DIVISION S-5—PEDOLOGY Modeling Soil–Landscape and Ecosystem Properties Using Terrain Attributes

demonstrated the successful implementation of quantitative soil–landscape modeling on a broad landscape Soil–landscape patterns result from the integration of shortand scale useful for soil resource inventory and as a framelong-term pedogeomorphic processes. A 2-ha hillslope catena in Caliwork for understanding soil–landscape function. Moore fornia shows short-distance variation in A horizon depth from 8 to 80 cm and in soil depth from 8 to .450 cm in convex to concave et al. (1993) and Gessler et al. (1995) showed strong positions. Similar variations in net primary productivity (NPP) and correlation and predictive utility between a digital tersoil C represent significant information often not captured by soil rain index, the compound topographic index (CTI), and survey maps. Strong correlations between these measured soil– several soil properties. The CTI, often referred to as landscape variables and explanatory digital terrain attributes are used the steady-state wetness index, is a quantification of to develop quantitative soil–landscape models. We were able to accatenary landscape position. It integrates both landform count for between 52 and 88% of soil property variance using easily position and context through an index defined as computed terrain variables such as slope and flow accumulation. Spatial implementation of the models suggest lateral redistribution proCTI 5 ln(As/tan b) [1] cesses resulting in differential accumulation of C and soil mass in where As is the specific catchment area [area (m2) per convergent and divergent landscape positions. The models are explicit unit width orthogonal to the flow direction] and b is and quantitative, which enables their use for testing hypotheses about the slope angle. Small values of CTI generally depict the spatial distribution of fine-scale landscape and ecosystem processes and for parameterizing spatially distributed hydrological and upper catenary positions and large values lower cateecosystem simulation models. nary positions with an overall range typically from 2 to 12 for zero-order upland areas. The purpose here was to continue development of T understanding that topography modifies water the quantitative catena concept at the hillslope scale for flow and material redistribution processes and rea detailed study site in southern California. Whereas sulting ecosystem and soil patterns in landscapes lies at the earlier catena study by Moore et al. (1993) focused the heart of the catena concept (Milne, 1935, 1936). It entirely on patterns of surficial soil properties (in the has been reformulated and modified by numerous studtop 15 cm), we began the process of integrating catena ies (Conacher and Dalyrymple, 1977; Hole and Camppatterns with pedologic processes. Specifically, we combell, 1985; Dikau, 1989; Moore et al., 1993; Gessler et bined developed methods of digital terrain analysis, soil al., 1995; Sommer and Schlichting, 1997) often with the and ecosystem field sampling, statistical modeling, and aim of refining a conceptual framework for soil mapping pedological and geomorphological process interpreor landscape process monitoring and interpretation (Sitation. monson, 1959; Smeck et al., 1983). Detailed hillslope characterization and modeling of The advent of geographic information system (GIS)digital terrain and soil property relationships is key to based digital elevation modeling has facilitated quantiunderstanding the biogeochemical cycling of nutrient tative modeling of catenas at local hillslope scales elements in terrestrial ecosystems because they often (McSweeney et al., 1994). Moore et al. (1993), building point to the spatial distribution of significant processes. on the work of Speight (1968) and others, developed Thus, one of our results is development of explicit soil– quantitative soil–landscape models using an integration landscape models describing the distribution of ecosysof digital terrain analysis and statistical modeling methtem net primary production of C and soil C along the ods to map predicted soil properties for a small hillslope catena. The models provide preliminary context for dein Colorado. In Australia, this work was continued and veloping hypotheses about soil–landscape function with expanded to a broader suite of parent materials and respect to C cycling and water balance. We believe that geomorphic settings to test an integration of tools for understanding soil–landscape dynamics at the local hilldigital terrain analysis, field sampling, remote sensing, slope catena level is key to understanding how to scale statistical modeling, and GIS (Gessler et al., 1995, 1996; C dynamics to broader regional and global levels. In Gessler, 1996; McKenzie et al., 2000). These workers this paper, we provide a methodological blueprint for quantitative catena studies and results that have regional significance for soil survey and the biogeochemiP.E. Gessler, Dep. of Forest Resources, Univ. of Idaho, Moscow, ID cal interpretation of landscapes. 83844-1133; O.A. Chadwick, F. Chamran, and K. Holmes, Dep. of Geography, Univ. of California, Santa Barbara, CA 93106; L. Althouse, Dep. of Ecology, Evolution & Marine Biology, Univ. of CaliforAbbreviations: CEC, cation-exchange capacity; CTI, compound toponia, Santa Barbara, California, 93106. Received 28 Sept. 1999. *Corregraphic index; Cmass, profile total C mass; DEM, digital elevation sponding author ([email protected]). model; GIS, geographic information system; GPS, global positioning system; NPP, net primary productivity. Published in Soil Sci. Soc. Am. J. 64:2046–2056 (2000).