Premium
Statistical approaches for morphological continuous characters: a conceptual model applied to Phytoseiidae (Acari: Mesostigmata)
Author(s) -
MarieStephane Tixier
Publication year - 2013
Publication title -
zoologica scripta
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.204
H-Index - 64
eISSN - 1463-6409
pISSN - 0300-3256
DOI - 10.1111/zsc.12004
Subject(s) - seta , mesostigmata , phytoseiidae , biology , acari , intraspecific competition , zoology , pointwise , taxonomy (biology) , ecology , genus , mathematics , predation , predator , mathematical analysis
Marie‐Stephane, T. (2012). Statistical approaches for morphological continuous characters: a conceptual model applied to Phytoseiidae (Acari: Mesostigmata). — Zoologica Scripta, 42 , 327–334. Species discrimination is certainly the most current and essential taxonomic task. Despite molecular development, species continue to be delimited using morphological characters. This study provides statistical approaches to assess decision rules, using continuous morphological characters, to determine whether specimens examined belong or not to a same species. As species discrimination is usually based on no overlapping between intraspecific distributions, a general statistical approach has been developed to assess, for a character x , the relation between intraspecific overlapping and (i) the differences between the means of specimen lots corresponding to two species and (ii) the differences between the values borne by two specimens belonging to two species. Then, this conceptual approach was applied to the predatory mite family Phytoseiidae, highlighting that the minimal difference between means of two specimen lots belonging to two species should be of 10.58 μm (for setae <65 μm) and 33.99 μm (for setae >65 μm). When no specimen sets are available but only two specimens compared, the model shows that a difference of 13.24 μm (for setae <65 μm) and 31.74 μm (for setae >65 μm) would be sufficient to conclude that these specimens belong to two species. The presently proposed decision rules are assumed to improve species discrimination and to limit synonymies. Further developments will consist in applying this approach to databases containing species features in order to automatically extract the putative synonyms. Furthermore, such decision rules would also be useful to determine whether a species newly described is really new for science.