NCBI GenBank Harvest, 4/3/2014 This is a partially literal copy of File S2 in the online supplementary material to Grímsson et al. (2016). Emendations/modifications are highlighted by blue font.

Search strings taxa and gene regions

Taxa

Loranthaceae: txid3963[Organism:exp]

Gene regions matK gene: … AND matK [gene] rbcL gene1: … AND rbcL [gene] NOT "atpB-rbcL" [title] trnLLF region (trnL intron, trnL-trnF spacer): … AND "trnL-trnF" [title]2 18S rDNA: … AND (18S [title] or "small subunit" [title]) NOT "internal" [title] ITS region of the 35S rDNA3: … AND “5.8S” [title] 25S rDNA: … AND 26S [title] NOT "internal" [title]

1 Four sequences of Benthamina sp. SH-2010 (Tobe et al., unpubl.; AB586377–AB586380) not caught by search and were manually added. 2 The covered region in these sequences includes also most of the upstream trnL intron 3 One ITS2 accession (Tristerix penduliflorus, DQ442975, Amico et al., Am. J. Bot., 2007) comprising 197 bp (ITS2 only) will not be caught. Being only a short fragment without extra information, the sequence was not manually added.

Processing

Gene bank flatfiles were transferred into FASTA-format using GBK2FAS (Göker et al. 2009, see batch file for options). Auto-alignments were generated with MAFFT v. 6.935b (Katoh et al. 2005; Katoh & Standley 2013), optimal algorithm chosen by the programme.

Dataset Algorithm [1] 18S L-INS-i [2] ITS FFT-NS-i [3] 25S FFT-NS-i [4] matK L-INS-i [5] trnLLF FFT-NS-i [6] rbcL L-INS-i

All alignments were inspected by eye, and their ends were trimmed. As a rule, we truncated the alignments for less than four accessions at the 5’ and 3’ ends.

Nuclear-encoded 18S rDNA (Fig. S1-1)

General—The data comprise near-complete 18S rDNA sequences showing only restricted length- polymorphism at pos. 184–189 (CT-dom. motif), 223–233 (upstream-GC, downstream CT) and within the terminal loop of stem 49 (tl49) at the 3’ end of the 18S rDNA. Several single-nucleotide (nt) gaps are introduced in which a single of the sequences shows an extra nucleotide typically the same then before or after the gap. Curation—Isolated single-nucleotide gaps are likely due to sequencing/editing artefact rather than representing actual mutations and have been eliminated from the alignment. Right-aligned the length- polymorphic motives at pos. 184ff; left-aligned pos. 229ff; centre-aligned tl49-motives. Several accessions (Gaiadendron punctatum, L24143; Helicanthes elastica, EU544328; Erianthemum dregei, L25679; Oedina pendens, EU544336; Ligaria cuneifolia, L24152; Struthanthus oerstedii, L24421) show increased amount of deviations within strongly conserved stem regions (pseudogeny?). Where conclusive comparative data was lacking, the sequence was kept as-is, otherwise blanked out (replaced by missing data symbol). Problematic data—Accession AF039074 (Decaisnina hollrungii, no voucher information included; Elytrantheae) differs from the sequence of its congener, hence, the misplacement of Decaisnina in the 18S tree based on the reduced data set.

Nuclear-encoded ITS region

General—The nrDNA spacers are highly divergent and length-polymorphic at the family level, resulting in a very gappy auto-alignment. The ITS1 cleavage site region, the 5.8S rDNA, and the generally more conserved parts of the ITS2 are successfully recognised as alignable blocks due to their high conservation level and overall lack of length-polymorphism. Note: The data was not included in family-wide analyses due to potential alignment artefacts on the overall topology. Curation—Except for right-aligning the start of the 5.8S rDNA, the alignment was left un-changed. Problematic data—Three accessions show likely pseudogenous mutations in the 5.8S rDNA (Dendrophthoe constricta, DQ333840, 3’ part; Helixanthera parasitica, DQ333823, central part; Psittacanthus schiedeanus, DQ333859, 3’ part, 5’ part with small missing data blocks). In case of DQ333840 mutational patterns linked to pseudogeny are also apparent in the ITS1 and start of ITS2 (identifiable by comparison with the second accession of the and other members of the Lorantheae).

In the case of DQ333823, pseudogeny is less obvious in the ITS1 and ITS2: the sequence appears to be a mix of ITS variants with different levels of pseudogeny as indicated by polymorphic calls (Y and R at overall conserved C and G positions). DQ333859 may be a chimera (or recombinant) of a non-pseudogenous and (weakly) pseudogenous ITS variants: the area of overlap between typical forward and reverse reads has been blanked out by the authors indicating problems with the direct PCR sequencing. Lacking comparative data of this genus/species, the sequence was kept as-is. The 3’ portion of the ITS2 in Notanthera heterophylla DQ333855 is entirely different from other accessions of the subtribe/tribe and markedly off-alignment. The strand portion finds no similar hits in gene banks. Based on the general distinctness of the ITS of this taxon, the 3’ ITS2 may be genuine and has been kept as-is. One of the two accessions of Taxillus (T. chinensis, JX177497) differs strongly from the other sequence and is essentially off-alignment within ITS1 and the larger part of ITS2. Subsequent MEGABLAST identifies the sequence as Viscum (same order, family Viscaceae).

Nuclear-encoded 25S rDNA data (Fig. S1-2)

General—Data comprise near-complete 25S rDNA sequences. Variation and minor length- polymorphism is limited to the D1–D4 regions of the 25S rDNA. Markedly divergent patterns in generally sequential- and structural-conserved regions are limited to a few accessions (→ Problematic data) Curation—Left-aligned C-rich motif at pos. 2078ff. Problematic data—Accessions of Diplatia furcata (EU544368), Helicanthes elastica (EU544375), Lepidaria cf. forbesii (EU544378), Ixocactus sp. (EU5443779), Spraguenella rhamnifolia (EU544401), Loxanthera speciosa (EU544382) show an increased number of substitutions in generally (clade or family) conserved regions. In the case of Ixocactus the sequence inflicts single-nt gaps. In case EU544368 many deviations are indicative for pseudogeny (G→A, C→T) in addition to deviations that may be linked to bad sequence reads/editing artefacts (e.g. AATTA at pos. 485–489 instead of the consensual ATTGA). The sequence was excluded from analysis. EU544375 (Helicanthes elastic; Nickrent 4906) represents a unique 25S rDNA fragment, with best

BLAST hits of 92–93% identity covering a wide range of taxa and clades (Solanales: Schizanthus, Apiales: Bolax; Saxifragales: Saxifraga, Telesonix; Ranuculales: Borax; etc.); top hits did not include members of . The sequence was tentatively excluded from analyses as putative contaminant. The 18S sequence from the same voucher is equally problematic. Accession EU544378 (Nickrent 4044, labelled as Lepidaria cf. forbesii) shows increasing amount of pseudogenous positions (G→A, C→T mutations in generally conserved regions) towards the 3’ part. The sequence was excluded from analysis as showing tendency towards pseudogeny.

The same, but to a lesser degree applies to accessions EU544368 (Diplatia furcata, Nickrent 2824) and EU544382 (Loxanthera speciosa, Nickrent 4026). The latter was tentatively kept for analysis due to the importance of the taxon for the phylogenetic framework. Accession EU5443779 (Nickrent 4332, labelled as Ixocactus sp. [= Phthirusa]) is composed of two strands. The first strand (170 nt from the 5’ end) finds best BLAST-hits among the Loranthaceae, but lower identity than usually found in when BLASTing a member of the family. Then comes a sequencing gap and the following remaining c. 1200 nt find highest identities (>90%; obtained with MEGABLAST) only with members of other families of the (Osyris, Santalaceae; Haloragis, Halogaraceae) and unrelated singletons such as Eucnide bartonioides (JF321124), Loasa vulcanica (JF321125; Cornales), Fendlera rupicola (AY260041), Dryas octopetala (JF317384; Rosales), and Eucryphia lucida (AF036494, Oxalidales). Based on this result it is uncertain whether the sequence is genuine or a mislabelling/artificial chimera; hence, it was tentatively excluded from all analyses. Accession EU544401 differs from the general sequences types within members of its clade and Loranthaceae in general by potential pseudogene-like mutations in addition to sequence deviations that may be linked to sequencing/editing artefacts. Because of its overall dissimilarity to other Loranthaceae sequences in family diagnostic regions, MEGABLAST (performed 9/2/2016) failed to retrieve any Loranthaceae 25S within the listed hits. Note that the pollen of one species formerly included in Ixocactus (Phthirusa hutchisonii) does not confirm with morphotypes typical for Loranthaceae (Grímsson, Grimm & Zetter 2017)

Plastid matK gene (Fig. S1-3)

General—Data includes up-to-near-complete matK sequences starting with the first codon in three accessions (stop codon seems to be not covered in any accession). More than 3-taxon coverage starts at the 27th codon (alignment was accordingly truncated). Length-polymorphism is restricted to singleton duplications or eliminations. Curation—In regions with duplications/eliminations, the general alignment was adapted for codon positions when necessary. Defect (due to missing single-nucleotides) and incomplete codons were blanked (replaced by missing data symbol, “?”). Problematic data—The 5’ part of accession EU544409 (Aetanthus nodosus) is markedly different from other members of its tribe and the entire family. MEGABLAST revealed that the first c. 260 nt are highly identical to accession of several species of Acacia (Fabales). The remainder of the sequence falls within the variation of the Loranthaceae. Thus, the accession is a (artificial) chimera of contaminated material and unverified Aetanthus (no comparative data available of the genus), and has to be excluded from analysis. Two accessions, EU544414 (Baratranthus axanthus, Nickrent 4029) and EU544435 (Lepeostegeres lancifolia, Nickrent 4041) are placed within wrong subtrees (cf. Nickrent et al. 2010). In case of Baratranthus, other gene regions (18S, 25S, trnLF) sequenced from the same voucher are unsuspicious.

Plastid trnLLF region (Fig. S1-4)

General—The data includes the trnL intron, the 3’ trnL exon (identified by comparison to Panax ginseng complete plastome, KC686331), the trnL-trnF intergenic spacer and up to nearly complete trnF genes (15–21 nt missing in longest sequences). Most length-polymorphism in the intron is linked to duplications in a single or few sequences and a prominent multiple-A motif (pos. 190ff) Curating—Alignment in singletons at pos 132–166 (in three accessions) were corrected for duplication patterns. Left-aligned pos. 252–266, and 633–648. Local alignments in the trnL-trnF spacer appear to be suboptimal but were kept as-is. Odd placements—The trnLLF region is the only gene region included in the final dataset, where genera (to some degree: species) are usually represented by more than a single accession. The single-gene tree reveals several issues with the currently available sequence data: Psittacanthinae clade: The putative sister genera Cladocolea and Strutanthus (Vidal-Russell & Nickrent 2008; Su et al. 2015; Grímsson, Grimm & Zetter 2017) are not resolved as distinct clades, despite a general relatively high branch support in the corresponding subtree. The only representative of genus Phthirusa, P. inorna (former Ixocactus inornus) is nested within the Tripodanthus subclade (genus’ root poorly supported) comprising multiple accessions for two of the three species in the genus. Most conspicuously, the accessions representing the putative sister genera Oryctanthus (O. occidentalis) and Passovia (P. pyrifolia) of Wilson & Calvin (2006) and Vidal-Russell & Nickrent (2008) are mixed up. A shorter branched subclade comprises accession DQ340608 of P. pyrifolia (Calvin & Wilson CR01- 3) and EU544501 of O. occidentalis (Nickrent 2763), whereas the other is remarkably long-branched and includes the respective counterparts (DQ340613, O. occidentalis, Calvin & Wilson CR01-11; EU544504, P. pyrifolia, Nickrent 2762). Comparative chloroplast data are only available for the Nickrent 2762 and 2763 vouchers; in the matK tree, P. pyrifolia is long-branched whereas O. occidentalis is shorter branched. Thus, we assume that the Calvin and Wilson sequences have been mixed-up during upload. Others: Visible distinct accessions separate the two Lepidaria samples (L. oviceps, DQ340602, Calvin & Wilson B02-19; L. cf. forbesii, EU544492, Nickrent 4044) and the three covered Amyema species. In the latter case, matK data are available for comparison and confirming the ‘non-monophyly’ of the genus. Accession DQ340578 (Taxillus wiensii; A. Robertson 7364) places in the wrong subtree, as sister to Vanwykia. The Scurrulinae Scurrula and Taxillus group in a distinct subclade in all single-gene trees. Thus, we have to assume that accession DQ340578 is either mislabelled, or that the according species is not a member of the Scurrulinae.

Plastid rbcL gene (Fig. S1-5)

General—The matrix covers the 469 codons of the rbcL gene (starting with the 8th codon, last three codons of complete rbcL not represented in GenBank data; reference sequence Panax ginseng KC686331). Most variable sites (~75%) refer to the 3rd codon position.

Curating—Alignment of rbcL is straightforward (generally no length polymorphism). The region is generally conserved at the genus/subtribe level, which facilitates the identification of putative mislabelled sequences. The data includes one unidentified Loranthacea rbcL, which was not considered for further analyses. Problematic data— Three sequences of Taxillus chinensis (JF949992; JN687568; KF447376) differ markedly from the other eight sequences of the species and other members of the genus/tribe (visible from pos. 250 onwards). MEGABLAST identified JF949992 as Elytrantheae/Gaiadendreae (99% identity; verified on the alignment) and JN687568 as Viscaceae, another family of the Santalales (most similar to Viscum, 97% identity; no Loranthaceae among the top hits). The sequences either represent mislabelled or misidentified specimens. KF447376 received no significant hit with identity >93% (best hits with Scurulla, the sister genus of Taxillus, generic affinity verified at hand of the full alignment). The sequence represents either a bad sequence read or a strongly aberrant rbcL gene (potential pseudogene). All three sequences were excluded from analyses. A similar sequence type than in JN687568 is found in the only accession of Oryctanthus cordifolius

(JQ592409), differing from the otherwise near-identical sequence of Psittacanthinae. MEGABLAST identifies the sequence accordingly as Viscaceae rbcL (Phoradendron; 98% identity). The sequence is either mislabelled or Oryctanthus is not a Loranthaceae and has been omitted. Another problematic sequence is the one representing Macrosolen brandisianus (JN687566) which differs increasingly towards the 3’ end from other accessions of the genus/tribe. Deviation is usually related to same-nucleotide repeats (e.g. GGTGGT instead GTGGGT in all other Lorantheae at pos. 349ff; AATTTG instead AAATTG as pos. 460ff). The problematic sequence part encompasses 50% of the total sequenced strand (c. 500 nt), the first half of the sequence is identical to other Lorantheae. The sequence was tentatively excluded from analyses. Odd placements—The rbcL data includes genuine sequences from phylogenetic studies as well as short(er) sequences from barcoding studies. The latter are often problematic, mislabelled accessions are not uncommon (see Problematic data). Moreover, and contrasting statements in many barcoding papers, there is so far no conclusive evidence that rbcL can be used as barcode at low hierarchical levels in . Reported high species-identification levels with rbcL fragments in barcoding studies are simply due to incomprehensive species sampling of large genera (G. Grimm, pers. obs.) The currently available rbcL data on Loranthaceae supports this observation: the resolution of the rbcL tree is generally poor. Identical sequences can be shared by species of the same genus, sometimes between genera. Based on the current, relatively taxon-limited data it is not possible to distinguish between potentially representative vs. mislabelled accession in critical cases (Dendrophthoe, Helixanthera, Struthanthus, Taxillus). With respect to this taxonomic uncertainty, the generally weak signal and the limited species coverage (23 species with genuine rbcL data compared to 44/45 covered by matK and trnLF) of the rbcL data, we excluded this gene region from further analyses.

Set-up for phylogenetic analyses and dating Phylogenetic tree inferences and branch support analyses relied on maximum likelihood (ML) as the optimality criterion, branch support was estimated using non-parametric bootstrapping. All analyses, files can be found in subfolder ML in OSA, were done using RAxML v.8.2.4 (Stamatakis 2014). RAxML was invoked using custom shell files (included in subsequent subfolders in the OSA). All inferences used the -f a option that performs a (in our case, fast; option -x) bootstrapping analyses with subsequent inference of the ‘best-known’ ML tree under a general-time-reversible substitution model allowing for site rate variation, modelled using a gamma distribution (GTR+ model) in the final optimisation step. During bootstrapping and the first (fast) tree inference step, the per-site approximation for the gamma distribution was used (option -GTRCAT; Stamatakis 2006). Number of necessary bootstrap replicates were determined by the extended majority rule consensus criterion (Pattengale et al. 2009), bootstrap analyses were automatically cut off when reaching a maximum of 1000 replicates.

We use a seven-step protocol Step 1: Preanalysis—Using the initial alignments (not consensed, not concatenated gene bank accessions), we inferred single-gene trees for five of the six harvested regions: 18S, 25S, matK, trnLLF, rbcL. This step allowed us to further check for problematic accessions not directly visible or overlooked during the visual inspection of the alignments and to assess the general level of discriminative signal in the single gene regions (see Processing). Step 2: Consensing—Modal and strict species-consensus sequences were computed from the initial alignments using the programme G2CEF (Göker & Grimm 2008); gaps were treated as missing data or fifth state (batch and files are included in the subfolder Step2_Consensi in this Online Supplementary Archive, OSA). For further analyses the strict consensus sequences with gaps treated as missing were used. Step 3: Inference of comprehensive species trees—Using the species-consensus alignments, we produced three concatenated matrices: (1) including all four gene regions (18S, 25S, matK, trnLLF), (2) a nuclear-only matrix including 18S and 25S rDNA, and (3) a plastid-only matrix including the matK gene and the trnLLF region. ML analyses were run partitioned and unpartitioned, including or excluding the relatively gappy, length-polymorphic trnL-trnF spacer (i.e. analyses used five matrices with different gene coverage). The partitioning scheme was as follows: ● 18S rDNA and 25S rDNA treated as independent partitions ● First, second and third codon position of matK gene treated as independent partitions ● Intron and intergenic spacer treated as independents partitions, exons (trnL 3’ exon, trnF gene) included in the same partition. Step 4: Clock rooting—Data sets including selected outgroups (Wilson & Calvin 2006; Vidal-Russell & Nickrent 2008) or representatives of all/most Santalales families (e.g. Su et al. 2015; see files provided in the OSA to Grímsson, Grimm & Zetter 2017) failed to produce unambiguous support for

the exact position of the Loranthaceae root, although trees based on combined data usually will recognise Nuytsia as the first diverging branch in the Loranthaceae. All potential outgroups are (extremely) long- branched and Nuytsia is one of the most distinct Loranthaceae. Thus, ingroup-outgroup long-branch attraction may be inevitable (as detailed by GWG in Grímsson, Grimm & Zetter 2017, file S6). Hence, a clock-rooting (cf. Renner et al. 2008), was tried using the five matrices from step three. For each of the matrices we performed a BEAST (v. 1.8.2; Drummond & Rambaut 2007; Drummond et al. 2012) run under partition specific substitution models, unconstrained tree topology, a Yule tree prior and uncorrelated log-normal clock prior [ucld.mean ~Gamma(0.001, 1000)]. The best fitting substitution models per partition, among the available in BEAST, were selected with JMODELTEST (Darriba et al. 2012). The selected model for most partitions (rDNAs, matK codon positions, trnL-trnF spacer) was

GTR+ whereas HKY+was suggested for trnL intron and HKY+I for the tRNA. Each BEAST analysis was conducted for 2*107 generations with a sampling frequency of 0.001. To explore the effect of missing data, we performed the clock-rooting also with the reduced dataset of 42 species covering the four best-sampled and alignable gene regions (18S, 25S, matK and trnLLF) that was used for the final dating step (Step 6). The rbcL gene was excluded since it was represented only by a minority of the samples. The analysis was run under the same priors as before. Step 5: Inference of preliminary dated phylogenies including all data—Using the two all- comprehensive matrix from Steps 3 and 4, i.e. the matrix including all four gene regions (18S, 25S, matK, trnLLF), dated phylogenies were inferred for three different rooting scenarios: traditional (following earlier studies) outgroup-based root as informed by Nuytsia, the same root is found via clock- rooting of the reduced 42-taxon, 4-gene region dataset; the clock-informed root based on the all-taxon datasets recognising two main clades: Lorantheae vs. Psittacantheae+root parasites; and a pollen-type- informed root recognising Tupeia as sister to all other Loranthaceae (cf. Grímsson, Grimm & Zetter 2017). Table S1-1 lists the used age constraints including short discussions (see also subsections Use as age constraint included in the pollen descriptions in the main-text). Dating was performed with BEAST v. 1.8.2 (Drummond & Rambaut 2007; Drummond et al. 2012). Bayesian optimisation used four data partitions with the following nucleotide substitution models: GTR+I for 18S and GTR+ for 25S, matK, and trnLLF, respectively. The Monte Carlo Marko chains (MCMC) were run twice for 5*107 million generations under an uncorrelated lognormal clock model. All age calibration priors were modelled as normal distributions around the midpoint of the known time intervals. For rooting scenario 2, the analysis of the comprehensive matrix did not converge after the set number of generations, hence, no result tree is included in the according subfolder. For the other two rooting scenarios, the high gappyness of the comprehensive matrix caused similar problems (relatively low ESS for several parameters, late convergence of runs). For this reason, the final dating analysis used a reduced species set, only including those members of each lineage that were best-covered by data for all four gene regions.

Table S1-1 MRCA Rooting scenario (r.sc.) 1 Rooting scenario 2 Rooting scenario 3 Scenario 4: ‘true tree’ (Su et al., 2015) enforced Fossil Age Fossil Age Fossil Age Fossil Age All Loranthaceae Miller Clay Pit MT1 47.8–41.2 Stolzenbach MT 47.8– (not constrained) Same as for r.sc. 1/2 41.2 All Loranthaceae except (not fixed)a (incompatible with tree) (incompatible with (not fixed)a Nuytsia tree) All Loranthaceae except (incompatible with tree) (incompatible with tree) Miller Clay Pit 47.8–41.2 (incompatible with tree) Tupeia MT1(–3), Stolzenbach MT Elytrantheae + Lorantheae (incompatible with tree) (incompatible with tree) (incompatible with tree) Profen MT2–5 41.2–38 (incompatible with tree) (incompatible with tree) (incompatible with tree) Lorantheae Changchang MT 47.8–37.8 Same as for r.sc.1 Same as for r.sc.1 Same as for r.sc. 1 Psittacantheae + root (incompatible with tree) Miller Clay Pit 47.8– (incompatible with tree) (incompatible with tree) parasites MT1 41.2 Notanthera + Profen MT2–5 41.2–38 Same as for r.sc.1 Same as for r.sc.1 (incompatible with tree) Elytrantheae Psittacanthinaec (Miller Clay Pit MT2, 3) (47.8–41.2) Same as for r.sc.1 Same as for r.sc.1 (not fixed)d Aamaruutissaa MT 42–40 Notanthera + (incompatible with tree) (incompatible with tree) (incompatible with tree) Aamaruutissaa MTd 42–40 Psittacanthinae a Could be constrained to the same minimum age than MRCA of all Loranthaceae (co-eval: Miller Clay Pit MT2-MT3 → Psittancanthinae, and Stolzenbach MT → Lorantheae) b The Theiss MT represents a putative ancestral, primitive (plesiomorphic) pollen of the Lorantheae lineage, less derived than the Changchang MT. Being younger but linked to a deeper node (formation of Lorantheae vs. radiation of core Lorantheae, Clade I/J according Vidal-Russell & Nickrent), it has little use as additional minimum age prior. c All pollen addressed as Miller Clay Pit MT2 and MT3 and Aamaruutissaa MT are very similar to that of two species of modern Tripodanthus and could be used to constrain the divergence of the Tripodanthus lineage from the remainder of the Psittacanthinae. Nevertheless, they are used as minimum age priors for a deeper node: the MRCA of all Psittacanthinae. This was done to compensate for missing information (pollen type of Phthirusa inorna, the potentially first branching modern Psittacanthinae is unknown; Tripodanthus pollen may represent a primitive type within the lineage) and with respect to ambiguous signal regarding deep relationships in the Psittacanthinae subtree. d Although the Miller Clay Pit MT2 and MT3 are possibly older, we opted to use the more precisely dated Aamaruutissaa MT as age constraint, placed one node up in the other trees constraining for the root age of the Psittacanthinae. This may result in underestimated (slightly too young) ages (see Results in the main text), but avoids having to use similar age constraints for two nodes in a line (i.e. the MRCA of Notanthera and Psittacanthinae and the root node of Psittacanthinae). Step 6: Final inference of dated phylogenies—To compensate for eventual missing data artefacts on the estimates, we performed the same set of analyses with a taxon-reduced data set, a data set that only included 42 species with data for all five gene regions. The general set-up (priors etc.) was the same as for the preceding step, further details can be extracted from the xml-files provided in the according subfolder in the OSA. At this step, we performed an additional analysis (Scenario 4, Table S1-1) with a partly constrained backbone phylogeny. During the review of Grímsson, Grimm & Zetter (2017), one anonymous reviewer, an expert on parasites in general and Loranthaceae in particular, informed us about the “correct” placement of several taxa (essentially the topology seen in Su et al. 2015, fig. 1B; but see figs S6 and S7 in the supplement to that paper). The topology of the “correct tree” differs in several aspects from the topologies that can be inferred from the here used data set(s). The data basis regarding the covered (coding) genes was similar but not identical in both studies, the additionally used genes (RPB2, accD, matR) in Su et al. (2015) cover very few of the Loranthaceae and mainly (RPB2, matR) enforce the placement of the Mystropetalaceae as sister of Loranthaceae. At the same time these additional gene regions inflict matK-conflicting signals (Grímsson, Grimm & Zetter 2017, file S6), the latter being the gene region that informs most splits seen in the Su et al. (2015) tree. Our dataset includes instead the non-coding trnLLF region, which has a broader taxon coverage and a much higher divergence than the coding regions included in Su et al. Our dataset also lacks distant, (extremely) long-branching outgroups (there is e.g. no plastid data available for the Mystropetalaceae). Furthermore, we use species-consensus sequences instead of selected single individuals used as placeholder, thus, our matrix has a lower gappyness. This may affect directly the optimised topology because the signal from individual gene regions is apparently not fully compatible in the Loranthaceae (see Grímsson et al. 2016, file S1). To compensate for the incapability of our data set to infer “correct" relationships, we hence did an additional dating analysis in which we constrained the topology according to that reviewer’s views. The three root parasites were constrained to form a grade: following up from the most distinct root parasite and sister to the remainder of the Loranthaceae (Nuytsia), come the relatively long-branched Atkinsonia and the short-branched Gaiadendron. Within the aerial parasite clade, most New World taxa are collected within an according clade. In this ‘New World’ clade, the putative sister taxa Ligaria and Tristerix (Ligarinae) are constrained as sister to Notanthera and the Psittacanthinae. The conservative pinning of Elytrantheae fossils found at Profen requires the identification of the sister group of the Elytrantheae (which is Notanthera in all optimised topologies, apparently a “wrong” placement). Following the topology of Su et al. (2015), we constrained the Elytrantheae as sister to the Lorantheae.

Step 7: Post-analyses test (comparison of Bayes factors) For each of the four topological configurations we estimated the marginal likelihood with two approaches, stepping-stone and path-sampling, both of which are implemented in BEAST (Baele et al. 2012; Baele et al. 2013), with 100 path steps with 20*107 chain length for each. For selecting the rooting

scenario that best fit our data, we compared the marginal likelihood estimate of each run using Bayes factors (Kass & Raftery 1995).

Documentation All analysis files are included in the OSA in corresponding subfolders. See ReadMe.txt in top folder for file labels and brief description of content. The OSA is mirrored at www.palaeogrimm.org/data; if you have questions regarding the data, files, and inferences contact GWG (mail-to: [email protected]).

Figures (appended at the end of the file) Single-gene trees based on the harvested, unconsensed/-concatenated sequence data and optimised under ML; visibly problematic – as directly deduced from the alignment – sequences not included. Branches, subtrees, etc. referring to major groups (tribes/subtribes) of Loranthaceae are coloured accordingly; branch thickness indicates bootstrap support under ML (see in-figure legends). Highlighted by red font and in bold, mislabelled accessions detected via the tree inference and excluded from subsequent analyses; accessions highlighted by orange font (bold) may be problematic but tentatively kept due to the unavailability of comparative data.

References Baele G, Lemey P, Bedford T, Rambaut A, Suchard MA, Alekseyenko AV. 2012. xxx. Molecular Biology and Evolution 29:2157–2167. Baele G, Li WLS, Drummond AJ, Suchard MA, Lemey P. 2013. xxx. Molecular Biology and Evolution 30:239–243. Darriba D, Taboada GL, Doallo R, Posada D. 2012. jModelTest 2: more models, new heuristics and parallel computing. Nature Methods 9:772–772. Drummond AJ, Rambaut A. 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology 7:214. Drummond AJ, Suchard MA, Xie D, Rambaut A. 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29:1969–1973. Göker M, García-Blázquez G, Voglmayr H, Tellería MT, Martín MP. 2009. Molecular of phytopathogenic fungi: a case study in Peronospora. PLoS ONE 4:e6319. Göker M, Grimm GW. 2008. General functions to transform associate data to host data, and their use in phylogenetic inference from sequences with intra-individual variability. BMC Evolutionary Biology 8:86. Grímsson F, Grimm GW, Zetter R. 2017. Evolution of pollen morphology in Loranthaceae. Grana DOI:10.1080/00173134.2016.1261939. Kass RE, Raftery AE. 1995. Bayes factors. Journal of the American Statistical Association 90:773–795. Katoh K, Kuma K, Toh H, Miyata T. 2005. MAFFT version 5: improvement in accuracy of multiple sequence alignment. Nucleic Acids Research 33:511–518. Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution 30:772–780. Pattengale ND, Masoud A, Bininda-Emonds ORP, Moret BME, Stamatakis A. 2009. How many bootstrap replicates are necessary? In: Batzoglou S, ed. RECOMB 2009. Berlin, Heidelberg: Springer-Verlag, 184–200.

Renner SS, Grimm GW, Schneeweiss GM, Stuessy TF, Ricklefs RE. 2008. Rooting and dating maples (Acer) with an uncorrelated-rates molecular clock: Implications for North American/Asian disjunctions. Systematic Biology 57:795-808. Stamatakis A. 2006. Phylogenetic models of rate heterogeneity: A high performance computing perspective. Proceedings of 20th IEEE/ACM International Parallel and Distributed Processing Symposium (IPDPS2006), High Performance Computational Biology Workshop. Rhodos, Greece, April 2006. p [on CD, no page nos]. Stamatakis A. 2014. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30:1312–1313. Su H-J, Hu J-M, Anderson FE, Der JP, Nickrent DL. 2015. Phylogenetic relationships of Santalales with insights into the origins of holoparasitic Balanophoraceae. Taxon 64:491–506. Vidal-Russell R, Nickrent DL. 2008. Evolutionary relationships in the showy family (Loranthaceae). American Journal of Botany 95:1015–1029. Wilson CA, Calvin CL. 2006. An origin of aerial branch parasitism in the mistletoe family, Loranthaceae. American Journal of Botany 93:787–796.

Oncella_ambigua|EU544338|Nickrent_4673 ML tree based on harvested Oncocalyx_sulfureus|EU544339|Nickrent_2850 Fig. S1-1. Phragmanthera_crassicaulis|EU544341|Nickrent_3037 Socratina_bemarivensis|EU544347|Nickrent_4179 18S rDNA data Agelanthus_sansibarensis|U59946|UNKNOWN Actinanthella_menyharthii|EU544313|Nickrent_4375 Moquiniella_rubra|AF039078|UNKNOWN BS≥ 90 Helixanthera_kirkii|AF039077|UNKNOWN Misplaced Globimetula_dinklagei|AF039076|UNKNOWN BS = [70,90[ Emelianthe_panganensis|EU544326|Nickrent_4889 Emelianthinae BS = [40,70[ Oliverella_rubroviridis|EU544337|Nickrent_4330 Tapinanthus_quinquangulus|L24422|UNKNOWN Tapinanthinae BS < 40 Englerina_woodfordioides|L24140|UNKNOWN Berhautia_senegalensis|EU544320|Nickrent_4576 Erianthemum_dregei|L25679|UNKNOWN Dendrophthoe_curvata|AY957441|Nickrent_2787 Dendrophthoe_curvata|L26491|UNKNOWN Decaisnina_hollrungii|AF039074|UNKNOWN Mislabelled Dendrophthoinae Dendrophthoe_curvata|EU544325|Nickrent_4012 Bakerella_sp_|EU544318|Nickrent_4161 Plicosepalus_sagittifolius|EU544342|Nickrent_2852 Vanwykia_remota|EU544351|Nickrent_4331 Scurrula_parasitica|EU544345|Nickrent_4004 Scurrula_pulverulenta|EU544344|Nickrent_4159 Scurrula_ferruginea|EU544343|Nickrent_4008 Scurrulinae Taxillus_chinensis|EU544350|Nickrent_4032 Benthamina_alyxifolia|EU544319|Nickrent_4127 Baratranthus_axanthus|EU544317|Nickrent_4029 Helixanthera_cylindrica|EU544327|Nickrent_4037 Amyema_glabrum|AF039073|UNKNOWN Amyeminae Diplatia_furcata|L24088|UNKNOWN Dactyliophora_novae_guineae|EU544323|Nickrent_4563 Amyema_queenslandicum|EU544315|Nickrent_2788 Loranthus_kaoi|JQ613221|UNKNOWN Loranthus_odoratus|EU544331|Nickrent_4977 Lorantheae Loranthus_delavayi|JQ613220|UNKNOWN Loranthinae Loranthus_europaeus|L24153|UNKNOWN Cecarria_obtusifolia|EU544321|Nickrent_4562 Ileostylus_micranthus|EU544329|Nickrent_2741 Ileostylinae Muellerina_eucalyptoides|EU544335|Nickrent_4310 Oedina_pendens|EU544336|Nickrent_4329 Pseudogeny artefact? Desmaria_mutabilis|EF464465|Nickrent_4510A Aetanthus_nodosus|EU544314|Nickrent_4561 Notantherinae Psittacanthus_angustifolius|L24414|UNKNOWN Cladocolea_gracilis|EU544322|Nickrent_3066 Struthanthus_woodsonii|EU544349|Nickrent_2761 Struthanthus_oerstedii|L24421|UNKNOWN Psittacanthinae Missing data/ Dendropemon_bicolor|AF039075|UNKNOWN Oryctanthus_occidentalis|L24408|UNKNOWN pseudogeny Phthirusa_pyrifolia|L24412|UNKNOWN Tripodanthus_acutifolius|L24424|UNKNOWN artefact Atkinsonia_ligustrina|EF464464|Nickrent_4344 Gaiadendreae Branch reduced by factor 2 Helicanthes_elastica|EU544328|Nickrent_4906 Ligaria_cuneifolia|L24152|UNKNOWN Tristerix_corymbosus|EF464467|Nickrent_4129 Ligarinae Tupeia_antarctica|L24425|UNKNOWN Tupeinae Alepis_flavida|L24139|UNKNOWN Peraxilla_tetrapetala|EU544340|Nickrent_2744 Notanthera_heterophylla|EF464466|Nickrent_4372 Notantherinae Septulina_glauca|EU544346|Nickrent_4089 Problematic placement Lepidaria_cf_forbesii|EU544330|Nickrent_4044 Gaiadendron_punctatum|L24143|UNKNOWN Gaiadendreae Lysiana_exocarpi|L24154|UNKNOWN Lysiana_filifolia|EU544333|Nickrent_4449 Sogerianthe_sessiflora|EU544348|Nickrent_4467 Problematic placement Elytrantheae Amylotheca_duthieana|EU544316|Nickrent_4022 Loxanthera_speciosa|EU544332|Nickrent_4026 Macrosolen_cochinchinensis|EU544334|Nickrent_4038 Decaisnina_triflora|EU544324|Nickrent_4491 Nuytsia_floribunda|DQ790103|D_Nickrent_2747 Nuytsieae 0.01 expected subst./site ML tree Fig. S1-2. Actinanthella_menyharthii|EU544352|Nickrent_4375 based on harvested Agelanthus_sansibarensis|EU544353|Nickrent_2987 Berhautia_senegalensis|EU544360|Nickrent_4576 25S rDNA data Oncocalyx_sulfureus|EU544388|Nickrent_2850 Bakerella_sp_|EU544358|Nickrent_4161 BS≥ 90 Phragmanthera_crassicaulis|EU544391|Phragmanthera_3037 Moquiniella_rubra|DQ790207|D__Nickrent_3042 BS = [70,90[ Septulina_glauca|EU544398|Nickrent_4089 Emelianthinae BS = [40,70[ Globimetula_dinklagei|EU544372|Nickrent_3087 BS < 40 Emelianthe_panganensis|EU544369|Nickrent_4889 Tapinanthinae Oedina_pendens|EU544386|Nickrent_4329 Erianthemum_dregei|EU544371|Nickrent_2985 Englerina_ramulosa|EU544370|Nickrent_2984 Tapinanthus_constrictiflorus|EU544404|Nickrent_3088 Oliverella_rubroviridis|EU544387|Nickrent_4330 Plicosepalus_sagittifolius|EU544393|Nickrent_2852 Vanwykia_remota|EU544407|Nickrent_4331 Socratina_bemarivensis|EU544399|Nickrent_4179 Helixanthera_cylindrica|EU544374|Nickrent_4037 Dendrophthoe_curvata|EU544367|Nickrent_4012 Dendrophthoe_longituba|EU544366|Nickrent_4010 Dendrophthoinae Helixanthera_coccinea|EU544373|Nickrent_4019 Scurrula_pulverulenta|EU544396|Nickrent_4159 Scurrula_parasitica|EU544397|Nickrent_4004 Scurrula_ferruginea|EU544395|Nickrent_4008 Scurrulinae Taxillus_chinensis|EU544405|Nickrent_4032 Dactyliophora_novae_guineae|EU544363|Nickrent_4563 Benthamina_alyxifolia|EU544359|Nickrent_4127 Amyema_queenslandicum|EU544355|Nickrent_2788 Amyeminae Amyema_glabrum|EU544354|Nickrent_2794 Loran- Diplatia_furcata|EU544368|Nickrent_2824 Sogerianthe_sessiflora|EU544400|Nickrent_4467 theae Baratranthus_axanthus|EU544357|Nickrent_4029 Muellerina_eucalyptoides|EU544385|Nickrent_4310 Ileostylinae Ileostylus_micranthus|EU544376|Nickrent_2741 Sequences show evidence of Loranthus_europaeus|EU544380|Nickrent_2849 Loranthus_odoratus|EU544381|Nickrent_4977 degradation (pseudogeny) Cecarria_obtusifolia|EU544361|Nickrent_4562 Loranthinae Amylotheca_duthieana|EU544356|Nickrent_4022 Loxanthera_speciosa|EU544382|Nickrent4026 Lepidaria_cf__forbesii|EU544378|Nickrent_4044 Elytrantheae Macrosolen_cochinchinensis|EU544384|Nickrent_4038 Decaisnina_triflora|EU544364|Nickrent_4491 Lepeostegeres_lancifolius|EU544379|Nickrent_4041 Lysiana_filifolia|EU544383|Nickrent_4449 Alepis_flavida|EF464474|Nickrent_2743 Sequence of unclear affinity Peraxilla_tetrapetala|EU544390|Nickrent_2744 and quality Notanthera_heterophylla|EF464478|Nickrent_4372 Notantherinae Spragueanella_rhamnifolia|EU544401|Nickrent_4674 Branches reduced by factor 2 Desmaria_mutabilis|EF464476|Nickrent_4510A Ixocactus_sp_|EU544377|Nickrent_4332 Ligaria_cuneifolia|EF464477|Nickrent_4567 Sequence of unclear affinity, Oryctanthus_occidentalis|EU544389|Nickrent_2763 possibly artificial chimera Dendropemon_bicolor|EU544365|Nickrent_2172 Branch reduced by factor 5 Helicanthes_elastica|EU544375|Nickrent_4906 Phthirusa_pyrifolia|EU544392|Nickrent_2762 Struthanthus_oerstedii|EU544402|Nickrent_2728 Short sequence of unclear affinity Struthanthus_woodsonii|EU544403|Nickrent_2761 Cladocolea_gracilis|EU544362|Nickrent_3066 Psittacanthinae Psittacanthus_calyculatus|EU544394|Nickrent_4043 Tripodanthus_acutifolius|EU544406|Nickrent_2969 Atkinsonia_ligustrina|EF464475|Nickrent_4344 Tupeia_antarctica|DQ790208|D__Nickrent_2742 Gaiadendreae Gaiadendron_punctatum|DQ790209|D__Nickrent_2729 Tupeinae Tristerix_corymbosus|EF464479|Nickrent_4572 Ligarinae Nuytsia_floribunda|DQ790210|D__Nickrent_2747 Nuytsieae 0.01 exp. subst./site Oncella_ambigua|EU544443|Nickrent_4673 Fig. S1-3. ML tree based on harvestedmat K data Oncocalyx_sulfureus|EU544444|Nickrent_2850 Agelanthus_sansibarensis|EU544410|Nickrent_2987 Englerina_ramulosa|EU544427|Nickrent_2984 BS≥ 90 Tapinanthus_constrictiflorus|EU544459|Nickrent_3088 Actinanthella_menyharthii|EU544408|Nickrent_4375 BS = [70,90[ Berhautia_senegalensis|EU544417|Nickrent_4576 BS = [40,70[ Oedina_pendens|EU544441|Nickrent_4329 Vanwykia_remota|EU544463|Nickrent_4331 BS < 40 Socratina_bemarivensis|EU544454|Nickrent_4279 Helixanthera_coccinea|EU544430|Nickrent_4019 Dendrophthoe_longituba|EU544423|Nickrent_4010 Emelianthinae Dendrophthoe_curvata|EU544424|Nickrent_4012 Tolypanthus_involucratus|EU544461|Nickrent_4907 Tapinanthinae Plicosepalus_sagittifolius|EU544449|Nickrent_2852 Bakerella_sp_|EU544415|Nickrent_4161 Spragueanella_rhamnifolia|EU544456|Nickrent_4674 Dendrophthoinae Erianthemum_dregei|EU544428|Nickrent_2985 Tapinanthus_sp_|EU214301|OM825 Oliverella_rubroviridis|EU544442|Nickrent_4330 Globimetula_dinklagei|EU544429|Nickrent_3087 Emelianthe_panganensis|EU544426|Nickrent_4889 Phragmanthera_crassicaulis|EU544447|Nickrent_3037 Moquiniella_rubra|DQ790171|D__Nickrent_3042 Scurrula_ferruginea|EU544451|Nickrent_4008 Scurrula_parasitica|EU544453|Nickrent_4004 Scurrulinae Scurrula_pulverulenta|EU544452|Nickrent_4159 Taxillus_chinensis|EU544460|Nickrent_4032 Helixanthera_cylindrica|EU544431|Nickrent_4037 Baratranthus_axanthus|EU544414|Nickrent_4029Placement at odds with current systematics Amyema_glabrum|EU544411|Nickrent_2794 Diplatia_furcata|EU544425|Nickrent_2824 Dactyliophora_novae_guineae|EU544420|Nickrent_4563 Lorantheae Amyeminae Benthamina_alyxifolia|EU544416|Nickrent_4127 ‘Non-monophyletic’ genus Helicanthes_elastica|EU544432|Nickrent_4906 Amyema_queenslandicum|EU544412|Nickrent_2788 Sogerianthe_sessiflora|EU544455|Nickrent_4467 Loranthinae Loranthus_europaeus|EU544436|Nickrent_5082 Cecarria_obtusifolia|EU544418|Nickrent_4562 Ileostylus_micranthus|EU544433|Nickrent_2741 Muellerina_eucalyptoides|EU544440|Nickrent_4310 Ileostylinae Struthanthus_orbicularis|JQ589640|RE_5125_1 Struthanthus_oerstedii|EU544457|Nickrent_2728 Struthanthus_sp_|JQ589641|RE_5128_1 Struthanthus_woodsonii|EU544458|Nickrent_2761 Cladocolea_gracilis|EU544419|Nickrent_3066 Tripodanthus_acutifolius|EU544462|Nickrent_2969 Psittacanthinae Oryctanthus_occidentalis|EU544445|Nickrent_2763 Dendropemon_bicolor|EU544422|Nickrent_2172 Phthirusa_pyrifolia|EU544448|Nickrent_2762 Ligarinae Psittacanthus_calyculatus|EU544450|Nickrent_4043 Ligaria_cuneifolia|EF464510|Nickrent_4567 Tristerix_corymbosus|EF464512|Nickrent_4129 Notanthera_heterophylla|EF464511|Nickrent_4372 Desmaria_mutabilis|EF464509|Nickrent_4510A Notantherinae Macrosolen_cochinchinensis|EU544439|Nickrent_4038 Lepidaria_cf__forbesii|EU544434|Nickrent_4044 Decaisnina_triflora|EU544421|Nickrent_4491 Amylotheca_duthieana|EU544413|Nickrent_4022 Loxanthera_speciosa|EU544437|Nickrent_4026 Lysiana_filifolia|EU544438|Nickrent_4449Elytrantheae Alepis_flavida|EF464508|Nickrent_2743 Peraxilla_tetrapetala|EU544446|Nickrent_2744 Lepeostegeres_lancifolius|EU544435|Nickrent_4041Placement at odds with current systematics Tupeia_antarctica|DQ790172|D_Nickrent_2742 Gaiadendron_punctatum|DQ787445|Nickrent_2729 Tupeinae Atkinsonia_ligustrina|DQ787444|Nickrent_4343 Gaiadendreae Nuytsia_floribunda|DQ787446|Nickrent_2747 0.01 exp. subst./site Nuytsieae Dendrophthoe_pentandra|HQ317759|LS27 Dendrophthoinae Dendrophthoe_pentandra|HQ317760|LS32 Fig. S1-4. ML tree based on harvestedrbc L data Dendrophthoe_pentandra|JN687564|A_Chveerach_P51_2 Dendrophthoe_lanosa|JN687563|A_Chveerach_P38_1 Taxillus_sclerophyllus|JQ933498|Chase_17408_K Helixanthera_sampsonii|HQ317766|LS34 BS≥ 90 Helixanthera_sampsonii|HQ317765|LS33 Tolypanthus_maclurei|HQ317758|LS02 Emelianthinae/Tapinanthinae BS = [70,90[ Tolypanthus_maclurei|HQ317757|LS01 Bakerella_sp_|EU544466|Nickrent_4161 BS = [40,70[ Tapinanthus_sp_|EU213529|OM825 Septulina_glauca|AM235022|JC_Manning_2925_NBG BS < 40 Moquiniella_rubra|AM235021|RBGK_5742 Agelanthus_sansibarensis|EU544464|Nickrent_2987 Dendrophthoe_nilgherrensis|JQ933297|Chase_17409_K Englerina_ramulosa|EU544470|Nickrent_2984 Taxillus_chinensis|HQ317787|LS53 Taxillus_chinensis|HQ317786|LS52 Taxillus_chinensis|HQ317780|LS03 Mislabelled or Taxillus_chinensis|HQ317784|LS08 Taxillus_chinensis|HQ317782|LS06 not congeners Taxillus_chinensis|HQ317790|LS56 Taxillus_chinensis|HQ317785|LS09 ‘Non-monophyletic’ Helixanthera Taxillus_chinensis|HQ317788|LS54 Taxillus_chinensis|HQ317783|LS07 or mislabelled Taxillus_chinensis|HQ317789|LS55 Taxillus_chinensis|HQ317781|LS04 Scurrula_sp_|JQ933474|MF_Watson_et_al_20812115_E Taxillus_sutchuenensis|HQ317796|LS16 Taxillus_sutchuenensis|HQ317797|LS17 Taxillus_nigrans|HQ317794|LS20 Taxillus_limprichtii|HQ317793|LS15 Taxillus_nigrans|HQ317795|LS21 Scurrulinae Taxillus_limprichtii|HQ317792|LS13 Taxillus_limprichtii|HQ317791|LS12 Taxillus_sutchuenensis|HQ317798|LS10 Taxillus_sutchuenensis|HQ317799|LS11 Scurrula_chingii|HQ317774|LS26 10 Scurrula_parasitica|HQ317779|LS58 Scurrula_chingii|HQ317773|LS25 Scurrula_parasitica|HQ317775|LS18 Scurrula_parasitica|HQ317778|LS23 Lorantheae Scurrula_parasitica|HQ317777|LS57 Scurrula_atropurpurea|KF114862|UNKNOWN Scurrula_ferruginea|KF114863|UNKNOWN Scurrula_parasitica|HQ317776|LS19 Helixanthera_parasitica|HQ317763|LS30 Helixanthera_parasitica|HQ317762|LS29 Helixanthera_parasitica|HQ317761|LS28 Helixanthera_parasitica|HQ317764|LS31 Helixanthera_parasitica|JN687565|A_Chveerach_P93_3 Benthamina_alyxifolia|AY957440|Nickrent_4127 Amyema_glabrum|EU544465|Nickrent_2794 Amyeminae Loranthus_europaeus|JQ933393|Chase_14152_K Cecarria_obtusifolia|EU544467|Nickrent_4562 Loranthinae Loranthus_delavayi|HQ317767|LS39 Muellerina_eucalyptoides|EU544472|Nickrent_4310 Ileostylinae Tupeia_antarctica|DQ790133|D_Nickrent_2742 Moquiniella_rubra|DQ790132|D_Nickrent_3042 Tupeinae Peraxilla_tetrapetala|EU544473|Nickrent_2744 Peraxilla_tetrapetala|JQ933439|P_Douglass_8714_RNG Desmaria_mutabilis|EF464527|Nickrent_4510A Macrosolen_cochinchinensis|HQ317770|LS36 Notantherinae Macrosolen_cochinchinensis|HQ317769|LS35 Macrosolen_cochinchinensis|HQ317768|LS05 Elytrantheae Macrosolen_cochinchinensis|JQ933395|IWJ_Sinclair_DG_Long_5781_E Macrosolen_cochinchinensis|JN687567|A_Chveerach_P31 Macrosolen_tricolor|HQ317771|LS37 Macrosolen_tricolor|HQ317772|LS38 Ligarinae Decaisnina_triflora|EU544468|Nickrent_4491 Tristerix_corymbosus|EF464530|Nickrent_4575E Gaiadendron_punctatum|L26072|UNKNOWN Struthanthus_orbicularis|JQ592415|BioBot05048 Gaiadendreae Struthanthus_orbicularis|JQ592414|BioBot05046 Struthanthus_orbicularis|JQ592413|BioBot05045 Other two Psittacanthinae nested in Struthanthus Psittacanthinae Tripodanthus_acutifolius|EU544475|Nickrent_4927 Struthanthus_orbicularis|JQ594621|RE_5125_1 ‘non-monophyletic’ S. orbicularis Struthanthus_sp_|JQ594622|RE_5128_1 Struthanthus_woodsonii|EU544474|Nickrent_2761 Dendropemon_bicolor|EU544469|Nickrent_2700 Ligarinae Ligaria_cuneifolia|EF464528|Nickrent_4567 Notanthera_heterophylla|EF464529|Nickrent_4372 Atkinsonia_ligustrina|EF464526|Nickrent_4344 Notantherinae Gaiadendreae Nuytsia_floribunda|DQ790134|D_Nickrent_2747 0.01 exp. subst./site Nuytsieae Diplatia_furcata|EU544486|Nickrent_2824 Diplatia_grandibractea|DQ340581|Calvin_and_Wilson_AU01_05 Amyema_glabrum|EU544476|Nickrent_2794 Fig. S1-5. ML tree Amyema_cambagei|DQ340582|Calvin_and_Wilson_AU99_01 ‘Non-monophyletic’ Dactyliophora_novae_guineae|EU544483|Nickrent_4563 Dactyliophora_novae_guineae|DQ340586|Calvin_and_Wilson_AU02_10 based on harvested Amyema Sogerianthe_sessiflora|EU544508|Nickrent_4467 Amyema_queenslandicum|DQ340583|Calvin_and_Wilson_AU02_22 Benthamina_alyxifolia|EU544480|Nickrent_4127 trnLLF data Amyeminae Benthamina_alyxifolia|DQ340584|Calvin_and_Wilson_AU99_14 Baratranthus_axanthus|EU544478|Nickrent_4029 Baratranthus_axanthus|DQ340587|Calvin_and_Wilson_B02_14 Bakerella_sp_|EU544479|Nickrent_4161 BS≥ 90 Plicosepalus_curviflorus|DQ340575|Calvin_and_Wilson_Y88_03 Dendrophthoe_longituba|EU544485|Nickrent_4010 BS = [70,90[ Dendrophthoe_vitellina|DQ340588|Calvin_and_Wilson_AU98_08 Dendrophthoe_constricta|DQ340589|Calvin_and_Wilson_B02_03 Dendrophthoinae Helixanthera_coccinea|EU544490|Nickrent_4019 BS = [40,70[ Scurrula_ferruginea|EU544505|Nickrent_4008 Scurrula_ferruginea|DQ340576|Calvin_and_Wilson_B02_02 BS < 40 Taxillus_chinensis|EU544512|Nickrent_4032 Erianthemum_dregei|EU544488|Nickrent_2985 Erianthemum_dregei|DQ340580|A__Robertson_7366 Oedina_pendens|EU544499|Nickrent_4329 Emelianthinae/ Emelianthe_panganensis|EU544487|Nickrent_4889 Globimetula_dinklagei|EU544489|Nickrent_3087 Phragmanthera_regularis|DQ340579|Calvin_and_Wilson_Y88_01 Tapinanthinae Septulina_glauca|EU544506|Nickrent_4089 Moquiniella_rubra|EF464489|Nickrent_3042 Phragmanthera_crassicaulis|EU544503|Nickrent_3037 Agelanthus_sansibarensis|DQ340573|A_Robertson_7365 Oncocalyx_sulfureus|EU544500|Nickrent_2850 Oncocalyx_schimperi|DQ340574|Calvin_and_Wilson_Y88_02 Tapinanthus_constrictiflorus|EU544511|Nickrent_3088 Englerina_ramulosa|DQ340577|A_Robertson_7280 Socratina_bemarivensis|EU544507|Nickrent_4179 Lorantheae Taxillus_wiensii|DQ340578|A_Robertson_7364 Vanwykia_remota|EU544514|Nickrent_4331 Wrong subtree Helixanthera_parasitica|DQ340572|Calvin_and_Wilson_B02_01 Loranthus_grewingkii|HQ917115|GUH_3998 Loranthus_europaeus|HQ917114|GUH_3997 Loranthinae Loranthus_europaeus|EU544493|Nickrent_2849 Loranthus_odoratus|EU544494|Nickrent_4977 Cecarria_obtusifolia|DQ340585|Calvin_and_Wilson_AU02_16

Muellerina_eucalyptoides|DQ340591|Calvin_and_Wilson_AU99_05 Cecarria_obtusifolia|EU544481|Nickrent_4562 Muellerina_eucalyptoides|EU544498|Nickrent_4310 Muellerina_celastroides|DQ340590|Calvin_and_Wilson_AU99_11 Ileostylus_micranthus|EU544491|Nickrent_2741 Ileostylus_micranthus|DQ340592|Calvin_and_Wilson_NZ00_02 Ileostylinae Tristerix_corymbosus|DQ442922|G_Amico_84_BCRU Tristerix_corymbosus|DQ442923|G_Amico_96_BCRU Tristerix_aphyllus|DQ442919|G_Amico_166_BCRU Tristerix_aphyllus|DQ442917|L_Suarez_sn_BCRU Tristerix_corymbosus|DQ442921|G_Amico_83_BCRU Tristerix_corymbosus|DQ340605|Calvin_and_Wilson_CL03_02 Tristerix_aphyllus|DQ442916|G_Amico_97_BCRU Tristerix_aphyllus|DQ442918|G_Amico_162_BCRU Tristerix_corymbosus|DQ442920|G_Amico_80_BCRU Tristerix_corymbosus|EF464493|Nickrent_4597A Tristerix_penduliflorus|DQ442928|M_Lewis_35236_4661844_MO Tristerix_penduliflorus|DQ442929|P_Nunez_8288_720879_MO Tristerix_verticillatus|DQ442926|G_Amico_158_BCRU Tristerix_verticillatus|DQ442925|G_Amico_157_BCRU Tristerix_verticillatus|DQ442924|G_Amico_154_BCRU Ligarinae Tristerix_verticillatus|DQ442927|G_Amico_155_BCRU Tristerix_grandiflorus|DQ442930|Palacios_Neil_Ceron_2547_3672311_MO Tristerix_grandiflorus|DQ442931|V_Zar_1368_3907304_MO Tristerix_longebracteatus|DQ442933|D_Fernandez_M_Cerna_P_Villacres_371_5889268_MO Tristerix_longebracteatus|DQ442932|DN_Smith_8288_722877_MO Tristerix_pubescens|DQ442937|DN_Smith_11388_3337809_MO Tristerix_pubescens|DQ442936|DN_Smith_M_Buddensiek_11090_3303268_MO Tristerix_peruvianus|DQ442934|DN_Smith_M_Buddensiek_10858_3307944__MO Tristerix_chodatianus|DQ442938|DN_Smith_Valencia_Gonzales_9536_3337838_MO Tristerix_peytonii|DQ442935|P_Nunez_J_Arque_8287_3632650_MO Peraxilla_tetrapetala|EU544502|Nickrent_2744 Peraxilla_tetrapetala|DQ340597|Calvin_and_Wilson_NZ98_03 Alepis_flavida|DQ340598|Calvin_and_Wilson_NZ98_04 Alepis_flavida|EF464481|Nickrent_2743 Atkinsonia_ligustrina|DQ788714|Nickrent_4343 Atkinsonia_ligustrina|DQ340616|Calvin_and_Wilson_AU00_01 Gaiadendreae Decaisnina_congesta|DQ340594|Calvin_and_Wilson_AU02_01 Decaisnina_signata|DQ340596|Calvin_and_Wilson_AU02_05 Decaisnina_brittenii|DQ340595|Calvin_and_Wilson_AU01_11 Decaisnina_triflora|EU544484|Nickrent_4491 Amylotheca_duthieana|EU544477|Nickrent_4022 Elytrantheae Lepidaria_cf__forbesii|EU544492|Nickrent_4044 Macrosolen_melintangensis|DQ340593|Calvin_and_Wilson_B02_17 Macrosolen_cochinchinensis|EU544497|Nickrent_4038 ‘ Lepidaria Lysiana_maritima|DQ340599|Calvin_and_Wilson_AU98_19 Non-monophyletic’ Lysiana_filifolia|EU544496|Nickrent_4449 Amylotheca_subumbellata|DQ340600|Calvin_and_Wilson_AU98_21 Loxanthera_speciosa|EU544495|Nickrent_4026 Lepidaria_oviceps|DQ340602|Calvin_and_Wilson_B02_19 Gaiadendron_punctatum|DQ788715|Nickrent_2729 Gaiadendron_punctatum|DQ340617|Calvin_and_Wilson_CR01_08 Tripodanthus_acutifolius|HM010441|Nickrent_5548 Gaiadendreae Tripodanthus_acutifolius|HM010425|Nickrent_4983 Tripodanthus_acutifolius|HM010439|Nickrent_5349 Tripodanthus_acutifolius|HM010438|Nickrent_5346 Tripodanthus_acutifolius|HM010437|Nickrent_5345 Tripodanthus_acutifolius|HM010426|Nickrent_4998 Tripodanthus_acutifolius|DQ340615|Calvin_and_Wilson_AR03_04 Tripodanthus_acutifolius|EU544513|Nickrent_4927 Tripodanthus_acutifolius|HM010436|Nickrent_5330 Tripodanthus_flagellaris|HM010446|Nickrent_5553 Tripodanthus_flagellaris|HM010442|Nickrent_5549 Tripodanthus_flagellaris|HM010445|Nickrent_5552 Tripodanthus_flagellaris|HM010429|Nickrent_5210 Tripodanthus_flagellaris|HM010430|Nickrent_5211 Tripodanthus_flagellaris|HM010428|Nickrent_5204 Tripodanthus_flagellaris|DQ340614|Calvin_and_Wilson_AR03_10 Tripodanthus_acutifolius|HM010433|Nickrent_5323 Tripodanthus_acutifolius|HM010432|Nickrent_5321 Ixocactus_inornus|DQ340609|Calvin_and_Wilson_MX03_04 Phthirusa Tripodanthus Tripodanthus_belmirensis|HM010427|Nickrent_5050 nesting in Tripodanthus_acutifolius|HM010434|Nickrent_5325 Tripodanthus_acutifolius|HM010431|Nickrent_5318 Psittacanthinae Tripodanthus_acutifolius|HM010444|Nickrent_5551 Tripodanthus_acutifolius|HM010440|Nickrent_5350 Tripodanthus_acutifolius|HM010443|Nickrent_5550 Tripodanthus_acutifolius|HM010435|Nickrent_5326 Struthanthus_woodsonii|EU544510|Nickrent_2761 Cladocolea_gracilis|EU544482|Nickrent_3066 Struthanthus_oerstedii|EU544509|Nickrent_2728 Cladocolea_cupulata|DQ340612|Calvin_and_Wilson_MX03_08 Cladocolea/Struthanthus not sorted Cladocolea_mcvaughii|DQ340611|Calvin_and_Wilson_MX03_09 Struthanthus_orbicularis|DQ340607|Calvin_and_Wilson_CR01_02 Oryctanthus_occidentalis|DQ340613|Calvin_and_Wilson_CR01_11 Passovia pyrifolia/ Phthirusa_pyrifolia|EU544504|Nickrent_2762 Phthirusa_pyrifolia|DQ340608|Calvin_and_Wilson_CR01_03 Oryctanthus_occidentalis|EU544501|Nickrent_2763 Oryctanthus cross-placed! Psittacanthus_schiedeanus|DQ340610|Calvin_and_Wilson_CR01_09 Ligaria_cuneifolia|DQ340604|Calvin_and_Wilson_CL03_01 Ligaria_cuneifolia|DQ442940|G_Amico_76_BCRU Notanthera_heterophylla|DQ340606|Calvin_and_Wilson_Cl03_03 Notanthera_heterophylla|DQ442939|G_Amico_151_BCRU Ligarinae Tupeia_antarctica|EF464494|Nickrent_2742 Tupeinae Tupeia_antarctica|DQ340601|Calvin_and_Wilson_NZ98_02 Desmaria_mutabilis|EF464486|Nickrent_4510A Desmaria_mutabilis|DQ340603|Calvin_and_Wilson_CL03_07 Notantherinae Nuytsia_floribunda|DQ788716|Nickrent_3080 Nuytsieae 0.01 exp. subst./site Nuytsia_floribunda|DQ340618|Calvin_and_Wilson_AU01_22