Obvious growing interest to multiscale organization of landscapes is encouraged by the necessity to transfer information between hierarchy levels (Wu & David, 2002; Burnett & Blaschke, 2003). Dominant landscape-ecological models focus either on the hierarchical organization of a single natural geocomponent or phenomenon (most commonly that being plant cover, land use or relief, which can be easily detected from DEM and space imagery) or they describe relations at a single hierarchical level. In this study, a new tool is proposed, in order to reveal multiple independent hierarchies based on interactions between properties of geocomponents. Constraints from higher-order systems are interpreted as emergent effects, evolving as a result of interactions between neighboring spatial units. The properties of a landscape unit are treated as a product of both interior interactions between geocomponents and spatial relations with surrounding units. The proposed procedure enables one to compare a set of hypotheses about the spatial extent of a higher-order system, while choosing the appropriate size. The spatial properties of a higher-order geosystem can be described in a number of ways. Here, we use a combination of landforms as a surrogate for characterizing spatial heterogeneity and describe it by morphometric values (standard deviation of elevations, summary length of thalwegs, vertical curvature, and horizontal curvature). Hierarchical levels are not postulated a priori but are induced on the basis of an evaluation of linkages between the soil-vegetation properties of the focus unit and spatial emergent properties of the higher-order geosystem. Next, response surface regression equations were compared in order to evaluate contributions of external (inter-level) and internal (intra-level) interactions to spatial variability of soil and vegetation. We propose a set of both deterministic and probabilistic multiscale cartographic models of partial geosystems. Probabilistic landscape mapping resulted in the identification of areas with perfect adaptation of soils and vegetation to abiotic environment as well as of areas with high possibility of sustaining several stable states. We identified three principal ways of subordination to higher-order geosystems for the properties of soil and plant cover. First, the property can be controlled by influence of the other geocomponents without contribution from external factors. Second, the property can be influenced by ecological processes of a single higher-order geosystem that imposes strict constraints on the possible range of values. Third, the property can be influenced by emergent effect of the several higher-order geosystems, i.e. to a set of broad-scale processes. The research confirmed the hypothesis that the combined effects of several higher-order geosystems provide emergent effects in low-order landscape units.
Burnett, C., & Blaschke, T. (2003). A multi-scale segmentation/object relationship modeling methodology for landscape analysis. Ecological Modeling, 168, 233-249.
Wu, J., & David, J.L. (2002). A spatially explicit hierarchical approach to modelling complex ecological systems: theory and applications. Ecological Modelling, 153, 7-26.