BOTANICAL BRIEFINGCalibrating the Tree of Life: fossils, molecules and evolutionary timescalesFe´lix Forest*Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UKReceived: 9 April 2009 Returned for revision: 11 June 2009 Accepted: 9 July 2009 Published electronically: 8 August 2009† Background Molecular dating has gained ever-increasing interest since the molecular clock hypothesis was pro-posed in the 1960s. Molecular dating provides detailed temporal frameworks for divergence events in phyloge-netic trees, allowing diverse evolutionary questions to be addressed. The key aspect of the molecular clockhypothesis, namely that differences in DNA or protein sequence between two species are proportional to thetime elapsed since they diverged, was soon shown to be untenable. Other approaches were proposed to takeinto account rate heterogeneity among lineages, but the calibration process, by which relative times are trans-formed into absolute ages, has received little attention until recently. New methods have now been proposedto resolve potential sources of error associated with the calibration of phylogenetic trees, particularly those invol-ving use of the fossil record.† Scope and Conclusions The use of the fossil record as a source of independent information in the calibrationprocess is the main focus of this paper; other sources of calibration information are also discussed. Particularlyerror-prone aspects of fossil calibration are identified, such as fossil dating, the phylogenetic placement of thefossil and the incompleteness of the fossil record. Methods proposed to tackle one or more of these potentialerror sources are discussed (e.g. fossil cross-validation, prior distribution of calibration points and confidenceintervals on the fossil record). In conclusion, the fossil record remains the most reliable source of informationfor the calibration of phylogenetic trees, although associated assumptions and potential bias must be takeninto account.Key words: Calibration, fossil, incompleteness, molecular dating, rate heterogeneity, relaxed molecular clock,uncertainty.INTRODUCTIONThe use of DNA sequences to estimate divergence times onphylogenetic trees (molecular dating) has gained increasinginterest in the field of evolutionary biology in the pastdecade. The abundance of publications on the subject, thenumerous alternative methods proposed and the often heateddebates on various aspects of the discipline demonstrate theinterest it generates. The molecular clock hypothesis wasfirst proposed by Zuckerkandl and Pauling (1965); they pro-posed that differences in DNA (or protein) sequencesbetween two species are proportional to the time elapsedsince the divergence from their most recent common ancestor.The subsequent inclusion of temporal frameworks in manyevolutionary studies has influenced the way results are inter-preted and significantly modified the way in which conclusionsare drawn from these findings. Linking the evolution of par-ticular morphological characters or key ecological innovationsto geological, climatic or biotic events is much improved in thelight of an evolutionary timescale. The development of mol-ecular dating tools became particularly valuable to the disci-pline of historical biogeography; it added a temporal gaugeto the directionality of events demonstrated by the topologyof phylogenetic trees. Inferences on observed distribution pat-terns were rendered significantly more plausible under a tem-poral framework, even if only descriptive. Furthermore, newmethods of biogeographical reconstruction have been devel-oped such as Lagrange, which uses a likelihood frameworkto infer the evolution of geographical ranges and incorporatesdivergence times as well as constraining the connectionsbetween areas to specific times (Ree and Smith, 2008).The rationale of the molecular clock hypothesis, thatevolutionary rates are constant, was shown to be invalid inthe majority of examined cases; the clock does not tick regu-larly. The heterogeneity of substitution rates among differentlineages in a phylogenetic tree explains this irregularity(Britten, 1986) and is a result of species-specific factors suchas generation time, metabolic rate, effective population sizeand mutation rates (see Rutschmann, 2006). The extent ofinfluence of some such factors, however, remains in dispute(e.g. Whittle and Johnston, 2003).Rutschmann (2006) classified the most commonly employedmethods for estimating divergence times into three categoriesdepending on how they handle rate heterogeneity, namely (1)assuming a global substitution rate (standard molecular clock);(2) correcting for rate heterogeneity (e.g. by deleting branchesor incorporating several rates categories before the dating pro-cedure), and (3) incorporating rate heterogeneity (i.e. integrat-ing rate heterogeneity into the dating procedure using ratechange models; relaxed molecular clock). The four most com-monly used methods in the literature all fall into the third cat-egory; these are non-parametric rate smoothing (NPRS;Sanderson, 1997), penalized likelihood (PL; Sanderson,2002), the Bayesian method implemented in the Multidivtimepackage (Thorne et al., 1998) and Bayesian evolutionaryanalysis by sampling trees (BEAST; Drummond andRambaut, 2007). The first three of these methods assume* E-mail [email protected]# The Author 2009. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved.For Permissions, please email: [email protected] of Botany 104: 789–794, 2009doi:10.1093/aob/mcp192, available online at www.aob.oxfordjournals.orgrate changes between ancestral and descendant lineages areautocorrelated, i.e. that substitution rates in descendantlineages are to an extent inherited from ancestral lineages;these methods differ in the way that rate autocorrelation ishandled. BEAST does not assume rate autocorrelation;instead, it samples rates from a distribution. Additional flexi-bility is found in BEAST in its optional tree topology require-ment that can incorporate phylogenetic uncertainty, and thepossibility of assigning distributions to the calibrationprocess a prior i (see below). More details on these methodsand several others are available elsewhere (Rutschmann,2006, and references therein).Two main topics have fuelled the controversy associatedwith molecular clocks: these are how to handle rate heterogen-eity and calibration. At its outset,
View Full Document
Unlocking...