vol. 170, no. 4 the american naturalist october 2007 ൴ Species Richness and Evolutionary Niche Dynamics: A Spatial Pattern–Oriented Simulation Experiment Thiago Fernando L. V. B. Rangel,1,* Jose´ Alexandre F. Diniz-Filho,2,† and Robert K. Colwell1,‡ 1. Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut 06269; 2. Departamento de Biologia Geral, Instituto de Cieˆncias As early as the eighteenth and nineteenth centuries, nat- Biolo´gicas, Universidade Federal de Goia´s, CP 131, 74001-970 uralists described and documented what we today call geo- Goiaˆnia, Goiaˆnia, Brasil graphical gradients in taxon diversity (species richness), Submitted November 27, 2006; Accepted May 14, 2007; especially the general global pattern of increase in species Electronically published August 9, 2007 richness toward warm and wet tropical regions (Whittaker et al. 2001; Hawkins et al. 2003b; Willig et al. 2003; Hil- Online enhancements: appendixes. lebrand 2004). Initial hypotheses explaining this pattern were deduced solely by observing and describing nature and were based on nothing more rigorous than intuitive correspondence between climatic and biological patterns abstract: Evolutionary processes underlying spatial patterns in (Hawkins 2001). Surprisingly, even after 200 years of re- species richness remain largely unexplored, and correlative studies search in biogeography and ecology, the most common lack the theoretical basis to explain these patterns in evolutionary framework used in such investigations still relies on sta- terms. In this study, we develop a spatially explicit simulation tistical measurements of the concordance between the spa- model to evaluate, under a pattern-oriented modeling approach, whether evolutionary niche dynamics (the balance between niche tial patterns in species richness and multiple environmen- conservatism and niche evolution processes) can provide a parsi- tal factors. The generally strong relationship between monious explanation for patterns in species richness. We model species richness and some of these environmental factors the size, shape, and location of species’ geographical ranges in a (e.g., Hawkins et al. 2003a) has led many researchers to multivariate heterogeneous environmental landscape by simulating conclude that environment (e.g., climate) is the main an evolutionary process in which environmental fluctuations create driver of species richness, based on a purely correlative geographic range fragmentation, which, in turn, regulates specia- tion and extinction. We applied the model to the South American view (see Hawkins et al. 2003b). Until recently, the pro- domain, adjusting parameters to maximize the correspondence be- cesses underlying these correlations were usually defined tween observed and predicted patterns in richness of about 3,000 by verbal arguments without formal modeling or even bird species. Predicted spatial patterns, which closely resemble ob- clear epistemological structures (but see Currie et al. 2004; 2 served ones (r p 0.795 ), proved sensitive to niche dynamics pro- Scheiner and Willig 2005). However, in the past decade, cesses. Our simulations allow evaluation of the roles of both evo- more complex models were developed to evaluate envi- lutionary and ecological processes in explaining spatial patterns in species richness, revealing the enormous potential of the link be- ronmental or climate-based hypotheses, based on the effect tween ecology and historical biogeography under integrated the- of environmental variables on organismal metabolism and oretical and methodological frameworks. physiology or on population dynamics (see Allen et al. 2002; Field et al. 2005; O’Brien 2006). Despite these ad- Keywords: niche conservatism, birds, South America, simulation model, latitudinal gradient, diversity gradients. vances, the evolutionary components underlying these correlations between climate and species richness remain poorly understood (Ricklefs 2006). * Corresponding author; e-mail: [email protected], tf.rangel@gmail .com. In contrast with climate-based hypotheses, evolutionary † E-mail: [email protected], [email protected]. hypotheses have developed since the early 1960s as almost ‡ E-mail: [email protected]. independent sets of explanations (Ricklefs 2004, 2006; Mit- telbach et al. 2007). In general, evolutionary hypotheses Am. Nat. 2007. Vol. 170, pp. 602–616. ᭧ 2007 by The University of Chicago. 0003-0147/2007/17004-42244$15.00. All rights reserved. invoke a geographical (i.e., latitudinal) bias in net diver- DOI: 10.1086/521315 sification rates (for recent evaluations, see Cardillo et al. Species Richness and Evolutionary Niche Dynamics 603 2005; Weir and Schluter 2007) driven by area availability during or immediately following speciation events, in the (Rosenzweig 1995) or climatic factors (Wright 1983; Rohde traits that define the niche, allowing descendant species to 1992, 1999; Wright et al. 1993). A geographical bias in net adapt rapidly to new environmental conditions (Wiens and diversification rates has been suggested to be caused by a Donoghue 2004). Thus, it has been hypothesized that if spatially patterned buffer against extinction (Hawkins et niche conservatism prevails over niche evolution in regions al. 2005, 2006) or by an acceleration of tropical speciation of highly heterogeneous and/or asynchronously fluctuating rates due to metabolic activation (Rohde 1992; Bromham environments, diversification might occur predominantly and Cardillo 2003; Allen et al. 2006). The most common by a process of range fragmentation, caused by the inability approach to studying these evolutionary hypotheses at- of species to adapt to changing environmental conditions tempts to decouple the effects of “ecological” (contem- in portions of the ancestral range (Wiens 2004; Wiens and porary) and “evolutionary” (historical) components of the Donoghue 2004). In this scenario, once environmental fluc- spatial patterns in species richness (e.g., Hawkins et al. tuation has fragmented a species’ geographical range into 2003b; Diniz-Filho et al. 2004; Bjorholm et al. 2006), al- isolated populations (Dynesius and Jansson 2000; Jackson though usually one or the other is simply ignored or cred- and Overpeck 2000; Jansson and Dynesius 2002; Ackerly ited with unexplained variation. 2003; Wiens 2004), geographic (and eventually reproduc- However, as highlighted by Wiens and Donoghue (2004) tive) isolation of those isolated populations would lead to and Hawkins et al. (2005), historical biogeography and allopatric speciation (see also Kirkpatrick and Barton 1997; ecology have much to offer each other, and perhaps a Holt 2003) and possibly adaptive radiation (Gavrilets and better approach would be to join the two perspectives into Vose 2005). a unified theoretical and analytical framework capable of The balance between niche conservatism and niche evo- resolving the tangled and/or interactive effects of both eco- lution may also have played an important role in deter- logical and evolutionary processes affecting species rich- mining contemporary geographical gradients in species ness (Ricklefs 2006). Furthermore, Currie et al. (2004), richness (Wiens and Donoghue 2004; Wiens and Graham who recently reviewed those hypotheses that invoke cli- 2005; Ricklefs 2006). According to this hypothesis, most matic factors to explain spatial patterns in species richness clades originated under a tropical climatic regime, which under a hypothetical-deductive, Popperian approach, con- had, for most of Earth’s history, and still has, greater geo- cluded that these hypotheses are still very difficult to eval- graphical extent and environmental stability than extra- uate and that the biological underlying mechanisms that tropical climates (Wallace 1878; Stephens and Wiens 2003; link climate to species richness remain to be discovered. Wiens and Donoghue 2004; Hawkins et al. 2005; Jablonski Phylogenetic niche conservatism (Harvey and Pagel et al. 2006). If descendant lineages tend to conserve an- 1991; Peterson et al. 1999; Ackerly 2003; Holt 2003) refers cestral niche characteristics (e.g., Ricklefs and Latham to an evolutionary pattern in which descendant species 1992), then clades may have been slow to spread toward tend to share, by common descent, a substantial propor- extratropical regions or are now extinct in those regions tion of the biological and physiological characteristics that that are not tropical anymore (e.g., Hawkins et al. 2005, determine their fundamental ecological niches (Hutch- 2006). Such a process would tend to generate a higher inson 1957). Thus, the adaptations of a common ancestor richness in the ecological zone of origin (tropical regions) to a particular set of environmental conditions (i.e., the because of a higher net diversification rate, whereas species ecological zone of origin; Holt and Gaines 1992; Holt 1996; in more recently derived clades could potentially become Wiens 2004) tend to be conserved in descendant species, adapted to temperate conditions, breaking away from phy- with little adaptive biological modification (e.g., Ricklefs logenetic constraints on niche characteristics (the “out of and Latham 1992; Ricklefs 2006). Whether a particular tropics” model; e.g., Jablonski et al. 2006). Thus, the trop- pattern of niche conservatism is the result of constraints ics would be both the “cradle” and the “museum” of spe- on adaptive evolution (e.g., absence of appropriate genetic cies diversity (Jablonski et al. 2006). variation)