A MOLECULAR PHYLOGENY OF SANTALALES

Daniel L. Nickrent

Valéry Malécot

Sequences of SSU rDNA and/or chloroplast rbcL have to date been obtained from 66 of the 155 total genera in the order: Viscaceae (7 of 7), Santalaceae (25 of 37), Opiliaceae (6 of 10), Misodendraceae (1 of 1), Loranthaceae (3 of 74), and Olacaceae (24 of 26). In global analyses using multiple genes, Santalales appears (unresolved) at the base of the core eudicots, thus sequences from six outgroup representatives were used. Sequences of both genes were manually aligned and phylogenetic analyses using maximum parsimony (MP) were conducted.

MP analysis of the 81 taxon matrix resulted in six trees of length 3783, the strict consensus of which was congruent with trees previously published. Strongly monophyletic families are Opiliaceae, Loranthaceae and Viscaceae. Santalaceae (including Eremolepidaceae) and Olacaceae are paraphyletic. As previously reported, Schoepfia is sister to Misodendrum and not closely related to the remaining Olacaceae. These analyses resulted in seven clades within the paraphyletic "Olacaceae." Converting these data to a classification results in one that differs in several aspects from the tribal classification of Sleumer (1984). Positive reports of parasitism and nonparasitism suggest that only two clades (tribes Olaceae and Ximeneae) are parasitic, thus indicating that this life form may have evolved only once in Santalales. Seven clades within "Santalaceae" are discussed, some of which agree well with previously published subfamilial classifications. Although greater sampling of taxa and genes has increased resolution of phylogenetic relationships in the order, complete sampling and likely other gene sequences will be required to fully resolve relationships in Santalales.

Among angiosperms, the most diverse assemblage of life forms is found in the sandalwood order (Santalales) because members may be nonparasites, root parasites or aerial parasites (mistletoes). Clarification of phylogenetic relationships among the component genera and families of this order thus presents opportunities to better understand the evolution of this heterotrophic mode. As traditionally defined, Santalales consist of seven families: Eremolepidaceae, Loranthaceae, Misodendraceae, Opiliaceae, Olacaceae, Santalaceae and Viscaceae. Previous molecular phylogenetic investigations have utilized nuclear small-subunit (SSU) rDNA sequences (Nickrent and Duff 1996, Nickrent and Franchina 1990) as well as sequences from plastid rbcL (Nickrent, et al. 1998). In each of these studies, however, sampling was limited, particularly in Santalaceae and Olacaceae. We report here our continuing studies whereby greater generic sampling has allowed increasingly detailed inferences of phylogenetic relationships among members of the sandalwood order.

MATERIALS AND METHODS

DNA extraction, PCR conditions, primer sequences, and sequencing methodology have been previously reported (Nickrent 1994). The sequences used here were obtained using both manual and automated methods. In global analyses using multiple genes (Soltis, et al. 2000) Santalales appears (unresolved) at the base of the core eudicots. For this reason, six taxa (different genera and families) were chosen to serve as outgroups. Sequences of SSU rDNA, rbcL, or both genes were obtained for 75 ingroup Santalales taxa and the six outgroups (81 total taxa). For taxa lacking one gene, these sites were scored as missing (9 rDNA, 10 rbcL sequences). Separate alignments for each gene was made and these were then combined into a single matrix and analyzed using maximum parsimony (MP). All phylogenetic analyses were conducted using PAUP* (Swofford 1998). Bootstrap (BS) analyses were conducted to test the support for various clades using 100 replicates. The entire data set, with non-Santalales outgroups, was called Santalales matrix. To test the stability of tree topologies and to allow more thorough bootstrap analyses, alternate (smaller) data sets were constructed using different outgroups. These were 1) Santalales2 matrix: non-Santalales outgroups removed, tree rooted with Olacaceae clade A (Stromboseae), 2) Olacaceae matrix: non-Santalales outgroups, Opiliaceae, Santalaceae, and Viscaceae removed, tree rooted with Olacaceae clade A (Stromboseae), 3) Santalaceae matrix: non-Santalales outgroups and Olacaceae removed, tree rooted with Misodendrum and Loranthaceae.

RESULTS AND DISCUSSION

Complete generic sampling was achieved for the mistletoe families Viscaceae (7 genera), "Eremolepidaceae" (3 genera), and Misodendraceae (1 genus) whereas sampling was incomplete for the remaining four families: Olacaceae (24 of 26 genera) Santalaceae (25 of 37 genera), Opiliaceae (6 of 10 genera) and Loranthaceae (3 of 74 genera). Thus, 66 genera (75 species) of the 155 total genera in the order were included in the 81 taxon data matrix. Because nearly half of the genera in Santalales are Loranthaceae, sampling to date constitutes 78% of the remaining generic diversity (63 of 81 genera).

Analyses of SSU rDNA and rbcL separately. Given that the nuclear and chloroplast genes used in this study represent different subcellular genomes (each with different modes of inheritance), they were treated as process partitions and analyzed separately. The strict consensus trees of each were similar in overall arrangement of major clades, but differed in detail. As previously observed (Nickrent, et al. 1998), analysis of the rbcL data alone (1408 nucleotides) places Opiliaceae as one of several clades within a paraphyletic Santalaceae. Other differences between the SSU rDNA and rbcL partition analyses can be seen among Olacaceae relationships; SSU rDNA data (1710 nucleotides) suggest monophyly for the family whereas rbcL indicates most relationships are unresolved (a basal polytomy). In general, the results from analyses of these separate partitions should be viewed with caution given the low number of parsimony-informative characters derived from each (272 and 379, respectively). Although formal congruence tests were not conducted, a combined matrix was constructed to increase the total number of informative characters.

Analysis of two gene matrix with MP. Upon analysis of the Santalales.matrix data set with MP, the combined rbcL and nuSSU rDNA matrix (3118 characters) yielded six most parsimonious trees of length 3783 (C.I. minus uninformative sites = 0.3951, R.I. = 0.6040, R.C. = 0.2933), the strict consensus of which is shown here. The topology of this tree is generally congruent with ones previously reported (Nickrent and Duff 1996, Nickrent, et al. 1998, Nickrent and Malecot 2000) with Olacaceae basal followed by a Loranthaceae/Misodendraceae clade, Opiliaceae, a grade comprising Santalaceae/Eremolepidaceae and then Viscaceae. Analysis of the Santalales2 matrix gave essentially an identical topology as that matrix incorporating non-Santalales outgroups. Moreover, relationships were also retained for the Olacaceae matrix and Santalaceae matrix, thus these were used to obtain bootstrap support values discussed below.

Bootstrap support for a monophyletic Santalales is moderate (61%). Santalales minus Olacaceae are monophyletic with high BS support (> 90%). High BS support (100%) was obtained for Loranthaceae, Opiliaceae and Viscaceae whereas Olacaceae and Santalaceae (including Eremolepidaceae) are paraphyletic. On the strict consensus trees, these "families" each show a series of clades that derive from a basal polytomy. Given that the increased taxon sampling reported here mainly concerns Olacaceae and Santalaceae, only these families will here be discussed in detail. Clades for Olacaceae and Santalaceae are given letter designations (A-G) as well as tribal names (see discussion of Classification below).

Clade A - Strombosieae. In analyses where non-Santalales are used, this clade (minus Maburea) receives high (100%) BS support. This clade is here composed of five genera traditionally classified in tribe Anacoloseae: Diogoa, Strombosiopsis, Strombosia, Scorodocarpus and Tetrastylidium (but not Anacolosa, Cathedra, and Phanerodiscus as circumscribed by Sleumer (1984)). Maburea, a recently described genus from Guyana, is seen to be basalmost among this clade (rbcL only), albeit with low BS support (51%). This BS value increases to 64% when the Olacaceae matrix is analyzed. Although not strongly associated with Clade A, these results also do not support a relationship between Maburea and Couleae as previously suggested (Maas, et al. 1992). Further clarification of the position of Maburea may increase when the nuSSU rDNA partition is available. Morphological synapomorphies for Clade A can be found in pollen morphology, specifically tricolporate pollen with rectangular endoaperture (Lobreau-Callen 1981, Feuer 1977). Four of these genera (Strombosia, Strombosiopsis, Diogoa, and Tetrastylidium) show anatomical similarities including presence of epidermal crystals, petiole vascularization, and ray type (Baas 1982, van den Oever 1984). Moreover, all six genera also share an insertion of two to four additional amino acid codons at the 3' end of the rbcL gene, just 5' of the stop codon (Nickrent and Malecot 2000).

Clade B, Heisterieae. Sequences for both genes were available for Heisteria concinna whereas only rbcL sequences were available for two other species. The Heisteria clade receives high BS support (90-100%) in all analyses, however, the placement of this clade within the family is less certain. The Santalales, Santalales2, and Olacaceae matrices suggest a relationship to clade D (Couleae), albeit with low BS support. Palynological features in Heisteria are similar to Couleae (Feuer 1977, Lobreau-Callen 1980), yet wood anatomy links the genus to Brachynema (not sampled here) (van den Oever 1984). Its leaf anatomy is intermediate between tribes Couleae on the one hand and clade E (Aptandreae) on the other (Baas 1982). A relationship with Chaunochiton (the only other genus placed in tribe Heistereae by Sleumer) is not supported by this study. Previous palynological and leaf/wood anatomical studies (Baas 1982, Feuer 1977, Lobreau-Callen 1980, Reed 1955, van den Oever 1984) and the mobile nature of the Heisteria clade in these molecular analyses call for additional data to test the placement of tribe Heistereae in the family.

Clade C, Erythropaleae. Clade C, represented solely by Erythropalum, did not receive high BS support, thus its position remains volatile within the family. The position of this genus changes depending upon the choice of outgroup. For example, when using only Glycine and Arabidopsis as outgroups, Erythropalum is sister (with low BS support) to Heisteria. Using the Santalales matrix, this genus is sister to clade D (Couleae). The position of this genus also differs when the nuSSU rDNA and rbcL gene partitions are analyzed separately, thus demonstrating a possible case of incongruence between the two process partitions. Having tendrils, large, lax, and slender inflorescences, and peculiar anatomy (Baas 1982, Sleumer 1984), Erythropalum is aberrant in Olacaceae, and has therefore at times been classified in its own family (Erythropalaceae Blume). However, the genus does possess the free-central, pendulous placentation that is characteristic of Olacaceae (Fagerlind 1946) and the molecular analyses reported here demonstrate affinity with Santalales and Olacaceae. Ling (1982) considered this genus to belong to the most evolved subfamily of Olacaceae, Erythropaloideae, a relationship not supported by our molecular data. At present, it is premature to state whether the molecular data support or refute the concept of separate familial status for Erythropalum.

Clade D, Couleae. Moderate BS support (58%) was obtained for Clade D composed of Coula, Minquartia, and Ochanostachys - a clade that corresponds exactly with the traditional classification of Couleae by Sleumer (1984). Morphological synapomorphies for the tribe include the presence of dendritic hairs, lignified epidermal cells, laticifers (Baas 1982), long vessels (van den Oever 1984) and the absence of an infratectum on the pollen (Lobreau-Callen 1980). This tribe is traditionally recognized as very homogeneous and is often considered the most primitive one in the family (Michaud 1966, Stauffer 1961). Although near the base, this tribe appears from molecular data to be more derived than Stromboseae.

Clade E, Aptandreae. This clade received moderately high (81%) BS support following analysis of the Olacaceae matrix and here consists of five genera: Aptandra, Ongokea, Chaunochiton, Cathedra and Phanerodiscus. Tribe Aptandeae, as defined by Sleumer (1984), included Aptanda, Ongokea and Harmandia (latter not sampled here). These molecular data suggest adopting an expanded concept for the tribe to include Chaunochiton (formerly Heistereae) and Cathedra and Phanerodiscus (formerly Anacoloseae). These taxa all have fruits with expanded floral structures. In the case of Aptandra, Ongokea,Harmandia, and Chaunochiton the fruit is surrounded by an accrescent calyx. For Phanerodiscus, the fruit is surrounded by membranous sack derived from the receptacular cupule (not the discal cupule). In Cathedra, the fruit is closely enveloped by the accrescent discal cupule. The clade of Chaunochiton plus Ongokea and Aptandra gains further support from common leaf and wood features.

Clade F, Ximenieae. This clade is strongly supported by this molecular analysis (99% BS) as well as by several morphological and anatomical features such as inflorescence type, stamen number, and wood characteristics. It may be mentioned that Sleumer (1984) placed Malania in tribe Olaceae (but with some doubts) whereas the Chinese descriptors (Lee 1980, Ling 1982) clearly assigned Malania to tribe Ximenieae. If the position of Ximeneae is correct with the family and order, this clade represents the first occurrence of haustorial root parasites in Santalales. Two other genera, Curupira and Douradoa, remain to be sampled but are placed in this tribe based on morphological features.

Clade G, Olaceae. This well-supported clade (100% BS) is composed of Anacolosa, Dulacia, Olax, and Ptychopetalum. The latter three genera were all classified in tribe Olaceae by Sleumer (1984), but molecular data do not support the inclusion of Malania which is clearly a component of clade F (Ximeneae). The strong association of Anacolosa with Olax (100% BS) is surprising given the absence of any obvious palynological, leaf, or wood anatomical characters that link these genera. Because Anacolosa does not appear to be closely related to taxa now classified as Stromboseae, the tribal name Anacoloseae of Sleumer must be abandoned. This name cannot be used for Clade G because this clade contains Olax, the type genus for the family.

Non-Olacaceae. As previously reported (Nickrent, et al. 1998, Nickrent and Malecot 2000), Schoepfia is not closely related to Olacaceae. Schoepfia is sister to the mistletoe genus Misodendrum, and this clade is itself sister to Loranthaceae, although both of these clades receive only moderate BS support (73% and 65%, respectively) when the Santalales matrix is analyzed. The relationship of Schoepfia to Loranthaceae was noted as early as 1830 by De Candolle, and traditional infrafamilial classifications have since placed this genus in a separate subfamily (Engler 1897, Sleumer 1984) or its own family, Schoepfiaceae (Blume 1851, Gagnepain 1910, Tieghem 1896). More recently, a relationship of this genus with Santalaceae was suggested based upon wood anatomy (Norverto 1993), thus further indicating its derived position relative to Olacaceae. Unlike any other Olacaceae, Schoepfia shares with Loranthaceae a reduced calyx on its monochlamydous epigynous flowers. Sequences of the Australian root parasitic loranthaceous genera Atkinsonia and Nuytsia are in progress and their inclusion may help clarify the position of Schoepfia within Santalales.

Three other genera have previously been proposed to be members of Olacaceae: Brachynema, Octoknema, and Worcesterianthus. To date, no molecular data are available to document whether or not these genera are components of Santalales. Octoknema has been placed in its own family (Octoknemaceae). Brachynema may be a component of Ebenaceae. Worcesterianthus is a synonym for Microdesmis, a genus in Pandaceae (van Steenis 1955).

Analyses of the Santalales and Santalales2 matrices place Opiliaceae at the base of Santalaceae, but with low BS support. This affinity with Santalaceae is further shown upon analysis of the Santalaceae matrix where Opiliaceae is one of a number of clades that derive from a basal polytomy. Using Loranthaceae and Misodendraceae as outgroups, the clade composed of Opiliaceae, Santalaceae and Viscaceae receives strong BS support (88%). As mentioned above, the association of Opiliaceae with Santalaceae derives mainly from the rbcL partition which contributes more parsimony informative sites than nuSSU rDNA. These results highlight the need for complete taxon sampling as well as the use of another gene that is evolving at a higher rate to fully resolve relationships in the family.

Clade A, Comandreae. The relationship between Comandra and Geocaulon has long been noted (Pilger 1935) and is supported by this molecular study (100% BS). Although this clade occurs at the base of a paraphyletic Santalaceae in the shortest trees from the Santalales and Santalales2 matrices, this position did not receive high BS support. Thus, Clade A should not be considered the basalmost clade in the Santalaceae but only one among many contenders for this position.

Clade B, Thesieae. High BS support (90%) was obtained for the clade containing Osyridocarpus and Thesium. This relationship is in agreement with the classification of Thesieae by Pilger (1935) and Stauffer (1961) and both possess heteropolar pollen (Stearn 1972). Arjona and Quinchamalium were also classified in this tribe by Pilger, but the former remains unsampled and the latter (nuSSU rDNA sequence only) is not included in this clade but in Clade C. The inclusion of Buckleya in this clade is tenuous and received only moderate BS support (60%). This position appears to be arising from the rbcL partition; nuSSU analyzed separately places this genus with members of Clade E (Santaleae).

Clade C, Pyrularieae. This clade received moderately strong BS support (83%) and contains a well supported (91%) clade of two South American genera, Acanthosyris and Jodina and a moderately strongly supported (85%) clade composed of Pyrularia (U.S. and Asia), Quinchamalium (S. America) and Okoubaka (Africa). Aside from Quinchamalium, all of these genera have all been classified in tribe Osyrideae by Pilger (1935) (which should be called Santaleae DC, see (Stearn 1972)). When first described by Normand and Pellegrin (1946), Okoubaka was classified in Octoknemaceae whose sole genus (Octoknema) was transferred to Olacaceae by Fagerlind (1948). The placement of the genus in Santalaceae was made by Stauffer (1957) who astutely noted an affinity with Pyrularia and Scleropyrum (latter not sampled) based on habit, fruit and wood anatomy.

Clade D, Anthoboleae. The clade containing Exocarpos and Omphacomeria received 99% BS support and contains two of the three genera of tribe Anthoboleae (Pilger 1935, Stauffer 1959) (Anthobolus not sampled). These genera are distinct in the family owing to the presence of hypogynous vs. epigynous flowers and fruits that occur at ends of swollen pedicels.

Clade E, "Santaleae". Although here comprising seven genera, it is likely that this "tribe" is not monophyletic. Four of the genera (Colpoon, Osyris, Rhoiacarpos, and Nestronia) formed a well supported clade (97% BS). The first three genera are African wheres Nestronia is endemic to the eastern U.S. A relationship between the African genera has been recognized (Stauffer 1961) and all were classified in Santaleae (Osyrideae) by Pilger (1935). The remaining three genera of "Santaleae" (Santalum, Myoscilos, and Mida) arise from the basal polytomy within Santalaceae. Presently unsampled genera that are potentially part of this clade are Kunkeliella and Nanodea. With some exceptions (e.g. Nestronia), most members of this clade have bisexual flowers.

Clade F, "Eremolepideae". Bootstrap support is low for the clade containing Antidaphne, Eubrachion, and Lepidoceras, however, these three genera have traditionally been classified as part of the family Eremolepidaceae (Kuijt 1988), hence they are here treated together. As has been previously discussed (Nickrent and Duff 1996, Nickrent, et al. 1998, Nickrent and Soltis 1995), this group of three genera is not monophyletic, however, 76% BS support is obtained for the Eubrachion and Lepidoceras clade. The paraphyletic "Eremolepideae" arises from the polytomy at the base of the Santalaceae clade. If one chooses to recognize only monophyletic groups as families (APG 1998), then to recognize Santalaceae requires the inclusion of Viscaceae. Alternatively, one could name all the monophyletic groups (clades) within Santalaceae s. lat. as separate families, an action that would be required if one wishes to maintain consistency and recognize Eremolepidaceae as a family.. The state of knowledge about relationships among the various clades ("tribes") of Santalaceae is currently incomplete, thus we consider it prudent to abstain from reclassification until more information is available.

Clade G, Amphorogyneae. At present, the most poorly sampled tribe in Santalaceae is Amphorogyneae, where only three of the nine genera have been sequenced. Clade G received high BS support (89%) as did the clade composed of two stem parasites Dufrenoya and Dendrotrophe (93% BS). Choretrum is a root parasite from Australia and is (currently) basal within the tribe. Tribe Amphorogyneae was proposed by Stauffer (1969) and validly published by Stearn (1972). Unsampled genera for this tribe include Amphorogyne, Daenikera, Dendromyza, Leptomeria, Phacellaria, and Spirogardnera. Given a possible relationship of these genera to Viscaceae, increased taxon density in this tribe is essential to fully understand the evolution of the mistletoe habit.

Classification and Sampling

The molecular phylogenetic results reported here have been used to construct modified classifications of Olacaceae and Santalaceae. Attempts were made to preserve tribal names as proposed in the classification of Olacaceae by Sleumer (1984). Similarly, the classification of Santalaceae by Pilger (1935), later modified in the Santalales Studien series by Stauffer (see references) was used as a starting point. In both cases, modifications of these existing classifications was warranted given the molecular phylogenetic results. Unsampled taxa have been placed (provisionally) in the tribes dictated by the traditional classifications. For Olacaceae, only four genera remain to be sequenced and included in the analyses (underlined = no DNA samples currently available): Engomegoma, Harmandia, Curupira and Douradoa. For Santalaceae, 12 of the 37 genera remain to be sequenced, 6 of which are in Amphorogyneae: Arjona, Cervantesia, Anthobolus, Kunkeliella, Nanodea, Scleropyrum, Amphorogyne, Daenikera, Dendromyza, Leptomeria, Phacellaria, and Spirogardnera. We feel it is important to have complete or nearly complete taxon sampling, as well as sequences from a nuclear (rDNA) and plastid (rbcL) gene before publishing a revised classification of Olacaceae and Santalaceae. These data will be particularly informative when viewed in parallel with data from biogeography, morphology, anatomy and palynology. Given the size of the order and the evolutionary diversity among its component taxa, sequence data from two genes (greater than 3000 nucleotides) is insufficient to resolve all relationships. Thus, additional sequences will be required, as well as complete or nearly complete taxon sampling, to advance our state of knowledge on this important group of parasitic plants.

Acknowledgements

With respect to sampling, the authors wish to acknowledge the numerous individuals who have collected tissue samples for DNA extraction; without such assistance this research would not be possible. For a table listing collection information for all the species used in the study reported here, see this web page. Thanks also are extended to the following individuals who provided laboratory assistance: Miguel A. García, Jonathan Cabrera, Erica Nicholson, Mark O'Dell, and R. Joel Duff. This work has been supported by grants from the US National Science Foundation (DEB-9407984 and MCB-9808752).

APG (1998) An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85. 531-553.

BAAS P (1982) Leaf anatomy and classification of the Olacaceae, Octoknema and Erythropalum. Allertonia 3. 155-210.

BLUME CL (1851) Flora Javae nec non insularum adjacentium. (ed. Fischer JB), Frank, J., Brussels.

ENGLER A (1897) Olacaceae in Die Naturlichen Pflanzenfamilien, Nachträge zu Teil II-IV, (ed. Engler A and Prantl K), pp. 144-149.

FAGERLIND F (1946) Gynöceummorphologische, embryologie und systematische stellung der gattung Erythropalum. Svensk Botanisk Tidskrift. 40. 9-14.

FAGERLIND F (1948) Beiträge zür Kenntnis der Gynäceummorphologie und Phylogenie der Santalales-Familien. Svensk Botanisk Tidskrift 42. 195-229.

FEUER SM (1977) Pollen morphology and evolution in the Santalales s. str., a parasitic order of flowering plants. Chapter I (Olacaceae). Ph.D. Thesis. University of Massachusetts. 1-144, fig. 1-184.

GAGNEPAIN F (1910) Comment faut-il comprendre la famille des Olacacées? Bulletin de la Société Botanique de France 57. 373-380.

KUIJT J (1988) Monograph of Eremolepidaceae. Systematic Botany Monographs 18. 1-60.

LEE S-K (1980) Malania, a new genus of oil-yielding plant. Bulletin of the Botanical Laboratory of the North-East Forestry Institute 6. 67-72.

LING Y-R (1982) A revision and addition of Olacaceae from China and a primitive discussion for the taxonomy and floristics of the family. Bulletin of Botanical Research 2. 7-36.

LOBREAU-CALLEN D (1980) Caractéres comparés du pollen des Icacinaceae et des Olacaceae. Adansonia 20. 29-89.

MAAS PJM et al. (1992) The identity of "Unknown Z": Maburea Maas, a new genus of Olacaceae in Guyana. Botanische Jahrbucher fur Systematik 114. 275-291.

MICHAUD M (1966) Contribution à l'étude des Olacacées d'Afrique tropicale. Mémoires de l'Institut Français d'Afrique Noire 75. 157-290.

NICKRENT DL (1994) From field to film: rapid sequencing methods for field collected plant species. Biotechniques 16. 470-475.

NICKRENT DL and DUFF RJ (1996) Molecular studies of parasitic plants using ribosomal RNA in Advances in Parasitic Plant Research, (eds. Moreno MT et al.), pp. 28-52. Junta de Andalucia, Dirección General de Investigación Agraria.

NICKRENT DL et al. (1998) Molecular phylogenetic and evolutionary studies of parasitic plants. In: Molecular Systematics of Plants II. DNA Sequencing, (eds. Soltis DE, Soltis PS and Doyle JJ), 2nd. edn. 211-241 Kluwer Academic Publishers, Boston, MA.

NICKRENT DL and FRANCHINA CR (1990) Phylogenetic relationships of the Santalales and relatives. Journal of Molecular Evolution 31. 294-301.

NICKRENT DL and MALECOT V (2000) Phylogenetic relationships of Santalales based on nuclear small-subunit (18S) rDNA and rbcL with special reference to Olacaceae. (Abstract, Botany 2000, Portland, Oregon USA).

NICKRENT DL and SOLTIS DE (1995) A comparison of angiosperm phylogenies based upon complete 18S rDNA and rbcL sequences. Annals of the Missouri Botanical Garden 82. 208-234.

NORMAND D and PELLEGRIN F (1946) Une nouvelle localité africaine du genre nouveau Okoubaka. Bulletin de Societie Botanique Francois 93. 138-139.

NORVERTO C (1993) Wood anatomy and relationships of Santalaceae: 1. Acanthosyris, Jodina, and Myoschilos. Aliso 13. 499-511.

PILGER R (1935) Santalaceae In: Die Natürlichen Pflanzen Familien, 16b (ed. Engler A and Prantl K), pp. 52-91, Engelmann, W, Leipzig.

REED CF (1955) The comparative morphology of the Olacaceae, Opiliaceae, and Octoknemaceae. Memorias da Sociedade Broteriana 10. 29-79.

SLEUMER HO (1984) Olacaceae. In: Flora Neotropica Monograph No 38, 1-159, New York Botanical Garden, New York.

SOLTIS DE et al. (2000) Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnaean Society 133. 381-461.

STAUFFER HU (1957) Santalales-Studien I. Zur Stellung der Gattung Okoubaka Pellegrin et Normand. Berliner Schweize Botanische Geselschaft 67. 422-427.

STAUFFER HU (1959) Santalales-Studien IV. Revisio Anthobolearum; eine Morphologische Studie mit Einschluss der Geographie, Phylogenie und Taxonomie. Mittleitung Botanische Museum Zürich 213.

STAUFFER HU (1961) Santalales-Studien V. Afrikanische Santalaceae I. Osyris, Colpoon und Rhoiacarpos. Vierteljahrsschr. Naturf. Ges. Zürich 106. 387-400.

STAUFFER HU (1961) Santalales-Studien VI. Afrikanische Santalaceae II. Osyridicarpos. Vierteljahrsschr. Naturf. Ges. Zürich 106. 400-406.

STAUFFER HU (1961) Santalales-Studien. VIII. Zur Morphologie und Taxonomie der Olacaceae-Tribus Couleae. Vierteljahrsschr. Naturf. Ges. Zürich 106. 412-418.

STAUFFER HU (1969) Santalales Studien X. Amphorogyneae eine Neue Tribus der Santalaceae. Vierteljahrsschr. Naturf. Ges. Zürich 114. 49-76.

STEARN WT (1972) Kunkeliella, a new genus of Santalaceae in the Canary Islands. Cuaderno Botanica Canariensis 16. 11-26.

SWOFFORD DL (1998) PAUP*: phylogenetic analysis using parsimony, version 4.0b6, Sinauer Associates, Sunderland, Mass.

TIEGHEM PV (1896) Sur les phanérogames à ovule sans nucelle, formant le groupe des innucellées ou Santalinées. Bulletin de Societie Botanique Francois 43. 543-577.

VAN DEN OEVER L (1984) Comparative wood anatomy of the Olacaceae. In: Proceedings of the Pacific Regional. Wood Anatomy Conference, Tsukuba, Japan, 177-178.

VAN STEENIS CGGJ (1955) Reductions of the genera Schizopremna Baill. (Verbenaceae) and Worcesterianthus Merr. (Olacaceae). Acta Botanica Neerlandica 4. 477-480.

SIUC / College of Science / Parasitic Plant Connection
URL: http://www.parasiticplants.siu.edu/Sants.IPWC.html
Last updated: 13-Jun-06 / dln