Assignment of relevés to pre‐defined classes by supervised clustering of plant communities using a new composite index

Abstract

Question: How does a newly designed method of supervised clustering perform in the assignment of relevé (species composition) data to a previously established classification. How do the results compare to the assignment by experts and to the assignment using a completely different numerical method? Material: Relevés analysed represent 4186 Czech grassland plots and 4990 plots from a wide variety of vegetation types (359 different associations or basal communities) in The Netherlands. For both data sets we had at our disposal an expert classification, and for the Czech data we also had available a numerical classification as well as a classification based on a neural network method (multi‐layer perceptron). Methods: Two distance indices, one qualitative and one quantitative, are combined into a single index by weighted multiplication. The composite index is a distance index for the dissimilarity between relevés and vegetation types. For both data sets the classifications by the new method were compared with the existing classifications. Results: For the Czech grasslands we correctly classified 81% of the plots to the classes of an expert classification at the alliance level and 71% to the classes of the numerical classification. Correct classification rates for the Dutch relevés were 64, 78 and 83 % for the lowest (subassociation or association), association, and alliance level, respectively. Conclusion: Our method performs well in assigning community composition records to previously established classes. Its performance is comparable to the performance of other methods of supervised clustering. Compared with a multi‐layer perceptron (a type of artificial neural network), fewer parameters have to be estimated. Our method does not need the original relevé data for the types, but uses synoptic tables. Another practical advantage is the provision of directly interpretable information on the contributions of separate species to the result.

An objective method based on assemblages of subfossil plant macro-remains to reconstruct past natural vegetation: a case study at Swifterbant, The Netherlands

“…Concurrence networks have recently been used by Araújo et al (2011) in studies on climate. For the taxon groups that we created, we identified the most similar plant association(s) described by Schaminée et al (1995aSchaminée et al ( , b, 1996Schaminée et al ( , 1998 and Stortelder et al (1999) via the analytic software package ASSOCIA (Van Tongeren et al 2008). The results of the three samples are consistent with one Fig.…”

“…These taxa lists were then exported as a Cornell Condensed file (cc!) to be able to import the data into the analytic software package ASSOCIA (Van Tongeren et al 2008). …”

“…Because the subfossil taxon lists are incomplete, we modified the ASSOCIA routines so that the list of possible syntaxa was based on the weirdness index only (Van Tongeren et al 2008). The weirdness index is calculated as the sum of all contributions to -2ln (likelihood) for the taxa present in the sample.…”

“…The methodology can be applied in every region for which synoptic tables of vegetation types are available, preferably containing all taxa retrieved in the palaeobotanical sample(s). For the non-palaeo version of the ASSOCIA package, available as a built-in identification tool in TURBOVEG for Dutch vegetation, a dataset of Czech grasslands has already been used to test its applicability in non-Dutch regions (Van Tongeren et al 2008).…”

Classification of mesic grasslands and their transitions of South Transdanubia (Hungary)

“…Though, these associations are not clearly delimited from Pastinaco-Arrhenatheretum (STAN^I] 2008) and the differences are probably so slight that the applied hierarchical classification methods are not able to distinguish them as separate clusters of our current data set. Here we suggest defining these associations by supervised classification tools (BRUELHEIDE 2000) and/or calculating similarity measures (TICHÝ 2005, VAN TONGEREN et al 2008) between reliable reference plot data and the unidentified relevés, and using the resulting assignments for mapping their distribution. We expect that a few relevés would be identified as any of these rarer associations but the majority of Group 8 should be assigned to Pastinaco-Arrhenatheretum.…”