Carbon Storage and Oregon’s Land-Use Planning Program

storage (e.g., Wayburn et al. 2000 and Perez-Garcia et al. 2004). Urban forests also retain some capacity to store carbon in shade trees and associated landscaping and play an important role in conserving energy in home and business heating and cooling (McPherson and Simpson 1999, Nowak and Crane 2002). Alig (2003) suggests that a mixture of forest-based greenhouse gas reduction strategies nationally may be beneficial, including afforestation and reforestation incentives, and land-use policies focused on retaining forestland. Although it is important to consider how forests and forest products can be managed for carbon storage, it may simply be the act of maintaining or increasing the amount of land area in forest cover and use that is the most important determinant in avoiding increases in greenhouse gas emissions. In Oregon, forestland conservation occurs primarily through the State’s land-use planning program (Oregon Department of Land Conservation and Development 1997). Enacted in 1973, the state directs counties and municipalities to focus new development inside urban growth boundaries and to restrict development outside of urban growth boundaries by zoning those lands for agricultural or forest use. We estimate the amount of carbon loss avoided in western Oregon as a result of conserving forest and agricultural lands from development under Oregon’s land-use planning program. We identify average carbon stock values for different land-use categories based on a review of literature and available data. We apply these values to estimates of low-density and urban development in western Oregon described in previous studies. We estimate carbon benefits for two scenarios: (1) assuming Oregon’s land-use planning program as enacted in 1973 and (2) assuming Oregon’s land-use planning program was not enacted in 1973. Calculations are in terms of units of carbon because forests store carbon, but we report the landuse planning effects on stored carbon in terms of CO2 because CO2 equivalency is the standard for emissions reporting (1 unit of carbon is equivalent to 3.67 units of CO2). Our calculations only account for changes in carbon stocks arising from development and do not account for changes in CO2 emissions arising from the change in land use. For example, the calculations show a gain in carbon storage when cultivated agriculture land is developed because of increased carbon storage in shade trees and landscaping ascribed to development; however, we ignore that the developed land-use activity may emit more CO2 emissions than agricultural activity. Our results suggest that Oregon’s land-use planning program has played a significant role in avoiding CO2 emissions from development by avoiding losses of forest carbon, highlighting the importance of maintaining Oregon’s forestland base to any strategy designed to reduce greenhouse gas emissions. Oregon’s Land-Use Planning Program Enacted in 1973, Oregon’s land-use planning program often is cited as a pioneering approach to maintaining resource lands for agricultural and forest uses, by concentrating human populations and business development within urban growth boundaries (Gustafson et al. 1982, Abbot et al. 1994). The foundation of the program is a set of 19 statewide planning goals that define the state’s policies on land use and related topics, including maintaining productive agricultural lands (goal 3) and commercial forestlands (goal 4) and conserving open spaces, scenic and historic areas, and natural resources such as fish and wildlife habitats (goal 5). The program directs local governments to plan and adopt land-use regulations by developing local comprehensive plans, based on standards set by the state. This process has resulted in a system of stateapproved local comprehensive plans that cover the entire state and define where development of different types can occur on forest, agricultural, and other resource lands. Oregon’s land-use planning program generally is considered successful with respect to its original purpose as the state has incurred relatively modest losses of farm and forests to development since the program was implemented. In western Oregon, e.g., losses of forest, mixed forest/agricultural, and agricultural lands to development from 1973 to 2000 are estimated at 348,000 ac (Lettman 2002). In eastern Oregon, losses of forest, range, mixed forest/range, mixed forest/agriculture, and agricultural lands to development from 1975 to 2001 are estimated at 192,000 ac (Lettman 2004). In total, Oregon has lost just 540,000 ac of agricultural, range, and forestlands to development from the mid-1970s to the turn of the century; roughly 21,600 ac/year. Sixty-four percent of those losses occurred in western Oregon—primarily in the Willamette Valley and in southern Oregon communities along Interstate 5. Different data and analysis from the US Department of Agriculture, Natural Resource Conservation Service’s (NRCS) National Resources Inventory (Nusser and Goebel 1997) of land cover for nonfederal lands in Oregon found comparable losses of resource lands to development, with 19,560 ac lost annually from 1982 to 1997 (NRCS 2003). The National Resource Inventory (NRI) estimates of conversion of forestland to urban land within Oregon were developed by the Oregon NRCS utilizing queries to the national NRI database through the web-based Online Analysis System. All estimates were derived from the 1997 NRI database (revised December 2000). Oregon’s land-use planning program is not without its opponents. Since its inception, the program has created tension between its advocates, who see land-use planning as necessary to the long-term conservation of forest and farmlands, and its detractors who argue that land-use regulations unduly burden private landowners. That tension culminated with the passage of a 2004 ballot measure—Measure 37— which mandates “just compensation” to landowners when land-use regulations enacted after the current owner or a family member became the owner of the property, reduces the fair market value of the property by restricting its use. In lieu of compensation, Measure 37 allows government jurisdictions to not apply the regulation, effectively nullifying the conservation intent of the program. Projecting Forestland Development and Estimating Stored Carbon Previous research has evaluated changes in forest and farmland development patterns from 1974 to 1994, resulting from implementation of the forest and farmland zoning and urban growth boundaries found in Oregon’s land-use planning program (Kline 2005). The evaluation relies on relatively well-documented spatial land-use modeling techniques created and tested for western Oregon (Kline 2003, Kline et al. 2003). The model describes changes in building densities as a function of prevailing development pressures, existing building densities, slope, elevation, and land-use zoning. Development pressures are represented in the model using a gravity index describing the spatial 168 Journal of Forestry • June 2007 location of land relative to western Oregon cities of varying population sizes. As computed, the gravity index describes the commuting opportunities offered by land in different locations, which are assumed to be a primary factor affecting development. The model was used to estimate future building densities on forest and farmlands based on historical populations of cities included in the gravity index computation. The conservation of Oregon’s forest and farmland base under the State’s land-use planning program was evaluated using zoning variables included in the empirical model, which enable computing estimated distributions of forest and farmlands among building density classes, with and without land-use zoning in effect. The land-use projections define three stages (or classes) of increasing development that affect the resource land-use categories of forest, agriculture, and mixed forest/agriculture. These development classes include relatively undeveloped (0–16 buildings/mi), low density (16–64 buildings/mi), and developed (greater than 64 buildings/mi). The 0to 16-buildings/mi density class describes relatively undeveloped lands consistent with average minimum parcel sizes of 40 ac/building (house). The 16to 64buildings/mi density class describe lowdensity developed lands consistent with average parcels sizes ranging from 10 to 40 ac. The greater than 64-buildings/mi density class describes relatively developed lands consistent with maximum average parcel sizes of 10 ac or less (Kline 2005). To estimate potential future forest carbon benefits resulting from forest and agricultural land conservation, land-use change estimates for the period of 1973–94 provided in Kline (2005) were extended to 2024 and used to compute potential gravity index values for future years. The extended land-use change estimates were based on reported population numbers for the 1994–2004 period and projected population numbers for the 2005–24 period (Office of Economic Analysis 1997). The potential gravity index values were then entered into the land-use change model reported in Kline (2005) to project future distributions of forest and agricultural lands among building density classes for 2004, 2014, and 2024, with and without Oregon’s land-use planning program in effect (Figure 1). The estimates of land-use change from the 1974 baseline year to 2004 are based on numbers depicting actual population growth, while the land-use estimates for 2014 and 2024 are based on state-reported population projections. We assumed that forestlands contain carbon stocks approaching live tree stocks found in young, second-growth Douglas-fir (Pseudotsuga menziesii) forest conditions. An average carbon stock value of 45.4 metric tons of carbon per acre was used (Table 1). The value corresponds to the live tree carbon stocks listed in the 2006 US Department of Energy’s voluntary reporting of greenhouse gas program technical guidelines for forests for a 25-year old Douglas-fir stand created from afforestation for the Pacific Northwest Region–West (Smith et al. 2006 as reported in US Department of Energy 2006). This is a simplifying assumption that ignores species composition, forest structure and type, dead and down woody material, and other factors (including any effects from climate change) that determine the actual amount of carbon storage on a given forested site. We feel that this value provides a lower bound for estimating avoided emissions from maintaining the forestland base, under the assumption that most forestland converted to nonforest use is likely older than 25 years. Further, projections of climate change impacts to Oregon’s vegetation suggest that Oregon is going to experience an overall increase in vegetative cover and stored biomass overtime, not less (USDA Forest Service 2004). However, our analysis does not give credit to the fact that some of the stored carbon assumed lost from forestland conversion to other land uses actually continues to be stored in wood products derived from the converted forest (Row and Phelps 1996, Perez-Garcia et al. 2004). In forecasting the CO2 emission reduction benefits from afforestation activity in western Oregon, the Oregon Office of Energy awarded 35% of the preharvest carbon stock as continuing in storage in forest products; although this carbon pool was further debited for wood decay emission of 1.5%/year thereafter (Oregon Office of Energy 1996). Agricultural land use in western Oregon was assumed to be cultivated cropland use. The average carbon stock factor for cultivated agriculture was 1.8 metric tons of nonsoil carbon per acre (Table 1). This value is consistent with the nonforest vegetation cover carbon stock reported in the voluntary reporting of greenhouse gas technical forestry guidelines for afforestation in the Pacific Northwest–West region (US Department of Energy 2006). We assumed that the carbon stocks for low-density residential use to be 2.5 times the carbon stock value for cultivated agriculture since the class of development contain a mixture of land cover in pasture, noncultivated agricultural, and woody agricultural (e.g., orchards, Christmas trees, and berry farms) uses. This value, 4.5 metric tons of nonsoil carbon per acre (Table 1), falls between the cropland and pastureland nonsoil carbon stocks calculated from values reported for the Pacific Coast region in Birdsey (1996). The average carbon stock value for urban lands in western Oregon was compiled from a study of changes in forest canopy cover and associated impacts to carbon stocks for the Willamette and Lower Columbia Rivers of western Oregon (American Forests 2001). Specifically, the value in Table 1 is the estimated per acre carbon stocks for urban forest canopy cover for the cities of Portland (and surrounding metropolitan Figure 1. Loss of undeveloped forest, agriculture, and mixed forest/agriculture lands with and without land-use planning. Journal of Forestry • June 2007 169 area), Albany, Corvallis, Eugene, Salem, and Wilsonville in Oregon and the City of Vancouver in Washington. In all cases, the average carbon stock values applied were aboveground stocks associated with aboveground, live vegetation. We did not account for differences in dead wood and soil organic carbon pools under agriculture and forestland use. The implication is that the results presented here likely underestimate the avoided carbon emissions from conserving forestland because both of these pools, especially for forests, store more carbon than under a developed land use. Average carbon stocks were assigned to the land-use and building development classes used in the land-use projections according to Table 1. Incremental avoidance of forest carbon stock losses from conserving forest, agricultural, and mixed forest/ agricultural lands were estimated by computing the difference between the carbon stock equivalent for the mix of land uses found under Oregon’s land-use planning and the carbon stock equivalent for the mix of land uses that might have occurred had the program not been implemented. In general, carbon stocks were reduced on forest and agricultural lands falling into the low-density developed and developed classes, relative to those lands remaining in the undeveloped class. For example, every acre of forestland that was converted to low-density development results in a decrease in carbon stocks of 40.9 metric tons of carbon because the original forest contained 45.4 metric tons carbon per acre and the resulting carbon stock after development is 4.5 metric tons of carbon per acre. Similarly, a loss of 1,000 ac of undeveloped forest to low-density development results in a loss of carbon stocks of 40,900 metric tons of carbon, or 150,103 metric tons of CO2 equivalent.