Biogeographic Analysis of the Woody Plants of the Southern Appalachians

Home , Buckleya, Catha (plant), Decumaria barbara, Dirca, Gaylussacia dumosa, Hudsonia

AMERICAN JOURNAL OF BOTANY RESEARCH ARTICLE

B IOGEOGRAPHIC ANALYSIS OF THE WOODY PLANTS OF THE SOUTHERN APPALACHIANS: IMPLICATIONS FOR THE ORIGINS OF A REGIONAL FLORA 1

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 analyze 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 paleofl 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 clades of pseudocongruent events and (2) that late Miocene climate the northern hemisphere. However, the range of splitting times change produced temporally congruent splits in clades with in- reported by Wen et al. (2010) suggests that multiple geophysi- tracontinental disjunctions from ENA to the areas of western cal events produced the same pattern of area relationships. The North America and Mesoamerica in the general area of eastern phenomenon of pseudocongruence ( Cunningham and Collins, Mexico. 1994 ), in which shared patterns of disjunction have been caused by temporally distinct historical events, was fi rst introduced in the context of north hemisphere plant biogeography by Xiang et al. (2000) . It continues to reinforce the critical role of temporal MATERIALS AND METHODS data in any attempt to generate a biogeographical synthesis of clades that share similar biogeographic patterns (Donoghue and Area of study— Our study area corresponds to the southern Blue Ridge Smith, 2004 ; Sanmartín and Ronquist, 2004 ; Donoghue, 2008 ; within the broader Level II Ecoregion called the Ozark Ouachita-Appalachian Forest. It is the most fl oristically homogeneous of the subregions within the Wen and Ickert-Bond, 2009 ; Kadereit and Baldwin, 2012 ). eastern temperate forest because of its higher elevation and level of endemism Intracontinental disjunct patterns of species between the area (A. S. Weakley, UNC Botanical Garden, personal communication). This area pairs, ENA-Mesoamerica and ENA-WNA, also have been de- corresponds to the southeastern unglaciated region of the United States defi ned scribed using fl oristic data (Miranda and Sharp, 1950; Braun, by the Blue Ridge Level III Ecoregion 66, ( US EPA, 2013 ), specifi cally the 1955 ; Wood, 1970 ), but the temporal context of these splits southern half, which lies south of the Roanoke River where the region has a across the temperate fl ora of North America remains unre- North-South disjunction. The area we sampled consists of level IV Ecoregions solved. Microfossil data suggest a late Miocene incursion of 66c through 66m ( Fig. 1A ). temperate elements into Mesoamerica (Graham, 1999b), and Sampling —A county-level query roughly approximating the shape of limited molecular data support the notion that recent climate Ecoregion 66c through 66m was carried out using the USDA Plants database change has produced vicariant sister species (e.g., Ruiz-San chez (USDA, NRCS, 2013). A woody plant species list was generated for the set of and Ornelas, 2014 ). For assessing the relative timing of con- counties in states (GA, TN, SC, NC, and VA) where species of the Southern nectivity among North American refugial areas, surveys of the Appalachian fl ora are known to occur (Appendix S1, see Supplemental Data available molecular data for both area patterns are needed with the online version of this article). The query was set for woody plants, in- across temperate clades. cluding trees, treelets, shrubs, subshrubs, and vines because they are generally better studied phylogenetically. Herbaceous annuals and perennials and all non- We specifi cally address the origin and relationships of tem- native species were excluded. Additional woody species that are typically over- perate woody plants in the montane areas of ENA by using its looked in surveys were added (e.g., Hypericum spp.), as well as several most diverse component—the Southern Appalachian fl ora—as a rare-occurrence species (e.g., Taxus canadensis) known from distributional test case to analyze larger patterns of biogeographic connectivity. data within the study area ( Weakley, 2012 ). 782 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

Fig. 1. (A) Area sampled for woody species of the Southern Appalachians. Dark gray areas show the southern portion of the Blue Ridge level III ecoregion (EPA ecoregion 66), consisting of level IV ecoregions 66c-66m. Counties that were queried to generate the species list are outlined in lighter gray. Bolded black lines show state borders. (B) Major areas of endemism for temperate ﬂ ora in the northern hemisphere. The study area is indicated with a box. Area abbreviations: eastern North America (ENA), eastern Mexico (EMEX), western North America (WNA), Europe (EU), and eastern Asia (EA). The maps were produced using ArcMap 10.2 ( Esri, 2013 ).

Phylogenies —We surveyed the phylogenetic literature using criteria that accommodate clades with widespread distributions from southern Mexico, included adequate sampling and at least moderate level of resolution such that south to Panama. SAMER and CARIB were broadly defi ned. each Southern Appalachian focal species, i.e., those species retrieved in our For the temperate Old World areas of the northern hemisphere, we also rec- query, could be placed within a focal clade (see Table 1, Fig. 2; Appendix ognized broad distributions. Sister taxa scored for EU included the geographic S1). Exceptions were made for large genera as long as the focal species are area at the eastern border of Europe and western Asia, corresponding to the placed within a reasonably sampled subclade. In rare cases where no phylog- well-established refuge known as the Caucasus Mountains, but also south to eny was available and the genus has a single species in the study area (e.g., cover taxa occurring in the Mediterranean vegetation zone. The area of EA in- Pyruleria), we used existing taxonomies and assumed the monophyly of the cludes Japan and other biogeographic infra-areas within continental Asia (see genus to identify the sister-area relationship. Taxa lacking phylogenetic data Wu, 1979 ). and monotypic genera were not part of the analysis (see Appendix S1 for the complete list). Clade defi nition and scoring biogeographic data— We derived the biogeographic data using a clade-based approach at the level of genus or below. Area defi nitions — The distributions of related species outside of the study Focal species placed within focal clades were classifi ed into three main biogeo- area were obtained by consulting original studies, fl oras, and other botanical graphic types: (1) North American clades that have a sister group in Eurasia, (2) sources. We followed the treatment of well-established areas of endemism used North American endemic clades, and (3) widespread clades combining a mix- in previous studies of northern hemisphere phytogeography (e.g., Donoghue ture of North American and Eurasian species. and Smith, 2004 ). We also treated several broadly construed areas throughout We recognized the separate biogeographic histories of focal species as much the analysis (Fig. 1B), including eastern North America (ENA), western North as possible and addressed the nonindependence of related species (Fig. 2). If America (WNA), Europe (EU), eastern Asia (EA), and eastern Mexico (EMEX), focal species are sister taxa or form groups of closely related species, they were along with several less frequent distribution areas including South America treated as a single clade, and ENA was scored once in the analysis (e.g., Aescu- (SAMER), Central America (CAMER), the Caribbean (CARIB), and circum- lus , Acer 3, and Carya ). If focal species are resolved in separate clades, we boreal (CIRCUM). recognized multiple clades, and treated them separately (e.g., Viburnum , with For North America, ENA was defi ned as the region roughly east of the Mis- clades 1–4). If the relationships of multiple focal species are unresolved with sissippi River, including the Bermuda Islands, north into Canada, and south to respect to related species outside the study area, we recorded the sister distribu- eastern Texas and northern Florida. The Southern Appalachian woody fl ora is tion to refl ect the combined areas of all species within the polytomy. treated as a subset of this broader area. WNA refl ects distributions that included Area patterns were recorded by overlaying the established area distributions the U.S. Pacifi c Northwest and U.S. Southwest, sometimes extending into west- of the species within clades onto published phylogenetic trees to generate area ern Mexico. EMEX delimits a more restricted area of endemism as part of the cladograms ( Fig. 2 ). A primary analysis was performed to identify sister taxa Sierra Madre Oriental of Mexico roughly east in the highlands southeast of and sister areas. Sister areas were recorded as single or multiple areas depend- Monterrey and Mexico City in the states of Tamaulipas, Hildago, and Vera ing on the distribution of sister taxa and sister clades. For those clades with Cruz, often extending farther south to the mountains of Guatemala. This area immediate connections to Eurasian sister taxa, the biogeographic analysis was has long been identifi ed as a refuge for temperate woody plants ( Miranda and complete and ready for synthesis. If focal species were unresolved within a Sharp, 1950; Braun, 1955; Wood, 1970). The area of CAMER was used to clade that included North American and Eurasian taxa, we also considered the MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 783 and Thomas, 2008 Donoghue et al., 2001 Fritsch et al., 2001 Cuenoud ; et al., 2009 et al., 2000 ) analysis (see Fig. 3A ), New ° times References (CARIB), and Circumboreal (CIRCUM); The (CARIB), and Circumboreal (CIRCUM); OW divergence divergence OW 25.182009 , Lo et al., 2007 61 (55.81, 71.51)Harris et al., 2009 Lo and Donoghue, 2012 Xiang et al., 2006; 10.97 (5.23, 17.6)Mao et al., 2010 Li, 2008 14 (8, 20)Renner et al., 2008 Li, 17 (11, 26) 39.14 (24.09, 53.02)Lang et al., 2007 Selbach-Schnadelbach 5.19 (4.69, 5.69) ; Kim and Kim, 1999 4.3 (3.9, 4.7) Schnabel et al., 2003 7.1 (3.1, 13.6)Xie et al., 2010 ; Fritsch and Lucas, 2000 Bush et al., 2010 EU EA EU EA EA EA EA EA EA EA and Old World (OW) deeper area connections scored in the (OW) World and Old EU,EA EU,EA EU,EA EU,EA EU,EA WNA,EA WNA,EA a abbreviations: Eastern North America (ENA), Eastern Mexico Eastern Mexico America (ENA), Eastern North a abbreviations: es, sister areas scored in the primary (1 (40.28, 55.90) ENA ENA ENA ENA ENA ENA ENA ENAENA ||WNA 49.04 ENA ENA ENA ENA ENA ENA

L., L.

L. Walter,

L. L. Beadle,

(L.) Mill. (Ehrh.)

Mill., C. intricata

C. pruinosa Mill.

L. (L.) Medik., Buckley L. Schrad. Michx.,

A. sylvatica W. A. sylvatica Sol., Jacq., Gatt., (Warder) A. saccharum C. viridis Willd., C. unifl ora C. unifl Wieg. species Primary sister areas NW areas times NW divergence areas OW C. crus-galli L., Aiton, C. macrosperma A. nigrum C. phaenopyrum A. rubrum Southern Appalachian Lange, Ashe, (L. f.) Medik., K. Koch, Wendl.) (H. L. C. punctata C. spathulata C. succulenta Link, ex Münchh., C. fl ava C. fl L., (Marshall) Borkh., Castanea pumila C. speciosa Engelm. & Savol. (L.) M. Powell, S. Andrews (Steud.) Sleumer C. calpodendron Medik., Small, Michx., F. Marshall D. rivularis D. sessilifolia A. glabra A. pavia Bartram. Michx.) Fernald, (F. A. canadensis A. laevis Acer saccharinum Castanea dentata Catalpa bignonioides Cornus amomum Halesia carolina mucronata Ilex Leucothoe fontanesiana Crataegus boyntonii Crataegus Gleditsia triacanthos virginiana Juniperus Acer leucoderme Diervilla lonicera Aesculus fl ava Aesculus fl arborea Amelanchier Hamamelis virginiana 3 2 2 2 2 2 2 2 6 4 3 5 3 50 11 1

2 1. (TD) within focal clad American taxon diversity Columns: North clades and focal species sampled in the area of study. Focal World (NW) areas scored as intermediate North American disjunctions (||) and deeper area connections in the secondary analysis, (NW) areas scored as intermediate North World Are ). Fig. 3B and underlined area patterns formed the basis for secondary analysis (see All boldfaced secondary analysis. America (CAMER), Caribbean America (SAMER), Central Asia (EA), South Europe (EU), Eastern America (WNA), North Western (EMEX), groupings. solid lines distinguish between the main sister-area 3 1 2 ABLE Diervilleae Acer Aesculus Amelanchier Castanea Catalpa Cornus Halesia Ilex Leucothoe Crataegus Gleditsia Hamamelis Juniperus Acer T Clade TD 784 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY et al., 2009 ; Cuenoud ; et al., 2009 et al., 2000 and A. S. Weakley, and A. S. Weakley, unpublished data 2005 ; Moore and ; Donoghue, 2007 Clement and Donoghue, Clement et al., 2014 ; 2011 Li et al., 2007 Yi et al., 2004 , 2007 , et al., 2004 Yi Gillespie and Kron, 2010 Weakley, Weakley, 2012 Bush et al., 2010 Weakley, 2012 Floyd, 2002 Arrington, 2004 Selbach-Schnadelbach times References OW divergence divergence OW 12.44 36 (28, 44) Lo and Donoghue, 2012 27.12Lo and Donoghue, 2012 14.15 (8.69, 21.01); K. Fetter Nie et al., 2008 26 (15, 37) Xiang et al., 2011 16 (10, 22.5) and Donoghue, Winkworth Li et al., 2002 14 (8.4, 20.6) 35.8Lo and Donoghue, 2012 ; Jarvinen et al., 2004 Samain et al., 2010 ? ? EA EA EA EA EA EA EU,EA EU,EA EU,EA EU,EA WNA, SAMER CAMER,

ENA ENA ENA ENAENA ENA ||WNA ||EMEX, 10 3 (1.5, 8) ENA ENA ||WNA ENA ENA ENA ||WNA 3 (1.5, 8) ENA ENA ENA ENA ENA ENA ENA ENA ENA ENA L.

L., L.

L.,

L., L., (L.) (L.)

Nutt. (Aiton) (Michx.) (Michx.) Marshall (L.) Pers., Lodd. (L.) Mill. R. typhina Fernald (A. Gray) Marshall Raf. (Cav.) Weath. Small Small, Walter Torr. & L., (M.A. Curtis) L., Alexander, species Primary sister areas NW areas times NW divergence areas OW B. lenta L. A. triloba A. prunifolia E. recurva S. ovata Southern Appalachian V. nudum V. prunifolium L., V. dulum rufi V. Nutt., Britton (Buckley) A. Gray Britton, I.montana L., recognitum V. A. melanocarpa Elliot, (Marshall) Rehder Dunal, Dunal Michx. H. cinerea H. radiata (Wangenh.) K. Koch, K. Koch, (Wangenh.) G. tomentosa Small, Pursh ex G. ursina A. Gray & Torr. Schult. K. carolina Viburnum cassinoides Viburnum Sorbus americana Sorbus Eubotrys racemosa major Fothergilla Hudsonia montana collina Ilex Betula alleghaniensis Malus coronaria Nyssa sylvatica malacodendron Stewartia dentatum Viburnum arbutifolia Aronia ora Asimina parvifl Malus angustifolia Hydrangea arborescens Hydrangea Liriodendron tulipifera Liriodendron Gaylussacia baccata Viburnum rafi nesquianum rafi Viburnum Rhus glabra Kalmia angustifolia 7 4 2 2 4 3 3 3 2 4 2 4 3 2 4 33 10

4 2a 2b

1 2 1 1. Continued.

2b 1 ABLE Viburnum Sorbus Eubotrys Fothergilla Gaylussacia Hudsonia Ilex Betula Malus Nyssa Rhus Stewartia Viburnum Viburnum Aronia Asimina Malus Hydrangea Kalmia Liriodendron Clade TD T MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 785 2012 Parks et al., 2012 Parks Denk and Grimm, 2009 and Thomas, 2008 Meireles J. E. unpublished Milne, 2004 data; ; Morris et al., 2008 Ruiz-Sanchez and Ornelas, 2014 Meireles, unpublished Milne, 2004 data; Gillespie and Kron, 2010 Wiegrefe et al., 1994 Wiegrefe Poon et al., 2007 ; Weakley, Weakley, ; Poon et al., 2007 times References OW divergence divergence OW 30 (fossil)2007 , et al., 2004 Yi Simmons, 2002 Simmons, Wen, 2011 Wen, 62-71 45-60 (fossil)Feng et al., 2005 ; and Hall, 2006 Eckert 2012 15 (10, 21) et al., 2005 Forest ; Li, 2008 7.14 (5.3, 8.94); Goetsch et al., 2005 McCarthy, Li, 2008 ; 2002 Wen, and Yoo 13 (6.9, 19) Fritsch and Cruz, 2012 31 30.2 (22.2, 38.43) ; Manos and Stanford, 2001 33.8 (30.7, 36.9)2007 , et al., 2004 Yi 2001 2010 16 26.9Xiang Zhou et al., 2006 ; Xiang et al., 2006 Fritsch, 2012 Nepal and Ferguson, Bell, 2005 Oh and Potter, 10 (6, 14)Smith and Donoghue, 2010 22.89 (12.65, 33.13); 2006 Wen, and Ickert-Bond Herbert, 2005 Herbert, ; J. E. Goetsch et al., 2005 ? ? EU EA EA EA EA EA EA EA EU EA EA EU EU,EA EU,EA EU,EA EU,EA EU,EA WNA,EA WNA,EA WNA,EU WNA,EA WNA, WNA, EA ENA,EU,EA ENA,EU,EA 10Lo and Donoghue, 2012 ENA ENA WNA 11.02 (7.02, 15.02) WNA WNAWNA ||CAMERWNA 16.4 4.62 (1.12, 8.12) WNA WNA 9.65 (4.92, 14.79) WNA WNA WNA 6 (3.8, 8.2) WNA WNA WNA EMEX EMEXEMEX 5 EMEX ||WNA 4.5 (1.6, 8.2); 10 (5.3, 16) EMEX 3.3 (2, 4.8) EMEX 7 (5.3, 8.5) EMEX 8 EMEX 8Nie et al., 2008 EMEX,WNA EMEX,WNA EMEX,WNA 4 EMEX,WNA EMEX, CAMER 8.34 (3.14, 15.97) EMEX,WNA,CAMER

Michx. L. L. Walter

L. Lam., L. Ehrh. (Mill.) L. Aiton L. L. Wangenh. L. L. L. Aiton Sarg., L. Sarg. L. L. species Primary sister areas NW areas times NW divergence areas OW Southern Appalachian (Michx.) Sweet K. Koch Michx. S. grandifolius (Mill.) Small (L.) Maxim. Moench. U. serotina U. (Pursh) DC. Rhododendron canescens Rhododendron Cercis canadensis Cercis Ptelea trifoliata Kalmia polifolia Rhus copallinum racemosa Aralia oridus Calycanthus fl Platanus occidentalis Fagus grandifolia Fagus Pinus strobus Staphylea trifolia orida Cornus fl Ostrya virginiana catawbiense Rhododendron americanus Styrax Carpinus caroliniana ua Liquidambar styracifl Morus rubra Rhus aromatica Lonicera dioica Lonicera Morella caroliniensis Morella Physocarpus opulifolius Symphoricarpos orbiculatus americana Tilia Magnolia macrophylla Magnolia Magnolia grandifl ora grandifl Magnolia Ulmus thomasii Amelanchier sanguinea Amelanchier 2 4 2 2 2 4 2 5 3 2 2 2 3 5 2 2 3 9 5 3 2 9 3 11 10 10

2 3

2b 3a

2 3

1. Continued.

1 2a 1

ABLE Rhododendron Cercis Ptelea Kalmia Rhus Aralia Calycanthus Platanus Fagus Pinus Staphylea Cornus Ostrya Physocarpus Rhododendron Styrax Symphoricarpos Tilia Carpinus Liquidambar Magnolia Morus Rhus Ulmus Lonicera T Clade TD Morella Amelanchier Magnolia 786 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY et al., 2011 unpublished data; Sattarian, 2006 ; Bortiri et al., 2001 et al., 2008 Wen 2004 Pearse and Hipp, 2009 and Thomas, 2008 times References OW divergence divergence OW 33.8Xiang ; Xiang et al., 2006 A. T. Whittemore, T. A. Chen et al., 2010 ; 17 (10, 24) and Small, 2005 Shaw ; Manos and Stanford, 2001 Schultheis and Donoghue, 21.58 (11.07, 35.51)Zhang et al., 2013 21.64 (10.23, 34.89)Nie et al., 2010 Nie et al., 2008 Ren ; et al., 2010 Trondle 30.4 +/− 1.8 Hinsinger et al., 2013 13.46 (7.95, 19.42) Nie et al., 2009 6Xiang et al., 2009 Smith and Sytsma, 1990 EA EA EA EA EA EA EA EA EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA ENA,EU,EA ENA,EU,EA WNA, CAMER ENA ||EMEX, ENA,WNA ENA,WNA ENA,WNA ENA,WNA ENA,EMEX ENA,EMEX ENA,EMEX 30 ENA,EMEX ENA,EMEX ENA,EMEXENA,EMEX 17-20 (fossil) ENA,EMEX,WNA ENA,EMEX,WNA ENA,EMEX, WNA ENA,EMEX,

L.,

(Mill.)

L.,

(L.) Michx. Michx., Grauer, L., Q. bicolor (Ashe) (Wangenh.) Walter Ehrh., L., Willd., Lam. Nutt. (L.) Mill. (F. Michx.) (F. Michx., V. labrusca V. a L. L., (Pursh) Poir., (Pursh) Poir., F. profunda F. (Ashe) Engl. & L. (Engelm. Q. prinoides C. glabra C. tomentosa C. ovalis species Primary sister areas NW areas times NW divergence areas OW Marshall (Mill.) K. Koch, (Mill.) K. Koch, Q. stellata Q. michauxii Q. muehlenbergii C. illinoinensis Q. montana T. radicans T. Southern Appalachian Willd., Willd., Wangenh. (Wangenh.) Sarg., Sarg., (Wangenh.) C. ovata C. pallida Graebn., (Lam) Nutt. Willd., Willd., Nutt., Willd., Engelm., (Wangenh.) K. Koch, K. Koch, (Wangenh.) C. laciniosa G. Don, C. racemosa Mill, A. balsamea Marshall Bartram ex virginian P. Kuntze C. occidentalis C. tenuifolia pennsylvanica F. Marshall, (Bush) Bush R. glandulosum R. rotundifolium septentrionalis Engl. & Graebn., C. cordiformis K. Koch, Sweet, V. vulpina V. (L.) Planch. V. cinerea V. A. Gray) Engelm. ex Millard, ex Cornus asperifolia Salix nigra pubescens Toxicodendron Abies fraseri W. deltoides Populus Prunus serotina Celtis laevigata americana Fraxinus Ribes cynosbati alba Quercus Carya. carolinae- Magnolia fraseri Magnolia Vitis aestivalis Vitis Magnolia virginiana Magnolia Parthenocissus quinquefolia Parthenocissus 6 3 8 4 3 2 6 4 3 10 10 12 15 130

3b 2a 1

4 3

1. Continued.

1 ABLE Cornus Magnolia Salix Vitis Toxicodendron Abies Populus Prunus Parthenocissus Celtis 1 Fraxinus Ribes Quercus T Clade TD Carya Magnolia MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 787 Peterson, 2013 Cuenoud ; et al., 2009 et al., 2000 1997 ; Moore and Donoghue, 2007 2007 et al., 2012 Bortiri et al., 2001 ; Bortiri et al., 2001 et al., 2008 Wen et al., 2011 et al., 2009 Smith and Donoghue, 2010 Eriksson and Donoghue, Gillespie and Kron, 2010 ; and Small, 2005 Shaw Ren ; et al., 2010 Trondle Simmons et al., 2008 times References OW divergence divergence OW 90-130 Parks ; and Hall, 2006 Eckert Appelhans et al., 2014 ? EU,EA EU,EA Selbach-Schnadelbach

Floyd, 2002 Floyd,

Stone ; Manos et al., 2007 et al., 2011 Burge

; 2004 Schrader and Graves, et al., 1994 Wiegrefe ENA, WNA WNA, WNA, EMEX CARIB EMEX, CARIB, SAMER SAMER SAMER CAMER CAMER, CAMER, CAMER, Smedmark and Anderberg, Anderberg, and Smedmark Nurk et al., 2013

ENA ENA ENA ||EMEX, EMEX EMEX EMEX CARIB SAMER ENA,EMEX ENA,EMEX EMEX,WNA ENA,CARIB ENA,CARIB EMEX,WNA EMEX,WNA WNA,CAMER WNA,CAMER ENA,EMEX,WNA ENA,EMEX,CARIB,CAMER ENA,EMEX,CARIB,CAMER ENA,EMEX,CARIB,CAMER

L.,

L. L. L. L.

(L.) Marshall, P. taeda P. Muscadinia W. Wight & Wight W. A.Gray Marshall,

L. L., W.P. Mill. L. Mill., H. lloydii L.H. Bailey, L.H. Bailey, L. Lamb., Michx. Mill., species Primary sister areas NW areas times NW divergence areas OW H. hypericoides Aiton H. densifl orum H. densifl H. frondosum Southern Appalachian (Andrews) (Andrews) A. Gray & Torr. (Planchon) Small) Michx. Small ( Michx., (L.) Crantz, Adams, P. W. (Svenson) cum H. prolifi H. stragulum Adams & N. Robson Mill. M. A. Curtis, H. crux-andreae Crantz, Pursh, P. pungens P. rigida P. P. virginiana P. P. angustifolia P. hortulana P. munsoniana P. Hedrick Lonicera sempervirens Lonicera Ceanothus americanus Pinus echinata Ulmus alata Gaylussacia dumosa Ilex opaca Ilex Dirca palustris Dirca Kalmia latifolia Vitis rotundifolia Vitis Zanthoxylum americanum Paxistima canbyi Paxistima canadensis Sambucus Hypericum buckleyi Juglans nigra Juglans Sideroxylon lycioides Sideroxylon Prunus americana 6 3 8 2 2 2 4 55 44 45 11 50 16 25 17

) Bumelia

2 (

1b 2 1a 1 2a 1. Continued.

2 2 3 ABLE Lonicera Ceanothus Gaylussacia Pinus Ulmus Ilex Dirca Sideroxylon Kalmia Vitis Hypericum (Myriandra) Clade Zanthoxylum TD Paxistima Sambucus Juglans Prunus T 788 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY lez et al., 2014 á and Paule, 2010 and Paule, š data et al., 2001 et al., 2007 2008 Meireles, unpublished Milne, 2004 data; 2004 1997 ; Moore and Donoghue, 2007 Gonz Bruneau et al., 2007 Whitcher and Wen, 2001 Wen, Whitcher and Chen and Li, 2004 Wen, 2011 Jarvinen et al., 2004 ; Li ; Jarvinen et al., 2004 Li et al., 2001 ; Hao et al., ; Li et al., 2001 Chen and Li, 2004 Gillespie and Kron, 2010 Bolmgren and Oxelman, Bolmgren and Oxelman, ; and Ritz, 2005 Wissemann times References OW divergence divergence OW Herbert, 2005 Herbert, 6.5Mao et al., 2010 Eriksson and Donoghue, CIRCUM CIRCUM CIRCUM

S. Manos, unpublished P.

et al., 2003 Lavin WNA WNA, EMEX, SAMER CAMER,

3 Dobe 3

11.5 (8.95, 13.79); J. E. Goetsch et al., 2005

9.67 (4, 16), et al., 2006 Ohi-Toma 25.3 (21.3, 28.9)Li et al., 2009 31.43 (22.97, 39.89) et al., 2008 Havill 9 (5, 13) 25 (15, 36) Renner et al., 2008 9.64 (7.85, 11.44)Donoghue ; Li et al., 2001 3.62 (1.52, 5.72)Xiang et al., 2000 EA EA EA EA EA EA EA EA ENA ENA WNA WNA ENA,EA ENA,EA ENA,EA ENA,EA ENA, ENA, EA WNA,EA WNA,EA CIRCUM ENA,EU,EA ENA,EU,EA WNA,EU,EA WNA,EU,EA ENA,WNA,EA ENA,WNA,EA EU,EA,CIRCUM EU,EA,CIRCUM ENA,WNA,EU,EA ENA,WNA,EU,EA WNA,EMEX,EU,EA WNA,EMEX,EU,EA

L., L.

Lam.

L. L. L. (Aiton) L., Marshall Raf.) (L.) Münchh., (L.) Münchh., Marshall L. (Pursh) (Bergius) Marshall Michx., L., k (Aiton) Q. shumardii Q. phellos L. á Lam.

L., (Vill.) DC. (Vill.) Vent. Michx., L. Q. velutina L., species Primary sister areas NW areas times NW divergence areas OW L., Isotrema Southern Appalachian R. pseudoacacia R. viscosa Buckley, Buckley, Q. rubra (Walter) A.Gray (Walter) Benth. & Hook. f. Willd. Beureau Seem. ex Q. falcata Q. imbricaria Q. marilandica Q. nigra Lam. ( R. palustris Marshall Carrière (Nutt.) Torr. Gift, Kron & P.F. Stevens Gift, Kron & P.F. & Soj Paule Alnus viridis Robinia hispida Frangula caroliniana Frangula oribunda Pieris ﬂ Acer spicatum Alnus serrulata nudicaulis Aralia spinosa Aralia Campsis radicans Quercus coccinea Quercus Corylus cornuta Aristolochia macrophylla Aristolochia Betula populifolia communis Juniperus canadensis Taxus Rosa carolina Rhododendron maximum Rhododendron Acer pensylvanicum Myrica gale racemosa Sambucus Tsuga canadensis Buckleya distichophylla Buckleya Kalmia buxifolia Sibbaldia tridentata 1 8 2 1 1 1 1 1 1 2 1 1 1 2 1 1 1 2 2 1 1 1 120 4

2 1

2 1b

2 1 3 1a 1. Continued.

2 1 1 4 2 ABLE Alnus Robinia Frangula Pieris Acer Alnus Aralia Aralia Buckleya Campsis T Clade TD Quercus Corylus Aristolochia Betula Juniperus Taxus Rosa Rhododendron Acer Kalmia Myrica Sambucus Sibbaldia Tsuga MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 789 unpublished data and Thomas, 2008 Cuenoud ; et al., 2009 et al., 2000 et al., 2009 Judd, unpublished data Meireles, unpublished Milne, 2004 data; et al., 2015 2005 ; Moore and Clement ; Donoghue, 2007 ; and Donoghue, 2011 Clement et al., 2014 and X. Lou, unpublished data Mu et al., 2012 M. Wojciechowski, Fior et al., 2003 Selbach-Schnadelbach S. W. ; Kron and Judd, 1997 Jacques and Bertolino, 2008 Guo et al., 2013 Weakley, 2012 Liu et al., 2006 Kron et al., 2002 J. Jiang, Li, C. Fu, M. times References OW divergence divergence OW 10.57(5.98, 15.91) 10.57(5.98, 8.4) 2.38(0.69, 4.07) 21.7 18.2(17.5, 18.8)Xiang et al., 2000 Schnabel et al., 2003 Xiang ; Xiang et al., 2006 7.2(5.3, 9.1) Stone ; Manos et al., 2007 28.3(20.58, 36.55)Nie et al., 2008 7.73(3.14, 12.53) 5(1.4, Huang et al., 2013 14.6(9.1, 20.1)Jiao and Li, 2009 18.2(14.5, 22.68); J. E. Goetsch et al., 2005 13.8(11.51, 16.09)Nie et al., 2007 31.5(27, 36) 20.3(15.5, 25.1)Fritsch ; et al., 2004 Wang 7.53(2.76, 12.86) Li and Xiang, 2005 Nie et al., 2009 23.03(14.95, 31.11) et al., 2008 Havill 6(3, 9) and Donoghue, Winkworth EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA EA

L. Nutt.

Michx. (L.) L. L.

L. L. f.

(L.) L. Michx. L. (Bickell)

Engelm. Michx. (L.) DC. L. L.

species Primary sister areas NW areas times NW divergence areas OW Southern Appalachian (Dum.Cours.) Rudd (L.) A. Gray Michx. Loisel. (Nutt.) Nees Rehder. (L.) Kuntze Michx. Michx. (L.) Poir. (L.) K. Koch (L.) L'Hér. Cladrastis kentukea Cladrastis acuminata Clethra Cornus alternifolia verticillata Ilex cinerea Juglans ligustrina Lyonia acuminata Magnolia tripetala Magnolia repens Mitchella procumbens Pachysandra Philadelphus hirsutus Philadelphus pubescens Pyrularia pubera minus Rhododendron albidum Sassafras glabra Schisandra vernix Toxicodendron Tsuga caroliniana erythrocarpum Vaccinium acerifolium Viburnum lantanoides Viburnum frutescens Wisteria Celastrus scandens Decumaria barbara Gymnocladus dioicus Menispermum canadense Symplocos tinctoria Thuja occidentalis 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 5

1 2

1 3

1 4

1b 1. Continued.

4 ABLE Cladrastis Clethra Cornus Decumaria Gymnocladus Ilex Juglans Lyonia Magnolia Magnolia Menispermum Mitchella Pachysandra Philadelphus Philadelphus Pyrularia Rhododendron Sassafras Schisandra Symplocos Thuja Toxicodendron Tsuga Vaccinium Viburnum Viburnum Wisteria T Clade TD Celastrus 790 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

analysis complete. We calculated the percentages of sister areas based on the raw area patterns and the relative contribution of each major component area across patterns (see below). A secondary analysis scored the new data for the clades with deeper area connections, most often to Eurasia ( Fig. 2 ). These updated area data were combined with the completed biogeographic results in the primary analysis. For the clades endemic to North America, we also scored additional area data for connections at the deepest node in the area cladogram. In a few cases, additional disjunctions within North America were recorded at an intermediate phyloge-

Bortiri et al., 2001 ; Wen Wen ; Bortiri et al., 2001 et al., 2008 Meireles, unpublished Milne, 2004 data; Bruneau et al., 2007 netic depth, and these data points were added to the calculation of the map of Wiegrefe et al., 1994 Wiegrefe Kim et al., 2004 2001 Wen, Whitcher and Duangjai et al., 2009 Ran et al., 2006 ; and Small, 2005 Shaw ; and Ritz, 2005 Wissemann Oh et al., 2010 et al., 1994 Wiegrefe Kron et al., 2002 component areas derived from the secondary analysis (see below). We then recalculated both the percentages of combined area patterns and the relative contribution of each component area across patterns. This approach is designed to reﬂ ect the change in going from the results of the primary analysis (i.e., sister areas) to the results of the secondary analysis (synthesis).

times References To evaluate the phylogenetic structure of intercontinental disjunctions between ENA and Eurasia, we performed a permutation test using (1) a phylog-

OW divergence divergence OW eny of the North American clades in the study area (Appendix S2a, see online Supplemental Data) derived from the APG tree ( APG III, 2009 ) with a few modiﬁ cations based on the individual clade phylogenies, and (2) a binary vector indicating whether a clade has its sister in Eurasia (see Table 1 ). The permutation test was performed by contrasting the observed phylogenetic diversity (PD; sensu Faith, 1992 ) of the clades with direct Eurasian disjunctions to a null distribution of PD. We estimated the null distribution of PD by shufﬂ ing the binary indicator vector 999 times and computing the PD without changing the structure of the phylogenetic tree. The test was conducted using both an ultra- metric tree and a unit branch length tree (all branch lengths set equals to one). All computations were performed in R ( R Core Team, 2013 ) using the pd function in the package picante ( Kembel et al., 2010 ).

Taxon diversity within clades — We quantifi ed the number of taxa in each focal clade using both phylogenetic and fl oristic literature. For most clades, this was the number of North American species in the clade or the number of taxa that have accumulated since the disjunction to the OW (see Fig. 2). For the American endemic clades, it is the number of taxa in the genus or subclade scored in the analysis. Whenever possible, we recognized the taxon diversity within the subclade being analyzed to accurately identify North American diversifi cation patterns within several complex clades (e.g., Hypericum , Myrian- dra clade; and Sideroxylon , Bumelia clade).

Divergence times — We recorded estimates of divergence times within clades as they were presented in the literature. The most commonly estimated 18.0 +/− 1.3Hinsinger et al., 2013 value was the splitting time between 8.05(5.05, 11.39); J. E. Goetsch et al., 2005 ENA and Old World species. When avail-

able, intracontinental disjunction times also were recorded. Node ages repre-

EU senting fossil constraints were plotted when appropriate. Point estimates and EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA EU,EA EU, EU, EA conﬁ dence intervals, when available, were compiled based on commonly used approaches, such as penalized likelihood using the program R8s (Sanderson, 2002) and Bayesian inference using the program BEAST (Drummond and Rambaut, 2007 ).

Data charts and area plots — We tabulated and visualized the complete set of biogeographic using pie charts and maps that represent the most relevant

patterns of area connections. The slices in the pie charts show the proportion of L. L. f. Mill.

Walter raw area patterns in the primary and secondary phase of analysis. Area compo- Britton L. (Michx.) nent maps were constructed by using the radius of each circle to visually repre- Marshall

Michx. sent a proportional count of each area of endemism within the raw area patterns. Sarg. Muhl.

species Primary sister areas NW areas times NW divergence areas OW For instance, a clade that has the combination of sister areas ENA-EMEX was given a score of one to count in the formation of each of the ENA and EMEX circles. The component map based on the secondary analysis combined sister- Southern Appalachian

calendulaceum Torr. Aiton area and deeper-area connections where appropriate, including the cases with Picea rubens Corylus americana nigra Fraxinus Prunus pensylvanica Rhododendron Rosa setigera virginiana Spiraea Ulmus rubra macrocarpon Vaccinium Ulmus americana Berberis canadensis virginiana Diospyros intermediate disjunctions within North America (see Fig. 2). Although some clades contribute more than one area in this procedure, plotting the data in this 2 1 1 1 1 1 1 1 1 1 1 1 manner is designed to show the relative importance of each area across all patterns.

1 RESULTS 2

2 3 1

1. Continued. The ﬁ nal species list of woody ﬂ ora includes 302 species.

2 Forty-eight taxa (16.6%) were excluded because they lack ad- ABLE Picea Berberis Corylus Diospyros Fraxinus Prunus Rhododendron Rosa Spiraea Ulmus Vaccinium T Clade TD Ulmus equate phylogenetic and biogeographic data (Appendix S1). MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 791

Fig. 2. Example area cladograms derived from the focal clades identiﬁ ed in the study. Southern Appalachian focal species of eastern North America are shown with circles. The node connecting areas scored in primary (1° ) and secondary (2° ) analyses are labeled. Nodes labeled 1° and 2 ° indicate that area connections identi- ﬁ ed in the primary analysis are complete (see Table 1 and Methods). (A) Area connections to the Old World, from left to right: Crateagus , Calycanthus , Acer 2, and Frangula . (B) Area connections to the Old World with intermediate disjunctions within North America (|| and gray branches), from left to right: Viburnum 2 (a and b) and Aesculus . (C) Area connections within North American endemic clades, from left to right: Vitis 2, Juglans 2, and Hypericum ( Myriandra s.s.). Area abbreviations: eastern North America (ENA), eastern Mexico (EMEX), western North America (WNA), Europe (EU), eastern Asia (EA), South America (SAMER), Central America (CAMER), and Caribbean (CARIB).

The remaining 252 species (Table 1) were delimited into 158 represented by three broad patterns: regional diversifi cations in clades distributed widely across the seed plant phylogeny (APG ENA, intracontinental disjunctions (e.g., from ENA to EMEX, III, 2009 ) in the following major groups and grades: Pinales WNA), and intercontinental disjunctions from ENA to Eurasia (10), magnoliid grade (12), eudicot grade (11), rosids (68), and (EA, EU). In summary, 62% of the clades have sister species asterids (57) (see Appendix S2a). and clades that occur within North America (shades of blue) In Table 1, we present blocks of clades that share biogeo- and 30.4% have sisters in Eurasia (shades of red). The remain- graphic patterns detected in the primary analysis of sister-area ing clades (7.6%) have sister species resolved in clades that connections. For example, all clades with a shared single-area combine northern hemisphere areas (gray = Eurasia–N. Ameri- sister area are grouped together and alphabetized. Similarly, can combinations). clades with related combinations of areas are grouped. These We found that 27.2% of Southern Appalachian species patterns form the basis for Figs. 3A and 7A , whereas the bolded have sister taxa that also occur in the single region of ENA. and underlined areas in Table 1 highlight the synthesis of re- This is a minimum estimate because we scored the sister sults in the secondary analysis. area—ENA—only once for cases with multiple ENA species, In Table 2, we present 18 clades with disjunctions between but poor phylogenetic resolution among them (e.g., Crataegus , ENA and WNA for the direction of movement between the two Hypericum , Quercus 1 & 2). Connections to sister taxa in the areas using the dispersal–vicariance analysis optimization pro- areas of EMEX and WNA accounted for 6.96% and 8.86% of tocol (DIVA; Ronquist, 1997 ). We based the analysis on either the intracontinental disjunctions, respectively. The remaining the area cladograms we generated, or by following published biogeographic patterns involve infrequent single areas (e.g., results on ancestral areas. For each clade, we also evaluated SAMER), combinations of areas in pairs (e.g., ENA, EMEX directionality for the deeper area relationships. and EMEX, WNA), and presence in three or more areas (see Fig. 3A ). Biogeographic patterns: Primary analysis of sister areas— For For the area of Eurasia (EU, EA), sister-area connections the 158 clades, we identifi ed 29 distinct biogeographic patterns, with EA were the most abundant pattern (22.78%), with the including 8 single areas and 21 unique patterns of various combined area of EU-EA accounting for another 6.98%, and area combinations ( Table 1 , Fig. 3A ). Phylogenetic resolution just a single clade was identifi ed with a direct sister-area con- largely determined the areas that were included in circumscrib- nection to EU alone (Ulmus 1). Noteworthy are the 12 North ing the distribution of sister taxa and clades. Sister areas are American clades (7.59%) with sister clades that combine the 792 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

Fig. 3. Pie charts based on single and combined area connections for the clades studied (see Table 1; see Materials and Methods and Results). (A) Percentage breakdown of sister-area connections based on the primary analysis for 158 clades. (B) Percentage breakdown of area relationships based on the secondary analysis for 150 clades. Area abbreviations: eastern North America (ENA), western North America (WNA), Europe (EU), eastern Asia (EA), eastern Mexico (EMEX), South America (SAMER), Central America (CAMER), and Circumboreal (CIRCUM). The pie charts were made with Excel 2010 (Microsoft, Redmond, Washington, USA). areas of Eurasia and North America (e.g., Alnus 2, Betula 2, disjunctions from ENA to Eurasia (see Methods; Kalmia 1b, Kal- Pieris , Sibbaldia , and Tsuga 1a). mia 2, Lonicera 1b, Magnolia 2b and 3b, Rhus 2b, Ulmus 2b, and Viburnum 2b). Biogeographic patterns: Secondary analysis— Our synthe- We used a combination of phylogenetic scales to synthesize sis of biogeographic patterns is based on 150 clades, including area patterns ( Fig. 3B , Table 1 ). The secondary analysis in- 23 clades endemic to North America (shades of blue; 15.3%, cludes results from the following biogeographic patterns: (1) excluding Aronia , and Pinus 2, see below), 122 clades (81.3%) North American clades with immediate connections to Eur- with biogeographic connections that include Eurasia (shades of asian sister taxa, (2) widespread clades combining a mixture of red and gray), and ﬁ ve additional clades (3.33%) with uncertain North American and Eurasian species, (3) North American en- deeper phylogenetic relationships (Fig. 3B, Table 1). The total demic clades with deeper area connections ( Fig. 3B , e.g., North number of clades in the secondary analysis was adjusted to 150 American combinations), and (4) North American clades con- because eight clades have nonindependent deeper intercontinental sisting of more than one species with deeper area connections

T ABLE 2. Clades in the area of study with sister taxa or related clades distributed in both eastern North America (ENA) and western North America (WNA). Clade names and relevant literature follow Table 1. Assessed direction of movement between areas is indicated with an arrow. Deeper areas patterns with equivocal results have no arrrows.

Clade name Related WNA taxa Directionality Assessment Deeper area pattern

Aesculus A. parryi A. Gray WNA → ENA from paper (EA ← (WNA-ENA)) Amelanchier 2 A. alnifolia M. Roem. WNA → ENA inferred from tree endemic Aralia 2 A. californica S. Wats. WNA → ENA from paper (EA→ (WNA-ENA)) Calycanthus C. occidentalis Hook. & Arn. equivocal from paper (EA (WNA-ENA)) Cercis C. occidentalis Torr. ex Gray equivocal from paper (EU (WNA-ENA)) Cornus 3C. nuttallii Audubon ex Torr. & A. Gray WNA → ENA from paper (EA→ (WNA-ENA, CAMER)) Crataegus Douglasianae clade WNA ← ENA from paper (ENA→ (WNA+EA)) Lonicera 1a L. arizonica Rehder equivocal from paper (EU (WNA-ENA)) Malus 1 M. fusca (Raf.) C.K. Schneid. WNA → ENA inferred from tree (EA→ (WNA-ENA)) Morus M. microphylla Buckley, M. celtidifolia Kunth equivocal inferred from tree (EA (WNA-ENA)) Ostrya O. knowltonii Coville, O. chisosensis Correll equivocal inferred from tree (EA→ (WNA-ENA)) Physocarpus a P. capitatus (Pursh) Kuntze, P. monogynus (Torr.) J.M. Coult. WNA → ENA from paper (EA (WNA-ENA)) Rhododendron 2R. occidentale (Torr. & A. Gray) A. Gray equivocal from paper ? Rhododendron 3R. macrophyllum G. Don equivocal inferred from tree (EA→ (WNA-ENA)) Rhus 1 R. trilobata Nutt. WNA → ENA from paper (EA← (WNA-ENA)) Robinia R. neomexicana A. Gray equivocal inferred from tree (ENA-WNA) Sorbus S. scopulina Greene WNA → ENA inferred from tree (EA (WNA-ENA)) Viburnum 2bV. ellipticum Hook. equivocal inferred from tree see Table 1 a The focal taxon Physocarpus opulifolius (L.) Maxim. is believed to be of hybrid origin involving either of the western North American species listed and an eastern Asian species. ( Oh & Potter, 2005 ) MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 793

to Eurasia. The secondary analysis also detected nine clades 2 and Aronia as part of the total number of species for endemic with intermediate disjunctions within North America. ( Table 1 , clades, although their deeper phylogenetic connections to Eur- Aesculus , Cercis , Cornus 3, Malus 1, Pinus 2, Quercus 1, Sor- asia are treated in the analysis of area patterns and divergence bus, and Viburnum (2a and 2b). times. Figure 4B shows the distribution of taxon diversity in the Five additional clades with likely Eurasian connections were 119 clades with disjunctions to the Old World. The percentage denoted by a question mark because of uncertain relationships of low-diversity clades is much higher than that across the total ( Betula 1, Hydrangea , Morella , Rhododendron 2, and Zanthox- number of clades ( Fig. 4A ). ylum ). For Tsuga , two clades (1a and 1b) were maintained because each contained unique sister-area data and divergence Divergence times — We summarized the available diver- times (see Materials and Methods; Table 1 ). For three clades, gence time data for intracontinental and intercontinental dis- Juniperus 1, Myrica and Sambucus 2, each focal species is cir- junctions with 95% confi dence intervals when available ( Table cumboreal with sister taxa distributed in EA and WNA, respec- 1 ; Figs. 5, 6 ). A total of 20 clades with ENA–EMEX and ENA– tively. To avoid recognizing all three clades as endemic to WNA splitting times were identifi ed, ranging from 3 Ma (Vi- North America, we chose an updated area of CIRCUM to rep- burnum 2a and 2b) to 49.04 Ma (Aesculus ) (Fig. 5A, B). A resent each clade as northern hemisphere in distribution. total of 65 clades with ENA-EA splitting time were identifi ed, Testing for phylogenetic structure among the clades with dis- ranging from 2.38 Ma (Decumaria ) to 130 Ma ( Pinus ) ( Fig. 6 ). junctions to Eurasia showed that they are randomly distributed For Aesculus, the consistently older divergence times associ- relative to the higher-level phylogeny of seed plants represented ated with the ENA–WNA and ENA, WNA–EA disjunctions by the Southern Appalachian species (Appendix S2b). were calculated with the program Multidivtime ( Thorne et al., 1998 ). The older splitting times within Pinus and Platanus are Taxon diversity across clades — Approximately 978 North compatible with the Cretaceous fossil record for these clades American woody taxa are encompassed by the 150 clades (see ( Miller, 1976 ; Crane et al., 1988 ). Table 1 , sum of taxon diversity). Of these, 442 taxa (45.2%) belong to 25 endemic clades (but see below), whereas the re- Mapping component areas — We summarized biogeographic maining 536 taxa (54.8%) are associated with the 125 clades connections between ENA and other component areas of ende- with sister groups that include taxa distributed in Europe and mism across all patterns (Fig. 7A, B ). Four clades with phylo- eastern Asia. Most of these clades are low in diversity, includ- genetic connections to widespread, circumboreal taxa required ing 56 single-taxon clades, 58 clades with 2–4 taxa, and 28 a revised treatment to correctly represent component areas: for clades with 5–25 taxa ( Fig. 4A ). Over 50% of the total taxa Kalmia 1b, a circumboreal sister taxon required the scoring of occurred within seven clades that range from 49–130 species ENA, WNA, EA, and EU; for Juniperus 2, Myrica and Sambucus ( Crataegus , Ceanothus , Gaylussacia 2, Hypericum , Pinus 2, 2, a circumboreal distribution detected in the secondary analy- and Quercus 1 and 2). We counted the taxon diversity of Pinus sis was treated as EA and EU; and for Betula 2, the combined

Fig. 4. Histograms showing clade number and taxon diversity across clades (see Table 1 ). (A) Taxon diversity for all 150 clades. (B) Taxon diversity for clades with disjunctions to the Old World. Histograms were produced with QtiPlot version 0.9.8.9 ( Vasilef, 2013 ). 794 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

Fig. 5. Divergence times plots for 20 intracontinental disjunctions. Point and error estimations were taken directly from the literature (see Table 1 ). (A) Sister taxa distributed in eastern North America (ENA) and eastern Mexico (EMEX). (B) Sister taxa distributed in eastern North America and western North America (WNA). Plots were produced with QtiPlot version 0.9.8.9 ( Vasilef, 2013 ). sister area including circumboreal taxa required the addition of sister taxon in the Eurasian fl ora, most often in EA. The domi- WNA and ENA to the component area score. nant pattern of Eurasian connections in over 80% of the clades Overall, the percentage of single-area connections (not com- results from our secondary analysis, providing strong support binations) showed that the areas of ENA and EA, in roughly for Gray’s initial hypothesis (Gray, 1846) that the woody fl ora equal proportion ( Fig. 7A ), are the most important for locating of ENA is most closely related to that of EA. Connections to the closest relatives of the Southern Appalachian woody fl ora. WNA and EMEX underscore more recent movement and di- Next in equal proportion are the areas of WNA and EMEX, versifi cation. At deeper phylogenetic scales, the increase in with EU notably lower, and limited connections to areas south phylogenetic affi nities with EU is noteworthy, as only one case of Mexico. The relative importance of single-area connections of an ENA–EU connection was observed in the primary analy- derived from deeper biogeographic disjunctions is summa- sis. This increase is due in part to deeper connections between rized in Fig. 7B . Single area connections now refl ect the area ENA and EU taxa present in the following clades: Acer 1, Cas- pattern results of the secondary analysis of 68 clades, plus the tanea , Cercis , and Lonicera 1. The net increase of EU as a com- intermediate areas for nine clades (see above). The increase in ponent area is mostly derived from combined areas patterns that the number of clades with deeper Eurasian biogeographic dis- include EA. junctions resulted in a proportional loss of North American sister areas. Specifi cally, the diminished connections of sister Patterns of endemic diversifi cation — Endemic clades of areas of WNA and ENA, and combinations of ENA–EMEX North America— One set of patterns revealed by the primary and EMEX–WNA contributed to the increase in deeper area analysis of sister areas highlights the contribution of North connections to EU (2.8-fold) and EA (2.0-fold) in the second- American endemic clades to both the phylogenetic diversity of ary analysis (Fig. 7A, B). ENA and to its broader fl oristic connections. The 25 endemic clades provide a range of geographic distributions from widespread, i.e., often extending throughout the Americas, to re- DISCUSSION gional, and local, i.e., restricted to the area of ENA (Table 1). Although most of the clades are representative of northern Biogeography and diversity of a regional fl ora across phy- hemisphere families, both Hypericum (Hypericaceae, Myrian- logenetic scales — The results of our primary analysis in combi- dra s.s. clade) and Sideroxylon (Sapotaceae, Bumelia clade) nation with estimates of taxon diversity show that over 25% of contribute several new lineages and support for south-to-north the clades with representative taxa in the Southern Appalachian migration in the assembly of the ENA fl ora. Bumelia likely woody fl ora have sister taxa that occur strictly within the area of originated in the neotropics and contributes limited diversity to ENA. With the addition of combined areas patterns that contain the regional fl ora of ENA ( Smedmark and Anderberg, 2007 ). In ENA, we suggest that well over one third of the clades have contrast, Myriandra s.s. is the most species-rich clade of shrubs likely undergone speciation within this broadly defi ned regional within ENA, extending south to Central America and the Carib- fl ora. A large component of apparent in situ diversifi cation, bean, and ultimately shares a deep ancestry with Hypericum 45% of the overall taxon diversity accumulated by all clades in subclades that form a cosmopolitan distribution (Nurk et al., the sample, is distributed among just 25 endemic clades. Sev- 2013 ). Less diverse, but equally novel in its phylogenetic con- eral of the larger species radiations, in particular the clades Hy- tribution to the regional endemism of ENA is Asimina (An- pericum ( Myriandra s.s. clade), Pinus 2 (Trifoliae clade), and nonaceae), with a southeastern distribution and somewhat Quercus 2 (Lobatae clade) are key sources of taxon richness obscure broader phylogenetic relationships to the pantropical and notable contributors to the fl ora of the Americas. Annona clade ( Couvreur et al., 2008 ). The unusual disjunct pat- Not surprisingly, our primary analysis determined that approx- tern of ENA–SAMER shown by Gaylussacia 2 represents an- imately 30% of the North American clades have an immediate other clear example of neotropical connections to the ENA fl ora MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 795

Fig. 6. Divergence times plot for intercontinental disjunctions between eastern North America (ENA) and eastern Asia (EA) for 65 clades. Point and error estimations were taken directly from the literature (see Table 1). For the clades plotted, solid circles represent clades with 1 species; open circles represent clades with more than 1 species, and solid squares represent clades with one or more species sister to a combined New and Old World areas (OW/NW SA). Plots were produced with QtiPlot version 0.9.8.9 ( Vasilef, 2013 ).

with phylogenetic support for a south-to-north movement most these, including Quercus, have fi gured prominently in the de- likely via long-distance dispersal ( Floyd, 2002 ). velopment of a robust hypothesis for a north-to-south migration Several other endemic clades form broader distributions en- of the ENA temperate fl ora into northern Mesoamerica during compassing North America, with pronounced extensions south the Miocene (Graham, 1999b). According to this scenario, ele- into Mexico, and Central and South America. Of these clades, ments of the fl ora of ENA reached their southern limit, some Quercus 2 ( Lobatae clade), Ceanothus, and Pinus 2 ( Trifoliae attaining their greatest latitudinal range and infl uence on the clade) show exceptional taxon diversity. Pinus 2 comprises the composition of adjacent fl oras to the south. One extreme ex- North American hard pines, the largest radiation in the genus, ample of this pattern is the Juglans 2 clade (Rhysocaryon ) that with a broad distribution throughout the Americas, and a deep ranges from southern Canada to northern Argentina (44° N to Cretaceous split to Eurasia (see below; Herná ndez-León et al., 28 ° S latitude) to form a recent and widespread diversifi cation 2013 ). Although Rhamnaceae are well established throughout throughout riparian zones of the Americas ( Manos et al., 2007 ; temperate fl oras, Ceanothus is best known as a Madrean group Stone et al., 2009 ). within WNA. The presence of Ceanothus in ENA, despite limited phylogenetic resolution, supports the notion that the mesic Endemic clades of eastern North America— In contrast to the habitat for the clade is plesiomorphic, the likely condition for well-established high levels of regional endemism within the other Madrean clades that diversifi ed recently within WNA, no- Eastern Asian woody fl ora (for review, see Harris et al., 2013 ), tably in the California Floristic Province ( Burge et al., 2011 ). the number of clades endemic to ENA with representation in The remaining endemic clades consist of modest temperate the fl ora is low, and all but one belong to temperate families. radiations ranging from 6 to 17 species, including Juglans 2 Our estimate of 12—Aronia , Asimina , Eubotrys , Fothergilla , ( Rhysocaryon clade), Prunus 1 ( Prunocerasus clade), Robinia , Gaylussacia 1, Hudsonia , Ilex 1, and fi ve monotypic clades—is and Ulmus 2 ( Trichoptelea + Chaetoptelea clades). Some of consistent with fl oristic surveys showing that the coastal plain 796 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

Fig. 7. Component area maps of the biogeographic connections for the clades studied. The radius of each circle visually represents a proportional count of each area of endemism. (A) Results of the primary analysis of sister areas for 158 clades. (B) Results of the secondary analysis for 150 clades. Area abbreviations: eastern North America (ENA), eastern Mexico (EMEX), western North America (WNA), Europe (EU), eastern Asia (EA), South America (SAMER), Central America (CAMER), Caribbean (CARIB), and circumboreal (CIRCUM). The maps were produced with ArcMap 10.2 ( Esri, 2013 ). of ENA harbors a higher level of vascular plant endemism rela- species distributed in the Southern Appalachians, including tive to upland areas (Sorrie and Weakley, 2001; Weakley, 2005, Piedmont monadnocks, and the coastal plain, but mostly absent 2012 ). A case could be made to include Diervilla, but phyloge- from the intervening piedmont (Fernald, 1931; e.g., Triplett netic analysis suggests that the regionally endemic ENA spe- et al., 2006 ). In the case of Eubotrys , long considered part of Leu- cies are nested within the Asian group Weigela ( Kim and Kim, cothoe , phylogenetic analysis shows the two ENA species to be 1999 ; Donoghue et al., 2001). Notwithstanding, the complete the morphologically disparate sister group to the circumboreal picture of regional woody endemism includes the monotypic species Chamaedaphne calyculata and more distant from the clades Comptonia (Myricaceae), Gaylussacia (Ericaceae, sec- newly circumscribed clades Leucothoe s.s. and Eubotryoides tion Vitis-idaea ), Nestronia (Santalaceae), Oxydendrum (Erica- (Bush et al., 2010). Like Eubotrys , Fothergilla is morphologi- ceae), and Xanthorhiza (Ranunculaceae). A New World origin cally divergent from its nearest relative Hamamelis , the two ac- seems likely for most of the monotypic taxa on the basis of counting for the genus-level breadth of New World temperate current distribution alone; however, Comptonia is clearly pa- Hamamelidaceae. For Hudsonia , the pattern is similar for the leoendemic, as judged by fossils located throughout the north- two species H. montana and H. ericoides L., but the third ern hemisphere (see Manchester, 1999 for review). species, H. tomentosa Nutt. may be a more distant relative Taxon diversity within the regionally endemic clades of ( Arrington and Kubitski, 2003 ; Weakley, 2012 ). ENA is also low, ranging from 2 to 10 species. With two spe- Other ENA endemics include Gaylussacia 1 (Ericaceae, cies each, Eubotrys (Ericaceae) and Fothergilla (Hamamelida- Decamerium clade; Floyd, 2002 ) and Ilex 1 (deciduous Ameri- ceae) stand out as part of an established and somewhat can clade, in part; see Selbach-Schnadelbach et al., 2009). The anomalous biogeographic pattern consisting of a pair of sister case for Gaylussacia 1 is more clear-cut, whereas Ilex occurs MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 797

worldwide, and as one of the most complex genera studied Molecular studies have mostly revealed some level of diver- here, the designation of Ilex 1 as a regionally endemic sublade gence and estimated splitting times that fall within the low is provisional and awaits improved resolution to other groups range of values recovered between intercontinentally disjunct within the Ilex clade. The confusing taxonomic history of Aro- taxa (compare Figs. 5A and 6). Temporal data are needed to nia , placed by different authors among various maloid groups broaden the support for congruence in vicariant ENA–EMEX has been resolved by analyses including species of Pourthiaea , pairs from other clades (e.g., Dirca, Pinus 1, and Ulmus ). an eastern Asian group of shrubs (Lo and Donoghue, 2012). This is the only example in our analysis of a deep split involv- Eastern North America–western North America— The origin ing another genus. We chose to recognize Aronia as a regional of the ENA–WNA disjunction follows the general paradigm of endemic, but also include its sister relationship and divergence widespread temperate clades disrupted by increasing range of time data to Pourthiaea . temperature and shifts in seasonal precipitation in association The biogeographic connections of the endemic clades are with the uplift of the Rocky Mountains and development of most apparent once the clades sharing nodes to Eurasia are deserts and grasslands across the Midwest (Wood, 1970; Raven identifi ed (see Fig. 7A, B). As shown in the results of the sec- and Axelrod, 1978 ; Graham, 1999a ). Accordingly, the Late Eo- ondary analysis, endemic clades share similar levels of phylo- cene fossil record of WNA provides evidence for a diverse col- genetic affi nity with other ENA species and disjunct WNA lection of temperate clades with modern equivalents that are species, whereas connections to EMEX fall off as deeper nodes now restricted to EA and ENA (e.g., Manchester, 1999 ). It reveal widespread clades spanning the North Hemisphere. seems likely that most of the woody temperate fl ora of WNA There are slight increases into tropical areas following range was driven to extinction by cooling and drying events of the extensions at deeper nodes within several clades (e.g., Juglans Oligocene and Miocene (Axelrod and Schorn, 1994; Graham, 2, Ilex 3, Ceanothus , and Quercus 2). This small set of ENA 1999a ), consistent with the modern, cold-intolerant tree fl ora of clades is the source of considerable taxon diversity throughout WNA ( Hawkins et al., 2014 ). the Americas and noteworthy additions to the phylogenetic pro- We also note that many of the remaining woody sister-taxon fi le of the Southern Appalachian fl ora. connections with the fl ora include clades of shrubs (e.g., Caly- canthus , Rhododendron 3) and understory trees (e.g., Cercis, Patterns of diversifi cation and disjunction— Disjunct Cornus 3, and Morus ). Exceptions include the Quercus 1 and 2 clades within North America— Clades consisting of a single clades, as preliminary data show the earliest split within both ENA species with disjunctions to sister species located in the involve WNA taxa and successive sister clades distributed in single areas of EMEX and WNA, respectively, include a few ENA and Mesoamerican areas ( McVay et al., 2014 ). Overall, North American endemic clades (e.g., Amelanchier 2, Dirca , the contribution of clades with tree species connecting ENA and Ptelea ) and over 30 clades with deeper disjunctions to Eur- and WNA areas is quite low, suggesting that very few clades asia (Table 1; Fig. 3A, B). About 20 case studies within the adapted to the evolution of the current Madrean climate. latter category have phylogenies and temporal information, Our primary analysis of sister areas also recovered over 20 both fossil and molecular, to address the hypothesis that the disjunctions between the single areas of ENA and WNA, 10 area disjunctions—ENA–EMEX and ENA–WNA—in differ- with temporal data points, and several other clades with WNA ent sets of widespread North American clades are congruent as part of combined area patterns ( Table 1 , Fig. 5B ). The dy- with Miocene to Pliocene cooling and drying trends, one of the namics of this pattern can be better understood by drawing in- primary drivers of the extinction of intervening temperate com- ferences on the geographic origin of clades and the direction of munities and sources of broad barriers to gene fl ow ( Graham, migration prior to vicariance (Donoghue and Smith, 2004; Wen 1999a ). et al., 2010 ). For a set of 18 clades ( Table 2 ), we found that eight clades originated in WNA and moved to ENA, one origi- Eastern North America–eastern Mexico— Although EMEX nated in ENA and moved to WNA (Crataegus , see below), and has been treated as an extension of ENA (see Donoghue and nine were equivocal. Of the eight cases suggesting movement Smith, 2004 ), we recognize it here as a distinct area of ende- from WNA to ENA, seven show that the ENA–WNA clade is mism because of the growing number of cases available for most closely related to one or more species from EA. Further- study. A common framework of north-to-south movement fol- more, the ancestors of these clades must have moved to ENA as lowed by vicariance due to range contraction is suggested by elements of mesic-adapted communities prior to the onset of this disjunction pattern in over 50 temperate lineages (Miranda drier conditions in WNA (Fritsch and Cruz, 2012; Baldwin, and Sharp, 1950 ; Graham, 1999b, 2010 ; Morris et al., 2008; 2014 ). The combination of data supports the “out of Asia” hy- Ruiz-Sanchez and Ornelas, 2014 ). We fi nd strong phylogenetic pothesis with movement from EA into WNA, presumably support for Graham’s microfossil-derived hypothesis for a mid- through the Bering Land Bridge ( Donoghue et al., 2001 ; Wen dle Miocene incursion of the eastern North American temperate et al., 2010 ). In contrast, three other cases with identical area fl ora into the highlands and cloud forests of Mesoamerica. Esti- patterns (Aesculus , Crataegus, and Rhus 1) suggest origins in mates for the node age of ENA–EMEX disjunctions generally North America with movement to the Old World. fall within the last 10 Myr ( Fig. 5A ), coinciding with climate One possible explanation for the maintenance of ENA– change associated with drying and the incursion of the Chihua- WNA fl oristic connections is that as temperate communities in huan Desert of southern Texas and northeastern Mexico (Graham, WNA decayed through climate change, persistent understory 2010 ). elements with broad ecological amplitude moved to suitable hab- The high degree of morphological similarity across a range itats among newly emerging community types. While perhaps of disjunct taxa sharing this pattern (e.g., Carpinus , Cercis , slightly older, the relative timing of the ENA–WNA disjunctions Fagus , Liquidambar , Magnolia 2b, Nyssa , Pinus 1, Platanus , generally coincides with the node-age dates of ENA–EMEX dis- and Tilia ) also suggests relatively recent boundary formation. junctions. An additional set of contemporaneous splitting events 798 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY also can be derived from the intercontinental ENA–EA disjunc- complexes of sister taxa, two other patterns involve clades with tions within the last 10 million years (see Fig. 6, lower portion). older, more divergent sister clades distributed in complex pat- Thus, a broad assemblage of ENA clades across three pairs of terns spanning the areas of ENA + EA and WNA + EA. One set area disjunctions (ENA–WNA, ENA–EMEX, ENA–EA) show of cases includes taxonomically distinct ENA taxa that are en- general temporal congruence, likely as part of the widespread demic to the Southern Appalachians (e.g., Aristolochia , Pieris , response and disruption of temperate biotic connections to cli- and Tsuga ). Phylogenies of these clades yield unusual area pat- mate change in the late Miocene to early Pliocene. terns and a few suggest lineage diversifi cation within North America followed by movement to Eurasia. In Aristolochia , reconstruction of the ancestral area for the node in question is Northern American clades with widespread sister clades— á Our primary analysis also reveals that just over 7.6% of the fo- equivocal ( Gonz lez et al., 2014 ), but in Pieris , two dispersals cal clades have immediate sisters that encompass distribution from the New to Old World are suggested ( Li et al., 2009 ). patterns combining areas of North America and Eurasia ( Fig. A second biogeographic pattern includes clades with some 3A ). When clades sharing regional patterns of North American level of diversifi cation and deep sister clades that often span the diversifi cation and intracontinental disjunctions are added to areas of WNA and Eurasia (e.g., Crataegus, Pinus 1, Platanus , the total (e.g., Crataegus , Pinus 1, Platanus , and Staphylea ), Rhus 2a, Staphylea, and Aralia 2). For Crataegus, one of the the percentage of clades with widespread sisters increases to most taxon-rich clades within ENA, area reconstructions sug- over 14% (Fig. 3B). For these clades, the range of area combi- gest an origin in ENA and successive migrations, fi rst to WNA nations and divergence times refl ect at least three distinct tem- and second to EA (Lo et al., 2009). Divergence time data esti- poral events and a mixed signal of origination and movement. mate that the split between the ENA and the WNA + EA clade Thus, pseudocongruent area relationships are strongly sus- occured approximately 25 Ma, more than twice as long ago as pected from the combination of evidence including fossil oc- the larger set of ENA–WNA disjunctions (Fig. 5B) that connect currence, phylogenies with branch lengths, and estimations of to deeper sister clades in Eurasia. This difference suggests that divergence times ( Table 1 , Fig. 6 ). the Crataegus pattern is temporally distinct from the (ENA– WNA) + EA pattern found in Calycanthus , Cercis, and Vibur- num 2 (see Table 2 ). Circumboreal clades and species— One clear pattern of A subset of cases includes the gymnosperm clades, Taxus sister-area connections highlights clades with shallow diver- and Tsuga , which exhibit nonmonophyletic North American gences, seemingly recent high-latitude movement, and relatively species. For Taxus , phylogenetic analysis suggests that T. ca- low levels of diversity. For many of these clades, taxonomic nadensis is sister to the rest of the genus (Hao et al., 2008); thus, distinction between New and Old World taxa is negligible, and the deepest split within the genus likely took place within the the general lack of phylogenetic resolution within species com- New World. For Tsuga , analyses show that the two ENA spe- plexes and recognized geographic varieties supports low levels cies are neither sister taxa nor closely related to WNA taxa. The of genetic differentiation (e.g., Betula 2, Frangula , and Sibbal- widespread T. canadensis ( Tsuga 1a) is sister to a clade that dia). Instructive examples include Alnus 2 and Corylus 2, each contains the Southern Appalachian endemic, T. caroliniana with unresolved species from ENA, WNA, and EA ( Chen and ( Tsuga 1b), plus all of the species from eastern Asia ( Havill Li, 2004 ; Whitcher and Wen, 2001 ). et al., 2008). Biogeographic inferences drawn from this discon- A similar biogeographic pattern involves circumboreal spe- tinuous northern hemisphere clade support successive vicari- cies that have been variously delimited geographically into spe- ance events leading to disjunct patterns that isolate North cies or varieties, such as Myrica gale and Sambucus racemosa American taxa from larger clades containing combinations of ( Sambucus 2). In these two cases, WNA appears to harbor the new and Old World species. Reconstruction of ancestral areas most morphologically divergent sister lineages. In Myrica , the within the Tsuga crown clade suggests the disjunction between disjunct and narrow WNA endemic M. hartwegii S. Watson is T. caroliniana and other EA taxa occurred around 20 Ma, pre- found in the Sierra Nevada of California, but it has yet to be ceded by movement from Asia ( Havill et al., 2008 ). studied with molecular data, and combined molecular and mor- The role of temporal data is critical when assessing biogeo- phological data support the more widely ranging S. melano- graphic histories in several clades that diversifi ed during the carpa to be sister to the species complex that includes S. Cretaceous, especially those containing ENA taxa that are not racemosa s.l. ( Eriksson and Donoghue, 1997 ). Another inter- sister to WNA taxa. In the case of Platanus , three different time esting case study is Juniperus communis ( Juniperus 1) as it too periods are refl ected in its calibrated phylogeny with bearing fi ts the pattern of circumboreal distribution with recognized on North American plant biogeography ( Feng et al., 2005 ). geographic varieties, such as J . communis var. depressa Pursh Whereas the sister taxa Platanus occidentalis (ENA) and in the Southern Appalachians. Phylogenetic analyses suggest P. mexicana Moric. (EMEX) show temporal congruence with modest differentiation within the J. communis complex and evi- other temperate ENA–EMEX disjunctions (see Fig. 6A), the dence for successive sister lineages within the area of EA, and divergence time of this subclade to its Madrean sister group support for an Asian origin of the Juniperus 1 clade (Mao et al., disjunctly distributed in WNA and EU predates most of the bio- 2010 ). A counter example of movement from North America to geographic patterns that have shaped the ENA fl ora ( Fig. 6 ). Eurasia is evident from the largely North American pattern of This older pattern is thought to refl ect dry-adapted parallel evo- diversifi cation within Kalmia . A recent origin from North lution in formerly widespread temperate clades ( Kadereit and America can be inferred from the sister relationship of K . buxi- Baldwin, 2012 ). For Pinus 1 (Quinquefoliae clade), there is no folia ( Kalmia 1b) to the circumboreal species K. procumbens divergence date available for the ENA–EMEX disjunct pair P. (L.) Gift & Kron ( Gillespie and Kron, 2010 ). strobus and P. chiapensis (Mart.) Andresen, but as in Platanus , a deeper intercontinental disjunction (WNA–EA) within this Clades with deeper divergences in Eastern North America— clade is dated to be at least 60 Ma (Eckert and Hall, 2006). In In contrast to the pattern of recent differentiation and widespread the case of Pinus 2 ( Trifoliae), the deepest node, ca. 90–120 Ma MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 799

( Eckert and Hall, 2006 ), marks an intercontinental split with et al., 2010 ), we also conclude that maximizing independent section Pinus that stands out as an extremely pseudocongruent biogeographic histories has interpretative advantages over area pattern linking New and Old World diversifi cations (Fig. 6). checklists and calculations of fl oristic similarities using the presence and absence of genera, especially as the number of North American clades with disjunctions to Eurasia (ENA–EA; relevant phylogenies increases. Nonphylogenetic studies ad- ENA–EU, EA; ENA–WNA, EA; ENA–ENA, EA) — As expected, dressing the affi nities of the ENA fl ora have done so by either a large number of the clades recorded in our study represent the combining montane elements with other habitat groupings to classic disjunct clades of the temperate fl ora of the northern cover all of ENA (e.g., piedmont, coastal plain; Qian, 2002b ) or hemisphere. In addition to the 46 clades comprising one South- by analyzing them separately through various habitat qualifi ers ern Appalachian species with immediate sister taxa in the Old ( Weakley, 2005 ). World, our secondary analysis of deeper biogeographic patterns In an analysis of the fl ora of southeastern and mid-Atlantic identifi ed 50 additional clades with Old World disjunctions, states, Weakley (2005) attributed over 21% of the overall af- most to EA, with a minority combining EA with other areas. This fi nities to the “Tertiary relict” distribution type, a rough equiva- latter assemblage of clades consists of two components (see Fig. lent of the patterns we recognize as clades with disjunctions to 6 ): a larger group characterized by ENA focal clades composed Eurasia. The higher percentage of clades with Old World con- of more than one species, evidence for diversifi cation that post- nections detected in our study is no doubt a function of our re- dates disjunction, and a smaller group of clades that share the stricted sample in area and plant habit (e.g., woody species). A property of having one or more species that are sister to wide- range of distribution types were reported for the associated spread clades (e.g., Aristolochia , Crataegus , Pieris , and Tsuga ). “higher taxa” that encompass over 1700 mountain species and Temporal information for 65 intercontinentally disjunct clades estimated the following set of biogeographic affi nities: 23.5% varies considerably across the sample suggesting no relation- circumboreal, 14.3% “Tertiary relict”, 1.9% western North ship between clade age and taxon diversity. Furthermore, the America, 7.2% tropical, and ca. 20% North America. Although distribution of divergence times is continuous throughout the it is diffi cult to compare the results of his approach with ours, it Cenozoic, providing strong support for the hypothesis of pseudo- is likely that the woody subset of diversity in our study contains congruent biogeographic histories in the formation of the ENA– a higher proportion of Tertiary relicts. By expanding this study EA pattern. The combination of intercontinental patterns and to include herbaceous clades, it should be possible to produce a temporal data suggests that the response of north hemisphere clade-based synthesis for the entire vascular fl ora of the South- clades to climate change has been a dynamic and continuous ern Appalachians based on biogeographic patterns. process, with widespread distributions forming via migration Recent studies using phylogenetic approaches to determine during favorable periods, and barriers to gene fl ow, range ex- the origin and affi nities of particular taxa in geographic space tinctions, and eventual disjunctions resulting repeatedly during (bryophytes on the Pantepui Region, Desamore et al., 2010; cooling trends. ferns on Mascarene Islands, Hennequin et al., 2014 ) or as ap- Although many of the core ENA–EA disjunctions have been plied to regional fl oras (e.g., Australia, Australian rainforest, discussed in previous meta-analyses of northern hemisphere California Floristic Province, Madagascar, Pacifi c Islands, tem- biogeography ( Wen, 1999 ; Donoghue and Smith, 2004 ; Wen perate South America) have provided valuable biogeographic et al., 2010), our study is the fi rst to assess their geographic insights ( Sanmartín and Ronquist, 2004 ; Yoder and Nowak, connections in the context of a whole woody fl ora in time and 2006 ; Price and Wagner, 2011; Sniderman and Jordan, 2011; space. One notable feature of the distribution of divergence Crisp and Cook, 2013; Baldwin, 2014). Future directions in- times is that most estimates are less than 30 Myr, suggesting clude the exploration of phylogenies to address specifi c ques- that barriers to gene fl ow and likely extinctions began at the end tions in the origins and evolution of Southern Appalachian of the Eocene when estimated global temperature sharply de- communities, as well as new perspectives on the role biogeo- creased within in a narrow time frame (Zachos et al., 2001). graphic history in community structure in relation to niche and Older divergence times (e.g., Pinus 1, Pinus 2, Platanus ) are functional traits. more likely a function of phylogenetic divergences and physio- graphic events that predate the primary windows of fl oristic as- LITERATURE CITED sembly within ENA. Paleoclimate reconstructions also suggest that the tempera- APG III [ANGIOSPERM PHYLOGENY GROUP . 2009 . An update of the ture decrease at the end of the Eocene was followed by a long Angiosperm Phylogeny Group classifi cation for the orders and fami- period of stasis and slight global warming, roughly from 26 to lies of fl owering plants: APG III. Botanical Journal of the Linnean 12 Ma ( Zachos et al., 2001 ). This condition would have pro- Society 161 : 105 – 121 . vided an opportunity for movement and diversifi cation across A PPELHANS , M. S. , J. WEN , K . W OOD , G . A LLAN , E . Z IMMER , AND W . W AGNER. 2014 . Molecular phylogenetic analysis of Hawaiian the northern hemisphere. Divergence times for over 60% of the Rutaceae (Melicope, Platydesma and Zanthoxylum) and their differ- clades studied split from their Old World sisters within the last ent colonization patterns. Botanical Journal of the Linnean Society 20 Myr, with about half diverging between 20 and 10 Ma. The 174 : 425 – 448 . remaining intercontinental disjunctions took place within the A RRINGTON , J. M. 2004 . Systematics of the Cistaceae. Ph.D. dissertation, last 10 Myr, possibly as part of temporally congruent and wide- Duke University, Durham, North Carolina, USA. spread barriers to gene fl ow followed by Late Miocene range A RRINGTON , J. M. , AND K . K UBITSKI . 2003 . Cistaceae . In K. Kubitzski extinctions that further shaped and modernized the temperate and C. Bayer [eds.], Families and genera of vascular plants, vol. 5, refuges of today. Flowering plants: Dicotyledons: Malvales, Capparales and non-beta- lain Caryophyllales, 62–70. Springer, Berlin, Germany. A XELROD , D. I. 1966 . Eocene Copper Basin fl ora of northeastern Nevada. Conclusions — Like other synthetic studies of northern hemi- University of California Publications in Geological Sciences, vol. 59. sphere phytogeography (e.g., Donoghue and Smith, 2004 ; Wen University of California Press, Berkeley, California, USA. 800 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

A XELROD , D. I ., AND H. E . SCHORN . 1994 . The 15 Ma fl oristic crisis at D AVIS , M. B. 1983 . Quaternary history of deciduous forests of eastern Gillam Spring, Washoe County, northwestern Nevada. PaleoBios 16 : North America and Europe. Annals of the Missouri Botanical Garden 1 – 10 . 70 : 550 – 563 . B ALDWIN , B. G. 2014 . Origins of plant diversity in the California D ELCOURT , P. A. , AND H. R. DELCOURT . 1987 . Long-term forest dynamics Floristic Province hot spot. Annual Review of Ecology, Evolution and of the temperate zone . Springer, New York, New York, USA. Systematics 45 : 347 – 369 . D ENK , T . , AND G. W. GRIMM. 2009 . The biogeographic history of beech B ELL , C. D. 2010 . Towards a species level phylogeny of Symphoricarpos trees. Review of Palaeobotany and Palynology 158 : 83 – 100 . (Caprifoliaceae), based on nuclear and chloroplast DNA. Systematic D ÉSAMORÉ , A . , A . V ANDERPOORTEN , B . L AENEN , S. R. GRADSTEIN , AND P. J. Botany 35 : 442 – 450 . R. KOK . 2010 . Biogeography of the Lost World (Pantepui region, B OLMGREN , K . , AND B . O XELMAN . 2004 . Generic limits in Rhamnus L. s.l. northeastern South America): Insights from bryophytes. Phytotaxa 9 : (Rhamnaceae) inferred from nuclear and chloroplast DNA sequence 254 – 265 . phylogenies. Taxon 53 : 383 – 390 . D OBE Š , C . , AND J . P AULE . 2010 . A comprehensive chloroplast DNA- B ORTIRI , E. , S. H. OH , J. G. JIANG , S . B AGGETT , A . G RANGER , C . W EEKS , M . based phylogeny of the genus Potentilla (Rosaceae): Implications for B UCKINGHAM , D . P OTTER , AND D. E. PARFITT . 2001 . Phylogeny and its geographic origin, phylogeography and generic circumscription. systematics of Prunus (Rosaceae) as determined by sequence analysis Molecular Phylogenetics and Evolution 56 : 156 – 175 . of ITS and the chloroplast trnL-trnF spacer DNA. Systematic Botany D ONOGHUE , M. J. 2008 . A phylogenetic perspective on the distribution 26 : 797 – 807 . of plant diversity. Proceedings of the National Academy of Sciences, B RAUN , E. L. 1950 . Deciduous forests of eastern North America . Hafner, USA 105 ( supplement 1 ): 11549 – 11555 . New York, New York, USA. D ONOGHUE , M. J. , C. D. BELL , AND J. H. LI . 2001 . Phylogenetic patterns in B RAUN , E. L. 1955 . The phytogeography of unglaciated eastern United Northern Hemisphere plant geography. International Journal of Plant States and its interpretation. Botanical Review 21 : 297 – 375 . Sciences 162 : S41 – S52 . B RUNEAU , A. , J. R. STARR , AND S . J OLY . 2007 . Phylogenetic relationships in D ONOGHUE , M. J. , AND S. A. SMITH. 2004 . Patterns in the assembly of the genus Rosa: New evidence from chloroplast DNA sequences and temperate forests around the Northern Hemisphere. Philosophical an appraisal of current knowledge. Systematic Botany 32 : 366 – 378 . Transactions of the Royal Society of London, B, Biological Sciences B URGE , D. O. , D. M. ERWIN , M. B. ISLAM , J . K ELLERMANN , S. W. KEMBEL , 359 : 1633 – 1644 . D. H. WILKEN , AND P. S. M ANOS . 2011 . Diversifi cation of Ceanothus D RUMMOND , A. J. , AND A . R AMBAUT . 2007 . BEAST: Bayesian evolution- (Rhamnaceae) in the California Floristic Province. International ary analysis by sampling trees. BMC Evolutionary Biology 7 : 214 . Journal of Plant Sciences 172 : 1137 – 1164 . D UANGJAI , S . , R . S AMUEL , J . M UNZINGER , F . F OREST , B . W ALLNÖFER , M. H. J. B USH , C. M. , W. S. JUDD , L . F RAZIER , AND K. A. KRON . 2010 . The phy- B ARFUSS , G . F ISCHER , AND M. W. CHASE . 2009 . A multi-locus plastid logeny of Leucothoe s. l. (Ericaceae: Vaccinioideae) based on mor- phylogenetic analysis of the pantropical genus Diospyros (Ebenaceae), phological and molecular (ndhF , matK, and nrITS) data. Systematic with an emphasis on the radiation and biogeographic origins of the Botany 35 : 201 – 206 . New Caledonian endemic species. Molecular Phylogenetics and C HANEY , R. W. 1947 . Tertiary centers and migration routes. Ecological Evolution 52 : 602 – 620 . Monographs 17 : 139 – 148 . E CKERT , A. J. , AND B. D. HALL . 2006 . Phylogeny, historical biogeography, C HEN , J. H. , H. SUN , J . W EN , AND Y. P. YANG . 2010 . Molecular phylogeny and patterns of diversifi cation for Pinus (Pinaceae): Phylogenetic tests of Salix L. (Salicaceae) inferred from three chloroplast datasets and its of fossil-based hypotheses. Molecular Phylogenetics and Evolution systematic implications. Taxon 59 : 29 – 37 . 40 : 166 – 182 . C HEN , Z . , AND J . L I . 2004 . Phylogenetics and biogeography of Alnus E RIKSSON , T . , AND M. J. DONOGHUE . 1997 . Phylogenetic relationships of (Betulaceae) inferred from sequences of nuclear ribosomal DNA ITS Sambucus and Adoxa (Adoxoideae, Adoxaceae) based on nuclear region. International Journal of Plant Sciences 165 : 325 – 335 . ribosomal ITS sequences and preliminary morphological data. C LARK , J. S. 2010 . Individuals and the variation needed for high species Systematic Botany 22 : 555 – 573 . diversity in forest trees. Science 327 : 1129 – 1132 . ESRI [ENVIRONMENTAL SYSTEMS RESOURCE INSTITUTE] . 2013. ArcMap 10.2. C LEMENT , W. L. , M. ARAKAI , P. W. SWEENEY , E . J. EDWARDS , AND M. J. ESRI, Redlands, California, USA. D ONOGHUE . 2014 . A chloroplast tree for Viburnum (Adoxaceae) and E STILL, J. C., AND M. B. C RUZAN. 2001 . Phytogeography of rare plant its implications for phylogenetic classifi cation and character evolu- species endemic to the southeastern United States. Castanea 6 6 : tion. American Journal of Botany 101 : 1029 – 1049 . 3 – 23 . C LEMENT , W. L. , AND M. J. DONOGHUE . 2011 . Dissolution of Viburnum F AITH , D. P. 1992 . Conservation evaluation and phylogenetic diversity. section Megalotinus (Adoxaceae) of Southeast Asia and its implica- Biological Conservation 61 : 1 – 10 . tions for morphological evolution and biogeography. International F ENG , Y. , S. H. OH , AND P. S. MANOS . 2005 . Phylogeny and historical bio- Journal of Plant Sciences 172 : 559 – 573 . geography of the genus Platanus as inferred from nuclear and chloro- C OUVREUR , T. L. P. , J. E. RICHARDSON , M. S. M. SOSEF , R. H. J. ERKENS , AND plast DNA. Systematic Botany 30 : 786 – 799 . L. W. CHATROU . 2008 . Evolution of syncarpy and other morphologi- F ERNALD, M. L. 1931 . Specifi c segregations and identities in some cal characters in African Annonaceae: A posterior mapping approach. fl oras of eastern North America and the Old World. Rhodora 8 8 : Molecular Phylogenetics and Evolution 47 : 302 – 318 . 25 – 63 . C RANE , P. R. , S. R. MANCHESTER , AND D. L. DILCHER . 1988 . Morphology F IOR , S. , P. O. KARIS , AND A. A. ANDERBERG . 2003 . Phylogeny, taxonomy, and phylogenetic signifi cance of the angiosperm Platanites hebridi- and systematic position of Clethra (Clethraceae, Ericales) with notes cus from the Paleocene of Scotland. Palaeontology 31 : 503 – 517 . on biogeography: Evidence from plastid and nuclear DNA sequences. C RISP , M. D. , AND L. G. COOK . 2013 . How was the Australian fl ora as- International Journal of Plant Sciences 164 : 997 – 1006 . sembled over the last 65 million years? A molecular phylogenetic F LOYD, J. W. 2002 . Phylogenetic and biogeographic patterns in Gaylussacia perspective. Annual Review of Ecology, Evolution and Systematics (Ericaceae) based on morphological, nuclear DNA, and chloroplast 44 : 303 – 324 . DNA variation. Systematic Botany 27 : 99 – 115 . C UENOUD , P . , M . M ARTINEZ , P. A. LOIZEAU , R . S PICHIGER , S . A NDREWS , AND F OREST , F . , V . S AVOLAINEN , M. W. CHASE , R . L UPIA , A . B RUNEAU , P. R. J. F. MANEN . 2000 . Molecular phylogeny and biogeography of the C RANE , AND M . L AVIN. 2005 . Teasing apart molecular- versus fossil- genus Ilex L. (Aquifoliaceae). Annals of Botany 85 : 111 – 122 . based error estimates when dating phylogenetic trees: A case study in C UNNINGHAM , C. W. , AND T. M. COLLINS . 1994 . Developing model sys- the birch family (Betulaceae). Systematic Botany 30 : 118 – 133 . tems for molecular biogeography: Vicariance and interchange in ma- F RITSCH , P. W. 2001 . Phylogeny and biogeography of the fl owering plant rine invertebrates . In B. Schierwater, B. Streit, G. P. Wagner, and R. genus Styrax (Styracaceae) based on chloroplast DNA restriction DeSalle [eds.], Molecular ecology and evolution: Approaches and ap- sites and DNA sequences of the internal transcribed spacer region. plications, 405–433. Birkhauser Verlad, Basel, Switzerland. Molecular Phylogenetics and Evolution 19 : 387 – 408 . MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 801

F RITSCH , P. W. , AND B. C. CRUZ . 2012 . Phylogeny of Cercis based on H AVILL , N. P. , C. S. CAMPBELL , T. F. VINING , B . L EPAGE , R. J. BAYER , AND DNA sequences of nuclear ITS and four plastid regions: Implications M. J. DONOGHUE. 2008 . Phylogeny and biogeography of Tsuga for transatlantic historical biogeography. Molecular Phylogenetics (Pinaceae) inferred from nuclear ribosomal ITS and chloroplast DNA and Evolution 62 : 816 – 825 . sequence data. Systematic Botany 33 : 478 – 489 . F RITSCH , P. W. , AND S. A. LUCAS. 2000 . Clinal variation in the Halesia H AWKINS , B. A. , M. RUEDA , T. F. RANGEL , R . F IELD , J. A. F. DINIZ-FILHO , carolina complex (Styracaceae). Systematic Botany 25 : 197 – 210 . AND P . L INDER . 2014 . Community phylogenetics at the biogeographi- F RITSCH , P. W. , S. R. MANCHESTER , R. D. STONE , B. C. CRUZ , AND F . A LMEDA . cal scale: Cold tolerance, niche conservatism and the structure of 2015 . Northern hemisphere origins of the amphi-Pacifi c tropical plant North American forests. Journal of Biogeography 41 : 23 – 38 . family Symplocaceae . Journal of Biogeography 42 : 891 – 901 . H ENNEQUIN , S . , M . K ESSLER , S . L INDSAY , AND H . S CHNEIDER . 2014 . F RITSCH , P. W. , C. M. MORTON , T . C HEN , AND C . M ELDRUM . 2001 . Phylogeny Evolutionary patterns in the assembly of fern diversity on the oceanic and biogeography of the Styracaceae. International Journal of Plant Mascarene Islands. Journal of Biogeography 41 : 1651 – 1663 . Sciences 162 : S95 – S116 . H ERBERT , J. 2005 . Systematics and biogeography of Myricaceae. Ph.D. G ILLESPIE , E . , AND K . K RON . 2010 . Molecular phylogenetic relationships dissertation, University of St. Andrews, St. Andrews, Scotland, UK. and a revised classifi cation of the subfamily Ericoideae (Ericaceae). H ERN ÁNDEZ -LEÓN , S. , D. S. GERNANDT , J. A. PÉREZ DE LA ROSA , AND L . Molecular Phylogenetics and Evolution 56 : 343 – 354 . J ARDÓN-BARBOLLA. 2013 . Phylogenetic relationships and species de- G OETSCH , L. , A. J. ECKERT , AND B. D. HALL . 2005 . The molecular system- limitation in Pinus section Trifoliae inferred from plastid DNA. PLoS atics of Rhododendron (Ericaceae): A phylogeny based upon RPB2 ONE 8 : e707501 . gene sequences. Systematic Botany 30 : 616 – 626 . H INSINGER , D. D. , J. BASAK , M . G AUDEUL , C . C RUAUD , P . B ERTOLINO , N . G ONZ ÁLEZ , F. , S. T. WAGNER , K . S ALOMO , L . S YMMANK , M. S. SAMAIN , F RASCARIA-LACOSTE , AND J . B OUSQUET . 2013 . The phylogeny and bio- S . I SNARD , N. P. ROWE , C . N EINHUIS , AND S . W ANKE . 2014 . Present geographic history of ashes (Fraxinus , Oleaceae) highlight the roles trans-Pacifi c disjunct distribution of Aristolochia subgenus Isotrema of migration and vicariance in the diversifi cation of temperate trees . (Aristolochiaceae) was shaped by dispersal, vicariance and extinction. PLoS ONE 8 ( 11 ): e80431 . Journal of Biogeography 41 : 380 – 391 . H UANG , W. P. , H. SUN , T . D ENG , S. G. RAZAFIMANDIMBISON , Z. L. NIE , AND J . G RAHAM , A. 1972 . Outline of the origin and historical recognition of W EN . 2013 . Molecular phylogenetics and biogeography of the east- fl oristic affi nities between Asia and eastern North America. In A. ern Asian–eastern North American disjunct Mitchella and its close Graham [ed.], Floristics and paleofl oristics of Asia and eastern North relative Damnacanthus (Rubiaceae, Mitchelleae). Botanical Journal America, 1–18. Elsevier, Amsterdam, Netherlands. of the Linnean Society 171 : 395 – 412 . G RAHAM , A. 1999a . Late Cretaceous and Cenozoic history of North I CKERT-BOND, S. M., AND J . W EN. 2006 . Phylogeny and biogeography of America vegetation north of Mexico. Oxford University Press, New Altingiaceae: Evidence from combined analysis of fi ve non-coding chlo- York, New York, USA. roplast regions. Molecular Phylogenetics and Evolution 39 : 512 – 528 . G RAHAM , A. 1999b . The Tertiary history of the northern temperate el- J ACKSON , S. T. , R. S. WEBB , K. H. ANDERSON , J. T. OVERPECK , T . W EBB III , ement in the northern Latin American biota. American Journal of J. W. WILLIAMS , AND B. C. S. HANSEN . 2000 . Vegetation and environ- Botany 86 : 32 – 38 . ment in Eastern North America during the Last Glacial Maximum. G RAHAM , A. 2010 . Late Cretaceous and Cenozoic history of Latin Quaternary Science Reviews 19 : 489 – 508 . American vegetation and terrestrial environments . Missouri Botanical J ACQUES, F. M. B., AND P . B ERTOLINO. 2008 . Molecular and morphological Garden Press, St. Louis, Missouri, USA. phylogeny of Menispermaceae (Ranunculales). Plant Systematics and G RAY , A. 1846 . Analogy between the fl ora of Japan and that of the United Evolution 274 : 83 – 97 . States. American . Journal of Science and Arts II : 135 – 136 . J ARVINEN , P . , A . P ALME , L. O. MORALES , M . L ANNENPAA , M . K EINANEN , G RAY , A. 1859 . Diagnostic characters of new species of phanerogamous T . S OPANEN , AND M . L ASCOUX . 2004 . Phylogenetic relationships of plants, collected in Japan by Charles Wright, botanist of the U. S. Betula species (Betulaceae) based on nuclear ADH and chloroplast North Pacifi c Exploring Expedition. With observations upon the rela- mat K sequences. American Journal of Botany 91 : 1834 – 1845 . tions of the Japanese fl ora to that of North America, and other parts J IAO , Z . , AND J . L I . 2009 . Phylogenetics and biogeography of eastern Asian– of the northern temperate zone. Memoirs of the American Academy of North American disjunct genus Pachysandra (Buxaceae) inferred from Arts and Sciences. 6 : 377 – 453 . nucleotide sequences. Journal of Systematics and Evolution 47 : 191 – 201 . G RAY , A. 1878 . Forest geography and archeology; a lecture delivered be- K ADEREIT , J. W. , AND B. G. BALDWIN . 2012 . Western Eurasian–western fore the Harvard University Natural History Society, April 18, 1878. North American disjunct plant taxa: The dry-adapted ends of formerly American Journal of Science, series 3 16 : 183 – 196 . widespread north temperate mesic lineages—and examples of long- G UGGER , P. F. , J. S. MCLACHLAN , P. S. MANOS , AND J. S. CLARK . 2008 . distance dispersal. Taxon 61 : 3 – 17 . Inferring long-distance dispersal and topographic barriers during K EMBEL , S. W. , P. D. COWAN , M. R. HELMUS , W. K. CORNWELL , H . M ORLON , post-glacial migration from the genetic structure of red maple (Acer D. D. ACKERLY , S. P. BLOMBERG , AND C. O. WEBB . 2010 . Picante: rubrum L.) in New England. Journal of Biogeography 35 : 1665 – 1673 . R tools for integrating phylogenies and ecology. Bioinformatics 26 : G UO , Q. F ., AND R. E. RICKLEFS . 2000 . Species richness in plant genera dis- 1463 – 1464 . junct between temperate eastern Asia and North America. Botanical K IM , Y. , AND S . K IM . 1999 . Phylogeny of Weigela and Diervilla (Caprifoliaceae) Journal of the Linnean Society 134 : 401 – 423 . based on nuclear rDNA ITS sequences: Biogeographic and taxonomic G UO , Y. L. , A. PAIS , A. S. WEAKLEY , AND Q. Y. XIANG . 2013 . Molecular implications. Journal of Plant Research 112 : 331 – 341 . phylogenetic analysis suggests paraphyly and early diversifi cation K IM , Y . , S . K IM , AND L. R. LANDRUM . 2004 . Taxonomic and phytogeo- of Philadelphus (Hydrangeaceae) in western North America: New graphic implications from ITS phylogeny in Berberis (Berberidaceae). insights into affi nity with Carpenteria. Journal of Systematics and Journal of Plant Research 117 : 175 – 182 . Evolution 51 : 545 – 563 . K RON , K. A. , AND W. S. JUDD . 1997 . Systematics of the Lyonia group H AO , D. C. , B. L. HUANG , AND L . Y ANG . 2008 . Phylogenetic relationships of (Andromedeae, Ericaceae) and the use of species as terminals in the genus Taxus inferred from chloroplast intergenic spacer and nuclear higher-level cladistic analyses. Systematic Botany 22 : 479 – 492 . coding DNA. Biological & Pharmaceutical Bulletin 31 : 260 – 265 . K RON , K. A. , W. S. JUDD , P. F. STEVENS , D. M. CRAYN , A. A. ANDERBERG , H ARRIS , A. J. , J. WEN , AND Q. Y. XIANG . 2013 . Inferring the biogeographic P. A. GADEK , C. J. QUINN , AND J. L. LUTEYN . 2002 . Phylogenetic origins of inter-continental disjunct endemics using a Bayes-DIVA classifi cation of Ericaceae: Molecular and morphological evidence. approach. Journal of Systematics and Evolution. 51 : 117 – 133 . Botanical Review 68 : 335 – 423 . H ARRIS , A. J. , Q. J. XIANG , AND D. T. THOMAS . 2009 . Phylogeny, origin, L ANG , P . , F . D ANE , T. L. KUBISIAK , AND H . H UANG . 2007 . Molecular evi- and biogeographic history of Aesculus L. (Sapindales)—An update dence for an Asian origin and a unique westward migration of spe- from combined analysis of DNA sequences, morphology, and fossils. cies in the genus Castanea via Europe to North America. Molecular Taxon 58 : 108 – 126 . Phylogenetics and Evolution 43 : 49 – 59 . 802 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

L AVIN , M. , M. F. WOJCIECHOWSKI , P . G ASSON , C . H UGHES , AND E . W HEELER . M CCARTHY , D. 2012 . Systematics and phylogeography of the genus 2003 . Phylogeny of robinioid legumes (Fabaceae) revisited: Coursetia Tilia in North America. Ph.D. dissertation, University of Illinois at and Gliricidia recircumscribed, and a biogeographical appraisal of the Chicago, Chicago, Illinois, USA. Caribbean endemics. Systematic Botany 28 : 387 – 409 . M CLACHLAN , J. S. , J. S. CLARK , AND P. S. MANOS . 2005 . Molecular indica- L I , H. 1952 . Floristic relationships between eastern Asia and eastern tors of tree migration capacity under rapid climate change. Ecology North America. Transactions of the American Philosophical Society 86 : 2088 – 2098 . 42 : 371 – 429 . M CVAY , J . , A . H IPP , AND P. S. MANOS . 2014 . RAD-Seq-based phylogenet- L I , J. H. 2008 . Phylogeny of Catalpa (Bignoniaceae) inferred from se- ics of New World oaks (Quercus L.) . In Abstracts, 129, Evolution quences of chloroplast ndhF and nuclear ribosomal DNA. Journal of 2014 conference, Raleigh, North Carolina, USA. Available at http:// Systematics and Evolution 46 : 341 – 348 . evolution2014.org/wp-content/uploads/2014/04/ABSTRACTS-printing- L I , J. H. , D. E. BOUFFORD , AND M. J. DONOGHUE . 2001 . Phylogenetics of fi nal.pdf. Buckleya (Santalaceae) based on ITS sequences of nuclear ribosomal M ILLER , C. N. 1976 . Early evolution in the Pinaceae. Review of DNA. Rhodora 103 : 137 – 150 . Palaeobotany and Palynology 21 : 101 – 117 . L I , J. H. , S. SHOUP , AND Z . C HEN . 2007 . Phylogenetic relationships of dip- M ILNE , R. I. 2004 . Phylogeny and biogeography of Rhododendron sub- loid species of Betula (Betulaceae) inferred from DNA sequences of section Pontica , a group with a tertiary relict distribution. Molecular nuclear nitrate reductase. Systematic Botany 32 : 357 – 365 . Phylogenetics and Evolution 33 : 389 – 401 . L I , J. H. , P. D. TREDICI , S . Y ANG , AND M. J. DONOGHUE . 2002 . Phylogenetic M ILNE , R. I. 2006 . Northern Hemisphere plant disjunctions: A window relationships and biogeography of Stewartia (Camellioideae, on Tertiary land bridges and climate change? Annals of Botany 98 : Theaceae) inferred from nuclear ribosomal DNA ITS sequences. 465 – 472 . Rhodora 104 : 117 – 133 . M IRANDA , F . , AND A. J. SHARP. 1950 . Characteristics of the vegetation in L I , J. H. , AND Q. P. XIANG. 2005 . Phylogeny and Biogeography of Thuja certain temperate regions of Eastern Mexico. Ecology 31 : 313 – 333 . L. (Cupressaceae), and eastern Asian and North American disjunct M OORE , B. R. , AND M. J. DONOGHUE . 2007 . Correlates of diversifi cation in genus. Journal of Integrative Plant Biology 47 : 651 – 659 . the plant clade Dipsacales: Geographic movement and evolutionary L I , M . , J . L I , J . J IANG , C . F U , AND W . J UDD. 2009 . Phylogenetics and innovations. American Naturalist 170 : S28 – S55 . biogeography of Pieris (Lyonieae, Ericaceae) Inferred from se- M ORRIS , A. B. , S. ICKERT-BOND , D. B. BRUNSON , D. E. SOLTIS , AND P. S. quences of nuclear and chloroplast genomes. Systematic Botany S OLTIS . 2008 . Phylogeographical structure and temporal complex- 34 : 553 – 560 . ity in American sweetgum (Liquidambar styracifl ua ; Altingiaceae). L ITTLE , E. L. 1970 . Endemic, disjunct and northern trees in the Southern Molecular Ecology 17 : 3889 – 3900 . Appalachians In P. C. Holt and R. A. Paterson [eds.], The distribu- M U , X. Y. , L. C. ZHAO , AND Z. X. ZHANG . 2012 . Phylogeny of Celastrus tional history of the biota of the Southern Appalachians, part II, Flora, L. (Celastraceae) inferred from two nuclear and three plastid markers. 249–290. Research Division Monograph 2, Virginia Polytechnic Journal of Plant Research 125 : 619 – 630 . Institute, Blacksburg, Virginia, USA. N EPAL , M. P. , AND C. J. FERGUSON . 2012 . Phylogenetics of Morus L ITTLE , S. A. , S. W. KEMBEL , AND P . W ILF . 2010 . Paleotemperature proxies (Moraceae) Inferred from ITS and trnL-trnF sequence data. Systematic from leaf fossils reinterpreted in light of evolutionary history. PLoS Botany 37 : 442 – 450 . ONE 5 : e15161 . N IE , Z. L. , H. SUN , Z. D. CHEN , Y . M ENG , S. R. MANCHESTER , AND J . W EN . L IU , Z . , G . H AO , Y. B. LUO , L. B. THIEN , S. W. ROSSO , A. M. LU , AND Z. D. 2010 . Molecular phylogeny and biogeographic diversifi cation of C HEN. 2006 . Phylogeny and androecial evolution in Schisandraceae, Parthenocissus (Vitaceae) disjunct between Asia and North America. inferred from sequences of nuclear ribosomal DNA ITS and chloro- American Journal of Botany 97 : 1342 – 1353 . plast DNA trnL-F regions. International Journal of Plant Sciences N IE , Z. L. , H. SUN , Y . M ENG , AND J . W EN . 2009 . Phylogenetic analysis 167 : 539 – 550 . of Toxicodendron (Anacardiaceae) and its biogeographic implications L O , E. Y. Y. , AND M. J. DONOGHUE . 2012 . Expanded phylogenetic and on the evolution of north temperate and tropical intercontinental dis- dating analyses of the apples and their relatives (Pyreae, Rosaceae). junctions. Journal of Systematics and Evolution 47 : 416 – 430 . Molecular Phylogenetics and Evolution 63 : 230 – 243 . N IE , Z. L. , J. WEN , H . A ZUMA , Y. L. QIU , H . S UN , Y . M ENG , W . S UN , AND L O , E. Y. Y. , S. STEFANOVI Ć , K. I. CHRISTENSEN , AND T. A. DICKINSON . 2009 . E. A. ZIMMER. 2008 . Phylogenetic and biogeographic complexity of Evidence for genetic association between East Asian and western Magnoliaceae in the Northern Hemisphere inferred from three nuclear North American Crataegus L. (Rosaceae) and rapid divergence of the data sets. Molecular Phylogenetics and Evolution 48 : 1027 – 1040 . eastern North American lineages based on multiple DNA sequences. N IE , Z. L. , J. WEN , AND H . S UN . 2007 . Phylogeny and biogeography of Molecular Phylogenetics and Evolution 51 : 157 – 168 . Sassafras (Lauraceae) disjunct between eastern Asia and eastern L O , E. Y. Y. , S. STEFANOVI Ć , AND T. A. DICKINSON . 2007 . Molecular reap- North America. Plant Systematics and Evolution 267 : 191 – 203 . praisal of relationships between Crataegus and Mespilus (Rosaceae, N ÜRK , N. M. , S. MADRIÑ ÁN , M. A. CARINE , M. W. CHASE , AND F. R. Pyreae)—Two genera or one? Systematic Botany 32 : 596 – 616 . B LATTNER. 2013 . Molecular phylogenetics and morphological evo- M ANCHESTER , S. R. 1999 . Biogeographical relationships of North American lution of St. John’s wort (Hypericum ; Hypericaceae). Molecular Tertiary fl oras. Annals of the Missouri Botanical Garden 86 : 472 – Phylogenetics and Evolution 66 : 1 – 16 . 522 . O H , S. H. , L. CHEN , S. H. KIM , Y. D. KIM , AND H . S HIN . 2010 . Phylogenetic M ANOS , P. S. , P. S. SOLTIS , D. E. SOLTIS , S. R. MANCHESTER , S . O H , C. D. Relationship of Physocarpus insularis (Rosaceae) endemic on Ulleung B ELL , D. L. DILCHER , AND D. E. STONE . 2007 . Phylogeny of extant Island: Implications for conservation biology. Journal of Plant Biology and fossil Juglandaceae inferred from the integration of molecular and 53 : 94 – 105 . morphological data sets. Systematic Biology 56 : 412 – 430 . O H , S. H. , AND D . P OTTER . 2005 . Molecular phylogenetic systematics and M ANOS , P. S. , AND A. M. STANFORD . 2001 . The historical biogeography of biogeography of tribe Neillieae (Rosaceae) using DNA sequences of Fagaceae: Tracking the Tertiary history of temperate and subtropical cpDNA, rDNA, and LEAFY. American Journal of Botany 92 : 179 – 192 . forests of the Northern Hemisphere. International Journal of Plant O HI-TOMA , T . , T . S UGAWARA , H . M URATA , S . W ANKE , C . N EINHUIS , AND Sciences 162 : S77 – S93 . J . M URATA . 2006 . Molecular phylogeny of Aristolochia sensu lato M AO , K . , G . H AO , J . L IU , R. P. ADAMS , AND R. I. MILNE . 2010 . Diversifi cation (Aristolochiaceae) based on sequences of rbcL , matK , and phyA and biogeography of Juniperus (Cupressaceae): Variable diversifi ca- genes, with special reference to differentiation of chromosome num- tion rates and multiple intercontinental dispersals. New Phytologist bers. Systematic Botany 31 : 481 – 492 . 188 : 254 – 272 . P ARKS , M . , R . C RONN , AND A . L ISTON . 2012 . Separating the wheat from M ARTIN , W. H. , S. G. BOYCE , AND A. C. ECHTERNACHT [ EDS .]. 1993 . the chaff: mitigating the effects of noise in a plastome phylogenomic Biodiversity of the southeastern United States: Upland terrestrial data set from Pinus L. (Pinaceae). BMC Evolutionary Biology 12 : communities. John Wiley, New York, New York, USA. 100 – 117 . MANOS AND MEIRELES—WOODY PLANTS OF SOUTHERN APPALACHIANS • VOL. 102 , NO. 5 MAY 2015 • 803

P EARSE , I. S. , AND A. L. HIPP. 2009 . Phylogenetic and trait similarity to a S ELBACH-SCHNADELBACH , A. , S. S. CAVALLI , J . M ANEN , G. C. COELHO , AND native species predict herbivory on non-native oaks. Proceedings of T. T. D. SOUZA-CHIES . 2009 . New information for Ilex phylogenet- the National Academy of Sciences, USA 106 : 18097 – 18102 . ics based on the plastid psbA-trnH intergenic spacer (Aquifoliaceae). P ETERSON , B. J. 2013 . Phylogeography of eastern leatherwood (Dirca Botanical Journal of the Linnean Society 159 : 182 – 193 . palustris L.) resolved by chloroplast sequencing and microsatellite S HAW , J . , AND R. L. SMALL. 2005 . Chloroplast DNA phylogeny and phy- genotyping. Ph.D. dissertation, Iowa State University, Ames, Iowa, logeography of the North American plums (Prunus subgenus Prunus USA. section Prunocerasus , Rosaceae). American Journal of Botany 92 : P OON , W. S. , P. C. SHAW , M. P. SIMMONS , AND P. P. H. BUT . 2007 . 2011 – 2030 . Congruence of molecular, morphological, and biochemical profi les S IMMONS , S. L. 2002 . A molecular phylogenetic investigation of the in Rutaceae: A cladistic analysis of the subfamilies Rutoideae and Staphyleaceae (DC.) Lindl.: With implications for its taxonomy and Toddalioideae. Systematic Botany 32 : 837 – 846 . biogeography. Ph.D. dissertation, University of Texas at Austin, P RICE , J. P. , AND W. L. WAGNER . 2011 . A phylogenetic basis for species- Austin, Texas, USA. area relationships among three Pacifi c Island fl oras. American Journal S IMMONS , M. P. , J. J. CAPPA , R. H. ARCHER , A. J. FORD , D . E ICHSTEDT , AND of Botany 98 : 449 – 459 . C. C. CLEVINGER. 2008 . Phylogeny of the Celastreae (Celastraceae) Q IAN , H. 2002a . A comparison of the taxonomic richness of temperate and the relationships of Catha edulis (qat) inferred from morphologi- plants in East Asia and North America. American Journal of Botany cal characters and nuclear and plastid genes. Molecular Phylogenetics 89 : 1818 – 1825 . and Evolution 48 : 745 – 757 . Q IAN , H. 2002b . Floristic relationships between eastern Asia and North S IMON , S. A. , C. K. THOMAS , G. L. KAUFFMAN , W. H. MCNAB , AND C. J. America: Test of Gray’s hypothesis. American Naturalist 160 : 317 – U LREY. 2005 . Ecological zones in the Southern Appalachians: First 332 . approximation. Research Papers United States Forest Service, Southern R AN , J . , X . W EI , AND X . W ANG . 2006 . Molecular phylogeny and bioge- Research Station, Asheville, North Carolina, USA. ography of Picea (Pinaceae): Implications for phylogeographical S MEDMARK , J. E. E. , AND A. A. ANDERBERG . 2007 . Boreotropical migra- studies using cytoplasmic haplotypes. Molecular Phylogenetics and tion explains hybridization between geographically distant lineages in Evolution 41 : 405 – 419 . the pantropical clade Sideroxyleae (Sapotaceae). American Journal of R C ORE TEAM. 2013 . R: A language and environment for statistical com- Botany 94 : 1491 – 1505 . puting. R Foundation for Statistical Computing , Vienna, Austria. S MITH , R. L. , AND K. J. SYTSMA . 1990 . Evolution of Populus nigra (sect. R AVEN , P. H. , AND D. I. AXELROD. 1978 . Origin and relationships of the Aigeiros): Introgressive hybridization and the chloroplast contribu- California fl ora. University of California Publications in Botany 72 : tion of Populus alba (sect. Populus ). American Journal of Botany 77 : 1 – 34 . 1176 – 1187 . R EN , H . , L . L U , A . S OEJIMA , Q . L UKE , D . Z HANG , Z . C HEN , AND J . W EN . S MITH , S. A. , AND M. J. DONOGHUE . 2010 . Combining historical bioge- 2011 . Phylogenetic analysis of the grape family (Vitaceae) based on ography with niche modeling in the Caprifolium clade of Lonicera the noncoding plastid trnC-petN , trnH-psbA, and trnL-F sequences. (Caprifoliaceae, Dipsacales). Systematic Biology 59 : 322 – 341 . Taxon 60 : 629 – 637 . S NIDERMAN , J. M. , AND G. J. JORDAN . 2011 . Extent and timing of fl oris- R ENNER , S. S. , G. W. GRIMM , G. M. SCHNEEWEISS , T. F. STUESSY , AND R. E. tic exchange between Australian and Asian rain forests. Journal of R ICKLEFS . 2008 . Rooting and dating maples ( Acer ) with an uncorre- Biogeography 38 : 1445 – 1455 . lated-rates molecular clock: Implications for north American/Asian S OLTIS , D. E. , A. B. MORRIS , J. S. MCLACHLAN , P. S. MANOS , AND P. S. disjunctions. Systematic Biology 57 : 795 – 808 . S OLTIS. 2006 . Comparative phylogeography of unglaciated eastern R ICKLEFS , R. E. , AND R. E. LATHAM . 1992 . Intercontinental correlation North America. Molecular Ecology 15 : 4261 – 4293 . of geographical ranges suggests stasis in ecological traits of rel- S ORRIE , B. A. , AND A. S. WEAKLEY . 2001 . Coastal Plain vascular plant ict genera of temperate perennial herbs. American Naturalist 139 : endemics: Phytogeographic patterns. Castanea 66 : 50 – 82 . 1305 – 1321 . S PIRA , T. P. 2011 . Wildfl owers and plant communities of the southern R ONQUIST , F. 1997 . Dispersal–vicariance analysis: A new approach to Appalachian Mountains and Piedmont: A naturalist’s guide to the the quantifi cation of historical biogeography. Systematic Biology 46 : Carolinas, Virginia, Tennessee, and Georgia. University of North 195 – 203 . Carolina Press, Chapel Hill, North Carolina, USA. R UIZ-SANCHEZ , E . , AND J. F. O RNELAS. 2014 . Phylogeography of Liquidambar S TONE , D. E. , S. OH , E. A. TRIPP , L. E. RÍOS G , AND P. S. MANOS . 2009 . styracifl ua (Altingiaceae) in Mesoamerica: Survivors of a Neogene Natural history, distribution, phylogenetic relationships, and conserva- widespread temperate forest (or cloud forest) in North America? Ecology tion of Central American black walnuts (Juglans sect. Rhysocaryon ). and Evolution 4 : 311 – 328 . Journal of the Torrey Botanical Society 136 : 1 – 25 . S AMAIN , M . , S . W ANKE , AND P . G OETGHEBEUR . 2010 . Unraveling extensive T HORNE , J. L. , H. KISHINO , AND I. S. PAINTER . 1998 . Estimating the rate of paraphyly in the genus Hydrangea s. l. with implications for the sys- evolution of the rate of molecular evolution. Molecular Biology and tematics of tribe Hydrangeeae. Systematic Botany 35 : 593 – 600 . Evolution 15 : 1647 – 1657 . S ANDERSON , M. J. 2002 . Estimating absolute rates of molecular evolution T IFFNEY , B. H. 1985a . Perspectives on the origin of the fl oristic similar- and divergence times: A penalized likelihood approach. Molecular ity between eastern Asia and eastern North America. Journal of the Biology and Evolution 19 : 101 – 109 . Arnold Arboretum 66 : 73 – 94 . S ANMARTÍN , I . , AND F . R ONQUIST . 2004 . Southern Hemisphere biogeog- T IFFNEY , B. H. 1985b . The Eocene North Atlantic land bridge: Its im- raphy inferred by event-based models: Plant versus animal patterns. portance in tertiary and modern phytogeography of the Northern Systematic Biology 53 : 216 – 243 . Hemisphere. Journal of the Arnold Arboretum 66 : 243 – 273 . S ATTARIAN , A. 2006 . Contribution to the biosystematics of Celtis L. T IFFNEY , B. H. , AND S. R. MANCHESTER . 2001 . The use of geological and (Celtidaceae) with special emphasis on the African species. Ph.D. dis- paleontological evidence in evaluating plant phylogeographic hy- sertation, Wageningen Universiteit, Wageningen, Netherlands. potheses in the Northern Hemisphere Tertiary. International Journal S CHNABEL , A. , P. E. MCDONEL , AND J. F. WENDEL . 2003 . Phylogenetic of Plant Sciences 162 : S3 – S17 . relationships in Gleditsia (Leguminosae) based on ITS sequences. T RIPLETT , J. K. , A. S. WEAKLEY , AND L. G. CLARK . 2006 . Hill cane American Journal of Botany 90 : 310 – 320 . ( Arundinaria appalachiana ), a new species of bamboo (Poaceae: S CHRADER , J. A. , AND W. R. GRAVES . 2004 . Systematics of Dirca (Thyme- Bambusoideae) from the southern Appalachian Mountains. SIDA, laeaceae) based on ITS sequences and ISSR polymorphisms. Sida Contributions to Botany 22 : 79 – 95 . 21 : 511 – 524 . T RÖNDLE , D . , S . S CHRÖDER , H. H. KASSEMEYER , C . K IEFER , M. A. KOCH , S CHULTHEIS , L. M. , AND M. J. DONOGHUE . 2004 . Molecular phylogeny and AND P . N ICK . 2010 . Molecular phylogeny of the genus Vitis biogeography of Ribes (Grossulariaceae), with an emphasis on goose- (Vitaceae) based on plastid markers. American Journal of Botany berries (subg. Grossularia ). Systematic Botany 29 : 77 – 96 . 97 : 1168 – 1178 . 804 • VOL. 102 , NO. 5 MAY 2015 • AMERICAN JOURNAL OF BOTANY

T SAI , Y. H. , AND P. S. MANOS. 2010 . Host density drives the postglacial W OLFE, J. A. 1993 . A method of obtaining climatic parameters from leaf migration of the tree parasite, Epifagus virginiana . Proceedings of the assemblages. U.S. Geological Survey Bulletin 2040. U.S. Government National Academy of Sciences, USA 107 : 17035 – 17040 . Printing Offi ce, Washington D.C., USA. USDA, NRCS. 2013 . The PLANTS Database. National Plant Data Team, W OLFE , J. A. 1994 . An analysis of Neogene climates in Beringia. Greensboro, North Carolina, USA. Website http://plants.usda.gov Palaeogeography, Palaeoclimatology, Palaeoecology 108 : 207 – 216 . [accessed 18 September 2013]. W OOD , C. E. 1970 . Some fl oristic relationships between the southern US EPA [U. S. ENVIRONMENTAL PROTECTION AGENCY] . 2013 . Level III Appalachians and western North America. In P. C. Holt and R. A. Ecoregions of the Continental United States [online map]. National Paterson [eds.], The distributional history of the biota of the southern Health and Environmental Effects Research Laboratory, U. S. Appalachians, part II, Flora, 331–404. Research Division Monograph Environmental Protection Agency. Western Ecological Division, 2, Virginia Polytechnic Institute, Blacksburg, Virginia, USA. Corvallis, Oregon, USA. Available at ftp://ftp.epa.gov/wed/ecoregions/ W OOD , C. E. 1972 . Morphology and phytogeography: The classical ap- us/Eco_Level_III_US.pdf . proach to the study of disjunctions. Annals of the Missouri Botanical V ASILEF , I. 2013 . QtiPlot—Data analysis and scientifi c visualization , Garden 59 : 107 – 124 . Universiteit Utrecht, Utrecht, Netherlands. W U , C. 1979 . The regionalization of Chinese Flora . Acta Botanica W ANG , Y. , P. W. FRITSCH , S . S HI , F . A LMEDA , B. C. CRUZ , AND L. M. KELLY . Yunnanica 1 : 1 – 22 [In Chinese with English abstract] . 2004 . Phylogeny and infrageneric classifi cation of Symplocos X IANG , Q. P. , J. Q. XIANG , Y . G UO , AND X . Z HANG . 2009 . Phylogeny (Symplocaceae) Inferred from DNA sequence data. American Journal of Abies (Pinaceae) inferred from nrITS sequence data. Taxon 58 : of Botany 91 : 1901 – 1914 . 141 – 152 . W EAKLEY, A. S. 2005 . Change over time in our understanding of the fl ora X IANG , Q.-Y. , D. E. SOLTIS , P. S. SOLTIS , S. R. MANCHESTER , AND D. J. of the southeastern United States: Implications for plant systematics, C RAWFORD. 2000 . Timing the Eastern Asian–Eastern North American bioinformatics and conservation. Ph.D. dissertation, Duke University, fl oristic disjunction: Molecular clock corroborates paleontological es- Durham, North Carolina, USA. timates. Molecular Phylogenetics and Evolution 15 : 462 – 472 . W EAKLEY , A. S. 2012 . Flora of the Southern and mid-Atlantic states, X IANG , Q.-Y. , AND D. T. THOMAS . 2008 . Tracking character evolution and working draft of November 2012. University of North Carolina biogeographic history through time in Cornaceae—Does choice of Herbarium, North Carolina Botanical Garden, Chapel Hill, North methods matter? Journal of Systematics and Evolution 46 : 349 – 374 . Carolina, USA. X IANG , Q.-Y. , D. T. THOMAS , AND Q. P. XIANG . 2011 . Resolving and dat- W EN , J. 1999 . Evolution of eastern Asian and eastern North American ing the phylogeny of Cornales—Effects of taxon sampling, data parti- disjunct distributions in fl owering plants. Annual Review of Ecology tions, and fossil calibrations. Molecular Phylogenetics and Evolution and Systematics 30 : 421 – 455 . 59 : 123 – 138 . W EN , J. 2011 . Systematics and biogeography of Aralia L. (Araliaceae): X IANG , Q.-Y. , D. T. THOMAS , W. H. ZHANG , S. R. MANCHESTER , AND Z . Revision of Aralia sects. Aralia, Humiles, Nanae, and Sciadodendron. M URRELL. 2006 . Species level phylogeny of the genus Cornus Contributions from the United States National Herbarium 57 : 1 – 172 . (Cornaceae) based on molecular and morphological evidence— W EN , J. , S. T. BERGGREN , C . L EE , S . I CKERT-BOND , T . Y I , K . Y OO , L . X IE , Implications for taxonomy and Tertiary intercontinental migration. J . S HAW , AND D . P OTTER . 2008 . Phylogenetic inferences in Prunus Taxon 55 : 9 – 30 . (Rosaceae) using chloroplast ndhF and nuclear ribosomal ITS se- X IANG , Q.-Y . , W. H. ZHANG , R. E. RICKLEFS , H. QIAN , Z. D. CHEN , J. WEN , quences. Journal of Systematics and Evolution 46 : 322 – 332 . AND J. H. LI . 2004 . Regional differences in rates of plant speciation W EN , J . , S . I CKERT-BOND , Z . N IE , AND R . L I . 2010 . Timing and modes of and molecular evolution: A comparison between eastern Asia and evolution of Eastern Asian–North American biogeographic disjunc- eastern North America. Evolution 58 : 2175 – 2184 . tions in seed plants. In M. Long, H. Gu, and Z. Zhou [eds].), Darwin’s X IE , L . , T . Y I , R . L I , D . L I , AND J . W EN. 2010 . Evolution and bio- heritage today: Proceedings of the Darwin 200 Beijing International geographic diversification of the witch-hazel genus (Hamamelis Conference, 252–269. Higher Education Press, Beijing, China. L., Hamamelidaceae) in the Northern Hemisphere. Molecular Phylo- W EN , J . , AND S. M. ICKERT-BOND . 2009 . Evolution of the Madrean- genetics and Evolution 56 : 675 – 689 . Tethyan disjunctions and the North and South American amphitropi- Y I , T. , A. J. MILLER , AND J . W EN . 2004 . Phylogenetic and biogeographic cal disjunctions in plants. Journal of Systematics and Evolution 47 : diversifi cation of Rhus (Anacardiaceae) in the Northern Hemisphere. 331 – 348 . Molecular Phylogenetics and Evolution 33 : 861 – 879 . W HITCHER , I. N. , AND J . W EN . 2001 . Phylogeny and biogeography of Y I , T. , A. J. MILLER , AND J . W EN . 2007 . Phylogeny of Rhus (Anacardiaceae) Corylus (Betulaceae): Inferences from ITS sequences. Systematic based on sequences of nuclear Nia-i3 intron and chloroplast trnC- Botany 26 : 283 – 298 . trnD . Systematic Botany 32 : 379 – 391 . W HITE , P. S. 1983 . Eastern Asian–North American fl oristic relations: The Y ODER , A. D. , AND M. D. NOWAK . 2006 . Has vicariance or dispersal been plant community level. Annals of the Missouri Botanical Garden 70 : the predominant biogeographic force in Madagascar? Only time 734 – 747 . will tell. Annual Review of Ecology, Evolution and Systematics 37 : W HITTAKER , R. H. 1956 . Vegetation of the Great Smoky Mountains . 405 – 431 . Ecological Monographs 26 : 1 – 80 . Y OO , K. O. , AND J . W EN. 2002 . Phylogeny and biogeography of Carpinus W IEGREFE , S. J. , K. J. SYTSMA , AND R. P. GURIES . 1994 . Phylogeny of elms and subfamily Coryloideae (Betulaceae). International Journal of ( Ulmus , Ulmaceae): Molecular evidence for a sectional classifi cation. Plant Sciences 163 : 641 – 650 . Systematic Botany 19 : 590 – 612 . Z ACHOS , J . , M . P AGANI , L . S LOAN , E . T HOMAS , AND K . B ILLUPS . 2001 . W INKWORTH , R. C. , AND M. J. DONOGHUE . 2005 . Viburnum phylogeny Trends, rhythms, and aberrations in global climate 65 Ma to present. based on combined molecular data: Implications for taxonomy and Science 292 : 686 – 693 . biogeography. American Journal of Botany 92 : 653 – 666 . Z HANG , J. B. , R. Q. LI , X. G. XIANG , S. R. MANCHESTER , L . L IN , W . W ANG , W ISSEMANN , V . , AND C. M. RITZ . 2005 . The genus Rosa (Rosoideae, J . W EN , AND Z. D. CHEN. 2013 . Integrated fossil and molecular data Rosaceae) revisited: Molecular analysis of nrITS-1 and atpB -rbcL reveal the biogeographic diversifi cation of the eastern Asian–eastern intergenic spacer (IGS) versus conventional taxonomy. Botanical North American disjunct hickory genus (Carya Nutt.). PLOS ONE 8 : Journal of the Linnean Society 147 : 275 – 290 . e70449 . W OLFE , J. A. 1975 . Some aspects of plant geography of the Northern Z HOU , S. , S. S. RENNER , AND J . W EN . 2006 . Molecular phylogeny and in- Hemisphere during the late Cretaceous and Tertiary. Annals of the tra- and intercontinental biogeography of Calycanthaceae. Molecular Missouri Botanical Garden 62 : 264 – 279 . Phylogenetics and Evolution 39 : 1 – 15 .