Current discussion on ecological trait selection in functional diversity researchHow plant trait selection can influence analysis results

7 July 2023, by Website Team Biologie

Photo: UHH/Bio

The characteristics of roots are thought to be critical to understanding the form and function of plants in their ecological context. A study involving Universität Hamburg sheds light on how the selection of plant traits, known as functional traits, can significantly alter the conclusion about the ecological role of roots. This discussion was published in the journal Nature.

Plant roots generally remain hidden below the soil surface. Yet their role in plant survival should not be underestimated, as they serve to absorb water and nutrients. Little has been known about how functional root properties relate to functional leaf properties and whether there are global patterns in their interactions. An international research team has now outlined the importance of ecological trait selection for inferring this. 

Principles of trait selection crucial for inference

Recently, two publications (Carmona et al. 2021, Weigelt et al. 2021) examined the importance of roots in understanding plant form and function and reached fundamentally opposite conclusions based on largely overlapping data. The researchers in the current publication in Nature identified two reasons for this. “We found that the number of traits studied and their ecological roles were critical to the outcome of the analysis”, says Prof. Dr. Ina Meier of the Department of Biology at Universität Hamburg and co-author of the study “When only traits that serve the uptake of resources, i.e. carbon in the case of leaves and water and nutrients in the case of roots, were included in the analysis, a correlation between leaf and root traits could be found. However, when traits not directly related to resource uptake were additionally analyzed, such as stem specific density, plant height, or seed mass, no relationship was found between leaf and root traits.” Therefore, the researchers conclude, that the type and number of plant traits studied can steer global statements about plant function in different directions. 

Biological understanding more important than statistical generalization

Another reason for the contrasting results of the two studies mentioned above lies in the statistical analysis of the results. In so-called multivariate statistics, which examines several variables simultaneously, variables in multidimensional spaces can be rotated until they meet a previously defined criterion. However, this rotation also changes the position of the underlying variables, even if this is not directly visible, and thus their interpretation. In addition, the association of the variables in the two studies above with several major axes of the multidimensional space also suggests flexibility in the biological coordination of the plant traits under study. "If we better understand the underlying principles of coordination between leaf and root traits, we can also make better predictions about the response of little-studied or unknown plant species and diverse and complex plant populations to changing climate and global change," says Meier. 
 
The research was funded by the German Research Foundation, among others. It emerged from the sDiv working group sROOT, in which international scientists from Australia, Estonia, Germany, France, the Netherlands, the UK and the USA worked together on scientific questions relating to the functionality of root traits. The German Center for Integrative Biodiversity Research was in charge of the described study.

Original publication
Alexandra Weigelt, Liesje Mommer, Karl Andraczek, Colleen M. Iversen, Joana Bergmann, Helge Bruelheide, Grégoire T. Freschet, Nathaly R. Guerrero-Ramírez, Jens Kattge, Thom W. Kuyper, Daniel C. Laughlin, Ina C. Meier, Fons van der Plas, Hendrik Poorter, Catherine Roumet, Jasper van Ruijven, Francesco M. Sabatini, Marina Semchenko, Christopher J. Sweeney, Oscar J. Valverde-Barrantes, Larry M. York, M. Luke McCormack, The importance of trait selection in ecology,
Nature (2023) 618: E29-E30.
DOI: https://doi.org/10.1038/s41586-023-06148-8