal conductance and mesophyll photosynthesis independently. The result with relatively mild water deficit in this field plot experiment was in agreement with the previous findings. Water stress had a significant effect on stomatal conductance, but had little influence on photosynthesis. It is surprising that the effect of nitrogen fertilization on growth was not significant, even though unfertilized plants grew less. It has been reported that nitrogen fertilization improved the growth of eastern cottonwood in the field. For example, Blackmon and White (1972) found that applying nitrogen fertilizer (150 lb/acre) to a six-year old eastern cottonwood increased diameter, basal area, and volume growth by 200 %. Curlin (1967) observed that a large increase in growth at one-year of age 105 resulted from nitrogen fertilization. Blackmon (1977) reported that eastern cottonwood’s response to nitrogen was related to age. When nitrogen fertilizer was applied to a plantation at age four (336 kg N/ha) diameter growth increased by 33 % over unfertilized treatments. Fertilizing at ages two and three resulted in no responses. Nitrogen fertilization was reported to increase photosynthetic capacity in a Douglas-fir stand (Brix, 1971; Brix, 1972). However, nitrogen fertilization has also been reported to have no significant effect on photosynthetic rate (Brix and Ebell, 1969; Helms, 1964). Since the nitrogen content in the soil of the field plot in this study was relatively low (Table 1), it was expected that the added nitrogen would result in increased growth rates. This discrepancy could in part be due to the high coefficient of variation in the sub-plot that resulted in an inability to detect any significant differences between nitrogen levels. The plantation might also be merely unresponsive to the added nitrogen at such a young age, as was also observed by Blackmon (1977). At this young age competition among plants was not a significant factor, and the nitrogen demand might be met by indigenous soil nitrogen. The greenhouse study was designed to be similar to the field experiment, but with more severe water stress in the SWS and MWS treatments. As expected, water stress affected all growth and physiological characteristics measured. The 106 water deficit occurring in this experiment reduced the growth and inhibited physiological processes of eastern cottonwood in a dramatic fashion. This is not surprising since a lack of moisture inhibits enzyme activities, affects membrane conformation, and influences all other physiological processes, the end result being a decrease in growth (Teskey and Hinckely, 1986; Kramer, 1983; Kramer and Kozlowski, 1979). In the greenhouse study most of the effects of water status were not independent, but rather varied according to the rate of nitrogen applied. Nitrogen fertilization was of little significance in influencing growth and physiological processes under water stress, but had considerable effects under favorable water conditions. Plants respond to nitrogen fertilization depending on other environmental factors, such as the availability of water supply (Kramer and Kozlowski, 1979). Water serves as the medium for diffusion and mass flow of nutrients to plant roots. Nutrient movement may be seriously limited in soils with a low moisture content, since that reduces hydraulic conductivity and thereby mass flow and pathways for nutrient diffusion (Ballard and Cole, 1974; Viets, 1972). Moisture deficiency affects nitrogen metabolism of plants both directly and indirectly in many complex ways, often resulting in a moisture-nitrogen interaction in growth. The direct effects are through inhibition of the 107 biosynthesis of nitrogen-dependent compounds such as protein. Indirect effects include reduction in nitrogen uptake, because nitrogen cannot be absorbed from dry soil. (Brix, 1979; Hsiao, 1973; Naylor; 1972). An interaction of moisture and nutrients in tree growth depends on whether moisture changes affect the relationship of nutrient supply and demand in plants. It is conceivable that demand is reduced more than availability, thus improving the mineral nutrient status of plants. Also, nutrient storage within plants may overcome brief limiting periods of nutrient uptake, providing a clear case of moisture-fertilizer interaction in which fertilization would affect growth under favorable soil moisture but not when moisture becomes deficient (Brix, 1979). The results of the greenhouse study basically verified the finding in other fertilization studies. Growth and physiological processes are more affected by nitrogen when plants are under well-watered conditions, except that root biomass declined as rdtxogen levels increased irrespective of water status. Water stress profoundly affected shoot biomass and root biomass in this study. Root-shoot ratios for plants under water stress were higher than those under favorable soil moisture conditions. Root growth is oftentimes less affected by water stress than shoot growth, resulting in an increase in root-shoot ratios (Kramer, 1983). This increase in root-shoot ratio observed under water deficit is 108 generally believed to be due to- greater water stress developing in the shoot (Kramer, 1983). A higher root-shoot ratio under water stress was also reported in eastern cottonwood (Farmer, 1970 b) and hybrid poplar (Mazzoleni, 1985). It has also been demonstrated in many species that shoot growth is affected more by water stress than root growth, resulting in a higher root-shoot ratio (Kramer, 1983; Kramer and Kozlowski, 1979). As :mentioned. before, root growth. declined. with. the increase of nitrogen rates regardless of water status. Root—shoot ratios tended to decline as the rate of nitrogen applied increased, particularly under favorable moisture conditions. This was shown by the negative slopes of regression lines. It has been known that fertilization, particularly heavy fertilization, causes a reduction in carbohydrates and an increase in nutrient content in the plant. Plants respond by producing proportionally more shoot and less root materials, resulting in a low root-shoot ratio. By contrast, a lack of nutrient (nitrogen) availability leads to low concentrations of limiting nutrients and. to accumulation. of‘ carbohydrates. Plants respond by increasing proportional allocation to root growth, resulting in a higher root-shoot ratio (Bloom gt _;., 1985). In many forest tree species nutrient availability affects carbon allocation, with more carbon being 109 allocated to the root in poor nutrient soil (Keyes and Grier, 1981; Grier gt _t., 1981). As with the field plot study, water stress reduced leaf area in the greenhouse study. Unlike the field experiment, however, nitrogen fertilization increased leaf area significantly. The effect of nitrogen was more profound under favorable soil moisture conditions than under water deficit. The water status and nitrogen treatments also had a significant influence on SLW. Plants lacking water and nitrogen had a lower SLW than those under favorable conditions. An inverse relationship between SLW and nitrogen availability has been observed in other species (Gulmon and Chu, 1981; Longstreth and Nobel, 1980; Osman gt gt., 1977). A lack of available nitrogen has been related to a reduction in the proportions of sugar and protein. Therefore, the remaining insoluble materials (cell wall materials) constitute a much greater portion of the dry matter in nitrogen-deficient plants (Shimsi, 1970; Radin and Parker, 1979). In ‘the greenhouse study' water stress inhibited all physiological processes. The effect of water deficit upon stomatal conductance, transpiration, and photosynthesis in eastern cottonwood has been previously documented. Under water stress, stomatal conductance, transpiration, and photosynthesis are reported to decrease (Bonner, 1967; Farmer, 1969; Kelliher and Tauer, 1980; Regehr gt g_., 1975; Scarascia-Mugnozza gt gt, 1986; Schulte, 1985). 110 Water deficit can decrease photosynthesis either by decreasing conductance to diffusion of carbon dioxide, or by affecting the photosynthetic and respiratory mechanisms (Boyer, 1976; Hinckley gt _a_t. 1981). Changes in stomatal conductance during water stress have a major impact upon photosynthesis. It has been reported in a number of studies that there is a strong correlation between stomatal conductance and photosynthesis during severe drought. Generally, as leaf water potential becomes more negative, both stomatal conductance and photosynthetic rate decrease (Boyer, 1976). In eastern cottonwood, transpiration at various degrees of water stress was found to be primarily controlled by stomatal conductance (Kelliher, gt a_l., 1980). Regehr gt gt.(1975) found that stomatal conductance and transpiration of eastern cottonwood paralleled the decline in net photosynthetic rate. The result of the greenhouse study seems in agreement with those found in other experiments. Leaf water potential, stomatal conductance, transpiration, and photosynthesis were strongly intercorrelated. However, the complete lack of photosynthesis in the SWS treatment indicates that it was due to more than just stomatal closure. The photosynthetic process itself was likely affected (Boyer, 1971; Hsiao, 1973; Jone, 1985; Teskey g gt., 1986). Despite the fact that stomatal conductance is believed to be a factor limiting photosynthetic rate, some 111 researchers have suggested otherwise (Farquar and Sharkey, 1982) . Even though there is a strong correlation between stomatal conductance and photosynthetic rate, it appears more likely to represent an adjustment of stomatal conductance to match the intrinsic photosynthetic rate rather than ,a causal relationship. According to a theoretical. model for* stomatal control, the stomata. may minimize daily transpiration for a given daily carbon gain. In other words, if a certain amount of water can be acquired for transpiration, stomata should act to maximize photosynthesis within this constraint (Cowan and Farquhar, 1977). Under favorable soil moisture conditions nitrogen fertilization seemed to enhance stomatal conductance and transpiration. Nitrogen deficiency has been reported to reduce stomatal conductance in :maize (Ryle and. Hesketh, 1969), cotton (Longstreth and Nobel, 1980; Ryle and Hesketh, 1969, sugar beet (Nevins and Loomis, 1970) and rice (Yoshida and Coronel, 1976). Childers and Cowart (1935) found a 30 % decrease in the rate of transpiration of nitrogen-deficient apple leaves. This reduction in transpiration occurred despite greater pore area per ‘unit leaf area in 'the nitrogen-deficient leaves. The leaves of nitrogen-deficient bean (Phaseoulus vulgaris L) transpired less than nitrogen-supplied plants (Shimsi, 1970). 112 Photosynthetic rate was enhanced by increasing the nitrogen level in pine seedlings (van den Driessche and Wareing, 1966), Douglas-fir (Brix, 1971, 1972) and cotton (Wong, 1979). In other studies, however, no apparent increase of photosynthetic rate was reported. Kozel gt gt. (1983), for example, found that wheat plants having nitrogen deficient chloroplasts are photosynthetically at least as active as those with normal chloroplasts. A study with maize plants grown with different nitrogen concentrations also indicated no significant differences in photosynthetic rate. In this later study the photosynthetic rate was low due to the low level of illumination during the growth period, which could possibly explain the lack of significant differences between nitrogen concentrations (Fernandez and Manero, 1983; Bouma, 1970). It is not yet clear as to how nitrogen affects stomatal conductance, transpiration and photosynthesis, but it is likely to be mediated through chlorophyll content of the leaves. An accumulating evidence indicates that there is an intimate relationship between chlorophyll content of the leaf and stomatal function. Nitrogen deficiency causes the following chain of effects: low chlorophyll content, low photosynthetic rate, high concentration of C02 in mesophyll spaces, and stomatal closure (Shimsi, 1970). Disruption of nitrogen nutrition has been known to affect all nitrogen- dependent plant constituents such as chlorophyll (Kramer and Kozlowski, 1979). 113 In the greenhouse experiment, the lack of significant differences in photosynthetic rates between the different nitrogen treatments may have been due to high temperatures that resulted from the lack of air movement (Table 19). During the measurement of photosynthesis, temperature was around 33 °C. High temperatures have direct effects on the synthesis and activity of enzymes, and have indirect effects through changing stomatal conductance for 002. Stomata tend to close with increasing temperature, the closure results from a stomatal response to an increased vapor deficit (Kramer' and. Kozlowski, 1979; Berry’ and Bojrkman, 1980). Under the same well-watered condition, photosynthetic rates of the greenhouse-grown plants were considerably lower than those of field-grown plants. The low photosynthetic rates of the greenhouse-grown plants may also have been due to low SLWs, which resulted from the high temperature and relatively low level of illumination during leaf development. The average SLW of the field-grown plants under well— watered conditions was 1.13 times greater than that of the greenhouse-grown plants under the same soil moisture condition. This difference was probably due to low levels of illumination and higher temperatures in the greenhouse environment during leaf development. PARS measured in August were 2510 and 1165 umol m-zs-1 for the field and greenhouse environments, respectively. The average daily 114 temperatures in the greenhouse were higher than those in the field (See Table 3 and 19). It has been well documented that SLW or leaf thickness is influenced by light intensity' and temperature. IFor example, Nelson and Ehlers (1984) found that SLW of populus hybrids grown in the field was 1.5 to 1.8 times than that of greenhouse-grown plants. This was primarily due to the greater average PARs in the field. The thickness of Populus x euramericana leaves increased about 30 % when the temperature was decreased from 25 to 16 ° C (Pieter 1974 cited by Nelson and Ehlers, 1984). The SLW of Festuca arundinaca decreased as much as 13 % due to an increase in temperature from 10 to 25 ° C (Nelson gt _t., 1978). Both thickness and SLW decreased as growth temperatures were raised for Glycine _gg and Gossypium hirsutum (van volkenburg and Davis, 1977). SLW is often associated with photosynthetic rate (Nelson. and Ehlers, 1984; McMillan. and. McClendon, 1983; Chabot gt gt. 1979). When the data of well-watered treatments from both field and greenhouse experiments were pooled, the correlation coefficient for the relationship between photosynthetic rate and SLW was 0.69. Nelson and Ehlers (1984) found slightly higher coefficients of correlation for hybrid poplar. The results of this experiment indicated that one of the major effects of the growth environment on photosynthetic rate is likely a result of changes in leaf thickness or SLW. 115 The SWS and MWS treatments of the greenhouse study were more severe than the corresponding treatments in the field. In the greenhouse experiment, GE-interactions for several growth and physiological characteristics were detected. Despite the fact that GE-interactions were significant, plants from the dry site grew more slowly than those from the wet site, indicating that the nature of interaction was not changes in direction, but rather changes in magnitude of differences. In addition, the variance components of GE- interactions appeared to be small in relation to the total variation. Similar results were reported by Farmer (1970b), who found. that. clone-moisture interactions 'were significant for growth, but the variance components for GE- interaction were relatively small. In another study, clone- moisture interaction for growth was not detected (Broadfoot and Farmer, 1969). In both field and greenhouse studies, however, GE- interactions with regard to nitrogen fertilization were not of great importance. This finding was in conflict with that reported by Curlin (1967), who found that the clone-nitrogen interaction was very strong. It is interesting to note that plants collected from the wet site consistently grew faster than those from the dry site in both field plot and greenhouse studies, even though in the greenhouse study the differences between the two populations were statistically significant only in the well-watered condition. 116 Compared with the plants from wet habitats, plants growing in dry habitats usually are smaller and their leaves are usually smaller and thicker. In addition, they tend to have deeper root systems and lower transpiration rates (Kramer and Kozlowski, 1983). A number of studies that have been done in forest tree species indicate that there are genetic differences between plants growing in dry habitats and those growing in wet habitats with regard to growth and physiological characteristics. For example, Ptggg gp. from xeric locations had a lower rate of shoot growth than those from mesic habitats (Venator,1976; Wells and wakely, 1966; Woesner, 1972a, 1972b; Wright and Bull, 1963). Transpiration of Douglas-fir from dry sites was found to be lower than that from wet site (Zavitkovski and Ferrell, 1968). On the other hand, Feret (1982) found that the growth of gtggg ponderosa grown under water stress did not differ significantly between xeric site type and mesic site type. Photosynthetic rate of Douglas-fir from dry sites was not significantly different from that of wet sites (Zavitkovski and Ferrell, 1968). Kelliher and Tauer (1980) found that there were differences between clones of eastern cottonwood from dry sites and those from wet sites for growth and stomatal conductance when they were grown under water stress. Height and stomatal conductance were higher for clones from the dry site than those from the wet site. 117 In the present study plants from the dry site grew more slowly than those from the wet site, particularly in the field plot experiment. It appears that there is genetic differentiation between the two populations with regard to growth. The existence of genetic differentiation between these populations is substantiated by the amount of variation attributable to clones between sites. For example, in experiment two of the field plot study, more than 45 % and more than 30 % of the total phenotypic variation for height and diameter, respectively, were accounted for by between population differences. By contrast, SLW and physiological traits were hardly affected at all by the site origin of clones. In other words, the original site of the population had little or no effect on the variation of these traits. Environmental factors, on the other hand, were the major cause of the existing variation. These traits may be of less significance for the adaptation of eastern cottonwood in the habitats being studied. Plants from the dry site appear to be better adapted for slower growth than those from the wet site, due to the lack of water and nutrient availabilities. In sand dune environments plants always experience water deficit, particularly during dry weather conditions. The primary source of water for plants in dune habitats is from direct rainfall, but because the mechanical nature of the soils restricts their ability to retain water, the major factors 118 limiting their growth in these situations are the water- holding capacity of the soil and the soil resistance to surface evaporation (Ranwell, 1972). Furthermore, dune soil is very poor in nutrients (McGee gt gt., 1981; Grime, 1977). By contrast, the floodplain from which the wet-site plants were collected possesses different soil conditions. In floodplains plants never experience severe water stress and floodplain soils are nutrient rich. Eastern cottonwood found in this habitat grows very rapidly (McGee e_t a_l., 1981). McGee gt gt. (1981) reported somewhat similar results. They found the existence of genetic differentiation between populations of eastern cottonwood originated from sand dune, strip mine and floodplain for several growth characteristics such as height, root weight, specific leaf area, and shoot root ratio. They also found that the pattern of transpiration and photosynthesis as influenced by water deficit was different between the three populations studied. Slower growth and higher root-shoot ratios for plants from the dry site seem to have adaptive significance. Sand- dune populations experience drought and nutrient limitation more frequently than floodplain plants and selection favors slower growth and higher root-shoot ratios for survival in the sand dune (McGee gt gt, 1981). It has been well-documented that plants adapted to poor habitats have slow growth rates (Chapin, 1980; Grime and Hunt, 1975) . Slow growth rates may enable the plant to 119 survive between occasional pulses of nutrient supply, whereas more rapidly growing plants may exhaust their nutrient reserves, leading to a complex of nutrient deficiency symptoms (Bradshaw, 1969). In this present study it was found that there was no significant difference in root-shoot ratio between dry site and wet site plants. The importance of high root-shoot ratios for plants growing in poor habitats has been elaborated by Chapin (1980). Plants from infertile habitats maximize nutrient intake to a greater extent through high root-shoot ratios than through high root-absorption capacities. Many plants occurring in poor habitats are found to have high root-shoot ratios as a response, in part, to a lack of nutrient availability. 0n the other hand, plants from nutrient-rich habitats show considerable phenotypic plasticity in root-shoot ratio and generally have higher ratios at low availability and low ratios at high availability than do plants from a poor-nutrient habitat. In this study it was found that plants from the wet site were more plastic than those from the dry site for characteristics such as height, diameter, and shoot biomass. Higher growth and higher plasticity values for plants from the wet site indicate that this population tends to adapt to more favorable conditions, in the present case to more favorable soil moisture conditions. According to Bradshaw (1965) phenotypic plasticity is one of the mechanisms by 120 which a species or genotype can maximize fitness. Phenotypic plasticity could itself be under genetic control. The result of this experiment has an important implication from a practical standpoint. Selection of plant materials for improvement programs or plantation establishment in. marginal lands needs to be undertaken cautiously. Populations from what we classify as dry sites may not be as drought resistant as expected. In addition, plants from dry sites may not grow better in marginal sites than those from wet sites. Plants from dry sites seem unresponsive to changes in more favorable conditions. Ideally, genotypes desirable for plantations are those having a high growth rate and a low' plasticity across different.lenvironmental. conditions. (Shelbourne, 1972). As far as the populations used in this study are concerned, there was no evidence that plants collected from the dry site outperformed those from the wet site under any water conditions. LIST OF REFERENCES LIST OF REFERENCES Abou-El-Fittouh, H. A., J. 0. Rawling, and P. A. Miller. 1969. Classification of environments to control genotype by environment interactions with an application to cotton. Crop Science. 9:135-140. Ballard, T. M. and D. W. Cole. 1974. Transport of nutrients to tree root system. Can. J. For. Res. 4:563-565. Baldocchi, D. D., S. B. Verma, and N. J. Rosenberg. 1985. Water use efficiency in soybean field. Influence of plant water stress. Agri. For. Meteorol. 34:53-65. Bates, L. M. and A. E. Hall. 1981. Stomatal closure with soil water depletion not associated with changes in bulk leaf water status. 0ecologia 50:62-65. Blackman, P. G. and W. G. Davies. 1985. Root to shoot communication in maize plants of the effects of soil drying. J. Exp. Bot. 36:39-48. Blackmon, B. G. 1977. Cottonwood response to nitrogen related to plantation age and site. U.S. Forest Service Research Note 80-229, 3 pp. Blackmon, B. G. and E. H. White. 1972. Nitrogen fertilization increases cottonwood growth on old-field soil. U.S. Forest Service Research Note S0-143, 5 pp. Bloom, A. J., F. S. Chapin III, and H. A. Mooney. 1985. Resource limitation in plants - an economy analogy. Ann. Rev. Ecol.Syst. 16:363-392. Bonner, F. T. 1967. Response of 1 year old cottonwood to increasing soil moisture tension. U.S. Forest Service Research Note S0-56, 3 pp. Bouma, D. 1970. Effects of nitrogen nutrition on leaf expansion and photosynthesis of Trifolium subterraneum L. 1. Comparison between different levels of nitrogen supply. Ann. Bot. 34:1131-1142. Boyer, J. S. 1976. Water deficit and photosynthesis. In Water deficit and plant growth. Vol. 4. Ed. T. T. Kozlowski. Academic Press Inc., New York, pp. 153-190. 1971. Nonstomatal inhibition of photosynthesis in sunflower at low leaf water potentials and high light intensities. Plant Physiol.48:532-536. 121 122 Berry, J. and O. Bjorkman. 1980. Photosynthetic response and adaptation to temperature in higher plants. Ann. Rev. Plant. Physiol. 31:491-543. Bradshaw, A. D. 1969. An ecologist’s viewpoint. In Ecological aspects of the mineral nutrition of plant. Ed. I. H. Rorison. Blackwel, Oxford, pp. 415-427. 1965. Evolutionary significance of phenotypic plasticity in plants. Advances in Genetics 13:115-156. Bridgwater, F. E. 1972. Multiple trait selection in a population of eastern cottonwood . Unpublished Ph . D Dissertation, Oklahoma State University, 72 pp. Broadfoot, W. M. and R. E. Farmer, Jr. 1969. Genotype and moisture supply influence nutrient content of eastern cottonwood foliage. Forest Science 15: 46-48. Brix, H. 1979. Moisture-nutrient interrelationship. Proc. Forest Fertilization Conf. Union, Washington, pp. 48- 52. 1972. Nitrogen fertilization and water effects on photosynthesis and earlywood-latewood production in Douglas-fir. Can. J. For Res. 2:467-478. 1971. Effect of nitrogen fertilization on photosynthesis and respiration in Douglas fir. Forest Science 4:407-414. Brix, H. and L. F. Ebell. 1969. Effects of nitrogen fertilization on growth, leaf area, and photosynthetic rate in Douglas-fir. Forest Science 15:189-196. Chabot, B. F., T. W. Jurik, and J. F. Chabot. 1979. Influence of instaneous and integrated light-flux density on leaf anatomy and photosynthesis. Am. J. Bot. 66:940-945. Chapin, F. S. III. 1985. Adaptation and physiological responses of wild plants to nutrient stress. In Genetic aspects of plant mineral nutrition. Ed. W. H. Gabelmand and B. C.) Loughman. Martinus Nijhoft Pub. Boston, pp. 15-25. 1980. The mineral nutrition of wild plants. Ann. Rev. Ecol. Syst. 11:233-260. Childers, N. F. and F. F. Cowart. 1935. The photosynthesis, transpiration and stomata of apple leaves as affected by certain nutrient deficiencies. Proc. Amer. Soc. Hort. Sci. 33:160—163. 123 Cock, J. H., M. C. M. Porto, and M. A. El-Sharkawy. 1985. Water use efficiency of cassava. III. Influence of air humidity and water stress on gas exchange of field grown cassava. Crop Science 25:265-272. Comstock, R. E. and R. H. Moll. 1963. Genotype-environment interactions. In Statistical genetics and plant breeding. Eds. W. D. Hanson and H. F. Robinson. NAS- NRC Publ. No. 912, pp. 161-196. Cowan, I. R. and G. D. Farquar. 1977. Stomatal function in relation to leaf metabolism and environment. Symp. Soc. Exp. Biol. 31:471-505. Crossa, J. 1988. A comparison of results obtained with two methods for assessing yield stability. Theor. Appl. Genet. 75:460-467. Crossa, J.,_B. Westcott, and C. Gonzales. 1988. Analyzing yield. stability' of :maize. genotypes using' a spatial model. Theor. Appl. Genet. 75:863-868. Curlin, J. W. 1967. Clonal differences in yield. Responses of Populus deltoides to nitrogen fertilization. Soil Sci. Soc. Amer. Proc. 31: 276-280. Dickmann, D. I. and K. W. Stuart. 1983. The culture of poplars in eastern north America. Department of Forestry, Michigan State University, East Lansing, Michigan, 168 pp. Drew, A. P. and F. A. Bazzaz. 1978. Variation in distribution of assimilate among plant parts in three populations of Pogulus deltoides. Silvae Genetica 27:189-193. Eberhart, S. A. and W. A. Russell. 1966. Stability parameters for comparing varieties. Crop Science 6:36- 40. Farmer, R. E. Jr. 1970 a. Genetic variation among open- pollinated progeny of eastern cottonwood. Silvae Genetica 19:149-151. 1970 b. Variation and inheritance of eastern cottonwood growth and wood properties under two soil moisture regimes. Silvae Genetica 19:5-8. 1969. Transpiration and leaf temperature in eastern cottonwood. Forest Science 15:151-153. 124 Farmer, R. E. Jr. and J. R. Wilcox. 1968. Preliminary testing of eastern cottonwood clones. Theor. Appl. Genet. 38:197-201. 1966. Variation in juvenile growth and wood properties in half-sib cottonwood families. U.S. For. Serv. Res. Pap. NC-6, 4 pp. Farquhar, G. D. and T. D. Sharkey. 1982. Stomatal conductance and photosynthesis. Ann. Rev. Plant Physiol. 33:247-274. Feret, P. 1982. Effect of moisture stress on the growth of Pinus ponderosa Dougl. ex. Laws. seedlings in relation to their field performance. Plant soil 69:177-186. Fernandez, J., M. P. Mazon, and J. Manero. 1983. Effect of nitrogen nutrition on photosynthesis and dark respiration of maize plants. In Photosynthesis and plant productivity. Ed. Helmut Metzner. Wissens Chafttliche Verlagsgesellschaft mbH. Stuttgart, pp. 143-147. Finlay, K. W. and G. N. Wilkinson. 1963. The analysis of adaptation in a plant breeding programme. Australian J. of Agric. Res. 14:742-754 Freeman, G. H. and J. A. Perkins. 1971. Environmental and genotype-environmental components of variability. VIII. Relations between genotypes grown in different environments and measures of the environments. Heredity 27:15-23. Ghaderi, A., M. W. Adams, and A. W. Saether. 1982. Environmental response patterns in commercial classes of common bean (Phaseolus vulgaris L.). Theor. Appl. Genet. 63:17-22. Grier, C. C., A. V. Kristiina, M. R. Keyes, and R. L. Edmonds. 1981. Biomass distribution and above-and below-ground production in young and mature Abies amabilis zone ecosystems of the Washington cascades. Can. J. For. Res. 11:155-166. Grieu, P., J. M. Guehl, and G. Aussenac. 1988. The effect of soil and atmospheric drought on photosynthesis and stomatal control of gas exchange in three coniferous species. Physiol. Plantarum 73:97-104. Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Am. Nat. 111:1169- 1194. 125 Grime, J. P. and R. Hunt. 1975. Relative growth rate: its range and adaptive significance in local flora. J. Ecol. 63:393-422. Gulmon, S. L. and C. C. Chu. 1981. The effects of light and nitrogen on photosynthesis, leaf“ characteristics and dry matter allocation in the chaparral shrub, Diplacus gurantiacus. 0ecologia 49:207-212. Helms, J. A. 1964. Apparent photosynthesis of Douglas-fir in relation to silvicultural treatment. Forest Science 10:432-442. Hinckley, T. B., R. O. Teskey, F. Duhme, and H. Richter 1981. Temperate hardwood forests. In Water deficit and plant growth. Vol. 1. Ed. T. T. Kozlowski. Academic Press Inc., New York, pp. 153-208. Hsiao, T. C. 1973. Plant responses to water stress. Ann. Rev. Plant Physiol. 24:519-570. Hsiao, T. C. and E. Acevedo. 1974. Plant responses to water deficits, water-use efficiency and drought resistance. Agric. Meteorol. 14:59-84. Jones, H. G. 1985. Partitioning stomatal limitations to and non-stomatal photosynthesis . Plant , Cel and Environment 8:95-104. Kelliher, F. M. and C. G. Tauer. 1980. Stomatal resistance and growth of drought-stressed eastern cottonwood from a wet and dry site. Silvae Genetica 29:166-171. Kelliher, F. M., M. B. Kirkhan, and C. G. Stomatal resistance, transpiration drought-stressed eastern cottonwood. 10:447-451. Tauer. 1980. and growth of Can. J. For. Res. Kelly, R. P., D. I. Dickmann, J. W. Wright, and W.A. Lemmien. 1978. Genetic variation of eastern cottonwood planted in southern Michigan. Michigan State University, Agric. Exp. Sta. Res. Rep. No. 362, 6 PP- Kempton, R. 1984. The use of biplots in interpreting variety by environment interactions. J. Agric. Sci. 103:123-135. Keyes, M. R. and C. C. Grier. 1981. Above-and below-ground net production in 40-year-old Douglas-fir stands on low and high productivity sites. Can. J. 604. For. Res. 11:599- 126 Kozel, U., C. Buschmann, and H. K. Lichtenthaler. 1983. Influence of nitrogen deficiency on chlorophyll accumulation and photosynthesis of wheat seedlings grown at high and low light-quanta fluence rates. In Photosynthesis and plant productivity. Ed. Helmut Metzner. Wissens Chafttliche Verlagsgsellschaft mbH, Stuttgart, pp. 122-142. Kramer, P. J. 1983. Water relation of plants. Academic Press Inc., New York, 489 pp. Kramer, P. J. and T. T. Kozlowski. 1979. Physiology of woody plants. Academic Press Inc., New York, 811 pp. Lin, C. S. 1982. Grouping genotypes by cluster method directly related to lgenotype-environment. interaction mean square. Theor. Appl. Genet. 62:277-280. Lin, C. S. and B. Thompson. 1975. An empirical methods of grouping genotypes based on a linear function of the genotype-environment interaction. Heredity 34:255-263. Longstreth, D. J. and P. S. Nobel. 1980. Nutrient influences on leaf photosynthesis . Effects of nitrogen, phosphorus, and potassium for Gossypium hirsutum L. Plant Physiol. 65:541-543. Mazzoleni, S. 1985. Growth and water relations of two populus clones under changing levels of water stress. MS Thesis. Michigan State University, East Lansing, Michigan, 40 pp. McGee, A. B., M. R. Schmierbach, and F. A. Bazzaz. 1981. Photosynthesis and growth in populations of Populus deltoides from contrasting habitats. The American Midland Naturalist 105:305-311. McMillen, G. G. and J. H. McClendon. 1983. Dependence of photosynthetic rates on leaf density thickness in deciduous woody plants grown in sun and shade. Plant Physiol. 72:674-678. Mohn, C. A. and W. K. Randal. 1973. Interaction of cottonwood clones with site and planting year. Can. J. For. Res. 3:329-332. ' Mohn, C. A. and S. S. Pauley. 1969. Early performance of cottonwood seed sources in Minnesota. Minn. For. Res Notes No. 207, 4 pp. Mungomery, V. E., R. Shorter, and D. E. Byth. 1974. Genotype x environment interactions and environmental analysis. I. Pattern analysis. Application to soya bean populations. Australian J. Agric. Res. 25:59-72. 127 Naylor, A. W. 1972. Water deficits and nitrogen metabolism. In Water deficit and plant growth. Vol. 3. Plant responses and control of water balance. Ed. T. T. Kozlowski, Academic Press Inc., New York, pp. 241- 254. Nelson, C. D. and C. G. Tauer. 1987. Genetic variation in juvenile characters of Populus deltoides Bartr. from the Southern Great Plains. Silvae Genetica 36:216-221. Nelson, N. D. and P. Ehlers. 1984. Comparative carbon dioxide exchange for two populus clones grown in growth room, greenhouse, and field environments. Can. J. For. Res. 14:924-932. Nelson, C. J., K. J. Threharne, and J. P. Cooper. 1978 Influence of temperature on leaf growth of diverse populations of tall fescue. Crop Sci. 18:217-220. Nevins, D. J. and R. S. Loomis. 1970. Nitrogen nutrition and photosynthesis in sugar beet (Beta vulgaris L.). Crop Science 10:21-25. Nicolodi, C., A. Massacci, and G. D. Marco. 1988. Water status on net photosynthesis in field-grown alfalfa. Crop Science 28:945-949. Osonubi, O. 1985. Response of coneas (Vigna unguiculata (Io Walp.) to jprogressive soil drought" 0ecologia 66:554-557. Osman, A. M., P. J. Goodman, and J. P. Cooper. 1977. The effects of nitrogen, phosphorus and potassium on rates of growth and photosynthesis of wheat. Photosynthetica 11:66-75. Passioura, J. B. 1980. The transport of water from soil to shoot in wheat seedlings. J. Exp. Bot. 31:333-345. Plasteid, R. L. and L. C. Peterson. 1959. A technique for evaluating the ability of selection to yield consistently in different locations or seasons. Am. Potato J. 36:381-384. Radin, J. W., L. L. Parker. 1979. Water relations of cotton plants under nitrogen deficiency. I. Dependence upon leaf structure. Plant Physiol. 64:495-498. Randall, W. K. 1973. Mississippi cottonwoods outperform local clones near Cairo, Illinois. U.S. Forest Service Research Note SO-164, 3 pp. 128 Randall, W. K. and D. T. Cooper. 1973. Predicted genotypic gain from cottonwood clonal tests. Silvae Genetica 22:165-167. Randall, W. K. and C. A. Mohn. 1969. Clone-site interaction of eastern cottonwood. Proceedings Tenth Southern Conference Forest Tree Improvement, pp.81-91. Randwell, D. S. 1972. Ecology of salt marches and sand dunes. Chapman and Hall, London. Regehr, D. L., F. A. Bazzaz, and W. R. Boggess. 1975. Photosynthesis, transpiration and leaf conductance of Populus deltoides in relation to flooding and drought. Photosynthetica 9:52-61. Ryle, G. J. A. and J. D. Hesketh. 1969. Carbon dioxide uptake in nitrogen-deficient plants. Crop Science 9:451-454. Richard, L. A. 1965. Physical condition of water in soil In Method of soil analysis Part 1. Monograph 9. Am. Soc. Agronom., Madison, Wisconsin, pp. 128-152. Scarascia-Mugnozza, G., T. M. Hinckley, and R. F. Stettler 1986. Evidence for nonstomatal inhibition of net photosynthesis in rapidly dehydrated shoots of Populus. ~Can. J. For. Res. 16:1371-1374. Schulze, E. D. 1986. Carbon dioxide and water vapor exchange in response to drought in the atmosphere and in the soil. Ann. Rev. Plant Physiol. 37:247-274. Shelbourne, C.J.A. 1972. Genotype-environment interaction: its study and its implications in forest tree improvement. UFRO Genetics-Sabrao Joint Symposia, Tokyo, 27 pp. Shimshi, D. 1970. The effect of nitrogen supply on some indices of plant water relations of beans (Phaseolus vulgaris L.). New Phytol. 69:413-424. Spraque, G. F. and W. F. Federer. 1951. A comparison of variance components in corn yield trials. II. Error, year x variety, location x variety and variety components. Agron. Journal 43:535-541. Schulte, P. J. 1985. Stomatal response to leaf water potential in Populus . Unpublished Ph . D Dissertation . Univ. of Washington, Seattle, Washington, 157 pp. Tai, G. C. C. 1971. Genotype stability analysis and its applications to potato regional trials. Crop Science 11:184-190. 129 Teskey, R. O. and T. M. Hinckley. 1986. Moisture: effects of water stress on trees. In Stress physiology and forest productivity. Eds. T. C. Hennessey, P. M. Dougherty, P. M., S. V. Kossuth, and J. 0. Johnson. Martinus Nijhoft Pub. Dordrecht, pp. 9-33. Teskey, R. O, J. A. Fites, L. J. Samuelson, and B. C. Bongarten. 1986. Stomatal and nonstomatal limitations to net photosynthesis in Pinus taeda L. under different environmental conditions. Tree Physiology 2:131-142. Turner, N. C. 1974. Stomatal behaviour and water status of maize, sorghum and tobacco under field conditions. II. A low soil water potentials. Plant Physiol. 53:360- 365. Turner, N. C., E. D. Schulze, and T. Golan. 1985. The response of stomata and leaf gas exchange to vapour pressure in herbaceous species Helianthus annus. 0ecologia 65:348-355. United States, Dept. of Commerce, 1988. Local climatological data, Lansing, Michigan. Ying, C. C. and W. T. Bagley. 1976. Genetic variation of eastern cottonwood in eastern Nebraska provenance study. Silvae Genetica 25:67-73. Yoshida, S. and V. Coronel. 1976. Nitrogen nutrition, leaf resistance and leaf photosynthetic rate of the rice plant. Soil Sci. Plant Nutr. 22:207-211. van den Drieche, R. and P. F. Wareing. 1966. Dry-matter production and photosynthesis in pine seedling. IUUL Bot. 30:673-682. van Volkenburgh, E. and W. J. Davies. 1977. Leaf anatomy and water relations of plants grown in controlled environments and in the field. Crop Science. 17:353- 358. Venator, C. R. 1976. Natural selection for drought resistance in Pinus caribaea Morelet. Turrialba 26:381-387. Viets, F. G., Jr. 1972. Water deficits and nutrient availability. In Water deficits and plant growth. Vol. 3. Plant responses and control of water balance. Ed. T. T. Kozlowski. Academic Press Inc., New York, pp. 217-239. Well, C. O. and P. C. Wakeley. 1966. Geographic variation in survival, growth and fusiform. rust infection of planted loblolly pine. For. Sci. Monogr. No.11, 40 pp. 130 Westcott, B. 1987. A method of assessing the yield stability of crop genotypes. J. Agric. Sci. 108:267- 274. 1986. Some methods of analyzing genotype- environment interaction. Heredity 56:243-253. Woesner, R. A. 1972 a. Crossing among loblolly pines indigenous to different areas as a means of genetic improvement. Silvae Genetica 21:35-39. 1972 b. Growth patterns of one-year old crosses under contrasting edaphic conditions. Forest Science 18:205-210. Wong, S. C. 1979. Elevated atmospheric partial pressure of CO and plant growth. I. Interactions of nitrogen nugrition. and. photosynthetic. capacity’ in. C3 and. C4 plants. 0ecologia 44:68-74. Wright, J. W. and W. T. Bull. 1963. ‘Geographic variation in scotch pine: results of a 3-year Michigan study. Silvae Genetica 12:1-25. Zavitkovski, J. and W. K. Ferrell. 1968. Effect of drought upon rates of photosynthesis, respiration, and transpiration of seedlings of two ecotypes of Douglas- fir. Bot. Gaz. 129:346-350. Zobel, B. J. and J. T. Talbert. 1984. Applied tree improvement. John Wiley and Sons, New York, 505 pp. milliliiiiiiiiiiiiii