Levels and limits in artificial selection of communities. (11th August 2015)

• Record Type: Journal Article
• Authors: Blouin, Manuel; Karimi, Battle; Mathieu, Jérôme; Lerch, Thomas Z.; Grover, James
• Is Part Of: Ecology letters. Volume 18:Number 10(2015:Oct.)
• Page Start: 1040
• Page End: 1048
• Publication Date: 2015-08-11
• Subjects: Ecology -- Periodicals; 577
• DOI: 10.1111/ele.12486
• Languages: English
• ISSNs: 1461-023X

Abstract

<abstract abstract-type="main" id="ele12486-abs-0001"> <title>Abstract</title> <p>Artificial selection of individuals has been determinant in the elaboration of the Darwinian theory of natural selection. Nowadays, artificial selection of ecosystems has proven its efficiency and could contribute to a theory of natural selection at several organisation levels. Here, we were not interested in identifying mechanisms of adaptation to selection, but in establishing the proof of principle that a specific structure of interaction network emerges under ecosystem artificial selection. We also investigated the limits in ecosystem artificial selection to evaluate its potential in terms of managing ecosystem function. By artificially selecting microbial communities for low CO₂ emissions over 21 generations (<italic>n</italic> = 7560), we found a very high heritability of community phenotype (52%). Artificial selection was responsible for simpler interaction networks with lower interaction richness. Phenotype variance and heritability both decreased across generations, suggesting that selection was more likely limited by sampling effects than by stochastic ecosystem dynamics.</p> </abstract>