1998 Spotlight on Graduate and Undergraduate Research at ESF
Forestry Abstracts

ASSESSING NITRATE AND AMMONIUM LEACHING IN ORGANICALLY AMENDED SOILS.  Hector G. Adegbidi and Russell D. Briggs.

NON-DESTRUCTIVE BIOMASS ESTIMATION OF SHORT ROTATION BIOENERGY CROPS USING ALLOMETRIC EQUATIONS: MODEL SELECTION AND EVALUATION FOR FIVE WILLOW CLONES AND ONE POPLAR CLONE. Benjamin D. Ballard, Steve V. Stehman, and Russell D. Briggs.

REGULATORY CLIMATE AND FOREST PRODUCTS INDUSTRY COMPETITIVENESS: PERSPECTIVES FROM THE NORTHEAST. Megan Carroll, Valerie Luzadis, John Wagner, and Don Floyd.

GROWTH OF SAPLINGS FOLLOWING A SELECTION CUT IN CENTRAL NEW YORK NORTHERN HARDWOODS. Pablo J. Donoso, Ralph D. Nyland, Lianjung Zhang, and Mariano Durn.

THE ROLE OF THE MANAGER'S EMOTION IN ENVIRONMENTAL DECISION-MAKING: APPLICATION OF FAMILY SYSTEMS THEORY TO THE MANAGER'S DECISION MAKING PROCESS IN THE FOREST SERVICE. Joan Kennedy and Don Floyd.

A LONGITUDINAL ANALYSIS OF TREE VOLUME DATA FROM STEM-ANALYSIS MEASUREMENTS. Fasheng Li and Lianjun Zhang.

RIPARIAN ZONE FLOWPATH DYNAMICS IN A SMALL HEADWATER CATCHMENT.   Brian L. McGlynn, Jeffrey J. McDonnell, James B. Shanley, and Carol Kendall.

THE EFFECT ON SOIL pH AND BASE CATION STATUS OF SAMPLE STORAGE AND PROCESSING. WHY DO WE CARE?  David McMillan and Ruth Yanai.

SCREENING WILLOW AND POPLAR BIOMASS PRODUCTION CLONES FOR INSECT RESISTANCE. Erik E. Nordman, Daniel J. Robison, and Lawrence P. Abrahamson.

MODELING LOCAL GOVERNMENT DECISION-MAKING IN WATERSHED MANAGEMENT. Timothy Schaeffer and Valerie Luzadis.

FIRST YEAR GROWTH AND DEVELOPMENT OF WILLOW AND POPLAR BIOENERGY CROPS AS RELATED TO PHOTOSYNTHETIC CHARACTERSTICS. Pradeep J. Tharakan, Lawrence P. Abrahamson, Daniel J. Robison and Jud G. Isebrands.

RESULTS OF A SPECIES SELECTION TRIAL ON THE JOS PLATEAU, NIGERIA. Timothy Volk, Andrew Kidd, Philip Godwill, and Lawrence Abrahamson.

INTEGRATION OF REMOTE SENSING, GIS AND ON-SITE PHOTOGRAPHY FOR MODELING URBAN INFLUENCES ON SUMMER MICROCLIMATE. Yingjie Wang, Gordon M. Heisler, and Lee. P. Herrington.

ASSESSING NITRATE AND AMMONIUM LEACHING IN ORGANICALLY AMENDED SOILS. Hector G. Adegbidi and Russell D. Briggs, Faculty of Forestry, 342 Illick Hall, SUNY-College of Environmental Science and Forestry.

     Wood biomass of willow clones (Salix spp.) grown in short-rotation intensive cultural systems (SRIC) has potential as source of renewable energy. The achievement of high biomass yield requires that adequate quantities of nutrient be supplied to plants. The use of biosolids and other waste products, as soil amendments in willow bioenergy plantations, have potential as inexpensive source of nutrients while it also providing for a safe disposal of the waste products. A greenhouse experiment was conducted at Lafayette Road Station to assess available nitrate and ammonium in soils amended with various waste products. Soil, in pots, was top-dressed with New York City sludge cake, Syracuse sludge cake, sludge compost, poultry manure compost and lime-stabilized sludge, each at a rate of 2.5 cm thick layer of material. Ammonium nitrate fertilizer and a slow-release nitrogen fertilizer were also applied at 200 kg N/ha. Anion and cation exchange resins were used as sinks for ammonium and nitrate leaching through columns of the organically amended and fertilized soil. The mass decrease of the applied materials was also recorded. After five months of incubation the treatments of New York City sludge, Syracuse sludge, sludge compost and poultry manure compost have lost respectively 25, 52, 27 and 47% of their initially applied masses. Nitrogen leaching as nitrate was variable and ranged from 56 to 110 kg N/ha. Nitrogen leaching as ammonium was negligible, ranging from 0.4 to 1.5 kg/ ha. Nitrogen collected by the ion-exchange resin columns suggested that the applied rates of sludge and compost materials could supply as much N for plants as mineral N fertilizers at a rate of 200 kg N/ha.

NON-DESTRUCTIVE BIOMASS ESTIMATION OF SHORT ROTATION BIOENERGY CROPS USING ALLOMETRIC EQUATIONS: MODEL SELECTION AND EVALUATION FOR FIVE WILLOW CLONES AND ONE POPLAR CLONE. Benjamin D. Ballard, Faculty of Forestry, 343 Illick Hall, SUNY-College of Environmental Science and Forestry, Syracuse, NY, 13210. (Major Professors: Steve V. Stehman and Russell D. Briggs.)

     Current research at the SUNY College of Environmental Science and Forestry (ESF), in a coordinated effort with other Salix Consortium members, is aimed at establishing an operational level of willow biomass production on a regional level to supply wood for co-firing with coal at power plants in the region. At an operational level, willow plantations will be harvested, chipped and transported to power facilities. Chips will have to be delivered on an as needed basis because of the limited amount of storage space at the plants. In order to ensure a steady supply of willow biomass to the plant, estimates of biomass production before the time of harvest will be necessary. Landowners will likely find estimation of biomass useful for decision-making. For example, rotations are typically three years, but could be extended to four or even five years if the amount of standing biomass does not make harvesting profitable. In addition, current research often requires estimation of aboveground biomass prior to harvesting. At the time of harvest, biomass can simply be weighed, but obtaining estimates prior to that time requires the use of an alternate non-destructive estimation technique. The use of regression equations is a reasonable non-destrucive estimation approach, because there is a strong allometric relationship between stem diameter and dry stem biomass.
     The objective of this study was to develop equations that accurately predict biomass in a cost effective manner. Stem samples of willow clones S25, S301, S546, SA2 and SV1, and poplar clone NM6 were measured in the field, then cut, oven dried, and weighed. For each clone, 27-94 stems were collected from a total of six to nine stools from at least two sites. Log-transformed, diameter-squared, and non-linear models were considered for stems of each clone individually and combined. Several variables were considered, including height (H), diameter (D), D² and D²H. Data splitting was used for obtaining a validation data set, and mean squared error of prediction (MSEP) was used as a validation criterion. Transformational bias can be an issue when using log-transformed models and various bias correction approaches were tested.
     Regression equations based solely on diameter can adequately estimate tree biomass. Adding height to the stem-based regression models does not appear to improve the models significantly, and increases the cost of data collection. Clone-specific models appear to be necessary. The performance of the log-transformed model and that of the non-linear model were comparable, and there was little justification for using a bias correction for log transformed models. With the exception of the poplar clone, NM6, the non-linear and log-transformed models performed equally well, suggesting that the two models are both reasonable. For NM6, the log-transformed model performed better.

REGULATORY CLIMATE AND FOREST PRODUCTS INDUSTRY COMPETITIVENESS: PERSPECTIVES FROM THE NORTHEAST. Megan Carroll, Valerie Luzadis, John Wagner, and Don Floyd, Faculty of Forestry, Bray Hall, SUNY College of Environmental Science and Forestry, Syracuse, NY.

     In recent years, New York’s regulatory climate has been cited as a critical factor hindering the competitiveness of its forest products industry (FPI) relative to other states in the Northeast region (ESFPA 1995, Canham 1995, Canham and Smith 1994). Despite estimates that New York’s forest resources could support an expanding FPI, such an expansion is inhibited by the perception that New York is a difficult state in which to operate (Canham and Stegemann 1989). FPI representatives throughout the Northeast share similar concerns that stringent state regulatory climates threaten the competitiveness of the industry and jeopardize the sustainability of forest resources (Northern Forest Lands Council 1994). Unfortunately, little in the way of rigorous comparative research is available to support or negate these assertions.
     This study attempts to fill this gap in the research by developing a methodology for making interstate comparisons of the regulatory climates faced by primary forest products firms. Based on input from FPI representatives and the relevant literature, an operational definition of regulatory climate was developed that identified five key components. These components of regulatory climate served as the basis for measures of actual and perceived regulatory climate. The construction of a regulatory climate index based on published data allowed us to examine the relative stringency of the actual regulatory approaches that different states have taken toward the primary FPI. The index provided a grounding measure against which to compare primary forest products representatives’ perceptions of several states’ regulatory climates as measured by a mail survey instrument.
     The methodology was implemented in a comparative assessment of the regulatory climates faced by loggers and sawmillers in New York, Maine, Pennsylvania, and Vermont. The regulatory climate index suggests that these states have enacted similarly stringent policies in some critical areas and not in others. Preliminary results indicate that survey respondents tend to perceive the regulatory climates of their home states as significantly more stringent than those of the other states included in the study. Further analysis will reveal which components of regulatory climate tend to drive perceptions of regulatory climate in each state.

GROWTH OF SAPLINGS FOLLOWING A SELECTION CUT IN CENTRAL NEW YORK NORTHERN HARDWOODS. Pablo J. Donoso, Ralph D. Nyland, Lianjung Zhang, and Mariano Durn, Faculty of Forestry, Marshall Hall, SUNY College of Environmental Science and Forestry, Syracuse, NY 13210.

     An assessment of height growth was conducted in a 23-acre northern hardwood selection stand cut to a residual density of 73 ft²/ac. Twenty four years later, during the second cutting cycle of the stand (the first lasted 19 years), 20 1-2 in. sugar maple (Acer saccharum) and 20 beech (Fagus grandifolia) saplings were randomly selected, cut, and subjected to stem analysis to determine height growth and time to reach different diameters and heights. Both species where pooled after conducting a dummy variable analysis. The saplings ranged from 26 to 52 years of age. They were classified according to the proportion of their lives under managed conditions, e.g. more than 80% (young trees), 67-80% (intermediate trees), and <67% (old trees). Five 5-yr. periods were analyzed, one representing the period before the cut (1968-1973), and the other four periods afterward. Trees of the different categories were compared for time to reach 1 in. in dbh. Young trees reached 1 in. in dbh almost one decade sooner than old trees (25,7 vs. 33.8 yrs.), and were not significantly different in height when reached 1 in. dbh. Growth of saplings for the second half of the cutting cycle did not differ significantly from these in the unmanaged stand (2.4 ft.), but differed from the growth during the first 10 years after cutting (2.85 ft.). Young and intermediate trees always grew more than old trees, but especially during the first 10 years after cutting (6.1 vs. 4.9 ft for the 10 years), and 20-25% more in height during the entire 19 year cutting cycle. For a shorter cutting cycle (10-15 years), significant decreases in height growth would not occur as stand density increases. Matching of the residual density and the cutting cycle is of importance as reflected in maintaining good rates of height growth among young age classes in an uneven-aged stand.

THE ROLE OF THE MANAGER'S EMOTION IN ENVIRONMENTAL DECISION-MAKING: APPLICATION OF FAMILY SYSTEMS THEORY TO THE MANAGER'S DECISION MAKING PROCESS IN THE FOREST SERVICE. Joan Kennedy, Faculty of Forestry, 320 Bray Hall, SUNY College of Environmental Science and Forestry, Syracuse, New York 13210, Professor Don Floyd.

     Family Systems Theory (FST) is used to explain the influence one's emotions have on natural resource decision making processes. The manager's decision making process is defined as person to person interactions made up of explicit and implicit messages. Implicit messages are full of emotional content, and not paid attention to in natural resource decision making. FST is an approach for looking at person to person interactions, of which emotion is an influential component. The context of this qualitative study is the Forest Service. Data are drawn from taped interviews, management meetings, public hearings, and workshops, written documents, policies and regulations. Situations of conflict are selected as illustrations of decision-making offering potential to study the influence of the manager's emotions on decision-making using FST.

A LONGITUDINAL ANALYSIS OF TREE VOLUME DATA FROM STEM-ANALYSIS MEASUREMENTS. Fasheng Li and Lianjun Zhang, Faculty of Forestry, 320 Bray Hall, SUNY College of Environmental Science and Forestry, Syracuse, NY 13210.

     Longitudinal data or repeated measurements over time refer to data sets with multiple measures of a response variable on the same experimental unit or subject over time (e.g., permanent-plot remeasurements and stem-analysis data). Two measurements taken at adjacent time points are more likely correlated than two measurements taken several time points apart. Misspecification of covariance structure for temporally correlated data will produce biased standard error estimators, consequently, affecting the hypothesis tests, and confidence and prediction intervals of the model and its coefficients. This project applied mixed-effects models to analyze Hondo spruce (Picea jezoensis Carr.) tree volume data from stem-analysis measurements collected in northeastern China. A repeated measures model and a random coefficients model were used to relate the logarithm of tree volume to the logarithm of tree diameter and total height. Various covariance structures were utilized to model the within-subject correlations. The Likelihood Ratio Test (LRT) showed that all covariance structures used were highly significant, indicating a significant gain in model fitting over incorrectly using Ordinary Least-Squares (OLS). Akaike's Information Criterion (AIC) revealed that the power spatial covariance structure was the best for the repeated measures model, and the completely general (unstructured) covariance structure was the best for the random coefficients model. These two covariance structures increased AIC up to 51% and 58%, respectively, compared to the OLS model. The regression coefficients and their standard errors in parentheses for the OLS model and the two mixed-effects models are:

Model 1. Fixed-Effects Model using OLS:
Log(Volume) = -9.31 + 1.21*Log(DBH) + 1.52*Log(HT)
                         (0.054)      (0.056)                (0.072)

Model 2. Mixed-Effects Model - Repeated Measures with power spatial covariance structure:
Log(Volume) = -9.19 + 0.70*Log(DBH) + 1.997*Log(HT)
                         (0.098)      (0.066)                (0.095)

Model 3. Mixed-Effects Model - Random Coefficients with unstructured covariance matrix:
Log(Volume) = -9.43 + 1.12*Log(DBH) + 1.66*Log(HT)
                         (0.107)      (0.143)                (0.170)

RIPARIAN ZONE FLOWPATH DYNAMICS IN A SMALL HEADWATER CATCHMENT. Brian L. McGlynn, Faculty of Forestry, 208 Marshall Hall, Jeffrey J. McDonnell, Faculty of Forestry, 207 Marshall Hall, James B. Shanley, USGS, Montpelier VT, and Carol Kendall, USGS Menlo Park, Ca.

     The hydrology of the near-stream riparian zone is poorly understood. We examined the spatial and temporal aspects of riparian flowpath convergence during snowmelt in a headwater catchment within the Sleepers River watershed in northern Vermont. A transect of 15 piezometers along a flowpath were sampled for Ca, Si, DOC, other major cations, and d¹⁸O. Daily piezometric head values reflected variations in the stream hydrograph induced by melt and rainfall. The riparian zone exhibited strong upward discharge gradients. An impeding layer was identified between the till and surficial organic soil. Soil water solute concentrations increased toward the stream throughout the melt. Ca concentrations increased with depth and DOC concentrations decreased with depth. The concentrations of Ca in all piezometers were lower during active snowmelt than during post-melt low flow. Ca data suggest snowmelt infiltration to depth, however, only upslope piezometers exhibited snowmelt infiltration and consequent low d¹⁸O values, while d¹⁸O values varied less than 0.5 permil in the deep near stream piezometers throughout the study period. Ca and d¹⁸O values in upslope piezometers during low streamflow were comparable to Ca and d¹⁸O in near stream piezometers during high streamflow. The upslope soil water Ca and d¹⁸O may explain the deep near-stream Ca dilution and consistent d¹⁸O composition. The temporal pattern in Ca and d¹⁸O suggest that upslope soil water moves to the stream via a lateral displacement mechanism that is augmented by pronounced layering in the subsurface. Snowmelt thus initiates the flux of older stored low Ca soil water to depth in the near stream zone. Our work suggests that combined use of chemical and isotopic tracers and hydrometric techniques is vital for the resolution of complex flowpaths in the near-stream zone.

THE EFFECT ON SOIL pH AND BASE CATION STATUS OF SAMPLE STORAGE AND PROCESSING. WHY DO WE CARE?  David McMillan, Faculty of Forestry, 342 Illick Hall, SUNY College of Environmental Science and Forestry, Syracuse, New York 13210 (Major Professor: Ruth Yanai).

     The measurement of long-term changes in soil pH and base cations offers us an opportunity to assess the impacts of anthropogenic stresses on forested ecosystems. Long-term changes in soil pH may be subtle and difficult to detect given inherently variable soil conditions and the strong buffering capacity of most forest soils. However, these changes may be biologically very significant due to the logarithmic increases in hydrogen ion concentrations indicated by small changes in pH. For this reason, it is important to be able to detect small changes in soil pH. Long-term changes in base cation status may reflect the impacts of acid rain on forest ecosystems. While these changes are not always as subtle as soil pH, any change introduced through storage and processing may obfuscate environmental changes.
     Soil pH has been called a master variable because it affects many of the physico-chemical properties of the soil. These changes are not always manifested in forest ecosystems, but in neighboring less well buffered aquatic systems where aluminum toxicity, a by-product of increasingly acidic conditions in the forest floor, occurs. Small changes introduced in sample storage and the processing stage of data collection may mask, or alternatively, exaggerate pH change over time.
     When conducting long-term studies, it common practice to archive samples for future paired analyses with the following round of samples. In a study of long-term change (1930 – 1997) in the acid-base status of the forest floor from the central Adirondacks, we have archived samples from an earlier study conducted in the mid-1980s. We are currently analyzing these samples together with the samples collected during our study (1995 – 1997). Samples from our study were brought into the lab still somewhat moist within 2 days of collection. Samples were then split and processed accordingly: The majority of the samples were air dried, but sub-sets were frozen and freeze dried. We evaluated the effect on soil pH and base cation status of storing samples air dried, frozen and freeze dried. Prior to analysis, samples are often oven dried at temperatures that range from 65 to 100 ^oC. We also examined the effect on soil pH and base cation status of oven drying samples at both 65 and 100 ^oC. Preliminary data suggest that base cation concentrations are quite robust to changes in storage and sample processing, but that pH is affected.

SCREENING WILLOW AND POPLAR BIOMASS PRODUCTION CLONES FOR INSECT RESISTANCE. Erik E. Nordman, Daniel J. Robison, and Lawrence P. Abrahamson, 211 Marshall Hall SUNY College of Environmental Science and Forestry, Syracuse, NY (315) 470-6775, eenordma@mailbox.syr.edu.

     Wood biomass from trees is a potentially useful source of fuel. Whether co-fired with coal or processed into a gas, burning wood fuels emits less sulfur emissions than fossil fuels and is a renewable resource. There are many different willow and poplar clones with utility for biomass production. Damage by herbivorous insects can significantly reduce production of biomass. In the field, these plants will be exposed to multiple pest complexes. It is helpful to understand the resistance and susceptibility of clones to probable pests before making planting decisions and in support of tree improvement efforts.
     Laboratory feeding assays tested the preference of seven insect pests for 19 willow and 6 poplar clones. Defoliation of branch tips was measured using a leaf area meter. Results were standardized to leaf area consumed per insect per day.
     Significant differences in resistance/susceptibility were found between clones. All six of the poplars and willow clones SH3 and SP3 suffered little feeding damage by these insects. Willow clones S185, SV1, and S25 were among those most preferred. Clones may differ in their resistance to feeding by generalist and specialist insects.
     A greenhouse study was conducted to test the hypothesis that mosaics of different clones may help reduce damage by pests. Boxes were planted with only SV1, only SH3, both SH3 and SV1 together in two large blocks ("mixed block"), and both SH3 and SV1 together in alternating rows ("mixed rows"). Leaf area consumed by gypsy moth was measured for each treatment. Statistically significant differences in the leaf areas eaten were found. The boxes containing the "mix block" pattern suffered the most defoliation and those with only SH3 the least. The pure SV1 pattern had slightly less defoliation than the mixed block. The "mixed rows" pattern had significantly less defoliation than both the pure SV1 and mixed block, but more than the pure SH3.
     The results of these experiments will support the development of screening techniques for evaluating host plant resistance and making outplanting -deployment decisions.

MODELING LOCAL GOVERNMENT DECISION-MAKING IN WATERSHED MANAGEMENT. Timothy Schaeffer, Graduate Program in Environmental Science, and Valerie Luzadis, Faculty of Forestry, 320 Bray Hall, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210.

     Many factors influence the decisions made by local government officials in rural areas. These include the types of decision-making strategies employed by the officials, the contexts in which they are forced to act, the limitations placed upon them by the nature of governmental structures, their individual characteristics, and the sources upon which they rely for information. The cumulative effects of these factors will be expressed in the form of a model of local government decision-making. This poster presents and explains the components of the model as an organizing concept.
     While past research has examined local government decisions in narrowly focused areas such as zoning and infrastructure, a broader perspective will guide this inquiry. The analysis particularly seeks to determine those characteristics of watershed management which make it similar to or different than other decisions facing local governments. The decision-making model will be tested among officials from New York’s Catskill/Delaware watershed in an attempt to gain an understanding of the relationship among these factors.

FIRST YEAR GROWTH AND DEVELOPMENT OF WILLOW AND POPLAR BIOENERGY CROPS AS RELATED TO PHOTOSYNTHETIC CHARACTERSTICS. Pradeep J. Tharakan¹, Lawrence P. Abrahamson¹, Daniel J. Robison² and Jud G. Isebrands^{3, 1}Faculty of Forestry, 340 Illick Hall, SUNY College of Environmental Sciences and Forestry, Syracuse, NY, ²Department of Forestry, North Carolina State University. Raleigh , NC, ³United States Forest Service. North Central Forest Experiment Station, North Central Forest Experiment Station, Rhinelander, WI.

     Biomass energy is being considered in several countries as an alternative energy source for reducing dependence on fossil fuels and achieving greater energy independence. Economic analyses have shown that increasing yields of these plantations is one of the most important factors that would ensure commercial viability. In order to achieve this objective, an understanding of the eco-physiological characteristics and processes that influence and thereby predict productivity is essential. For example, earlier studies have proved that a linear relationship exists between leaf area index and dry matter production for hybrid poplar and willow. In 1997, a genetic selection trial of forty willow and poplar clones was established by SUNY-ESF¹in central New York. Simultaneously a collaborative study was set up at the USFS North Central Forest Experiment Station, Rhinelander, WI². Periodic morphological measurements of leaf area and biomass partitioning were used to monitor growth and development of ten clones with contrasting morphology and phenology. After the onset of dormancy in the fall, final growth and biomass yield values were obtained for all the forty clones. At the end of the first growing season, mean survival percentages for the clones were 87.14 percent in New York and 97 percent in Wisconsin. Statistically significant differences were found between clones, both in the case of individual tree characteristics (height and diameter growth), stand productivity (leaf area, biomass production) and biomass partitioning. In New York, the maximum leaf area per plant ranged from 0.12m² to 0.53m² for willow clones Pur12, S365, and 1.2 m² in the case of the poplar clone NM6. This poster reports on the preliminary findings at the end of the first growing season.

RESULTS OF A SPECIES SELECTION TRIAL ON THE JOS PLATEAU, NIGERIA. Timothy Volk¹, Andrew Kidd², and Philip Godwill^{3, 1}Faculty of Forestry, 133 Illick Hall, SUNY College of Environmental Science and Forestry, Syracuse, NY 13210 (Major Professor: Lawrence Abrahamson), ² University of Hohenheim, Stuttgart, Germany, ³Plateau State Afforestation Program, Jos, Nigeria.

     In the early 1900s open woodlands covered the Jos Plateau, located in central Nigeria. The development of the tin industry, and its associated population increases, resulted in extensive deforestation of the area during the first half of the 1900s. Refforestation efforts initiated in the late 1940s had limited success. The Plateau State Afforestation Project’s (PTPU) was established in 1987 with the goal of reducing erosion and preventing further soil degradation by raising and distributing tree seedlings. The program’s efforts have focused primarily on Eucalyptus camaldulensis seedlings. The current study assessed the survival and growth of 38 different local and exotic species and provenenaces of trees on three marginal sites -- typical of those available for tree planting.
     Trees were planted in 1992 at 2 x 2m spacing in a completely randomized design with three replications at three locations – Wereng, Kai, and Fobur. Wereng was a particularly harsh site consisting of mine spoils from the tin industry. The spoils have a high proportion of clay and low nutrient contents. Mine spoils cover about 316 km² on the Jos plateau. The soils at Kai and Fobur were very acidic, eroded soils, with low nutrient contents, and limited buffering capacity. The soils on these sites are typical of the land available for reforestation. Precipitation averages 1200mm per year and follows a unimodal pattern. Survival and height measurements were taken after each of the next three rainy seasons.
     The least successful site, Wereng, was established on a tin mine spoil. Survival after three seasons ranged from 0 - 70% at this site. The top ten species had survival rates >50%. Five of the species were from the genus Casaurina, and four were legumes (Acacia auriculiformis, A. nilotica, Leucaena leucocephala, and Albizia lebbeck). Only one provenance of E. camaldulensis was in this group. Heights of the ten best species on this site ranged from 30.9 – 51.2cm after three years. The difficult site conditions and some browsing damage limited growth. Survival at Kai, ranged from 0 - 100% with three species - Albizia lebbeck, Enterolobium cyclocarpum, and Senna siamea - exceeding 80%. Survival of all four provenances of E. camaldulensis was less than 50%. After three years, height growth was the best at this site. Four provenances of E. camaldulensis and two Acacia species had height growth >100cm. Survival at Fobur was similar, with a range of 0 - 100%. Three provenances of E. camaldulensis and two other species - Grevillia robusta and Pterocarpus santaloides - had survival rates greater then 80%. A. auriculiformis, Casaurina cunninghamiana, E. torreliana, and all four provenances of E. camaldulensis attained heights >100cm. Results indicate that other species can be established as successfully as E. camaldulensis on these sites; however, the success of each species varies from site to site. Differences in establishment success and growth, the variety of uses for the species tested, and input from local communities should be taken into consideration before expanding the mix of tree species that are raised and distributed by the PTPU program.

INTEGRATION OF REMOTE SENSING, GIS AND ON-SITE PHOTOGRAPHY FOR MODELING URBAN INFLUENCES ON SUMMER MICROCLIMATE. Yingjie Wang and Gordon M. Heisler, Faculty of Forestry, USDA Forest Service, 5 Moon library, SUNY-ESF, Syracuse, NY 13210. (Major professor: Dr. Lee. P. Herrington).

     The objective of the study was to explore the possibility of using semi-empirical analysis to arrive at algorithms to predict urban air temperatures in below-canopy spaces. The wind and cloud conditions at the National Weather Service meteorological station at the Atlanta airport were assumed to represent the general synoptic weather conditions over the city during below-canopy air temperature measurements in July 1996. The below-canopy measurements were made at three points: in a botanical garden, a residential area, and a downtown location.
     In the analysis, we related air temperature differences between sites to several different kinds of descriptors of city morphology and weather conditions. The morphology descriptors included tree, grass, and impervious cover for a range of distances in the upwind direction from each of the three sites, as derived by analysis of a satellite image. Other descriptors categorized nearby tree and building structure. General weather condition descriptors included estimates of solar radiation inputs and long-wave radiation exchange, an index of atmospheric stability, vapor pressure deficit, and accumulated antecedent precipitation over a range of time periods. The dependent variable was temperature difference (D T) between sites. Statistical procedures were used to explore the relationship between the D T and the independent descriptor variables. The average air temperature were associated with average tree cover as determined from the satellite image. The greater the surrounding tree cover density, the lower the average temperatures. The values of D T were also highly correlated with land cover over a distance of at least 2000 meters.
     The independent variables that we used could be generated for any location on the Earth for which an estimate of surface synoptic weather can be given in terms of cloud cover and wind speed and direction. The empirical methods that were tested here show promise as a means to develop models of the effects of urban structure on below-canopy microclimate at these scales. Required input variables include the distribution of vegetation and non-vegetated surfaces surrounding the points of interest, an assessment of structure immediately surrounding the point from on-site photography, and synoptic weather from a nearby weather station.