Evolution of Leaf Anatomy and Photosynthetic Pathways in Portulacaceae Author(S): Gilberto Ocampo, Nuria K

Evolution of leaf anatomy and photosynthetic pathways in Portulacaceae Author(s): Gilberto Ocampo, Nuria K. Koteyeva, Elena V. Voznesenskaya, Gerald E. Edwards, Tammy L. Sage, Rowan F. Sage and J. Travis Columbus Source: American Journal of Botany , December 2013, Vol. 100, No. 12 (December 2013), pp. 2388-2402 Published by: Wiley Stable URL: https://www.jstor.org/stable/23596763

REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/23596763?seq=1&cid=pdf- reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

Wiley is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Botany

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms ■ A"rr" AN )l 1 "NAL American Journal of Botany 100(12): 2388-2402. 2013.

Evolution of leaf anatomy and photosynthetic PATHWAYS IN PORTULACACEAE1

Gilberto Ocampo2-6-7, Nuria K. Koteyeva3, Elena V. Voznesenskaya3, Gerald E. Edwards4, Tammy L. Sage5, Rowan F. Sage5, and J. Travis Columbus2

2Rancho Santa Ana Botanic Garden and Claremont Graduate University, 1500 North College Avenue, Claremont, California 91711 USA; laboratory of Anatomy and Morphology, V.L. Komarov Botanical Institute of the Russian Academy of Sciences, Prof. Popov Street 2 197376, St. Petersburg, Russia; 4School of Biological Sciences, Washington State University, Pullman, Washington 99164 USA; department of Ecology and Evolutionary Biology, University of Toronto, 25 Willcocks Street, Toronto, Ontario M5S 3B2 Canada

Premise of the study: Portulacaceae is a family with a remarkable diversity in photosynthetic pathways. This lineage not only has species with different C4 biochemistry (NADP-ME and NAD-ME types) and C3-C4 intermediacy, but also displays different leaf ana tomical configurations. Here we addressed the evolutionary history of leaf anatomy and photosynthetic pathways in Portulacaceae. Methods: Photosynthetic pathways were assessed based on leaf anatomy and carbon isotope ratios. Information on the NA DP-ME and NAD-ME C4 variants was obtained from the literature. The evolutionary relationships and trait evolution were estimated under a Bayesian framework, and divergence times were calibrated using the ages obtained in a previous study. Key results: C4 photosynthesis is the main pathway in Portulacaceae. One clade (Cryptopetala), however, includes species that have non-Kranz anatomy and C3 type isotope values, two of which are C3-C4 intermediates. The ancestral leaf anatomy for the family is uncertain. The analysis showed one origin of the C4 pathway, which was lost in the Cryptopetala clade. Nevertheless, when a second analysis was performed taking into account the limited number of species with NAD-ME and NADP-ME data, a secondary gain of the C4 pathway from a C3-C4 intermediate was inferred. Conclusions: The C4 pathway evolved ca. 23 Myr in the Portulacaceae. The number of times that the pathway evolved in the family is uncertain. The diversity of leaf anatomical types and C4 biochemical variants suggest multiple independent origins of C4 photosynthesis. Evidence for a switch from C4 to C3-C4 intermediacy supports the hypothesis that intermediates represent a distinct successful strategy.

Key words: C3-C4 intermediacy; C4 photosynthesis; Cactineae; Kranz anatomy; leaf anatomy; Portulaca; Portulacaceae; Portulacineae.

C4 photosynthesis is documented as one of the best examples of flowering plants (Sage et al., 2011), likely as a response to of convergent evolution. This photosynthesis pathway is esti the decreasing levels of atmospheric C02 during the Oligocene mated to have evolved independently in more than 60 lineages (Ehleringer et al., 1997; Sage, 2004; Besnard et al., 2009; Christin et al., 2011a; Kadereit et al., 2012). To increase intracellular C02 1 Manuscript received 7 March 2013; revision accepted 28 August 2013. around Rubisco, most of these lineages have developed a set of We are grateful to Lucinda McDade and Wendy Applequist for reviewing earlier versions of this manuscript, two anonymous reviewers for providing leaf anatomical features, called Kranz anatomy as a whole, that helpful comments on the manuscript, James André, Jennifer Cruse-Sanders, perform C4 photosynthesis (Gutiérrez et al., 1974; Hatch, 1987; Urs Eggli, Patricia Jaramillo, James Matthews, David Orr, Ernesto Sandoval, Dengler and Nelson, 1999). In plants with Kranz anatomy, atmo and John Trager for providing samples for this study, Héctor Osorio at spheric the C02 capture is accomplished in the elongated mesophyll Museo Nacional de Historia Natural (Uruguay), Bryan Simon at the Queensland cells (M; Figs. 1A-C) with the formation of C4 acids (malate or Herbarium, Femando Zuloaga at the Instituto de Botánica Darwinion, aspartate) the that are transported to the bundle sheath (BS) cells sur Missouri Botanical Garden, the National Plant Germplasm System (USA), roundingand the vascular bundles (VB). Subsequently, C02 is released the University of Arizona Herbarium for access to herbarium collections, James within the BS by decarboxylation of C4 acids, thus increasing the André, Stephen Dreher, Amanda Ingram, Tasha LaDoux, Oscar Morales, Sarah C02 concentration and minimizing the oxygenation of ribulose Siedschlag, and Valerie Sosa for their companionship and assistance during fieldwork. This study was supported by Rancho Santa Ana Botanic Garden, 1,5-bisphosphate by the enzyme Rubisco (Edwards and Walker, the Cactus and Succulent Society of America, Claremont Graduate University, 1983; Hatch, 1987; Kanai and Edwards, 1999). This strategy is the Claremont University Club, and The Community Foundation serving particularly advantageous in dry, warm, high-light intensity envi the Riverside and San Bernardino Counties. Financial support to G. O. wasronments (Ehleringer and Monson, 1993; Guralnick et al., 2002), provided by Rancho Santa Ana Botanic Garden, The Fletcher Jones Foundation, where photosynthesis can be inhibited by up to 33% due to the ef Comisión Nacional de Ciencia y Tecnología (Mexico), Fundación Prywer fects of photorespiration (Ehleringer and Monson, 1993). (Mexico), and the Instituto de Ecología, A. C. (Mexico). G. E. acknowledges It is not well understood how the C4 pathway evolved, al support by the National Science Foundation grant MCB 1146928. though there are some models that propose that it originated 6Corresponding author (e-mail: [email protected]); fax: 1-415 379-5737 from C3 plants through a series of gradual biochemical and ana 7 Present address: Department of Botany, California Academy of Sciences, tomical modifications (Kennedy and Laetsch, 1974; Monson 55 Music Concourse Drive, San Francisco CA 94118 USA et al., 1984; Sage, 2004; Sage et al., 2012). Plants with transitional characteristics between the C3 and C4 pathways are known as doi:10.3732/ajb. 1300094 C3-C4 intermediates (Ehleringer and Monson, 1993; Monson et al.,

—r

IBS iww. ^ / *

* 'TBS r- V # t ? M '■ »;; • *u ji *, M w y \ 7 A ;£ ^ HI ' • - A* */ j'■ H'/ ¿L '\wA*j-•• ' ? - _1 : • -'A

Fig. 1. Light micrographs of transectional leaf anatomy showing Kranz and C3-C4 features in Portulacaceae. Clade affiliation (Ocampo and Columbus, 2012) and anatomical type (Voznesenskaya et al., 2010) within parentheses. (A) Portulaca olerácea subsp. nítida (Olerácea clade; Atriplicoid). (B) P. hali moides (Pilosa clade; Pilosoid). (C) P. cf. bicolor (OL clade; Portulacelloid). (D) P. cryptopetala (Cryptopetala clade; C3-C4). BS = bundle sheath; M = mesophyll cells; WS = water storage cells. Scale bar = 0.1 mm.

1984) and have been detected in 21 plant lineages (Sage et al., photosynthetic variants (Guralnick and Jackson, 2001; Ocampo 2011). In comparison with C3 plants, photosynthesis in the and Columbus, 2010). Previous reports suggested that C4 C3-C4 intermediates is more effective under CÓ2 limiting con- photosynthesis may have evolved twice within the suborder, ditions because they have an efficient system for recapturing particularly in Anacampserotaceae and Portulacaceae s.s. (treated photorespired C02, and some intermediates have a partially as the traditional Portulacaceae in Guralnick and Jackson, 2001); functioning C4 cycle (Edwards and Ku, 1987; Monson and however, it has been shown that the C4 strategy is only present in Rawsthorne, 2000; Sage, 2004). In addition, compared to C3 Portulacaceae s.s. (i.e., limited to Portulaca species; Guralnick plants, intermediates have certain anatomical and structural et al., 2008; Ocampo and Columbus, 2010). Recently it was dis modifications including reduction in the M/BS cell ratio, in- covered that the family not only has members with Kranz anat creased numbers of organelles in BS cells, and in some genera omy and C4 photosynthesis, but also has at least one species increased vein density (Brown and Hattersley, 1989; Monson which is a C3-C4 intermediate (P. cryptopetala Speg.; Fig. ID) and Moore, 1989; Rawsthorne, 1992; McKown and Dengler, based on structural and functional features (Voznesenskaya 2007; Voznesenskaya et al., 2013). The C3-C4 intermediates et al., 2010), a situation that has been interpreted as a loss of the have played an important role in investigations on the origins of C4 condition (Ocampo and Columbus, 2010). C4 photosynthesis; also, a number of studies provide evidence Portulacaceae s.s. (hereafter simply referred to as Portu that, in some cases, it should be considered as a distinct strat- lacaceae; Nyffeler and Eggli, 2010) has ca. 100 species distributed egy, particularly in warm and hot environments (Monson, worldwide, mainly in tropical and subtropical regions (Legrand, 1999; Christin et al., 201 lb; Sage et al., 2012). 1953; Matthews, 2003). Phylogenetic analyses have shown that The suborder Cactineae (Portulacineae) of the Caryophylla- the family has two major lineages (Ocampo and Columbus, les (Thome and Reveal, 2007) is a group rich in species with 2010; 2012). One clade contains species with opposite leaves

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2390 American Journal of Botany [Vol. 100

(OL clade) that are distributed almost exclusively in the Old 1993; Ocampo and Columbus, 2010; Voznesenskaya et al., World (P. quadrifida L. is a pantropical weed that has reached 2010). the Caribbean region), whereas the other comprises alternate- In this investigation, we surveyed photosynthetic pathways leaved taxa (AL clade) with representatives on almost every inferred by stable carbon isotope ratios (813C) and leaf anatomy continent. Within the AL clade, four major subclades can be of 70 taxa of Portulacaceae, including a number of subspecies distinguished: the Olerácea clade, whose species have incon- and cultivars. The increased taxon sample allowed us to exam spicuous hairs in the leaf axils and almost always have pseudo- ine the diversity of these two traits and resulted in the discovery opposite leaves (e.g., the nonmonophyletic P. olerácea L.); P. of a clade consisting of three members that are not C4 species cryptopetala, with axillary hairs (Ocampo and Columbus, 2010) (including the C3-C4 intermediate P. cryptopetala). Using the and C3-C4 intermediate type anatomy with Kranz-like BS cells molecular data and divergence times estimates provided by (Voznesenskaya et al., 2010), an atypical condition in this oth- Ocampo and Columbus (2012), we explored the diversification erwise C4 genus; the Umbraticola clade, with taxa having in- patterns of leaf anatomy types and photosynthetic pathways, conspicuous leaf axillary hairs and a wing around the dehiscence which allowed us to obtain insights into the evolutionary his line of the capsule (P. umbraticola Kunth); and the Pilosa clade, tory of C4 photosynthesis in Portulacaceae. with members that almost always have conspicuous leaf axil lary hairs (e.g., P. pilosa L.). Relationships among the AL sub- MATERIALS AND METHODS clades are strongly supported, where the Olerácea clade and P. cryptopetala are sisters to the Umbraticola and Pilosa clades 0 . , , . „ , • ^ ' Sampling—The samples used are a subset of those from Ocampo and (Ocampo and Columbus, 2012). Columbus (2012), and represent all the major clades identified in that study Biochemical studies have shown that Portulacaceae has spe- (Appendix 1). New stable carbon isotope ratio (813C) and leaf anatomy data cies with either NAD-malic enzyme (NAD-ME) or NADP- were generated, except for nine samples and the outgroup taxa which were malic enzyme (NADP-ME) type C4 cycles (e.g., Gutiérrez taken from Ocampo and Columbus (2010) and Voznesenskaya et al. (2010). et al.. 1974; Ku et al., 1981; Voznesenskaya et al., 2010), although Herbarium samples from additional species lacking DNA sequence data and

the. , . ,NAD-ME r» i rrn, bletype non-C4 seems species to be(Appendix restricted \). Pereskiaaculeata to the species Mill. that (Cactaceae), are îff Talinop ana*my Preparations were subjected to 8'3C analysis to detect any possi closely related to P. olerácea (Voznesenskaya et al., 2010). The sisfrutescens A. Gray (Anacampserotaceae), and Talinumpaniculatum (Jacq.) presence of these two biochemical variants in the same family Gaertn. (Talinaceae) were designated as outgroups. is somewhat unusual, as it is only known to occur in six other angiosperm families (Sage et al., 2011 ). It also has been shown Sample preparation and data acquisition—813C data—Leaf samples were that Kranz leaf anatomy in Portulaca is diverse, mainly with taken from silica gel-dried material or herbarium specimens, using multiple respect to the arrangement of the VB and the position of the samples of the same taxon when available (Appendix 1). Samples were ana water Storage tissue (WS) within the leaf blade (Welkie and Xyztà with a PDZ EuroPa ANCA-GSL elemental analyzer interfaced to a PDZ iCaldwell, c\nc\ T- .i 1 c\or\1970; XT Carolin irvoo r^- it— et i al.,i at the1978; University Prabhakar of California and at DavisRamayya, Stable Isotope EuroPa Facility, 20"20or with aratio GV mass spectrometer (Sercon limited, Cheshire, UK) 1979, Lin et al., 1982, Nyananyo, 1988, Kim and Fisher, 1990, Instruments IsoPrime continuous flow IRMS (IsoPrime, Cheadle, UK) interfaced Kim, 1993; Landrurn, 2002; Muhaidat et al., 2007; Voznesenskaya to a Costech elemental analyzer (Costech, Valencia, USA) at the Washington et al., 2010). Voznesenskaya et al. (2010) characterized this di- State University at Pullman Stable Isotope Core facility. Measures of the car versity by reporting three different types of Kranz anatomy in bon isotope composition were determined using a standard procedure relative the family: Atriplicoid, Pilosoid, and Portulacelloid (illustrated t0 PDB (Pee Dfe Belemnite> limestone as the carbon isotope standard (Bender in Appendix S1, see Supplemental Data with the online versionet al., 1973). 8I3C values were determined where 8 = 1000 x (Rsampie/Rstandard) -1. The three photosynthetic pathways discriminate in different proportions of this article). In the Atriplicoid type, the VB are arranged the inisotope a 13C (isotope fractionation; O'Leary, 1988), so the following scale horizontal row (P. umbraticola) or in a zigzag pattern (P. molo- was used to identify photosynthetic pathways: C3, typically < -25 per mil kiniensis Hobdy and P. olerácea) with the hypodermal tissue (%o; O'Leary, 1988; Guralnick et al., 2008; Raven et al., 2008); C4: -10 to (H) in both adaxial(l layer) and abaxial( 1-2 layers) sides of the -16%o (O'Leary. 1988; Sage et al., 2007); Crassulacean acid metabolism flat leaves, serving as WS (Edwards and Voznesenskaya, 2011); (CAM): lar§ely c" ranëe values' but can vary from "910 ~20%" (O'Leary, 1988; Winter and Holtum, 2002). Kranz anatomy (M and BS) completely surrounds the second ary veins. The Pilosoid type (reported to occur in P. amilis Speg., P. grandiflora Hook., and P. pilosa L.) has the VB Leaf ar anatomy—The midportion of living mature leaf blades was cut trans versely into small segments ca. 5 mm wide and fixed and stored in a solution of ranged in a peripheral ring, the H is formed by one layer of FPA cells (Ruzin, 1999). When fresh material was not available, leaf samples from or is absent, and the WS is located in the central part of the flat herbarium specimens or dried in silica gel (Appendix 1) were treated in a solu or terete to subterete leaves; Kranz anatomy is formed on the tion of 10% Aerosol OT or boiled in water for 10 min for rehydration. Leaf external side of the peripheral veins. Finally, the Portulacelloid samples were subsequently prepared following the methods in Ocampo and type (P. cf. bicolor F. Muell.) has a horizontal row of VB below Columbus (2010). Slides were examined with a light microscope and images were recorded with a SPOT digital camera (Diagnostic Instruments, Sterling the adaxial epidermal layer, the H is absent, and well-developed WS tissue is found in the center and abaxial side of the sub Heights, MI) to characterize leaf anatomy. Kranz leaf anatomical types were identified following Voznesenskaya et al. (2010). A set of slides was deposited terete leaves, Kranz anatomy largely surrounds the secondary at Rancho Santa Ana Botanic Garden (RSA). veins, although it is more developed on the adaxial side. Al- Leaf anatomical sections of two of the Portulaca samples with C3-like 813C though there have been a number of studies concerning leaf values, representing P. hirsutissima Cambess. and P. mucronata Link, could anatomy and photosynthesis in Portulacaceae, taxon sampling not be obtained- Therefore, the lamina of those species was examined to deter has been limited to a few species. Portulaca grandiflora, P. ol- !"in? whether they had ™merous dark veins and sma11 areoles- a Pattem tbat j n i , j « , ,. . . has been associated with Kranz anatomy (see Sage et al., 2007). Images were eracea, and P pilosa have received the most attention, prob- acquired with a QImaging Go.5 camJa (Qima|ing, SulTcy, Canada) at the ably because these species are Widespread and plants are easily Department of Botany of the California Academy of Sciences (CAS), obtained. Besides these species, ca. 20 others have been in cluded in different studies (Carolin et al-, 1978; Prabhakar and DNA sequencing andphylogenetic analysis—A combined data matrix of Ramayya, 1979; Nyananyo, 1988; Kim and Fisher, 1990; Kim, DNA chloroplast (ndhF, trnT-psbD spacer, and ndhA intron) and nuclear (ITS)

This content downloaded from 86.59.13.237 on Mon, 21 ff on Thu, 01 Jan 1976 12:34:56 UTC All use subject to https://about.jstor.org/terms December 2013] Ocampo et al.—Portulacaceae leaf anatomy and photosynthesis 2391

Table 1. Leaf anatomy, 513C values, and photosynthetic pathway inferred for the specimens used in this study.

Taxon Leaf anatomy (Kranz type if applicable) 813C %o (sample size if N > 1) C4 biochemical type (if known)

Portulacaceae Portulaca amilis" Kranz (Pilosoid)d -12.5 (3) C4 (NADP-ME)' P. armitlii" Kranz (Portulacelloid) -13.2 c4 P. australis " Kranz (Pilosoid) -13.5 c4 P. bicolor" Kranz (Portulacelloid)d -13.7 c4 P. cf. bicolora Kranz (Portulacelloid) -15.4' C4 (NADP-ME)' P. brevifolia NA -13.0 C4 P. caulerpoides NA -12.9 (2) C4 P. chacoana NA -10.5 C4 P. confertifolia " Kranz (Pilosoid) -11.1(2) C4 P. conzattii NA -11.8 C4 P. cryptopetala " C3-C4 d< -25.5(11) C3-C4' P. cubensis NA -13.75 (2) C4 P. decipiens a Kranz (Pilosoid) -12.9 C4 P. digyna " Kranz (Portulacelloid) -12.2 C4 P. echinosperma " Kranz (Atriplicoid) d -11.6 (2) C4 P. elatior " Kranz (Pilosoid)d -12.7 C4 P. elongata NA -12.3 c4 P. eruca" Kranz (Pilosoid) -15.4 (2) C4 P. filifolia " Kranz (Pilosoid) -13.3 C4 P. fluvialis " Kranz (Pilosoid) -11.6 (3) C4 P. foliosa " Kranz (Pilosoid) -15.0 C4 P. frieseana NA -13.6 C4 P. fulgens " Kranz (Atriplicoid) -11.0 C4 P. gilliesii" Kranz (Pilosoid) -12.6 (3) C4 P. grandiflora NA -12.2 C4 P. grandiflora cv." Kranz (Pilosoid) -11.7 C4 (NADP-ME)' P. guanajuatensis " Kranz (Atriplicoid) d -14.2 c4 P. halimoides a Kranz (Pilosoid) -13.4 Q P. hirsutissima " C3-C4V -24.5 (6) c3-c4/ P. howellii NA -10.2 c4 P. intraterranea " Kranz (Atriplicoid) -11.3 c4 P. johnstonii ° Kranz (Atriplicoid) -13.7 c4 P. lutea NA -11.5 C4 P. massaica " Kranz (Pilosoid) -15.2 c4 P. matthewsii" Kranz (Pilosoid) -12.8 c4 P. mexicana " Kranz (Pilosoid) -12.5 (2) c4 P. minuta NA -12.4 c4 P. molokiniensis " Kranz (Atriplicoid)d -15.5 C4 (NAD-ME)' P. mucronata " non-Kranz (provisionally scored as C3-C4)b -26.2 (7) putative C3-C4 P. mucronulata " Kranz (Pilosoid) -12.0 (2) c4 P. obtusa " Kranz (Pilosoid) -13.0 C4 P. oleracea subsp. impolita " Kranz (Atriplicoid) -12.4 C4 (NAD-ME)« P. oleracea subsp. nitida " Kranz (Atriplicoid) -12.6 C4 (NAD-ME)« P. oleracea subsp. papillatostellulata " Kranz (Atriplicoid) -12.5 C4 (NAD-ME) P. oligosperma " Kranz (Portulacelloid) -13.7 C4 P. papulifera " Kranz (Pilosoid) -12.6 (2) C4 P. papulosa NA -12.5 (2) C4 P. perennis" Kranz (Pilosoid) -12.7 (2) C4 P. pilosa " Kranz (Pilosoid) d -13.0 (3) C4 (NADP-ME)' P. pusilla NA -14.6 c4 P. pygmaea NA -15.6 C4 P. quadrifida " Kranz (Portulacelloid) -12.5 (3) C4 P. retusa " Kranz (Atriplicoid) -12.3 C4 P. rotundifolia a Kranz (Atriplicoid) -12.4 (2) c4 P. rubricaulis NA -12.8(4) c4 P. rzedowskiana " Kranz (Pilosoid) -12.4 C4 P. samoensis NA -11.8(2) C4 P. sclerocarpa NA -12.8(3) c4 P. sedifolia NA -14.2 (2) c4 P. smallii" Kranz (Pilosoid) NA C4 P. suffrutescens " Kranz (Pilosoid) -12.1 (3) c4 P. tingoensis " Kranz (Pilosoid) -12.5 c4 P. tuberosa " Kranz (Pilosoid) -11.3 C4 P. umbraticola subsp. umbraticola " Kranz (Atriplicoid) -12.5 C4 P. umbraticola subsp. coronata a Kranz (Atriplicoid) NA Q P. umbraticola subsp. lanceolata " Kranz (Atriplicoid)d -12.1 (3) C4 P. umbraticola cv. "wildfire mixed" " Kranz (Atriplicoid) -12.6 C4 (NADP-ME)' P. villosa NA -11.8 C4

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2392 American Journal of Botany [Vol. 100

Table 1. Leaf anatomy, 813C values, and photosynthetic pathway inferred for the specimens used in this study. Continued.

Taxon Leaf anatomy (Kranz(Kranz typetype ifif applicable)applicable) 813C %o%c (sample size ifif NN >> 1)1) C4 biochemical typetype (if(if known)known) P. wedermannii NA -12.3 (2) c4 P.P. sp.sp. nov. nov. " 0 Kranz (Portulacelloid) (Portulacelloid) -14.5 C4 Outgroups Pereskia aculeataaculeata" " (Cactaceae)(Cactaceae) c3¿C,' -27.5 c3 Talinopsis frutescens " c3c3« d -27.1 c3 ((Anacampserotaceae) Anacampserotaceae) Talinum paniculatum " (Talinaceae) CC3« -28.2 c3

Notes: NA = not available, a = taxon included in the phylogenetic analysis; b = from leaf venation pattern (see text for details); c = NAD-ME type determined at species level. References: d = Ocampo and Columbus (2010); e = Voznesenskaya et al. (2010);/= Koteyeva, Voznesenskaya, and Edwards, unpublished results.

sequences generated by Ocampo and Columbus (2010,2012) was prepared and the gamma distribution of the characters under study were sampled by an analyzed to estimate the evolutionary relationships among the species of Portu- MCMC analysis performed in SIMMAP. Then, the best-fitting parameters were laca. A combined analysis of the molecular markers was preferred because this estimated using its posterior distribution in the program R (R Development approach provides a more robust phylogeny than analyses of individual loci Core Team, 2006) employing the script provided with the SIMMAP package. (Ocampo and Columbus, 2012). Portulaca hirsutissima and P. mucronata, Posterior probabilities for ancestral character states were obtained using 1 000 which were not included in the Ocampo and Columbus (2012) study, were random trees generated in the MrBayes analysis after the burn-in period, found to have C3 8nC values and were therefore sampled for the current study. In addition, we explored the diversification patterns of C4 subtypes. Because Samples from herbarium specimens of these species (Appendix 1) were used for the biochemical subtypes are not characterized for most species of Portu DNA extraction, amplification, and sequencing as described in Ocampo and lacaceae, we pruned the phylogenies to include only the species reported by Columbus (2012). Sequences were aligned using MUSCLE version 3.7 (Edgar, Voznesenskaya et al. (2010). Evolution of C4 subtypes (NADP-ME and 2004), followed by manual alignment in Se-Al version 2.0al 1 (Rambaut, 2002). NAD-ME) and C3-C4 intermediacy was estimated using a Bayesian approach as The combined data matrix (archived in TreeBASE, study accession 13963) explained above, was partitioned by locus and analyzed using Bayesian inference under Markov Chain Monte Carlo (MCMC; Yang and Rannala, 1997) in MrBayes version 3.1.2 (Ronquist et al., 2012) and maximum likelihood (ML; Felsenstein, 1973) RESULTS in RAxML version 7.2.6 (Stamatakis, 2006). Bayesian analysis was conducted

MrModeltest™ xf îe„be,St~fit version modo1 2.3 (Nylander, ^rr,0" lrad;2004) undermf^r1mfrperProvid^by the Akaike Information This Cnte- study , / includes. . . . f , . 73fc taxaPL , ,(counting \ three outgroups), rep rion (AIC; Akaike, 1974). The model selected for ITS and ndhF was a general resenting 64 species, SIX subspecies, and two cultivais (Table 1 ; time reversible model (GTR; Tavaré, 1986) plus parameters for proportion Appendix 1). of invariant sites (I; Reeves, 1992) and a gamma-distributed rate variation (G; Yang. 1993). For trnT-psbD and ndhA the model selected was GTR + G. 5I3Ç data—We gathered 813C data for 64 samples that were Because RAxML can only employ one model of evolution for a partitioned data newly generated or obtained from Ocampo and Columbus GTRGAMMA model c<'mhiped(GTR + G). dataBayesian maírix analyses was performed were run with using two tbe repli- (2010) J and Voznesenskaya et al. (2010). Data for P. smallii cates for 10 000 000 generations. Trees were saved every 1000th R generation, Wilson and P. umbraticola subsp. coronata (Small) J. F. and the burn-in value for obtaining an allcompat consensus tree was set to Matthews & Ketron were not obtained due to lack of leaf mate ignore the first 30% of trees. Clade support was determined by Bayesian poste- rial (Table 1). Samples with neither molecular nor leaf anatomy rior probabilities (p.p.; Rannala and Yang, 1996; Li et al., 2000) and nonpara- data were sampled from herbarium specimens in the collections metric bootstrapping (bs; Felsenstein, 1985)^ from 1000 replicates performed 0f the Missouri Botanical Garden (MO), the New York Botani simultaneously with the ML search using the "-f a" option. Phylogenetic cal Garden analy (NY), and the V. L. Komarov Botanical Institute ses were executed using the High Performance Computing Cluster at CAS. /T i-\ r ?iL~. , • /» 4- ,, . n . , (LE) for 513C analysis (Appendix 1). All species had 813C values within the C4 pathway range (-10.2 to -15.6), except for Divergence times—The ambiguous fossil record of the suborder Cactineae p Cryptopetala (-25.5), P. hirsutissima (-24.5), P. mucronata ^etSnS^V2lnCe timeH t Challengin,g task' Rf,T st"dies T USf (-26.2), and the outgroup taxa, all of which had values charac different calibration strategies and taxon sampling resulted in dissimilar esti- k . . fr , • / oqi mates for the most recent common ancestor (MRCA) of Portulacaceae (e.g., tenstic Ot C3 photosynthesis (—28.2 to —25.5). Ocampo and Columbus. 2010; Arakaki et al., 2011; Ocampo and Columbus, 2012). Therefore, in this study we chose to estimate the ages of the MRCA of Leaf anatomy—We examined 51 leaf anatomical prepara dle two new samples, as well as P. cryptopetala with which they formed a clade tions of Portulaca (Table 1 ), 40 of which were newly generated (see below), building upon the results found in Ocampo and Columbus (2012), for this study (Figs 2 and 3). Material treated in 10% Aerosol which represents the best sampled phylogeny of Portulacaceae known to date. at u -i a- . .r j- i ■. We used the ages recovered in that study (the mean of the 95% highest posterior °T,OT h°lled,in WateJ m0Stly d'Splayed limited tissue expansion density) for fixing the dates of major clades in a Bayesian majority-rule consen- was inadequate for anatomical characterization. The excep sus tree using the program PATHd8 (Britton et al., 2007). We used the ages tions were the samples for P. howellii (D. Legrand) Eliasson, of the MRCA of the following clades (ages in millions of years; Myr): P. quadrifida, and P. retusa Engelm. (the first from silica-preserved 7a/mops¡s+Portulacaceae (29.8), Portulacaceae (23), OL clade (18.6), Austra- leaves, the latter two from herbarium specimens), which provided lian clade (17.2) AL clade (17.5), Pilosa clade (10.7), Olerácea clade (6.2), and acceptable material for anatomical examination, m ratico a c a e ( . ). All of the Portulaca samples analyzed have Kranz anatomy, except P. cryptopetala, which has a C3-C4 type anatomy with Character evolution-Both leaf anatomy types and photosynthesis path- Kranz-like BS cells containing numerous organelles arranged ways estimated by the oIJC data were subjected to ancestral character recon- • u- u- u * • ? r r* • T struction. A Bayesian approach was used for estimating ancestral characters 1 ^ centripetal position, which IS characteristic of C3-C4 lnter (Bollback, 2006) as implemented in the program SIMMAP version 1.5 (Bollback, mediates (Fig. 4K in Ocampo and Columbus, 2010; Figs. 2U-V 2009). The priors on the overall evolutionary rate were obtained using a two- in Voznesenskaya et al., 2010). Members of the outgroup display step approach, as recommended in Bollback (2009). The a and p parameters of a C3 leaf anatomy, which is characterized by having palisade

Fig. 2. Light micrographs of transectional leaf anatomy in Portulacaceae. (A-K) Atriplicoid type. (A) Portulaca fulgens. (B) P. intraterranea. (C) P. johnstonii. (D) P. olerácea subsp. impolita. (E) P. olerácea subsp. nítida. (F) P. olerácea subsp. papillatostellulata. (G) P. retusa. (H) P. rotundifolia. (I) P. umbraticola subsp. umbraticola. (J) P. umbraticola subsp. coronata. (K) P. umbraticola cv. 'wildfire mixed'. (L-X) Pilosoid type. (L) P. australis. (M) P. confertifolia. (N) P. decipiens. (O) P. eruca. (P) P.filifolia. (Q) P. fluvialis. (R) Pfoliosa. (S) P. gilliesii. (T) P. grandiflora cv. (U) P. halimoides. (V) P. mas saica. (W) P. matthewsii. (X) P. mexicana. Black scale bar = 0.75 mm; gray scale bar = 1 mm.

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2394 American Journal of Botany [Vol.100

Fig. 3. Light micrographs of transactional leaf anatomy in Portulacaceae. (A-J) Pilosoid type. (A) Portulaca mucronulata. (B) P. obtusa. (C) P. papu lifera. (D) P. perennis. (E) P. pilosa. (F) P. rzedowskiana. (G) P. smallii. (H) P. suffrutescens. (I) P. tingoensis. (J) P. tuberosa. (K-P) Portulacelloid type. (K) P. armitti. (L) P. cf. bicolor. (M) P. digyna. (N) P. oligosperma. (O) P. quadrifida. (P) P. sp. nov. Black scale bar = 0.75 mm; gray scale bar = 1 mm.

mesophyll cells on the adaxial side and spongy mesophyll cells a C4 species (Fig. 4A). The C4 species has a higher venation den on the abaxial face, as seen in Figs. 4C, 5G, and 5K in Ocampo sity, with a relatively more reduced areole space when compared and Columbus (2010). to P. cryptopetala, P. hirsutissima, and P. mucronata (Figs. 4B-D). In all samples with Kranz anatomy, chloroplasts are located The first two species are known to have C3-C4 leaf anatomy centripetally in the BS cells (e.g., Figs. 1A-C). Among all of the (Ocampo and Columbus, 2010; Voznesenskaya et al., 2010; C4 species analyzed, we identified the three leaf anatomical Koteyeva, Voznesenskaya, and Edwards, unpublished results); types characterized by Voznesenskaya et al. (2010): Atriplicoid therefore, because of the resemblance in vein pattern, the leaf (Figs. 2A-K), Pilosoid (Figs. 2L-X and 3A-J), and Portulacel- anatomical type of P. mucronata was provisionally scored as C3-C4. loid (Figs. 3K-P) (Table 1); however, minor variations in the It is noteworthy to mention that P. hirsutissima (Fig. 4D) has Atriplicoid and Pilosoid types were detected. In the case of the trichomes on the leaf surface, a rare feature in the family. Atriplicoid type, one to four layers of WS cells are conspicu ously separating the VB, which are in a zig-zag pattern and are Phylogenetic analysis—Our sampling includes 53 taxa, count located toward the central part or the adaxial side of the leaves ing the outgroup taxa. The analyses estimated that P. hirsutis of P.fulgens Griseb. (Fig. 2A), P. intraterranea J. M. Black sima and P. mucronata are sisters to P. cryptopetala (Cryptopetala (Fig. 2B), P. johnstonii Flenrickson (Fig. 2C), and P. echino- clade, Fig. 5). The Bayesian and ML topologies are essentially sperma Flauman (Fig. 4L in Ocampo and Columbus, 2010). identical to the one retrieved in the analysis of the combined mo in addition, the latter species presented a subterete leaf shape in lecular data in Ocampo and Columbus (2012), where the family cross section. With respect to the Pilosoid type, our sample of and major clades are strongly supported as monophyletic. P. grandiflora does not possess FI (Fig. 2T). Images of the venation pattern of the species with C3 ô13C val- Divergence times—A chronogram showing the estimated ues are shown in Fig. 4, and they are compared to P. umbraticola, ages of the MRCA of the major clades in Portulacaceae is found

» , »

%t, ; m ^ - ■-.•t-u:

¿ ¿i »«*».•- ,

. ¿Ï& y. •*'■-- "" .. n íiVs*. mm

Fig. 4. Light micrographs comparing venation patterns of the lamina of some species of Portulaca. (A) P. umbraticola, a C4 species displaying the high venation density characteristic of Kranz anatomy. (B) P. cryptopetala, a C3-C4 intermediate species. (C-D) Portulaca species with 513C values resem bling C3 plants and with increased interveinal distance when compared to a C4 species. (C) P. mucronata. (D) P. hirsutissima (note the pubescence on the surface of the leaf, rare in Portulacaceae). Scale bar = 0.5 mm.

in Fig. 6A. The stem and crown node ages for the Cryptopetala Photosynthetic pathways—Most of the Portulaca samples clade are 14.4 and 8 Myr, respectively. had C4-like S'3C values and Kranz anatomy; therefore, they were coded as C4 plants. However, there were three remarkable exceptions: P. cryptopetala, P. hirsutissima and P. mucronata, Character evolution—Leaf anatomy—Results are not con- the only members of the Cryptopetala clade (Fig. 5), have C3 elusive about the ancestral condition of the leaf anatomy for type values. It has been shown that P. cryptopetala is a C3-C4 Portulacaceae (Atriplicoid, 0.61 ; Pilosoid, 0.23; Portulacelloid, intermediate (Voznesenskaya et al., 2010), a condition that can 0.14), but the ancestral character states for the OL and the AL not be detected with the use of stable isotope or leaf anatomy clades were recovered with high probability as Portulacelloid data alone. Portulaca hirsutissima is also a C3-C4 intermediate and Atriplicoid, respectively (Fig. 5). All samples from the OL with reduced loss of C02 by photorespiration and selective clade display a Portulacelloid leaf anatomy, while the AL clade localization of glycine decarboxylase in mitochondria of is more diverse and its members have three different anatomical Kranz-like BS cells (Koteyeva, Voznesenskaya, and Edwards, types. Both the Pilosoid (Pilosa clade) and the C3-C4 (Crypto- unpublished results). Thus, we consider the Cryptopetala clade pétala clade) leaf anatomical types were recovered as derived to be composed of C3-C4 intermediates; however, further stud from the Atriplicoid anatomy. The Atriplicoid type was found ies are required to determine whether P. mucronata is a C3 or in the Olerácea and Umbraticola clades, although there is un- C3-C4 intermediate species. certainty about the number of times that this anatomical type Character evolution analyses showed that the C4 strategy evolved in Portulacaceae (ancestral character state probabilities evolved once within Portulacaceae ca. 23 Myr, with one transi for the Pilosa + Umbraticola clades: Pilosoid, 0.51 ; Atriplicoid, tion to a C3-C4 intermediate pathway in the Cryptopetala clade 0.48). Interestingly, the arrangement of the VB in both clades ca. 8 Myr (Fig. 6A). With regard to C4 biochemical type diver shows some differences: in the Umbraticola clade they are ar- sification, the analyses recovered the C4 NADP-ME type as the ranged in a nearly horizontal straight line, while in the Olera- ancestral condition for Portulacaceae. It was shown that there is cea clade there is a tendency to find the VB in a zigzag pattern a transition from the NADP-ME type to C3-C4 intermediacy in (Figs. 2A-K). To investigate an alternative scenario where the the Cryptopetala clade and subsequently to the NAD-ME type Atriplicoid configuration is not homologous, we carried out an in the Olerácea clade (Fig. 6B). in contrast, the estimation of exploratory analysis assuming the Atriplicoid leaf anatomies evolution of C4 in Portulacaceae showed only one origin of the found in the Olerácea and Umbraticola clades represent two C4 pathway (Fig. 6A). However, p.p. for the ancestral C4 bio different character states. The results (not shown) did not sig- chemical type are not conclusive for the MRCA of the Crypto nificantly differ from those of the first analysis, nor did they pétala and Olerácea clades (C3-C4 = 0.57, NAD-ME = 0.37, resolve the ancestral leaf anatomy type for Portulacaceae. NADP-ME = 0.052, and C3 < 0.01).

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2396 American Journal of Botany [Vol.100

- ■■ Talinum paniculatumpaniculatum . ■ PereskiaPereskia aculeata aculeata . ■■ TalinopsisTalinopsis frutescens frutescens - □ Portulaca quadrifidaquadrifida . □□ PortulacaPortulaca digyna digyna . □□ PortulacaPortulaca armittii armittii . □□ PortulacaPortulaca bicolor bicolor Australian - □ Portulaca cf.cf. bicolor bicolor P clade . □ PortulacaPortulaca oligospermaoligosperma o o> Portulacelloid - □ Portulaca sp.sp. nov. now lo - ■ Portulaca hirsutissimahirsutissima Cryptopetala - ■ Portulaca cryptopetalacryptopetala - ■ Portulaca mucronata clade C - ■ Portulaca intraterranea CrC4C^-C4 - ■ Portulaca rotundifolia - ■■ PortulacaPortulaca echinospermaechinosperma ■ I PortulacaPortulaca oleráceaoleracea subsp.subsp. papillatostellulata papillatostellulata ■ I PortulacaPortulaca oleráceaoleracea subsp.subsp. nítida nitida OleráceaOleracea ■ I PortulacaPortulaca guanajuatensis clade ■ I PortulacaPortulaca fulgensfulgens ■ I PortulacaPortulaca retusaretusa AtriplicoidAtriplicoid ■ I PortulacaPortulaca johnstoniijohnstonii ■ I PortulacaPortulaca molokiniensismolokiniensis ■ I PortulacaPortulaca oleráceooleracea subsp.subsp. impoliimpolita ta 5 ■ [ PortulacaPortulaca umbraticolaumbraticola subsp. coronata ■ I PortulacaPortulaca umbraticola subsp. lanceolatalanceolata Umbraticola ••••#••• •• • ••• ■ [ Portulaca umbraticolaumbraticola cv cv. 'wildfire 'wildfire mix' mix' clade ■ PortulacaPortulaca umbraticola umbraticola subsp. subsp.umbraticola umbraticola ■ PortulacaPortulaca elatiorelatior [>• Atriplicoid Atriplicoid O ■ Portulaca foliosa foliosa r * ■ PortulacaPortulaca massaica o BranchBranch support support ■ PortulacaPortulaca decipiens decipiens STp ■ PortulacaPortulaca australisaustralis Q O) a >i 90% > bootstrap 90% bootstrap ■ PortulacaPortulaca ftlifolia filifolia Portulaca tuberosa > >0.95 0.95p.p. p.p. ■ Portulaca tuberosa ■ PortulacaPortulaca papulifera papulifera q ^75-89% 75-89% bootstrap bootstrap ■ PortulacaPortulaca JluvialisJluvialis > >0.95 0.95p.p. p.p. ■ PortulacaPortulaca tingoensistingoensis ■ PortulacaPortulaca halimoides O O>0.95 >0.95p.p. p.p. ■ Portulaca smalliismallii Pilosa ■ Portulaca pilosa Ocampo 1718 clade 0 075-89% 75-89% bootstrap bootstrap ■ PortulacaPortulaca matthewsii matthewsii ■ PortulacaPortulaca pilosapilosa Nortrup Nortrup s.n. s.n. Anatomy types ■ Portulaca sufjrutescenssufjrutescens Anatomy types ■ PortulacaPortulaca mexicana CD Pilosoido ■ PortulacaPortulaca rzedowskianarzedowskiana ■ C3 ■ PortulacaPortulaca amiami lis ■ C3-C4 ■ PortulacaPortulaca grandiflora grandiflora cv. cv. ■ PortulacaPortulaca confertifolia confertifolia H AtriplicoidAtriplicoid ■ PortulacaPortulaca eruca EH PilosoidPilosoid ■ Portulaca perennisperennis ■ Portulaca mucronulata mucronulata □ Portulacelloid ■ Portulaca gilliesiigilliesii "^1 ■ PortulacaPortulaca obtusa Fig. 5. Phylogenetic relationships and ancestral character reconstruction for leaf anatomy types in Portulacaceae. Bayesian allcompat tree from the analysis of a combined matrix of ndhF, trnT-psbD, ndhA intron, and ITS sequences. Branch support for phylogenetic relationships and posterior probabili ties (p.p.) for leaf anatomy types are shown at each node. The arrangement of vascular bundles within the leaf (cross section) is illustrated on the right (green circles represent enlarged mesophyll cells surrounding the bundle sheaths). Leaf anatomy type for Portulaca hirsutissima and P. mucronata is esti mated from their venation patterns and unpublished results (see text).

DISCUSSION was present in the group ca. 23 Myr. The survey of Muhaidat et al. (2007) provides evidence that the Atriplicoid anatomy is This investigation expands taxon sampling in Portulacaceae the predominant type in C4 eudicots, due in part by the preva enabling further examination of the diversity of leaf anatomy lence of laminate leaves in C3 ancestral taxa. Cactaceae is the and photosynthesis pathways. This information, along with the closest relative of Portulacaceae (Arakaki et al., 2011), and its analysis of DNA sequence data in a phylogenetic framework, most early divergent members have laminate leaves (Pereskia allowed us to explore diversification of these two traits in the spp.; Edwards et al., 2005) and C3 anatomy (Ocampo and Colum family and to hypothesize about the origins of C4 biochemical bus, 2010). Therefore, it is plausible that the Atriplicoid type was types and C3-C4 intermediacy in the group. The interpretation present in the ancestor of the Portulacaceae. The species of Portu of these results has to be considered with some prudence given laca with Atriplicoid and Pilosoid leaf anatomies have cotyledons the potential difficulties in the estimation of the evolutionary with the Atriplicoid type (Voznesenskaya et al., 2010), which pro history of complex traits (Wiens et al., 2007; Christin et al., 2010; vides further evidence to support the Atriplicoid anatomy as the Wiens, 2011 ). ancestral condition in the family. Our analysis indicates that the Portulacelloid type appeared ca. Leaf anatomy diversification in Portulacaceae—Our survey 18 Myr in the Old World (Ocampo and Columbus, 2012), and it is shows the presence of C3-C4 anatomy and three different C4 restricted to the species of the OL clade. This anatomical type may Kranz anatomical types (Atriplicoid, Pilosoid, and Portulacel- have resulted from the developmental suppression of the adaxial loid) in the family, as observed among representative species hypoderm of an Atriplicoid leaf, leaving the VB in close proximity by Voznesenskaya et al. (2010). We were unable to assign a to the adaxial epidermal cells. specific ancestral anatomical type for Portulacaceae, even if the The Atriplicoid anatomy is recovered as the ancestral type of the Atriplicoid anatomies in the Olerácea and Umbraticola clades AL clade. Although the Atriplicoid type is found in species of the were not considered homologous. However, the combined pos- Olerácea and Umbraticola clades, the configuration found in the lat terior probabilities for a C4 anatomy, as opposed to a C3-C4 inter clade may have derived from a Pilosoid anatomy ca. 2.2 Myr. termediate type, suggested that some form of Kranz anatomy This hypothesis of independent origins of the Atriplicoid anatomy

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms OCAMPO ET AL. PORTULACACEAE LEAF ANATOMY AND PHOTOSYNTHESIS 2397

I |Taknum Tali paruculatum ram paniculatum I ■I PereskiaPereskia aculeata aculeata loutgroup I OUtgFOlip I ITalinopsts Talinopsis fiutescens frutescens I i

!IOL OL clade

i| CryptopetalaCryptopetala ' claae

OleráceaOleracea clade

]IUmbraticola Umbraticola clade *0 0 1 5* S

Pilosa clade

30.0 20.0 10.0 0.0 Myr

B I TalinumTalinum paniculatumpaniculatum

I PereskiaPereskia aculeataaculeata outgroup

I Talinopsis frutescensfrutescens

I PortulacaPortulaca cf.cf. bicolorbicolor OL clade

I Portulaca cryptopetalacryptopetala CryptopetalaCryptopetala clade

I Portulaca oleraceaolerácea ^3 OleráceaOleracea Oo clade x £ I Portulaca molokiniensis > I r £ 5" o £ S" E5£ Umbraticola o Ii Portulaca umbraticola o cv. 'wildfire mix' clade

Photosynthetic I Portulaca pilosapilosa biochemical type type m C3 m C4, NAD-ME I Portulaca amilisamilis Pilosa □ C4, NADP-ME clade m C3-C4

1I PortulacaPortulaca grandiflora cv.cv.

Fig. 6. Chronogram and ancestral character reconstructions for photosynthetic pathways in Portulacaceae. Posterior probabilities in the form of a pie chart are shown at each node of the Bayesian reconstruction. (A) Chronogram derived from a PATHd8 analysis (Britton et al., 2007) based on the Bayesian allcompat tree from the combined matrix, showing the estimated evolutionary history of the C4 and C3-C4 pathways. Dates in millions of years (Myr). (B) Pruned Bayesian allcompat tree, showing the estimated evolutionary history of the C4 variants (NADP-ME and NAD-ME) and C3-C4 intermediacy. Only taxa with reported photosynthetic biochemical data are included.

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2398 American Journal of Botany [Vol. 100

is supported by the presence of different C4 biochemical sub- the C4 condition could have been lost once, with a switch to a types (see below) and VB arrangement between these two C3-C4 intermediate pathway (Figs. 6A-B; see below), clades. While the species of the Umbraticola clade have the VB Although relevant data on photosynthetic biochemical C4 arranged in a horizontal plane (Ocampo and Columbus, 2010; variants in Portulacaceae remains scarce, it is sufficient for Voznesenskaya et al., 2010), the members of the Olerácea clade drawing some preliminary observations about its diversification tend to have the VB in a zigzag pattern (e.g., Prabhakar and within the family. Our results show that the NADP-ME type is Ramayya, 1979; Kim, 1993; Ocampo and Columbus, 2010; the ancestral variant in the Portulacaceae, though it may have Voznesenskaya et al., 2010). It is possible that the zigzag con- switched ca. 14 Myr toward a C3-C4 intermediate or toward the figuration provides increased vein density by displacement of C4 NAD-ME type (Fig. 6B); however, the former scenario im the VB in both abaxial and adaxial directions due to the devel- plies a secondary gain of the C4 pathway in the Olerácea clade. opment of additional WS among them, which may be an advan- In any event, transitions between different photosynthesis path tageous strategy in habitats with low water availability. ways would require extensive genomic adjustments (e.g., The Pilosoid type is characteristic of the Pilosa clade, and Brâutigam et al., 2008) and the benefits of switching from one may have been already established at least ca. 10.7 Myr. Be- C4 biochemical type to another are not apparent, especially cause of the peripheral arrangement of the VB in the Pilosoid since there is no strong evidence that demonstrates the advan type, Edwards and Voznesenskaya (2011) considered that this tages of one C4 variant over the other from a bioenergetic per form could resemble some evolutionary stages of leaf anatomi- spective (Edwards and Voznesenskaya, 2011 ; Furbank, 2011). cal diversification that occurred in Chenopodiaceae. There are models that hypothesize that the C3 pathway was The C3-C4 anatomy is found in the flat leaves of P. cryptopetala the precursor of C4 photosynthesis, and that this transition was (Voznesenskaya et al., 2010) and P. hirsutissima (Koteyeva, accomplished through a C3-C4 intermediate stage (Kennedy and Voznesenskaya and Edwards, unpublished), which are mem- Laetsch, 1974; Monson et al., 1984; Sage et al., 2012). How bers of the Cryptopetala clade. They are C3-C4 intermediates, ever, the current study provides evidence for a reversal from C4 with Kranz-like BS cells having intracellular distribution of or- to C3-C4 intermediacy, supporting the proposal that the C3-C4 ganelles similar to that found in other C3-C4 intermediate taxa intermediate pathway can be considered as an independent (e.g., Monson, 1999; Ku et al., 1983; Rajendrudu et al., 1986; strategy for increasing fitness in C02-limited environments Brown and Hattersley, 1989; Christin et al., 2011b; Muhaidat (Monson, 1999; Sage et al., 2011; Christin et al., 201 lb), et al., 2011; Voznesenskaya et al., 2001). The other member of While reversions from C4 to C3 has been suggested in Che this clade, P. mucronata, is assumed to lack Kranz anatomy as nopodiaceae and Poaceae (Pyankov et al., 2001; Bouchenak it has a lower vein density than observed in C4 species. The loss Khelladi et al., 2009; Kadereit et al., 2012) it has been difficult of the Kranz anatomy is estimated to have occurred only once to differentiate whether these are reversals or multiple cases of within Portulacaceae ca. 8 Myr, likely from an ancestor with evolution oftheC4 pathway (Christin et al., 2010; Voznesenskaya Atriplicoid leaf anatomy. The species of the Cryptopetala clade et al., 2013). Our analysis shows that the C4 condition could are endemic to South America (Legrand, 1962), and are found have been lost in the species of the Cryptopetala clade ca. 8 in dry and semihumid habitats (Legrand, 1962; Coelho and Myr, which is consistent with the estimated age of the C3-C4 Giulietti, 2010). This transition in the Cryptopetala clade is of intermediacy in the verticillata group of the Molluginaceae special interest because it was not recovered as a precursor of (Christin et al., 201 lb); however, evidence for a possible switch the Kranz anatomy as proposed by some models of C4 evolution from C4 to C3-C4 intermediacy has only been documented in (see Sage, 2004), but rather a derived condition. A transition Portulacaceae (Ocampo and Columbus, 2010; this study), from C4 photosynthesis to C3-C4 intermediacy implies not only The C3-C4 intermediacy is a rare condition, known to occur in major changes in leaf anatomical features, but also in the bio- only 13 angiosperm families (Sage et al., 2011). The Portu chemical machinery for reducing photorespiration without a C4 lacaceae are known to have type IC3-C4 intermediate species pathway (i.e., recovery of the C3 cycle in mesophyll chloroplasts, (Voznesenskaya et al., 2010), a system that enhances photosyn while maintaining refixation of photorespired C02 in BS cells). thesis by refixing photorespired C02 in the BS cells and has The potential advantages of developing a C3-C4 type anatomy C3 type carbon isotope composition without any development from a C4 configuration are obscure. In general, the anatomical of C4 biochemistry (for discussion of type I and II intermediates variants (C3-C4, C4/Kranz) may each offer advantages in terms see Edwards and Ku, 1987; Sage et al., 2012). of internal light distribution and water and C02 allocation (see The species of the Cryptopetala clade prefer open sites in Tholen et al., 2012). These factors certainly need further explo- sandy or rocky soils, and it is common to find P. cryptopetala ration to understand leaf anatomical and photosynthetic varia- associated with rivers or streams, as well as ruderal habitats tion in Portulacaceae. (especially roadsides). The potential advantage of C4 vs. C3-C4 intermediacy needs to be evaluated, especially because the Por Evolution of photosynthetic pathways in Portulacaceae—The tulaca species that use these two pathways may be equally corn results from this broad survey provide strong evidence for C4 pho- petitive in the same habitats (G. Ocampo, personal observation), tosynthesis being the predominant pathway in the family (Table 1 ; The results of this investigation indicate that the C4 pathway Fig. 6A). Our analysis estimates that the C4 strategy was acquired originated only once in Portulacaceae and that C3-C4 interme ca. 23 Myr, which is consistent with other studies that propose the diacy is a derived trait. However, some studies argue that cur emergence of the C4 pathway in other plant families in response to rent methods of ancestral character reconstruction may mislead the declining atmospheric C02 concentration in the Oligocene the interpretation of the evolutionary transitions of complex (e.g., Besnard et al., 2009; Christin et al., 2011a; Kadereit et al., traits (Wiens et al., 2007; Christin et al., 2010; Wiens, 2011). 2012). Although the geographical origin of Portulacaceae is uncer- Under this consideration, an alternative scenario of photosyn tain (Ocampo and Columbus, 2010; Ocampo and Columbus, thesis evolution in Portulacaceae would imply the existence of 2012), Sage et al. (2011) speculated a South American origin for an ancestor with C3-C4 anatomical and photosynthetic charac the C4 pathway in the family. In addition, our results indicate that teristics; subsequent dispersal and speciation events into harsh

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms December 2013] Ocampo et al.—Portulacaceae leaf anatomy and photosynthesis 2399 environments would have promoted the gradual, Christin, independent P. A., C. P. Osborne, R. F. Sage, M. Arakaki, and E. J. Edwards. consolidation of the C4 pathway in different lineages, 2011a. C4 aseudicots sug are not younger than C4 monocots. Journal of gested in other groups (e.g., Christin et al., 2011b). Experimental Therefore, Botany 62: 3171-3181. Christin, P. A., T. L. Sage, E. J. Edwards, R. M. Ogburn, R. Khoshravesh, the existence of the C3-C4 condition would represent a con and R. F. Sage. 2011b. Complex evolutionary transitions and served strategy under this scenario. The presence of multiple the significance of C3-C4 intermediate forms of photosynthesis in leaf anatomical types is a signature of independent Molluginaceae. origins Evolution; of International Journal of Organic Evolution the C4 pathway in Amaranthaceae (Kadereit et al., 65: 2003), 643-660. so the existence of different Kranz and C4 biochemical Coelho, types A. A.in O.Portu P., and A. M. Giulietti. 2010. O género Portulaca L. lacaceae may support the hypothesis of independent (Portulacaceae) acquisition no Brasil. Acta Botánica Brasilica 24: 655-670. of C4 photosynthesis in the group. Dengler, N. G., and T. Nelson. 1999. Leaf structure and development This study provides insights into the evolutionary in C4history plants. In of R. F. Sage & R. K. Monson [eds.], C4 plant biology, leaf anatomy and photosynthesis pathways in Portulacaceae, 133-172. Academic al Press, San Diego, California, USA. though a number of questions still remain open. ItEdgar, is evident R. C. 2004. that MUSCLE: multiple sequence alignment with high ac additional biochemical data are needed to clarify the curacyevolutionary and high throughput. Nucleic Acids Research 32: 1792-1797. Edwards, E. J., R. Nyffeler, and M. J. Donoghue. 2005. Basal cactus history of C4 variants and C3-C4 intermediacy. In addition, it is phytogeny: Implications of Pereskia (Cactaceae) paraphyly for the not clear whether the C4 pathway has only one origin in the fam transition to the cactus life form. American Journal of Botany 92: ily, or if it was secondarily acquired from a C3-C4 ancestor 1177-1188. in the Olerácea clade, where the NAD-ME variant could Edwards, be a signature G. E., and M. S. B. Ku. 1987. The biochemistry of C3-C4 inter of that transition. The estimation of evolutionary mediates.patterns In mayM. D. Hatch and N. K. Boardman [eds.], The biochemis also be obscured by insufficient taxon sampling ortry byof plants, the vol.ab 10, Photosynthesis, 275-325. Academic Press, New sence of extinct taxa in the analyses. Increased York, taxon New York,sam USA. pling and further biochemical and ecological Edwards, studies G. E., andE.will V. Voznesenskaya. 2011. C4photosynthesis: Kranz contribute to understanding the relative advantages forms of theseand single-cell C02 C4 in terrestrial plants. In A. S. Raghavendra and concentration mechanisms in concert with different leafR. F. anatomiSage [eds.], C4 photosynthesis and related C02 concentrating mechanisms, 29-61. Springer, Dordrecht, The Netherlands. cal configurations, and to clarifying their evolutionary history in Edwards, G. E., and D. A. Walker. 1983. C3, C4: mechanisms, and cellu Portulacaceae. lar regulation, of photosynthesis. Blackwell Scientific, Oxford, UK. Ehleringer, J. R., T. E. Cerling, and B. R. Helliker. 1997. C4 photosyn LITERATURE CITED thesis, atmospheric C02, and climate. Oecologia 112: 285-299. Akaike, H. 1974. A new look at the statistical model identification. Ehleringer, IEEE J. R., and R. K. Monson. 1993. Evolutionary and ecologi Transactions on Automatic Control 19: 716-723. cal aspects of photosynthetic pathway variation. Annual Review of Arakaki, M., P. A. Christin, R. Nyffeler, A. Lendel, U. Eggli, R. Ecology and Systematics 24: 411-439. M. Ogburn, E. Spriggs, M. J. Moore, and E. J. Edwards. 2011. Felsenstein, J. 1973. Maximum likelihood and minimum-steps methods Contemporaneous and recent radiations of the world's major succu for estimating evolutionary trees from data on discrete characters. lent plant lineages. Proceedings of the National Academy of Sciences, Systematic Zoology 22: 240-249. USA 108: 8379-8384. Felsenstein, J. 1985. Confidence limits on phylogenies: An approach Bender, M. M., I. Rouhani, H. M. Vines, and C. C. Black Jr. 1973. using the bootstrap. Evolution; International Journal of Organic Evolution 39: 783-791. 13C/I2C ratio changes in Crassulacean acid metabolism plants. Plant Physiology 52: 427^130. Furbank, R. T. 2011. Evolution of the C4 photosynthetic mechanism: Besnard, G., A. M. Muasya, F. Russier, E. H. Roalson, N. Salamin, Are there really three C4 acid decarboxylation types? Journal of and P. A. Christin. 2009. Phylogenomics of C4 photosynthesis in Experimental Botany 62: 3103-3108. sedges (Cyperaceae): Multiple appearances and genetic convergence. Guralnick, L. J., A. Cline, M. Smith, and R. F. Sage. 2008. Evolutionary Molecular Biology and Evolution 26: 1909-1919. physiology: the extent of C4 and CAM photosynthesis in the gen Bollback, J. P. 2006. Simmap: Stochastic character mapping of discrete era Anacampseros and Grahamia of the Portulacaceae. Journal of traits on phylogenies. BMC Bioinformatics 7: 88. Experimental Botany 59: 1735-1742. Bollback, J. P. 2009. Simmap version 1.5. [computer program]. Website Guralnick, L. J., G. E. Edwards, M. S. B. Ku, B. Hockema, and V. http://www.simmap.com/pgs/docs.html [accessed 30 April 2012]. Franceschi. 2002. Photosynthetic and anatomical characteristics in Bouchenak-Khelladi, Y., G. A. Verboom, T. R. Hodkinson, N. Salamin, the C4-Crassulacean acid metabolism-cycling plant Portulaca gran O. Francois, G. N. Chonghaile, and V. Savolainen. 2009. The ori diflora. Functional Plant Biology 29: 763-773. gins and diversification of C4 grasses and savanna-adapted ungulates. Guralnick, L. J., and M. D. Jackson. 2001. The occurrence and phy Global Change Biology 15: 2397-2417. logenetics of crassulacean acid metabolism in the Portulacaceae. Braütigam, A., S. Hoffmann-Benning, and A. P. M. Weber. 2008. International Journal of Plant Sciences 162: 257-262. Comparative proteomics of chloroplast envelopes from C3 and C4 Gutiérrez, M., V. E. Gracen, and G. E. Edwards. 1974. Biochemical and plants reveals specific adaptations of the plastid envelope to C4 photo cytological relationships in C4 plants. Planta 119: 279-300. synthesis and candidate proteins required for mantaining C4 metabo Hatch, M. D. 1987. C4 photosynthesis, a unique blend of modified bio lite fluxes. Plant Physiology 148: 568-579. chemistry, anatomy, and ultrastructure. Biochimica et Biophysica Britton, T., C. L. Anderson, D. Jaquet, S. Lundqvist, and K. Bremer. Acta 895: 81-106. 2007. Estimating divergence times in large phylogenetic trees. Kadereit, G., D. Ackerly, and M. D. Pirie. 2012. A broader model for C4 Systematic Biology 56: 741-752. photosynthesis evolution in plants inferred from the goosefoot family Brown, H. R., and P. W. Hattersley. 1989. Leaf anatomy of C3-C4 spe (Chenopodiaceae s.s.). Proceedings of the Royal Society of London. cies as related to evolution of C4 photosynthesis. Plant Physiology 91 : Series B, Biological Sciences 279: 3304—3311. 1543-1550. Kadereit, G., T. Borsch, K. Weising, and H. Freitag. 2003. Phylogeny Carolin, R. C., S. W. L. Jacobs, and M. Vesk. 1978. Kranz cells and meso of Amaranthaceae and Chenopodiaceae and the evolution of C4 pho phyll in the Chenopodiales. Australian Journal of Botany 26: 683-698. tosynthesis. International Journal of Plant Sciences 164: 959-986. Christin, P. A., R. P. Freckleton, and C. P. Osborne. 2010. Can Kanai, phy R., and G. E. Edwards. 1999. The biochemistry of C4 photosyn logenetics identify C4 origins and reversals? Trends in Ecology thesis.& In R. F. Sage and R. K. Monson [eds.], C4 plant biology, 49-87. Evolution 25: 403-409. Academic Press, San Diego, California, USA.

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2400 American Journal of Botany [Vol.100

Kennedy, R. A., and W. M. O'Leary,Laetsch. M. H. 1988. 1974.Carbon isotopes Plant in photosynthesis. species Bioscience intermediate 38: for C3, C4 photosynthesis. Science 328-336. 184: 1087-1089. Kim, I. 1993. Biosystematics Prabhakar, of M., andPortulaca N. Ramayya. 1979. Anatomicalspecies studies inin the Indiannortheast Asia. Korean Journal of Plant Taxonomy Portulacaceae. II. Leaf 23: lamina. 71-96.In Histochemistry, Developmental and Kim, I., and D. G. Fisher. 1990. Structural Structural Anatomy of Angiosperms: aspects a Symposium, of 159-69. the P & Hleaves of seven species of Portulaca growing Publications, in Hawaii. Tiruchirapalli, India.Canadian Journal of Botany 68: 1803-1811. Pyankov, V. I., E. G. Artyusheva, G. E. Edwards, C. C. Black Jr., Ku, M. S. B., R. K. Monson, R. O. Littlejohn Jr., H. Nakamoto, D. and B. P. S. Soltis. 2001. Phylogenetic analysis of tribe Salsoleae Fisher, and G. E. Edwards. 1983. Photosynthetic characteristics (Chenopodiaceae) of based on ribosomal ITS sequences: Implications C3-C4 intermediate Flaveria species. Plant Physiology 71: 944-948. for the evolution of photosynthesis types. American Journal of Botany Ku, M. S. B., Y. J. Shieh, B. J. Reger, and C. C. Black. 1981. 88:1189-1198. Photosynthetic characteristics of Portulaca grandiflora, a succulent R Development Core Team. 2006. R: A language and environment for C4 dicot. Plant Physiology 68: 1073-1080. statistical computing [computer program]. Website www.R-project. Landrum, J. V. 2002. Four succulent families and 40 million years org of [accessed 30 April 2012], evolution and adaptation to xeric environments: what can stem Rajendrudu, and G., J. S. R. Prasad, and V. S. Rama Das. 1986. C3-C4 inter leaf anatomical characters tell us about their phylogeny? Taxon 51mediate : species in Alternanthera (Amaranthaceae). Plant Physiology 463-473. 80: 409-414. Legrand,D. 1953. Desmembración del género Portulaca. Comunicaciones Rambaut, A. 2002. Se-Al. Sequence alignment editor [computer pro Botánicas del Museo de Historia Natural de Montevideo 31: 1-14. gram]. Website http://tree.bio.ed.ac.uk/software/seal/ [accessed 21 Legrand, D. 1962. Las especies americanas de Portulaca. Anales del March 2012], Museo de Historia Natural de Montevideo, Ser. 2 1: 1-147. Rannala, B., and Z. H. Yang. 1996. Probability distribution of molecular Li, S., D. K. Pearl, and H. Doss. 2000. Phylogenetic tree construction evolutionary trees: A new method of phylogenetic inference. Journal using Markov chain Monte Carlo. Journal of the American Statistical of Molecular Evolution 43: 304-311. Association 95: 493-508. Raven, J. A., C. S. Cockell, and C. L. De La Rocha. 2008. The evolu Lin, C., J. Hsue, and J. Tsay. 1982. The anatomy of the noxious weeds tion of inorganic carbon concentrating mechanisms in photosynthesis. on cultivated land of Taiwan. Proceedings of the National Science Philosophical Transactions of the Royal Society of London. Series B, Council, Republic of China. Part B, Life Sciences 6: 452^160. Biological Sciences 363: 2641-2650. Matthews, J. F. 2003. Portulaca. In Flora of North America Editorial Rawsthorne, S. 1992. C3-C4 intermediate photosynthesis: linking physi Committee [eds.], Flora of North America, vol. 4, 496-501. Oxford ology to gene expression. The Plant Journal 2: 267-274. University Press, New York, USA. Reeves, J. H. 1992. Heterogeneity in the substitution process of amino McKown, A. D., and N. G. Dengler. 2007. Key innovations in the evolu acid sites of proteins coded for by mitochondrial DNA. Journal of tion of Kranz anatomy and C4 vein pattern in Flaveria (Asteraceae). Molecular Evolution 35: 17-31. American Journal of Botany 94: 382-399. Ronquist, F., M. Teslenko, P. van der Mark, D. L. Ayres, A. Darling, Monson, R. K. 1999. The origins of C4 genes and evolutionary pattern S. Hóhna, B. Larget, et al. 2012. MrBayes 3.2: Efficient Bayesian in the C4 metabolic phenotype. In R.F. Sage and R.K. Monson [eds.], phylogenetic inference and model choice across a large model space. C4 Plant Biology, 377-410. Academic Press. San Diego, California, Systematic Biology 61: 539-542. USA. Ruzin, S. E. 1999. Plant microtechnique and microscopy. Oxford Monson, R. K„ G. E. Edwards, and M. S. B. Ku. 1984. C3-C4 intermedi University Press, New York, New York, USA. ate photosynthesis in plants. Bioscience 34: 563-574. Sage, R. F. 2004. The evolution of C4 photosynthesis. The New Phytologist Monson, R. K., and B. D. Moore. 1989. On the significance of C3-C4 161:341-370. intermediate photosynthesis to the evolution of C4 photosynthesis. Sage, R. F., P. A. Christin, and E. J. Edwards. 2011. The C4 plant lineages Plant, Cell & Environment 12: 689-699. of planet Earth. Journal of Experimental Botany 62: 3155-3169. Monson, R. K., and S. Rawsthorne. 2000. C02 assimilation in C3-C4 Sage, R. F., T. L. Sage, and F. Kocacinar. 2012. Photorespiration and intermediate plants. In R. C. Leegod, T. D. Sharkey, and S. C. von the evolution of C4 photosynthesis. Annual Review of Plant Biology Caemmerer [eds.]. Photosynthesis: physiology and metabolism, 553 63: 19^47. 550. Kluwer, Dordrecht, Netherlands. Sage, R. F., T. L. Sage, R. W. Pearcy, and T. Borsch. 2007. The taxo Muhaidat, R., R. F. Sage, and N. G. Dengler. 2007. Diversity of Kranz nomic distribution of C4 photosynthesis in Amaranthaceae sensu anatomy and biochemistry in C4 eudicots. American Journal of Botany stricto. American Journal of Botany 94: 1992-2003. 94: 362-381. Stamatakis, A. 2006. RAxML-VI-HPC: Maximum likelihood-based Muhaidat, R., T. L. Sage, M. W. Frohlich, N. G. Dengler, and R. F.phylogenetic analyses with thousands of taxa and mixed models. Sage. 2011. Characterizarion of C3-C4 intermediate species in Bioinformaticsthe (Oxford, England) 22: 2688-2690. genus Heliotropium L. (Boraginaceae): Anatomy, ultrastructure Tavaré, and S. 1986. Some probabilistic and statistical problems on the anal enzyme activity. Plant, Cell & Environment 34: 1723-1736. ysis of DNA sequences. Lectures on Mathematics in the Life Sciences Nyananyo, B. L. 1988. Leaf anatomical studies in the Portulacaceae 17: 57-86. (Centrospermae) with regards to photosynthetic pathways. Tholen, Folia D., C. Boom, and X.-G. Zhu. 2012. Opinion: Prospects for im Geobotánica et Phytotaxonomica 23: 99-101. proving photosynthesis by altering leaf anatomy. Plant Science 197: Nyffeler, R., and U. Eggli. 2010. Disintegrating Portulacaceae: A 92-101.new familial classification of the suborder Portulacineae (Caryophyllales) Thorne, R. F., and J. L. Reveal. 2007. An updated classification of based on molecular and morphological data. Taxon 59: 227-240. the class Magnoliopsida ("Angiospermae"). Botanical Review 73: Nylander, J. A. A. 2004. MrModeltest, version 2 [computer program]. 67-181. Website http://www.abc.se/~nylander/mrmodeltest2/mrmodeltest2 Voznesenskaya, E. V., E. G. Artyusheva, V. R. Franceschi, V. I. Pyankov, [accessed 30 April 2012], O. Kiirats, M. S. B. Ku, and G. E. Edwards. 2001. Salsola arbuscu Ocampo, G., and J. T. Columbus. 2010. Molecular phylogenetics of subliformis, a C3-C4 intermediate in Salsoleae (Chenopodiaceae). Annals order Cactineae (Caryophyllales), including insights into photosyn of Botany 88: 337-348. thetic diversification and historical biogeography. American Journal Voznesenskaya, E. V., N. K. Koteyeva, H. Akhani, E. H. Roalson, and of Botany 97: 1827-1847. G. E. Edwards. 2013. Structural and physiological analyses in Ocampo, G., and J. T. Columbus. 2012. Molecular phylogenetics, histori Salsoleae (Chenopodiaceae) indicate multiple transitions among C3, cal biogeography, and chromosome number evolution in Portulaca intermediate and C4 photosynthesis. Journal of Experimental Botany (Portulacaceae). Molecular Phylogenetics and Evolution 63: 97-112. 64: 3583-3604.

VoZNESENSKAYA, E. V., N. K. KOTEYEVA, G. E. EDWARDS, DoesAND ancestralG. OcAMPO. trait reconstruction mislead? Evolution; International 2010. Revealing diversity in structural and biochemical Journal forms of ofOrganic Evolution 61: 1886-1899. C4 photosynthesis and a C3-C4 intermediate in genus Winter, Portulaca K., and L. J. A. M. Holtum. 2002. How closely do the §13C (Portulacaceae). Journal of Experimental Botany 61: 3647-3662.values of Crassulacean acid metabolism plants reflect the propor Welkie, G. W., and M. Caldwell. 1970. Leaf anatomy of tionspecies of inC02 some fixed during day and night? Plant Physiology 129: dicotyledon families as related to the C3 and C4 pathways 1843-1851.* of carbon fixation. Canadian Journal of Botany 48: 2135-2146. Yang, Z. 1993. Maximum-likelihood estimation of phylogeny from DNA Wiens, J. J. 2011. Re-evolution of lost mandibular teeth sequences in frogs when substitution after rates differ over sites. Molecular Biology more than 200 million years, and re-evaluating Dollo' s andlaw. Evolution Evolution; 10: 1396-1401. International Journal of Organic Evolution 65: 1283-1296. Yang, Z., and B. Rannala. 1997. Bayesian phylogenetic inference using Wiens, J. J., C. A. Kuczynski, W. E. Duellman, and T. W. DNA Reeder. sequences: 2007. a Markov Chain Monte Carlo Method. Molecular Loss and re-evolution of complex life cycles in marsupial Biology and frogs: Evolution 14: 717-724.

Appendix 1. Specimens used for this study. Information within parentheses indicates the carbon isotope value and the collection holding the specimen, followed by the GenBank accession numbers when available (ITS, ndhF, trnT-psbD, and ndhA intron). Herbaria acronyms: ARIZ = Department of Plant Sciences, University of Arizona, Tucson, Arizona, USA; BRI = Queensland Herbarium, Brisbane, Queensland, Australia; IEB = Centro Regional del Bajío, Instituto de Ecología, A.C., Pátzcuaro, Michoacán, Mexico; LE = Russian Academy of Sciences, V.L. Komarov Botanical Institute, Saint Petersburg, Russia; MO = Missouri Botanical Garden, Saint Louis, Missouri, USA; NY = New York Botanical Garden, Bronx, New York, USA; SI = Museo Botánico, San Isidro, Buenos Aires, Argentina; UNCC = Mecklenburg County Park and Recreation Herbarium, Charlotte, North Carolina, USA. Symbols: § = voucher for anatomical sections; * = data from Ocampo and Columbus (2010); # = data from Voznesenskaya et al. (2010); NA, not available.

Portulaca amilis Speg.; Fortunato et al. 6563, Argentina (-12.0, MO); Macedo (-11.6, MO). Portulaca grandiflora Hook.; Fortuna 20, Argentina (-12.2, 2719, Brazil (-13.6, MO); Ocampo et al. 1556, Argentina (-11.9*, RSA, LE). Portulaca grandiflora Hook, cv.; §Ocampo 1403cv, cultivated SI; JF508527, JF508674, JF508757, HQ241593). Portulaca armitii F. (-11.7, RSA; JF508550, JF508697, JF508780, JF508626). Portulaca Muell.; 5Ocampo et al. 1750, Australia (-13.2, BRI, RSA; JF508529, guanajuatensis G. Ocampo; Ocampo 1482, Mexico (-14.2*, RSA; JF508676, JF508759, JF508610). Portulaca australis Endl.; «Ocampo JF508551, JF508698, JF508781, HQ241599). Portulaca halimoides L.; et al. 1747, Australia (-13.5, BRI, RSA; JF508531,1F508678, JF508761, Acampo 1474, Mexico (-13.4, RSA; JF508552, JF508699, JF508782, JF508612). Portulaca bicolor F. Muell.; Ocampo et al. 1753, Australia JF508627). Portulaca hirsutissima Cambess.; Anderson et al. 35377, (-13.7*, BRI, RSA; JF508532, JF508679, JF508762, HQ241594). P. cf. Brazil (-25.8, MO); Belem 3754, Brazil (-27.8, NY); Irwin et al. 27580, bicolor; Acampo et al. 1726, Australia (-15.4#, BRI, RSA; JF508530, Brazil (-22.3; MO); Laessoe & Silva H52581, Brazil (-24.3, NY); JF508677, JF508760, JF508611). Portulaca brevifolia Urb.; Broadway Macguire et al. 44744, Brazil (-24.8, NY); Irwin et al. 27580, Brazil 204, Venezuela (-13.0, NY). Portulaca caulerpoides Britton & P. Wilson; (-22.3, NY; KC690150, -, -, -). Portulaca howellii (D. Legrand) Eliasson; Acevedo-Rodriguez & Siaca 4371, Puerto Rico (-13.0, NY); Wodbury et Howell 9782, Galápagos (-10.2, MO; JF508555, JF508702, JF508785, al. WI-122, Virgin Islands (-12.9, NY). Portulaca chacoana D. Legrand; HQ241601). Portulaca intraterranea J.M. Black; «Ocampo et al. 1748, Schinini & Bordas 18043, Argentina (-10.5, MO). Portulaca confertifolia Australia (-11.3, BRI, RSA; JF508556, JF508703, JF508786, JF508630). Hauman; «Ocampo et al. 1619, Argentina (-12.2, RSA, SI; JF508536, Portulaca johnstonii Henrickson; «Columbus 5076, Mexico (-13.7, RSA; JF508683, JF508766, JF508615); West 6127, Argentina (-10.0, MO). JF508557, JF508704, JF508787, JF508631). Portulaca lútea Sol. ex G. Portulaca conzattii P. Wilson; Standley 25053, Mexico (-11.8, NY). Forster.; Fosberg 36735, Marshall Islands (-11.5, LE). Portulaca massaica Portulaca cryptopetala Speg.; Conrad 2264, Paraguay (-24.8, MO); Degen S.M. Phillips; «Cruse-Sanders s.n., Tanzania (-15.2*, RSA; JF508559, 1209, Paraguay (-26.9, MO); Fernández Casas & Molero FC 4497, JF508706, JF508789, HQ241600). Portulaca matthewsii G. Ocampo; Paraguay (-23.8, MO); Jardim 1964, Bolivia (-27.0, MO); Kellogg s.n., «Ocampo 1425, Mexico (-12.8, RSA; JF508560, JF508707, JF508790, 14/Aug/1909, Venezuela (-26.1, MO); Maranta & Arenas 84, Argentina JF508633). Portulaca mexicana P. Wilson; «Ocampo & Morales 1461, (-26.4, NY); Morong 1053, Paraguay (-24.7, LE); Ocampo et al. 1540, Mexico (-12.1, RSA; JF508561, JF508708, JF508791, JF508634); Panti Argentina (-26.5*, RSA, SI; JF508538, JF508685, JF508768, HQ241596); Madero 187, Mexico (-13.0, MO). Portulaca minuta Correll; Correll & Pozner & Belgrano 332, Argentina (-24.9, MO); Saravia 575, Bolivia Sauleda 50027, Bahamas (-12.4, NY). Portulaca molokiniensis Hobdy; (-25.4, MO); Schinini & Palacios 25826, Argentina (-24.7, MO). Portulaca Perlman 12643, Hawaii (-15.5*, RSA; JF508562, JF508709, JF508792, cubensis Britton & P. Wilson; León & Loustalol 11349, Cuba (-13.6, NY); HQ241602). Portulaca mucronata Link; Balcazar 92, Bolivia (-25.6, Smithetal. 3138,Cuba(-13.9,NY).PortulacadecipiensPoelln.; «Ocampo MO); Irwin et al. 28781, Brazil (-27.3, NY); Irwin et al. 32639, Brazil et al. 1758, Australia (-12.9, BRI, RSA; JF508539, JF508686, JF508769, (-23.0, MO); Lewis 40712, Bolivia (-24.4, MO; KC690151, -, -, -); Liesner JF508617). Portulaca digyna F. Muell.; «Ocampo et al. 1749, Australia 6812, Venezuela (-27.9, MO); Plowman et al. 9358. Brazil (-29.9, NY); (-12.2, BRI, RSA; JF508540, JF508687, JF508770, JF508618). Portulaca Zardini 8211, Paraguay (-25.1, MO). Portulaca mucronulata D. Legrand; echinosperma Hauman; Ocampo et al. 1638, Argentina (-10.4*, RSA, SI; «Ocampo et al. 1598, Argentina (-11.3, RSA, SI; JF508563, JF508710, JF508541, JF508688, JF508771, HQ241597); Pedersen 15198, Argentina JF508793, JF508635); Schinini & Palacios 25724 (-12.7, MO). Portulaca (-12.9, MO). Portulaca elatior Mart, ex Rohrb.; Ocampo 1708cv, obtusa Poelln.; «Ocampoetal. 1591, Argentina (-13.0, RSA, SI; JF508565, Caribbean, cultivated (-12.7, RSA; JF508542, JF508689, JF508772, JF508712, JF508795, JF508637). Portulaca olerácea L. subsp. impolita HQ241598). Portulaca elongata Rusby; Bang 1140, Bolivia (-12.3, LE). Danin & H.G. Baker; «André 8501, USA (-12.4, RSA; JF508570, Portulaca eruca Hauman; Lossen 292, Argentina (-15.8, LE); «Ocampo JF508717, JF508800, JF508642). Portulaca olerácea subsp. nitida et al. 1645, Argentina (-15.0, RSA, SI; JF508543, JF508690, JF508773, Danin & H.G. Baker; «Ocampo et al. 1553, Argentina (-12.6, RSA, SI; JF508619). Portulaca filifolia F. Muell.; «Ocampo et al. 1733, Australia JF508572, JF508719, JF508802, JF508644). Portulaca olerácea subsp. (-13.3, BRI, RSA; JF508544, JF508691, JF508774, JF508620). Portulaca papillatostellulata Danin & H.G. Baker; «Ocampo & Columbus 1512, fluvialisD. Legrand; Brunner 1655, Paraguay (-11.0, MO); «Ocampo et al. Mexico (-12.5, RSA; JF508566, JF508713, JF508796, JF508638). 1581, Argentina (-12.5, RSA, SI; JF508545, JF508692, JF508775, Portulaca oligosperma F. Muell.; «Ocampo et al. 1751, Australia (-13.7, JF508621); Pedersen 3814, Argentina (-11.5, LE). Portulaca foliosa Ker BRI, RSA; JF508579, JF508726, JF508809, JF508651). Portulaca Gawl.; «Ocampo 1772cv, Tropical Africa, cultivated (-15.0, RSA; papulifera D. Legrand; Herter 50722, Uruguay (-11.4, MO); «Ocampo JF508546, JF508693, JF508776, JF508622). Portulaca frieseana Poelln.; et al. 1569, Argentina (-13.8, RSA, SI; JF508580, JF508727, JF508810, Irwin & Soderstrom 7297, Brazil (-13.6, MO). Portulaca fulgens Griseb.; JF508652). Portulaca papulosa Schltdl.; Biganzoli et al. 1733, Argentina «Ocampo et al. 1636, Argentina (-11.0, RSA, SI; JF508547, JF508694, (-12.6, MO); Gallinal et al. PE-4760, Uruguay (-12.4, MO). Portulaca JF508777, JF508623). Portulaca gilliesii Hook.; «Ocampo et al., 1545, perennis R.E. Fr.; Horn 48, Bolivia (-12.4, MO); «Ocampo et al. 1606, Argentina (-11.8, RSA, SI; JF508548, JF508695, JF508778, JF508624); Argentina (-13.1, RSA, SI; JF508581, JF508728, JF508811, JF508653). Pedersen 3743, Argentina (-14.4, LE); Schinini & Cristóbal, Argentina Portulaca pilosa L.; Curtiss 352, USA (-11.4, LE); Nortrup s.n., USA

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms 2402 American Journal of Botany

(-14.0*, UNCC; JF508585, JF508732, JF508815, HQ241603); Ocampo JF508665); Ricketson 4582, USA (-12.5, MO); Ricketson & Raechal et al. 1718, Australia (-13.8, BRI, RSA; JF508582, JF508729, JF508812, 4241, USA (-12.4, MO). Portulaca tingoensis J.F. Macbr.; 5Ocampo et al. JF508654). Portulaca pusilla Kunth; Davidse & Huber 15161, Venezuela 1615, Argentina (-12.5. RSA, SI; JF508598, JF508745, JF508828, (-14.6, MO). Portulacapygmaea Steyerm.; Groger 350, Venezuela (-15.6, JF508666). Portulaca tuberosa Roxb.; §Ocampo et al. 1737, Australia MO). Portulaca quadrifida L.; §Cruse-Sanders s.n., Tanzania (-13.0*, (-11.3, BRI, RSA; JF508599, JF508746, JF508829, JF508667). Portulaca RSA; JF508588, JF508735, JF508818, HQ241604); Schimper 1006, umbraticola Kunth subsp. umbraticola; §Ocampo et al. 1586, Argentina Arabia (-12.0, LE); Schweinfurth & Riva 2082, Ethiopia (-12.5, LE). (-12.5, RSA. SI; JF508603, JF508750, JF508833, JF508670). Portulaca Portulaca retusa Engelm.; 8Baker 16325, USA (-12.3, ARIZ; JF508590, umbraticola subsp. coronata (Small) J.F. Matthews & Ketron; 8Faircloth JF508737, JF508820, JF508659). Portulaca rotundifolia R.E. Fr.; s.n., USA (NA, UNCC; JF508601, JF508748, JF508831, JF508669). §Ocampo et al. 1611, Argentina (-13.2, RSA, SI; JF508591, JF508738, Portulaca umbraticola subsp. lanceolata J.F. Matthews & Ketron; Ocampo JF508821, JF508660); Taylor et al. 11195, Argentina (-11.6, MO). &Columbus 1527,Mexico(-14.0*,RSA;JF508602.JF508749,JF508832, Porta/aca ™bricauZ/.s Kunth; Gaumer603,Mexico(-13.7,NY); Hitchcock HQ241605); Palmer 6587, USA (-10.2, MO); Pringle s.n.. USA (-12.1, 6, USA (-11.0. MO); Proctor 35140, Cayman Islands (-13.3, MO); LE). Portulaca umbraticola cv. 'wildfire mixed'; §Ocampo 1485cv, Steyermark & Espinoza 112741, Venezuela (-13.2, MO). Portulaca cultivated (-12.6, RSA; JF508600, JF508747, JF508830, JF508668). rzedowskiana G. Ocampo; §Ocampo 1124, Mexico (-12.4, IEB; JF508593, Portulaca vi/ZosaCham.; Greenwell 19642, Hawaii (-11.8, NY). Portulaca JF508740, JF508823, JF508662). Portulaca samoensis Poelln.; Anderson wedermannii Poelln.; Harley et al. 15921, Brazil (-12.1. MO); Thomas 1117, Micronesia (-11.4, NY); Fosberg 25955, Micronesia (-12.2, NY). et al. 12818, Brazil (-12.5, MO). Portulaca sp. nov.; §Ocampo et al. 1754. Portulaca sclerocarpa A. Gray; Degener & Mull 35207, Hawaii (-14.5, Australia (-14.5, BRI, RSA; JF508596, JF508743, JF508826, JF508664). NY); Forbes 243, Hawaii (-11.9, MO); Fosberg 46059, Hawaii (-12.1, Outgroups: Pereskia aculeata Mill. (Cactaceae); NA, Americas, cultivated NY). Portulaca sedifolia N.E. Br.; Groger 1261, Venezuela (-14.3, MO); (-27.5*, ZSS; JF508526, JF508673, JF508756, HQ241587). Talinopsis Jansen-Jacobs et al. 4806, Guyana (-14.1, MO). Portulaca smallii P. frutescens A. Gray (Anacampserotaceae); Ocampo 1480, Mexico (-27.1*, Wilson; 8Herkenham s.n., USA (NA, UNCC; JF508595, JF508742, RSA; JF508607, JF508754, JF508837, HQ241613). Talinum paniculatum JF508825, JF508663). Portulaca suffrutescens Engelm.; §Ocampo & (Jacq.) Gaertn. (Talinaceae); Ocampo & Morales 1458, Mexico (-28.2*, Columbus 1505, Mexico (-11.6, RSA; JF508597. JF508744, JF508827, RSA; JF508608, JF508755, JF508838, HQ241618).

This content downloaded from 86.59.13.237 on Mon, 21 Jun 2021 11:25:32 UTC All use subject to https://about.jstor.org/terms