Towards a standardization of new functional trait measurements by assessing intraspecific variation: the case of springtail mandibles

Abstract

Novel quantitative functional traits, such as feeding traits, are increasingly described or applied to new soil invertebrate taxa, but their standardization remains deficient. A better systematic estimation of intraspecific trait variation (ITV) would not only bring conceptual advances (e.g. identifying population and community filters) but also allow for more robust protocols to compare trait data across studies. The aim of this study was to determine ITV of newly described mandibular traits of springtails (Collembola) to improve current measurement protocols and ultimately to demonstrate how such simple assessment can favour trait standardization both at a conceptual and methodological levels. To do so, three common cryptic springtail species (Isotomiella minor, Parisotoma notabilis and Protaphorura armata) were sampled at multiple forest sites in two distinct geographical zones given their potentially higher ITV than non-cryptic species. We assessed factors potentially influencing intraspecific variation of seven feeding traits (i.e. mandibular asymmetry, geographic zone, sex and life stage). Using linear mixed models, we determined the sample size needed to generate a representative mean trait value with bootstrap resampling analyses. Intraspecific trait variation was mostly related to body length differences among individuals of different sex, life stage and/or geographic zone. Our results allowed us to better standardize these springtail mandibular traits by (1) determining the adequate number of replicates needed for each trait, (2) proposing adjustments to the measurement protocol, (3) identifying a potential methodological error, (4) highlighting potential phenotypic plasticity and (5) validating the adequacy of using data derived from the literature and/or databases. Despite being an initial time investment, estimating ITV ultimately speeds up the standardization process. We encourage researchers to estimate ITV when describing new traits or measuring traits on new taxa in order to ensure reliable data and accelerate large-scale comparative efforts.