American Journal of Botany 91(6): 943±952. 2004. NORTHERN HEMISPHERE BIOGEOGRAPHY OF CERASTIUM (CARYOPHYLLACEAE): INSIGHTS FROM PHYLOGENETIC ANALYSIS OF NONCODING PLASTID NUCLEOTIDE SEQUENCES1 ANNE-CATHRINE SCHEEN,2,6 CHRISTIAN BROCHMANN,3 ANNE K. BRYSTING,3 REIDAR ELVEN,3 ASHLEY MORRIS,4 DOUGLAS E. SOLTIS,4 PAMELA S. SOLTIS,5 AND VICTOR A. ALBERT2 2Botanical Garden, Natural History Museums and Botanical Garden, University of Oslo, P.O. Box 1172 Blindern, N-0318 Oslo, Norway; 3National Centre for Biosystematics, Natural History Museums and Botanical Garden, University of Oslo, P.O. Box 1172 Blindern, N-0318 Oslo, Norway; 4Department of Botany, University of Florida, P.O. Box 118526, Gainesville, Florida 32611-5826 USA; and 5Florida Museum of Natural History, P.O. Box 117800, University of Florida, Gainesville, Florida 32611-7800 USA Phylogenetic relationships and biogeography of the genus Cerastium were studied using sequences of three noncoding plastid DNA regions (trnL intron, trnL-trnF spacer, and psbA-trnH spacer). A total of 57 Cerastium taxa was analyzed using two species of the putative sister genus Stellaria as outgroups. Maximum parsimony analyses identi®ed four clades that largely corresponded to previously recognized infrageneric groups. The results suggest an Old World origin and at least two migration events into North America from the Old World. The ®rst event possibly took place across the Bering land bridge during the Miocene. Subsequent colonization of South America occurred after the North and South American continents joined during the Pliocene. A more recent migration event into North America probably across the northern Atlantic took place during the Quaternary, resulting in the current circumpolar distribution of the Arctic species. Molecular clock dating of major biogeographic events was internally consistent on the phylogenetic trees. The arctic high-polyploid species form a polytomy together with some boreal and temperate species of the C. tomentosum group and the C. arvense group. Lack of genetic variation among the arctic species probably indicates a recent origin. The annual life form is shown to be of polyphyletic origin. Key words: biogeography; Cerastium; cpDNA; molecular phylogeny; Northern Hemisphere; psbA-trnH; support weighting; trnL- trnF. The historical biogeography of the Northern Hemisphere are mostly evergreen or deciduous angiosperms adapted to a has long been of interest to botanists and zoologists. Evidence warm temperate, subtropical, or tropical climate (e.g., Paeon- for different biogeographic patterns has increased in recent ia, Sang et al., 1997; Quercus, Manos et al., 1999; Arbutus, years with the increasing number of molecular phylogenetic Hilman et al., 2001; Lycium, Tatsuya et al., 2001; but see studies being published. A recent synthesis of non-marine an- Lindqvist and Albert, 2002). Thus far, little is known about imal phylogenies, for example, suggests that the Northern the biogeographic history of more cold-adapted Northern Hemisphere has a complex biogeographic history involving Hemisphere taxa. vicariance events that span time intervals from the Cretaceous Two major migration routes have been discussed with re- onwards (Sanmartin et al., 2001). gard to the historical biogeography of the Northern Hemi- Biogeographic studies of the Northern Hemisphere ¯oras sphere ¯oras: the North Atlantic land bridge (NALB) and the have generally focused on taxa with restricted and disjunct Bering land bridge (BLB). The NALB is thought to have been distributions that are thought to have been broader in the past most important during the early Tertiary, connecting the ¯oras (i.e., the eastern Asian±eastern North American disjunctions, of North America and southern Europe via Scotland and e.g., Wen, 1999; Donoghue et al., 2001; Xiang and Soltis, southern Greenland, or via northern Greenland and Fennos- 2001; Milne and Abbott, 2002). Those few taxa studied that candia (Tiffney, 1985). It is generally accepted that the south- have wider distributions throughout the Northern Hemisphere ernmost connection was severed during the early Eocene, ap- proximately 50 million years ago (Mya), when the climate was 1 Manuscript received 17 July 2003; revision accepted 30 January 2004. subtropical at these latitudes (Tiffney, 1985; Tiffney and Man- The authors thank the curators at the ILLS, MO, O, W, WU herbaria, Clive chester, 2001). The northernmost connection is thought to have Stace, and Danny Gustafson for providing material, Matthew Gitzendanner persisted until ca. 40 Mya (Milne and Abbott, 2002). This and Charlotte Lindqvist for help in the lab, and Steve Farris for permission to use the XAC parsimony jackkni®ng application and the ZAC support second connection may have been less important because Fen- weighting application. Most of the laboratory work was performed at the noscandia was separated by seaways from southern Europe as Soltis laboratory, University of Florida, and some sequencing was performed well as from Asia at the time (Tiffney, 1985). at the MegaBACE-laboratory, University of Oslo. This project was initiated The connection between Asia and North America via the and supported in part by a Fulbright Fellowship to A. C. S. Additional funding Bering land bridge persisted throughout most of the Tertiary was provided by the National Centre for Biosystematics at The Natural His- tory Museums and Botanical Garden, University of Oslo, and by the Univer- and was severed at approximately 5.5±4.8 Mya (Marincovich sity of Florida Research Foundation. and Gladenkov, 1999). It is thought that the BLB was most 6 E-mail: [email protected]. important for intercontinental migration after the NALB dis- 943 944 AMERICAN JOURNAL OF BOTANY [Vol. 91 appeared (Tiffney, 1985; Tiffney and Manchester, 2001) and Polymerase chain reactions (PCR) were performed in volumes of 25 mL that it was more important for deciduous angiosperms than for containing 13 PCR buffer (0.05 mol/L KCl, 0.02 mol/L Tris [pH 8.0], 1.5 m evergreen angiosperms because the high latitude of the region mmol/L MgCl2, 0.001% Tween), 200 mol/L of each dNTP (USB, Cleveland, would have served as a ®lter to the migration of evergreen Ohio, USA), 0.5 mmol/L of each primer, 5% DMSO, and 1.25 units Taq DNA angiosperms due to winter darkness (Tiffney and Manchester, Polymerase (Eppendorf, Hamburg, Germany), and 2.5 mL of unquanti®ed 2001; Milne and Abbott, 2002). The BLB reappeared during genomic DNA. In a few cases, PCR reactions were performed in volumes of the Quaternary glacial epochs (Elias et al., 1996). During the 25 mL using the AmpliTaq DNA polymerase buffer II kit (Applied Biosys- Quaternary glaciations, migration across the BLB was only tems, Foster City, California, USA) containing 0.2 mmol/L of each dNTP, 0.04% bovine serum albumen (BSA), 0.01 mmol/L tetramethylammonium possible for taxa of arctic or boreal af®nity (HulteÂn, 1937; m m Murray, 1981). The Bering area was an important refugium chloride (TMACl), 0.8 mol/L of each primer, and 2 L unquanti®ed geno- for the arctic ¯ora during the Quaternary glaciations (e.g., Hul- mic DNA. Ampli®cations were performed in a Biometra T3 thermal cycler (Whatman Biometra Biomedizinishe Analytik GmbH, GoÈttingen, Germany) teÂn, 1937; Abbott et al., 2000; Abbott and Brochmann, 2003), or in a Gene Amp PCR System 9700 (Applied Biosystems) using a program and the BLB must have in¯uenced the biogeography of the consisting of 4 min at 958C followed by six cycles of 20 s denaturation (958C), current circumpolar ¯ora. 1 min annealing (starting at 588C; temperature decreasing by 18C per cycle) Divergence time estimates may become the most reliable and 1 min extension (728C), and 35 (in a few cases only 25) cycles of 20 s method for distinguishing between migration via the BLB and denaturation (958C), 1 min annealing (528C) and 1 min extension (728C), via the NALB (Donoghue et al., 2001). Even when an as- ending with a ®nal 5 min extension (728C). Successful PCR reactions were sumption of a molecular clock cannot be calibrated with fossil puri®ed with exonuclease I and shrimp alkaline phosphatase (ExoSAP). data, age estimates of divergence times based on well-studied Cycle sequencing, using the same primers as in the PCR, was performed geological events are better than none at all. Furthermore, in- with CEQ 2000 Dye Terminator Cycle Sequencing kit (Beckman Coulter, ternal consistency of several divergence time estimates within Fullerton, California, USA) in quarter reactions (i.e., 2 mL DTCS Quick Start the same molecular phylogeny, i.e., several independent geo- Master Mix, 3.2 pmol primer and 1 mL cleaned PCR product), using the logically correlated migration/vicariance events, provide self- program suggested by the manufacturer on an Eppendorf Mastercycler (Brink- corroboration even in the absence of a strict molecular clock. mann Instruments, Westbury, New York, USA). Sequenced products were The genus Cerastium L. (Caryophyllaceae) can be used for precipitated in ethanol and sodium acetate to remove excess dye terminators testing different hypotheses of biogeographic relationships in before running them on a CEQ 8000 Sequencer (Beckman Coulter, Fullerton, the Northern Hemisphere. Cerastium is a genus of approxi- California, USA). In a few cases, cycle-sequencing reactions were prepared mately 100 perennial or annual, herbaceous, or rarely slightly with DYEamic ET dye terminator kit for MegaBACE (Amersham Bioscienc- woody species (e.g., Jalas et al., 1993; Dequan and Morton, es, Piscataway, New Jersey, USA) following the manufacturer's recommen- 2001), with an almost cosmopolitan distribution. The genus is dations and were run on a MJ Dyad 96-block thermal cycler. In these cases, most abundant in temperate and cold regions, especially at the products were puri®ed using Sephadex and run on a MegaBACE 500 (Amersham Biosciences, Piscataway, New Jersey, USA). high elevations, with a center of diversity in Eurasia (Dequan and Morton, 2001). A few species are cosmopolitan weeds Alignment and indel codingÐSequences were assembled and edited using (e.g., C.