Northeastern Area

Forest Health Protection

Forest Health Management

Urban Forest Structure:
The State of Chicago's Urban Forest
David J. Nowak, Research Forester,USDA Forest Service, Northeastern Forest Experiment Station, Chicago, IL

Abstract
Information on urban forest structure (species composition, tree size and location, etc.) provides the basis for understanding the urban forest functions that affect urban inhabitants and for improving management to maximize the environmental and social benefits of urban forests. There are an estimated 4.1 million trees in the City of Chicago, with an estimated 50.8 million trees across the study area of Cook and DuPage Counties. Most of these trees are small and on institutional, residential, and vacant lands.

Relatively short-lived pioneer species contribute significantly to the Chicago area urban forest. The invasive buckthorn is the most common tree, accounting for 12.7 percent of the total tree population but only 2.9 percent of total leaf-surface area. Other common trees are green/white ash, Prunus spp., boxelder, and American elm. The most dominant species in leaf area are silver maple, green/white ash, white oak, American elm, and boxelder. Native pioneer tree species (e.g., boxelder, green ash, willow, cottonwood) and buckthorn are most prevalent on land uses with minimal or naturalistic management (e.g., forest stand conditions) and may constitute an even more important component of the Chicago area's urban forest structure in the future.

Streets trees are a significant part of Chicago's landscape, accounting for 10 percent of the city's trees and 24 percent of the total leaf-surface area. Street trees are less significant in more suburban or rural areas. Common ground surfaces in the study area are maintained grass, tar, herbaceous cover (e.g., crops) and buildings. This paper presents formulas for estimating the leaf-surface area of urban trees and discusses the importance of urban forest structure, particularly leaf-surface area, and how managers and planners can direct urban forest structure to a desired outcome.

Introduction
Urban forest structure is the three-dimensional spatial arrangement of vegetation in urban areas (species composition, tree size and health, number and location of trees, etc.). Information on this structure provides the basis for understanding the urban forest functions that affect urban inhabitants (air temperature modifications, human stress reduction, air pollution mitigation, improved sense of community, etc.) and for improving management to maximize the environmental and social benefits of urban forests.

Urban forest structure is determined by three broad factors: urban morphology, which creates the spaces available for vegetation; natural factors, which influence the amounts and types of biomass likely to be found within cities; and human management systems, which account for intraurban variations in biomass configurations according to land use distributions (Sanders 1984). There are significant variations in urban forest structure both within and among cities. Aerial photographic analyses of urban tree canopy cover reveal that tree cover varies between 5 and 60 percent among land-use types within four eastern U.S. cities, while overall urban tree cover ranged from 24 to 37 percent among the cities (Rowntree 1984).

There has been little ground-based research evaluating the urban forest structure of an entire city. Many researchers have evaluated the street-tree component of the urban forest (Impens and Delcarte 1979; Richards and Stevens 1979; Dawson and Khawaja 1985; Talarchek 1985; Jim 1986; Stevens and Richards 1986; McPherson and Rowntree 1989) or limited portions of non-street tree urban forests (e.g., Derrenbacher 1969; Schmid 1975; Whitney and Adams 1980; Airola and Buchholz 1982; Boyd 1983; Buhyoff et al. 1984; Dorney et al. 1984; McBride and Froehlich 1984; Miller and Winer 1984; Richards et al. 1984; Schroeder and Green 1985; Schroeder and Cannon 1987; Profous et al. 1988, Profous and Rowntree 1993), but ground-based urban forest structural analyses of an entire urban area have been conducted only for the Los Angeles Basin (Horie et al. 1991) and Oakland, California (Nowak 1991). The Los Angeles study focused on leaf biomass and volatile organic emissions from vegetation. The Oakland study focused on variations in urban forest structure and its overall effect on forest compensatory value, atmospheric carbon storage and volatile organic emissions from vegetation (Nowak 1993a,b).

Since many environmental functions are related to leaf-surface area (e.g., reductions in air temperature, air pollution removal, volatile organic emissions, carbon dioxide sequestration), understanding the leaf-area contribution of various tree species is important to urban-forest researchers, managers and planners. The measure of tree-species dominance reflects the relative contribution of a species to the overall leaf-surface area of the forest. Species with the greatest proportion of leaf-surface area are the most dominant and likely have the greatest influence on the local environment. Many social benefits of trees also may be related to leaf-surface area. For example, large trees contribute more scenic beauty than smaller ones (Buhyoff et al. 1984; Schroeder and Cannon 1987).

Leaf-area indices (LAI) are another common means of comparing the relative contribution of leaf area among different areas or tree species on an equal-area basis. LAI is the total leaf area (one surface only) divided by the ground area occupied by the plant. A LAI of 4 means that for every square meter of ground below the tree canopy, 4 m2 of leaves lie above it. Net primary productivity (individual plant growth) of forests is greatest at a LAI of approximately 4. However, the yield (growth) per unit of ground area is low in such open stands (LAI < 4). Maximum gross productivity usually occurs at LAI values of 8 to 10 (Kramer and Kozlowski 1979); LAI varies with plant size, age, spacing, species, and site characteristics.

Typical LAI's are 10 to 11 for tropical rain forests, 5 to 8 for deciduous forests, and 9 to 11 for boreal coniferous forests (Barbour et al. 1980). The LAI of some Piedmont hardwood forests range from 4.5 to 7.4 (Hedman and Binkley 1988), and LAI's of a subalpine Sierra Nevada forest range from 3.6 to 11.7 (Peterson et al. 1988). Little research has been conducted on the LAI of urban trees. Data from individual urban trees and shrubs in Warsaw, Poland, show LAI's for individual trees ranging from 1 to 15 with an average LAI of individual trees for various areas in Warsaw of 3.5 to 4.8 (Gacka-Grzesikiewicz 1980).

Because information is scarce on the variation in forest structure within urban areas, on how urban forest structure combines to create an urban forest ecosystem, and on leaf-surface area of urban trees, the objectives of this study were to: 1) quantify urban forest structure and its variation by land-use type in the Chicago area; and 2) measure the leaf-surface area of individual open-grown urban trees and develop predictive equations of leaf-surface area to estimate tree species dominance in the Chicago area. This information will be used to reveal key urban forest characteristics and aid in quantifying various environmental functions (see Nowak 1994a,b: Chapters 5 and 6, this report).

Methods
Study Area:
The study area encompasses Cook and DuPage Counties (3,350 km2; 1,292 mi2) and contains nearly six million people. To reveal regional variation within the Chicago area, the study area was subdivided into the City of Chicago, Cook County exclusive of Chicago (hereafter referred to as suburban Cook County), and DuPage County (Figure 1). Chicago is the most densely populated sector, accounting for 18 percent of the entire study area and 47 percent of the total population. Suburban Cook County contains 56 percent of the study area and 40 percent of the total population, and many of the older suburban communities in the Chicago region. DuPage County is the least densely populated, most agricultural, and most rapidly urbanizing sector within the study area. It contains 13 percent of the population and occupies 26 percent of the study area. Tree crowns cover an average of 11 percent of the land area in Chicago, 23 percent in suburban Cook County, and 19 percent in DuPage County (McPherson et al. 1993). Crown cover also varies by individual land-use types within each sector (Table 1.

Ground Sampling of Vegetation
Urban vegetation and other surface data were collected on 652 randomly located plots established as a sample of grid points (213 plots in Chicago, 222 in suburban Cook County and 217 in DuPage County). Because the focus of this study is on urban trees, the number of sample plots allocated to each land-use type was proportional to the estimated tree cover in the land use.1

Plot structure varied by land-use type.2 Residential plots were subdivided into smaller ground units, whose area was measured to aid in estimating ground-surface cover (to the nearest 5 percent). Building size on each residential plot was measured and building-surface characteristics were noted. The amount of ground area occupied by various materials (tar, cement, buildings, small structures, other impervious material, maintained or unmaintained grass, shrubs, soil, herbaceous, rock, duff, water, wood) was measured or estimated on each plot.

Table 1. Mean percent tree cover and standard error by land-use type in Chicago, suburban Cook County, DuPage County, and entire study area.

The size and species of individual shrub masses were recorded (length, width, height). On every 10th plot measured, stem diameters of individual shrubs at 15 cm (6 inches) above groundline were measured. Data were collected on 8,996 trees and shrubs that were growing in tree form (i.e., relatively large open-grown individuals). The data included species, trunk diameter at breast height (d.b.h. diameter at 1.37 m or 4.5 ft), total tree height, height to base of crown, crown width, crown shape, percent of crown occupied by leaves, tree location (street-tree locations between sidewalk and road, or on median, were noted), and condition. Estimates of tree condition were based on foliage characteristics. Trees were rated as excellent if less than 5 percent of the crown showed dieback or leaf discoloration. Other ratings were good (5 to 25 percent dieback or discoloration), moderate (26 to 50 percent), poor (51 to 75 percent), dying (76 to 99 percent), and dead (no leaves).

Plot information was combined to produce aggregate estimates on vegetation and other urban-forest attributes by land-use type in each sector of the study area (Gerald Walton, USDA Forest Service, 1992, pers. commun.).

Leaf Area of Urban Trees
To estimate leaf-surface area of urban trees, data were collected from 54 healthy, open-grown park trees in Chicago that were selected specifically for their excellent condition (10 American elm, 10 green ash, 10 hackberry, 10 honeylocust, and 14 Norway maple). The crown height (base of crown to crown top) of sampled trees ranged from 3.4 to 9.1 m (11.2 to 29.9 ft); crown width ranged from 4.1 to 12.0 m (13.5 to 39.4 ft) and individual LAI's ranged from 0.7 to 12.5. The volume of each tree crown was mapped (including areas devoid of leaves) using a telescoping pole.3 Crown height and distance from the tree base were measured at crown boundary points every 1.5 m (5 ft) vertically and at every 45° angle radially (i.e., eight points around the tree at every 1.5 m vertically). Ten 0.4 m3 (14.1 ft3) samples of foliage were collected from random points within the tree crown using a high-lift truck.4 The number of leaves per sample were counted and approximately 30 leaves were randomly subsampled for analysis of leaf area. For samples with 50 leaves or less, all leaves were analyzed for leaf area. Individual leaf areas were measured with a leaf-area meter (CID Inc., Conveyor Area Meter C1251). Average sample leaf area (one-surface only) per unit crown volume (m2/m3) was extrapolated using the total crown volume (m3) to estimate total leaf area for each tree. Following leaf-area analyses, all leaves were dried at 65°C (149°F) for 24 hours and then weighed.

Total leaf-surface area for smaller urban trees was obtained from Gacka-Grzesikiewicz (1980). Data from 34 trees (12 species) that ranged in crown height (H) from 0.7 to 12.8 m (2.3 to 42.0 ft) and in crown width (D) from 0.5 to 4.6 m (1.6 to 15.1 ft) were combined with field data on leaf-surface area of individual trees to produce equations for estimating total leaf-surface area of individual urban trees based on crown parameters. Other variables included in the predictive equations were a factor for leaf-surface area based on the outer surface of the tree crown (S= nD(H + D)/2) (Gacka-Grzesikiewicz 1980) and average shading coefficients for individual species (percent sunlight intercepted by foliated tree crowns) (McPherson 1984).

Least-squares linear regression was used to produce two regression equations for estimating total leaf area of individual urban trees. One equation included shading coefficients, the other excluded shading coefficients to aid in estimating leaf area of species for which shading coefficients are unknown (40 percent of the total population). Because logarithmic equations slightly underestimate leaf area (Crow 1988) a correction factor of one-half of the estimated variance of the estimate was added to the untransformed value (y = ex + var(x) /2) for each equation (G. Walton, 1993, pers. commun.).

The regression formula estimated for log-leaf area of trees with measured shading coefficients was:

In Y = -4.3309 + 0.2942H + 0.7312D + 5.7217Sh - 0.0148S (r2 = 0.91),

where Y = total leaf area (m2), H = crown height (m), D = crown diameter (m), Sh = shading coefficient (Appendix A, Table 1), and S=D(H + D)/2. The correction factor (0.1159), added to the untransformed estimate, resulted in the following estimate for leaf area:

Y = e-4.3309 + 0.2942H + 0.7312D + 5.7217Sh - 0.0148S + 0.1159

For trees for which shading coefficients are unknown, the estimated log-leaf area relationship was:

In Y = 0.6031 + 0.2375H + 0.6906D - 0.0123S (r2 = 0.86)

The correction factor added to the untransformed estimated value was 0.1824.

Total leaf area, derived from trees in excellent condition, was adjusted according to the condition class of the tree. Estimates of total leaf area were multiplied by 1 for trees in excellent condition, by 0.85 for trees in good condition, by 0.625 for moderate trees, by 0.375 for poor trees, by 0.125 for dying trees, and by 0 for dead trees.

For trees with characteristics outside the range of conditions under which the regression equations were derived (H > 12 m, D > 12 m, H/D > 3, S > 500 or S < 1; n = 759, 8.4 percent of the sample), leaf area was estimated using a volumetric approach. The volume of individual crowns occupied by leaves (foliated-crown volume) was estimated based on measured crown height, width, shape, and percent of crown occupied by leaves. Average leaf dry weight (g/m3) was calculated based on measured data and information from the literature on individual tree species (Winer et al. 1983; Nowak 1991). Factors for average leaf dry weight were applied to the foliated-crown volume to estimate total leaf dry weight of the tree. This estimate was converted to leaf area using conversion factors (m2/g) calculated from measured data and from the literature (McLaughlin and Madgwick 1968; Monk et al. 1970; Gacka-Grzesikiewicz 1980; Box 1981; Shelton and Switzer 1984; Bacon and Zedaker 1986; Vose and Allen 1988; Reich et al. 1991; Cregg 1992). If no conversion data were found for an individual species, the genera average was substituted; if no genera data were found, the average conversion value for the hardwood or conifer group was used.

Relative dominance of a tree species was calculated as the total leaf-surface area of all trees of one species as a percentage of the total leaf-surface area of trees of all species. Reliable estimates of error of leaf area estimates could not be made because it was not possible to determine the amount of error regarding factors associated with estimates of leaf area, for example, regression formula transformations, conversions used in the volumetric approach, and adjustments for crown condition. Thus, standard errors are not reported for estimates of species dominance.

Average LAI's for individual trees were calculated by dividing the sum of leaf-surface areas by the sum of crown projections (individual ground area = D2/4). The total LAI for the study area was calculated by dividing the estimate of the total leaf-surface area in the study area by the total area occupied by trees (from aerial photograph interpretation) (McPherson et al. 1993). Ground projections based on aerial photographs account for the multiple layering effect of trees (combined effect of overstory and understory trees).

Results
There are approximately 50.8 million trees in the study area, with 4.1 million trees in Chicago, 31.8 million in suburban Cook County, and 14.9 million in DuPage County (Table 2). The largest proportion of trees (49 percent) is on institutional lands dominated by vegetation (e.g., parks, forest preserves, cemeteries, golf courses), followed by 1-3 family residential land (25 percent), and vacant land (21 percent) (Table 2). These land uses also have the highest tree densities with institutional lands dominated by vegetation having 563 trees/ ha (228 trees/acre). Vacant lands have 488 trees/ha (197 trees/acre) and 1-3 family residential lands have 93 trees/ha (38 trees/acre) (Table 3). Overall tree density is highest in DuPage County at 173 trees/ha (70 trees/acre), followed by suburban Cook County with 169 trees/ha (68 trees/acre) and Chicago with 68 trees/ha (28 trees/acre) (Table 3). Most of the estimated leaf-surface area (87.5 percent) is on 1-3 family residential lands and institutional lands dominated by vegetation (Table 4).

Cottonwood and green/white ash are the most common species in Chicago. Buckthorn and green/white ash are most common in suburban Cook County, and willow and boxelder are the most common species in DuPage County (Table 5; Appendix A, Tables 2-6). Species that dominate in leaf area are cottonwood and green/white ash in Chicago, silver maple and American elm in suburban Cook County, and white oak and silver maple in DuPage County (Table 5; Appendix A, Tables 2-6). Composition and leaf-area dominance of tree species by land-use type for each sector of the study area are given in Appendix A, Tables 7-14.

Table 2. Estimated number of trees (In thousands) by land-use type in Chicago, suburban Cook County, DuPage County, and entire study area.

Table 3. Tree density (no. trees/ha) by land-use type in Chicago, suburban Cook County, DuPage County, and entire study area (divide by 2.471 to convert stems/ha to stems/acre).

Table 4. Percentage of land area, total number of trees (tree population), and total leaf area within the study area, by land-use type.

Table 5. Tree-species compostition in Chicago, suburban Cook County, DuPage County, and entire study area; includes top 20 species in number and percentage of trees and species dominance based on percentage of total leaf-surface area in each sector.

Table 5 Continued...

Common and/or dominant species that contribute the most leaf area on a per tree basis are white oak, swamp white oak, Norway spruce, silver maple, and Norway maple (Table 6). Species that contribute the most large-diameter trees to the study area are silver maple, white oak, American elm, bur oak, and cottonwood (Table 7). Common small-diameter tree species are buckthorn, Prunus spp., green/white ash, boxelder, and willow (Table 8).

Fifty-six percent of the trees in the study area are less than 7 cm (3 inches) in diameter and 76.9 percent are less than 15 cm (6 inches) d.b.h. (Table 9). Chicago has the highest proportion of large trees greater than 46 cm (18 inches) d.b.h. (7.5 percent). Land uses with the highest proportion of large trees are institutional land dominated by buildings (29 percent) and 1-3 family residential land (10 percent) (Appendix A, Table 15).

About 55 percent of the trees in the study area were rated in good condition and 10.5 percent were rated as dead or dying (Table 10). Land uses with the highest proportion of dead and dying trees are institutional land dominated by vegetation (16 percent), followed by institutional lands dominated by buildings (11 percent), and vacant land (9.5 percent) (Appendix A, Table 16).

Table 6. Average leaf-surface area (m2) per tree for top 20 species (in number and species dominance) in entire study area (index value is average species leaf area per tree divided by average leaf area per tree for entire population (81 m2))

The average LAI of individual trees is 4.3 in Chicago, 4.2 in suburban Cook County, 4.5 in DuPage County and 4.3 in the study area. The maximum LAI calculated using the regression equations for an individual tree was 18.1 with only 0.05 percent of the estimated LAI's for individual trees greater than 15. The estimated LAI for the entire study area, which accounts for the multiple layering of trees, is 6.3. The overall LAI may be slightly overestimated because of a likely conservative estimate of tree cover in Chicago. The large amount and size of buildings in Chicago tend to obscure small trees. This obstruction likely results in an underestimation of tree cover and consequently a slight overestimation of the overall LAI. Thus, an overall LAI of 6.0 is probably more likely for the Chicago area. Conifers account for 6 percent of the leaf-surface area in the study area.

Table 7. Most common large trees given as percentage of total number of trees larger than 46 cm (18 inches) d.b.h.

Table 8.Most common small trees given as percentage of total number of trees less than 7 cm (3 inches) d.b.h.

Table 9. Distribution of tree diameters in Chicago, suburban Cook County, DuPage County, and entire study area.

aPercentage of population

Table 10. Distribution of trees by condition in Chicago, suburban Cook County, DuPage County, and the entire study area.

aPercentage of population

Populations of Street Trees
There are an estimated 1,463,700 street trees in the study area (SE =151,900), with 416,000 in Chicago (SE = 48,500), 854,300 in suburban Cook County (SE =139,400), and 193,400 in DuPage County (SE = 35,700). Norway maple and honeylocust are the most common street trees in Chicago, silver maple and green/white ash in suburban Cook County, and green/white ash and Norway maple in DuPage County (Table 11). Street trees in the study area tend to be larger than trees in general 51.5 percent of all street trees are 16 to 46 cm (6 to 18 inches) d.b.h. (Table 12). Chicago has the highest proportion of large street trees with 28.7 percent larger than 46 cm d.b.h. (Table 12).

Most street trees in the study area were rated as good (46 percent) or excellent (34 percent) (Table 13). Only 0.5 percent were rated as dead or dying. No dead or dying street trees were found in Chicago or suburban Cook County. Street trees account for only 2.9 percent of the total tree population but 9.5 percent of the total leaf-surface area (Table 14). Street trees are most significant in Chicago where they account for 10.1 percent of the total population and 24 percent of total leaf-surface area. Dominance of street trees varies by land-use type with the greatest proportion occurring on residential lands in Chicago where street trees account for 27.9 percent of the trees and 43.7 percent of leaf-surface area (Table 14).

Urban Ground Cover
The most common ground surfaces in the study area are' maintained grass, tar, and herbaceous plants; common surfaces in Chicago are tar, maintained grass, and buildings (Table 15). Ground cover varied by land-use type with maintained grass the most common ground cover type on institutional and 1-3 family residential lands, tar most common on commercial/industrial and transportational lands, herbaceous cover most abundant on agricultural and vacant lands, and building cover most common on multifamily residential lands (Appendix A, Table 17).

Table 11. Top 25 street tree species in study area by sector.

Table 11 continued...

aPercentage of population

Table 12. Diameter distribution of street trees in Chicago, suburban Cook County, DuPage County, and entire study area.

aPercentage of population

Table 13. Distribution of street trees by condition in Chicago, suburban Cook County, DuPage County, and entire study area.

aPercentage of population

Table 14. Street trees as a percentage of total tree population (%POP) and percentage of total leaf-surface area (%LSA) in Chicago, suburban Cook County, DuPage County, and entire study area

Table 15. Distribution of ground-surface materials in Chicago, suburban Cook county, DuPage County, and entire study area.

a Percentage of population

Discussion
Urban Forest Structure in the Chicago Area
The Chicago area's urban forest is composed mostly of small trees less than 15 cm d.b.h. (76.9 percent). Small trees also account for the majority of trees in other cities. In Shorewood, Wisconsin, and Oakland, California, 67 percent and 60.9 percent of the trees are less than 15 cm d.b.h., respectively (Dorney et al. 1984; Nowak 1993a). However, the distribution of tree sizes varies among and within land-use types depending on the duration and intensity of vegetation management. Less-managed (e.g., vacant) or naturalistically managed lands (e.g., forest preserves) had the highest proportion of small trees. Highly managed areas, particularly those managed for a relatively long period (e.g., street trees, residential areas), tend to have a higher proportion of large trees. However, there are some large old remnant trees throughout the Chicago area, particularly in forest preserves.

Most of the trees in the study area were classified as being in good condition. Ratings on tree condition are affected by urban-environmental stresses (e.g., salt, soil compaction, vandalism, injury), plant competition (related to tree density) and natural aging processes (tree size), all of which tend to increase crown discoloration and dieback (e.g., Nowak and McBride 1991). Consequently, relatively few trees were rated as excellent. Most of the dead and dying trees are in areas with minimal maintenance, naturalistic management, or in areas with more large trees that are not intensively managed (institutional land dominated by buildings). Dead and dying trees tend to be removed in the more intensively managed areas.

Species Composition
The most common species is the exotic and highly invasive buckthorn, accounting for 12.7 percent of the tree population. Seven of the 10 most common trees are native; three are genera of both native and exotic species. Four of the eight most common species are native pioneer species: green ash, boxelder, willow, cottonwood. These species have a propensity to colonize sites but have a shorter lifespan than more shade-tolerant species (Spurr and Barnes 1980; Burns and Honkala 1990). These species are common on all land uses but most common on vacant lands where they account for 47 percent of the population. Buckthorn is common on the three land uses that contain 95 percent of the trees (institutional lands dominated by vegetation, 1-3 family residential, and vacant lands). These land uses include many areas with relatively low maintenance (e.g., tree stands), which facilitates invasion by buckthorn. The most common ornamental species, exclusive of major pioneer species, planted on residential lands are silver maple, Prunus spp., blue spruce, crabapple, mulberry, Norway maple, arborvitae, honeylocust, American elm, and junipers.

The most common trees in Chicago are cottonwood and green/white ash, which make up 25 percent of the city's tree population. Green/white ash, both a pioneer and common ornamental tree, is common on most land uses in Chicago and accounts for 12 percent of all trees in the city. Cottonwood, which generally is not planted as an ornamental species, is the most common tree on vacant lands and institutional lands dominated by vegetation in Chicago. These land uses contain many low maintenance sites which facilitate invasion by cottonwood.

Species and Individual Tree Dominance
The most dominant species in total leaf area are silver maple, green/white ash, white oak, and American elm. These four species most likely have the greatest impact on the surrounding environment and constitute 34.8 percent of total leaf-surface area. Institutional lands dominated by vegetation are dominated by American elm, white oak, green/white ash, and red/black oak (39.8 percent of total leaf-surface area); 1-3 family residential areas are dominated by silver maple, green/white ash and white oak (31.7 percent); and vacant lands are dominated by the pioneer species of cottonwood, boxelder, willow, and poplar (other) (50.7 percent). Although buckthorn is the most common tree in the study area, it accounts for only 2.9 percent of total leaf-surface area due to its relatively small size.

The greatest average leaf-surface area on a per tree basis occurs on white oak, swamp white oak, Norway spruce, silver maple, and Norway maple. Management activities should be directed toward preserving dominant individuals in a healthy condition so that their large environmental and social benefits, relative to smaller trees, are sustained (e.g., Schroeder and Cannon 1987; Nowak 1994a,b).

Diameter-growth rates of individual open-grown urban trees are relatively high (Nowak 1994b) and these growth rates are explained partially by the average LAI of individual trees in the study area (4.3), which is near the index level of maximum net growth. The overall urban tree LAI of 6.0 is at the low end of the normal range of LAI's exhibited for deciduous forests (Barbour et al. 1980). This relatively low index level is understandable considering the relative lack of lower level canopy (understory trees) in some urban areas that are common in deciduous forests. The urban forest understory of more intensively managed land uses often is occupied by grass or impervious surfaces.

Street Trees
Street trees in Chicago constitute 1 of every 10 trees overall and 1 of every 4 trees in 1-3 family residential areas. Chicago's street trees contribute 24 percent of the total city leaf-surface area, and 44 percent of total leaf area on 1-3 family residential lands. Street trees play a less important role in less urbanized areas, but can still contribute significantly to the street-corridor environment (Schroeder and Cannon 1987).

In suburban Cook County, street trees constitute 1 of every 37 trees (9.5 percent of total leaf-surface area) and 1 of every 10 trees on residential land. In the least urbanized sector, DuPage County, street trees account for 1 of every 77 trees (3.6 percent of total leaf-surface area) and 1 of every 26 trees on residential land. Thus, street trees become a more important component of the urban forest in more urbanized areas as artificial surfaces and land-use activities compete for tree space.

A high percentage of street trees in the Chicago area are greater than 46 cm d.b.h. (Chicago: 28.7 percent; suburban Cook County: 14.9 percent; DuPage County: 17.1 percent). There is a 4 to 6 times higher percentage of large street trees than non-street trees. Large trees are important to the urban environment, contributing significantly more air quality and carbon dioxide sequestration benefits than small trees (see Nowak 1994a,b: Chapters 5 and 6, this report).

Urban Ground Surfaces
Besides trees, a wide range of other urban surfaces interact with the surrounding environment and affect local gas and energy exchanges, visual quality, human stress, etc. The most abundant urban ground surfaces in the study area are maintained grass, tar, herbaceous plants (e.g., agriculture crops) and buildings. Impervious surfaces cover 60 percent of Chicago, 33 percent of suburban Cook County, and 25 percent of DuPage County. Tar generally is the most common ground-surface cover of commercial/industrial and transportation lands. Maintained grass often is the most abundant surface on residential and institutional lands. Converting non-essential impervious surfaces (e.g., abandoned parking lots) to more pervious surfaces (e.g., soil) could facilitate the formation of vegetation and reduce surface runoff. Understanding how various urban surfaces interact to affect the local environment and city inhabitants remains to be investigated.

Factors Influencing Current Vegetation Patterns
Vegetation within urban and urbanizing areas changes through time and space. Land use is one of the most significant factors affecting local vegetation patterns and distribution. In conjunction with its associated patterns of buildings and other artificial surfaces, land use influences the space available for trees and to some extent whether those spaces will be filled with trees and how they will be managed. Most of the nearly 51 million trees in Cook and DuPage Counties are on institutional lands dominated by vegetation, 1-3 family residences, and vacant land. This distribution pattern is similar to that for trees in Oakland, California (Nowak 1993a). These land uses generally are the most amenable to tree growth in urban areas and are likely where most of the trees exist in U.S. cities. Management plans should consider differences in tree distribution among land-use types to optimize tree configurations across the entire urban area. By understanding tree variations among land-use types, managers could focus planting efforts in areas typically lacking trees and direct species composition in more heavily-treed areas to meet specific management objectives and enhance the local environment.

In regions such as the Chicago area where trees are readily established through natural seeding, available planting space that is not filled with trees often has been actively managed to prohibit trees (e.g., mowing, use of herbicides, planting of herbs, selective tree removal). Such activities are necessary for land uses such as agriculture, airports, prairies, and sporting fields, but uses such as residential, commercial, and some transportation corridors could be used to increase tree cover if desired.

Tree cover can be increased through education and other promotional efforts that support tree planting and maintenance and/or encourage reducing management activities that prohibit trees and thereby allow trees to become established on the site naturally. Natural tree establishment can facilitate the development of invasive species so management activities should be directed toward altering species composition if certain invasive species are deemed undesirable.

The intensity of urban development also influences the amount of trees in a city, with tree density generally decreasing with urbanization. Average tree density in the Chicago area ranged from 68 trees/ha (28 trees/acre) in Chicago to 173 trees/ha (70 trees/acre) in DuPage County. There are two primary reasons for the decrease in tree density with increased urbanization. First, in more heavily urbanized areas, more of the land is occupied by uses that preclude tree establishment (e.g., commercial/industrial, transportation).5 Second, tree space tends to be more limited in highly urban areas (i.e., residential lots tend to be smaller; impervious surfaces occupy a higher proportion of the ground area).

Tree density on residential and commercial land in Chicago is comparable to those in Shorewood, Wisconsin, for the same land uses (Dorney et al. 1984). Tree density from other urban areas are 120 trees/ha (49 trees/acre) in Oakland, California (Nowak 1993a) and 373 and 40 trees/ha (151 and 16 trees/acre) for portions of South Lake Tahoe and Menlo Park, California, respectively (McBride and Jacobs 1986). By contrast, the average live tree density on timberland in Illinois is 1,186 trees/ha (480 trees/acre) (Raile and Leatherberry 1988).

Besides affecting management and various environmental functions, tree density affects visual quality of a landscape. Optimal foreground density for aesthetic quality in municipal parks has been estimated at approximately 125 trees/ha (51 trees/acre) (Schroeder and Green 1985). High tree densities and large trees are also preferred along streets (Schroeder and Cannon 1987).

Most of the differences in vegetation patterns within the study area are due to differences in land-use distribution, intensity of urbanization, and age of development. Chicago is the oldest, most urbanized area while DuPage County is the most suburban to rural area with newer residential developments and the highest proportion of agricultural areas.

Directing Future Urban Forest Structure in the Chicago Area
The future urban forest in the Chicago area, as indicated by the distribution of tree species less than 7 cm d.b.h., is likely to be dominated by green/white ash, boxelder, willow, cottonwood, black locust, and shagbark hickory. Other common species (buckthorn, Prunus spp., hawthorns, alders) in this smallest d.b.h. class generally do not reach a dominating size. American elm also is a common small tree, but sanitation programs and/or the planting of cultivars that resist Dutch elm disease must be continued or utilized if American elms are to maintain a dominant position in the Chicago area's urban forest.

This probable future forest will mean a shift from silver maple and white oak that codominate today toward more invasive pioneer species. While silver maple, white oak, and bur oak account for one-third of the trees greater than 46 cm d.b.h., they make up only 3.3 percent of the trees less than 7 cm d.b.h. However, planners and managers can alter or direct future species composition and structure (Nowak 1993c).

Education and management can influence the amount, type, and location of urban vegetation (e.g., tree planting in backyards and parking lots) and thereby direct future urban forest structure to a desirable outcome. Trees are not appropriate in all locations or land uses. However, where trees are desirable, planning and management can facilitate proper urban forest structure. The more space available for tree planting that is not inhibited by the existing land, use, the more the natural environment and local planning and management can influence vegetation structure (e.g., vacant lands, parks).

Management plans should consider directing current urban forest structure toward a future structure that enhances healthy, functional leaf-surface area and optimizes species composition to maximize both social and environmental benefits of trees. Management plans should be developed to meet specific local needs, for example, enhancing the scenic beauty of a park or reducing air pollution in a certain area. Managing for one need or to maximize one benefit may reduce some other benefits derived from urban trees, so local and regional management priorities and plans must be developed. Besides preserving large trees, multilayer forest structures (stand conditions) should be sustained where appropriate, and healthy canopies should be maintained to maximize many tree benefits. Also, ample water should be supplied to trees to optimize benefits that are linked with transpiration (e.g., removal of gaseous pollutants and reduced air temperatures).

Implications for Research
The equations developed to predict the leaf-surface area of individual urban trees appear to yield reasonable estimates when applied within the bounds in which the regression equations were developed. However, more work is needed on developing shading coefficients and leaf area predictions for individual species, particularly for large trees and coniferous species. Also needed is additional research on urban forest structure and its link to various functions for other U.S. cities to help clarify and determine existing urban forest patterns and processes. Finally, researchers need to investigate changes in urban forest structure and functions through time to better predict and understand the dynamics of these ecosystems, and to determine how urban surfaces interact in affecting the local environment and inhabitants.

Conclusion
Urban forest planning and management can direct urban forest structure toward a desired outcome. One of the first steps in properly directing urban forest structure is to understand if, and what, changes are necessary by analyzing the existing urban forest structure. By understanding forest structure and determining the relationships between structure and forest functions, various social and environmental benefits can also be quantified. The Chicago area urban forest contains 50.8 million trees, approximately 9 trees per resident. Most of the trees are small and predominantly found on institutional, residential and vacant lands.

The current pattern of urban vegetation has been formed through both present and past human and environmental factors. Education of both the public and private sectors can facilitate directing future urban forest structure toward desired results as dictated by urban forest management plans. However, the urban environment (e.g., land uses) presents many constraints on urban forest structure that managers and planners must consider.

Relatively short-lived pioneer species contribute significantly to the Chicago area urban forest and are most prevalent on land uses with minimal or naturalistic management (e.g., forest stand conditions). Street trees are also important elements of the urban forest, particularly in the City of Chicago. Trees are just one of many surfaces that interact to influence the urban environment; other prominent ground surfaces include tar and grass.

Acknowledgments
I sincerely thank AI Neiman, Ron Nemchausky, and The Chicago Park District for use of a high-lift truck and assistance in data collection; Henry Henderson, Edith Makra, Chicago Department of Environment, Steve Bylina Jr., Jerry Dalton, Bill Brown, and the Chicago Bureau of Forestry for local technical assistance; Gerald Walton for statistical advice and assistance; John Dwyer, Marty Jones, Rowan Rowntree, and Gerald Walton for reviewing this article; and Scott Prichard, Rachel Mendoza, Steve Wensman, Merle Turner, Joanna Mignano, Wendy Vear-Hanson, Brad Bonner, Marcia Henning, Paul Sacamano, Hyan-kil Jo, Jennifer Lixey, Lisa Blakeslee, and Jennifer Lee for assistance in field data collection and entry.

Literature Cited
Airola, T. M.; Buchholz, K. 1982. Forest community relationships of the Greenbrook Sanctuary, New Jersey. Bulletin of the Torrey Botanical Club. 109: 205-218.

Bacon, C. G.; Zedaker, S. M. 1986. Leaf area prediction equations for young southeastern hardwood stems. Forest Science. 32(3): 818-821.

Barbour, M. G.; Burk, J. H.; Pitts, W. D. 1980. Terrestrial plant ecology. Menlo Park, CA: Benjamin/Cummings Publishing Co. 604 p.

Box, E. O. 1981. Foliar biomass: data base of the International Biological Program and other sources. In: Bufalini, J. J.; Arnts, R. R. eds. Atmospheric biogenic hydrocarbons. Volume 1. Emissions. Ann Arbor, MI: Science Publishers: 121-148.

Boyd, J. B. 1983. Natural reproduction of exotic and indigenous trees in three urban environments. Milwaukee, WI: University of Wisconsin. 97 p. M.S. thesis.

Buhyoff, G. J.; Gauthier, L. J.; Wellman, J. D. 1984. Predicting scenic quality for urban forests using vegetation measurements. Forest Science. 30(1): 71-82.

Burns, R. M.; Honkala, B. H. 1990. Silvics of North America, Volume 2, hardwoods. Agric. Handb. 654. Washington, DC: U.S. Department of Agriculture. 877 p.

Cregg, B. M. 1992. Leaf area estimation of mature foliage of Juniperus. Forest Science. 38(1): 61-67.

Crow, T. R. 1988. A guide to using regression equations for estimating tree biomass. Northern Journal of Applied Forestry. 5: 15-22.

Dawson, J. O.; Khawaja, M. A. 1985. Change in street-tree composition of two Urbana, Illinois neighborhoods after fifty years: 1932-1982. Journal of Arboriculture. 11(11): 344-348.

Derrenbacher, W. E. 1969. Plants and landscape: an analysis of ornamental planting in four Berkeley neighborhoods. Berkeley, CA: University of California, Berkeley, CA. 231 p. M.S. thesis.

Dorney, J. R.; Guntenspergen, J. R.; Stearns, F. 1984. Composition and structure of an urban woody plant community. Urban Ecology. 8: 69-90.
Gacka-Grzesikiewicz, E. 1980. Assimilation surface of urban green areas. Ekologia Polska. 28(4): 493-523.

Hedman, C. W.; Binkley, D. 1988. Canopy profiles of some Piedmont hardwood forests. Canadian Journal of Forest Research. 18: 1090-1093.

Horie, Y.; Sidawi, S.; Ellefsen, R. 1991. Inventory of leaf biomass and emission factors for vegetation in California's South Coast Air Basin. Tech. Rep. III-C. El Monte, CA: South Coast Air Quality Management District. 149 p.

Impens R.; Delcarte, E. 1979. Survey of urban trees in Brussels, Belgium. Journal of Arboriculture. 5: 169-176.

Jim, C. Y. 1986. Street trees in high-density urban Hong Kong. Journal of Arboriculture. 12: 257-263. Kramer, P. J.; Kozlowski, T. T. 1979. Physiology of woody plants. New York: Academic Press. 811 p.

McBride, J. R.; Froehlich, D. 1984. Structure and condition of older stands in parks and open space areas of San Francisco, California. Urban Ecology. 8: 165-178.

McBride, J. R.; Jacobs, D. F. 1986. Presettlement forest structure as a factor in urban forest development. Urban Ecology. 9: 245-266.

McLaughlin, S. B.; Madgwick, H. A. I. 1968. The effects of position in crown on the morphology of needles of loblolly pine (Pinus taeda L.). American Midland Naturalist. 80(2): 547-550.

McPherson, E. G. 1984. Planting design for solar control. In:McPherson, E.G. ed. Energy-conserving site design. Washington, DC: American Society of Landscape Architects: 141-164.

McPherson, E. G.; Nowak, D. J.; Sacamano, P. L.; Prichard, S. E.; Makra, E. M. 1993. Chicago's evolving urban forest: initial report of the Chicago Urban Forest Climate Project. Gen. Tech. Rep. NE-169. Radnor, PA: U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station. 55 p.

McPherson, E. G.; Rowntree, R. A. 1989. Using structural measures to compare twenty-two U.S. street tree populations. Landscape Journal. 8(1): 13-23

Miller, P. R.; Winer, A. M. 1984. Composition and dominance in Los Angeles Basin urban vegetation. Urban Ecology. 8: 29-54.

Monk, C. D.; Child, G. I.; Nicholson, S. A. 1970. Biomass, litter and leaf-surface area estimates of an oak-hickory forest. Oikos. 21:138-141.

Nowak, D. J. 1991. Urban forest development and structure: analysis of Oakland, California. Berkeley, CA: University of California. 232 p. Ph.D. dissertation.

Nowak, D. J. 1993a. Compensatory value of an urban forest: an application of the tree-value formula. Journal of Arboriculture. 19(3): 173-177.

Nowak, D. J. 1993b. Atmospheric carbon reduction by urban trees. Journal of Environmental Management. 37: 207-217.

Nowak, D. J. 1993c. Historical vegetation change in Oakland and its implications for urban forest management. Journal of Arboriculture. 19(5): 313-319.

Nowak, D. J. 1994a. Air pollution removal by Chicago's urban forest. In: McPherson, E. G., Nowak, D. J. and Rowntree, R. A. eds. Chicago's urban forest ecosystem: results of the Chicago Urban Forest Climate Project. Gen. Tech. Rep. NE-186 . Radnor, PA: U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station.

Nowak, D. J. 1994b. Atmospheric carbon dioxide reduction by Chicago's urban forest. In: McPherson, E. G., Nowak, D.J. and Rowntree, R. A. eds. Chicago's urban forest ecosystem: results of the Chicago Urban Forest Climate Project. Gen. Tech. Rep. NE-186. Radnor, PA: U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station.

Nowak, D. J.; McBride, J. R. 1991. Comparison of Monterey pine stress in urban and natural forests. Journal of Environmental Management. 32: 383-395.

Peterson, D. L.; Arbaugh, M. J.; Lardner, M. A. 1988. Leaf area of lodgepole pine and whitebark pine in a subalpine Sierra Nevada forest. Canadian Journal of Forest Research. 19: 401-403.

Profous, G. V.; Rowntree, R. A. 1993. The structure and management of the urban forest in Prague, Czechoslovakia. I. Growing space in metropolitan Prague. Arboricultural Journal. 17: 1-13.

Profous, G. V.; Rowntree, R. A.; Loeb, R. E. 1988. The urban forest landscape of Athens, Greece: aspects of structure, planning and management. Arboriculture Journal. 12: 83-107.

Raile, G. K.; Leatherberry, E. C. 1988. Illinois forest resource. Resour. Bull. NC-105. St, Paul, MN: U.S. Department of Agriculture, Forest Service, North Central Forest Experiment Station. 113 p.

Reich, P. B.; Walters, M. B.; Ellsworth, D. S. 1991. Leaf age and season influence the relationships between leaf nitrogen, leaf mass per area and photosynthesis in maple and oak trees. Plant, Cell and Environment. 14: 251-259.

Richards, N. A.; Mallette, J. R.; Simpson, R. J.; Macie, E. A. 1984. Residential greenspace and vegetation in a mature city: Syracuse, New York. Urban Ecology. 8: 99-125.

Richards, N. A.; Stevens, J. C. 1979. Streetside space and street trees in Syracuse -1978. Syracuse, NY: State University of New York, College of Environmental Science and Forestry. 66p.

Rowntree, R. A. 1984. Forest canopy cover and land use in four eastern United States cities. Urban Ecology. 8: 55-67.

Sanders, R. A. 1984. Some determinants of urban forest structure. Urban Ecology. 8: 13-27.
Schmid, J. A. 1975. Urban vegetation: a review and Chicago case study. Res. Pap. (vol. 161). Chicago, IL: University of Chicago, Department of Geography. 266 p.

Schroeder, H. W.; Cannon, W. N. 1987. Visual quality of residential streets: both street and yard trees make a difference. Journal of Arboriculture. 13(10): 236-239.

Schroeder, H. W.; Green, T. L. 1985. Public preference for tree density in municipal parks. Journal of Arboriculture. 11(9): 272-277.

Shelton, M. G.; Switzer, G. L. 1984. Variation in the surface area relationships of loblolly pine fascicles. Forest Science. 30(2): 355-363.

Spurr, S. H.; Barnes, B. V. 1980. Forest ecology. New York: John Wiley and Sons. 687 p.

Stevens, J. C.; Richards, N. A. 1986. Village and city street tree resources: a comparison of structure. Arboricultural Journal. 10: 45-52.

Talarchek, G. M. 1985. The New Orleans urban forest: structure and management. New Orleans, LA: Xavier University of Louisiana, University Research Center. 104 p.

Vose, J. M.; Allen, H. L. 1988. Leaf area, stemwood growth, and nutrition relationships in loblolly pine. Forest Science. 34(3): 547-563.

Whitney, G. G.; Adams, S. D. 1980. Man as a maker of new plant communities. Journal of Applied Ecology. 17: 431-448.

Winer, A. M.; Fitz, D. R.; Miller, P. R.; Atkinson, R.; Brown, D. E.; Carter, W. P.; Dodd, M. C.; Johnson, C. W.; Myers, M. A.; Neisess, K. R.; Poe, M. P. Stephens, E. R. 1983. Investigation of the role of natural hydrocarbons in photochemical smog formation in California. Riverside, CA: Statewide Air Pollution Research Center. 234 p.

Footnotes:
1Overall, 249 plots were located on 1-3 family residential lands, 26 plots on multifamily residential lands (apartments with four or more units), 194 plots on institutional lands dominated by vegetation (e.g., parks, cemeteries, golf courses, forest preserves), 22 plots on institutional lands dominated by buildings (e.g., schools, churches), 52 plots on commercial/industrial lands, 45 plots on vacant lands, 39 plots on transportational lands (e.g., airports, freeways), and 25 plots on agricultural lands.

2On 1-3 family residential lands, the entire residential lot (mid-road to mid-alley) was measured. For other land use types, 0.04-hectare (ha) (0.1-acre) plots were measured.

3A sliding pole that displays the height at the top of the pole.

4A computer program was written to map the measured tree-crown dimensions and calculate crown volume. Random distances along x, y, and z coordinates from the tree base were selected to determine sampling locations within each tree crown. Sample locations in the tree crown were approached with the high-lift truck bucket so as not to disturb the sample prior to leaf collection.

5 Rural areas also can have land uses where low tree densities are typical (e.g., agriculture, vacant land in desert areas).

Hot Topics