Diversification of Caiophora (Loasaceae Subfam

Org Divers Evol (2017) 17:29–41 DOI 10.1007/s13127-016-0312-4

ORIGINAL ARTICLE

Diversification of Caiophora (Loasaceae subfam. Loasoideae) during the uplift of the Central Andes

Marina Micaela Strelin1 & José Ignacio Arroyo2,3 & Stella Fliesswasser5 & Markus Ackermann4,5

Received: 25 May 2016 /Accepted: 16 November 2016 /Published online: 29 November 2016 # Gesellschaft für Biologische Systematik 2016

Abstract Andean orogeny and the ecological changes that solve phylogenetic relationships among 19 Caiophora species followed promoted diversification in plant and animal line- (including different accessions). We also included 10 species ages since the Early Miocene. The angiosperm genus of the allied genera Blumenbachia and Loasa. Aosa rostrata Caiophora (Loasaceae, subfam. Loasoideae) comprises and Xylopodia klaprothioides were used as outgroups. around 50 species that are endemic to South America. These Phylogenetic reconstruction strongly supports the monophyly are distributed from southern Ecuador to Central Chile and of Caiophora, and although several clades within this genus Argentina. Bee pollination and distribution at low- are poorly supported, our study yielded a better infra-generic intermediate elevations probably represent the ancestral con- resolution than previous studies. The origin of Caiophora is dition in the lineage that includes Caiophora and its allied dated to the Early-Middle Miocene and can be related to the genera. The majority of Caiophora species grow at high ele- uplift of the Cordilleras Frontal and Principal and to Pacific vations in the Andes, where some depend on vertebrate polli- marine transgressions. According to our estimations, nation. Previous studies did not resolve phylogenetic relation- Caiophora began to diversify during the Middle-Late ships within Caiophora, which precluded the dating of the Miocene and this unfolding proceeded eastwards during the origin and divergence of this group. We used markers of one Pliocene and the Pleistocene, in parallel to the uplift of differ- nuclear (ITS) and one plastid region (trnSGCU-trnGUUC)to ent Andean mountain ranges.

Keywords Diversification . Caiophora . Loasaceae, subfam. Electronic supplementary material The online version of this article Loasoideae . Diversification . Central Andes . Orogeny (doi:10.1007/s13127-016-0312-4) contains supplementary material, which is available to authorized users.

* Marina Micaela Strelin Introduction [email protected] The diversification of a lineage can be promoted by intrinsic factors, e.g. the acquisition of an evolutionary novelty 1 Laboratorio Ecotono, INIBIOMA (Universidad Nacional del Comahue-CONICET), Pasaje Gutierrez 1125, 8400 Bariloche, Rio (Minelli and Fusco 2012), or by extrinsic factors. The latter Negro, Argentina can be either abiotic, e.g. orogeny and climate change (Bryson 2 Departamento de Ecología, Facultad de Ciencias Biológicas, et al. 2012;Solàetal.2013; Meng and Zhang 2013), or biotic, Pontificia Universidad Católica de Chile, Santiago, Chile e.g. changes in the selective regimes (Nylin et al. 2014), and 3 Instituto de Ecología & Biodiversidad (IEB-Chile), Casilla 653, emergence of new ecological niches (Muschick et al. 2012). Santiago, Chile Among the most striking examples of lineage diversification 4 Nees Institut für Biodiversität der Pflanzen, Rheinische are those related to Andean orogeny (Gamble et al. 2008; Friedrich-Wilhelms-Universität, Meckenheimer Allee 170, Antonelli et al. 2009;Chavesetal.2011;Drummondetal. 53115 Bonn, Germany 2012a). 5 Institut für Integrierte Naturwissenschaften – Biologie, Universität The Andean uplift started about 40 Ma during the Middle Koblenz-Landau, Universitätsstr. 1, 56070 Koblenz, Germany Eocene, but the most pronounced uplift events took place in 30 Strelin M.M. et al. the Late Oligocene-Middle Miocene (∼25 to 17 Ma), in the high-Andean environments. The recent origin of a two- Late Miocene-Early Pliocene (∼9 to 5 Ma) and in the Early species clade in Schizanthus (Solanaceae) from a presum- Pliocene-Pleistocene (∼5–less than 5 Ma) (Lamb et al. 1997). ably bee-pollinated ancestor is related to plant specializa- These events are linked to considerable diversification in an- tion on hummingbird pollination in high-Andean environ- giosperm lineages, e.g. Smith and Baum (2006), Särkinen ments (Perez et al. 2006). et al. (2012), Drummond et al. (2012a), Baranzelli et al. Loasaceae subfamily Loasoideae is a monophyletic lineage (2014) and Moré et al. (2015). In addition to acting as a dis- largely restricted to Central and South America. It comprises persion corridor for plant lineages, the rising Andes resulted in 13 genera and more than 220 species, being the most species- the formation of vicariant barriers fostering allopatric specia- rich group within Loasaceae. Nasa Weigend, the largest genus tion in lineages growing on the eastern and the western within the subfamily, has its centre of diversity in the Andean slopes or inhabiting different inner Andean valleys Amotape-Huancabamba Zone (northern Peru) (Mutke et al. (Luebert and Weigend 2014). Glaciations, which resulted 2014). The second largest genus, Caiophora C. Presl from the interplay between the Andean uplift and periods of (Fig. 1), shows its highest diversity in the area from northern intense cooling (Mercer and Sutter 1982; Clapperton 1983; Peru to northern Argentina. The ca. 50 species in this genus Zech et al. 2009; Rubioldo et al. 2001; Martini et al. 2012), are distributed from Ecuador to Central Chile and Argentina, constituted additional vicariant barriers between the middle between 2000 and 5000 m a.s.l. This genus is associated to Miocene and the Pliocene and during the Pleistocene (e.g. different Andean mountain ranges, and only a single species, Cosacov et al. 2010; Sérsic et al. 2011). Glaciations also fos- Caiophora arechavaletae Urb. & Gilg., is found in southern tered northward dispersal and displacement of different plant Brazilian and Uruguayan grasslands (Zuloaga et al. 2008). lineages (Palazzesi et al. 2012). Some Caiophora species, e.g. Caiophora cirsiifolia C. Presl, The rising of the Andes fostered plant lineage diversi- present phenotypic variants, which were defined on the basis fication not only through the generation of vicariant bar- of flower characteristics, e.g. corolla shape and nectar compo- riers, but also through the generation of high-elevation sition, as well as inflorescence and vegetative traits areas, which were vacant for colonization, e.g. high- (Ackermann and Weigend 2006). Andean species of Lupinus (Hughes and Eastwood Bee pollination is presumably the ancestral condition in 2006). Besides the origin of these high-elevation areas, Caiophora and its allied genera (Strelin et al. 2016a). In addi- unprecedented ecological conditions in these new envi- tion to bee-pollinated species, Caiophora also includes spe- ronmentsmayhavealsocontributedtoplantlineagedi- cies that rely on hummingbird pollination (Ackermann and versification (Antonelli and Sanmartín 2011; Särkinen Weigend 2006; Weigend et al. 2004)aswellasasingle et al. 2012; Luebert and Weigend 2014). For instance, rodent-pollinated species, Caiophora coronata (pollinated while altitude is believed to act as an environmental filter by the small rodent G. griseoflavus)(CocucciandSérsic on the diversity of bee communities (Hoiss et al. 2012) 1998). Despite the absence of pollinator records for and insect pollination tends to be negatively affected by Caiophora pentlandii (PaxtonexGraham)G.Donex altitude in the Andes (Arroyo et al. 1985), hummingbirds Loudon, this species is also believed to be pollinated by small diversified in a context of Andean uplift since the middle rodents, based on its floral and inflorescence morphology Miocene (∼15 Ma) (McGuire et al. 2014). Furthermore, (Ackermann and Weigend 2006). Previous work showed that some of these hummingbird lineages, e.g. Ensifera flower shape in Caiophora and allied taxa evolved in response ensifera and Patagona gigas, adapted their flight dynam- to bee, hummingbird and rodent pollination (Strelin et al. ics and their body size to high-elevation conditions in the 2016a). Andes (Altshuler et al. 2004). The Andean uplift also Although the monophyly of Caiophora was demonstrated fostered diversification in other vertebrate lineages, e.g. in previous phylogenetic analyses (Hufford et al. 2003, 2005; small rodents of the Sigmodontinae subfamily (Martinez Weigend et al. 2004), these studies included a restricted sam- et al. 2012 ). In fact, there is evidence that some rodent pling of this genus (10 out of the ∼50 Caiophora species in lineages in this subfamily, e.g. Graomys griseoflavus Hufford et al. 2005). Phylogenetic relationships between spe- (Cavides-Vidal et al. 1987)andPhyllotis xanthopygus cies of Caiophora are poorly resolved, presumably because (Nespolo et al. 1999), evolved adaptations to resist the previous phylogenetic inferences were based only on partial extreme environmental conditions characteristic of high- plastid DNA evidence. This impedes dating the origin of Andean environments. Since transitions from insect to Caiophora and the diversification events within the genus, hummingbird pollination are likely to take place in envi- which is crucial to better understand the effects that orogenic, ronments where insects are poorly represented (Cruden climatic and ecological factors may have had on the evolution 1972), adaptation to hummingbird pollination, and even- of this genus. tually to pollination by other vertebrates, may have been In this study, we present a phylogenetic reconstruction of an important factor in driving lineage diversification in the genus Caiophora, including 19 species (with four Diversification of Caiophora (Loasaceae subfam. Loasoideae) 31

Fig. 1 Flower diversity in Caiophora (Loasaceae, subfam. Loasoideae) (Ancash/Huaraz form), j C. canarinoides (Lenné & K. Koch) Urb. & and in allied lineages. a Loasa acerifolia Dombey ex Juss., b L. bergii Gilg., k C. hibiscifolia (Griseb.) Urb. & Gilg., l C. lateritia Klotzsch, m Hieron., c Caiophora arechavaletae (Urb.) Urb. & Gilg, d C. cernua C. coronata (Gillies ex Arn.) Hook. & Arn., n C. pentlandii (Paxton ex (Griseb.) Urb. & Gilg ex Kurtz, e C. stenocarpa Urb. & Gilg., f Graham) G. Don ex Loudon, o C. chuquitensis (Meyen) Urb. & Gilg and C. cirsiifolia C. Presl (Arequipa form), g C. carduifolia C. Presl (Cuzco p C. clavata Urb. & Gilg. Scale bar equals 10 mm form), h C. carduifolia C. Presl (Apurimac form), i C. cirsiifolia C. Presl

phenotypic variants of C. cirsiifolia and two of Caiophora the origin of Caiophora and the divergence events within this carduifolia). We used both nuclear and plastid DNA evidence. genus and (4) to reconstruct the evolution of pollination Our specific aims are (1) to explore whether monophyly of modes in the studied species. Finally, the results obtained for Caiophora holds after estimating a phylogeny that includes Caiophora were compared to the results of other studies ad- DNA samples and more species of the genus, (2) to solve dressing diversification in South American angiosperm line- phylogenetic relationships within Caiophora, (3) to date both ages during Andean uplift. 32 Strelin M.M. et al.

Materials and methods MgCI2 and 1 μl of 1:20 diluted freshly prepared genomic DNA, and a final volume of 25 μl was obtained with the Taxon sampling and outgroup selection addition of ddH2O. The PCR for trnSGCU-trnGUUC was initi- ated with a 95 °C denaturing cycle for 5 min, followed by A total of 19 species of Caiophora (including four forms of 34 cycles of 95 °C for 30 s, 50 °C for 1 min and 72 °C for C. cirsiifolia and two forms of C. carduifolia)wereincluded 90 s, and the reaction was ended with an extension cycle of in this study (Supplementary file 1: Table S1). The sampling 72 °C for 4 min. PCR amplicons of ITS1 and ITS2 and of the species was aimed at having a good coverage of the trnSGCU-trnGUUC were sequenced in both directions by distribution range of Caiophora (Supplementary file 2: GATC (Constance, Germany) with the same primers used Table S2) and of the pollination modes presented by species for amplification. Forward and reverse sequences were exam- in this genus, i.e. bee, hummingbird and small rodent pollina- ined, compared and corrected using the program PhyDe tion (Ackermann and Weigend 2006; 2012; Supplementary (Müller et al. 2010). The GenBank accession numbers of the file 3: Table S3). These pollination modes are in turn markers included in this study are shown in Supplementary associated with different flower morphologies (Strelin et al. file 1: Table S1. 2016a). Studied species belong to 9 out of the 10 infra-generic Caiophora groups, which were defined on the basis of morphological characters by Ackermann (2012), i.e. Alignment and phylogenetic reconstruction C. arechavaletae group, Caiophora nivalis group, Caiophora pterosperma group, Caiophora chuquitensis group, Caiophora Alignments were performed using the program Mafft ver- contorta group, C. coronata group, Caiophora clavata group, sion 7 (Katoh and Standley 2013), with the default strat- C. cirsiifolia group, Caiophora lateritia group and C. clavata egy L-INS-i. To test for substitution saturation in the group. The only group that we were not able to include was the markers used in the phylogenetic analyses described in C. contorta group, composed of two species (Ackermann the following, we employed the substitution saturation 2012). The C. chuquitensis group is poorly represented, since approach of Xia et al. (2003). This approach consists in we were able to include only a single representative, calculating an index of substitution saturation (ISS) for a C. chuquitensis (Meyen Urb. & Gilg), of this eight-species set of markers, which represents the observed value of group. Furthermore, eight species of Loasa Adans. and three substitution saturation. ISS is then statistically compared species of Blumenbachia Schrad, representing different infra- to a critical saturation value (ISS.C), beyond which the generic groups, were included. Aosa rostrata (Urb.) Weigend markers start failing to recover the correct tree (Xia and Xylopodia klaprothioides Weigend were used as et al. 2003). In practice, if ISS is significantly lower than outgroups. ISS.C, saturation is assumed to be low enough to reconstruct the phylogenetic history of a lineage using that set Sequence data of markers. Technical details about this approach can be consulted in Xia et al. (2003). Furthermore, we visually ex- We used the DNeasy Plant Mini Kit (Qiagen, Hilden, amined the relationship between the uncorrected genetic dis- Germany) to extract total genomic DNA from silica-dried leaf tances in our alignment and the distances corrected with the material. The internal transcribed spacer of the nuclear ribo- F84 model of DNA substitution (Felsenstein 1984). somal repeat (ITS1 and ITS2) and the trnSGCU-trnGUUC Uncorrected and corrected distances are expected to scale lin- intergenic spacer were amplified using the following proto- early if there is no substitution saturation. Saturation analyses cols: For ITS1 and ITS2, we used the forward primer were done with the software DAMBE (Xia and Xie 2001). ITS5_bryo (Stech et al. 2013) and the reverse primer We estimated phylogenetic relationships among Loasaceae ITS4_bryo (Stech et al. 2013). Occasionally, we used the in- species based on a combined ITS/ trnSGCU-trnGUUC align- ternal forward primer SEQITS2_Angio (5′-AACG ment and on each separate marker, using Bayesian inference ACTCTCGGCAACGG-3′) and the internal reverse primer (BI) as implemented in MrBayes v3.1.2 (Ronquist and SEQITS1_Angio (5′-TTGCGTTCAAAGACTCGATGG-3′). Huelsenbeck 2003). The program was used to simultaneously Polymerase chain reactions (PCRs) were done according to estimate tree topology and parameter values for an indepen- the protocol of Wen and Zimmer (1996). For trnSGCU- dent GTR+Γ model of nucleotide substitution. Two simulta- trnGUUC, we used the forward primer trnS and the reverse neous and independent runs were performed for 10 × 106 it- primer trnG (Hamilton 1999). All PCR reactions used 1× erations of a Markov Chain Monte Carlo (MCMC) algorithm. GoTaq®Flexi buffer (Promega GmbH, Mannheim, Each run included six simultaneous chains, sampling every Germany), 0.75 units of GoTaq® Flexi DNA polymerase 1000 generations. Support for the nodes and parameter esti- (Promega GmbH, Mannheim, Germany), 0.05 nmol of for- mates were derived from a majority-rule consensus of the last ward and reverse primers, 0.1 mM of each dNTP, 1 mM of 5000 trees sampled after convergence. The average standard Diversification of Caiophora (Loasaceae subfam. Loasoideae) 33 deviation of split frequencies remained <0.01 after the burn-in non-parametric rate smoothing (NPRS) (Jobb et al. 2004; threshold. Sanderson 1997). Briefly, while NPRS maintains the rates at

To account for the sensitivity of the marker and the recon- each node (ri), as similar as possible to their respective ances- struction method, we performed separate phylogenetic recon- tral rates (r(A)i), LRMD maintains real rates as similar as pos- structions for the ITS and the trnSGCU-trnGUUC alignments, sible to ideal local rates, accounting for the similarity of rates using alternative methods of phylogenetic reconstruction: among sister groups. For NPRS and LRMD, we used maximum parsimony (MP), neighbour joining (NJ) and max- Treefinder version March 2011 (Jobb et al. 2004). The sec- imum likelihood (ML), as implemented in the programs ondary calibration points remain fixed in the LRMD and the PAUP version 4 (for MP and NJ) (Swofford 2003)and NPRS estimations. Since the calibration point corresponding Treefinder version March 2011 (Jobb et al. 2004), respective- to the common ancestor of C. chuquitensis and L. pinnatifida ly. For MP, a heuristic search, saving up to 1000 trees, was represents the origin of Caiophora, this event was not estimat- performed, using the branch-swapping BTree bisection and ed with LRMD and NPRS. BEAST v1.8.1 (Drummond et al. reconnection^ (TBR) algorithm. For NJ, a GTR+Γ model 2012b) was used in the Bayesian approach. Secondary cali- was used, with a gamma shape parameter of 0.5 and five rate bration points were included in the form of a normal distribu- categories. For ML, we searched for the best fitting model for tion with a standard deviation of 10 Ma, to account for the each marker and used a partition when estimating the phylog- range of the distribution estimated in Schenk and Hufford eny with the concatenated alignment. The best fitting models (2010). The Bayesian MCMC analysis was run for were TN+Γ (for ITS) and TVM+Γ (for trnSGCU-trnGUUC). 50.0 × 106 iterations, using a GTR substitution model (empir- We finally obtained consensus trees for ITS, trnSGCU- ical base frequencies), a gamma site heterogeneity model with trnGUUC and concatenated markers, in order to summarize four gamma categories, an exponential log-normal relaxed the tree topologies recovered with the four methods of phylo- clock (default) and a Yule speciation process tree prior. We genetic reconstruction. Although we based our conclusions on sampled every 1000 generations, which yielded a total of the concatenated alignment using BI, since this is a highly 50,000 trees, of which the first 50% was subsequently consistent and efficient method of phylogenetic reconstruction discarded. The maximum posterior credibility tree was there- (Yang and Rannala 2012), we also briefly discussed the topol- fore constructed with the remaining 25,000 trees. The ogies obtained with separate markers, using other approaches BEAUTI file is provided as supplementary information of phylogenetic reconstruction. (Supplementary file 4: Appendix S1).

Molecular clock test and molecular dating Reconstruction of the evolution of pollination modes

Before dating, we first tested concatenated markers for the We obtained pollinator data from Ackermann and Weigend existence of a molecular clock, using the test of molecular (2006), Weigend and Gottschling (2006) and other litera- clock implemented in MEGA version 5.2.2 (Tamura et al. ture (Supplementary file 3: Table S3). A pollination mode 2011). Three calibration points were employed for molecular (hummingbird, bee and rodent pollination) was assigned to dating, which were obtained from a previous study that in- all species based on these records. Species with ambiguous cluded species of the Loasaceae family (Schenk and Hufford pollinator records, i.e. reported to be pollinated by hum- 2010): the common ancestor of C. chuquitensis (Meyen) Urb. mingbirds and bees, C. chuquitensis, C. lateritia and &Gilg.(Caiophora macrocarpa in that study) and A. rostrata Caiophora hibiscifolia, were assigned either to bee or to (40.21 Ma [16.95–50.43]), the common ancestor of hummingbird pollination, based on the study of Strelin C. chuquitensis and Loasa heterophylla Hook & Arn. et al. (2016a). We pruned the maximum posterior credibil- (28.8 Ma [12.92–40.62]) and the common ancestor of ity tree to the 25 species for which pollinator data was C. chuquitensis and Loasa pinnatifida Gillies ex Arn. available. To reconstruct the history of pollination modes (Loasa filicifolia in that study) (22.53 Ma [9.97–36.11]). The in the studied species, we conducted stochastic character numbers in square brackets are the minimum and the maxi- mappings (Nielsen 2002)usingthemake.simmap function mum of the probability distribution of estimated ages in from the R package phytools (Revell 2012). Given a phy- Schenk and Hufford (2010). Bayesian posterior probability logeny and discrete character states for extant terminal for the first node was 0.59 and above 0.95 for the remaining taxa, this Bayesian method applies a Monte Carlo algo- two. rithm to sample the posterior probability distribution of We used three molecular dating methods to estimate ancestral states on the branches of a phylogeny under a the divergence time among species of our phylogeny. Markov process of evolution. One thousand stochastic The obtained ML best tree topology, with a 0.0001 minimum character mapping simulations were performed on the edge length and 1000 time replicates, was used both for local pruned maximum posterior credibility BEAST tree, and rate minimum deformation (LRMD) (Jobb et al. 2004)and the posterior probability of each character state plotted on 34 Strelin M.M. et al. the corresponding node (Revell 2014), using the a relatively well-supported clade with C. coronata,and describe.simmap function from the R package phytools C. carduifolia (2) forms a relatively well-supported clade with (Revell 2012). C. chuquitensis. C. cirsiifolia (1) and C. cirsiifolia (2) constitute another relatively well-supported clade, which is sister to C. coronata-C. carduifolia (1). C. cirsiifolia (3) and Results C. cirsiifolia (4) also constitute a relatively well-supported clade, sister to C. chuquitensis-C. carduifolia (2). Phylogenetic relationships Furthermore, C. coronata and C. pentlandii (both assigned to the C. coronata group) are not monophyletic (Fig. 2). In the saturation test, the Xia index showed significant no Species in the C. lateritia and the C. clavata groups are not saturation of markers (Supplementary file 5: Table S4)and monophyletic but together constitute a monophyletic group saturation plots showed a linear relationship between the ob- (Fig. 2). Only the C. pterosperma group (Caiophora served and the marker distances expected under the F84 mod- stenocarpa Urb. & Gilg. and Caiophora dederichiorum el (Supplementary file 6: Fig. S1). This suggests that the Mark. Ackermann & Weigend) is monophyletic and highly markers in our data set can be used to reconstruct the evolu- supported (Fig. 2;Table1). tionary history of the species in this study. Although phylogenetic topologies varied when using Divergence times in Caiophora different markers and methods of phylogenetic reconstruction (Supplementary files 7–9: Figs. S2, S3 and S4), all According to the LRT, concatenated markers did not fit well to a phylogenetic reconstructions recovered the monophyly of strict clock model (lnL ratio test statistic (LRTS) = 79.93; df = 34; Caiophora (Supplementary file 10: Fig. S5), suggesting a P < 0.01). The same result was obtained for non-concatenated single origin for this group. Despite the fact that some ITS and trnSGCU-trnGUUC markers (LRTS = 343.55; df = 34; Caiophora clades are well supported, several infra- P < 0.01 and LRTS = 67.86; df = 34; P < 0.01, respectively). generic clades are poorly supported (Fig. 2). All phyloge- According to BI using relaxed clock, Caiophora originated netic reconstructions recovered a clade, which includes the spe- 16.05 Ma (95% highest posterior density (HPD) 7.02–26.12). cies Caiophora aconquijae Sleumer, Caiophora dumetorum The dating of the diversification of the genus using different Urb. & Gilg., C. clavata Urb. & Gilg. and C. lateritia calibration methods was estimated as follows: 10.43 Ma (95% Klotzsch. C. arechavaletae was always recovered as a HPD 4.37–17.64) (BI), 5.02 Ma (LRMD) and 15.17 Ma lineage diverging early from the remaining Caiophora (NPRS). According to BI, most of the divergence events in clade (Fig. 2, Supplementary files 7–9: Figs. S2, S3 Caiophora took place during the Late Pliocene and the and S4). Loasa was never recovered as a monophyletic Pleistocene, i.e. during the last 5 million years (Fig. 2;Table1). group, and Loasa Ser. Pinnatae was always recovered as sister For BI estimation, effect sample sizes (ESSs) varied between 458 to Caiophora (Fig. 2, Supplementary files 7–9: S2, S3 and and 2829. S4). Reconstructions with the ITS region using different methods yielded remarkably different tree topologies and re- vealed that the majority of the groups within Caiophora are Reconstruction of the evolution of pollination modes poorly supported (Supplementary file 8: Fig. S3). Reconstructions with trnSGCU-trnGUUC resulted in relatively Hummingbird pollination is not the ancestral condition within well-supported groups within Caiophora,withLoasa ser. Caiophora, since the species that diverged earlier from the Pinnatae being always retrieved as sister to Caiophora remaining Caiophora clade are bee pollinated, and their re- (Supplementary file 9: Fig. S4). constructed ancestors are likely to have been bee pollinated According to the Bayesian inference (BI) analysis (Fig. 2). There is uncertainty regarding how many times hum- using concatenated markers, Caiophora is recovered as a mingbird pollination evolved from bee pollination in monophyletic and highly supported genus (Fig. 2; Caiophora, since the common ancestor of the three lineages Table 1); species in Loasa ser. Pinnatae constitute a clade presenting hummingbird pollinated species (C. carduifolia that is sister to Caiophora. C. arechavaletae and (1)-C. chuquitensis; C. hibiscifolia-Caiophora canarinoides- C. nivalis Lillo are recovered as a lineage diverging early Caiophora madrequisa; C. carduifolia (2)-C. coronata)had from the remaining Caiophora species. Most of the infra- equal probabilities of being bee or hummingbird pollinated generic groups proposed for Caiophora (Ackermann (Fig. 2). It is likely that there was a transition from humming- 2012), which were defined on the basis of floral and veg- bird to rodent pollination in C. coronata (Fig. 2) and that there etative morphology, are not recovered as monophyletic was a reversion from hummingbird to bee pollination in the (Fig. 2). The C. carduifolia and the C. cirsiifolia groups are lineage giving rise to Caiophora cernua, C. clavata, not monophyletic (Fig. 2), since C. carduifolia (1) constitutes C. dumetorum and C. lateritia (Fig. 2). Diversification of Caiophora (Loasaceae subfam. Loasoideae) 35

C. aconquijae 344 * Assigned groupp 33 C. lateritia * and pollinator 32 C. dumetorum # Assigned group 31 C. clavata # & Pinnatae * Lateritia 29 C. cernua # # Clavata C. madrequisa $ Pterosperma 28 30 * C. buraeavii + Coronata * ¤ Loasa 26 C. canarinoides ¢ Angulatae * £ Gripidea 27 C. chuquisacana # § Macrospermae 22 C. hibiscifolia ¥ Floribundae 5.64 * C. cirsiifolia (2) ¶ Volubiles 25 ₣ ~ Pusillae 23 C. cirsiifolia (1) ₣ ₣ Cirsiifolia 24 C. carduifolia (1) Ω Carduifolia 17 Ω

╩ Arechavaletae C. coronata + Caiophora ◊ Nivalis C. cirsiifolia (3) € Chuquitensis 20 ₣ % Klaprothieae 19 C. cirsiifolia (4) ₣ Pollinator 15 18 C. pentlandii Puna + Plateau-Altipl Bees C. chuquitensis 21 € ano Hummingbirdsr C. carduifolia (2) Ω Small rodentsnt 13 C. dederichiorum Unknown pollinatoro 10.43 16 $ C. stenocarpa $ 10 14 C. arechavaletae ╩ 16.05 5.98 C. nivalis ◊ L. bergii (1) 9 12 & 11 L. bergii (2) & L. pinnatifida & Cordillera Occidental

Loasa Cordillera Central 6 L. martini ¶ Cordillera Oriental (arid) Cordillera Oriental (yungas) 8 B. insignis £ Atacama dessert 7 B. amana £ Sierras Subandinas 2 B. silvestris ¢ Sierras Pampeanas Precordillera L. heterophylla 5 § Cordillera Frontal 4 L. acerifolia Blumenbachia Cordillera Principal 1 § Patagonian Andes 3 L. pallida ¥ Pampas grasslands L. acanthifolia ¤ Loasa Chilean scrub Central Brazilian sabanas A. rostrata ~ Aosa X. klaprothioides % Xylopodia

Cenozoic Paleogene Neogene Qua Eocene Oligocene Mio Plio Plei 50 40 30 20 10 Myr ago 0

Fig. 2 Phylogenetic relationships, geographic distribution and markers. Node support and dating values in Caiophora and basal pollination modes of Caiophora (Loasaceae, subfam. Loasoideae) and genera can be consulted in Table 1, using the reference number on each allied genera (Loasa and Blumenbachia). Geographic distribution node. A reconstruction of ancestral pollination modes using stochastic corresponds to the information in Supplementary file 2: Table S2.The character mapping is shown. Pie charts on each node indicate the topology of the phylogeny is the maximum posterior credibility tree of a posterior probability of each pollinator, retrieved by 1000 stochastic Bayesian analysis performed with ITS1, ITS2 and trnSGCU-trnGUUC character mappings

Discussion reflect shared adaptations. Some of the traits used to define infra-generic groups in this genus, e.g. flower morphology Phylogenetic relationships and colouration, as well as inflorescence architecture (Ackermann 2012), may have been under selection to some This study supports the monophyly of Caiophora and, in ac- extent. For instance, variation in the morphology and the cordance with previous studies (Hufford et al. 2003, 2005), colouration of the corolla is related to different functional supports the fact that Loasa ser. Pinnatae is the Loasa group groups of pollinators in the literature, e.g. Fenster et al. more closely related to Caiophora (Fig. 2). Although some (2004). Further, the morphology and the colouration of the Caiophora clades are poorly supported, our study yielded a staminodial complex, a flower structure unique to Loasoideae better resolution of infra-generic clades than previous works that participates in nectar harvesting (Ackermann and Weigend (Weigend et al. 2004; Hufford et al. 2003, 2005). Low node 2006) and in flower-pollinator mechanical fit (Weigend and support values in several Caiophora clades may be a conse- Gottschling 2006; Weigend et al. 2010), differ between bee quence of the reduced number of species and genetic markers and hummingbird pollinated species (Weigend and used in this study, but it may also suggest an adaptive radiation Gottschling 2006). Indeed, Strelin et al. (2016a)demonstrated in this genus, since the same macroevolutionary pattern was that the shape of the corolla and the staminodial complex are recovered for other Andean angiosperm lineages undergoing under pollinator selection in Caiophora. Interestingly, while such diversification events (e.g. Smith and Baum 2006; Smith species in the C. clavata and the C. cirsiifolia groups are mainly et al. 2008;Moréetal.2015). pollinated by bees, hummingbird pollination appears to be im- The infra-generic Caiophora groups, which were de- portant for species in the C. lateritia and the C. carduifolia fined on the basis of floral and vegetative morphology groups (Fig. 2; Supplementary file 3: Table S3). Concluding, (Ackermann 2012), were not recovered in the phylogeny, some of the trait combinations used in the previous delimitation with the exception of the C. pterosperma group, which has a of Caiophora infra-generic groups may reflect adaptations to high support (Fig. 2; Table 1). Rather than reflecting phyloge- different functional groups of pollinators rather than shared an- netic proximity, infra-generic groups in Caiophora may cestry. Nevertheless, a lack of correspondence between plant 36 Strelin M.M. et al.

Table 1 Node support and dating values in Caiophora (Loasaceae, C. cirsiifolia (1)-C. cirsiifolia (2), C. cirsiifolia (3)- subfamily Loasoideae) and allied genera C. cirsiifolia (4), C. carduifolia (1) and C. carduifolia (2) rep- Node MP/NJ/ML/BI support Age 95% HPD resent lineages with independent evolutionary trajectories.

1 – 44.68 – Timing and context of Caiophora diversification 2 –/–/99.6/0.99 26.97 13.74–41.00 3 –/–/–/– 20.76 9.006–35.39 Three types of methods to estimate divergence times are 4 –/–/67.8/0.86 16.97 5.97–29.25 often recognized: methods that assume a global rate, 5 100/100/100/1 5.86 0.91–12.51 methods that assume local rates and methods that account 698.9/–/–/0.75 23.81 10.33–36.82 for rate heterogeneity through the phylogeny. Because our 7 98.97/100/99.9/1 11.83 3.29–22.18 molecular clock analysis showed that there is no global 8 100/100/100/1 3.72 0.60–8.53 rate, we used methods of local rates (LRMD) and rate het- 9 74.93/68.59/98.5/1 20.09 8.27–31.97 erogeneity. Among the latter, we used two different 10 100/100/100/0.99 16.05 7.02–26.12 methods, (1) NPRS, a method that models the rate accord- 11 100/100/99.7/1 4.81 0.59–10.71 ing to a standard Poisson process and the phylogenetic 12 94.62/100/98.6/0.97 2.12 0.20–5.38 autocorrelation of rates and (2) the Bayesian relaxed clock 13 58.53/56.09/71.4/1 10.43 4.37–17.64 approach implemented in BEAST. The differences that we 14 –/–/–/0.76 5.98 1.11–11.99 found in the divergence time estimation with these three 15 71.5/–/–/0.61 8.43 3.36–14.35 methods are attributable to the differences in their assump- 16 88.82/87.9/96/0.99 4.03 0.53–8.70 tions and statistical implementations (see Rutschmann 17 –/53.59/–/0.61 6.74 2.58–11.63 2006, for a more detailed discussion on methods of 18 –/50.29/–/0.74 3.85 3.36–14.35 divergence time estimations). 19 –/–/–/– 1.98 0.13–5.06 The origin and diversification of Caiophora can be re- 20 –/–/66.3/0.86 0.67 0.05–2.2 lated to a variety of environmental events taking place 21 –/–/51.1/0.98 1.77 0.22–4.21 since the Early Miocene (Table 2). Our results suggest that 22 –/–/–/0.61 5.64 2.12–9.77 Caiophora originated during the Early-Middle Miocene, 23 –/–/–/0.83 3.18 0.55–6.37 approximately 16 Ma. This period coincides with the de- ∼ 24 –/–/–/0.87 1.66 0.19–3.90 formation and uplift of the Cordillera Frontal ( 17 to 25 –/60.79/–/0.83 1.30 0.09–3.48 14 Ma) (Levina et al. 2014) and the Cordillera Principal ∼ 26 –/–/–/0.53 4.65 1.69–8.21 ( 20 to 9 Ma) (Ramos et al. 2002; Giambiagi and Ramos 27 –/–/–/0.76 2.51 0.38–5.34 2002). Interestingly, while the distribution of Caiophora 28 –/–/–/0.53 3.58 1.22–6.53 encompasses these two mountain ranges, which are in turn 29 –/–/–/0.53 2.76 1.22–6.53 the southern limit of the distribution of this genus (Fig. 2; 30 –/–/–/– 1.72 1.22–6.53 Supplementary file 2: Table S2), species in Loasa ser. 31 –/70.4/50.6/0.53 1.83 0.48–3.62 Pinnatae (sister to Caiophora in our phylogeny), as well as the majority of the sampled species in clades that are 32 86.1/70.4/50.06/0.99 1.14 0.22–2.46 successive sister groups to Caiophora,onlygrowtothe 33 96.31/–/87/1 0.34 0.03–0.94 south of the Cordillera Frontal and the Cordillera 34 –/–/–/– 0.15 0.03–0.45 Principal (Fig. 2; Supplementary file 2: Table S2). Caiophora group values are in italics. Only node support values >50%/ Moreover, Scyphanthus Sweet (not included in this study), 0.5 are shown which according to Hufford et al. (2003, 2005) is the genus MP maximum parsimony, NJ neighbour joining, ML maximum likeli- most closely related to Caiophora, also grows to the south hood, BI Bayesian inference of these mountain ranges (Ackermann 2012). This suggests that Caiophora may have originated as a consequence of phenotype and phylogeny may, at least in part, also be a con- lineage isolation in northern and higher-elevation areas, sequence of interspecific hybridization, which was already re- during the uplift of the Cordillera Frontal and the ported for some Caiophora species (Ackermann et al. 2008). Cordillera Principal. Pacific marine transgressions affect- DuetothelowsamplesizeofCaiophora species in this study, ing central and northern Patagonia during the Early and we are not able to arrive at taxonomic conclusions concerning Middle Miocene (∼23–16 Ma) (Bechis et al. 2014)may the different positions of the C. cirsiifolia and C. carduifolia have also fostered this split (Fig. 2; Supplementary file 3: accessions in the phylogeny. Nevertheless, since these species Table S3). were previously defined based only on plant morphology According to our estimations, Caiophora began to diversi- (Ackermann 2012), it would not be surprising that fy about 5–15 Ma, depending on the dating method. If the Diversification of Caiophora (Loasaceae subfam. Loasoideae) 37

Table 2 Summary of evolutionary events, i.e. origin and divergence of different groups, in Caiophora and allied genera, along with the environmental events hypothesized to be involved in these evolutionary events

Evolutionary event Explanatory environmental event Geological epoch (environmental event)

Origin of Caiophora (16.05 Ma [7.02–26.12]) (BI) Main uplift event of the Cordillera Principal (∼20 to 9 Early to Late Miocene Ma; Ramos et al. 2002 and Giambiagi and Ramos 2002). Uplift event of the Cordillera Frontal (∼17 to 14 Ma; Levina et al. 2014). Origin of Caiophora (16.05 Ma [7.02–26.12]) (BI) Pacific marine transgressions in central and northern Early to Middle Miocene Patagonia (∼23 to 16 Ma; Bechis et al. 2014) Diversification of Caiophora (10.43 Ma [4.37–17.64]) Diversification of the Andean hummingbird lineage Middle Miocene (BI); 5.02 Ma (LRMD); 15.17 Ma (NPRS) (∼15 Ma; McGuire et al. 2014). Diversification of Caiophora (10.43 Ma [4.37–17.64]) Uplift of the Cordillera Oriental and the Sierras Subandinas Middle Miocene to present (BI); 5.02 Ma (LRMD); 15.17 Ma (NPRS) (eastern Andean slope). The Puna-Altiplano Plateau rose from 1500 to 4000 m. (∼10 Ma to present: Armijo et al. 2015). Diversification of Caiophora (10.43 Ma [4.37–17.64]) The Puna-Altiplano Plateau rose ∼2.5 km (∼10 to 7 Ma; Late Miocene to present (BI); 5.02 Ma (LRMD); 15.17 Ma (NPRS) Garzione et al. 2006; Hoke and Garzione 2008). The Puna-Altiplano Plateau rose from 1500 to 2500 ma.s.l(∼9to5Ma:Lambetal.1997). Uplift of the Sierras Subandinas (east to the Puna-Altiplano Plateau) and the Cordillera Oriental. (∼10 to 6 Ma; Hoke and Garzione 2008). This uplift propagates into the Sierras Subandinas (∼9 Ma to present; Echavarria et al. 2003) Origin of Caiophora clade composed of species Uplift of the Cordillera Oriental and establishment of Late Miocene to Pliocene inhabiting vegetated areas (towards the east) of the the mountain rainforest (Yungas; ∼7 to 6 Ma; Graham sub-Andean mountain range (C. clavata, C. cernua, et al. 2001). Establishment of the Yungas on the eastern Clateritia, C. dumetorum, C. aconquijae), and the slope of the Sierras Subandinas (∼4 Ma; Starck and Cordillera Oriental (C. hibiscifolia, C. madrequisa, Anzótegui 2001). C. canarinoides, C. buraeavii, C. chuquisacana) (5.64 Ma [2.12–9.77]) (BI) Origin and diversification of the Caiophora clade Uplift and deformation of the Sierras Subandinas on the composed of species inhabiting the Sierras Brasilia Massif (∼5to0Ma;Lambetal.1997). Subandinas and the Sierras Pampeanas (C. clavata, C. cernua, C. lateritia, C. dumetorum, C. aconquijae) (2.76 Ma [1.22–6.53] and 1.83 Ma [0.48–3.62], respectively (BI) Pliocene to present Origin of the C. carduifolia (1)-C. coronata clade The Puna-Altiplano Plateau rose from 2500 to 4000 Pliocene to present (3.18 Ma [0.55–6.37]) (BI) ma.s.l.(∼5toMatopresent;Lambetal.1997;Barnes and Ehlers 2009) Origin of the C. carduifolia (2)-C. chuquitensis The Puna-Altiplano Plateau rose from 2500 to 4000 Pliocene to present clade (3.85 Ma [3.36–14.35]) (BI) m.a.s.l. (∼5 Ma to present; Lamb et al. 1997; Barnes and Ehlers 2009) Origin of the C. cirsiifolia (1)-C. cirsiifolia (2) and The Puna-Altiplano Plateau rose from 2500 to 4000 Pliocene to present C. cirsiifolia (3)-C. cirsiifolia (4) clades on the m.a.s.l. (∼5toMatopresent;Lambetal.1997;Barnes western Andean slope (3.85 Ma [3.36–14.35] and Ehlers 2009) and 3.18 Ma [0.55–6.37], respectively) (BI) Split between the lineage of Andean Caiophora Small dust particles are transported by dust storms form Late Pliocene to present species and C. arechavaletae (inhabiting the the Puna-Altiplano Plateau (NW of Argentina) to the Pampas grasslands) (5.98 Ma [1.11–11.99]) (BI) Pampas grasslands and deposited by rain (∼3Mato present; Gaiero et al. 2013) Origin of the lineage giving rise to B. insignis Paranense marine transgression (∼13 to 15 Ma; Donato Middle Miocene and B. amana (inhabiting the Pampas grasslands et al. 2003) and the Central Brazilian sabanas) (11.83 Ma [3.29–22.18]) (BI)

NPRS estimation (∼15 Ma) and the BI estimation (∼10 Ma) diversification of Caiophora. The reliance of many species in are considered, the diversification of Andean hummingbirds the genus on hummingbird pollination, as well as the high (∼15 Ma) (McGuire et al. 2014) may have contributed to the posterior probability of a relatively early acquisition of 38 Strelin M.M. et al. hummingbird pollination in Caiophora (Fig. 1), is in accor- (C. dumetorum, C. lateritia, C. cernua and C. clavata). dance with this expectation. Our estimation using BI suggests Since the Cordillera Oriental (towards the west) is higher than that most of the divergence events within Caiophora are rel- the Sierras Subandinas (towards the east) and insect pollina- atively recent (<5 Ma). The fact that several species in the tion is more frequent at lower-elevation areas (Arroyo et al. genus hybridize giving rise to fertile descendants 1985), reversion from hummingbird to bee pollination in this (Ackermann et al. 2008) and that important rearrangements eastern Caiophora lineage may be expected. of flower development, which are expected in adaptive radia- Orogenic activity during the Pliocene-Pleistocene did af- tion context (West-Eberhard 2005), apparently took place in fect not only the eastern slope of the Andes, but also the Puna- Caiophora (Strelin et al. 2016b) is consistent with a recent Altiplano Plateau, which rose from 2.500 m. to the current diversification. Interestingly, the Late Miocene, the Pliocene 4.000 m. (Lamb et al. 1997; Barnes and Ehlers 2009). This and the Pleistocene were periods of intense orogenic activity sudden uplift of the Puna-Altiplano Plateau presumably drove in the Central Andes and related areas (Armijo et al. 2015). diversification in Caiophora through the generation of high- The Puna-Altiplano Plateau rose ∼2500 m. during the Late elevation pollination niches, where insects are scarce and re- Miocene (∼10 to 7 Ma) (Garzione et al. 2006;Hokeand liance on vertebrate pollinators may therefore be important, Garzione 2008), and the uplift of the Cordillera Oriental (east e.g. Perez et al. (2006). Interestingly, two small and well- to the Puna-Altiplano Plateau) also took place during this pe- supported high-Andean Caiophora clades (2000 to riod (∼10 to 6 Ma) (Hoke and Garzione 2008). This uplift 5000 m.), C. coronata-C. carduifolia (1) and C. chuquitensis continued eastwards, into the Sierras Subandinas (∼5Mato and C. carduifolia (2) (Fig. 2; Supplementary file 2: Table S2), present, according to Lamb et al. (1997); ∼9Matopresent seem to have originated during the end of the Pliocene and to according to Echavarria et al. (2003)). Interestingly, although have diverged during the Pleistocene. Caiophora coronata- eastern Andean Caiophora clades are poorly supported, diver- C. carduifolia (1) originated approximately 3.18 Ma and di- gence within this genus also seems to follow this west to east verged approximately 1.66 Ma; C. chuquitensis-C. carduifolia direction (Fig. 2). (2) originated approximately 3.85 Ma and diversified approx- During the Miocene and the Pliocene, the rising Cordillera imately 1.77 Ma. Although bee pollination was reported for Oriental and the Sierras Subandinas began to act as a rain some of the species in these clades, all of them also rely on barrier, retaining and precipitating Atlantic humidity. This vertebrate pollinators. Hummingbirds play an important role led to the establishment of the mountain rain forests in the pollination of C. chuquitensis and C. carduifolia (Yungas) on the eastern slopes of both mountain ranges about (Ackermann and Weigend 2006; Fig. 2; Supplementary file 7 to 4 Ma (Graham et al. 2001; Starck and Anzótegui 2001). 3: Table S3), and small rodents are in turn pollinators of The origin of the clade composed of C. cernua (Griseb.) Urb. C. coronata (Cocucci and Sérsic 1998;Fig.2; & Gilg ex Kurtz, C. hibiscifolia (Griseb.) Urb. & Gilg, Supplementary file 3: Table S3). While the Puna-Altiplano C. madrequisa Killip, C. canarinoides (Lenné & C. Koch) Plateau uplift may have led to adaptation of high-Andean Urb. & Gilg, Caiophora buraeavii Urb. & Gilg, Caiophora Caiophora species to hummingbird and rodent pollination, chuquisacana Urb. & Gilg., C. clavata, C. lateritia, this same event may have led to the geographic isolation of C. dumetorum and C. aconquijae, which are mainly recorded the C. cirsiifolia (1)-C. cirsiifolia (2) and the C. cirsiifolia (3)- for rain forest areas of the Sierras Subandinas and the C. cirsiifolia (4) clades on the western Andean slope, more Cordillera Oriental (Fig. 2; Supplementary file 2: Table S2) specifically on the Cordillera Occidental (Fig. 2; is dated approximately 5 Ma. In turn, the origin and diversifi- Supplementary file 2: Table S2). Reconstructions of ancestral cation of the Caiophora subclade composed of C. clavata, pollination modes are ambiguous with regard to whether the C. cernua, C. lateritia, C. dumetorum and C. aconquijae, acquisition of hummingbird pollination evolved or not inde- some of which are recorded for the Cordillera Oriental, but pendently in the C. coronata-C. carduifolia (1) and the which are also recorded for mountain ranges to the east C. chuquitensis-C. carduifolia (2) lineages. (Sierras Subandinas and Sierras Pampeanas) (Fig. 2; The origin of C. arechavaletae deserves an explanation Supplementary file 2: Table S2), took place ∼3and∼2Ma, alternative to Andean orogeny. While C. arechavaletae is only respectively (Table 1). Although these clades are poorly sup- distributed in the Uruguayan and the Brazilian Pampas grass- ported, this divergence pattern is consistent with the eastward lands (Fig. 2; Supplementary file 2; Table S2), all remaining propagation of Andean uplift (Table 2). The remaining Caiophora species are associated with the Andes (Fig. 2; Caiophora species in this study were only recorded in arid Supplementary file 2; Table S2). The split between the environments, to the west of the distribution range of the pre- Andean Caiophora lineage and C. arechavaletae is dated viously mentioned species (Fig. 2; Supplementary file 2: about 6 Ma (Fig. 2). Neither orogeny nor climate change can Table S2). Reconstruction of ancestral pollination modes sug- account for this split (Jorge Strelin, personal communication), gests a reversion from hummingbird to bee pollination in the but aeolian transportation of Caiophora seeds from the Caiophora lineage inhabiting the Sierras Subandinas Altiplano region to the current Pampas grasslands is a Diversification of Caiophora (Loasaceae subfam. Loasoideae) 39 plausible explanation. Gaiero et al. (2013) described two re- (CICTERRA, Universidad Nacional de Córdoba, Instituto Antártico cent events of major dust transportation originating in the Argentino); Matías Ghiglione (Instituto de Estudios Andinos, – CONICET, Universidad Nacional de Buenos Aires, Argentina); and the high-altitude, subtropical Puna-Altiplano Plateau (15 26° S; two anonymous reviewers for contributing with their comments and sug- 65–69° W). According to paleoenvironmental evidence, these gestions to the quality of this manuscript. We thank Laura Gatica for events may have also taken place during the Late Pliocene, editing the English of this manuscript and Maximilian Weigend (Nees about 5 Ma. Large dust particles from these events were de- Institut für Biodiversität der Pflanzen, Rheinische Friedrich-Wilhelms- Universität, Bonn, Germany) for providing material, funds and data for tected in traps in the north of the province of Buenos Aires, this study. M.S. has a scholarship from the National Scientific and and according to the modelling of particle deposition and con- Technical Research Council (CONICET). centration, these dust storms may have also transported large dust particles from the Puna-Altiplano Plateau to the Uruguayan and Brazilian grasslands (Gaiero et al. 2013). References Interestingly, the seeds of most of the BSouth Andean Loasas^, which include Loasa, Caiophora, Scyphanthus and Ackermann, M. (2012). Studies on systematics, morphology and taxon- Blumenbachia, are dispersed by wind (Weigend et al. 2005). omy of Caiophora and reproductive biology of Loasaceae and Seeds present specialized testa structures that constitute an Mimulus (Phrymaceae). PhD Thesis. Free University of Berlin, adaptation for aeolic transportation (Weigend et al. 2005). Germany. Based on the weight, the linear measurements and the wing- Ackermann, M., & Weigend, M. (2006). Nectar, floral morphology and like structures of C. arechavaletae seeds (Weigend et al. pollination syndrome in Loasaceae subfam. Loasoideae (Cornales). Annals of Botany, 98,503–514. 2005), strong winds may have transported Caiophora seeds Ackermann, M., Achatz, M., & Weigend, M. (2008). Hybridization and from the Puna-Altiplano Plateau to the Pampas grasslands, crossability in Caiophora (Loasaceae, subfam. Loasoideae): are in- where these may have been deposited by rain (Gaiero, person- terfertile species and inbred populations results of a recent radiation? – al communication). AmericanJournalofBotany,95, 1109 1121. Although the sampling effort in this work concentrated on Altshuler, D. L., Dudley, R., & McGuire, J. A. (2004). Resolution of a paradox: hummingbird flight at high elevation does not come with- Caiophora and not on its allied lineages, it can be speculated out a cost. Proceedings of the National Academy of Sciences of the that the origin of the Blumenbachia insignis Schrad.- United States of America, 101, 17731–17736. Blumenbachia amana T. Henning & Weigend clade (about Antonelli, A., & Sanmartín, I. (2011). Why are there so many plant – 12 Ma), which are recorded for the lowland and grow closer species in the Neotropics? Taxon, 60,403 414. to the Atlantic coast than to the Andes (Fig. 2; Supplementary Antonelli, A., Nylander, J. A., Persson, C., & Sanmartín, I. (2009). Tracing the impact of the Andean uplift on Neotropical plant evolu- file 2; Table S2), is related to the Paranaense marine transgres- tion. Proceedings of the National Academy of Sciences, 106, 9749– sion that took place during the Middle-Late Miocene (15– 9754. 13 Ma), isolating the current Pampas grasslands from the west Armijo, R., Lacassin, R., Coudurier-Curveur, A., & Carrizo, D. (2015). of southern South America (Donato et al. 2003)(Fig.2). Coupled tectonic evolution of Andean orogeny and global climate. Earth-Science Reviews, 143,1–35. Concluding, this study suggests that Andean orogeny fos- Arroyo, M. T. K., Primack, R., & Armesto, J. J. (1985). Community tered the origin and diversification of Caiophora, through the studies in pollination ecology in the high temperate Andes of generation of vicariant barriers and novel ecological condi- Central Chile. I. Pollination mechanism and altitudinal variation. tions, e.g. emergence of mountain rain forests and dependence AmericanJournalofBotany,69,82–97. on vertebrate pollination in high-Andean environments. Baranzelli,M.C.,Johnson,L.J.,Cosacov,A.,&Sérsic,A.N. (2014). Historical and ecological divergence among popula- Unresolved infra-generic phylogenetic relationships, along tions of Monttea chilensis (Plantaginaceae), an endemic endan- with evidence that many divergence events in Caiophora have gered shrub bordering the Atacama Desert, Chile. Evolutionary taken place during the last 5 Myr, suggest a rapid diversifica- Ecology, 28,751–774. tion of the genus. Nevertheless, phylogeographic studies, Barnes, J. B., & Ehlers, T. A. (2009). End member models for Andean Plateau uplift. Earth-Science Reviews, 97,105–132. along with work looking more closely at ecological factors Bechis, F., Encinas, A., Concheyro, A., Litvak, V. D., Aguirre-Urreta, B., promoting lineage divergence in Caiophora,e.g.plant- & Ramos, V. A. (2014). New age constraints for the Cenozoic ma- pollinator interactions and seed dispersion, are essential to rine transgressions of northwestern Patagonia, Argentina (41°-43° better understand the overwhelming diversification of this ge- S): paleogeographic and tectonic implications. Journal of South – nus. Including more Caiophora species, along with species in American Earth Sciences, 52,72 93. Bryson Jr., R. W., García-Vázquez, U. O., & Riddle, B. R. (2012). its sister group, as well as using more genetic markers, is also Relative roles of Neogene vicariance and Quaternary climate change an essential step to accomplish this goal. on the historical diversification of bunchgrass lizards (Sceloporus scalaris group) in Mexico. Molecular Phylogenetics and Evolution, 62,447–457. Acknowledgements We thank Marcela Moré and Cristina Acosta Cavides-Vidal, E., Bozinovic, F., & Rosenmann, M. (1987). Thermal (IMBIV, CONICET, Universidad Nacional de Córdoba, Argentina); freedom of Graomys griseoflavus in a seasonal environment. Romina Vidal-Russel (INIBIOMA, CONICET, Universidad Nacional Comparative Biochemistry and Physiology Part A: Physiology, 87, del Comahue); Jorge Strelin, Mateo Martini and Diego Gaiero 257–259. 40 Strelin M.M. et al.

Chaves, J. A., Weir, J. T., & Smith, T. B. (2011). Diversification in of the Altiplano plateau. Earth and Planetary Science Letters, 271, Adelomyia hummingbirds follows Andean uplift. Molecular 192–201. Ecology, 20,4564–4576. Hufford, L., McMahon, M. M., Sherwood, A. M., Reeves, G., & Chase, Clapperton, C. H. (1983). The glaciation of the Andes. Quaternary M. W. (2003). The major clades of Loasaceae: phylogenetic analysis Science Reviews, 2,83–155. using the plastid matK and trnL-trnF regions. American Journal of Cocucci, A. A., & Sérsic, A. N. (1998). Evidence of rodent pollination in Botany, 90,1215–1228. Cajophora coronata (Loasaceae). Plant Systematics and Evolution, Hufford, L., McMahon, M. M., O’Quinn, R., & Poston, M. E. (2005). A 211,113–128. phylogenetic analysis of Loasaceae, subfamily Loasoideae based on Cosacov, A., Sérsic, A. N., & Sosa, V.(2010). Multiple periglacial refugia plastid DNA sequences. International Journal of Plant Sciences, in the Patagonian steppe and post-glacial colonization of the Andes: 166,289–300. the phylogeography of Calceolaria polyrhiza. Journal of Hughes, C., & Eastwood, R. (2006). Island radiation on a continental Biogeography, 37, 1463–1477. scale: exceptional rates of plant diversification after uplift of the Cruden, R. W. (1972). Pollinators in high-elevation ecosystems: relative Andes. Proceedings of the National Academy of Sciences of the effectiveness of birds and bees. Science, 176,1439–1440. United States of America, 103, 10334–10339. Donato, M., Posadas, P., Miranda-Esquivel, D. R., Jaureguizar, E. O., & Jobb, G., Haeseler, A. V., & Strimmer, K. (2004). TREEFINDER: a Cladera, G. (2003). Historical biogeography of the Andean region: powerful graphical analysis environment for molecular phyloge- evidence from Listroderina (Coleoptera: Curculionidae: netics. BMC Evolutionary Biology. doi:10.1186/1471-2148-4-18. Rhytirrhinini) in the context of the South American geobiotic sce- Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence align- nario. Biological Journal of the Linnean Society, 80,339–352. ment software version 7: improvements in performance and usabil- Drummond, C. S., Eastwood, R. J., Miotto, S. T. S., & Hughes, C. E. ity. Molecular Biology and Evolution, 30,772–780. (2012a). Multiple continental radiations and correlates of diversifi- Lamb, S., Hoke, L., Kennan, L., & Dewey, J. (1997). Cenozoic evolution cation in Lupinus (Leguminosae): testing for key innovations with of the Central Andes in Bolivia and northern Chile. Geological incomplete taxon sampling. Systematic Biology, 61,443–460. Society, London, Special Publications, 121,237–264. Drummond, A. J., Suchard, M. A., Xie, D., & Rambaut, A. (2012b). Levina, M., Horton, B. K., Fuentes, F., & Stockli, D. F. (2014). Cenozoic Bayesian phylogenetics with BEAUti and the BEAST 1.7. sedimentation and exhumation of the foreland basin system pre- Molecular Biology and Evolution, 29,1969–1973. served in the Precordillera thrust belt (31–32°S), southern central Echavarria, L., Hernndez, R., Allmendinger, R., & Reynolds, J. (2003). Andes, Argentina. Tectonics, 33,1659 –1680. Subandean thrust and fold belt of northwestern Argentina: geometry Luebert, F., & Weigend, M. (2014). Phylogenetic insights into Andean and timing of the Andean evolution. AAPG Bulletin, 87,965–985. plant diversification. Frontiers in Ecology and Evolution. Felsenstein, J. (1984). Distance methods for inferring phylogenies: a jus- doi:10.3389/fevo.2014.00027. tification. Evolution, 38,16–24. Martinez, J. J., Ferro, L. I., Mollerach, L. I., & Barquez, R. M. (2012). Fenster, C. B., Armbruster, W. S., Wilson, P., Dudash, M. R., & The phylogenetic relationships of the Andean swamp rat genus Thomson, J. D. (2004). Pollination syndrome and floral specializa- Neotomys (Rodentia, Cricetidae, Sigmodontinae) based on mito- tion. Annual Review of Ecology, Evolution and Systematics, 25, chondrial and nuclear markers. Acta Theriologica, 57,277–287. 375–403. Martini, M.A., Strelin, J.A., Kaplan, M.R., & Schaefer, J.M. (2012). Glacial Gaiero, D. M., Simonella, L., Gassó, S., Gili, S., Stein, A. F., Sosa, P., and periglacial geomorphology and chronology around the Nevado de Becchio, R., Arce, J., & Marelli, H. (2013). Ground/satellite obser- Chañi (Cordillera Oriental of Jujuy): implication for past climate in vations and atmospheric modeling of dust storms originating in the NW Argentina. Resource document. AGU Fall Meeting Abstracts. high Puna-Altiplano deserts (South America): implications for the http://adsabs.harvard.edu/abs/2012AGUFMEP53C1053M. Accessed interpretation of paleo-climatic archives. Journal of Geophysical 13 Oct 2016. Research: Atmospheres, 118,3817–3831. McGuire, J. A., Witt, C. C., Remsen Jr., J. V., Corl, A., Daniel, L., Gamble, T., Simons, A. M., Colli, G. R., & Vitt, L. J. (2008). Tertiary Rabosky, D. L., Altshuler, D. L., & Dudley, R. (2014). Molecular climate change and the diversification of the Amazonian gecko ge- phylogenetics and the diversification of hummingbirds. Current nus Gonatodes (Sphaerodactylidae, Squamata). Molecular Biology, 24,910–916. Phylogenetics and Evolution, 46,269–277. Meng, H. H., & Zhang, M. L. (2013). Diversification of plant species in Garzione, C. N., Molnar, P., Libarkin, J. C., & MacFadden, B. J. (2006). arid Northwest China: species-level phylogeographical history of Rapid late Miocene rise of the Bolivian Altiplano. Evidence from Lagochilus Bunge ex Bentham (Lamiaceae). Molecular removal of mantle lithosphere. Earth and Planetary Science Letters, Phylogenetics and Evolution, 68,398–409. 241,543–556. Mercer, J. H., & Sutter, J. (1982). Late Miocene—earliest Pliocene gla- Giambiagi, L. B., & Ramos, V. A. (2002). Structural evolution of the ciation in Southern Argentina: implications for global ice-sheet his- Andes in a transitional zone between flat and normal subduction tory. Palaeogeography, Palaeoclimatology, Palaeoecology, 38, (33°30′–33°45′S), Argentina and Chile. Journal of South 185–206. American Earth Studies, 15,101–116. Minelli, A., & Fusco, G. (2012). On the evolutionary developmental Graham, A., Gregory-Wodzicki, K. M., & Wright, K. L. (2001). Studies biology of speciation. Evolutionary Biology, 39,242–254. in Neotropical Paleobotany. XV. A Mio-Pliocene palynoflora from Moré, M., Cocucci, A. A., & Sérsic, A. N. (2015). Phylogeny and floral the Eastern Cordillera, Bolivia: implications for the uplift history of trait evolution in Jaborosa (Solanaceae). Taxon, 64,523–534. the Central Andes. American Journal of Botany, 88,1545–1557. Muschick, M., Indermaur, A., & Salzburger, W. (2012). Convergent evo- Hamilton, M. B. (1999). Four primer pairs for the amplification of chlo- lution within an adaptive radiation of cichlid fishes. Current Biology, roplast intergenic regions with intraspecific variation. Molecular 22, 2362–2368. Ecology, 8,521–523. Mutke, J., Jacobs, R., Meyers, K., Henning, T., & Weigend, M. (2014). Hoiss, B., Krauss, J., Potts, S. G., Roberts, S., & Steffan-Dewenter, I. Diversity patterns of selected Andean plant groups correspond to (2012). Altitude acts as an environmental filter on phylogenetic topography and habitat dynamics, not orogeny. Frontiers in composition, traits and diversity in bee communities. Proceedings Genetics.doi:10.3389/fgene.2014.00351. of the Royal Society B-Biological Sciences, 279,4447–4456. Müller, J., Müller, K., Quandt, D., & Neinhuis, C. (2010). PhyDe- Hoke, G. D., & Garzione, C. N. (2008). Paleosurfaces, paleoelevation, Phylogenetic Data Editor. Program distributed by the author. and the mechanisms for the late Miocene topographic development Available at: http://www.phyde.de/index.html. Diversification of Caiophora (Loasaceae subfam. Loasoideae) 41

Nespolo, R. F., Opazo, J. C., & Rosenmann, F. B. (1999). Thermal accli- Stech, M., Veldman, S., & Larraín, J. (2013). Molecular species delimi- mation, maximum metabolic rate, and nonshivering thermogenesis tation in the Racomitrium canescens Complex (Grimmiaceae) and of Phyllotis xanthopygus (Rodentia) in the Andes Mountains. implications for DNA barcoding of species complexes in mosses. JournalofMammalogy,80,742–748. PloS One.doi:10.1371/journal.pone.0053134. Nielsen, R. (2002). Mapping mutations on phylogenies. Systematic Strelin, M. M., Benitez-Vieyra, S., Ackermann, M., & Cocucci, A. Biology, 51,729–739. (2016a). Flower reshaping in the transition to hummingbird pollina- Nylin, S., Slove, J., & Janz, N. (2014). Host plant utilization, host range tion in Loasaceae, subfam. Loasoideae despite absence of corolla oscillations and diversification in Nymphalid butterflies: a phyloge- tubesorspurs.Evolutionary Ecology, 30,401–407. netic investigation. Evolution, 68,105–124. Strelin, M. M., Benitez-Vieyra, S., Fornoni, J., Klingenberg, C. P., & Palazzesi, L., Gottschling, M., Barreda, V., & Weigend, M. (2012). First Cocucci, A. A. (2016b). Exploring the ontogenetic scaling hypoth- Miocene fossils of Vivianiaceae shed new light on the phylogeny, esis during the diversification of pollination syndromes in divergence times, and historical biogeography of Geraniales. Caiophora (Loasaceae, subfam. Loasoideae). Annals of Botany, Botanical Journal of the Linnean Society, 107,67–85. 117,937–947. Perez, F., Arroyo, M. T. K., Medel, R., & Hershkovitz, M. A. (2006). Swofford, D.L. (2003). PAUP*. Phylogenetic Analysis Using Parsimony Ancestral reconstruction of flower morphology and pollination sys- (*and Other Methods). Version 4. Sinauer Associates, Sunderland, tems in Schizanthus (Solanaceae). American Journal of Botany, 93, Massachusetts. 1029–1038. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., & Kumar, S. Ramos, V. A., Cristallini, E. O., & Pérez, D. J. (2002). The Pampean flat- (2011). MEGA5: molecular evolutionary genetics analysis using slab of the Central Andes. Journal of South American Earth Studies, maximum likelihood, evolutionary distance, and maximum parsi- 15,59–78. mony methods. Molecular Biology and Evolution, 28,2731–2739. Revell, L. J. (2012). Phytools: an R package for phylogenetic compara- Särkinen, T., Pennington, R. T., Lavin, M., Simons, M. F., & Hughes, C. tive biology (and other things). Methods in Ecology and Evolution, E. (2012). Evolutionary islands in the Andes: persistence and isola- 3,217–223. tion explain high endemism in Andean dry tropical forests. Journal Revell, L. J. (2014). Graphical methods for visualizing comparative data of Biogeography, 39,884–900. on phylogenies. In L. Z. Garamszegi (Ed.), Modern phylogenetic Weigend, M., & Gottschling, M. (2006). Evolution of funnel-revolver comparative methods and their application in evolutionary biology flowers and ornithophily in Nasa (Loasaceae). Plant Biology, 8, (pp. 77–103). Berlin, Heidelberg: Springer. 120–142. Ronquist, F., & Huelsenbeck, J. P. (2003). MrBayes 3: Bayesian phylo- Weigend, M., Gottschling, M., Hoot, S., & Ackermann, M. (2004). A genetic inference under mixed models. Bioinformatics, 19, 1572– preliminary phylogeny of Loasaceae subfam. Loasoideae 1574. (Angiospermae: Cornales) based on trnL(UAA) sequence data, with Rubioldo, D., Seggiaro, R., Gallardo, E., Disalvo, A., Sanchez, M., Turel, consequences for systematics and historical biogeography. A., Ramallo, E., Sandruss, A., & Godeas, M. (2001). Hoja Organisms Diversity & Evolution, 4,73–90. Geológica 2366-II / 2166-IV, La Quiaca. Geología y Provincias de Weigend, M., Gröger, A., & Ackermann, M. (2005). The seeds of Jujuy y Salta. Instituto de Recursos Minerales, Servicio Geológico Loasaceae subfam. Loasoideae (Cornales) II: seed morphology of Minero Argentino. Boletín 246, p. Buenos Aires. BSouth Andean Loasas^ (Loasa, Caiophora, Scyphanthus and Rutschmann, F. (2006). Molecular dating of phylogenetic trees: a brief Blumenbachia ). Flora-Morphology, Distribution, Functional review of current methods that estimate divergence times. Diversity Ecology of Plants, 200,569–591. and Distributions, 12,35–48. Weigend, M., Ackermann, M., & Henning, T. (2010). Reloading the Sanderson, M. J. (1997). A nonparametric approach to estimating diver- revolver- male fitness as a simple explanation for complex reward gence times in the absence of rate constancy. Molecular Biology and partitioning in Nasa macrothyrsa (Loasaceae, Cornales). Biological Evolution, 14,1218–1231. Journal of the Linnean Society, 100,124–131. Schenk, J. J., & Hufford, L. (2010). Effects of substitution models on Wen, J., & Zimmer, E. (1996). Phylogeny and biogeography of Panax L divergence time estimates: simulations and an empirical study of (the ginseng genus, Araliaceae): inferences from ITS sequences of model uncertainty using Cornales. Systematic Botany, 35,578–592. nuclear ribosomal RNA. Molecular Phylogenetics and Evolution, 6, Sérsic, A., Cosacov, A., Cocucci, A. A., Johnson, L. A., Pozner, R., Avila, 167–177. L. J., Sites Jr., J. W., & Morando, M. (2011). Emerging West-Eberhard, M. J. (2005). Developmental plasticity and the origin of phylogeographical patterns of plants and terrestial vertebrates from species differences. Proceedings of the National Academy of Patagonia. Biological Journal of the Linnean Society, 103,475–494. Sciences, 102,6543–6549. Smith, S. D., & Baum, D. A. (2006). Phylogenetics of the florally diverse Xia, X., & Xie, Z. (2001). DAMBE: software package for data analysis in Andean clade Iochrominae (Solanaceae). American Journal of molecular biology and evolution. Journal of Heredity, 92,371–373. Botany, 93, 1140–1153. Xia, X., Xie, Z., Salemi, M., Chen, L., & Wang, Y. (2003). An index of Smith, S. D., Ané, C., & Baum, D. A. (2008). The role of pollinator shifts substitution saturation and its application. Molecular Phylogenetics in the floral diversification of Iochroma (Solanaceae). Evolution, 62, and Evolution, 26,1–7. 793–806. Yang, Z., & Rannala, B. (2012). Molecular phylogenetics: principles and Solà, E., Sluys, R., Gritzalis, K., & Riutort, M. (2013). Fluvial basin practice. Nature Reviews Genetics, 13,303–314. history in the northeastern Mediterranean region underlies dispersal Zech, J., Zech, R., Kubik, P. W., & Veit, H. (2009). Glacier and climate and speciation patterns in the genus Dugesia (Platyhelminthes, reconstruction at Tres Lagunas, NWArgentina,basedon10Besur- Tricladida, Dugesiidae). Molecular Phylogenetics and Evolution, face exposure dating and lake sediment analyses. Palaeogeography, 66,887–888. Palaeoclimatology, Palaeoecology, 284,180–190. Starck, D., & Anzótegui, L. M. (2001). The late Miocene climatic Zuloaga, F.O., Morrone, O., Belgrano, M.J., Marticorena, C., & change—persistence of a climatic signal through the orogenic strat- Marchesi, E. (2008). Catálogo de las plantas vasculares del Cono igraphic record in northwestern Argentina. Journal of South Sur. Monogr. Syst. Bot. Missouri Bot. Gard, 107(1–3), i–xcvi, 1– American Earth Sciences, 14,763–774. 3348.