Cover Page

The handle http://hdl.handle.net/1887/35731 holds various files of this Leiden University dissertation

Author: Mennes, Constantijn Title: Antiquities of the rainforest : evolution of mycoheterotrophic angiosperms growing on Glomeromycota Issue Date: 2015-10-06 Appendix B: Supplementary Figures.

181 Antiquities of the rainforest

182 Sciaphila albescens 1 Sciaphila albescens 4 18S Sciaphila albescens 2 Sciaphila albescens 3 Sciaphila purpurea 1 Sciaphila purpurea 2 Sciaphila nana Andruris sp. Seychellaria madagascariensis Seychellaria africana Triuris hyalina 1 Triuris hyalina 2 Triuris hexophthalma 100 Sciaphila sp. 96 Sciaphila quadribullifera Kupea martinetugei 1 Kihansia sp. Sciaphila ledermannii 2 Sciaphila ledermannii 1 Sciaphila tenella Sciaphila picta Pandanus sp. 2 Pandanus tectorius Pandanus baptistii Pandanus tectorius var. laevis Pandanus pacificus Pandanus labyrinthicus Pandanus veitchii Pandanus furcatus Pandanus stenophyllus Pandanus cf. dubius Pandanus purpurascens Pandanus sp. 1 Pandanus utilis Freycinetia cumingiana 1 100 Freycinetia cumingiana 2 Freycinetia philippinensis 88 Freycinetia sp. 1 94 Freycinetia sp. 3 100 Freycinetia sp. 2 * Pandanus montanus 70 Asplundia cf. venezuelensis Asplundia rigida Dicranopygium atrovirens Carludovica palmata 1 98 Carludovica palmata 2 Ludovia lancifolia 58 Cyclanthus bipartitus Croomia pauciflora 100 Stemona sp. Stemona tuberosa var. ternatensis 66 Pentastemona sumatrana Barbacenia elegans 100 Vellozia elegans 67 Acanthochlamys bracteata Dioscorea rockii Dioscorea communis Trichopus sempervirens Tacca integrifolia Nietneria paniculata Narthecium ossifragum Aletris lutea Orontium aquaticum

0.02 * This clade is not recovered in the ML analysis Fig. S1. Phylogeny of based on an ML analysis of the 18S rDNA dataset. Values above branches are Bayesian posterior probabilities (expressed as percentages); those below branches are bootstrap support percentages. The scale bar shows the number of substitutions per site. Clades which are not recovered in the ML analysis, but only in the BI analysis are indicated by a vertical bar. Freycinetia sp. 3 Freycinetia sp. 1 atpA Freycinetia cumingiana 1 Freycinetia sp. 2 Freycinetia cumingiana 2 Freycinetia philippinensis Pandanus montanus Pandanus purpurascens Pandanus utilis Pandanus stenophyllus Pandanus furcatus Pandanus veitchii 67 Pandanus tectorius 68 Pandanus cf. dubius Pandanus labyrinthicus Pandanus sp. 1 Pandanus pacificus Pandanus baptistii Pandanus sp. 2 54 Pandanus tectorius var. laevis Asplundia rigida ** 99 Asplundia cf. venezuelensis 78 Ludovia lancifolia 93 Carludovia palmata 2 60 Dicranopygium atrovirens Carludovica palmata 1 Cyclanthus bipartitus Stemona sp. 100 Croomia pauciflora Stemona tuberosa var. ternatensis 61 Pentastemona sumatrana* Sciaphila albescens 1 Sciaphila albescens 2 Sciaphila albescens 3 Sciaphila purpurea 1 Seychellaria africana Sciaphila nana Sciaphila sp. 99 Sciaphila quadribullifera 71 Triuris hyalina 1 Triuris hyalina 2 schismatica Sciaphila picta 100 Sciaphila ledermannii 2 96 Sciaphila ledermannii 1 Kupea martinetugei 2 Kupea martinetugei 1 Kihansia sp. Vellozia elegans 100 Barbacenia elegans 96 Acanthochlamys bracteata Aletris lutea Nietneria paniculata Narthecium ossifragum Trichopus sempervirens Tacca integrifolia Dioscorea rockii Dioscorea communis Triuris hexophthalma Orontium aquaticum * Pentastemona is embedded in Pandanaceae in the BI analysis 0.01 ** This clade is not recovered in the ML analysis Fig. S2. Phylogeny of Pandanales based on an ML analysis of the mitochondrial atpA dataset. Values above branches are Bayesian posterior probabilities; those below branches are bootstrap support percentages. The scale bar shows the number of substitutions per site. Clades which are not recovered in the ML analysis, but only in the BI analysis are indicated by a vertical bar.

183 Antiquities of the rainforest

Pandanus montanus matR Pandanus utilis Pandanus purpurascens Pandanus sp. 2 Pandanus tectorius var. laevis Pandanus baptistii Pandanus tectorius Pandanus veitchii Pandanus pacificus Pandanus stenophyllus Pandanus labyrinthicus Pandanus cf. dubius Pandanus sp. 1 100 Pandanus furcatus Freycinetia cumingiana 2 99 Freycinetia cumingiana 1 Freycinetia philippinensis Freycinetia sp. 2 Freycinetia sp. 3 100 Freycinetia sp. 1 98 Carludovica palmata 2 Carludovica palmata 1 Asplundia rigida Asplundia cf. venezuelensis 100 Dicranopygium atrovirens Ludovia lancifolia 85 Cyclanthus bipartitus Sciaphila albescens 4 Sciaphila albescens 3 Sciaphila albescens 1 Sciaphila albescens 2 Sciaphila purpurea 2 Sciaphila purpurea 1 100 Seychellaria madagascariensis 87 Sciaphila nana Seychellaria africana Andruris sp. Sciaphila ledermannii 1 Sciaphila ledermannii 2 Triuris hexophthalma Triuris hyalina 1 Sciaphila sp. 100 Sciaphila quadribullifera 100 Sciaphila picta 99 Sciaphila tenella 100 Kupea martinetugei 1 100 Kupea martinetugei 2 Kihansia sp. 75 Stemona tuberosa var. ternatensis Stemona sp. 100 Croomia pauciflora 99 Pentastemona sumatrana 100 Barbacenia elegans 100 Vellozia elegans Nietneria paniculata Narthecium ossifragum Aletris lutea Dioscorea rockii Dioscorea communis Trichopus sempervirens Tacca integrifolia Orontium aquaticum Fig. S3. Phylogeny of Pandanales based on0.02 an ML analysis of the mitochondrial matR dataset. Values above branches are Bayesian posterior probabilities; those below branches are bootstrap support percentages. The scale bar shows the number of substitutions per site.

Pandanus furcatus nad1b-c Pandanus sp. 1 Pandanus veitchii Pandanus tectorius var. laevis Pandanus labyrinthicus Pandanus stenophyllus Pandanus sp. 2 Pandanus utilis Pandanus tectorius Pandanus cf. dubius Pandanus montanus Pandanus baptistii 82 Pandanus purpurascens 57 Pandanus pacificus Freycinetia sp. 1 Freycinetia sp. 3 Freycinetia cumingiana 2 Freycinetia philippinensis Freycinetia cumingiana 1 Freycinetia sp. 2 Asplundia cf. venezuelensis Asplundia rigida Dicranopygium atrovirens Ludovia lancifolia 98 Carludovica palmata 1 88 Carludovica palmata 2 Cyclanthus bipartitus 92 Sciaphila albescens 3 Sciaphila albescens 1 71 Sciaphila purpurea 1 64 Sciaphila ledermannii 2 Sciaphila ledermannii 1 64 Kihansia sp. Kupea martinetugei 1 Stemona sp. 98 Stemona tuberosa var. ternatensis 61 Pentastemona sumatrana Sciaphila nana Nietneria paniculata Narthecium ossifragum 87 Aletris lutea 53 100 Barbacenia elegans 100 Vellozia elegans Dioscorea rockii Dioscorea communis Tacca integrifolia Trichopus sempervirens Orontium aquaticum

0.06 Fig. S4. Phylogeny of Pandanales based on an ML analysis of the mitochondrial nad1 b-c intron dataset. Values above branches are Bayesian posterior probabilities; those below branches are bootstrap support percentages. The scale bar shows the number of substitutions per site.

184 Pandanus sp. 2 Complete Pandanus tectorius Pandanus baptistii Pandanus tectorius var. laevis dataset Pandanus purpurascens Pandanus utilis Pandanus montanus Pandanus sp. 1 Pandanus furcatus Pandanus veitchii Pandanus stenophyllus Pandanus labyrinthicus 100 Pandanus pacificus Pandanus cf. dubius 100 Freycinetia cumingiana 2 Freycinetia philippinensis Freycinetia cumingiana 1 Freycinetia sp. 1 100 Freycinetia sp. 3 Freycinetia sp. 2 100 Asplundia cf. venezuelensis Asplundia rigida Dicranopygium atrovirens Carludovica palmata 2 65100 Carludovica palmata 1 51 Ludovia lancifolia 100 Cyclanthus bipartitus Stemona sp. Stemona tuberosa var. ternatensis 100 Croomia pauciflora 100 Pentastemona sumatrana * Sciaphila albescens 2 10086 Sciaphila albescens 1 Sciaphila albescens 4 100 100 Sciaphila albescens 3 100 100 Sciaphila purpurea 2 100 99 Sciaphila purpurea 1 100 100 Sciaphila nana 93 94 76 73 Andruris sp. 10065 96 Seychellaria africana 52 100 Seychellaria madagascariensis 100 100 Sciaphila ledermannii 2 Sciaphila ledermannii 1 86 100 100 Sciaphila sp. 100 Sciaphila quadribullifera 100 92 100 Triuris hyalina 1 100 93 Triuris hyalina 2 100 100 61 100 Triuris hexophthalma 91 Lacandonia73 schismatica 100 97 Sciaphila picta Sciaphila tenella 100 100 100 Kupea martinetugei 1 100 Kupea martinetugei 2 100 Kihansia sp. 100 Barbacenia elegans Vellozia elegans 100 Acanthochlamys bracteata Dioscorea rockii Dioscorea communis Trichopus sempervirens Tacca integrifolia Narthecium ossifragum Nietneria paniculata Aletris lutea Orontium aquaticum

0.02 * This clade is not recovered in the BI analysis

Fig. S5. A four-gene ML analysis of and relatives including several Triuridaceae taxa (in grey) with less than three genes (a portion of this is shown in Fig. 6). Values above branches are Bayesian posterior probabilities; those below branches are Bootstrap support percentages. The scale bar shows the number of substitutions per site.

185 Antiquities of the rainforest

Dioscoreales

Poales sp. Arecales sp. Zingiberales sp. Commelinales Asparagales

1 100 2 Liliales 100

cf. cf. 100 cf. 100

cf.

sp. 1

sp. 2

sp. 1 sp. 2

cf. // Pandanales

100 100 99 sp. 100 sp. sp.

Alismatales // Acorales

Fig. S6. Phylogeny of Liliales and related monocot orders based on a Bayesian-inference (BI) analysis of a three-locus molecular dataset (nuclear 18S rDNA and mitochondrial atpA and matR), including only spp. (blue) from Corsiaceae. The grey shaded area shows the Liliales taxa. Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. Support values are only given for Corsiaceae and the slightly conflicting Triuridaceae taxa. The scale bar indicates the number of substitutions per site.

186 Dioscoreales

Zingiberales sp. Commelinales sp. Poales sp. Arecales Asparagales

100 1 Liliales 100 2

100 100 Alismatales cf.

sp. 1

sp. 2

sp. 1 sp. 2 Pandanales

cf.

// 100 sp. 100100 100

sp. sp.

// Acorales

Fig. S7. Phylogeny of Liliales and related monocot orders based on a Bayesian-inference (BI) analysis of a three-locus molecular dataset (nuclear 18S rDNA and mitochondrial atpA, and matR), including only Arachnitis uniflora (red) from Corsiaceae. The grey shaded area shows the Liliales taxa. Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. Support values are only given for Corsiaceae and the slightly conflicting Triuridaceae taxa. The scale bar indicates the number of substitutions per site.

187 Antiquities of the rainforest

18S rDNA 96 97 97 100 90 98 100 97 100 100 98 72 100 100 91 * 100 76 99 * 100 10082 62 71 * 100 95 100 53 95 * 100 100 * 99 * *

100 5785 100 100 62 100 * 76 * 100 100 98 83 100 * 100 100 77 97 * 10062 77

62 85 * *

* These clades are not recovered in the ML analysis Fig. S8. Phylogeny of Liliales based on a Bayesian-inference (BI) analysis of a single-locus molecular dataset (nuclear 18S rDNA). Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. The scale bar indicates the number of substitutions per site.

atpA 87 100 71 81 96 52 100 100 100 91 100 * 95 89 81 53 75 100 69 100 71 83 * 100 * 76

100 68 96 98 82 70

100 87 96 100 57 83

* These clades are not recovered in the ML analysis Fig. S9. Phylogeny of Liliales based on a Bayesian-inference (BI) analysis of a single-locus molecular dataset (mitochondrial atpA). Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. The scale bar indicates the number of substitutions per site.

188 matR 100 79 99 69 100 100 98 100 96 * 100 100 96 62 93 10063 86 100 99 63 62 100 100

100 100 86 100

100 99 100 100 63 100 100 100 90 100 97 61 100 100 100 100 100 100 100 100 92 100 87 * 88 60 61 *

* These clades are not recovered in the ML analysis Fig. S10. Phylogeny of Liliales based on a Bayesian-inference (BI) analysis of a single-locus molecular dataset (mitochondrial matR). Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. The scale bar indicates the number of substitutions per site. cob 100 95 56 79 * *

100 100 100 98 91 100 77 81 89 84 54 79 63 *

100 99

99 64 100 99 100 96 96 * 93 *

* These clades are not recovered in the ML analysis Fig. S11. Phylogeny of Liliales based on a Bayesian-inference (BI) analysis of a single-locus molecular dataset (mitochondrial cob). Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. The scale bar indicates the number of substitutions per site.

189 Antiquities of the rainforest

nad5 69 *

53 * 92 61

100 95 93 57

100 100 96 69 99 90 * *

100 88 100 * 100 89 100 82

* These clades are not recovered in the ML analysis Fig. S12. Phylogeny of Liliales based on a Bayesian-inference (BI) analysis of a single-locus molecular dataset (mitochondrial nad5). Values above branches are posterior probabilities (expressed as percentages) resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the maximum-likelihood (ML) analysis. The scale bar indicates the number of substitutions per site.

190 Fig. S13. Maximum clade credibility tree resulting from the divergence-time estimation of monocots inferred using the uncorrelated lognormal relaxed-clock method, based on a two-locus molecular dataset (nuclear 18S rDNA and mitochondrial atpA), six fossil calibration points, and assuming a Yule process of speciation (see text). The scale bar and the numbers in the figure both represent time (million years ago; Ma); bars around nodes represent 95% confidence intervals.

191 Antiquities of the rainforest

Fig. S14. Phylogeny of Polygalaceae based on a Bayesian Inference (BI) analysis of the separate ITS dataset. Values above branches are posterior probabilities resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the Maximum Likelihood (ML) analysis. The scale bar indicates the number of substitutions per site.

Fig. S15. Phylogeny of Polygalaceae based on a Bayesian Inference (BI) analysis of the separate matK dataset. Values above branches are posterior probabilities resulting from the BI analysis; those below branches are bootstrap support percentages resulting from the Maximum Likelihood (ML) analysis. The scale bar indicates the number of substitutions per site.

192 193 Antiquities of the rainforest

194