Forest Distribution and Topographic Features -- A Cell-Based Modeling Application

Susan L. Ustin^a, Assistant Professor of Resource Science Minghua Zhang^a, Postdoctoral Associate Paul Grant^b, GIS Analyst

^aDepartment of Land, Air, and Water Resources ^bInformation Technology/ANSA

Author for Correspondence:
Susan L. Ustin
Department of Land, Air, and Water Resources
University of California
Davis, CA 95616
Phone: (530) 752-0621
FAX: (530) 752-5262
email: slustin@ucdavis.edu

Abstract

Significant gradients in climatic and vegetation exist with elevation and topographic features in the Sierra Nevada range. The Eldorado National Forest can serve as a "model forest" for studying some physiographic interactions, as it is located on the western slope of the central part of the range. Within the national forest a wide variety of ecological communities and conditions occur, which represent many of the features that characterize the entire Sierra Nevada range. The elevation gradient and topographically related conditions within the Eldorado National Forest produce abrupt climatic differences over relatively short distances which affect species distributions and abundances. To investigate these physiographic patterns and to illustrate the spatial relationships between distributions of specific forest types and topographic features, we compared their basic statistical relations in a Geographic Information System (GIS). Thirty-seven USGS Digital Elevation Models (DEM) at 1 :24,000 were joined in ARC/INFO GRID for topographic analysis of the forest. The USFS forest inventory data base was used to generate relationships between forest distributions, stand structure, and topographic features. These relationships will be used to develop a model of ecological dynamics within the forest. The analysis indicated that 90% of the forests were in the slope ranges between 0 and 50%, which were found throughout the elevation range of 300 to 3300 meters. About 70% of forests are composed of small and medium timber size classes.

Each forest type had its own elevation distribution but all have overlapping slope ranges. A tendency for forests to be distributed on north facing aspects at their lowest elevational limits and on west or south facing aspects at their highest elevation was observed. This study will discuss an ecological modeling application that interfaces statistical models and GIS output using GRID. The resulting ecological rules describe physiographic characteristics for each forest type and the procedures used for developing these rules may themselves provide a useful research tool for predicting new vegetation distributions under different climate scenarios. The new GIS developed maps will provide a direct view of the topographic dependence of different ecological communities and vegetation cover in a complex heterogeneous environment. This approach to integrated environmental analysis will provide better information for planning and management of environmental resources.

INTRODUCTION

Montane forests of the Sierra Nevada range are among the largest and most economically important vegetation types in California. An accurate spatial map of vegetation distribution is critical to understanding many urgent ecological questions, including knowledge of the distribution of late-successional forests, the sustainability of habitat for wildlife conservation, and for understanding physiological processes, e.g., annual primary production, energy balance, and water budgets. Ecologically sound management and planning of the forests of the Sierra Nevada require understanding of the factors that control vegetation distribution and ecological succession over the region. Although many factors, e.g., edaphic influences, competition, predation, atmospheric conditions, etc., influence the distribution of vegetation, topographic features are among the most important factors. Elevation, slope and aspect determine the total precipitation, net radiation, and air and surface temperatures, soil moisture and runoff, that directly and indirectly interact to affect the distribution of the vegetation. Although the distribution of native forests are expected to be closely related to topographic patterns (Rundel et al., 1988) in some of these forests, because of extensive commodity extraction or other resource use over the past 100 years, many may not express these patterns and there, current ecological distributions may be more closely related to patterns of human activities.

Much research has been done by California ecologists to define the relationships between forests and environmental variables. Because of the Mediterranean climate in California, precipitation (both timing and abundance) is one of the critical physical factors controlling the distribution of vegetation (Barbour, 1988; Rundel et al., 1988; Major, 1989). Soil moisture gradients are dependent on the site conditions that are themselves largely a function of aspect, slope, soil type, and profile (Hole and Campbell, 1985; Ward and Robinson, 1990). Although past reviews (e.g., Rundel et al., 1988) have summarized the physical, physiological and ecological relationships for these forests, they have not been examined in the synoptic view possible with GIS techniques.

This paper will identify and illustrate some spatial relationships between the distribution and abundance of Sierran forests and topographic features. This preliminary work is an initial step in a longer-term research effort to model dynamic ecosystem processes for predicting future ecological characteristics in the Sierra Nevada.

MATERIALS AND METHODS

The Eldorado National Forest (ENF) was chosen as the case study area (Figure 1) because of its central location in the Sierra Nevada range and because it includes most forest transitions typical of the larger Sierra Nevada range. The forest maintains habitats for diverse wildlife species (about 320 species of birds, mammals, amphibians and reptiles) and has also experienced a high level of human disturbance in recent decades. Because of its proximity to urban centers, Eldorado is an important year-round recreation area and has been a valuable source of forest commodities for more than 100 years.

Most data in this study were obtained from the U.S. Forest Service, including digitalisheries in the region. The model represents a new synthesis of GIS, remote sensing, and modeling on a critically important hydrologic problem.

Mr. Xiao has nine years of remote sensing research at the Chinese Academy of Sciences, Remote Sensing Application Center of Ministry of Water Resources P.R.C. (China), before coming to U.C. Davis as a graduate student in 1992. He has completed his MS degree in two years (GPA is 4.0 at UCD), during which time he has become proficient in using the ARC/Info GISing these data.

Slope was classified into four groups, low (0-30%), medium (31-50%), high (5170%), and very steep (>70%). Aspects were divided into 8 equal groups between 0 and 360 degrees (Table 2). Elevation contours at 100 meter intervals were used in the analysis. Furthermore, basic descriptive statistics were spatially plotted and other map overlays were used in the study.

RESULTS AND DISCUSSION

The ENF is dominated by mixed conifer forests (fir and pine), including ponderosa pine, red fir, and sub-alpine, which are distributed across the range of elevation from 300 to 3300 meters (Figure 2). When all forest ecosystems are considered within the ENF, they are found to exhibit a large range of slope classes from 0 to over 100% (Table 1), and show nearly equal distribution with aspect (Table 2). Forests found on steeper slopes are distributed mostly on the west and southwest aspects. These types of basic physiographic patterns provide useful information for the management of forest resources. The total annual precipitation and temperature are strongly related to the elevation range and aspect. The complex topographies and resulting large climatic variations directly affect the distribution of specific forest ecosystems within the ENF.

Eighteen conifer and broadleaf tree species grow within the ENF, and occur in a wide range of tree size classes and densities. However, only six species are dominant (Table 3) and these comprise 89.3% of all forested land in the EMF. The most abundant forest classes are mixed conifer pine, mixed conifer fir, ponderosa pine, red fir, sub-alpine conifer and non-forest, respectively (Table 3). Most of these forests (over 77% of the total vegetation cover) are composed of trees in the small and medium size classes and these units have a mean density class rating of more than 40% crown closure.

About 70% of all forests today occur with slope gradient less than 30%, and over 20% occurs in the next slope class, between 31 to 50% (Table 1). Eighty-six percent of forests are distributed between elevations of 1000 to 2600 meters. Table 3 shows that forests are distributed over the full elevation range with some, e.g., ponderosa pine, distributed primarily at the lower elevations and sub-alpine forest at highest elevations, and the rest in intermediate elevation zones. However, when the analysis is restricted to forests containing the largest trees (class 3G and 4N), they are found to be distributed primarily in the higher elevations. The distributions of younger tree classes were disproportionately abundant at lower elevations in the forest. The small and medium timber size classes are found throughout the range of density classes and stand density was not significantly related to elevation. The forest types, size classes, and the densities, were not significantly associated with aspect in the forest except that the eastern aspects had the lowest vegetation cover. We found that 78% of the total forest cover was within the size 3 and 4 timber classes (small and medium timber) shown in Table 4, and that 54% of the forest cover had medium and high (greater than 40%) crown closure. Most timber areas of small and medium size classes with dense crown closures were within the slope ranges between 0 and 50%. These findings are consistent with an ENF composed of predominantly younger forests or those in an early to mid-successional status relative to forests of predominantly late seral characteristics (lower density forests of high crown closure composed of largest size-class trees). Historic records indicate that the Sierra Nevada forests have always supported a mosaic of variable aged forests. Evaluation of the age distribution structures in relation to the historic distributions was not attempted in this study.

Figure 2 shows that the major forests differ from each other with respect to abundance, aspect, elevation, and slope. The patterns of forest distributions are somewhat asymmetric with aspect. These patterns could be related to climate and energy balance constraints or a both climate and land use.

On the basis of these findings, we believe that further research focused on modeling the ecological relationships between topographic features and vegetation distribution in the ENF will be useful. The vegetation distribution integrates the complexity of the physiographic factors, therefore, it is essential to characterize the environmental variables directly associated with distribution. Using a multivariate statistical approach, an integrated environmental index can be identified to quantify this response. This type of processes-based index is derived from intermediate variables obtained from a spatial analysis of climate data in the GIS, which is used to drive a model of physiological functioning (e.g., Running and Coughlan, 1988; Running and Grower, 1991, Running and Hunt, 1993; Bonan, 1989). A statistical model can be built using methods employed in the "Chipmunk" model (Ed Royce, UC Davis, personal communication). Such a model can be constructed to predict the vegetation cover for each forest species based on topographic and climate constraints. This simulation method can be useful for predicting the vegetation dynamics under different scenarios. In addition, various additional kinds of spatial relationships can be identified through the simulation modeling effort.

To conclude, we found that the distribution of forest age and size classes in the ENF was strongly related to the topographic features of elevation, slope, and aspect. Some current patterns appear to be related to the pattern of forest extraction in the past century and others to climate conditions. The analysis also provided us with a clear direction for developing a model to describe the spatial relationships among the factors. GIS is the critical tool to a landscape scale ecological analysis.

Acknowledgments: We wish to thank Ralph Warbington, USFS (Sacramento, CA) and the Sierra Nevada Ecosystem Project (SNEP) for providing us with the U.S. Forest Service Forest Timber Inventory data base for the Eldorado National Forest and Steve Beckwitt, Sierra Biodiversity Institute for assisting us with the AML elevation classification.

References

Barbour, M.G. 1988. California upland forests and woodlands in Barbour, M.G. and W.D. Billings, eds: North American Terrestial Vegetation, Cambridge University Press. p. 131-164.

Bonan, G.B. and Shugart, H.H. 1989. Environmental factors and ecological processes in boreal forests. Ann. Rev. Ecol. Systm. 20: 1-28.

Rundel, P.W., Parsons, D.J., and Gordon, D.T., 1988. Montane and subalpine vegetation of the Sierra Nevada and Cascade Ranges. in Barbour, M.G. and Major, J. eds: Barbour, M.G. and J. Major, eds: Terrestrial vegetation of California, California Native Plant Society. p. 559601.

Clements, E.S. 1920. Plant indicators: the relation of plant communities to process and practice. Carnegie Institution of Washington. p. 290.

ESRI, 1992. ARC/TNFO GRID Users Guide. Environmental Systems Research Institute, Redlands, CA.

Hole, F.D., and Campbell, J.B. 1985. Soil Landscape Analysis. Rowman and Allanheld Publ., Towata, NJ, p. 191.

Major, J. 1988. California Climate in relation to vegetation, in Barbour, M.G. and J. Major, eds: Terrestrial vegetation of California, California Native Plant Society. p. 11-74.

Major, J. 1982. Climate, in Burke, M.T., Curry, R., Major, J. and D.W. Taylor, eds: Natural landmarks of the Sierra Nevada, Department of Botany, University of California, Davis. p. 1228.

Royce, E.B. 1993. Climate effects on the distribution of vegetation in the upper montane zone of the southern and central Sierra Nevada range -- the Chipmunk climate model. (Unpublished materials) Department of Environmental Horticulture, University of California, Davis.

Running, S.W. and Coughlan, J.C. . 1988. FOREST-BGC, a general model of forest ecosystem processes for regional applications. I. Hydrologic balance, canopy gas exchange and primary production processes. Ecol. Model. 42: 125-154..

Running, S.W. and Grower, S.T. 1991. FOREST-BGC, a general model of forest ecosystem processes for regional applications. II. Dynamic carbon allocation and nitrogen budgets. Tree Physiol. 9: 147-160.

Running, S.W. and Hunt, E.R., Jr. 1993. Generalization of a forest ecosystem process model for other biomes, BIOME-BGC, and an application for global scale models. In Scaling Physiological Processes: Leaf to Globe (J. Elhringer and C. Fields, eds.) Academic Press, p. 141-158.

Ustin, S., Szeto, L.H.., Xiao, Q.F., Hart, Q.J., and E.S. Kasischke. 1994. Vegetation mapping of forested ecosystem in interior central Alaska. IGARSS 94: Proceedings of the International Geosciences and Remote Sensing Symposium, T. I. Stein, ed., Pasadena, CA August 8-12, 1994.

Ward, R.C., and Robinson, M. 1990. Principles of Hydrology. McGraw-Hill Book Co. San Francisco, CA, p. 365.

1998, Center for Spatial Technologies and Remote Sensing (CSTARS)
University of California, Davis