vol. 170, supplement the american naturalist august 2007Correlates of Diversification in the Plant Clade Dipsacales:Geographic Movement and Evolutionary InnovationsBrian R. Moore*and Michael J. Donoghue†Department of Ecology and Evolutionary Biology, Yale University,New Haven, Connecticut 06520abstract: We explore patterns of diversification in the plant cladesAdoxaceae and Valerianaceae (within Dipsacales), evaluating corre-lations between biogeographic change (i.e., movements into newareas), morphological change (e.g., the origin of putative key in-novations associated with vegetative and reproductive characters),and shifts in rates of diversification. Our findings indicate that ratesof diversification in these plants tend to be less tightly correlatedwith the evolution of morphological innovations but instead exhibita pronounced correlation with movement into new geographic areas,particularly the dispersal of lineages into new mountainous regions.The interdependence among apparent novelties (arising from theirnested phylogenetic distribution) and the correlation between mor-phological and biogeographic change suggests a complex history ofdiversification in Dipsacales. Overall, these findings highlight theimportance of incorporating biogeographic history in studies of di-versification rates and in the study of geographic gradients in speciesrichness. Furthermore, these results argue against a simple deter-ministic relationship between dispersal and diversification: like otherfactors that may influence the probability of speciation and/or ex-tinction, the impact of dispersal on diversification rates depends onbeing in the right place at the right time.Keywords: biogeographic movement, dispersal, diversification rates,key innovation, speciation, extinction.Differences in species richness among geographic regionsultimately reflect differences in the durations that theseareas have been occupied by the organisms of interest (i.e.,historical biogeography) and/or differences in rates of di-versification (i.e., speciation minus extinction) of thegroups in those areas. Our focus here is on shifts in di-versification rate as a factor underlying geographic patternsin species diversity. Both the origin of intrinsic traits—* Corresponding author; e-mail: [email protected].†Order of authorship is arbitrary.Am. Nat. 2007. Vol. 170, pp. S28–S55. 䉷 2007 by The University of Chicago.0003-0147/2007/1700S2-42201$15.00. All rights reserved.including morphological, behavioral, and physiologicalnovelties—and extrinsic events—including episodes ofbiogeographic or climatic change—have been hypothe-sized to promote shifts in diversification rate by alteringthe probability of speciation and/or extinction. A com-prehensive understanding of the causes of diversificationtherefore requires incorporation of both elements. Sur-prisingly, the relative impact of both biotic and abioticfactors on diversification rates has seldom been evaluatedin an explicitly phylogenetic context.In plants, the search for intrinsic correlates of diversi-fication has tended to focus on a few putative key inno-vations relating to floral morphology, growth form, andfruit type/dispersal mechanism. The role of extrinsic fac-tors in plant diversification has received less attention;however, an important class of abiotic factors pertains tothe impact of biogeography on diversification. Dispersalto new geographic areas could conceivably increase theprobability of diversification in several ways. For example,the dispersing species may diversify in response to eco-logical opportunities afforded by the new area if, for in-stance, it is well suited to exploit resources in the new areathat are underexploited by residents of that community.Alternatively, the new area may provide a measure of “eco-logical release” for the dispersing species if that area lacksantagonistic elements (competitors, predators, etc.) thatinhibit diversification (e.g., via competitive exclusion).Conversely, abiotic aspects of the new area may promotediversification either by increasing the opportunity for al-lopatric speciation (e.g., owing to greater topographiccomplexity; Carlquist 1974; von Hagen and Kadereit 2003;Xiang et al. 2004; Hughes and Eastwood 2006) or by de-creasing the incidence of extinction (e.g., owing to greaterclimatic stability/reduced environmental harshness; Cra-craft 1985; Qian and Ricklefs 1999).The relationship between geographic dispersal, mor-phological change, and shifts in diversification rate can beconceived of as a cube in which the cells delimit the setof possible interactions (fig. 1). Of special interest are out-comes in which shifts in diversification are correlated withgeographic movement (which we refer to as dispersifica-Correlates of Diversification in Dipsacales S29Figure 1: Graphical model of the relationship between morphologicaland biogeographic variables and diversification rate. The axes of this cubedesignate change or stasis in morphology, biogeographic distribution,andrates of diversification, such that its cells collectively circumscribe the setof possible correlations between each of the three variables. Our focusis on scenarios restricted to the upper tier of the cube, which entailssignificant shifts in diversification rate. Within this tier, shifts in diver-sification rate may occur independent of changes in the other two var-iables (cell 1, which may be correlated with some unexplored factor),rate shifts may be correlated with the evolution of a morphological trait(cell 2, which is consistent with a key innovation scenario), rate shiftsmay be correlated with the dispersal to a new biogeographic area (cell3, which we refer to as a dispersification scenario), or rate shifts may becorrelated with both morphological and biogeographic change (cell 4).Our goal is to explore the extent to which inferred diversification rateshifts populate the cells of this cube.tion), character evolution (key innovation), or changes inboth variables. Our goal is to explore how the cells of thiscube are occupied. Note that this exercise falls squarely inthe realm of data exploration. To the extent that we aretesting hypotheses, they are fairly prosaic: the null hy-potheses under evaluation are that rates of diversificationare stochastically constant and that any shifts in diversi-fication rate are not significantly correlated with morpho-logical or biogeographic change. We hope to show that anexploratory approach to
View Full Document
Unlocking...