Laserfiche WebLink
Arboriculture & Urban Forestry 34(6): November 2008 389 <br />Table 3. Effect of plot size and design on number of parcels per plot, number of access permissions required, and percent <br />tree cover in Syracuse, NY, using 2 ft resolution tree cover and land use/parcel boundary maps of 500 randomly located <br />plots. <br />Number of parcels <br />Plot size (ac)z Total Perm. req.Y Perm. quest.' No perm.' <br />1/24 <br />1/12 <br />1/10 <br />1/8 <br />1/6 <br />1/4 <br />FIA` <br />1.9 <br />2.3 <br />2.4 <br />2.6 <br />2.9 <br />3.4 <br />5.3 <br />0.9 <br />1.2 <br />1.3 <br />1.4 <br />1.6 <br />2.0 <br />3.2 <br />0.4 <br />0.4 <br />0.4 <br />0.4 <br />0.5 <br />0.5 <br />0.8 <br />0.6 <br />0.7 <br />0.7 <br />0.8 <br />0.8 <br />0.9 <br />1.3 <br />Percent of plot area <br />Additional parcels Percent tree cover <br />First <br />parcel" Penn. req.Y Perm. quest.' No perm.' Mean" SE` RSE` <br />84 <br />78 <br />76 <br />74 <br />70 <br />65 <br />48 <br />9 <br />13 <br />14 <br />15 <br />17 <br />20 <br />27 <br />2 <br />3 <br />3 <br />4 <br />4 <br />5 <br />8 <br />5 25.8 1.1 4.1 <br />7 26.1 0.9 3.3 <br />7 26.2 0.8 3.1 <br />7 26.3 0.8 2.9 <br />8 26.4 0.7 2.6 <br />8 26.6 0.6 2.2 <br />17 26.2 0.8 3.0 <br />'1/24 ac = 0.017 ha; 1/12 ac = 0.034 ha; 1/10 ac = 0.04 ha; 1/8 ac = 0.05 ha; 1/6 ac = 0.067 ha; 1/4 ac = 0.1 ha. <br />'Permission required (residential land). <br />"Permission requirement is questionable; uncertain if crew would need to obtain permission (commercial/industrial; institutional; utility/transportation). <br />"No permission needed (greenspace; street right-of-way; vacant). <br />"Average percent of plot within parcel where plot center is located. <br />"Average tree cover in Syracuse = 26.6%. <br />`Standard error. <br />`Relative standard error (SE/mean x 100). <br />`USDA Forest Service, Forest Inventory and Analysis plot design of four 1/24 ac (0.067 ha) subplots. <br />However, when subdividing the analysis into smaller units (e.g., <br />species, land use), the RSE will tend to increase. To increase <br />precision for various estimates, more crews could be used to <br />collect more plot data by either increasing plot size and/or in- <br />crease the number of plots. In addition, stratification of plots in <br />similar groups (e.g., land use classes, as done in the UFORE <br />analyses) tends to increase precision. Increasing the number of <br />plots from 200 to 500 will likely reduce the RSE on the total <br />number of trees to 7.7% (a 36% reduction). Thus, increasing the <br />number of plots enhances the precision of the estimate, but at an <br />increased cost. <br />A sampling of 150 to 200 plots is a reasonable sample size <br />given the costs associated with measuring field plots during a <br />summer season and a goal of maximizing reduction in SE of the <br />estimates per unit cost. If sample size increases to greater than <br />200 plots, it is likely a second field crew will be needed to collect <br />the additional plot data. Thus, increasing sample size to greater <br />than 200 plots increases costs (adding an additional crew) with <br />50.0 <br />u.0 <br />10 30 50 70 00 110 130 150 170 1E0 210 230 250 270 2E0 010 330 350 370 937 410 49C 450 -070 400 <br />Number or Plebs <br />Figure 1. Estimated relative standard error (SE/total x 100) of <br />total number of trees based on varying number of total one - <br />tenth acre (0.04 ha) field plots. <br />relatively minimal gains in the reduction in SE as compared with <br />the first 200 plots sampled. Increasing the plot size from one - <br />tenth acre (0.04 ha) to one -sixth acre (0.067 ha) will also likely <br />reduce the RSE by approximately 16% to 20%. However, in- <br />creasing the plot size will increase the number of permissions <br />needing to be obtained for the sample and thus the overall project <br />time required. <br />CONCLUSION <br />Data gathered on urban forest structure is essential to improve <br />urban forest management. Random sampling offers a relatively <br />easy means to accurately assess urban forest structure and sub- <br />sequently estimate its ecosystem services and values. The pre- <br />cision and cost of the estimate is dependent on sample and plot <br />size. Managers need to plan their data collection procedures <br />properly to ensure a desired precision of the estimate and ad- <br />equately plan for data collection costs. Ensuring that the proper <br />variables are collected will help guarantee that the data are useful <br />for urban forest management. Incorporating these data within <br />models to assess ecosystem services and values, and within long- <br />term management and monitoring plans, can help improve urban <br />forest health and sustain or increase urban tree cover and con- <br />sequently environmental and human health in urban and urban- <br />izing areas. <br />Acknowledgments. This work was funded, in part, by the USDA Forest <br />Service, Forest Health Monitoring Staff. We thank Sue Sissini for as- <br />sistance with field data collection. We also thank Drs. Jerry Bond and <br />John Stanovick for their review of an earlier draft of this manuscript. <br />LITERATURE CITED <br />Cumming, A.B., D.B. Twardus, and D.J. Nowak. 2008. Urban forest <br />health monitoring: Large scale assessments in the United States. Ar- <br />boriculture and Urban Forestry 34:341-346. <br />Dwyer, J.F., D.J. Nowak, M.H. Noble, and S.M. Sisinni. 2000. Assess- <br />ing our Nation's Urban Forests: Connecting People With Ecosystems <br />in the 21st Century. USDA Forest Service Gen. Tech. Rep. PNW- <br />460. 540 pp. <br />i-Tree. 2007. i-Tree Software Suite v1.2 User's Manual. www.itreetools. <br />org (accessed 7/23/2007). <br />©2008 International Society of Arboriculture <br />