RESEARCH ARTICLE A New Method for Delineating Tree Patches and Assessing Spatial Reference Conditions of Ponderosa Pine Forests in Northern Arizona Andrew J. S´ anchez Meador, 1,2,3 Pablo F. Parysow, 1 and Margaret M. Moore 1 Abstract Multiscale spatial patterns of forest reference conditions offer insight on how historical and environmental processes have influenced forest stand dynamics. Yet, spatial infor- mation is often either unavailable or partial, because many studies provide nonspatial reference condition information, whereas others only report the local (i.e. observed at plot extent) spatial arrangement of trees. However, knowledge of multiscaled spatial patterns, including stand-, among- patch-, and within-patch-level patterns, is needed to ensure that forest management strategies suit a wide range of objectives in ecosystem restoration. In this study, we pro- pose a new framework for delineating tree patches based on common stand attributes. We found that patch refer- ence conditions for presettlement ponderosa pine in north- ern Arizona ranged from 10 to 27 patches/ha, accounted for 62–75% of the total historic stand basal area (m 2 /ha) and varied in size from 0.01 to 0.15 ha. Lastly, discussion was made on the importance of using a patch-delineation framework, such as the one proposed in this study, as a basis for quantifying stand- and patch-level patterns of reconstructed presettlement and current forest conditions and how this information should guide spatially explicit restoration management (e.g. thinning prescriptions). Key words: Arizona, G. A. Pearson, presettlement, Pinus ponderosa, spatial reference conditions, T. S. Woolsey, Jr., tree patches, Woolsey plots. Introduction Vegetation spatial pattern is determined by a complex combi- nation of mechanisms including initial conditions, disturbance history (i.e. harvesting, livestock grazing, fire exclusion, nat- ural, or prescribed fires), species life history (e.g. seed dis- persal), and the underlying heterogeneity inherent to the envi- ronment (Law et al. 1993). Either natural or anthropogenic in origin, these processes influence change in the structure of forest communities through time and space (Oliver & Larson 1996), often obscuring prior patterns. Whereas it is not possi- ble to indisputably determine that the observed spatial pattern of vegetation has resulted from a specific process (Pielou 1961, 1974), quantitative descriptions of spatial patterns offer insight into the scale at which historical and environmental mech- anisms influence forest stand dynamics (Youngblood et al. 2004; Boyden et al. 2005; Fortin & Dale 2005). Numerous studies have examined the importance of spatial patterns in a wide variety of ecosystems (Watt 1947; Greig-Smith 1952; 1 School of Forestry, PO Box 15018, Northern Arizona University, Flagstaff, AZ 86011, U.S.A. 2 Address correspondence to A. J. S´ anchez Meador, email asanchezmeador@fs.fed.us 3 Present address: Forest Management Service Center, U.S. Forest Service, 3463 Las Palomas Road, Alamogordo, NM 88310, U.S.A.  2010 Society for Ecological Restoration International doi: 10.1111/j.1526-100X.2010.00652.x Kershaw 1959; Kenkel 1988; Rebertus et al. 1989; Duncan 1991; Leemans 1991; Camarero et al. 2000; S´ anchez Meador et al. 2009). The importance of understanding the spatial patterns of ref- erence conditions and their influence on stand dynamics is of particular interest throughout the frequent-fire conifer forests of the western United States. Throughout these systems, pre- settlement and remnant old-growth stands often exhibit an aggregated tree distribution (Harrod et al. 1999) in uneven- aged patches, except for the random tree arrangement found in a few ponderosa pine (Pinus ponderosa Laws. var. scop- ulorum ) stands in Oregon, United States (Youngblood et al. 2004). Canopy openings alternating with tree patches are thought to typify within-forest patterns in many frequent-fire conifer forests (Abella et al. 2007). Furthermore, this is par- ticularly true for the southwestern ponderosa pine ecosystem of the United States where grass openings account for the highest level of plant diversity (Laughlin et al. 2006) and spa- tial patterns have been shown to influence genetic diversity (DeWald 2003), growth of large trees (Biondi 1996; Ffolliott et al. 2000), forest dynamics (Youngblood et al. 2004; Boy- den et al. 2005; S´ anchez Meador et al. 2009), wildlife habitat (Graham et al. 1994; Ffolliott et al. 1977; Meyer & Sisk 2001; Waltz & Covington 2003; Dodd et al. 2006; Wightman & Ger- maine 2006), and risk of catastrophic, stand-replacing crown fires (Ful´ e et al. 2007). Restoration Ecology 1