76 Pearce et al.: Using Size Class Distributions of Species for the Private Urban Forest Data Analysis Tree species were classed as fast-growing or slow-growing on the basis of the descriptions in Botanica (2001). The upper value in the mature height range given for each spe- cies in Botanica (2001) was used in analyses as potential height, except where field observation suggested a higher value. The potential height for a species group was the po- tential height of the most commonly planted species. To determine the influence of tree characteristics on height class distributions, researchers tested whether growth rate or potential height affected either the number or propor- tion of individuals in the <3 m and >8 m height classes, using Pearson’s product moment correlation coefficient. A ‘weed’ tree was defined as an exotic invader of native vegetation in the region, whether or not it was sold to the public as a cultivated plant. Weeds, along with many native and exotic ornamentals, self-establish in gardens. The orna- mentals were defined as species grown or retained for human purposes other than food production. Native ornamentals were defined as species that occurred in Australia before British occupation in 1788, and planted or retained for human purpos- es other than food production. Food trees were defined to be those that produce fruit, bark, or leaves consumed by humans. The species were classified in order to allow the research- ers to be able to generalize size class distribution patterns. The percentages of front garden trees in each height class were used as the inputs for an agglomerative classification of the tree species using Euclidean distance and Ward’s method. The latter method avoids the chaining that results from the group averaging procedure. The dendrogram and error values were examined to determine the appropriate number of groups. One-way analysis of variance (ANOVA) was used to determine if classificatory groups differed at P < 0.05 in the potential heights of their component taxa and whether fast-growing trees differed from slow-growing trees in their percentages in each height class. Pearson’s product moment correlation coefficient was used to test the strength of linear relationships between potential height and both numbers and proportions of individuals in each height class. RESULTS The data indicated that private trees dominated the urban for- est in the suburbs. In Melbourne, there was an almost even distribution of >8 m tall trees between the street, front garden, and back garden. In Hobart, a majority of tall trees occurred in private gardens (Table 2). In both cities, the proportions of trees on public spaces, which included small parks, school grounds, and church grounds, were low. A low percentage of trees were located in the streets in the Hobart suburbs (Table 2). Fifteen taxa (species or species groups) were observed 100 or more times as plants <3 m tall (Appendix). Ten of these were taxa that had a height potential that just qualified them to be considered as trees. In contrast, only Betula, Eucalyptus, and Cupressaceae, all of which were potentially tall taxa, had more than 100 individuals in the >8 m height class (Appendix). The trees in group 4, including Acmena spp., Michelia spp., Magnolia grandiflora, and Leptospermum spp., had high propor- tions of individuals in the smaller height classes and few in the larger height classes (Figure 1), indicating recent planting. In con- trast, the trees in group 3, such as Liquidambar styraciflua, Schi- nus molle, and Eucalyptus spp., were concentrated in the taller height classes with few individuals in the shorter classes (Figure 1), indicating popularity in the past, but a lack of recent planting. Groups 1 and 2 were characterized by high percentages in the medium height classes. Group 1 had a smaller percentage of the two lowest height classes than group 2, and group 2 had smaller percentages than group 1 in the 3–5 m and 5–8 m height classes (Figure 1). The species that were the stron- gest outliers from other species for the ratio >5 m: <3 m trees were in groups 1 and 3, while the outliers for the ra- tio 3–5 m: 1–3 m trees were all in group 1 (Figure 2). Fast- growing trees did not differ (ANOVA, P > 0.05) from slow- er-growing trees in their percentages in any height class. The potential height of the tree taxa varied significantly = 44%), by group (ANOVA, F = 16.85, d.f. = 3, P < 0.001, r2 with taxa in group 3 being potentially much taller than trees in the three other groups (group 1 mean = 12.5 m, s.d. = 6.3 m; group 2 mean = 12.0 m, s.d. = 6.3 m; group 3 mean = 26.7 m, s.d = 9.8 m; group 4 mean = 12.7 m, s.d. = 6.4 m). Table 2. Percentages of >8 m tall trees in different locations, sorted by suburb and city. Suburb N Melbourne Fairfield Albert Park Croydon Clayton Malvern Hobart Sandy Bay North Hobart Lenah Valley South Hobart Risdon Vale City Melbourne Hobart 1320 1064 31 5 33 50 32 41 3 5 418 227 186 225 8 1 7 7 9 0 46 37 40 38 87 49 47 47 47 13 4 9 5 6 0 231 168 467 209 245 % street 24 64 15 22 33 % front 29 16 42 35 44 % back 37 19 39 43 23 % public spaces 10 1 4 0 0 ©2013 International Society of Arboriculture
March 2013
| Title Name |
Pages |
Delete |
Url |
| Empty |
Search Text Block
Page #page_num
#doc_title
Hi $receivername|$receiveremail,
$sendername|$senderemail wrote these comments for you:
$message
$sendername|$senderemail would like for you to view the following digital edition.
Please click on the page below to be directed to the digital edition:
$thumbnail$pagenum
$link$pagenum
Your form submission was a success. You will be contacted by Washington Gas with follow-up information regarding your request.
This process might take longer please wait
