Investigating Morphospace Occupation in Multi-Scale Ecological and Evolutionary Data Using Regression Tree: Case Studies and Perspectives

A key challenge in ecology and evolutionary biology is to explain the origin, structure and temporal patterns of phenotypic diversity. With regard to the potentially complex determinism of phenotypic differences, the issue should be comprehended in a general view, across multiple scales and an incre...

Full description

Saved in:
Bibliographic Details
Published in:Evolutionary biology 2017-03, Vol.44 (1), p.120-134
Main Author: Vignon, Matthias
Format: Article
Language:English
Subjects:
Analysis
Animal Genetics and Genomics
Biodiversity and Ecology
Biomedical and Life Sciences
Case studies
Developmental Biology
Ecology
Environmental Sciences
Evolution
Evolutionary Biology
Fishes
Human Genetics
Investigations
Learning algorithms
Life Sciences
Machine learning
Mathematical models
Morphology
Statistical analysis
Tools and Techniques
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A key challenge in ecology and evolutionary biology is to explain the origin, structure and temporal patterns of phenotypic diversity. With regard to the potentially complex determinism of phenotypic differences, the issue should be comprehended in a general view, across multiple scales and an increasing number of phenomic studies investigate shape variation through large taxonomic, biogeographic or temporal scales. In this context, there is an ever-increasing need to develop new tools for a coherent understanding of morphospace occupation by disentangling and quantifying the main determinants of phenotypic changes. The present study briefly introduce the possibility to use multivariate regression tree technique to cope with morphological data, as embedded in a geometric morphometric framework. It emphasizes that hierarchical partitioning methods produce a hierarchy between causal variables that may help analyzing complexity in multi-scale ecological and evolutionary data. I therefore suggest that morphological studies would benefit from the combined use of the classical statistical models with rapidly emerging and diversifying methods of machine-learning. Doing so allows one to primary explore in an extensive exploratory manner the hierarchy of nested organisational levels underlying morphological variation, and then conduct hypothesis-driven analysis by focusing on a relevant scale or by investigating the appropriate model that reflects hypothesized nested influence of explanatory variables. The outlined approach may help investigating morphospace occupation in an explicitly hierarchical quantitative context.
ISSN:0071-3260
1934-2845
DOI:10.1007/s11692-016-9386-3