“22.34:?“ . * At... :1 3.. . .1 ‘11:} ’r' .1” :6 m. v ~r 9;! :‘c €¢;.'L"""::v —' “I V , ‘ n 593 v. \! fig}? , .' PM” ’ ~- baa“ ‘ ‘ 9:; :5. “'13. 1“ x? ‘ I" K 33:55 «. w ‘ ,, a; .‘357‘ . ,‘i . I‘- e'é uSfl'zx‘, 4 ..; - * ~ w. Izerézfiflrfix‘» .f'EmM" ‘ ‘ '. .. , I. (ll. ‘ “5,2; ‘4: . , 11¢"; ‘. . I.“ ‘ ‘- ’ Ruiz} x33}. ‘3 “"‘ : .I '~' ,1 i Q: ‘frflk: ' ‘L'Q‘Vi \A A“). ‘5. k Jé‘fi‘“ “H" (“t ; . THEBVB This is to certify that the thesis entitled Urban Land Rehabilitation Using Fast Groufing Black Locust presented by Charleen Marie Buncic has been accepted towards fulfillment of the requirements for Masters degreein Forestry ,~”l , “l7 Ljéggcayugrjkfhfiéhfliaor- / (/ Major professor Date 3-24-94 0.7639 M5 U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State Unlversity PLACE N RETURN BOXtomnavothbdndloMrom yourrocotd. To AVOID FINES Mum on or baton dd. duo. DATE DUE DATE DUE DATE DUE URBAN LAND REHABILITATION USING FAST GROWING BLACK LOCUST TREES BY Charleen Marie Buncic A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Forestry 1994 ABSTRACT URBAN LAND REHABILITATION USING FAST GROWING BLACK LOCUST TREES BY Charleen Marie Buncic Detroit spends more than two ndllion dollars annually maintaining approximately 60,000 vacant lots that.it.ownsu It is an economically depressed city with limited funds for environmental projects. This thesis explores the success of black locust on vacant urban land, and its suitably for use in urban land rehabilitation/vacant lot renewal. The effects of soil compaction on tree performance was evaluated using a penetrometer. No relationship was found between tree performance and the degree of soil compaction. Direct seeding was evaluated as a possible reclamation method, but its suitability was unable to be determined through this study. Fall planting of containerized seedlings was also evaluated. Black locust planted at later times during the fall had better survival and growth. For Greg I am grateful that we have met in our journey through this life. Never forget me, for I will always love you. ACKNOWLEDGMENTS I am deeply grateful for the opportunity provided to me by my former advisor, the late Dr. James Hanover. When I needed guidance and support he was there. When I was feeling hopeless, he provided the encouragement to continue with.this project. Dr. Hanover was a very special person and he is sadly missed. To Dr. J. James Kielbaso, thank you for taking on the advisory role after Dr. Hanover's death; I know that it was difficult. For making me laugh when I needed it most, and for your many efforts, your support and encouragement, I thank you. To my other committee members, Dr. Boyd Ellis and Dr. Maureen McDonough, thank you for your support and prompt feedback. I appreciate your willingness to work with me through all that had gone wrong, and helping me to keep things in perspective. To Dr. Douglas Lantagne for his guidance with statistics, and Dr. Phu Nguyen for his assistance with my data and Harvard Graphics. To Dr. Daniel Keathley for seeing to it that both me and my research were supported, and for his personal support during difficult times. I would also like to thank Better Trees, Inc. for donating well over two-thousand seedlings for this research; iv the Department of Natural Resources in Brighton, Michigan for donating the signs for the research site; and to the city of Detroit for the opportunity to conduct research on their land and for donating and delivering materials for the study. A very special thank you to two individuals who were always willing to help, despite the nature of the request. Paul Bloese and Roy Prentice. They have helped me in innumerable ways, and I am indebted to them for life. I would certainly have been lost without their assistance. To all of my friends who were never too busy to help out in the field, as rough as it was; I am grateful. Ken Ames, Edith Bross, Jye Chang, Chishih Chu, Andy David, Lisa Finck, Abibou Gaye, Pat Guertin, Tapani Haapala, Mark Hare, Bill Linnell, Kathy Maas, Ernest Mastorides, Tesfai Mebrahtu, Mike Powers, Carolyn Randall, James Roshetko, Todd Snider, Eric Stafne, Chrys Yangs and Greg Zielke. To three very special people, Bob Morikawa, Paul Mueller, and Peggy Payne, who have helped me in many ways both academically and personally. Thank you for being so wonderful. For my family who has always supported me; I love you. TABLE OF CONTENTS Page LISTOFTABLESCOOOO0..OOOOOOCCOOCCCOOCCOOOOO0.0000000000CCix LISTOFFIGURESOOOOO0.00....0.0.000......OOOOOOOOOOCOOOOOOxi CHAPTER I. INTRODUflIONOO ........ 00....OOOOOOOOOOOOOOOOOOOOOOI THE PROBLEM.......................................2 OBJECTIVES........................................2 TEXT PREVIEW......................................3 PROJECT LIMITATIONS...............................4 II. LITERATURE REVIEW, URBAN SOILS....................5 URBAN SOILSOOOOOOO..0.000......OOOOOOIOOOOOOOO0.006 MANAGING URBAN SOILS FOR PLANTINGS...............15 III. PROJECT BACKGROUND...............................18 BLACK LOCUSTOOOOOOOOOOOOOOOOOOO0.0.00.0000000000019 RESTRICTIONS...OOOOOOOOOOOOOOOOOOOOOO..0.00......22 Iv. SPACING/SOIL COMPAflION STUDY.‘00.00000000000000025 INTRODUCTION.....................................26 LITERATURE REVIEW................................27 MATERIALS AND METHODS............................30 Site Description...............................30 Experimental Design............................35 Site Preparation...............................41 Soil Sampling..................................41 Weed Control...................................41 Planting.......................................42 Fertilizer Application.........................42 Site Maintenance...............................43 Data Collection................................43 New growth data...........................44 vi CHAPTER VI. VII. Total height data.........................44 Soil compaction data......................44 Data Analysis..................................46 RESULTS AND DISCUSSION...........................46 DIRECT SEEDING STUDYOOOOOOOOOOOOOOOOOOOOOOOO0.00.65 INTRODUCTION.....................................66 LITERATURE REVIEW................................66 MATERIALS AND METHODS............................68 Site Description...............................68 Seed Preparation...............................68 Site Preparation...............................68 Soil Sampling..................................69 Weed Control...................................69 Seeding........................................69 Data Collection................................71 RESULTS AND DISCUSSION...........................71 FALL PLANTING TIME STUDY.........................76 INTRODUCTION.....................................77 LITERATURE REVIEW................................78 MATERIALS AND METHODS............................80 Site Description...............................80 Experimental Design............................80 Site Preparation...............................80 Soil Sampling..................................82 Weed Control...................................82 Planting.......................................82 Data Collection................................83 Data Analysis..................................83 RESULTS AND DISCUSSION...........................84 COMPOST STUDY.0.0...0.0..0...0.0.00.000000000000092 vii CHAPTER VIII.SUMMARY AND RECOMMENDATIONS......................95 SWY OF FINDINGS. O O O O O O O O O O O O O O O O O O I O O C O O O O I O O 96 DISCUSSION. 0 O O O O O O O O O O I O O O O O O O O C I O O O O O C O O O O O O O O O .97 FUTURE RESEARCH RECOWENDATIONS. O O O O O O O O C O O O O O O O O 98 GmSSARYOOOOOOOOOOO0.00.00.00.00...00.0.0000000000000000099 APPENDIXOOOOOOOOOO0.0.0.0.0.0.0000...0.0.0.0....100 REFERENCES.O...00....OOOOOOOOOOOOOOOOO...0.000.0101 GENERAL REFERENCES..............................104 viii Table 1. LIST OF TABLES Page Results of twelve composite soil sample analyses collected from all four blocks from theSite in1991.00.00.000000000000.0.0000000000000036 Results of twelve composite soil sample analyses collected from all four lots at the site in 1992....36 Results of chi-square analyses, making various comparisons between the height and the non-compacted zone and compacted zone. The variables have been regrouped up to two times, and further analyses made.0.00000000000000000000000COOOOOOO0.00.00.00.00048 Results of a linear regression analysis relating the non-compacted zone and total height for black locust sampled from lot two during the 1992 growing seasonOOOO0....OOO0..OO...00......00.000.000.000000049 Results of a linear regression analysis relating the zone of compaction and total height for black locust sampled from lot two during the 1992 growing season0.0...O0.0000000000000000000000000.0.0.0000... Results of a linear regression analysis relating the non-compacted zone and total height from black locust trees sampled from lot two during the 1992 growing season. Data used for this analysis do not contain the outliers................................53 Results of a linear regression analysis relating the zone of compaction and total height for black locust trees sampled from lot two during the 1992 growing season. The data used for this analysis do not contain the outliers............................53 Survival and growth results for lots two and four for the 1991 growing season; according to the original (spacing study) experimental design........60 ix Table 10. 11. 12. Page Survival and growth results for lot two for the 1992 growing season; according to the original (spacing study) experimental design.................62 Analysis of variance of arcsin transformed data showing a significant difference between fall planting treatmentSOOOOOO0.0...00.00.000.0000000000085 Overwintering success of containerized black locust seedlings planted at two-week intervals and four planting times during the fall season...............85 Fisher's Protected Least Significant Difference (FPLSD) test of mean values of survival and planting times for containerized black locust seedlings planted during the fall season............88 Fertilizer recommendations for twelve composite soil samples collected from the study site in 1991 and 1992......O0.00..O0.0.0....0.0...OCOOOOOOOOOOOOOOCOIOO LIST OF FIGURES Figure Page 1. Natural Range of Black Locust (Fowells, 1965, p. 642)OO...0.0....O...0......OOOOOOOOOOOOOOOOO0.0.0.0.20 2. Map of Michigan, Locating Detroit...................31 3. Map of Downtown Detroit, Locating Research Site (in bOX)I.0....0....0....OOOOOOOOOOOOOOOOOOO0.00.0.032 4. Research Area from the Box in Figure 3, Locating the Four (1 through 4) Research Plots...............33 5. Diagram of the Spacing Study, Located on Lot f2.....37 6. Diagram of the Spacing Study, Located on Lot #4.....38 7. Diagram of the Spacing/Compaction Study, Located on Lot #200...0.0.000....0.IO...0.0.00.000000000000039 8. Diagram of the Spacing/Compaction Study, Located on LOt ’40....0....OO...0....OOOOOOOOOOOOOOOOOOOOO..40 9. Regression of non-compacted zone (x) on height of black locust (y) sampled from lot two of the research SiteOOOOOOOO OOOOOOOOOOOO 0.0.0....0.0.0.000000000000051 10. Regression of compacted zone (x) on height of black locust (y) sampled from lot two of the research 8 teOO0...0...0.0.0.0...0..0....OOOOOOOOOOOOOOOOOOOOSZ 11. Regression of non-compacted zone (x) on height of black locust (y) sampled from lot two of the research site. Data does not contain outliers...............54 12. Regression of compacted zone (x) on height of black locust (y) sampled from lot two of the research site. Data does not contain outliers......................55 13. Diagram of the Direct Seeding Study, Located on Lot #4 (Not drawn to scale)......................70 xi Figure 14. 15. 16. 17. Page Diagram of the Fall Planting Time Study, Located on Lot ’3...0.0000000000000000000000000.00000000000000081 Percent survival of black locust (y) sampled from the four treatments (x) in the fall planting time study in 1992.0.0...0......OOOOOOOOOOOOOOOOOCOOOOO0.0.0.0086 Average height of black locust (y) sampled from the four treatments (x) in the fall planting time study in 1992.000...O0.0...00.0.0000...0.00.00.00.0000000087 Diagram of the Weed Control Study, Located on Lot ’1.00.0..O0.0...0.0.0.000....000000000000000000094 xii CHAPTER I INTRODUCTION 2 IBE_EBQBLEM Detroit is a large city with a population of 1,027,974 that has many poverty-stricken areas. According to 1989 national census data, the average poverty threshold for a four-person household was $12,674. In 1989, there were more than 143,936 households in Detroit living below the poverty threshold (Bureau of Census, 1990) . Monies for urban renewal have been limited, and hopes of receiving funding for projects aimed at environmental enhancement are small. The city currently has approximately 65,000 vacant lots. In accordance with city ordinances for vegetation management on public lands, the lots must be maintained. This is accomplished by mowing them two to three times during the summer months, costing the city more than two million dollars each year (verbal communication with Detroit Department of Public Works). This research ‘was conducted. to ascertain some economically feasible alternatives for vacant lot renewal. W The goal of this study was to determine, through several research experiments, the performance of black locust (Robinia pseudoacacia L.) trees under urban conditions at a site in Detroit, Michigan. Black locust was chosen because it is a low maintenance species that is able to tolerate weather extremes, such as heat and drought. It grows rapidly, and competes with grasses and weeds, thereby reducing the amount of required maintenance on the vacant land, and money invested by the city. Currently, there is a considerable amount of 3 research being conducted on the multiple uses of black locust. Black locust was chosen to explore its performance under urban conditions. A compaction study investigated the effect that soil compaction had on the survival and growth of black locust. A direct seeding study evaluated the plausibility of direct seeded black locust. A fall planting time study investigated the best time during the fall season to plant black locust seedlings. .A compost study explored the effect of the surface application of compost on the performance of black locust. IEXI_EBE¥IEE The characteristics of urban soils and their impact on tree growth will be described in the literature review. Applications for future research and potential vacant lot renewal projects for the city' will be proposed in ‘the concluding chapter of this thesis. Chapter Two contains a literature review on urban soils and their management for planting. Chapter Three provides background information on the project. Chapter Four details a compaction study that looked at the effects of soil compaction on tree growth. Chapter Five focuses on a direct seeding study that determined the success of direct seeded black locust on an urban site. Chapter Six discusses a fall planting study which tested the best time during the fall season to plant black locust. Chapter Seven briefly discusses an additional study which failed to yield reliable results. It has been presented in this thesis for the purpose of 4 documentation. Chapter Eight will summarize all research activities and findings, and present recommendations for Detroit. EBQIE§I_LIMIIAIIQH§ Anyone conducting research in the urban area should be aware of possible circumstances that they may encounter. Dealing with city'politics and.bureaucracy“will.most likely’be an issue. Other considerations to be aware of are vandalism, limited funding, quality of materials, level of maintenance and infrequent site visitation which are directly related to project support. Project limitations will become evident throughout this thesis. CHAPTER II LITERATURE REVIEW URBAN SOILS 6 QBEAN_§QIL§ "Drastically disturbed urban soils are those in which human activities have dominated the soil-forming process" (Short, 1990, p.229). High bulk density, lithologic variation and variable organic matter content are some of the unique characteristics of these soils (Short, 1990). The ecological environment of an urban area varies tremendously from one site to the next, and is quite different from the ecology of areas outside of a city. ‘Urban ecological factors impacting plant life include: "fifteen to twenty percent less solar radiation, 0.5 to 1.5 degrees Celsius higher mean temperatures, ten to twenty percent less wind speed, two to ten percent less relative humidity, five to ten percent more clouds and five to ten percent more total rainfall" (Craul, 1990, p. 219). Not only do the atmosphere and plant growth factors differ in the urban setting, but soils, something often neglected when planning for "urban reforestation", also differ, perhaps even more (Craul, 1992). Urban soils are distinct from other soils because they often are completely disrupted by construction and urbanization and they no longer function in a natural way. Urban soils usually have one or more layers about fifty centimeters thick comprised of fill material that often contains contaminants (some of which may be toxic), at levels higher than found in natural, undisturbed soils. This "soil" is a result of construction activities, e.g., mixing, filling, scraping, compaction, and pulverization. In addition to these problems faced by the 7 urban forester, is the scarcity of information pertaining to urban soils and their management. In many urban areas, there is a lack of soil survey' data, making soil assessment difficult.because there is no information on the original soil conditions (i.e., topography, depth of water table, parent material, depth to bedrock, etc.) (Craul, 1992). Creating a taxonomic system for classifying urban soils has been extremely difficult. The extent of spatial and vertical variability, because of disturbance, makes it difficult to gain an understanding and knowledge of the pedogenic processes which played a role in original soil formation. It takes many years for pedogenic processes to result in profile evolution in disturbed soils. This difficulty in classification and mapping of urban soils was best summarized.by Indorante et al. (1984): "Pedologists have been slow to define and separate different soils within disturbed land areas because these soils seem inherently too variable; mappable patterns of order were not apparent. The soil genesis model does little to help one perceive spatial order. The time is so short that the active factors of soil formation have had little effect. Lacking an applicable conceptual model, one commonly finds the apparent complexity overwhelming, and fails to perceive order" (In Craul, 1992, p.116). The District of Columbia Soil Survey has listed a dozen soil associations for urban areas. Two of those associations seem applicable to the majority of Detroit's urban area. Those associations are the "Urban Land Association" and the "Udorthents Association". The Urban Land association is described as "nearly level to moderate sloping areas, most of 8 which are built up and occupied by structures and works; on all landscape positions" (Smith, 1976, p. 9). The‘Udorthents association is described as "deep to moderately deep, nearly level to steep, well drained soils that consist of cuts, fills, or otherwise disturbed land; on all landscape positions" (Smith, 1976, p. 8). It is possible, with the soil taxonomic system developed thus far, to infer general characteristics of an urban soil. A detailed system has yet to be developed, and may take years to do so. Common characteristics of urban soils, include "a high degree of vertical and spatial variability, a loss of structure leading to a tendency towards compaction, poor aeration status and generally impeded drainage, presence of a hydrophobic crust on the bare surface, elevated soil reaction (pH), increased temperatures, interrupted cycling of nutrients and organic matter, and the presence of anthropeic (human- made) contaminants and toxic substances" (Craul, 1990). Vertical variability is characterized by abrupt changes, both chemical and physical, as the soil deepens. The layers may'differ in structure, texture, bulk.density, nutrients, and pH. Differences in these soil properties can lead to altered aeration, drainage, fertility, and water-holding capacity of the soil. Drastic differences in the soil layers can create a microsite that may not be suitable for plant growth (Craul,1985). Added to the complexities of vertical variability, is spatial variability. Spatial variability is characterized by horizontal differences that are caused by earlier construction 9 activities that can leave one site ‘very different from another, despite their close proximity (Craul, 1986). Because natural soil forming processes are lacking in urban soils, they easily become compacted. Urban conditions (such as history of being disturbed, low organic matter content, heavy amounts of surface traffic, infrequency of the structure-enhancing cycles of freezing-thawing/wetting-drying , and low amounts of vegetation), all lead to the5% Ideal) Nitrogen 0.59 - 4.40 (ppm non Low Table 2. Results of twelve composite soil sample analyses collected from all four blocks from the site in 1992. Fertilizer Parameter Range Index Calcium 5474 - 6232 (lbs/acre) High Magnesium 304 - 496 (lbs/acre) High Potassium 232 - 320 (lbs/acre) Med Phosphorous 4 - 21 (lbs/acre) Low-Med pH 8.1 - 8.2 Very high CEC 15.0 - 17.9 me/lOOg High Organic matter 2.5% - 3.5% Low (>5% ideal) Nitrogen 2.6 - 8.45 (ppm NOQ Low-Med fi 2 4— = ==—= 37 BE AUBIEN ROAD >m:< gnome; .om 0.82.. N a 38‘ N n 332.. N a .2: . N .32.. p a .Sel . . "3.5.5:; Die-aw Nxam , n p N EN .. 2.5 N 53x5: \ _- -, Ill!!! L .. till lullll'lllli, . 1! .‘ll‘UIvl ..|.. Oo:< 22mm2> .om e.e.el N n e.e.el N N :80! N a 3.2- . N .8el — I .e.el — p i...el.ee.b use-en m N m :3.— nxaa Em p N p em 5”.pr3 .\ O3=< 0.2.503 .03 e.3el N a 3303 N o 33: N a .30! p N .30! . I .30! a — "3.....ae; 95923 1 m C(Om wEOZmI OVOU HSDUB Figure 7. Diagram 0! the Spacing/Compaction Study Located on Lot #2 40 BE AUBIEN ROAD 53:5 so: :5: _: moo... -3 52......» em: :55 .c mama .- I m v vsm 5.: $58923 0.0.0.3 N I 0.0... N a I‘D-OE N I h...‘ a N 90.0.! p l 50...! v w ‘l m. , "3:25-03. GEO-ow 0 14¢ 95mm. :2 2d OVOU HSRHB Diagram of the Spacing/Compaction Study Figure 8. Located on Lot #4 41 W 1991 — The twelve experimental plots of the original spacing study were marked. They were fourteen meters by fourteen ‘meters. A two meter buffer zone was left between the plots to minimize the effects of shading. The total size of the study area including buffer zones was 2,760 meters squared. W 1991 - Three composite samples per lot, each consisting of twenty sub-samples, were collected. Samples were taken to a twelve inch depth. Soil samples were analyzed by the Soil Testing Laboratory located at the campus of Michigan State University, East Lansing, Michigan. All samples were analyzed for calcium, magnesium, phosphorous, potassium, nitrogen, cation exchange capacity (CEC), pH, and organic matter content. Results were presented in Table 1. Fertilizer recommendations were also given based on trees and shrubs as crops. See appendix A. 1992 - Prior to the second growing season, composite soil samples were again collected. They were analyzed for the same nutrients by the soil testing lab at Michigan State University. Results were presented in Table 2. W 1991 - Weed control was established by broadcast spraying the experimental plots with the post-emergent herbicide glyphosate at.a rate of three pounds active ingredient per acre (a.i.a.). 42 Three meter buffer zones were left between plots to minimize edge shading effects. 1992 - All experimental plots were spot sprayed to the point of drip with a 1.5 percent solution of glyphosate. Because weed growth was excessive the previous year, the pre-emergent herbicide treflan 5G was applied at a rate of two pounds per 1,100 square feet, or four pounds a.i.a. The herbicide was broadcast to all plots during a rain in order to increase its effectiveness. Banting 1991 - The planting stock was 1-0 bare root black locust seedlings. The option of using a tree planter was determined to be inappropriate because of heavy compaction and the presence of foreign materials. Using this equipment in these conditions could. pose hazards ‘to the equipment and its operators. Therefore, the seedlings were planted by hand using a dibble bar. Seedlings were root-pruned to a uniform length of eight inches (twenty centimeters). They were then placed into buckets containing peat moss and water to keep the roots moist during planting. A total of 1,636 seedlings were planted. WW 1991.- Because soil nitrogen levels*were extremely low, it was decided to broadcast fertilize the plots. Nitrogen-fixing trees may gain some benefits from nitrogen fertilization (Hanover, 1989). Ashby, et al. (1985) also note that planted 43 and seeded locust respond positively to a fertilizer appli- cation at the time of planting, unlike many other species. Fertilizing the trees with a‘very low level of nitrogen may be enough to enhance survival. This was important because the trees were very small (because they were planted at a high density in the nursery beds), and were being outplanted to a harsh site. Along with attempting to increase soil nitrogen levels, the formulation used would also aid in reducing the soil pH. Although it has proven extremely difficult to decrease soil pH as opposed to increasing it, sulphur is used in effort to do so (Foth, et al., 1988). Ammonium sulfate fertilizer (21-0-0) was broadcast by hand to each of the twelve plots. It was applied at a rate of forty-four pounds per acre (1.01 pounds per 1,000 ft’) in three separate applications between June and July 1991. ten e 1991 and 1992 - The research site was located in a residential area and it was suggested by the city (Department of Public Works) to maintain the aesthetics of the site. The perimeter of all experimental plots as well as the aisles were maintained by periodic mowing throughout the growing seasons. Both a brush hog mower and hand mowers were used. Plots were kept clearly marked with paint and flags. Motion 1991 and 1992 - Following the original design, new growth and total stem height were measured for all trees at the end of 44 the first growing season. On the un-mowed trees, survival, total height and the amount of new growth were measured. For the trees that were mowed twice, survival, height, and new growth after the second mowing were measured. Soil compaction and total tree height were measured on randomly selected trees within the thirteen isolated groups during the second year. W 1991 - The amount of new growth was measured for all trees in the study. New'growth was measured on the longest branch, and was recorded in centimeters. For the trees which were mowed for the second time in August 1991 (those on lot two), new growth was measured from the scar of the second mowing (this was about six inches above ground level). W 1991 - Total stem height of each tree, from soil level to the terminal meristem, was measured. Total height was recorded in centimeters. 1992 - Total height was recorded in the same manner as the previous year, recorded in meters. 52W 1992 - During mid growing season, groups of trees displaying different growth patterns which were located adjacent to each other, were selected. A qualitative scale with two levels was used to measure vigor. Healthy and vigorous trees received one number, while those that appeared chlorotic and less vigorous, received a different number. Height was estimated using a six-level scale ranging from no height to above five feet. 45 In August, soil compaction was measured using a cone penetrometer (DICKEY-John Soil Compaction Tester) with a half- inch diameter cone-shaped tip and a probe length of twenty- seven inches. The most reliable penetrometer readings are obtained when a soil is at field capacity, therefore readings were collected when the soil was presumed to be at field capacity. Measuring at field capacity minimizes variability in soil moisture, which could affect penetrometer readings. Soil compaction was measured next to random trees within the previously isolated groups. Three subsamples were collected around the base of each chosen tree, each tree being a sample. Because the penetrometer needs to be pushed into the ground at a slow steady rate, two people performed the test procedure. While one person recorded the readings, the other operated the penetrometer. The operator was responsible for pushing the instrument into the ground and reporting the readings. A flat soil surface was created before each measurement was obtained. This was accomplished by brushing away any litter or debris. Doing this assured that there were no obstructions to interfere with the accuracy of the readings. Holding the instrument in a vertical position, it was uniformly pushed into the ground, and readings were observed and recorded. Two measurements per subsample were recorded. The first measurement was the depth at which 300 p.s.i. was first reached, and was designated the ”non-compacted zone". The second measurement was the depth at which 300 p.s.i. was maintained without exceeding that pressure, this was designated the ”total depth". This final reading was recorded 46 when penetration was no longer possible, to a maximum depth of twenty-seven inches (probe length). Depth was recorded to the nearest inch. Survival of all trees in the entire spacing study was also recorded. DAIA_AEAL1§I§ Absurv statistical package (Anderson Bell, Corp.) was used to run statistical analyses on all data. Percent survival and average height were calculated on data from both growing seasons. A ”zoneiof compaction" was.calculated by subtracting non— compacted zone values from the total depth values. Chi-square tests were performed on the data. The analyses compared height versus non-compacted zone and height versus the compacted zone. All three variables, (i.e. , height, compacted zone, non-compacted zone), were regrouped two times. Comparisons were made amongst the variables (non-regrouped, and regrouped two times), with a chi-square analysis being performed on all possible comparisons. Correlations were performed relating height and the non- compacted zone, and height and the compacted zone. Linear regression analyses were also performed on the data. Both the non-compacted zone and the zone of compaction were regressed on total stem height. BE§QLT§_ANQ_DI§QH§§IQE The portion of the study which was not vandalized (that portion located on lot two) had 56.28 percent survival with an 47 average height of 1.17 meters, a minimum height of 0.12 meters, and a maximum height of 2.99 meters (refer to Table 9) . Results from chi-square analyses were not significant, suggesting that compaction was not a significant factor in influencing tree growth on this site. The data for all three variables were grouped to form new variables. The new variables were height regrouped, non-compacted zone regrouped, and compacted zone regrouped. These regrouped variables were again condensed and labelled as new variables. The original and regrouped non-compacted zone and compacted zone variables were compared in all possible combinations to the original and regrouped height variables. Comparisons made between all variables still failed to yield significant results, and are not able to explain the growth differences observed in the field. Results from all chi-square analyses can be found in Table 3. Correlation analyses showed there to be no relationship between the non-compacted zone and height (correlation value of -0.00236) , nor between the compacted zone and height (correlation value of 0.27099). Linear regression analyses for both the non-compacted zone versus height and the compacted zone versus height indicated that there was not a significant linear relation- ship. The results suggest that the soil characteristic measured does not explain the performance of black locust on this site. Results from both analyses are presented in Tables 4 and 5. 48 Table 3. Results of chi-square analyses, making various comparisons between the height and the non-compacted zone and compacted zone. The variables have been regrouped up to two times, and further analyses made. Test Comparison df Chi—sq. Prob Compacted zone 225 215.21 0.668 vs. height Non-compacted zone 180 194.86 0.212 vs. height Compacted zone regrouped 72 86.62 0.115 vs. height Non-compacted zone re- 135 126.24 0.693 grouped vs. height Compacted zone regrouped 75 89.42 0.122 vs. height regrouped Non-compacted zone regrouped 40 39.82 0.478 vs. height regrouped Compacted zone twice re- 20 18.71 0.541 grouped vs. height regrouped Non-compacted zone twice re- 10 9.53 0.482 grouped vs. height regrouped Compacted zone regrouped 34 40.08 0.218 vs. height twice regrouped Non-compacted zone regrouped 18 23.33 0.178 vs. height twice regrouped Compacted zone twice re- 4 6.22 0.183 grouped vs. height twice regrouped Non-compacted zone twice re- 2 2.72 0.063 grouped vs. height twice regrouped 49 Table 4. Results of a linear regression analysis relating the non-compacted zone and total height for black locust sampled from lot two during the 1992 growing season. £9312: .fii §L§& E_!ilnfi * Regression 1 4.7022 6.4646 ns Residuals 27 19.639 ‘ Total 28 24.341 R2 - 0.1632 * ns - not significant Table 5. Results of a linear regression analysis relating the zone of compaction and total height for black locust sampled from lot two during the 1992 growing season. Source 51f. $.28... Luise * Regression 1 0.5643 0.6408 ns Residuals 27 23.776 Total 28 24.341 R2 = 0.0129 * ns - not significant 50 Graphs of the data for the non-compacted zone and the compacted zone are presented in Figures 9 and 10. When a regression line was fitted to these graphs, it seemed to be skewed by the presence of some outlying data points. Further investigation of these points revealed that all but two data points occurred in.a part of the field that was questionable. These particular trees were located in an area which supported little vegetation. This area was rectangular measuring four- teen meters long by five meters wide. It was speculated that there may be something obscure, other than compaction (e.g., past dumping of toxic substances, leaking sewage tiles. etc.) , occurring in that area which could possibly be influencing tree growth. These points were eliminated from the data sets, regression analyses relating the same variables were per- formed, and graphs of the regressions plotted. The second set of regression analyses indicated a non- significant relationship between the non-compacted zone versus height and the compacted zone versus height. The eliminated data points did not have a significant influence on the re- gression. Results from the second set of regression analyses are presented in Tables 6 and 7, and the graphs are shown in Figures 11 and 12. The data used for the above analyses included twenty-nine sampled trees; three subsamples per tree. Collection of more data points may have yielded more reliable results. It is im- portant to reiterate that more than half of the data were lost to vandalism. Had this not occurred, there would have been seventy samples, which may have yielded different results.' 51 20:30 9:53:00 3.5 30:25 oco~ 363800-52 m m6. v man a mVN N m# p mAv o 1 4 q q 1 J E d u q o msw .2825 220: £90: .m> ocoN moLomanoucoz .m 9:9... Regression of non-compacted zone (x) on height of black locust (y) sampled from lot two of the research site. Figure 9. 52 3 llvltl r|.1lll-ll oil-l c:||ll‘.til|',ll 0.2:.5 2.2.3.50 3.5 32.2.: ocoN ooaomofioo or m m N c m v n N w o a q q _ q q q q a - O md .8225 £22.. £92.. .m> ocoN mouomano .0. 2:9”. It... OI'II'IIII'II. on height of (y) sampled from lot two of the compacted zone (x) Regression of black locust research site. Figure 10. 53 Table 6. Results of a linear regression analysis relating the non-compacted zone and total height from black locust trees sampled from lot two during the 1992 growing season. Data used for this analysis do not contain the outliers. 529125 9: §1§1 £_!§lflfi * Regression 1 0.2568 0.3621 ns Residuals 22 15.605 Total 23 15.862 R2 - 0.0285 * ns - not significant Table 7. Results of a linear regression analysis relating the zone of compaction and total height for black locust trees sampled from lot two during the 1992 growing season. The data used for this analysis do not contain the outliers. Source 9f $1.5... Males Regression 1 0.0037 0.0052 Residuals 22 15.858 Total 23 15.862 R2 -= 0.0452 54 l 3 1 15 215 Figure 11. Non-Compacted Zone vs. Height (15 2 Non-Compacted Zone (Inches) L. Height (meters) (15" :15 Data does not contain outliers Figure 11. Regression of non-compacted zone (x) on height of black locust (y) sampled from lot two of the research site. 'Data does not contain outliers. 55 11 1 7 i l 5 6 Compacted Zone (inches) Figure 12. Compacted Zone vs. Height L s s . .1 (9 8 - .2 A O :3 3 0 -‘ N c ~ 0. o ' g . E C V O u d P 0 8 ‘a '3 _ s I 1 1 1 1 1 1 o 8 '0. to no N it) v- 10 o .8 0) CV 'P C’ s E O Figure 12. Regression of compacted zone (x) on height of black locust (y) sampled from lot two of the research site. Data does not contain outliers. 56 In an urban site, there are many factors influencing tree growth. This study did not indicate that compaction had a significant effect on tree growth. It would not be reasonable to completely rule out compaction as having an influence. The level of compaction which was found here was just a portion of the range of values for the degree of compaction. For example, if one portion of the site did not show a large degree of compaction while the other portion did, regression analyses may have yielded more significant results. Because a great portion of the site was compacted, obtaining significant results for a range of soils with varying degrees of compaction was difficult. The data revealed broad variation in the total depth measurement around each tree. Variations in the non-compacted zone readings ranged from a quarter of an inch to six inches around the same tree. For the total depth measurement, there were variations in depth up to fifteen inches at one tree. This much variation in compaction within one tree microsite seems unusual, considering that the maximum diameter was never more than three feet. This extreme variation may be explained by the degree of variability at this site, or it could be an indication of improper use of the penetrometer. Another contributing factor to the variation found around some trees and the non-significant relationship between height and compaction could be that penetrometer readings were not accurate. The most reliable readings are obtained when the jpenetrometer is inserted into the soil at a slow, steady rate. The presence of foreign material makes this difficult to 57 accomplish, and may have resulted in inaccurate readings which were not able to reflect if there was a significant inter- action between height and compaction. Although the city blocks used for this research are considered to be "open urban land" (they presently have no buildings on them), they still showed vertical and spatial variability from past human activities and they consist of miscellaneous fill and preexisting artifacts (foundations). The presence of anthropeic materials could have affected tree survival and growth. There was a substantial amount of foreign matter on the site used for this research, visible on the surface and when digging or planting. Planting with either the dibble bar or shovels was difficult. Planting in 1991 was done according to the design for the spacing study. In order to maintain the integrity of this study at the time of planting, treatment spacings were adhered to. Therefore, if an immovable object was encountered when planting, the planter had no choice but to inch away from the planting spot until a workable location was found. Many times it was not possible to find a suitable location without severely skewing the spacings (and hence the row effect), therefore the tree was planted as deep as possible at the intended location; even if that meant planting on top of a large piece of anthropeic matter. It is possible that improper planting lead to mortality or poor growth; but in many cases, finding a spot free from anthropeic material was unavoidable and the tree had to be planted at a shallow depth. 58 Soil conditions were probably the most influential factors affecting tree performance, as opposed to atmospheric conditions (pollution, heat island, precipitation). Performance of these trees on the site in Detroit, can most likely be attributed to a combination of factors found in an urban environment, emphasizing the soil. There are many confounding factors ‘that need. to be studied (like con- taminants, soil fertility, anthropeic materials, poor species- to-site selection), but this was beyond the realm of this project. Because the site was so variable, it would take a major research effort to determine the factors directly influencing growth. During data collection in 1991, it was noticed that many of the dead trees, had chewed roots. Because the symptoms indicated insect damage, several samples of dead trees were taken to the Pest Diagnostic Center in the Entomology De- partment at the campus of Michigan State University. It was diagnosed that white grubs had recently chewed and killed the roots of the trees sampled. Because white grubs 'were frequently encountered during planting, their activity could be a likely explanation for some mortality. It is also possible that their populations were present in isolated locations, resulting in the patchy tree growth patterns that were observed at the end of the first and during the second growing seasons. Another factor that may have influenced growth was that these trees were mowed twice during the first growing season. It is possible that survival and growth would have been 59 greater had this not occurred. However, these trees seemed to grow in spite of both mowings, hence the mowings may not have been a significant confounding factor. In fact, observations made after both mowings indicated that those trees that had been mowed, grew more vigorously than those that were not mowed. Survival and growth results from the 1991 growing season are presented in Table 8. The average amount.of new'growth (based on 1991 data) for mowed lot two (7.50 centimeters) is 1.85 centimeters less than that for unmown lot four (9.35 centimeters). It is important to stress that the lot two trees had been mowed, for the second time, in early August. They had about two months (sixty-nine days) to produce new growth before the first frost on October 11, 1991. Determining the growth potential (by dividing the number of days that each set of trees had.to»grow (lot two - 69; lot four - 155 days) into the average amount of new growth per lot), indicates that the trees which were mowed two times (lot two) grew nearly twice as much (1.82 times) as those that were never mowed. This outcome may be due to a better root-to-shoot ratio 'being established in the trees that were mowed. It was speculated that the lot two trees would have the reserves to put on more growth before winter, but not enough to make it through the winter and into spring. However, these trees were able to survive the winter and were able to increase their total height by almost.six:times (5.9 times) during the second growing season. Because all trees on lot four were vandalized, it was not possible to determine the amount of I Table 8. four for the 1991 growing season; according to the original (spacing study) experimental design. LQI_Z Percent* Avg.* Hgt.* New* N.G.* TIE; fiflllilil REESE QIQEID BADQE 1 74.67 20.07 12.5-77.8 7.65 0.1-65.9 2 75.00 20.74 12.5-71.1 7.59 0.1-56.7 3 62.50 16.87 12.5-50.0 6.80 0.1-40.5 Total 72.88 19.77 7.50 A m Trtommmhtmgsfizmnfiangs 1 70.99 32.20 13.8-72.0 9.04 0.1-38.1 2 65.40 37.09 15.5-102.0 9.17 0.1-73.9 3 77.30 42.00 14.1-97.2 10.78 0.1-56.0l Total 70.30 34.76 9.35 * Units are in centimeters 61 growth that those trees gained during the second growing season. Survival and growth results for the 1992 growing season for lot two are presented in Table 9. It is possible that some.mortality could be attributed to a high weed and grass population during the first growing season. Although the lots were treated with a post-emergent herbicide prior to planting, fertilizer applications may have encouraged weed growth. ”Black locust usually is more successful than other trees in becoming established" in heavy ground cover (Ashby, et al., 1985). It is possible that competition from weeds did not contribute to the majority of first year mortality that was calculated (28.41 8 average). The trees were able to grow well, if in a suitable place. At this point it is not possible to determine long term survival or vigor. In summary, the stated hypothesis that areas displaying poor tree survival and growth have a greater degree of soil compaction than those areas displaying better tree survival and growth was rejected. In the event that Detroit would have the resources to 'conduct city-wide planting operations of black locust, they would need some recommendations. Given the amount of information obtained from this research, it would be difficult to say for certain what the success of such an effort might be. Continuous site monitoring needs to be carried out for several years. This would provide information on long term survival and growth of black locust on this particular site. If the trees are able to thrive on this site with only two 62 Table 9. Survival and growth results for lot two for the 1992 growing season; according to the original (spacing study) experimental design. 1:11 W W Range: 1 58.70 123.0 12.0-287.0 2 58.30 115.0 15.0-299.0 3 44.00 102.0 12.0-192.0 Total 56.28 117.0 * Units are in centimeters years of maintenance (i.e., site preparation, weed control, planting, etc.) , then it would be possible to suggest that the city needs only two years of major maintenance investment. In this case, Detroit could begin, with little resources, to establish some urban forests of black locust.w Follow-up studies should focus on soil physical, biological, and chemical conditions. Extensive soil analyses ‘would provide information on nutrient and micro-nutrient availability and deficiencies, contaminants and . their toxicity, microfauna populations, and other physical, biological, and chemical properties of the soil. Establishing a grid system over the site and collecting soil samples for extensive analyses may provide more detailed information regarding present soil conditions and the extent of its variation. Penetrometer readings in conjunction with bulk 63 density readings may offer more information on the degree of compaction. Future research could also be conducted on the trees already established at the research site. A natural phenomenon with older black locust stands is deterioration, loss of trees, and canopy breakup. IFuture studies could focus on underplanting the site with different species and determining their success. Examining the condition of the soil and its influence on the new plantings would provide information on any soil amelioration that hopefully would have resulted from planting black locust on these sites. Examining root performance may also provide more information on soil conditions. A detailed study on plant roots (morphology, growth patterns, nutrient content), though tedious, can. reveal. a myriad information, if there are analytical data from both foliar and other tissues with which to correlate. A species comparison trial would provide information on the most suitable species for these urban sites. If locust exhibits adequate survival, it could serve as a nurse crop for less "site"-tolerant species. In terms of aesthetic appeal, wildlife value, and property value, the establishment of a mixed stand would be more beneficial than a monoculture. A stand of mixed species could provide several food sources for wildlife, resulting in variable wildlife species. Mixed stands can be more appealing aesthetically, which could potentially increase property values. Mixed stands are also usually healthier because they are less susceptible to devastating attacks by pests. The species 64 comparison trial could also consider any soil or atmospheric variations that exist within this urban area. It would also provide data on species mortality and this would provide increased confidence in selecting plant material for vacant lots in the city. If such data were applied across the city, soil analyses should be conducted per site to ensure that chosen species are site-suitable. CHAPTER V DIRECT SEEDING STUDY 65 66 IEIBQDHQIIQE Revegetating a large city such as Detroit can be time consuming and expensive. Without sufficient funds, it can be difficult to embark on a revegetation project. Direct seeding of woody species is an alternative revegetation method. It is economically feasible because it requires less labor than planting. If planned properly, direct-seeded trees may quickly revegetate derelict lands, such as those found within Detroit’s urban areas. Another benefit of direct seeding is rapid coverage of a large area. It seems worthwhile to consider direct seeding as a method of revegetation because Detroit has so much vacant land. This will provide rapid revegetation and offer the myriad benefits that are associated with trees in an urban area. The functions of trees in an urban area range from architectural and aesthetic uses to climatological (cooling, windbreaks, etc.) and emgineering uses (erosion control, sound control, etc.) (Miller, 1988). The objective of this study was to evaluate the success of direct seeded black locust in Detroit. This experiment attempted to test the hypothesis that direct seeding of black locust on an urban site will be successful. LIIEBAIHBE_BE¥IEH The use of direct seeding versus planting nursery stock for woody establishment on disturbed land, is controversial. Either method is certainly situation-dependent. Results 67 depend on site selection, species selection, seed source and treatment, planting time, and other cultural methods (Wakeley, 1954) . Direct seeding has both successes and failures. Direct seeding is usually favored to avoid the high costs of stock, shipping, and planting. Direct seeding eliminates dependence on a nursery and results in more time available for field work. The trees also have a better chance for more normal root development (Wakeley, 1954). The high costs associated with the use of nursery stock for revegetating large areas of land can be prohibitive. The initial cost of the stock, labor and maintenance, especially for the first year, can make planting uneconomical. Because planted seedlings have so little time to adapt and to recover from the stress of outplanting, they become more susceptible to climatic stresses (Hipkins and Coartney, 1987). In contrast, direct seeding offers a lower initial cost, lower maintenance costs and lower labor costs relative to using nursery stock. Newly emerged seedlings have an advantage over planted seedlings because they do not utilize their starch reserves to overcome the strains of lifting, planting, and adapting to a new environment (Hipkins and Coartney, 1987). The disadvantages of direct seeding are: 1) rodent damage and 2)germination problems. Seeds and young seedlings (new shoots) commonly fall prey to rodents because new seedlings are smaller and tender. Rodent damage can be extensive. Seeds that do germinate can be affected by insects and 68 disease. Weather can also be a problem. Despite careful planning and disease and rodent prevention, the weather is uncontrollable and not easily predicted. Low temperatures and more importantly, dry periods can prevent germination. Even if there is germination, extensive drought can kill almost everything (Wakeley, 1954). ELIEBIAL§_AHD_HEIHQQ§ This part of the study was initiated in mid-April 1991 and terminated in June 1992. te es t n A detailed description of the study site can be found in Chapter Four. W In early June 1991 about 900 black locust seeds were treated for sowing. The seeds were scarified by soaking them in a ninety-three percent solution of sulfuric acid for one hour, while stirring occasionally. Scarification, prior to planting, increases the percentage of early germination (Ashby, et al., 1985). Scarified seeds were then rinsed well with water and stored in a seed cooler at thirty-five degrees fahrenheit and fifty percent relative humidity until sowing time, three days later. Wise 1991 - A ten meter by ten meter area was marked off and was broadcast sprayed with the post-emergent herbicide glyphosate 69 at a rate of three pounds a.i.a. A diagram showing the plot layout can be found in Figure 13. Melina 1991 - Soils were sampled in the same manner as described in the compaction study. Refer to the soil sampling section in Chapter Four. Wee t o The use of certain herbicides on different soil types has been found to prevent or decrease germination in direct seeded black locust (Geyer, Melichar and.Long, 1987). The use of any potentially harmful herbicide was avoided in this study. 1991 - Weed control was established by broadcast spraying the experimental plot with the pre-emergent herbicide glyphosate at a rate of three pounds a.i.a. per acre. 1992 - The site was spot sprayed to the point of drip with a 1.5 percent solution of glyphosate. Seeding 1991 - Seeds were sown at a one meter by one meter spacing. Because of compaction, it was not feasible to mechanically cultivate the entire site. Using a hoe, a small area was cultivated where the seeds were to be sown. A planting hole, about a quarter of an inch deep, was formed. Five to seven seeds were planted in each hole, and were covered with a thin layer of soil. The seeds were then watered until about a ten inch diameter area was saturated. Orange, wooden stakes 7O BEAUBIEN ROAD >22 Sacco.) 5m Soon 2 .oz 9:06. .206. a .32... s 3 S6- spoon v... .3 so 6233 >35 3.83 9: .6 666.65 0...: O- I 0...: 9 6N Ev— !.I||'.u - " "0’1-‘l- OF Block 3 429.7869 143.2623 4.09 0.0436 Treatment 3 1909.637 636.5457 18.17 0.0004 Error 9 315.3406 35.03785 Total 15 2654.764 Table 11. Overwintering success of containerized black locust seedlings planted at two-week intervals and four planting times during the fall season. Treatment Survival and Growth 1 2 3 4 (8/27) (9/12) (9/25) (10/11) Avg. Percent Survival 16.7% 37.5% 53.5% 62.5% 42.5% Avg. Total Height (cm) 33.32 46.10 52.02 65.43 49.23 Height Gain (cm) 13.00 25.78 31.70 45.11 28.91 86 :>; v m \m\mw _._\‘w .op ON [on 10? ..0m 333.5 2.083. woocoaotfi 29:23:. 6.5 68E 8:526 :61 .9 659“. co Percent survival of black locust (y) sampled from the four treatments (x) in the fall planting time study in 1992. Figure 15. 87 P_\m: 36c mczcsn: E958; n N . ow mp .;\m 1cm Dex? 1mm .1 1 on 62w 9:: 8:8... =6“. .9 65?. .53 290: 80:05.20 Ego: Average height of black locust (y) sampled from the four treatments (x) in the fall planting time study in 1992. Figure 16. 88 Table 12. Fisher's Protected Least Significant Difference (FPLSD) test of mean values of survival and planting times for containerized black locust seedlings planting during the fall season. Treatment Mean (%) Mean Arcsin Survival Transformation 1 (8/27) 16.7 23.60 A * 2 (9/12) 37.5 37.40 B 3 (9/27) 53.5 46.93 c 4 (10/11) 62.5 52.40 C FPLSD 8 10.932 * means with the same letter are not significantly different from each other at the .025 probability level. outplanting, whereas treatments three and four had frost exposure. "Fall planting must be in perfect physiological synchronization with their intended environment" (Ellington, 1985) . All seedlings should have been acclimatized in Detroit where they would have been exposed to the same environmental conditions as the site. Although treatment comparisons are not reliable, discussion of treatment differences may be valuable. One would expect that survival and height for treatments one and two to be greater than those for treatments three and four. This is because one and two had more time after outplanting to establish a root system and accumulate stem height. Because three and four were outplanted later in the growing season, they would not.be able terestablish.a‘vigorous 89 root system nor gain much height before winter. However, treatments three and four showed better survival and growth than one and two (refer to Table 11 and Figures 15 and 16). This is probably because treatments one and two were severely stressed following outplanting, and did not have sufficient energy reserves for proper establishment before winter. There must be adequate soil moisture for the seedling after outplanting to encourage root growth. If adequate soil moisture is not provided, "the whole point of fall planting concept is lost" (Ellington, 1985). Soil moisture was low at the time of planting treatments one and two (August 27 and September 12, respectively). It is probable that a low soil moisture content affected the performance of the trees in these treatments, resulting in their survival and height differences with treatments three and four. It is possible to attribute some mortality to frost heaving. Although shovels were used to increase planting depth, ensuring a proper planting depth was still difficult. Large slabs of buried concrete were frequently encountered, and the amount and quality of topsoil sparse. Although planting'was carried out.to the best of ones ability, this was not enough to always ensure an appropriate planting depth because of site conditions. One report indicates frost heaving as being the cause for extensive amounts of initial seedling mortality (Anderson et al., 1983). Another report indicates that frost heaving was not a problem. But instead, insufficiently hardened off stock for fall planting or improperly planted seedlings were reasons for 90 poor survival (Van Sambeek et al., 1987). Planting stock must be sufficiently hardened off before it is planted during the fall. For hardwoods, this is after exposure to low temper- atures, at which point they become