Premium
Towards a more balanced combination of multiple traits when computing functional differences between species
Author(s) -
Bello Francesco,
BottaDukát Zoltán,
Lepš Jan,
Fibich Pavel
Publication year - 2021
Publication title -
methods in ecology and evolution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.425
H-Index - 105
ISSN - 2041-210X
DOI - 10.1111/2041-210x.13537
Subject(s) - trait , categorical variable , multivariate statistics , function (biology) , curse of dimensionality , trait theory , r package , quantitative trait locus , biology , ecology , statistics , computer science , mathematics , psychology , big five personality traits , evolutionary biology , social psychology , biochemistry , personality , gene , programming language
Functional trait differences between species are key drivers of community assembly and ecosystem functioning. Quantifying these differences routinely requires using approaches like the Gower distance to combine various types of traits into a multi‐trait dissimilarity. Without special care, the Gower distance can however produce a multi‐trait dissimilarity with a disproportional contribution of certain traits, particularly categorical traits and bundle of correlated traits reflecting similar ecological functions. These effects persist even after applying multivariate analyses traditionally used to reduce trait dimensionality. We propose the ‘gawdis’ R function, and corresponding package, to produce multi‐trait dissimilarity with more uniform contributions of different traits, including fuzzy coded ones. The approach is based on minimizing the differences in the correlation between the dissimilarity of each trait, or groups of traits, and the multi‐trait dissimilarity. This is done using either an analytical or a numerical solution, both available in the function. Properly taking into account the contribution of multiple traits into multi‐trait dissimilarity is key for interpreting the ecological effects of complex species differences. The gawdis r package in CRAN can be further applied to improve equitability in distance‐based measures in other field of research, such as social sciences or marketing surveys, which routinely analyse mixed type data.