Using geometric and non‐geometric internal evaluators to compare eight vegetation classification methods

Abstract

Questions: How similar are solutions of eight commonly used vegetation classification methods? Which classification methods are most effective according to classification validity evaluators? How do evaluators with different optimality criteria differ in their assessments of classification efficacy? In particular, do evaluators which use geometric criteria (e.g. cluster compactness) and non‐geometric evaluators (which rely on diagnostic species) offer similar classification evaluations? Methods: We analysed classifications of two vegetation data‐sets produced by eight classification methods. Classification solutions were assessed with five geometric and four non‐geometric internal evaluators. We formally introduce three new evaluators: PARTANA, an intuitive variation on evaluators which use the ratio of within/between cluster dissimilarity as the optimality criterion, an adaptation of Morisita's index of niche overlap, and ISAMIC, an algorithm which measures the degree to which species are either always present or always absent within clusters. Results and Conclusions: 1. With the exception of single linkage hierarchical clustering, classifications resulting from the eight methods were often similar. 2. Although evaluators varied in their assessment of best overall classification method, they generally favored three hierarchical agglomerative clustering strategies: flexible beta (β=– 0.25), average linkage, and Ward's linkage. 3. Among introduced evaluators PARTANA appears to be an effective geometric strategy which provides assessments similar to C‐index and Gamma evaluators. Non‐geometric evaluators ISAMIC and Morisita's index demonstrate a strong bias for single linkage solutions. 4. Because non‐geometric criteria are of interest to phytosociologists there is a strong need for their continued development for use with vegetation classifications.

Vegetation Heterogeneity and Life-Strategy Diversity in the Flora of the Heterogeneous Landscape of Nizanda, Oaxaca, Mexico

“…We performed three classifications of the floristic information from the 188 sampling plots (presence/absence data), using Ward's method and Euclidean distances (Kent and Coker 2003;Aho et al 2008). The analyses were performed at the species, genus and family levels.…”

A hierarchical analysis of vegetation on a Mojave Desert landscape, USA

“…An application proposed for ISA is to identify optimal levels of clustering (i.e. number of groups) in hierarchical community analyses (Dufrêne and Legendre, 1997;Aho et al, 2008). For example, when clustering beetle communities in Belgium, Dufrêne and Legendre (1997) found that the strength of indicator species had different peaks between two and 10 groups but was optimized at a three-group level.…”

When rare species are not important: linking plot-based vegetation classifications and landscape-scale mapping in Australian savanna vegetation

“…We determined classifications for datasets using agglomerative hierarchical clustering, and internal evaluators to determine the level of cluster division (Aho et al 2008). All analyses were undertaken in the software package PRIMER v6 (Clarke and Gorley 2006) or in the R environment (R Development Core Team 2014).…”

“…There are necessarily subjective choices inherent in any classification process (Kent 2012) and these will influence outcomes (Aho et al 2008, Tichý et al 2010, Lotter et al 2013, Lengyel and Podani 2015. To find species which indicate landscape level changes we have used species nominated by experts.…”