Biogeographic Analysis of the Woody Plants of the Southern Appalachians
Total Page：16
File Type：pdf, Size：1020Kb
AMERICAN JOURNAL OF BOTANY RESEARCH ARTICLE B IOGEOGRAPHIC ANALYSIS OF THE WOODY PLANTS OF THE SOUTHERN APPALACHIANS: IMPLICATIONS FOR THE ORIGINS OF 1 A REGIONAL FLORA P AUL S. MANOS 2 AND J OSÉ E DUARDO M EIRELES Department of Biology, Box 90338 Duke University, Durham, North Carolina 27708-0338 USA • Premise of the study: We investigated the origins of 252 Southern Appalachian woody species representing 158 clades to ana- lyze larger patterns of biogeographic connectivity around the northern hemisphere. We tested biogeographic hypotheses re- garding the timing of species disjunctions to eastern Asia and among areas of North America. • Methods: We delimited species into biogeographically informative clades, compiled sister-area data, and generated graphic representations of area connections across clades. We calculated taxon diversity within clades and plotted divergence times. • Key results: Of the total taxon diversity, 45% were distributed among 25 North American endemic clades. Sister taxa within eastern North America and eastern Asia were proportionally equal in frequency, accounting for over 50% of the sister-area connections. At increasing phylogenetic depth, connections to the Old World dominated. Divergence times for 65 clades with intercontinental disjunctions were continuous, whereas 11 intracontinental disjunctions to western North America and nine to eastern Mexico were temporally congruent. • Conclusions: Over one third of the clades have likely undergone speciation within the region of eastern North America. The biogeographic pattern for the region is asymmetric, consisting of mostly mixed-aged, low-diversity clades connecting to the Old World, and a minority of New World clades. Divergence time data suggest that climate change in the Late Miocene to Early Pliocene generated disjunct patterns within North America. Continuous splitting times during the last 45 million years support the hypothesis that widespread distributions formed repeatedly during favorable periods, with serial cooling trends producing pseudocongruent area disjunctions between eastern North America and eastern Asia. Key words: Blue Ridge fl ora; eastern North America; phylogeny; phytogeography; Southern Appalachians; woody plants. The rich fl ora of the montane areas of eastern North America the fl oras of the two areas that had become fragmented ( Gray, provides a natural experiment for understanding how one of the 1859 , p. 444). Support for this hypothesis was formalized with largest refuges of temperate plants has evolved over the last 65 fossil data documenting the presence of widespread, high-lati- Myr. The early observation that enough elements of this fl ora tude paleofl oras ( Chaney, 1947 ; Axelrod, 1966 ). The distribu- are strikingly similar to the temperate fl ora of eastern Asia tion and diversity of the “Tertiary boreotropical” and mixed prompted Gray (1846 , 1859 , 1878 ) to advance one of the earli- mesophytic paleofl oras that gave rise to the modern temperate est biogeographic hypotheses, positing that these similarities refuges have been well documented in fossil deposits (e.g., “suggest former continuity, migration, or interchange” between Wolfe, 1975 ; Tiffney, 1985 a , b ; Manchester, 1999 ). These pa- leofl oras provide temporal context to understand the origin of 1 Manuscript received 3 December 2014; revision accepted 21 April modern plant distributions ( Graham, 1999a ; Manchester, 1999 ) 2015. and allow inference of paleoclimatic conditions during the This study was conceived in part by the working group on the Phytogeography of the North Hemisphere (2006-08) with support by the Cenozoic Era ( Wolfe, 1993 , 1994 ; Little et al., 2010 ). Features National Evolutionary Synthesis Center (NESCent). The authors thank Jon such as light, temperature, moisture, and physical continuity Shaw and Alan Weakley for providing thoughtful suggestions on the continue to drive hypotheses on how the relative timing of mi- concept of the study and Ben Carter, Alan Graham, Diane Lennox, John gration and subsequent barrier formation have contributed to McVay, Tom Wentworth, and two anonymous reviewers for comments on the discontinuous distribution of temperate fl oras across the the manuscript. Duncan Hauser prepared the tables and fi gures. They also northern hemisphere ( Tiffney and Manchester, 2001 ; Milne, thank Dylan Burge, Michael Burgess, Christopher Campbell, Karl Fetter, 2006 ). Peter Fritsch, Walter Judd, Kathy Kron, Ann Powell, Alan Whittemore, The affi nities of temperate fl oras have been considered fur- and Marty Wojciechowski for assistance with particular clades. P.S.M. ther by assessing patterns of similarity in the current distribu- acknowledges the Highlands Biological Station for continued support for teaching the fl ora of the Blue Ridge. This paper is dedicated to the memory tion of genera around the northern hemisphere ( Wood, 1970 , and achievements of E. Lucy Braun, Lewis Anderson, and Bob Zahner. 1972 ; Qian, 2002b ). For the three most relevant areas, eastern 2 Author for correspondence (e-mail: [email protected]) North America (ENA), western North America (WNA), and eastern Asian (EA), Qian (2002b) surveyed over 3500 genera doi:10.3732/ajb.1400530 and determined that the most similar fl oras are WNA and ENA, American Journal of Botany 102 ( 5 ): 780 – 804 , 2015 ; http://www.amjbot.org/ © 2015 Botanical Society of America 780 MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 781 and the most different are WNA and EA. These results do not Species and communities of the Southern Appalachian fl ora are support Gray’s (1859) initial hypothesis that the fl oras of ENA among the best studied in the world, resulting in a long chain of and EA are more similar. Potentially confounding such a simi- research in ecology, fl oristics, demography, phylogeography, larity-based study is the broad defi nition of the study areas and and natural history ( Braun, 1950 , 1955 ; Whittaker, 1956 ; Little, limited use of phylogenetic and temporal information within 1970 ; Wood, 1970 ; Martin et al., 1993 ; Estill and Cruzan, 2001 ; and among individual clades. Simon et al., 2005 ; Weakley, 2005 , 2012 ; Soltis et al., 2006 ; During the last 20 years, the rise of molecular phylogenetics Clark, 2010 ; Spira, 2011 ). Sparked by the pattern of temperate and the integration of fossil data have generated a wealth of forest disjunction across the northern hemisphere ( Li, 1952 ; information on the relationships of the plants of ENA, particu- Graham, 1972 ), many have investigated the evolutionary and larly of clades representative of the classic disjunction between ecological response of various isolated sets of species in com- species of ENA and EA ( Wen, 1999 ; Donoghue and Smith, munities ( White, 1983 ) as a function of species richness and 2004 ; Milne, 2006 ; Wen et al., 2010 ). Investigations into the molecular diversity ( Guo and Ricklefs, 2000 ; Qian, 2002a ; Xiang origins of northern hemisphere disjunctions also have included et al., 2004 ) and in comparisons of geographical range size paleoclimatic models in the context of the two most likely land ( Ricklefs and Latham, 1992 ). connections for plant migration—the Bering Land Bridge Southern Appalachian tree species also have been central to (BLB) and the North Atlantic Land Bridge (NALB). Diver- developing biogeographic hypotheses on the capacity for range gence time estimates have been used to predict whether clades expansion. For example, the fossil records for ENA species migrated via the BLB or NALB, providing a useful platform for suggest that rapid migration was typical for trees during the hypothesizing origins and directions of migration ( Donoghue Quaternary Period ( Davis, 1983 ; Delcourt and Delcourt, 1987 ; and Smith, 2004 ; Donoghue, 2008 ; Wen et al., 2010 ). Age esti- Jackson et al., 2000 ), whereas molecular phylogeographic pat- mates of the ancestral node linking intercontinentally disjunct terns support slower rates of postglacial range expansion, in- clades that are younger than 30 Myr have been viewed as com- cluding evidence for northern refuges ( McLachlan et al., 2005 ; patible with migration through the BLB, whereas older dates Soltis et al., 2006 ; Gugger et al., 2008 ; Tsai and Manos, 2010 ). imply NALB exchanges. In this paper, we take a phylogenetic approach to address the Wen et al. (2010) compiled the largest set of divergence time affi nities, patterns of connectivity, and diversifi cation of the estimates to date on the North America–Eastern Asia disjunc- Southern Appalachian woody fl ora. First, we use phylogenies tion and showed that many of the splitting times fall within the and species distributions across individual clades to describe Miocene, roughly 23 to 5 Ma. That study suggested that over the biogeographic patterns of area connections across phyloge- 50% of the sampled disjunctions achieved their current distri- netic scales. Second, we address the evidence for regional pat- butions via the BLB and support the hypothesis of an Asian ori- terns of diversifi cation by quantifying the taxon diversity gin for many temperate clades within North America ( Donoghue encompassed by the clades sampled. Last, we test two biogeo- et al., 2001 ). Overall, there is solid evidence for Miocene cli- graphic hypotheses: (1) that the main pattern of disjunction be- mate change as the primary infl uence on the pattern of intercon- tween ENA and EA in seed plant clades is explained by tinental vicariance observed across many seed plant