Abstract
Background: An urban site index is an approach for identifying site quality for optimal matching of urban tree tolerances to site conditions and for determining the efficacy of soil management actions. The Rapid Urban Site Index (RUSI) was previously developed and found to significantly relate to urban tree performance. However, the RUSI needs further testing to verify its accuracy in other urban tree populations. Furthermore, calibration of the RUSI with parameter weighting and additional parameters might also improve its accuracy. Methods: The objectives of this study are to: (1) evaluate the RUSI in 3 Wisconsin cities; (2) evaluate RUSI parameter weighting models to improve its accuracy; and (3) examine the addition of a labile organic matter indicator to the RUSI for detection of a soil management action. Results: The RUSI was found to significantly correlate to urban tree metrics in 3 Wisconsin cities (r = 0.29 to 0.31; n = 90). Parameter weighting increased significant correlation values between urban tree metrics and the RUSI model (r = 0.24 to 0.37; n = 90). The Solvita® soil respiration test detected differences in soils from a biosolids application (P = 0.0275), and its addition to the RUSI model improved significant correlation values to urban tree metrics (r = 0.27 to 0.38; n = 90). Conclusions: This research demonstrates effective approaches for RUSI refinement. These findings show the RUSI to be a valid approach for urban site assessment and demonstrate how the RUSI can be tailored and refined for use in specific urban tree populations.
INTRODUCTION
Urban Site Assessments
Urban sites and soils are variable and influence tree species selection and performance. An urban site index helps arborists and urban foresters characterize this heterogeneity to increase species diversity of urban forests (Scharenbroch et al. 2017). Urban tree species have a range of site condition tolerances (Sjöman and Nielsen 2010). By planting trees with a low site condition tolerance on high-quality sites, new tree species may be successfully introduced to the urban environment. Trees with high site condition tolerance can then be planted on low-quality sites to maintain and improve forest canopy. An accurate and field-based site index may allow arborists and urban foresters to increase the health and benefit of urban forests.
An urban site index would also aid in the management of urban soils for individual tree performance. Due to the often degraded nature of urban soils, amendments have been found to enhance urban tree performance (e.g., Scharenbroch and Watson 2014). Industry standards recommend, but do not require, soil testing before and after performing management actions to monitor their necessity and impact (American National Standards Institute 2018). Current urban site assessments are limited in their ability to measure the efficacy of soil management actions (Scharenbroch and Watson 2014). Improving these assessments will allow for improved urban tree site management.
Rapid Urban Site Index
Recent efforts to create an urban site index include the Ohio urban site index (Siewert and Miller 2011), the soil quality minimum data set (Scharenbroch and Catania 2012), and the Rapid Urban Site Index (RUSI) (Scharenbroch et al. 2017). The RUSI was based on these previous urban and several nonurban site indices. The RUSI consists of 5 factors and 15 parameters. Factors include climate, urban, soil physical, soil chemical, and soil biological. Climate parameters include precipitation (PPT), growing degree days (GDD), and exposure (EXP). Urban parameters include traffic (TRA), infrastructure (INF), and surface (SUR). Soil physical parameters include texture (TEX), structure (STR), and penetration (PEN). Soil chemical parameters include pH, electrical conductivity (EC), and organic matter (SOM). Soil biological parameters include estimated rooting area (ERA), depth of the A-horizon (HOR), and wet aggregate stability (WAS). Each parameter is measured and scored from 0 to 3 using scoring functions which are described in the Appendix. After development, the RUSI was tested in 7 cities to determine its ability to predict urban tree performance. Initial testing was performed in Boston, MA, USA; Chicago, IL, USA; Cleveland, OH, USA; Springfield, MA, USA; Toledo, OH, USA; Ithaca, NY, USA; and New York City, NY, USA (Scharenbroch et al. 2017). This research showed a significant correlation between the RUSI and urban tree performance across all cities and species tested (P < 0.0001; R2 values of 0.18 to 0.40).
Initial RUSI testing showed the need for refinement to other urban tree populations, parameter weighting, and inclusion of dynamic parameters that would respond to soil management. To date the RUSI has only been tested in a limited number of cities and with a few urban tree species. Research is needed to test the RUSI model’s applicability in other cities and tree species.
The current RUSI assigns equal weights for all 15 parameters, but initial testing identified several parameters which appeared to be better predictors of urban tree performance. These parameters include those associated with soil volume and compaction, such as estimated rooting area (ERA), structure (STR), and wet aggregate stability (WAS). This importance was expected, as many urban tree health issues are due to limited soil volume and compaction (Jim 1998). Soil quality indices often utilize unequal parameter importance with weighting schemes (Andrews et al. 2002). In this approach, parameter weights are assigned based on available data, literature, and expert knowledge (Karlen et al. 2003).
Labile organic matter is a portion of total soil organic matter (SOM) that is readily available for decomposition by soil organisms. Consequently, it is proposed as an ideal indicator of dynamic soil properties, such as nutrient availability, and has been found to be responsive to soil management actions (Sharifi et al. 2008; van der Heijden et al. 2008). A variety of methods exist for determining labile organic matter including direct measures of organic matter pools (Marriott and Wander 2006) or microbial activity (Zou et al. 2005). Particulate organic matter (POM) is a measure of the low-density, sand-sized organic matter (Cambardella and Elliott 1992). Permanganate oxidizable carbon (POXC) is labile organic matter measured with a chemical reaction (Tirol-Padre and Ladha 2004). Total microbial biomass carbon (MBC) and nitrogen (MBN) are the total carbon (C) or nitrogen (N) contained in the microbial biomass pool and are measured with fumigation and extraction. Indirect measurements of labile organic matter include quantifying microbial respiration defined as the CO2 production of microbial communities within a soil sample that is placed in a sealed container (Alvarez and Alvarez 2000). These CO2 levels are often measured by observing a color change in chemical indicators. The inclusion of a more sensitive soil biological indicator may increase accuracy of the RUSI, allowing it to be used to assess soil management actions.
Objectives
This study investigated 3 knowledge gaps in the current RUSI. First, does the RUSI correlate to urban tree performance in other urban tree populations? Second, can customizing the RUSI with parameter weighting increase its correlation to tree growth and health? Third, is the RUSI sensitive to soil management actions and does the addition of a labile organic matter parameter increase this sensitivity? To address these knowledge gaps 3 specific hypotheses were developed: (1) the RUSI will significantly correlate to tree performance in 3 Wisconsin cities; (2) adjusting the parameter weighting will improve the correlation between RUSI and tree performance; (3) the addition of a labile organic matter parameter will increase the RUSI correlation to urban tree performance.
METHODS AND MATERIALS
Description of Study Cities and Plots
This research was conducted in Stevens Point, Green Bay, and Milwaukee, WI, USA. These cities were chosen due to funding available for travel to conduct the research, the cities’ willingness to participate, and the presence of accurate planting and tree inventories. Full descriptions and data on human and tree populations, climate, and native soils are provided in the Appendix. Thirty sample plots were randomly selected in each city from planting data and tree inventories. A target tree age of 5 to 12 years old was selected to avoid trees that might still be under transplant stress. The most common species planted in this age cohort in all 3 cities was Tilia spp. and thus was chosen as the tree species for this experiment. Sample plots were defined as a single tree and the surrounding 9.3-m2 circular or rectangular planting area. In Stevens Point and Green Bay, 15 plots were rectangular shaped between the street and the sidewalk, with the other 15 plots circular shaped (not bound by a sidewalk). In Milwaukee, all of the study sites were rectangular shaped between the street and sidewalk.
Field Assessments
Urban tree performance was assessed by a single primary investigator using urban tree growth and health metrics (Table 1). Tree performance evaluations were done independently of the site assessments to limit bias. The urban tree health metrics included tree condition (TC), tree condition index (TCI), and urban tree health (UTH), as used by Scharenbroch et al. (2017). Tree health was also assessed by measuring leaf chlorophyll contents of 12 leaves per tree using a SPAD meter (SPAD-502, Konica Minolta, Tokyo, Japan). These 12 leaves were collected on 4 sides of the tree from equally distributed branch tips throughout the bottom, middle, and top of the crown. Growth metrics included total tree height measured with a height pole and diameter at breast height, which was measured at 1.37 m and marked to ensure accurate follow-up readings. Crown volume was calculated by measuring the crown radius in each of the 4 cardinal directions and then calculated using the equations presented in Moser et al. (2015).
Mean (n = 90), standard deviations (SD), minimum values, and maximum values for tree diameter (DIA), tree height (HT), tree crown volume (CV), leaf greenness (SPAD), tree condition (TC), tree condition index (TCI), and urban tree health (UTH).
Site quality was assessed by a single primary investigator at each sample plot using the RUSI in the spring and fall of 2017 (Table 2). The RUSI uses climatic, urban, soil physical, soil chemical, and soil biological factors to provide an index (0 to 100) of urban site quality (Scharenbroch et al. 2017). Embedded in each of these main factors are 3 parameters. Individual parameters were assessed in the field and scored on a 0 to 3 scale using the scoring functions described in Scharenbroch et al. (2017). Observed scores were summed, divided by the maximum possible score, and then multiplied by 100 to compute the RUSI score.
Mean (n = 90), standard deviations (SD), minimum values, and maximum values for scores of precipitation (PPT), growing degree days (GDD), exposure (EXP), traffic (TRA), infrastructure (INF), surface (SUR), estimated rooting area (ERA), penetration (PEN), A-horizon (HOR), texture (TEX), structure (STR), wet aggregate stability (WAS), soil organic matter (SOM), electrical conductivity (EC), pH, and Rapid Urban Site Index (RUSI).
Soil Collection, Treatment, and Analyses
During each site visit, 20 soil cores 2.5 cm wide by 15 cm deep were randomly collected throughout each sample plot. Cores were composited by plot, placed in individually labeled plastic bags, and kept on ice in a cooler until being transported to the laboratory where they were then stored at 5 °C until analyses were performed.
Immediately after the first soil sampling, a top dressing of organic biosolids (Milorganite, Milwaukee, WI, USA) was applied by hand at 3 rates. Biosolids are high in carbon, nutrients, microbes, and microbial activity, and thus have been found to stimulate the biological communities and increase decomposition and nutrient mineralization (Sullivan et al. 2006). Application rates based on nitrogen (N) content were chosen in accordance with industry standards on urban tree fertilization (American National Standards Institute 2018). Ten sites per city received the maximum rate of 2.92 kg N 100 m−2, ten sites received the standard rate of 1.46 kg N 100 m−2, and the remaining ten sites received no soil amendment and served as the control.
In the laboratory, each soil sample was sieved through a 6-mm screen for homogenization and removal of coarse material. Soil particle-size analysis was performed using the hydrometer method (Gee and Or 2002) to verify the field assessment of soil texture. The total soil organic matter (SOM) was determined using the loss on ignition method at 360 °C for 6 hours (Nelson and Sommers 1996). The particulate organic matter (POM) was determined following particle size fractionation (Gregorich et al. 2006). Potassium permanganate oxidizable carbon (POXC) was determined colorimetrically (Weil et al. 2003). Potentially mineralizable carbon (PMC) was measured as the amount of CO2 in 0.25-M NaOH traps following a 7-day soil incubation, which was then titrated to a phenolphthalein endpoint using 0.25 N HCl (Parkin et al. 1997). Soil respiration was determined using the Solvita® gel system (Solvita, Woodsend Laboratories, Augusta, ME, USA) which incubates a color gel paddle in a container with a field moist soil sample for 24 hours, after which the paddle color indicates the quantity of CO2 present (Haney et al. 2008). Microbial biomass carbon and nitrogen were determined using a chloroform fumigation and extraction (Vance et al. 1987), using efficiency factors of kN = 0.54 (Joergensen and Mueller 1996) and kC = 0.45 (Beck et al. 1997). After fumigation, samples were extracted using 0.5 M K2SO4 and analyzed for microbial biomass nitrogen and carbon on a PerkinElmer C:N analyzer (PerkinElmer Inc., Waltham, MA, USA). The labile organic matter parameters tested in this study attempt to assess soil biological condition by measuring the microbial biomass (MBC and MBN), microbial activity (PMC and SOLV), or the microbial substrate (SOM, POM, or POXC).
Statistical Analyses
Statistical tests were conducted using SAS JMP 13.2.1 software (SAS Institute Inc., Cary, NC, USA) with significance determined at a 95% confidence level.
To answer the first research question, Pearson product-moment correlation analyses were conducted with the RUSI model and the tree metrics. The R-correlation and P-value statistics were used to evaluate the strength and significance of the correlations.
To answer the second research question, parameter weighting was applied (Table 3). For each model, all weights summed to one. Weights were developed using the data collected during the second sampling period and were tested on data collected during the first sampling period. The weighted RUSI models were compared to the nonweighted RUSI model, which had an equal weight distribution for the 15 parameters.
Property weights for RUSI, weighted RUSI (RUSIw), and organic weighted RUSI (RUSIow) models. Precipitation (PPT), growing degree days (GDD), exposure (EXP), traffic (TRA), infrastructure (INF), surface (SUR), estimated rooting area (ERA), penetration (PEN), A-horizon (HOR), texture (TEX), structure (STR), wet aggregate stability (WAS), soil organic matter (SOM), electrical conductivity (EC), pH, and Solvita® (Solvita, Woodsend Laboratories, Augusta, ME, USA) soil respiration (SOLV).
Parameter weights for the weighted RUSI (RUSIw) were assigned based on limiting factor rank, relative correlation strengths to tree metrics, and data distributions. The 15 parameters were ranked 1 to 15 based on their potential limitation for tree health and growth (Table 4). Parameters that were expected to be more limiting received a lower rank. The R-correlation values for the 15 RUSI parameters and each tree metric were determined (Table 5). Data distributions were examined to determine the mean, standard deviation, minimum, and maximum scores for each of the RUSI parameters (Table 2). Four weighting tiers (and weights) were established: none (0.00), low (0.04), moderate (0.09), and high (0.13). Parameters in the high tier were expected to be limiting, had relatively high correlation to tree metrics, and had relatively wide data distributions. Parameters in the none tier were not expected to be limiting, had relatively low correlation values, and had narrow data distributions. Parameters in the low and moderate tiers fell in between those extremes for these 3 weighting criteria.
Potential limitation to tree health and growth for each of the RUSI parameters. The lower the rank, the more limiting that parameter was expected to be for tree growth and health based on expert opinions and professional experiences of the primary investigators. Precipitation (PPT), growing degree days (GDD), exposure (EXP), traffic (TRA), infrastructure (INF), surface (SUR), estimated rooting area (ERA), penetration (PEN), A-horizon (HOR), texture (TEX), structure (STR), wet aggregate stability (WAS), soil organic matter (SOM), electrical conductivity (EC), and pH.
Pearson R-correlation values (n = 90) of tree diameter (DIA), tree height (HT), tree crown volume (CV), leaf greenness (SPAD), tree condition (TC), tree condition index (TCI), and urban tree health (UTH) with RUSI scores for precipitation (PPT), growing degree days (GDD), exposure (EXP), traffic (TRA), infrastructure (INF), surface (SUR), estimated rooting area (ERA), penetration (PEN), A-horizon (HOR), texture (TEX), structure (STR), wet aggregate stability (WAS), soil organic matter (SOM), electrical conductivity (EC), and pH.
For the third research question, analysis of variance (ANOVA) with Tukey-Kramer Honestly Significant Difference (HSD) testing was used to examine the responses of labile organic parameters (SOM, POM, POXC, PMC, SOLV, MBC, and MBN) as a result of the soil amendment (biosolids) application. Labile organic matter measurements were determined on soils from the spring and fall collections. Percent changes in each of the parameters were computed for each plot. The ANOVA analyses were conducted on data from the fall sampling and the percent change data for each parameter. The labile organic matter parameter that most significantly responded to treatments was included in the RUSI model as a 16th parameter for the organic weighted RUSI model (RUSIow). The RUSIow model with the labile organic matter parameter was then tested for correlation with urban tree condition metrics using the previously described methods.
RESULTS AND DISCUSSION
RUSI Significantly Correlates with Urban Tree Health in Wisconsin
Across all 3 Wisconsin cities, RUSI scores significantly (P ≤ 0.05) correlated with tree condition (R = 0.32), tree condition index (R = 0.29), and urban tree health (R = 0.29)(Figure 1). The RUSI scores were not significantly correlated with leaf greenness (SPAD), diameter, height, or crown volume. These results confirm findings of the original RUSI study (Scharenbroch et al. 2017) and again suggest that the model is a better predictor of tree health, not growth.
Matrix of scatterplots (black dots), data distributions (solid lines), and Pearson R-correlation values (Corr:) among tree diameter (DIA), tree height (HT), tree crown volume (CV), leaf greenness (SPAD), tree condition (TC), tree condition index (TCI), urban tree health (UTH), RUSI, weighted RUSI (RUSIw), and organic weighted RUSI (RUSIow) models. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
Correlation strength between RUSI scores and urban tree health metrics tended to be weaker than in the previous study (Scharenbroch et al. 2017). The observed tree performance and site quality ranges were narrower in this study (RUSI scores = 51.0 to 81.1) compared to the initial study (RUSI scores = 30.0 to 82.2). This reduced variability truncates the data distribution and may have led to a reduction in the strength of correlation. The limited geographic extent of the current study resulted in a decreased range of climate factors. Initial study sites occurred in 4 states with mean annual temperatures ranging from 6.7 to 12.9 °C (US Climate Data 2018), mean annual precipitations ranging from 830 to 1,219 mm yr −1 (US Climate Data 2018), and growing degree days ranging from 2,808 to 3,948 (Growing Degree Days 2014). Sites in this study occurred in a single state with mean annual temperatures ranging from 6.7 to 8.8 °C (US Climate Data 2018), mean annual precipitations ranging from 830 to 876 mm yr −1 (US Climate Data 2018), and growing degree days ranging from 2,378 to 2,696 (Growing Degree Days 2014). The observed decrease in the variability of climate factors related to the limited geographic extent of this study may have reduced the RUSI models’ ability to predict tree performance.
Weighting RUSI Parameters Improves Correlation to Urban Tree Health
Parameter weighting improved correlation to urban tree condition. The weighted RUSI model (RUSIw) improved the correlation to all urban tree health and growth assessments compared to the nonweighted model (RUSI)(Figure 1). The RUSIw was also significantly correlated with leaf color, which was not the case for the nonweighted RUSI model.
The RUSIw model applies the greatest weights to the depth of the A-horizon, soil texture, and soil structure. Surface condition, estimated rooting area, and penetration resistance were assigned the next greatest weights in the RUSIw model. The RUSIw applies no weight to the precipitation and growing degree days scores. The weighting in the RUSIw was developed based on limiting factor rank, relative correlation strengths to tree metrics, and data distributions. An urban tree manager will likely have a reasonable understanding of the limiting factors for the trees they are managing. They can, and should, utilize that information to assign greater weights to those parameters that are likely driving site quality differences. Furthermore, some of the RUSI parameters may not be important for separating site quality differences for a particular population of urban trees. This was the case for the current study in which all trees were in a similar climate with similar precipitation and growing degree days. Consequently, the parameter weighting removed those parameters from the RUSI model by assigning a 0.000 weight.
Adding a Labile OM Parameter Improves RUSI’s Sensitivity to Soil Management
The Solvita® (SOLV) respiration test was significantly greater with the high (33.9 mg kg−1 d−1) biosolids application rate compared to the null (32.4 mg kg−1 d−1) (Table 6). The SOLV responses for the low biosolids application rate were between the high and null although not significantly different from either (Table 6). Significant differences for the other 6 labile organic matter parameters were not detected among the treatments. It is unclear why significant differences were not detected with these other labile organic matter measurements. The standard errors appear relatively high for these measurements compared to SOLV, possibly suggesting that site variability may have masked treatment differences.
Mean (n = 30) and standard error of the mean (SE) for total soil organic matter (SOM), particulate organic matter (POM), potassium permanganate oxidizable carbon (POXC), potentially mineralizable carbon (PMC), Solvita® (Solvita, Woodsend Laboratories, Augusta, ME, USA) soil respiration (SOLV), microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN).
The Solvita® test appears to be the most accurate and most practical measurement for detecting response to soil management—in this case, biosolids amendment. All of the other labile organic matter measurements (MBC, MBN, SOM, PMC, POM, and POXC) involve laboratory analyses that are beyond the capabilities of typical urban foresters and arborists. Conversely, the Solvita® test is practical and can easily be utilized and interpreted by an urban tree manager without the need for expensive laboratory testing. Materials to conduct the Solvita® test can be purchased for approximately $10 (US dollars) per sample, and the test is conducted over a 24-hour period.
An organic weighted model (RUSIow) was created by adding SOLV as a 16th parameter and weighting it in the tier of greatest importance (Table 3). Improvements in correlation strength to urban tree health metrics were found with the RUSIow compared to the original RUSI model. Slight improvements in correlation strength to urban tree health were also observed for RUSIow compared to the RUSIw model. This finding was expected because the existing RUSI parameters and tree responses are likely not dynamic enough to respond to a soil amendment over the course of several months. The addition of the dynamic labile organic matter parameter that is more sensitive to a soil amendment did appear to provide an early indication of potential site quality improvements leading to improved tree growth and health.
CONCLUSION
This study showed that the RUSI can be used in Wisconsin to relate urban site conditions and urban tree performance. The study also demonstrated the value of parameter weighting to improve the RUSI model. Lastly, the study identified a labile organic matter parameter that might be used to make the RUSI model more dynamic and detect soil management.
It is important to recognize that the RUSI model was developed as an approach, not a “one-size-fits-all” model. The approach allows for sensible and meaningful tailoring of RUSI to specific site conditions and urban tree populations. The RUSI approach involves understanding the site conditions affecting an urban tree population, tailoring an assessment to those conditions, assessing those conditions, and then evaluating the results for management. The results from the current study demonstrate the value of parameter weighting, adding or removing parameters, and using labile organic matter for improving the RUSI model for more accurate site assessments for urban trees.
It is also important to recognize that nonsite factors also influence the health and growth of urban trees. Understanding site conditions is important for urban tree management, but other factors (e.g., nursery practices, pruning) also impact urban tree condition. Significant but relatively low correlations between RUSI scores and urban tree condition parameters in this study provide evidence for this statement.
Urban forests, soils, and sites are diverse. Our understanding for assessing those conditions, and then evaluating the results, is evolving. Future work on the RUSI model should be directed at tailoring and testing the RUSI model in more urban tree populations, soils, and site conditions. The corresponding author of the current study has begun working with individual urban forest managers to develop tailored RUSI models for specific cities. Data from these case studies will be critical for improving the RUSI model and further demonstration of how it can be practically applied. If interested in participating in this effort, please contact the corresponding author.
ACKNOWLEDGMENTS
This study was funded by a Hyland R. Johns Grant (No. 16-HJ-01) from the Tree Research and Education Endowment (TREE) Fund, as well as funding from the Wisconsin Arborist Association, the College of Natural Resources at University of Wisconsin–Stevens Point, Stevens Point, Wisconsin, and The Morton Arboretum, Lisle, Illinois. All field data was collected by Luke Scheberl. Laboratory work was performed by Joel Gebhard and Luke Scheberl and was assisted by Alyssa Gunderson at the University of Wisconsin–Stevens Point.
Appendix.
DESCRIPTION OF STUDY AREAS
Stevens Point, WI, USA (44.523483, −89.574814) has a total population of 26,670 people (US Census Bureau 2017) with an elevation of 331.9 m, average precipitation of 83.0 cm, and an average temperature of 6.7 °C. Native soils in Stevens Point are described as a Plainfield-Friendship association, which is moderate to excessively well-drained and formed in deep sandy glacial deposits (USDA NRCS 1978). Stevens Point has approximately 7,230 city trees distributed among 47 species, with dominant genera of Acer (25%), Fraxinus (15%), Malus (7%), Tilia (6%), and Pinus (6%)(Davey Resource Group 2010).
Green Bay, WI, USA (44.513287, −88.01326) has a total population of 104,779 people (US Census Bureau 2017) with an elevation of 177.0 m, average precipitation of 74.9 cm, and an average temperature of 6.7 °C. The native soils in Green Bay are described as an Oshkosh-Manawa association. These soils are well-drained to somewhat poorly drained with sand and loamy subsoil (USDA NRCS 1974). Green Bay has approximately 35,000 city trees, with dominant genera of Acer (31%), Fraxinus (21%), Tilia (19%), and Gleditsia (9%)(Freberg 2016).
Milwaukee, WI, USA (43.04181, −87.90684) has a total population of 599,164 people (US Census Bureau 2017) with an elevation of 188.0 m, average precipitation of 87.4 cm, and an average temperature of 8.7 °C. The native soils in Milwaukee are described as an Ozaukee-Marley-Mequon association. These soils are well-drained to somewhat poorly drained with clay subsoils (USDA NRCS 1971). Milwaukee’s total tree population is approximately 3,377,000 trees, with dominant genera of Rhamnus (23%), Acer (20%), Fraxinus (17%), Ulmus (6%), and Gleditsia (6%)(i-Tree 2008). It should be noted that native soils in all 3 cities have been significantly altered by urbanization.
TREE PERFORMANCE METRICS
Qualitative tree health was assessed using 3 metrics: tree condition (TC), tree condition index (TCI), and urban tree health (UTH). Equations and scoring functions for these metrics are as follows.
Tree condition (TC) scores were calculated following Scharenbroch et al. 2017 (Equation S1; Table S1). This method is a quick assessment of the relative growth and signs/symptoms of stress. It provides a 0 to 3 rating based on an ocular estimation of the presence of leaves and their condition, bark condition, and growth rate. The tree condition is considered dead when more than one-half of the crown is dead and bark is sloughing off. Trees are in poor condition when less than half of the crown is dead and there are signs of severely stunted growth. Trees are in fair condition if they have reduced growth, minor dieback, and/or are chlorotic. Trees are in good condition when there are no signs of stress present and exhibit high growth rates.
Equation S1.
Parameters and scoring function for the tree condition (TC) model.
Tree condition index (TCI) scores were calculated using the modified Webster (1979) method first used by Scharenbroch and Catania (2012)(Equation S2; Table S2). This method provides a rating on a 1 to 5 scale on the tree’s trunk, crown, and roots. The trunk factor rates how sound the tree is and the presence of damage or decay and its extent. Crown is the tree’s canopy density and balance or evenness. The roots factor is the presence of proper rooting habits represented by a large, evenly spaced structural root flare.
Equation S2.
Parameters and scoring function for the tree condition index (TCI) model. Adapted from Webster (1979).
Urban tree health (UTH) scores were calculated following the methods developed by Jerry Bond (2012)(Equation S3; Table S3). This method provides a 0 to 5 scale rating the tree’s live crown ratio, opacity, vitality, growth, and quality. The live crown ratio is the percent live crown height to the total live tree height. Opacity is the percent of light visibly blocked by branches, foliage, and reproductive structures of the actual live crown. Vitality is the percent of the upper crown that is free from recent mortality. Growth is the 3-year average terminal shoot extension on 3 random branches with the same sun exposure that have not been pruned or damaged. Quality measures the percent of the upper crown that is free from necrotic, chlorotic, or undersized foliage.
Equation S3.
Parameters and scoring function for the urban tree health (UTH) model. Adapted from Bond (2012).
RAPID URBAN SITE INDEX
Rapid Urban Site Index (RUSI) scores were calculated following Scharenbroch et al. (2017)(Equation S4; Table S4). A description of each of the 15 RUSI parameters is as follows.
Equation S4.
Parameters and scoring functions for the Rapid Urban Site Index (RUSI) model. Precipitation = PPT; growing degree days = GDD; exposure = EXP; traffic = TRA; infrastructure = INF; surface = SUR; texture = TEX; structure = STR; penetration = PEN; A-horizon = HOR; estimated rooting area = ERA; wet aggregate stability = WAS; soil organic matter = SOM; electrical conductivity = EC; sand = S; sandy clay = SC; silt = SI; sandy loam = SL; clay = C; loam = L; clay loam = CL; loamy sand = LS; sandy clay loam = SCL; silty clay = SIC; silt loam = SIL; silty clay loam = SICL; coarse fragment (>2 mm in diameter) = CF; massive = M; single-grained = SG; platy = PL; angular blocky = ABK; subangular blocky = SBK; granular = GR.
The climate factors of the RUSI model include precipitation (PPT), growing degree days (GDD), and exposure (EXP). For PPT and GDD scores, it is suggested to use the most recent, practical, and accurate local data available. The PPT score was calculated using data acquired from US Climate Data (2014). If irrigation was present on the site, then the PPT score was increased one point to a maximum score of 3. The GDD score is a measure of heat accumulation. The GDD units are calculated by mean daily temperature (maximum plus minimum divided by 2) minus base temperature (10 °C). The GDD units are summed for the year for annual GDD. The Growing Degree Days smartphone application was used to determine the GDD score for each location (Growing Degree Days 2014). The start date was 2016 January 01 and the end date was 2016 December 31 and the GDD50 was selected as the base temperature. The free application returns the GDD for the most recent 2 years, and a mean of this value was used to score GDD. The EXP score was assessed in the field based on the number of faces of the tree that were exposed to full sun.
The urban factors in the RUSI model are traffic (TRA), infrastructure (INF), and surface (SUR). The TRA score was based on the number of lanes and amount of parking available on the street. More lanes and less parking indicate more traffic, likely faster-moving automobiles, and more of an “urban” impact (e.g., road salts, recent soil disturbance) on the site. The INF score was based on the distance to the nearest hard-space or building from the main stem of the tree. The SUR score was based on the type of ground covering for the majority (greater than 50%) of the rooting area for the tree.
Soil physical factors include texture (TEX), structure (STR), and penetration (PEN). Texture reflects the relative particle size distribution and is determined by the feel method. Structure is the shape of the soil aggregates present. Methods for assessing soil texture by the feel method and structure shape are described in Schoeneberger et al. (2012) and Scharenbroch and Watson (2014). Penetration was assessed by recording the depth and ease that the core sampler went into the soil when collecting samples.
The soil chemical factors were pH, electrical conductivity (EC), and soil organic matter (SOM). Soil pH and EC were measured on homogenized subsamples at each site using a handheld combination pH/EC meter. For this research, the Oakton PCTestr 35 (OAKTON Instruments, Vernon Hills, IL, USA) was used. Soil organic matter was estimated using the Color Chart for Estimating Organic Matter in Mineral Soils of Illinois (Alexander 1971).
The soil biological factors were estimated rooting area (ERA), depth of the A-horizon or topsoil (HOR), and wet aggregate stability (WAS). Estimated rooting area was an evaluation of the surface permeable space for root growth. The ERA score was increased by 1 to a maximum of 3 if a breakout area of at least 50 m2 was present within 2 m of the tree. The HOR was the depth of the A-horizon or topsoil via visual inspection. The A-horizon was distinguished by darker color, a more well-developed structure, and a greater abundance of fine roots compared to the underlying horizon. Wet aggregate stability is an estimate of the strength of the aggregates to resist degradation (Nimmo and Perkins 2002). A modified field method was used to assess WAS. A total of 5 aggregates 2 to 5 mm in diameter were placed on a 1-mm screen. The aggregates were soaked in water for 30 seconds. After 30 seconds the screen was agitated (i.e., a vigorous swirl) for another 30 seconds. The number and amount of aggregates left after the soak and swirl were volumetrically estimated and scored.
Footnotes
Conflicts of Interest:
The authors reported no conflicts of interest.
- Copyright © 2023, International Society of Arboriculture. All rights reserved.
LITERATURE CITED
Jump to section
Related Articles
Cited By...
- No citing articles found.