Factors biasing the correlation structure of patch level landscape metrics

Possibilities of land use change analysis in a mountainous rural area: a methodological approach

“…A narrative description of class level data is an efficient method to demonstrate the landscape dynamics and can be supported by PCA biplot diagrams to reveal trends over time. Landscape metrics can be correlated but not always redundant: the provided information can be interpreted in different ways depending on the aims of the investigations (Uuemaa et al 2011, Szabó et al 2014.…”

“…In addition to this, patterns, that is the absolute and relative status of patches, the distance from the next patch belonging to the same land cover class, and linkage density, also influence species distribution (Pimm 1984, Williams et al 2002. Although there are several landscape metrics, they often correlate accounting for redundant information (Riitters et al 1995), thus, their universality should also be considered (Cushman et al 2008, Szabó et al 2014.…”

Effects of landscape heterogeneity on the elevated trace metal concentrations in agricultural soils at multiple scales in the Pearl River Delta, South China

“…Landscape heterogeneity has two major components: compositional and configuration heterogeneity (Fahrig et al, 2011;Szabo et al, 2014;Wu, 2012). Potential contributing factors for trace metal concentrations include landscape heterogeneity expressed as landscape matrix of bedrock geology (Deschenes et al, 2013), landscape composition with various cover types (Fritsch et al, 2010(Fritsch et al, , 2011, and configuration in terms of the distance to a potential pollutant source (e.g.…”

ShrinkShape2: A FOSS toolbox for computing rotation‐invariant shape spectra for characterizing and comparing polygons

“…Clearly, existing tools for characterizing shape are not to be replaced by ShrinkShape2 , but rather to be augmented by it. With ShrinkShape2 , since it provides detailed analyses of individual polygons, results are more similar to patch‐level pattern analyses (Szabó et al ) and results are local to individual shapes rather than average landscape changes, which are more appropriately assessed as pattern, not shape, alteration (Turner and Ruscher ). Forman and Collinge (1997) also discuss the need for explicit measurements of landscape entities to facilitate improved mimicking or preserving of natural processes and landscapes in a conservation context than relying on non‐spatial approaches.…”