Skip to Main content Skip to Navigation
Conference papers

Behavioral biology: Behavioral phylogeny in ruminants

Abstract : The use of behavioural characters in phylogenetics has a long and controversial history. Early approaches were clearly behaviour-based phylogenetic inferences, or mere self-contained macro-evolutionary scenarios. Decisive conceptual and methodological improvements in phylogeny inference led to a renewed interest in behavioural phylogenetics. Modern phylogenetics are best construed as a historical science, both for ‘phylogeny making’ and for ‘phylogeny use’ in comparative studies. Very contrasting views have emerged regarding the use of behaviour in such approaches. Is behaviour generally too plastic to carry reliable phylogenetic information? Should behavioural traits be used to establish phylogenetic relation-ships, or just mapped on an already made phylogeny based on supposedly more reliable grounds? Such problems invite analyses of the requirements for characters to be used in phylogenetics. Both phylogenetic inference and testing of evolutionary scenarios on a phylo-geny require reliable homology assessment. If we don't think that characters are likely homologous, we should not force them into a single evolutionary scenario, be it before or after phylogeny construction. The desirable criteria for establishing reliable hypotheses of homology involve special structure, connections, development and functions, as exemplified in the analysis of some behavioural characters of ruminants. Accurate resolution of phylogenetic relationships requires the use of all relevant evidence, analyzing data sets separately or simultaneously. Ideally, the relative reliability of different kinds of characters needs to be recognized and incorporated in any analysis to effectively combine different topologies produced from separate analyses or to produce topologies based on different types of characters (molecular, morpho-logical, behavioral) analyzed simultaneously. This is particularly important for the integration of behavioural analyses in phylogenetics. Relying on only one kind of character (e.g. molecular) is a popular but arbitrary solution. The notion of relative reliability of characters questions the evolutionary laws of character evolution (if any...), like possible differential probabilities of some character changes, or concerted versus independent evolutionary processes for similar characters in different hylogenetic lineages. These concepts are illustrated in comparative studies of behaviour in ruminants. We examined stereotyped courtship behaviours in Cervidae. We found that behavioural-based phylogenetic relationships revealed similar relationships to those based on morphological or molecular characters. The primary separation of Cervinae from Capreolinae/Odocoilei-nae was strongly supported. Our behavioural approach failed to resolve relationships among some clades but these were typically where other approaches show similar problems. This may be indicative of real phylogenetic polytomies due to rapid specific radiation. The question of when to stop phylogenetic investigations in cases of suspected 'true polytomy' is well worth being treated seriously to avoid possible arbitrary over-resolution of phylogenetic relationships and corresponding macroevolutionary scenarios. Behavioural characters have some particularities, just as other phenotypic characters, when compared to molecular ones: existing in a range of more or less precisely defined states, being more diverse and hence potentially resistant to saturation than molecular characters, while being less easy to delineate and classify. But they are not so difficult to classify as to be discarded from phylogenetic analysis. A major difficulty consists in documenting the behavioural repertoire of numerous species with sufficient completeness and precision. Exhaustive field studies are very demanding, and the available literature may be of doubtful reliability for some less well-studied species. Ideal investigations should discriminate between species-specific, stable differences and plastic, context-dependent, intraspecific behavioural variants. The potential epigenetic objections are not acceptable as a general argument against any comparative behavioural study, but should be considered as a reason to investigate behavioural plasticity. This often necessitates extensive, specialized field work or experimental work. This has been a major limitation in the use of behavioral characters in phylogenetic studies, much more so than objections based on the quality of behavioural characters per se. Behavioural phylogenetics is not unsound. It is intellectually attractive, but it requires a lot of work to be valid. Recently, we have seen impressive technical developments in the computing programs for phylogenetic analysis, including the capacity to treat ever enlarging sets of characters and taxa, and the growing possibility of implementing different evolutionary 'models' for different kinds of characters in a simultaneous analysis. In my view, these model-based approaches invite us to concentrate our theoretical reflections on our evolutionary biological assumptions, in an attempt to make them more explicit than implicit. Clearly, phylogenetic inferences cannot be made without implementing some notions of evolutionary processes. What do we know about such evolutionary processes, which allows us to interpret characters in a historical phylogenetic perspective? If our knowledge is limited, we must recognize the extent of our ignorance and remain correspondingly cautious regarding the risk of over-resolution of weakly supported phylogenetic relationships. I do not view this as a reason to renounce these analyses but rather as a way to improve our use and interpretation of all possible phylogenetic information. Given the present mix of success and continued difficulties in ruminant phylogenetics, macroevolutionary studies on this group provide a stimulating field of investigations for pushing forward the very young science of phylogenetics.
Document type :
Conference papers
Complete list of metadatas
Contributor : Umr6552 Ethos <>
Submitted on : Thursday, June 16, 2016 - 11:14:37 AM
Last modification on : Tuesday, March 17, 2020 - 3:14:36 AM


  • HAL Id : hal-01332581, version 1


Pierre Deleporte, Henri Cap. Behavioral biology: Behavioral phylogeny in ruminants. International Conference on Ruminant Phylogeny (ICRPM13), Université Louis-et-Maximilien de Munich, Sep 2013, Munich, Germany. ⟨hal-01332581⟩



Record views