Circumscription of Murraya and Merrillia (Sapindales: Rutaceae: Aurantioideae) and Susceptibility of Species and Forms to Huanglongbing

Home , Aeglopsis, Afraegle, Archidendron lucidum, Atalantia monophylla, Aurantioideae, Balsamocitrus

CIRCUMSCRIPTION OF MURRAYA AND MERRILLIA (SAPINDALES: RUTACEAE: AURANTIOIDEAE) AND SUSCEPTIBILITY OF SPECIES AND FORMS TO HUANGLONGBING

Student: Nguyen Huy Chung

Principal Supervisor: Professor G Andrew C Beattie, University of Western Sydney

Co-supervisors: Associate Professor Paul Holford, University of Western Sydney Dr Anthony M Haigh, University of Western Sydney Professor David J Mabberley, Royal Botanic Garden, Kew Dr Peter H Weston, National Herbarium of New South Wales

Date of submission: 31 August 2011

Declaration

The work reported in this thesis is the result of my own experiments and has not been submitted in any form for another degree or diploma at any university or institute of tertiary education.

Nguyen Huy Chung

31 August 2011

i Acknowledgements

I would first and foremost like to thank my supervisors, Professor Andrew Beattie, Associate Professor Paul Holford, Dr Tony Haigh, Professor David Mabberley and Dr Peter Weston for their generous guidance, academic and financial support.

My research required collection of pressed specimens and DNA of Murraya from within Australia and overseas. I could not have done this without generous assistance from many people.

I am thankful to Associate Professor Paul Holford and Ms Inggit Puji Astuti (Bogor Botanic Garden, Indonesia) who accompanied me during the collection of samples in Indonesia; to Mr Nguyen Huy Quang (Cuc Phuong National Park) and Mr Nguyen Thanh Binh (Southern Fruit Research Institute), who travelled with me during collecting trips in the southern Việt Nam and to Cuc Phuong National Park in northern Việt Nam; to Dr Paul Forster (Brisbane Botanic Garden) who accompanied me during the collection of samples in Brisbane; and to Mr Simon Goodwin who accompanied me during the collection samples in the Royal Botanic Garden, Sydney; to Dr Cen Yijing (South China Agricultural University) who travelled with Prof Beattie to collect specimens from Yingde, in Guangdong.

The following people provided Murraya specimens or DNA extracts of Murraya accessions:  Dr Trinh Xuan Hoat, Plant Protection Research Institute (PPRI), Hanoi, Việt Nam;  Ms Inggit Puji Astuti, Center for Plant Conservation, Bogor Botanic Garden, Indonesia;  Dr Siti Subandiyah, Faculty of Agriculture, Gadjah Mada University, Indonesia;  Dr Shahid Nadeem Chohan, Department of Biosciences, COMSATS Institute of Information Technology, Islamabad, Pakistan;  Dr Cen Yijing and Dr Deng Xiaoling and their students at South China Agricultural University;  Prof Zhang Dianxiang, South China Botanical Garden;

 Dr Hung Shih-Cheng, Plant Protection Department, Chiayi Agricultural Experiment Branch, Agricultural Research Institute, Council of Agriculture, Executive Yuan, Taiwan;  Dr Chandrika Ramadugu, University of California, Riverside, United States of America;  Dr Susan Halbert, Division of Plant Industry Entomology, Florida Department of Agriculture and Consumer Services, United States of America;  Dr Silvio Lopes, Fundo de Defesa da Citricultura, São Paulo, Brazil;  Mr Fanie Venter, Botanical & Environmental Consultant, Queensland, Australia;  Mr Graham Schultz, Biosecurity and Product Integrity, Department of Regional Development, Primary Industry, Fisheries and Resources, Darwin, Australia. I express my sincere thanks to all of them for their kind support.

I thank Dr Mark Williams, School of Natural Sciences, University of Western Sydney, for his guidance and support in gas chromatography (GC) analysis. I also would like to thank Elizabeth Kabanoff, Linda Westmoreland, and Gillian Wilkins, Centre for Plant and the Environment for their useful assistance. Elizabeth Kabanoff took the scanning electron micrographs in Chapter 4.

I am very grateful to the Australian Centre for International Agricultural Research (ACIAR) for offering me a ‘John Allwright Fellowship’, the Centre for Plants and the Environment, University of Western Sydney for use of its research facilities, and to Plant Protection Research Institute, Việt Nam, for supporting me to undertake the study overseas.

iii

Summary

As circumscribed by Swingle & Reece (1967), Murraya paniculata (L.) Jack [Rutaceae: Aurantioideae: Aurantieae] comprises Murraya paniculata var. paniculata (L.) Jack and three varieties, Murraya paniculata var. omphalocarpa (Hay.) Tan., Murraya paniculata var. ovatifoliolata (Engl.) Domin, and Murraya paniculata var. zollingeri Tan. The widely cultivated ornamental known variously as orange jasmine, orange jessamine or mock orange is often called Murraya exotica L. This latter epithet was regarded by Swingle & Reece (1967) as a junior synonym of Murraya paniculata and opinions of botanists on this issue have been divided for more than 200 years. My interest in the status of the ornamental form stemmed from the need for its status and origins to be resolved, as it is a transient host of a devastating disease of citrus known as huanglongbing. This disease is caused by phloem-limited putative species of bacteria ‘Candidatus Liberibacter spp.’ [α-Proteobacteria]. The vector of the disease, which throughout most of Asia and the Americas is caused by ‘Candidatus Liberibacter asiaticus’, is the Asiatic citrus psyllid (Diaphorina citri Kuwayama [Hemiptera: Psyllidae]), a species that is native to India and that may have evolved with a species of Murraya (sensu lato). Orange jasmine is the favoured host of the psyllid.

My project focused on the taxonomic status of Murraya exotica as a species but also included the relationship between Murraya and Merrillia caloxylon. My research was based on the molecular biology, morphology and phytochemistry of accessions collected from Asia, Australasia and the Americas. I also tested these accessions for the presence of the HLB pathogens.

Six regions of the maternally-inherited chloroplast genome (trnT-rps4, trnCGCA-ycf6, trnL-F, rps16, matK-5′trnK, psbM-trnDGUC) were amplified by PCR and sequenced. In addition, part of the internal transcribed spacer (ITS) region of the nuclear-encoded ribosomal RNA operon was also sequenced. The data obtained were subjected to phylogenic analysis using parsimony and Bayesian inference. The phylogenetic results derived from the chloroplast genome, as well as the nuclear ITS region, indicated that Murraya exotica is a species. This was supported by:  a phylogenetic tree based on characters of basal and terminal leaflets; principal component and redundancy analysis of elliptic Fourier descriptors based on the

shape of basal and terminal leaflets and a separate analysis of the dimensions of basal and terminal of leaflets, and the basal angles the leaflets;  discriminant function analysis of quantitative characters of basal and terminal leaflets; and  phytochemistry of leaflet and bark ethanol and n-hexane extracts.

The molecular and morphological studies also indicated that, in addition to Murraya exotica, other taxa formerly classified as Murraya paniculata comprise:  Murraya paniculata (L.) Jack from Indonesia (syn. Camunium vulgare Rumph. and Murraya sumatrana Roxb.);  Murraya asiatica n. sp. ineditus from mainland Asia; and  Murraya ovatifoliolata (Engl.) Domin., comprising small and large leaflet forms of Murraya ovatifoliolata var. ovatifoliolata (Engl.) Domin. n. comb ineditus, and Murraya ovatifoliolata var. zollingeri (Tan.) n. comb. ineditus and two hybrids:  Murraya × omphalocarpa Hay. ineditus, small and large leaflet forms, possibly with Murraya exotica as the male parent and a form of Murraya ovatifoliolata as the female parent; and  Murraya × cycloopensis ineditus from Papua, possibly with Murraya exotica as the female parent and a form of Murraya ovatifoliolata as the male parent.

I developed a key to the taxa based on leaf and leaflet characters.

I found no evidence to reunite Merrillia caloxylon with the genus Murraya.

I only detected ‘Candidatus Liberibacter asiaticus’ in Murraya exotica accessions from Brazil and China.

Table of Contents

Acknowledgements ...... ii Summary ...... iv Chapter 1: Introduction ...... 1 1.1. Aims, Hypotheses and Objectives ...... 6 Chapter 2: Literature Review ...... 7 2.1. The Aurantioideae sensu Swingle (Swingle & Reece 1967) ...... 7 2.2. The tribe Citreae (Aurantieae) sensu Swingle (Swingle & Reece 1967) ...... 8 2.3. The Tribe Clauseneae sensu Swingle (Swingle & Reece 1967) ...... 8 2.4. Current circumscription of the genus Murraya König ex L...... 9 2.5. Murraya exotica L. and Murraya paniculata (L.) Jack: one species or two? 11 2.6. Rumphius’s notes and descriptions of Camunium vulgare and Camunium japonense ...... 19 2.7. Descriptions of Murraya paniculata and Murraya exotica ...... 22 2.8. The genus Merrillia Swingle ...... 28 2.9. Huanglongbing (HLB) ...... 30 2.10. Molecular markers in HLB research ...... 35 Chapter 3: Molecular phylogenetics of Murraya and Merrillia ...... 38 3.1. Introduction ...... 38 3.2. Materials and methods ...... 40 3.2.1. Plant materials and DNA extraction ...... 40 3.2.2. DNA amplification ...... 42 3.2.3. DNA sequencing ...... 50 3.2.4. Phylogenetic analysis ...... 51 3.2.5. Incongruence length difference (ILD) test ...... 52 3.2.6. Determination of monophylogeny of Murraya ...... 52 3.3. Results ...... 54 3.3.1. Statistical summary of the sequence data ...... 54 3.3.2. Phylogenetic results ...... 56 3.4. Discussion ...... 92 3.4.1. Relationship between Murraya paniculata and Murraya exotica ...... 97 3.4.2. Relationship between Murraya and Merrillia and the monophyly of Murraya ...... 101 Chapter 4: Morphology: Derivation of Phenograms, Elliptic Fourier Descriptors and Description of Taxa ...... 103 4.1. Introduction ...... 103 4.2. Materials and methods ...... 105 4.2.1. Plant and tissue samples ...... 105

4.2.2. Morphological assessments ...... 105 4.2.3. Statistical analyses ...... 110 4.2.4. Principal component analysis, redundancy analysis and discriminant function analysis for separating taxa on the basis of quantitative characters of leaflets ...... 111 4.2.5. Principal component analysis and redundancy analysis of elliptic Fourier descriptors of leaflet shapes ...... 111 4.2.6. Morphological descriptions of putative taxa ...... 112 4.3. Results ...... 112 4.3.1. Morphological differences between Murraya accessions ...... 112 4.3.2. Dendrograms ...... 123 4.3.3. Principal component analysis and redundancy analysis for separating taxa on the basis of quantitative characters of leaflets ...... 126 4.3.4. Use of elliptic Fourier descriptors for separating taxa on the basis of leaflet shapes ...... 143 4.3.5. Morphological descriptions of putative taxa ...... 151 4.4. Discussion and conclusions ...... 182 Chapter 5. Phytochemistry of Murraya paniculata, Murraya exotica and Merrillia .. 189 5.1. Introduction ...... 189 5.2. Materials and methods ...... 190 5.2.1. Gas chromatography ...... 190 5.2.2. Statistical analyses ...... 191 5.3. Results ...... 192 5.4. Discussion ...... 213 Chapter 6: Susceptibility of taxa to huanglongbing ...... 216 6.1. Introduction ...... 216 6.2. Materials and methods ...... 216 6.2.1. Plant materials ...... 216 6.2.2. DNA extraction ...... 217 6.2.3. HLB detection ...... 217 6.3. Results ...... 218 6.4. Discussion ...... 219 Chapter 7: General Discussion ...... 220 7. 1. Hypothesis 1 ...... 220 7.2. Hypothesis 2 ...... 221 7.3. Placement of other Murraya taxa ...... 222 7.4. Conclusions ...... 223 References ...... 231

vii

Chapter 1: Introduction

My thesis stemmed from research at the University of Western Sydney on a bacterial disease of citrus known as huanglongbing (HLB). This disease is arguably the most serious threat facing citriculture and citrus germplasm (da Graça 1991, Bové 2006, Beattie et al. 2008, Gottwald 2010). When the HLB research commenced, it was assumed that the systematics of the family Rutaceae, including the genus Citrus, was well understood with limited nomenclatural inaccuracies and ambiguity (Beattie et al. 2006, 2008). As the research progressed, it became clear that this was far from the case (Beattie et al. 2006, 2008, Beattie & Barkley 2009). The number of species of Citrus range from 1 to 162 depending on the authority (Malik 1973, Tanaka 1977, Jones 1995). However, recent estimates and new discoveries suggest that there are some 26 species (Mabberley 2004, 2010, Forster & Smith 2010). The confusion is compounded by the use, in some instances, of several binomials for a species and use of specific names for hybrids (Scora 1975, Mabberley 1997). This taxonomic confusion is evidenced by any search of current literature.

When plant pathologists commence studies on a pathogen of a particular crop, one of the features of the pathogen that they need to determine is its host range. In particular, it is important to identify wild and weedy relatives of the crop as they may harbour the pathogen and act as a source of infection. Reviews of the literature have been made to identify species that host the bacteria that cause HLB. These records are incomplete and also suffer from the problems associated with our poor understanding of the taxonomy of the Rutaceae, i.e., the identification of species listed as host may be incorrect. The known host range of the HLB pathogens include species of Citrus, with all species likely to be susceptible (Garnier & Bové 1993), and other Rutaceae including the distantly related Vepris lanceolata1 (Lam.) G. Don [Rutoideae] (Korsten et al. 1996) and more closely related species in the subfamily Aurantioideae, tribe Aurantieae. These species include Atalantia (syn. Severinia) buxifolia (Poir.) Oliv. (Su et al. 1995, Hung et al. 2000, 2001), Limonia acidissima L. (Su et al. 1995, Koizumi et al. 1996, Hung et al. 2000, 2001), and Murraya paniculata (L.) Jack, sometimes cited as Murraya paniculata (L.) Jack var. exotica (L.) CC Huang or Murraya exotica L. (Li & Ke 2002, Lopes 2006, Lopes et al. 2010) and referred to by the common names orange jasmine, orange jessamine and mock orange.

1 For clarity, generic names of taxa mentioned in the thesis are not abbreviated. 1

Murraya paniculata is of particular importance. The widely cultivated ornamental form of it is the favoured host of the Asiatic citrus psyllid (Diaphorina citri Kuwayama [Hemiptera: Sternorrhyncha: Psyllidae]), the most widely distributed of two known vectors of the disease (Halbert & Manjunath 2004, Halbert & Núñez 2004, Weinert et al. 2004, OEPP/EPPO 2005a, b, Villalobos et al. 2005, Conant et al. 2007, Poe 2007, Poe & Shea 2007, Subandiyah et al. 2008). Secondly, there is uncertainty, as implied above, about its taxonomic status: whether the forms that are commonly grown as ornamental plants around the world, including in Australia, are simply selected cultivated forms of Murraya paniculata; another species (Murraya exotica L.); or a variable hybrid, derived from a wild form of Murraya paniculata and another species of Murraya (perhaps Murraya alata Drake). Common ornamental forms are known to be hosts of ‘Candidatus Liberibacter asiaticus’ and ‘Candidatus Liberibacter americanus’, two of three pathogens that cause HLB, but the host status of wild forms, including Murraya paniculata (L.) Jack var. ovatifoliolata Engl. in northern Australia, is not known. Therefore, my study focused on the systematic status of Murraya paniculata (L.) Jack ‘sensu lato’ 2.

Unfortunately, like some other genera in the Rutaceae, the taxonomy of Murraya, as indicated above, is confused. According to the circumscription of Swingle & Reece (1967), the genus Murraya comprises 11 species and 4 varieties within the tribe Clauseneae, one of two tribes in the sub-family Aurantioideae in the family Rutaceae in the order Sapindales. The other tribe is the Aurantieae (Citreae, as circumscribed by Swingle & Reece 1967) to which the genus Citrus belongs (Swinge & Reece 1967, Mabberley 2004). Species of Murraya, as circumscribed by Swingle & Reece (1967), and the closely related monotypic genus Merrillia (Merrillia caloxylon (Ridl.) Swingle), are assumed to be native to the tropics and subtropics of Asia and Australasia, where they grow in climates ranging from hot and humid to cool and temperate (Swingle & Reece 1967). Several species that have been included in the genus Murraya (s.l.) are used in traditional medicines; these species include Murraya paniculata, Murraya koenigii L.3 and Murraya tetramera C.C. Huang4. The roots of Murraya paniculata are used to treat influenza, headache, stomach ache, rheumatism and epidemic encephalitis- B (Kong et al. 1986, Rahman et al. 1997, Mondal et al. 2001). Despite the

2 Sensu lato – ‘s.l.’: ‘in the broadest sense’): inclusive of taxa that other authors would consider as distinct. 3 Murraya koenigii (L.) Spreng. is now considered, once again, to be Bergera koenigii L., as originally named by Linnaeus: see elsewhere in this review. 4 Li et al. (1988) recommended that Murraya tetramera be placed in the genus Bergera on the basis of its phytochemistry and morphology. 2 pharmacological importance of these plants to humans and the role of some species or forms as hosts of D. citri, and as reported hosts (Li & Ke 2002, Lopes 2006, Lopes et al. 2010), possible hosts (Tirtawidjaja 1981, Aubert 1988), or non-hosts (Miyakawa 1980, Koizumi et al. 1996, Dai et al. 2005) of HLB, the systematics and phylogeny of the species is controversial. This has been so since shortly after Linnaeus named one species as Chalcas paniculata in 1767 and another as Murraya exotica in 1771 (Oliver 1861, Tanaka 1929, Swingle & Reece 1967, Stone 1985, Mabberley 1998, Ranade et al. 2006), following Rumphius’ description of the plants as Camunium vulgare and Camunium japonense, respectively, in 1747 (Rumphius 1747). Swingle & Reece (1967) considered Murraya exotica to be a synonym of Murraya paniculata. Huang (1959) considered Murraya exotica to be a variety of Murraya paniculata. Stone (1985) considered Murraya exotica and Murraya paniculata to be distinct species. Huang (1997) in contrast to his earlier view (Huang 1959), also considered Murraya exotica and Murraya paniculata to be distinct species. Nevertheless, Mabberley (1998) suggested that Murraya exotica be best treated as a cultivar of Murraya paniculata.

Some progress towards resolving key issues concerning the systematic status of the genus Murraya ‘sensu stricto’5 (s.s.) has been made through molecular and pharmacological studies (But et al. 1986, 1988, Kong et al. 1986, Samuel et al. 2001). On the basis of phytochemistry, But et al. (1986) recommended separation of the genus Murraya (s.l.) into Murraya section (sect) Murraya and Murraya sect Bergera. This recommendation has been supported by molecular studies (Samuel et al. 2001, Bayer et al. 2009) that have shown that Murraya paniculata is more closely related to Citrus than previously thought. Samuel et al. (2001) suggested that Murraya ‘s.s.’ and Merrillia caloxylon belong with Citrus, within the tribe Aurantieae. These suggestions have not been fully accepted (Zhang et al. 2008), but the weight of evidence (Kong et al. 1986, But et al. 1988, Samuel et al. 2001, Bayer et al. 2009) suggests that placement of 4 species of Murraya, as circumscribed by Swingle & Reece (1967) (Murraya sect Murraya) and Merrillia caloxylon should be placed in the Aurantieae, with the other 7 species (Murraya sect Bergera) returning to, or being placed in, the genus Bergera and remaining in the tribe Clauseneae. My studies attempted to further clarify the systematic status of Merrillia caloxylon (Ridl.) Swingle.

5 Sensu stricto (s.s.: ‘in the stricter sense’): using a taxon restrictively, excluding taxa that other authors include. 3

The terms taxonomy and systematics have various usages. However, taxonomy is usually taken to mean the finding, describing and naming of organisms whilst systematics is a broader term that includes study of relationships between taxa. The kinds of relationships that contemporary biological systematics recognises are phylogenetic and tokogenetic relationships (sensu Hennig 1966). Both kinds of relationships are constructed by the analysis of characters: variable intrinsic features of organisms that allow taxa to be discriminated from each other. Two broad categories of characters can be drawn. The first is a class of phenotypic characters, most notably those of plant morphology, which have not been reduced to genotypic variables. For my study, I used a character set based on the morphology of Murraya species as circumscribed by Stone (1985). This character set was augmented with additional characters discovered from the literature pertinent to the taxonomy for the Rutaceae and with those discovered through my own studies. My studies in this area started with living specimens growing in Australia. This was augmented with herbarium specimens imported into the country. In addition, elliptical Fourier analysis (Kuhl & Giardina 1982) was used to examine overall differences in leaflet shape among the accessions of Murraya collected during my study. Therefore, a major aim of my studies was to discriminate between forms of Murraya paniculata (L.) Jack (s.s.) and Merrillia caloxylon (Ridl.) Swingle, based on their morphology.

One problem with studies of morphology is that plant phenotypes are influenced by the environment. Over the past few decades, we have gained the ability to clone and sequence genes. This has led to the creation of a new field of science called molecular systematics. These studies of plant genotype are not confounded by environmentally induced phenotypic variation. For the study of the molecular systematics of plants, gene sequences that reside in the chloroplast genome and of those that encode nuclear ribosomal RNAs are often informative. A recent study of the Rutaceae (Bayer et al. 2009) has identified a number of such genes that can be used to study the taxonomy of the Rutaceae. Therefore, for my studies I sequenced chloroplast and nuclear ribosomal RNA genes of Murraya and Merrillia. Data from molecular studies were subjected to phylogenetic analysis.

Phytochemistry is the study of chemicals derived from plants and different compounds occurring in different plants. Therefore, when looking at the taxonomy of plants, beside morphology and molecular techniques, phytochemistry is considered to be a potentially

4 useful source of phenotypic characters for separating taxa. Differences in the phytochemistry of Murraya, Merrillia and Bergera have been reported by But et al. (1986), Li et al. (1988) and Kong et al. (1988a, b). My study also looked at the phytochemistry of plants to provide additional evidence for classification of Merrillia and Murraya.

HLB is caused by members of the ‘Candidatus’ genus Liberibacter (α-Proteobacteria) (Jagoueix et al. 1994b, Bové 2006). These are fastidious, phloem-limited bacteria of uncertain origin (Beattie et al. 2008). They were once thought to have evolved in association with plants of the family Rutaceae (Bové 2006, Beattie et al. 2008). However, they have recently been reported to occur naturally in the families Solanaceae, Umbelliferae [Apiaceae] and Leguminosae [Fabaceae] (Liefting et al. 2009a, b, Munyaneza et al. 2010, Fan et al. 2011). Three species have been detected in Rutaceae, mostly commonly in Citrus: Liberibacter, ‘Candidatus Liberibacter asiaticus’, ‘Candidatus Liberibacter africanus’ and ‘Candidatus Liberibacter americanus’. A subspecies, ‘Candidatus Liberibacter africanus subsp. capensis’ has been recorded in the Cape chestnut, Calodendrum capense (L. f.) Thunb. [Rutaceae: Rutoideae] (Garnier et al. 2000, Bové 2006, Teixeira et al. 2005a). Murraya paniculata has been regarded as both a host and non-host of HLB. Most recent records (Li & Ke 2002, Lopes 2006, Lopes et al. 2010) appear to verify early observations by Bernard Aubert (Aubert et al. 1985, Aubert 1988) who recorded Murraya paniculata as a host of HLB, in contrast to views espoused by Garnier & Bové (1993). It is not known if infections by liberibacters can be asymptomatic within Murraya paniculata, but it seems infections in plants that express symptoms can be short-term and transient or long-term and permanent (Aubert 1988, da Graça 1991, Dai et al. 2005, Lopes 2006, Zhou et al. 2007, Lopes et al. 2010). This variability could be related to variability in the pathogenicity of forms (species or strains) of the pathogens and/or variation in the susceptibility of forms of Murraya paniculata.

I determined the presence of ‘Candidatus Liberibacter asiaticus’ in some wild and ornamental forms of Murraya paniculata (s.l.) as part of my studies. My goal was to determine if the pathogen was present in the naturally occurring wild forms that I used in my taxonomic studies. This knowledge may enhance current understanding of interactions between the HLB pathogens and their hosts, and may be useful for developing control strategies and for future work on HLB-resistant/tolerant plants. I

5 used molecular techniques to determine the presence of ‘Candidatus Liberibacter asiaticus’ within wild and cultivated forms of Murraya paniculata (s.l.)

1.1. Aims, Hypotheses and Objectives

The overall aims of my research were to resolve the taxonomic status of Murraya and Merrillia, the status of Murraya paniculata and Murraya exotica, and the susceptibility of the latter, as a single species or otherwise, to ‘Candidatus Liberibacter asiaticus’. The first component of my project was to test the status of species and forms of Murraya and Merrillia in Australasia, the United States of America, Brazil, China, Indonesia, Pakistan, Taiwan and Việt Nam. The results should improve our knowledge of systematics, contribute to conservation of germplasm, and help minimise the impact of HLB. Tests of the taxonomic limits of the species and forms of Murraya and Merrillia were based on phylogenetic and phenetic analyses of the morphology and anatomy of the plants, DNA sequence data and phytochemistry. The second component was to test and document the susceptibility of Murraya and Merrillia to HLB.

I hypothesised that Murraya paniculata and Murraya exotica (sensu Stone (1985) and Huang (1997) are species and that Murraya exotica is not a variety of Murraya paniculata.

I also hypothesised that Merrillia caloxylon may be a species of Murraya (s.s).

My specific objectives were to:  characterise and delimit species and forms of Murraya, focusing on Murraya paniculata and Murraya exotica, and to reconstruct phylogenetic relationships between species, based on chloroplast and ITS sequences (Chapter 3);  characterise and delimit species and forms of Murraya, focusing on Murraya paniculata and Murraya exotica, and to reconstruct phylogenetic relationships between species, based on morphology (Chapter 4);  test whether Murraya paniculata and Murraya exotica are differentiated on the basis of phytochemistry (Chapter 5); and  evaluate the susceptibility of accessions of Murraya and Merrillia caloxylon obtained during my study to HLB (Chapter 6).

Chapter 2: Literature Review6

2.1. The Aurantioideae sensu Swingle7 (Swingle & Reece 1967)

According to Swingle & Reece (1967), the subfamily Aurantioideae of the family Rutaceae contains the genus Citrus and 32 other genera related more or less closely to Citrus, within two tribes (Citreae and Clauseneae). Each tribe comprises three subtribes, as illustrated in Figure 2.1. Most genera of the Aurantioideae (29 of 33) are native to the monsoon region extending from West Pakistan to north-central China and thence south through the East Indian Archipelago to New Guinea, the Bismarck Archipelago, northeastern Australia, New Caledonia, Melanesia and the western Polynesian islands (Swingle & Reece 1967).

Figure 2.1. Swingle’s classification of the Aurantioideae (Swingle & Reece 1967)

All species of the Aurantioideae are trees or shrubs that produce partitioned berry fruits with either a leathery rind or a hard shell. This kind of fruit is called a hesperidium (Swingle & Reece 1967). Many of the genera bear subglobose fruit with a green, yellow or orange peel, dotted with numerous oil glands that often give an agreeable aroma when the fruit is handled. There is no endosperm in the seeds and the seeds sometimes contain 2 or more nucellar embryos. The leaves and bark contain schizolysigenous oil glands. The flowers are usually white and fragrant. Members of the subtribe Balsamocitrinae, within the tribe Citreae, have fruits as large as oranges or grapefruits but with a hard woody shell. These hard shelled fruits do not contain juicy

6 Within this review I have quoted most descriptions verbatim so as to avoid misinterpretation, and to allow the text to be accurately and easily referred to in later discussion. In some instances, footnotes are used where clarification is required. 7 Walter Tennyson Swingle (1871‐1952) was an eminent American botanist. 7 pulp-vesicles, although some of them are pleasantly aromatic. Many of the remote relatives8 of Citrus, in the tribe Clauseneae have extremely small fruits and are usually semidry and entirely inedible (Swingle & Reece 1967).

2.2. The tribe Citreae (Aurantieae) sensu Swingle (Swingle & Reece 1967)

Swingle & Reece (1967) placed Citrus, all of its near relatives, and many more remote relatives, which have not been hybridised with Citrus or capable of being grafted on to it, in the tribe Citreae with the three subtribes and the 28 genera listed in Table 2.1. This tribe comprises 124 species, distributed in Southeast Asia, the monsoon region to the southeast as far as eastern Australia, New Caledonia, New Guinea, and the Melanesian islands, and also in tropical (and subtropical) mountainous regions in eastern, central and western Africa. Swingle & Reece (1967) described the Citreae as:

‘small trees or woody, clambering lianas almost always with single or paired axillary spines at the nodes of the twigs; leaves persistent (except in three monotypic genera, Poncirus, Aegle, and Feronia), usually simple or 1‐3‐foliolate (usually pinnately 3‐foliolate with a short petiole and the middle leaf larger, but in Luvunga palmately 3‐foliolate with a very long petiole) or sometimes odd‐pinnate with strictly opposite pinnae; petioles often winged; flowers in axillary clusters, or in terminal or axillary panicles or corymbs; flowers white or greenish, usually fragrant; calyx with 3‐5 sepals or lobes; corolla imbricate in aestivation, with 3‐5 (rarely 6) petals; stamens 2‐4 or more times as many as the petals; ovary 2‐18‐locular, ovules 1‐18 (usually 1‐12) in each locule (22 or more in Burkillanthus); fruits usually orange‐ or lemon‐like in external appearance, often small, sometimes small red or nearly black juicy berries; in the subtribe Citrinae the locules contain pulp‐vesicles filled with juice but in the other two subtribes (Triphasiinae and Balsamocitrinae) the locules lack pulp‐vesicles, but are usually filled with a mucilaginous gum; seeds with one or several embryos; if with several, only one )(rarely two is a true embryos, the others being false embryos9 that develop from nucellar buds that grow into the embryo sac; germination hypogeous or epigeous.’

2.3. The Tribe Clauseneae sensu Swingle (Swingle & Reece 1967)

According to Swingle & Reece (1967), the tribe Clauseneae comprises 5 genera and 79 species, with the genera, in three natural subtribes (Table 2.2), being the ‘more primitive genera’ of the Aurantioideae. All of the species have no spines in the axils of the leaves

8 Including Murraya as circumscribed by Swingle & Reece (1967). 9 Apomictic. 8 and the odd-pinnate (imparipinnate) leaves are distinguished from those of the tribe Citreae by having the leaflets attached alternately to the rachis, which does not break up into segments when the leaves fall; the rachises are generally not winged (except in Merrillia caloxylon, Murraya alata, Murraya alternans, Clausena guillauminii, Clausena wallichii and Clausena luxurians). Trifoliolate leaves are rarely found exclusively in any species, but they can occur sporadically, often on the same plant, merely by reduction of odd-pinnate leaves to trifoliolate leaves. Such trifoliolate leaves do not clearly show the precise pairing of the lateral leaflets that is always shown in the tribe Citreae. Swingle & Reece (1967) describe the reproductive structures of the Aurantioideae as follows:

‘The ovary has two to five locules with only one or two ovules in each locule, except in Merrillia, which has five (rarely six) locules and eight to ten ovules in each locule. The fruits are usually small semidry or juicy berries, except in Merrillia, which has ovoid fruits the size of a lemon, with a tough leathery peel. The mature ovaries and young fruits of Micromelum have convolute locule walls. This genus has the valvate petals in aestivation10, differing thereby from all the other genera of the subfamily.’

2.4. Current circumscription of the genus Murraya König ex L.

The type species of the genus is Murraya exotica L. (under which Murraya paniculata (L.) Jack is usually synonymised). The type locality is India (Tanaka 1929), but information presented below suggests that the type specimen represents a plant introduced to India from China. Murraya initially comprised Murraya exotica L. before the addition of Chalcas paniculata L. based on Rumphius’ Camunium vulgare. Subsequently, additional species were added, including Bergera koenigii L. (curry leaf). Swingle & Reece (1967) listed 13 species/varieties, some regarded by authors (Tanaka 1929, But et al. 1986) as representing Murraya sect. Murraya (s.s.) and Murraya sect. Bergera. Recent molecular systematic studies (Samuel et al. 2001, Bayer et al. 2009) have supported this separation. Moreover, these studies have supported transfer of Murraya (s.s.) from the Clauseneae to the Aurantieae. Swingle & Reece (1967) regarded Murraya exotica L. as a synonym of Murraya paniculata (L.) Jack.

10 Aestivation: the arrangement of sepals and petals or their lobes relative to one another in an unexpanded flower bud. 9

Table 2.1. The genera of the tribe Citreae (Aurantieae) sensu Swingle (Swingle & Reece 1967).

Subtribe Genera Wenzelia Monanthocitrus Oxanthera Triphasiinae Merope Triphasia Pamburus Luvunga Paramignya Severinia Pleiospermium Burkillanthus Limnocitrus Hesperethusa Citropsis Citrinae Atalantia Fortunella Eremocitrus Poncirus Clymenia Microcitrus Citrus Swinglea Aegle Balsamocitrinae Afraegle Aeglopsis Balsamocitrus Feronia Feroniella

Table 2.2. The genera of the tribe Clauseneae sensu Swingle (Swingle & Reece 1967).

Subtribe Genera Micromelinae Micromelum Glycosmis Clauseninae Clausena Murraya Merrilliinae Merrillia

Of the taxa circumscribed by Swingle & Reece (1967) the genus Murraya within the Aurantieae currently comprises four species and three varieties (major dot points) and synonyms (minor dot points):  Murraya alata Drake (Swingle & Reece 1967, Zhang et al. 2008) o Chalcas alata (Drake) Tan. (Tanaka 1929) o Murraya alata var. hainanensis Swing. (Zhang et al. 2008);  Murraya alternans (Kurz) Swing. (Swingle & Reece (1967)—but it has not been recorded or listed since the 1870s and its existence is doubtful (Swingle & Reece 1967, and recent searches);  Murraya gleniei Thwaites ex Oliv.  Murraya paniculata (L.) Jack

o Chalcas camuneng Burm. f., Chalcas exotica (L.) Millsp., Chalcas paniculata L., Marsana buxifolia Sonner., Murraya exotica L., Murraya banati Elmer, Murraya exotica L., Murraya odorata Blanco, Murraya scandens Hassk., Murraya sumatrana Roxb., and Limonia lucida Forst., according to Swingle & Reece (1967) o Chalcas exotica Millsp., Camnium exoticum O. Kze., Chalcas sumatrana Roem., Murraya banati Elmer, Murraya chinensis Pavon, Murraya elongata S. DC ex Hook. f., Murraya exotica var. sumatrana Koord. & Valet., Murraya heptaphylla Spanog., Murraya sumatrana Roxb., and Murraya tavoyana A. DC. Ex. Wall., according to Tanaka (1929) o Murraya paniculata var. exotica (L.) C.C. Huang according to Huang (1959);  Murraya paniculata var. omphalocarpa (Hay.) Tan. o Murraya omphalocarpa Hay.;  Murraya paniculata var. ovatifoliolata Engl. o Murraya ovatifoliolata (Engl.) Domin.; and  Murraya paniculata var. zollingeri Tan. o Chalcas paniculata var. zollingeri Tan.

The genus Bergera within the Clauseneae, and included by Swingle & Reece (1967) in the genus Murraya, currently encompasses seven species (synonyms not listed) (Swingle & Reece 1967, Huang 1997, Zhang et al. 2009):  Bergera crenulata (Turcz.) comb. nov. ineditus;  Bergera euchrestifolia (Hayata) comb. nov. ineditus;  Bergera koenigii L.;  Bergera kwangsiensis (C.C. Huang) comb. nov ineditus.  Bergera microphylla (Merr. & Chun) comb. nov. ineditus;  Bergera siamensis (Craib) comb. nov. ineditus; and  Bergera tetramera (C.C. Huang) comb. nov. ineditus.

2.5. Murraya exotica L. and Murraya paniculata (L.) Jack: one species or two?

Among plants collected by Georg Eberhard Rumpf (Rumphius) in the 1600s from the Indonesian Archipelago were two plants he named Camunium vulgare and Camunium japonense (= Camunium javanicum) (Rumphius 1747) (Fig. 2.1) before the Linnaean

11 system of nomenclature was standardised in 1753. He collected Camunium vulgare from Makassar (Ujung Pandang) and other locations in southern Sulawesi and several islands in the Maluku Islands, including Ambon and Halmahera. He collected Camunium japonense from Java11. He noted that where he observed Camunium vulgare, it grew ‘everywhere in rocky places which are also dry, and the more stony the place, the better the Camunium [wood] is reckoned’.

Linnaeus renamed Camunium vulgare as Chalcas paniculata in 1767 (Linnaeus 1767), basing his description on Rumphius’ descriptions, and illustrations drawn by Rumphius’s colleagues after Rumphius became blind in 1670. Linnaeus had no specimens of Chalcas paniculata. de Loureiro (1790: pages 270-271) referred to the plants as Chalcas paniculata and Chalcas japonense, noting that they occurred in Cochinchina (southern Việt Nam) and China.

Figure 2.2. Camunium vulgare (left) and Camunium japonense (right) (Rumphius 1747).

11 During a survey of Ambon in 1913 in search of plants described by Rumphius, Charles Budd Robinson (1871‐1813) found cultivated trees ‘exactly Camunium japonense’ (Plantae Rumphianae Amboinensis 249, Herbarium, Bureau of Science, Manila: now the Philippines National Herbarium) in the town of Ambon (Amboina) (Merrill 1917). 12

Linnaeus published the genus Murraya (as Murraea) of König (Johann Gerhard König12) in 1771 (Linnaeus 1771). He based the genus on a bushy form of a plant he named Murraea exotica, the type specimen of which is LINN 539.1 in Herbarium Linnaeus (Murraea exotica: http://www.linnean-online.org/5864/) (Fig. 2.3). This specimen resembles Rumphius’s illustration of Camunium japonense (Fig. 2.2: right). König, a student of Linnaeus, joined a Moravian Mission in Tranquebar (now Tharangambadi: 11° 01' 36'' N, 79° 51' 15'' E, 6 m asl), a coastal town in Tamil Nadu in southern India that was leased by the Danish East India Company, in 1768 and stayed there until 1774 (http://apps.kew.org/floraindica/htm/biography_koenig.htm). I assume he collected the specimen in Tranquebar.

Figure 2.3. Murraea exotica, LINN 539.1 Herbarium Linnaeus (http://www.linnean‐online.org/5864/).

Coincidentally, or otherwise, a plant resembling ‘Murraea’ exotica (LINN 539.1 in Herbarium Linnaeus: Fig. 2.3) and Rumphius’s Camunium japonense was introduced by Mr Benjamin Tobin from India to the Royal Botanic Gardens at Kew in Britain in 1771 (Aiton 1811, Edwards & Lindley 1819-1820). An illustration (Edwards & Lindley 1819-1820) of foliage and flowers of the plant is reproduced in Fig. 2.4 (left). It proved

12 Johann Gerhard König (Koenig) (1728‐1785) was a missionary‐surgeon and botanist and a friend of William Roxburgh with whom he travelled and worked (http://apps.kew.org/floraindica/htm/biography_koenig.htm). 13 to be a most ‘desirable evergreen for either the conservatory or the greenhouse’ (Edwards & Lindley 1819-1820). In India, it was known to have been introduced many years beforehand from China to the coast of Coromandel (the eastern coast of Peninsular India) where it was universally cultivated in gardens (Edwards & Lindley 1819-1820; Roxburgh 1832). It was also known to occur in a wild state in the mountains of the Northern Circars (the coastal region of Andhra Pradesh on the northeast coast of Peninsular India) (Edwards & Lindley 1819-1820).

Figure 2.4. Left, Murraya exotica (Edwards’ Botanical Register 1819); right, Murraya exotica (Roxburgh drawing No. 48, Royal Botanic Gardens, Kew13).

A similar plant referred to as Murraya sumatrana Roxb. was introduced from Sumatra to the Botanic Gardens at Calcutta (Edwards & Lindley 1819-1820, Roxburgh 1832) not long before publication of the 1819-1820 edition of Edwards’ Botanical Register (Edwards & Lindley 1819-1820). According to Edwards & Lindley (1819-1820), this plant differed ‘from exotica in being much less bushy, with larger leaves, fewer and bigger flowers, and a very distinct habit’. It was described by William Roxburgh14

13 http://apps.kew.org/floraindica/img/illustration/large/26754.jpg. William Roxburgh went to India in 1776 in the service of the East India Company, being stationed first in Madras, and then, from 1793 onwards, at Calcutta, where he was Superintendent of the Botanic Garden, until ill‐health forced him leave India in 1813 to return to England via the Cape Town of Good Hope and St. Helena. Soon after his arrival in India he began making descriptions of about 2,600 native plants and had paintings made of more than 2,500 of them (Sealy 1956). This illustration (No. 48) was presumably commissioned shortly after he arrived in India. 14 William Roxburgh (1751‐1815) was a Scottish surgeon and botanist. 14

(Roxburgh 1832) who compared it with Murraya exotica (as illustrated in Fig. 2.4, right), noting Murraya sumatrana as having 5–7 ovate and ovate-oblong leaflets, terminal flowers, being a large shrub, native to Sumatra, much thinner of branches, with larger leaves and fewer but much larger flowers than Murraya exotica and, when growing together, conspicuously different in habit. William Jack (1820) proposed that Chalcas paniculata be transferred to Murraya and recognised it as being distinct from Murraya exotica15, 16. Thus began decades of debate and confusion about the status of both species.

Oliver (1861) noted that the ‘extreme states’ of the plants looked diverse but were connected by transitional links. He considered Murraya paniculata to be a form of Murraya exotica, noting that the former occurred in the south and northwest India (‘Peninsular Ind. or., Ind. bor.-occid!’), Sri Lanka and southern Borneo, and the latter in Kumaon (Uttarakhand in northern India), Assam (northeast India), Hong Kong, China, the Japanese Ryukyu Islands (‘Loo-choo’), Java and Vanuatu.

Hooker (1875) included Murraya paniculata as one of three forms of Murraya exotica present in British India. He noted that it occurred throughout the hotter parts of India, from Garwhal (now Uttarakhand) to Assam and Birma (Burma), and southward to Chittagong (south-eastern Bangladesh), Travancore (the Western Ghats region of southern India comprising most of Kerala and southern-most Tamil Nadu) and Ceylon (Sri Lanka), and outside of the subcontinent in China, Australia and the Pacific Islands. Within the subcontinent he noted that the exotica form was present in Sri Lanka within gardens and in northern India, and that the paniculata form occurred in Sri Lanka ascending to 914 m asl and in the ‘Peninsular’. The third form was Murraya gleniei Thwaites ex Oliv. from Sri Lanka. Hooker (1875) also noted that Murraya elongata A. DC. from Burma at ‘Taong-dong’ (located on a mountain range bordering the Mytinge River some 20 km southeast of Mandalay) was a very different looking plant from any of the forms of Murraya exotica.

15 According to Merrill (1952), Jack’s description was apparently based on material from Penang or Singapore. Merrill (1952) also commented that the plant was common, variable, and a widely distributed Malaysian species. 16 William Hunter (1755‐1812), surgeon and naturalist, visited Penang (then Prince of Wales Island) in 1802 (Forman 1989) where he recorded Murraya exotica growing in the garden of Lieutenant Colonel Thomas Polhill (Hunter 1909), the commanding officer (Hodson 1910). The tree was the only one on the island and had not yet ripened seeds (Hunter 1909). The plant was common in the neighbourhood of neighbouring Kedah, on the Malay Peninsula (Hunter 1909). 15

Tanaka (1929) returned Murraya to Chalcas. He commented that ‘There is considerable variation in the size of leaves and flowers, but the separation of larger type as C. paniculata, and smaller type as C. exotica, is meaningless.’

Tanaka (1929) described ‘Chalcas paniculata var. zollingeri’ from the ‘Malay islands’ (Sumbawa, the type locality, and Timor, both part of the Indonesian archipelago) on the basis that it appeared quite different from the type of Chalcas paniculata. He also renamed Murraya omphalocarpa Hayata from Taiwan as ‘Chalcas paniculata var. omphalocarpa Tanaka’.

Merrill17 (1935) listed Chalcas paniculata L., Chalcas japonensis Lour., Chalcas camuneng Burm. f. and Murraea exotica L. as synonyms of Murraya paniculata (L.) Jack. He (Merrill 1935) commented that Chalcas paniculata L. was correctly interpreted by de Loureiro (1790) as a synonym of Camunium vulgare Rumph., and that Chalcas japonensis Lour. is merely a cultivated form of Murraya paniculata Jack. with small leaves and well represented by Camunium japonicum Rumph. In 1943, Swingle (1943) reunited Chalcas with Murraya whilst Stone (1985) considered Murraya paniculata and Murraya exotica to be distinct species.

Huang (1959) considered Murraya exotica as a variety (Murraya paniculata (L.) Jack var. exotica C.C. Huang) of Murraya paniculata, but he later reinstated Murraya exotica as a species distinct from Murraya paniculata (Huang 1997) noting that:  leaflets of Murraya exotica are broader on the upper half; the tip obtuse, sometimes acute or nearly round; mature fruits rather equal in length and width, the tip acute or obtuse, and that it occurs naturally in China on sandy loams nearby the coast, with plenty of sunshine, and tolerant of dry heat whereas, within the same region,  leaflets of Murraya paniculata are often broader on the lower half, with acuminate tips, and mature fruits are longer than broad, mostly narrowly oblong, the tip conical, and that it occurs naturally in China in Taiwan, Fujian, Guangdong, Hainan and southern parts of Hunan, Guangxi, Guizhou, and Yunnan provinces sometimes as a dominant species in small patches on low rolling hills and in sparse or dense woods of higher altitudes, including woodlands in inland

17 Elmer Drew Merrill (1876-1956) was an eminent American botanist. 16

mountains, more common in limestone areas including humid, shaded sites and seasonally dry hills, and also in granite areas18.

Kong et al. (1986) noted that Murraya paniculata occurs in inland, hilly limestone areas of southern China whereas Murraya exotica is found primarily in maritime sites on acid, red soils. Li et al. (1988) found that major volatile oil components of Murraya exotica and Murraya paniculata leaves differed, with the major oil component of Murraya exotica and Murraya paniculata being t-caryophyllene (50%) and γ-elemene (31.7%), respectively19.

Jones (1995) separated Murraya paniculata var. paniculata and Murraya exotica growing in Sabah and Sarawak in Borneo, East Malaysia, on the basis of the following characters:

‘Leaflets obovate, less than 4 cm long, apex somewhat obtuse. Inflorescences terminal. Fruits elliptic to sub globose. Cultivated...... M. exotica L. Leaflets ovate or elliptic to rhomboid, 3–7 cm long, apex acuminate. Inflorescences axillary. Fruits ovoid with acuminoid apex. Wild species ...... M. paniculata.’

He (Jones 1995) noted that Murraya exotica is commonly cultivated throughout the tropics, that it is ‘Native to China and Taiwan (?)’, and that its natural geographical range (coastal areas of Hong Kong and China) is more restricted than that of Murraya paniculata. He commented that in Sabah and Sarawak Murraya paniculata was ‘occasionally found in lowland and hill forests, usually on rocky soils or limestone, from near sea-level to 600 m’ whereas Murraya exotica ‘was only found in cultivation’.

Mabberley (1998) suggested that Murraya exotica be best treated as a cultivar of Murraya paniculata. He noted that Australian workers have applied the name Murraya exotica L. var. ovatifoliolata Engl. to the wild plant (i.e., typical Murraya paniculata) in Australia, distinguishing it from the cultivated forms in its straggling habit, rather hairy shoots and broadly oval or ovate leaflets. Mabberley (1998) also noted that:  two other varieties were distinguished by Swingle (1943), one of them, var. zollingeri (Tanaka) Tanaka (1937), described from Timor, having small leaflets with deflexed margins;

18 Summary based on text kindly translated by Paul Pui Hay But. 19 Voucher specimens related to these studies are located in the Zhongshan (Sun Yat‐sen) University Herbarium, Guangzhou, Guangdong, China: http://biomuseum.sysu.edu.cn/ASP/search/plant/species.asp?familyid=RUTACEAE&id=Murraya 17

 in Australia, with very few intermediate exceptions, plants from dry semi- deciduous to deciduous vine thickets tend to be low sprawling shrubs less than 3 m tall with such leaflets, highly aromatic when crushed (the 'Small Leaves' plant of Brophy et al. (1994) ? = var. zollingeri); and  plants from 'less dry' semi-deciduous notophyll forests form small trees with larger less aromatic leaflets ('Big Leaves' of Brophy et al. 1994) and correspond to M. paniculata s.s. as represented elsewhere in Indomalesia.

Mabberley (1998) concluded that whether the two leaf-forms were genetically or merely phenotypically distinct variants remains to be demonstrated and their relationship to plants in neighbouring territories was not yet fully elucidated.

Ranade et al. (2006) applied the following polymerase chain reaction (PCR)-based methods, analysis of the internal transcribed spacers (ITS), randomly amplified polymorphic DNA (RAPD), and directed amplification of minisatellite DNA (DAMD), to the classification of Murraya species as circumscribed by Swingle & Reece (1967). They concluded that Murraya koenigii and Murraya paniculata are distinct species, but considered Murraya exotica as not distinctly different to Murraya paniculata and, therefore, a synonym of Murraya paniculata. In contrast, by using ISSR, DAMD, and RAPD, he (Ranade 2006) and his colleagues, three years later, concluded that Murraya paniculata and Murraya exotica were different taxa (Verma et al. 2009)20 and they also explained that the previous work (Ranade et al. 2006) had been incapable of detecting intraspecific variability in the Murraya paniculata complex due to the limited number of genotypes examined.

Zhang & Hartley (in Zhang et al. 2008) treated Murraya paniculata (千里香 qian li xiang) and Murraya exotica (九里香 jiu li xiang) as separate species on the basis of differences in morphology. They noted that:  Murraya paniculata occurred in thickets and montane forests, from near sea level to 1300 m in Fujian, Guangdong, Guangxi, southern Guizhou, Hainan, southern Hunan, Taiwan and Yunnan in China: elsewhere in Bhutan, Cambodia, India, Indonesia, Japan (Ryukyu Islands), Laos, Malaysia, Myanmar, Nepal, New Guinea, Pakistan, Philippines, Sri Lanka, Thailand, Việt Nam; Australia and south-west Pacific islands;

20 Voucher specimens for this studied were deposited in the National Botanical Research Institute, Lucknow, India (Verma et al. 2009). 18

 Murraya exotica occurred in thickets; near sea level in southern Fujian, Guangdong, Guangxi, southern Guizhou, Hainan and Taiwan in China: elsewhere, widespread in cultivation in tropical and subtropical areas.

Mou (2009) and Mou & Zhang (2009), who regarded Murraya exotica and Murraya paniculata as separate species, observed differences between palynological characteristics, with exine ornamentation of Murraya exotica being cross-striate with some micro-perforations and that of Murraya paniculata being cross-striate without micro-perforations. Mou (2009) also found morphological and anatomical differences between flowers. Additionally, she separated the taxa into two clades on the basis of molecular differences in the nuclear ITS region and two chloroplast spacer sequences (trnL-F, atpB-rbcL). However, both the number and source of the accessions studied was limited. Therefore, in this thesis, the nuclear ITS region and six chloroplast regions of 86 accessions sourced from Australia, Brazil, China, Indonesia, Pakistan, Taiwan, United States of America and Việt Nam were used to determine the phylogenetic relationships between them.

2.6. Rumphius’s notes and descriptions of Camunium vulgare and Camunium japonense

Rumphius (1747) described Camunium vulgare and Camunium javanicum (Fig. 2.1) as follows21:

‘Camunium seems to me to be the Indian equivalent of Buxus, certainly in respect of its wood, admittedly less so in respect of its shape, for it has great similarity with the Caju Lacca [=??] above, but it does not creep in the same way; rather it lifts itself up; it is divided into two species: the first the common or Ambonese, the second the Javanese; however they are not very different.

Firstly, the common or Ambonense Camunium [C. vulgare of the plate] (plate 17) is the closer to Caju Lacca, since it creeps forward in a certain manner: first it lifts itself up as high as a bush [or small tree], then it bends and creeps forward like a rope, supporting itself on other trees; its trunk is usually a foot thick, never round and solid, but for its whole length it is furrowed and pitted, so that it appears somewhat decayed and rarely solid, and their configuration makes them look as if several stems had coalesced; its furrows are much deeper than in the trunks of Caju Lacca, and moreover it is covered with a thin, pale yellow, fissured bark.

21 Kindly translated from Latin by David Mabberley, with his notations in square brackets. 19

The branches are round, with a thin and tough bark, which can easily be stripped off, as in Salix trees; from them large numbers of branches grow, straight and for the most part vertical, and divided into very many lateral twigs [i.e. leaves], which for the most part grow in pairs mfro a single point of origin. On these are situated the leaves [i.e. leaflets], arranged as in Caju Lacca, that is to say alternating, five, seven and eight on one twig [i.e. rachis]. They are firmer and smoother than those of Caju Lacca, ending in a long point. The end one, which is always single and very large, is bent [?geniculate] near its point of origin, and it is four transversal fingers long, and two wide.

The lower leaves [leaflets] are indeed hardly a knuckle’s width long, and a finger’s wide; the upper side is bright green and smooth, and the underside rather yellower, and soft to the touch, like the denser silk, particularly in the older leaves; the central [i.e. mid]vein on the underside forms a sharp ridge, and sub‐veins [i.e. laterals] are few, though on the underside more prominent than on the leaves of Caju Lacca; they have a pungent and unpleasant taste, which stings the tongue much less than the leaves of Limonum [i.e. Citrus x limon]. They turn white as they age and drop.

The flowers bloom late and are few; one or two from one point of origin; their five or six petals have the same shape as those of Jasmine, but are fewer and firmer, having a short and wide bell‐shaped tube, in whose centre appear ten stamens with in the middle a pistil with a yellow end; the smell of these is pleasant, though faint, similar to that of Jasmine, especially towards evening when the weather is clear and calm. The petals of the flowers easily fall, nor can they be handled much, though the central pistil remains firm, and grows into the fruit. This is a longish berry in every way resembling that of the lesser Capsicum, as much in the shape as the colour; it is one fingernail long and inside are located two longish little bones, joined together and to some extent downy; the surrounding flesh is reddish and soft, with a reddish skin, like that of the fruit of Cynosbatum; the odour is strong. The aforementioned bones are the seeds, which germinate readily when committed to the earth. It flowers in November and December, and the fruit ripens in the month of March.

Its wood is free from knots, and under the thin bark is found a homogeneous material, fine‐grained, hard and heavy, in colour paler than that of Buxus wood, which in its interior material and heaviness it resembles, especially around the base of mature trunks, where it is yellower; around the heart it is a dark honey colour, and splashes or spots of this colour appear here and there in the wood; that wood which has the most of such flecks is reckoned to be the best. Some trunks have none of this variegation at all, but have homogeneous and evenly yellow coloured wood, like ivory, certainly when worked, whereas others have few spots around the heart and near the root.

The wood is elegant to the eye, but large articles cannot be made from it, because it is so curved and when cut into square sections, it shows such a small quantity of solid wood, and especially because it splits so easily, if it is exposed to the sun for a very short time. That which has been recently worked gives off a strong unpleasant smell, like that of the wood of Cofassinus or Buxus.

Secondly, Camunium Javanicum [= C. japonense on the plate] (plate 18) is an elegant small tree, as high as Granatus [i.e. pomegranate]; it does not raise itself high above the earth on a single trunk, but soon divides into several thick and sinuous branches, which bear very many short branches, which form a dense mass like hair.

Its leaves [i.e. leaflets] are of the same shape and are arranged on the branches [i.e. rachides], five, six and seven in one twig [i.e. rachis], as the previous species; they are however shorter, firmer and smoother, and not downy underneath; the end leaf is the length of the little finger, the remainder below it are shorter, and among them some shorter than a knuckle; the taste is sharper than in the leaves of the previous species, more like that of the leaves of Limonum.

The flowers and fruits are as in the previous species, but the flowers of this species are more plentiful, and they give off a more pleasant odour, so that on warm evenings it fills the whole neighbourhood in which they are located; throughout the whole summer season flowers and fruits are produced in succession, now flowers, now fruits; however it is necessary to keep the chickens away from them, since they devour the flowers and fruits very eagerly, constantly flying up into these small trees.

Its wood is of the same colour and consistency as the previous one, but those dark flecks are not seen in this species, with the result that this species is especially sought after because of its shape and elegance, as is the previous one if he [the buyer] does not mind, but indeed desires, the markings. This Javanese wood is more solid and plentiful, since its trunks are not so irregular and furrowed as those of the Ambonese species, which spreads more, and solid sections of it as thick as a leg can be procured.

Although throughout the whole dry season of the year some flowers can be seen on this small tree, I have however observed it weighed down with flowers for three consecutive days around the middle of December; they bloomed on the 16th December, the driest time having ended on the 9th day of the same month; after those three days the flowers dropped off, and a large crop of fruits developed. I also observed all the grains and seeds shed below the tree, although in a dry and sandy place, germinated much better than in other places, which I think is duee to th slight shade provided by the dense canopy of foliage.’

2.7. Descriptions of Murraya paniculata and Murraya exotica

Linnaeus described Chalcas paniculata as a shrub, leaves alternate, petiolate, a little dull, scarcely crenate, flowers somewhat paniculate and terminal.

He (Linnaeus 1771) described Murraya exotica (Murraea exotica) as a tree, covered in white bark, appearing like a pepper tree [Schinus], leaves alternate, petiolate, somewhat rigid, imparipinnate, leaflets 7, petiolate, alternate, obovate, glabrous, slightly obtuse22.

William Jack (Jack 1820) said of Murraya paniculata L.:

‘This is an abundantly distinct species from M. exotica, though unaccountably confounded with it by later authors. Loureiro discriminates between them very well, and his description in on the whole good. Rumphius’s figure is bad, but preserves several of the distinguishing characters, particularly in the inflorescence and leaves, which however are not sufficiently acuminate. It grows to the size of a small tree, and the wood is much employed for the handles of cresses being capable of receiving a fine polish. The leaflets are generally five, ovate, terminating in a long acumen which is slightly emarginate at the point, shining and very entire, the terminal one considerably the largest. In M. exotica, the leaflets are numerous and closer, obovate, blunt, and of much firmer thicker substance. The flowers of M. paniculata are fewer and larger than those of M. exotica, and are sometimes terminal, generally one or two together from eth axils of the upper leaves. The ovarium is two celled; the berries are oblong, reddish, and mostly contain two seeds which are covered with silky hairs. The berries of M. exotica are ovate and generally one seeded. The whole habit of the two plants is very distinct. The specific name paniculata is objectionable, as the flowers are much less panicled than in the other species.’

As mentioned above, Oliver (1861) considered Murraya paniculata to be a form of

Murraya exotica. He noted that ‘The leaflets vary much in form and size, being lanceolate, ovate, or obovate‐lanceolate, and often acuminate.’ He also noted that paniculata flowers ‘are sometimes nearly solitary’ whereas in exotica ‘they often form close cymose corymbs’. He considered imperfect specimens labelled Murraya elongata A. De Candolle from ‘Toong Doong’ (Taong Dong) in Burma in the Wallichian Herbarium to be allied to the paniculata variety but added: ‘The leaflets are shining, acuminate, and 2 to 4 inches in length. The inflorescence would seem to have consisted of few‐flowered cymes. It is perhaps but a large‐leaved variety of M. exotica.’

22 Kindly translated by Peter Lister, University of Western Sydney. 22

Tanaka (1929), who regarded Murraya exotica as a synonym of Murraya paniculata (Chalcas paniculata) (as outlined above) gave the following description of Murraya paniculata:

‘Evergreen shrub generally with dense foliage. Branches are moderately thick, slender. Leaves unequally pinnate; leaflets alternate, 3, 5 or 7, obtusely acuminate at the apex, acuminate or tapering at the base, usually drying green, semi‐coriaceous, shining above, minutely and indistinctly punctate, veins few, rather indistinct. Inflorescens terminal, paniculate; flower bud about 1 × 3 cm, slender, broaden toward the upper part and acute‐ pointed, smooth or pubescent at apex; calyx comparatively small, well lobed, lobes elongate‐triangular, with more or less blunt tipped, densely pubescent or nearly naked, much glandular with convex oil cell dots; petals erect‐patent, not fully expand, linear oblanceolate, striated, dots obscure. Stamens free, unequal in length, about two third of the petal in length; filament linear‐subulate, quite slender, glabrous; anther small, broad, bi‐locular, dorsifix. Pistil nearly as long as the stamen; ovary fusoid, tapering and continuous to the style; style terete, long, cylindric, smooth or puberulent; stigma distinct, oblate, disk annular, collar‐like. Fruit a berry with a few seeds, oval and tapering toward the apex, base also somewhat pointed, red, minutely punctate. Both leaflets and flowers may be very large sized. Berry may also be sharp pointed or quite globose.’

He (Tanaka 1929) also described Chalcas paniculata var. zollingeri:

‘leaflets 3‐6, attached quite apart, small, about 3 × 1.5 cm, oval, thin, chartaceous, pellucid‐dotted, puberulent on both sides, ends obtuse, surface somewhat buckled, margin considerably reflexed; rachis very thin and declined, minutely soft tomentose. Inflorescence axillary, few‐flowered, often pendulous after flower; peduncle leafy, about 1.8 cm long, pedicel slender, continuous to calyx, about 1–1.5 cm long, tomentose with very short soft hairs. Calyx small, well‐lobed, lobes elongated‐triangular, copiously tomentose, not glandular. Ovary after flower oval in shape, tomentose, seated on a short disk. Immature fruit oboval, short, smooth, apex rounded, densely dotted with large pellucid glands, puberulent. Flowers not seen: much declined rachis, small thin leaves with reflexed margin, and non apiculate fruit, seem to be distinct from the type.’

Tanaka (1929) described Chalcas paniculata var. omphalocarpa Tanaka as:

‘branchlets yellow, terete, brownish. Leaves about 5‐foliolate, rachis 5–8 cm long, terete, striated. Leaflets broad‐oval, terminal one about 7 × 4.5 cm, lateral ones smaller, measuring about 5 × 4 cm, the largest being 8 × 4.7 cm; midrib prominent on lower side, side veins rather broad angled to the midrib, about 5 paired, texture thick, margin entire

or slightly crenulate, glabrous. Flowers solitary or fascicled, terminal or axillary, at least near the end of the branchlets; pedicels 1.3–1.5 cm long; calyx long‐lobed, body campanulate, lobes oval to linear oblong, about 3 mm long, obtuse and pubescent at the apex; petals oblong‐oboval, almost oblanceolate, 13 × 4 mm, attenuate at the base, obtuse at the apex. Stamens 10, filaments alternately shorter, longer ones about 9 mm, shorter ones about 6 mm. Pistil as long as the filaments; style fine, filiform; stigma capitate, about 1 mm diam. Berry ovoid, apiculate, about 21 × 12 mm, very much rostrate when immature, about 15 × 5 mm.’

Swingle & Reece (1967), who considered Murraya exotica a junior synonym of Murraya paniculata, described Murraya paniculata as:

‘An evergreen tree, 15–25 ft. [4.6–7.6 m] high, trunk 6–8 ft. [1.8‐2.4 m] high, 1½ –2 ft. [46–61 cm] diam., the young shoots puberulous; leaves unpaired‐pinnate or occasionally pinnately 3‐foliolate, glossy, glabrous, or sometimes the rachis puberulous; leaflets alternate, cuneate‐obovate or almost obliquely rhomboid, shortly petioluled, blunt or bluntish acuminate, 1–1½ in. [2.5–3.8 cm] long, coriaceous; flowers rather large, white, in dense but small, almost sessile terminal corymbs; petals about ½–¾ in. [13–18 mm] long, recurved; stamens 10, alternately shorter; ovary 2‐celled, the style long with a capitate glandular stigma; berries ovoid‐oblong, bluntish acuminate, nearly ½ in. [13 mm] long, orange‐coloured, 1–2‐seeded; seed villous.’

‘The orange jessamine is a handsome greenhouse ornamental that blooms profusely. It has large white fragrant flowers that are succeeded by small red fruits. In warm subtropical climates it thrives out of doors. It grows vigorously and can be propagated easily from cuttings.’ ‘There are many varieties and strains of M. paniculata in the Old World; possibly some of them would support Citrus better than the strain now grown in the United States, probably originally introduced from India.’

Swingle & Reece (1967) recognised three varieties: Murraya paniculata var. ovatifoliolata Engl., Murraya paniculata var. zollingeri Tan., and Murraya paniculata var. omphalocarpa (Hay.) Tan. For Murraya paniculata var. ovatifoliolata, they noted that Engler distinguished this variety from Murraya paniculata var. paniculata as having broadly oval or ovate leaflets and that Bailey23 described it as follows:

‘This, our indigenous form, is of a more straggling habit with more numerous and larger oil‐dots, and is often decidedly hirsute and tomentose, thus very distinct from the two

23 Frederic M Bailey (1927‐1915) was the Queensland Colonial Botanist from 1881 until his death; the description cited by Swingle & Reece (1967) pertains to Australia. 24

Indian ones of our gardens. The leaves are 3–9‐foliolate; the twigs, calyx, petals, and ovary hirsute.’

Murraya paniculata var. zollingeri Tan. was described by Swingle & Reece (1967) as:

‘Leaflets 3–6, small, chartaceous, margins much reflexed; rachis thin, declinous, slightly

pubescent; fruits not apiculate, often pendulous, sparsely puberulous’

They (Swingle & Reece 1967) diagnosed Murraya paniculata var. omphalocarpa (Hay.) Tan. from Murraya paniculata var. paniculata as follows:

‘Differs from the species in the following characters: fruits large with attenuate tips, about 21 × 12 mm (when immature, rostrate, about 15 × 5 mm); flowers larger; petals narrowed at the base; calyx lobes elongate, ovate to linear‐oblong, 3 mm long; leaflets broad, 5‐8 × 4‐4.7 cm.’

Stone (1985) regarded Murraya exotica and Murraya paniculata as distinct species. He described the genus as:

‘Trees or shrubs, unarmed; bark often smooth, sometimes pale. Leaves alternate, imparipinnate (very rarely―in one species―unifoliolate), with alternate leaflets. Inflorescences paniculate, axillary or terminal. Flowers often medium or rather large (over 1 cm long), 5‐merous; sepals ovate or lanceolate; petals linear to ovate‐lanceolate or oblanceolate, imbricate; stamens, 10 subequal or alternating longer and shorter, filaments sometimes compressed; anthers small, ovate or elliptic; disc annular to short‐cylindric; ovary 2 5‐celled, ovoid, often on a distinct gynophore; ovules 2, rarely 1, per cell: style elongate, slender, caducous; stigma capitate. Fruit ovoid to globose, baccate, pericarp thin, glandular, red or black, pulp mucilaginous; seeds I or few, the testa thin, glabrous or tomentellous. Cotyledons green, plano‐convex.’

Stone (1985) gave the following description of Murraya exotica:

‘An erect shrub; dense, later several‐stemmed, to c. 3 m tall; innovations with a very minute, short rather ephemeral puberulence of simple curved hairs 0.05–0.1 mm long, on the stems, petioles, leaf axes, petiolules and midribs also but sparser on the inflorescence axes near nodes, and margins of dbracts an sepals. Bark medium grey, shallowly fissured. Leafy branchlets 1–2 mm diam. Leaves to 9 cm long, leaflets mostly 3–7 (below inflorescences sometimes 1), alternate or the lowest pair opposite, petioles 4–12 mm long, petiolules 1–3 mm long; blades mostly obovate to subelliptic, obtuse to bluntly

acuminate with the apex minutely notched, base cuneate; mostly 1–3.5 cm long, 0.9–1.8 cm wide, subentire or the distal margins obscurely crenate; upper surface darker, glossy, lower surface medium semi‐glossy, both glandular punctate (pellucid by transmitted light). Inflorescences terminal, paniculate, 5–6 cm long, 3–6‐flowered, the axes slender, pedicels 5–8 mm long. Calyx‐lobes 1.5–1.9 mm long, deltoid, glandular. Petals 5 or 6, white, oblong‐subspathulate, to 21 mm long, about 6 mm wide, bluntly acute, midveins pale greenish dorsally, glabrous, entire, at anthesis gracefully recurved. Stamens usually 10, longer ones c. 11 mm long alternating with shorter ones c. 8 mm long; filaments thickish subulate, white, glabrous; anthers rounded, scarcely 1 mm long; gynophore torulose, c. 0.5 mm high; ovary 2 mm long, slightly 2–(3–)‐angled, pale green glabrous; style cylindric, 6.5–7 mm long; stigma flattened capitate, pale yellow, obscurely 2–3‐lobed, about 2–2.4 mm wide. Ovary 2–3‐celled. Fruit ellipsoid‐subglobose, not or but slightly acuminoid, ripening red; epicarp glandular, this and mesocarp fleshy, to 1.8 mm thick. Seeds 1–3, ellipsoid‐obovoid, c. 9 × 5 mm, seed coat greyish‐olive‐brown; cotyledons pale green; hypocotyl and epicotyl scarcely 1 mm long.’

Stone (1985) described Murraya paniculata as:

‘Shrub or small tree, glabrous or finely puberulent on innovations, sometimes also on the ovary. Bark pale to white, thin. Leaves alternate, imparipinnate, with, usually 3–5 (rarely 7) leaflets, rarely unifoliolate, to 10–17 cm long: leaflets petiolulate, the petiolules mostly 2–6 mm long: leaflets mostly 3–7 cm long, ovate or ovate‐elliptic, acuminate, at base cuneate to rounded, thinly coriaceous, densely glandular, glossy and darker above, the midrib slightly depressed above, slightly raised beneath, the main lateral nerves about 5–8 pairs, evident on both surfaces as are the reticulations; margins entire or obscurely crenate. lnflorescences axillary, paniculate, few‐flowered, the peduncle 1–2 (–4) cm long, the pedicels mostly 5–10 mm long; calyx usually 5–lobed (rarely 4–), sepals narrowly deltoid, c. 1 mm long, minutely puberulent, glandular with several raised, round glands; petals 5 (rarely 4), elliptic to ovate‐elliptic or slightly obovate, glabrous, to 15–21 mm long or more, 4‐6 mm wide, with a broad midvein and pinnate slender ascending nerves, white; stamens 10, alternately longer, the longer ones to 12 mm long, glabrous, shorter ones to 9 mm long, filaments dilated slightly downward, to 1 mm broad; anthers short ‐oblong, obtuse‐emarginate; disc glabrous, c. 1 mm high, no wider than ovary; ovary ovoid‐ ellipsoid, 2.5–3 mm long, glandular (slightly tuberculate); style columnar, 8–10 mm long, including the enlarged, capitate, bilobed stigma. Fruit an ovoid 1–2–seeded berry, apex acuminoid (especially when dry), pericarp glandular; seed about 1 cm long, the seedcoat densely hairy, hairs simple, c. 2 mm long; cotyledons thick, fleshy, c. 8 mm long; plumule glabrous.’

Stone (1985) restricted Murraya paniculata to the wild forms with ovate-acuminate, comparatively large leaflets and ovoid-acuminoid fruits, with larger flowers with a tendency to elongate petals. He considered it as distinct from Murraya exotica, which has usually been placed in synonymy with this species. He noted that in Ceylon the latter appears only in cultivation.

Jones (1995) described Murraya paniculata (var. paniculata) as:

‘Shrub or small to medium‐sized tree to 20 m tall, 25 cm diameter. Bark thin, pale to whitish. Young shoots, twigs, sepals, petals, and ovary glabrous to slightly hairy. Leaves pinnate, to 17 cm long; leaflets 3–7, rarely unifoliolate, ovate or ovate‐elliptic to rhomboid, 3–7 x 2–3.5 cm, chartaceous or thinly leathery, glossy and darker above, glabrous; base cuneate to rounded, margin entire or faintly crenulate, apex acuminate; lateral veins 5–8 pairs; petiolules 2–6 mm long. Inflorescences axillary panicles or cymes, few‐flowered; peduncle 1–4 cm long; pedicels 2–9 mm long. Flowers large; sepals (4–)5, narrowly deltoid, c. 1 mm long, glandular; petals (4–)5, elliptic to oblong‐obovate, to 15– 21 mm long or more, white; stamens alternately long and short, the longer to 12 mm long, filaments dilated below; ovary 2‐carpellate, glandular, style columnar, nearly 1 cm long, stigma enlarged, bilobed. Fruit an ovoid berry, c. 12 mm long, apex acuminoid; peel shiny red, gland‐dotted, glabrous. Seeds densely hairy; cotyledons thick, fleshy.’

Huang (1997) described Murraya paniculata as:

‘Small trees, up to 12 m tall. Trunks and branches greyish white or yellowish grey, slightly shiny; young twigs (of the year) green, cross‐section somewhat triquitrous, the lower surface nearly arc‐shaped. Seedling leaves unifoliate, subsequent ones unifoliate or bifoliate, mature leaves bearing 3–5 leaflets, rarely with 7 leaflets; leaflets dark green, shiny on upper surface, ovate or ovate‐lanceolate, 3–9 cm long, 1.5–4 cm wide, the tip narrowly acuminate, rarely mucronate, the base mucronate, the two sides symmetrical or slanted on one side, margins entire, undulating, lateral veins 4–8 on each side; petiolules less than 1 cm long. Inflorescences axillary and terminal, normally about 10 flowers, rarely up to 50 flowers; sepals ovate, up to 2 mm, sparsely hairy along margins, persistent; petals oblanceolate or long elliptic, up to 2 cm, slightly reflexed when in full bloom, scattered with pale yellow translucent oil glands; stamens 10, alternating short and long, filaments white, linear, slightly shorter than the stigma, rarely gland‐dotted on connective body and apex; style green, slender, together with the ovary up to 12 mm long, the stigma very large, wider or as wide as the ovary, the ovary bilocular. Fruits orange yellow to vermilion, narrowly long ellipsoid, rarely ovoid, tapering at the top, 1–2 cm long and 5–14 mm wide, oil glands raised when dry but depressed in the centre. Seeds 1–2; seed coat villous.’

He (Huang 1997) described Murraya exotica as:

‘Small trees, up to 8 m tall. Branches greyish white or yellowish grey, young branches green. Leaflets 3–5(–7), obovate or obovate elliptic, asymmetrical, 1–6 cm long, 0.5–3 cm wide, the apex rounded or obtuse, sometimes retuse, the base mucronate, slightly oblique on one side, the margins entire, flat, the petiolules very short. Inflorescences usually terminal or sometimes terminal and axillary, many flowers clustered as panicoid cymes; flowers white, fragrant; sepals ovate, about 1.5 mm long; petals 5, long elliptic, 10–15 mm long, reflexed in full bloom; stamens 10, unequal, slightly shorter than the petals, filaments white, anthers with 2 oil dots on the back; style slightly more slender than ovary, with no obvious demarcation from the ovary, both light green, the stigmas yellow, thick. Fruit orange to vermillion, broadly ovoid or ellipsoid, somewhat pointed at apex, slightly slanted, sometimes spherical, 8–12 mm long, 6–10 mm across, meat viscous, seeds with short villous hairs.’24

Zhang & Hartley (Zhang et al. 2008) also regarded Murraya exotica and Murraya paniculata as distinct species and described them as:

‘Murraya paniculata. Shrubs or trees, 1.8–12 m tall. Older branchlets grayish white to pale yellowish gray. Leaves 2–5‐foliolate; petiolules less than 1 cm; leaflet blades mostly suborbicular to ovate to elliptic, 2–9 × 1.5–6 cm, margin entire or crenulate, apex rounded to acuminate. Inflorescences terminal or terminal and axillary. Flowers 5‐merous, fragrant. Sepals ovate to lanceolate, to 2 mm, persistent in fruit. Petals white, narrowly elliptic to oblanceolate, to 2 cm. Stamens 10. Fruit orange to vermilion, narrowly ellipsoid or rarely ovoid, 1–2 × 0.5–1.4 cm. Seeds villous.

Murraya exotica. Trees to 8 m tall. Older branchlets grayish whitee to pal yellowish gray. Leaves 3–7‐foliolate; petiolules rather short; leaflet blades elliptic‐obovate or obovate, 1– 6 × 0.5–3 cm, margin entire, apex rounded or obtuse. Inflorescences terminal or terminal and axillary. Flowers 5‐merous, fragrant. Sepals ovate, ca. 1.5 mm. Petals white, oblong, 1–1.5 cm. Stamens 10. Fruit orange to vermilion, broadly ovoid, 8–12 × 6–10 mm. Seeds villous.’

2.8. The genus Merrillia Swingle

According to Swingle & Reece (1967), Merrillia belongs to subtribe Merrilliinae, tribe Clauseneae, with only one species, Merrillia caloxylon (Ridley) Swingle (Swingle 1918)25, that occurs in the Malay Peninsula (including southern Thailand), Sumatra and

24 Kindly translated by Paul Pui Hay But. 25 Named in honour of Elmer Drew Merrill (Swingle 1918). 28

Borneo (Swingle & Reece 1967, Stone & Jones 1988). It was originally named Murraya caloxylon by Ridley (1908) who described it as:

‘A tree of considerable size, the branches covered with a pale flaky bark. Leaves 8 inches [20.3 cm] or more long, with 13 leaflets; rachis flattened and winged, narrow; leaflets 3– 3½ inches [7.6–8.8 cm] long or less by 1½ inches wide [3.8 cm], alternate, oblanceolate, obtusely acuminate with a triangular base, minutely petiolate, inequilateral, thin, bright deep green. Flowers pale yellowish‐green, several together in small panicles, in the upper

1 axils of a branch, about an inch [2.5 cm] long. Sepals connate, ovate‐acute, /10 inch [2.5 mm] long. Petals and stamens not seen. Ovary stalked, hairy, style rather stout, hairy, stigma capitulate. Fruit oblong, rounded at both ends, 4 inches [10.2 cm] long and three inches [7.6 cm] in diameter, the pericarp dotted and warty, greenish, eventually becoming yellow, half an inch [12.7 mm] thick, lemon yellow inside, full of long resin cells narrowed at the mouth and dilated below, cells 5, with rather thick tough walls, pulp of transparent flattened sticky fibers, olive green in colour and tasteless. Seeds numerous, about 5 in a section, ovate‐flattened, half an inch [12.7 mm] long, ⅛ inch [3 mm] thick, olive grey.’

Swingle & Reece (1967) proposed that the subtribe Clauseneae derived as an offshoot from the ancestors of Murraya and evolved rapidly in flower and fruit characters while retaining the general facies of a Murraya of the Murraya paniculata group of species. In order to test Swingle’s hypothesis that the Merrillinae might have arisen from the same stock as Murraya section Murraya, But et al. (1988) studied the chemotaxonomic relationship between Murraya and Merrillia. They found that root and stem bark of Merrillia caloxylon contained the anti-implantation indole alkaloid, yuehchukene, and the 8-prenylated coumarins, sibiricin and phebalosin, as well as 3-(3-methyl-buta-1,3- diene) indole and eupatorin. But et al. (1988), therefore, confirmed a close relationship between Merrilliinae and Murraya section Murraya. Plants of both taxa contain yuehchukene and the 8-prenylated coumarins, but no carbazole alkaloid. Root and stem bark of Merrillia caloxylon, like those of plants of Murraya section Murraya, are straw- coloured to pale whitish, its leaves also bear wings along the rachis as in Murraya alata, and the seeds are also villous. However, Merrillia caloxylon has long trumpet-shaped flowers (55–60 mm long), much larger than those of other Rutaceous plants. Kong et al. (1988a) also concluded that Merrillia caloxylon has a chemical profile that is closely comparable to that of Murraya section Murraya. Therefore, Kong et al.’s (1988a) work supports Swingle’s hypothesis and brings into focus that polyoxygenated flavonoids could be another taxonomically useful chemical marker in this group. Samuel et al.

(2001) considered that Merrillia should be transferred to the Aurantieae (Citreae). Their study was based on non-coding plastid DNA sequences and phytochemical features. The further aim of this study is to test the proposal by Samuel et al. (2001) to re-unite Merrillia with Murraya (s.s.).

2.9. Huanglongbing (HLB)

HLB is the most serious disease of Citrus (Bové 2006). The International Organisation of Citrus Virologists adopted huanglongbing as the officia1 name of the disease in 1995 (Moreno et a1. 1996, van Vuuren 1996, Ohtsu et al. 2002) and the name means ‘yellow shoot disease’. It is or has been known as greening, yellow branch disease,26 and blotchy mottle in South Africa (Schwarz et al. 1973, Bové 2006), citrus decline and citrus dieback in India (Schwarz et al. 1973, Bové 2006), citrus vein-phloem degeneration in Indonesia (Schwarz et al. 1973, Supriyanto & Whittle 1991, Bové 2006), leaf mottling and leaf-mottle yellows in the Philippines (Schwarz et a1. 1973, Bové 2006), and likubin in Taiwan (Schwarz et al. 1973, Su & Huang 1990).

HLB is caused by three putative species of fastidious, phloem-limited, Gram-negative α-Proteobacteria in the genus ‘Candidatus27 Liberibacter’: a heat-tolerant form, ‘Candidatus Liberibacter asiaticus’, and two heat-sensitive forms, ‘Candidatus Liberibacter africanus’ and ‘Candidatus Liberibacter americanus’ (Bové 2006). The pathogens have also been recorded in some citrus relatives, in some instances causing HLB-like symptoms (Bové 2006, Beattie & Barkley 2009, Gottwald 2010). A subspecies, ‘Candidatus Liberibacter africanus subsp. capensis’ is known to infect Cape chestnut, Calodendrum capense (Garnier et al. 2000). The pathogens are transmitted by two psyllids, Diaphorina citri Kuwayama [Hemiptera: Sternorrhyncha: Psyllidae] and Trioza erytreae (del Guercio) [Hemiptera: Sternorrhyncha: Triozidae].

‘Candidatus Liberibacter asiaticus’ causes the most severe symptoms (Fig. 2.5) of the disease and the greatest devastation (Bové 2006): it can destroy orchards within 5 years of planting and 100% infection of initially pathogen-free trees can occur within 2 years of planting (Yang et al. 2006). It is also the most widespread of the three species. It occurs in South and Southeast Asia (from the Indian subcontinent to the Philippines, the Indonesia Archipelago and Japan), New Guinea, Arabia, Mauritius, Réunion, Ethiopia,

26 In honour of Prof Lin Kung Hsiang (Bové 2006). 27 Murray & Schleifer (1994) proposed the Candidatus designation as an interim taxonomic status for proper allocation of sequence‐based potentially new taxa at the genus and species level. 30 the United States of America (California, Florida, Georgia, Louisiana, South Carolina and Texas), Mexico, the Caribbean (Belize, Cuba, Dominican Republic, Jamaica and Puerto Rica), the Leeward Islands (US Virgin Islands), Central America (Costa Rica, Honduras, Nicaragua) and South America (Brazil) (Teixeira et al. 2005a, b, Bové 2006, Martínez et al. 2008, Beattie & Barkley 2009, Matos et al. 2009; and more recent internet sources). Diaphorina citri and ‘Candidatus Liberibacter asiaticus’ have spread rapidly after they were first reported in the United States of America in Florida in 1998 (Halbert & Manjunath 2004) and 2005 (Gottwald 2010), respectively, and since the pathogen was first reported in Brazil in 2004 (Teixeira et al. 2005a, Lopes et al. 2010). The only known host plants of the pathogen were species of Rutaceae until it was recently detected in Chinese apea ear-ring (Archidendron lucidum (Benth.) I. C. Nielsen [Fabales: Leguminosae]) (Fan et al. 2011).

Figure 2.5. Primary and secondary symptoms of huanglongbing in sweet orange (Citrus × aurantium L.) in Florida, United States of America (left: GAC Beattie, 24 August 2007), and severe dieback caused by the disease Siem mandarin (Citrus reticulata Blanco) trees in Java, Indonesia (right: GAC Beattie, 1 October 2003).

‘Candidatus Liberibacter africanus’ occurs in sub-Saharan Africa, Arabia, Mauritius and Réunion were it has been recorded in citrus (Bové 2006, Pietersen et al. 2010), false horsewood (Clausena anisata (Willd.) Benth.) (van den Berg et al. 1991-1992, van den Berg et al. 1992) and white ironwood (Vepris lanceolata) (Korsten et al. 1996, da Graça & Korsten 2004). ‘Candidatus Liberibacter americanus’ occurs in Brazil where it has been recorded in citrus and Murraya exotica, sensu Huang (1997) (Lopes et al. 2010). Additionally, ‘Candidatus Liberibacter africanus subsp. capensis' occurs in southern Africa where it has only been recorded in Cape chestnut (Pietersen et al. 2010, Phahahladira 2010).

As stated in Chapter 1, it is important for plant pathologists to determine the host range of a particular pathogen including the wild and weedy relatives of the crop as they may act as a source of the disease. Information presented by Garnier & Bové (1993), Jagoueix et al. (1994b), Halbert & Manjunath (2004) and Beattie & Barkley (2009) suggests that all species of Citrus species are hosts of HLB. Reports on the susceptibility of Murraya paniculata to the pathogens have been variable. All reports appear to be for the common ornamental form: Murraya exotica, sensu Huang (1997) and Zhang et al. (2008). Miyakawa (1980), Aubert (1990), Guo & Deng (1998), Hung et al. (2000), Dai et al. (2005) concluded that it was not a host plant of ‘Candidatus Liberibacter asiaticus’. In contrast, Li & Ke (2002), Deng et al. (2007a) and Zhou et al. (2007) reported it as a host of ‘Candidatus Liberibacter asiaticus’ and Lopes et al. (2010) reported it as host of both ‘Candidatus Liberibacter asiaticus’ and ‘Candidatus Liberibacter americanus’ in urban areas in Brazil. Symptoms of the disease in orange jasmine leaflets are evident in Fig. 2.6.

Figure 2.6. Symptoms of huanglongbing, in this instance yellowing and mottling, in leaflets of orange jasmine (accession 62), South China Agricultural University, Guangzhou, Guangdong, China. Powdery mildew caused by Oidium sp. is also evident in the photograph. (GAC Beattie, 31 March 2009).

There are no reports of Merrillia caloxylon being susceptible to HLB; however, it is a host of Diaphorina citri (Lim et al. 1989, 1990).

Symptoms of HLB vary among species and varieties of Citrus. However, the most common leaf symptoms are chlorotic patterns similar to those of Zn deficiency (yellowing between the veins and the retention of a narrow band of green along the veins). The leaves are reduced in size, upright and thickened. The symptoms sometimes appear as dwarfing, dulling and rolling of the foliage suggestive of B deficiency (Schwarz et al. 1973). Leaf symptoms are followed by a progressive decline of the upper foliage involving dieback of growing points initially, and then of branches. The size and yield of fruit of infected trees is reduced dramatically and the fruit are greened and worthless (McClean & Oberholzer 1965, Schwarz et al. 1973); often the fruit are lopsided. Ripening occurs at the stalk end first, and the central columella of the fruit is curved. Inside, fruit remain a dirty greenish brown with a bitter, unpleasant taste that makes them quite unsalable. Initially, ‘greening’ was thought to be a disease of the fruit, as the name was used to refer to certain abnormal fruit, often confined to a few branches, that failed to mature properly: the quality was so poor as to make them unmarketable (McClean & Oberholzer 1965). However, McClean & Oberholzer (1965) noted that ‘greening’ was not just a disease of the fruit but a disease of the tree and severely affected trees are stunted and unthrifty. Trees produce leaves with chlorotic leaf patterns. The most distinctive of these is a mottle caused by an irregular yellow discolouration following the course of the midrib and larger lateral veins (McClean & Oberholzer 1965). The disease can be diagnosed by its typical symptoms, graft transmission to indicator Citrus plants, electron microscopy, enzyme-linked immuno- sorbent assay (ELISA) and PCR (Ohtsu et al. 2002). However, the latter technique is the most commonly used method.

Symptoms resembling those of HLB were first recorded in Pakistan (then part of India) by Husain & Nath (1927) in their description of injury caused by Diaphorina citri. They attributed severe symptoms of infestation by the psyllid to a toxin in its saliva. In China, the disease was thought to be due to malnutrition, water injury or infection by species of Fusarium [Hypocereales: Nectriaceae] (Lin 1956, da Graça 1991, Bové 2006) until experiments undertaken by Lin (1956) suggested that a ‘virus’ was the primary causal agent. Oberholzer et al. (1965) described the symptoms in South Africa and concluded that the disease was not related to soil conditions, or mineral deficiency or toxicity. McClean & Oberholzer (1965) and Oberholzer et al. (1965) concluded from graft transmission studies and observations on insect transmission, respectively, that the

33 disease was caused by a virus. Molecular evidence subsequently proved that the disease is caused by Gram-negative bacteria (Jagoueix et al. 1994b).

Diaphorina citri is the most widely distributed of the two known vectors of HLB. It occurs throughout Southern and Southeast Asia from Iran to the southern islands of Japan, the Philippines, in American Samoa and Guam, throughout the Indonesian Archipelago to New Guinea in northern Australasia, Arabia, the Mascarenes (Réunion and Mauritius), the United States of America (Alabama, Arizona, California, Florida, Georgia, Hawaii, Louisiana, Mississippi, South Carolina and Texas), Mexico, the Caribbean (Bahamas, Belize, Cayman Islands, Cuba, Dominican Republic, Jamaica, Puerto Rico and St. Thomas), Leeward Islands (Guadeloupe, Virgin Islands), Central America (Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua), South America (Argentina, Bolivia, Brazil, Columbia, Paraguay, Uruguay and Venezuela) (Halbert & Manjunath 2004, Halbert & Núñez 2004, Weinert et al. 2004, OEPP/EPPO 2005a, Conant et al. 2007, Poe & Shea 2007, Beattie & Barkley 2009, Darderes 2009: and numerous internet sources). Trioza erytreae, the less widely distributed of two known vectors of HLB (McClean & Oberholzer 1965, Bové 2006, Beattie et a1. 2008), is native to, and occurs widely in, sub-Saharan Africa (Hollis 1984). It also occurs in Arabia, the Indian Ocean islands of Madagascar, Mauritius, and Réunion, and the Atlantic Ocean islands of Saint Helena, Madeira, Porto Santo, Tenerife and Gomera (OEPP/EPPO 2005b, Bové 2006). Bacterial cells can cross the insect gut membranes and multiply into the haemolymph, from which they reach the salivary glands (Jagoueix et al. 1994a, b). ‘Candidatus Liberibacter africanus’ and Trioza erytreae are both heat- sensitive, whereas ‘Candidatus Liberibacter asiaticus’ and Diaphorina citri are both heat-tolerant (Bové 2006). Both psyllids remain infective in a persistent manner (Jagoueix et al. 1994a, Hung et al. 2004).

Figure 2.7. Adult Diaphorina citri (left) and ovipositing female: both on orange jasmine, South China Agricultural University, Guangzhou, Guangdong, China (GAC Beattie, 28 June 2007). 34

Although Diaphorina citri was first described from specimens collected on Citrus sp. in Taiwan (Kuwayama 1908), the first record of it (as Euphalerus citri) as a serious pest of citrus was in India (Lal 1920a, b). In describing damage caused by the psyllid Husain & Nath (1927) were the first to describe what are now recognised as severe symptoms of HLB. In the early to mid 1930s, the psyllid did not assume such a destructive status in China, the Philippines, Malaya or Indonesia (Clausen 1933, Hoffmann 1936). Hoffmann and Clausen were both aware of the destruction wrought by the psyllid in India (Hoffmann 1936) as described by Husain & Nath (1927).

Diaphorina is a large homogenous genus typically associated with the warm arid regions of Africa and Asia (Hodkinson 1986), and the Diaphorininae probably evolved with the Sapindales in Gondwana (Hollis 1985, 1987, White & Hodkinson 1985) with an ecological preference for dry climates (Hollis 1987). Hollis (1987) and Halbert & Manjunath (2004) suggested that Diaphorina citri evolved in India in association with a species of Murraya where Bergera koenigii (cited as Murraya koenigii) was the first Citrus relative to be recorded as a host (Fletcher 1917, 1919, Husain & Nath 1927). Both primary parasitoids of the Diaphorina citri, the ectoparasitoid Tamarixia radiata (Waterston) [Hymenoptera: Eulophidae] and the endoparasitoid Diaphorencyrtus aligarhensis (Shafee, Alam & Agarwal) [Hymenoptera: Encyrtidae] were first described from India by Waterston (1922) and Shafee et al. (1975), respectively, and other records suggest that both parasitoids were introduced intentionally or unintentionally to Southeast Asia following spread of their host eastward (Beattie et al. 2008).

Murraya paniculata (sensu Swingle & Reece (1967)) was not reported as a host of Diaphorina citri in India until 1975 (Cheema & Kapur 1975), 60, 44, 40 and 10 years, respectively, after it was reported as a host in Taiwan by Maki (1915) and Kuwayama (1931), in China by He & Zhou (1935), and in the Ryukyu Islands of Japan by Miyatake (1965). It is regarded as a favoured host of the psyllid, but all records appear to be for the common ornamental form, Murraya exotica (sensu Huang (1997), Stone (1985) and Zhang et al. (2008)) aside from one report in Indonesia (Subandiyah et al. 2008) for Murraya paniculata (L.) Jack (sensu Jack (1820)).

2.10. Molecular markers in HLB research

Cloning of DNA fragments from a periwinkle (Catharanthus roseus (L.) G. Don [Gentianales: Apocynaceae]) plant infected with an Asian strain of ‘Candidatus

Liberibacter’ allowed the production of a probe that could detect Asian forms of disease (Villechanoux et al. 1992). Villechanoux et al. (1993) sequenced the fragment used as a probe and showed that it encoded part of the rplKAJL-rpoBC operon. When used at lower stringencies, the fragment would also detect African strains of the pathogen (Villchanoux et al. 1993). Planet et al. (1995) designed primers for PCR amplification of the equivalent genes from an African strain. The amplification product was able to detect African strains by Southern and dot hybridisations. By sequencing comparison of the operon, Planet et al. (1995) suggested that there were two bacteria species causing the disease, an African form and an Asiatic form. Jagoueix et al. (1996) produced PCR primers for detecting the bacteria: the primer pair OI1/OI2c amplify both the Asiatic and African forms, while the pair OI2c/OA1 only amplify the African form. They suggested that in countries where the two forms are known or suspected to be present, the use of the three primers OI2c/OI1/OAl in the same PCR mixture should be applied. Hocquellet et al. (1999) developed a new PCR detection method based on the amplification of ribosomal protein genes rplA and rplJ. These primers, named A2 and J5, allow specific amplification of a 669 bp fragment from the African form and a 703 bp fragment from the Asiatic form. These primers were used to detect the Asiatic form in Citrus and Diaphorina citri in Pakistan (Chohan et al. 2007).

In South Africa, ‘Candidatus Liberibacter africanus subsp. capensis’ was recognised in Cape chestnut by sequencing of its 16S rDNA, intergenic 16S/23S rDNA and ribosomal protein genes of the ß operon (Garnier et al. 2000). More recently, ‘Candidatus Liberibacter americanus’ was reported in Brazil (Teixeira et al. 2005a). It was detected with new primers (GBl/GB3) after the 16SrDNA primers OI1+OAl/OI2C used to detect the African and Asiatic forms of the disease failed to detect bacteria in symptomatic leaf samples. It has also been detected in Diaphorina citri, which was first recorded in South America in Brazil in about 1940 (Teixeira et al. 2005b). Recently, quantitative real-time (RTi)-PCR, was used for diagnosis of HLB (Li et al. 2006, Wang et al. 2006). The RTi-PCR assay is rapid and has the greatest sensitivity. The resu1ts also indicated that among various HLB symptoms (leaf or vein yellowing, Zn deficiency, small leaves, symptomless, and mottling) used for PCR diagnosis, mottling of leaves had the highest positive rate (96.5%) which indicates that this symptom is the most reliable symptom for field surveys. According to Li et al. (2006), the RTi-PCR assays do not cross-react with other pathogens or endophytes commonly resident in citrus plants. This method is very sensitive and specific for the detection, identification and quantification of

‘Candidatus Liberibacter spp.’. Nested PCR is also sensitive and has also been used for detecting HLB pathogens (Li & Ke 2002, Ding & Wang 2004, Ding et al. 2005, Deng et al. 2007a, b, Faghihi et al. 2009, Fan et al. 2011). This latter technique has been used in this study to determine whether the accessions used in the morphological, chemical and phylogenetic studies were infected with liberibacters and the host status of the Murraya accessions is discussed in Chapter 6.

Chapter 3: Molecular phylogenetics of Murraya and Merrillia

3.1. Introduction

According to the circumscription of Swingle & Reece (1967), the genus Murraya belongs to the tribe Clauseneae (comprising 11 species and 4 varieties) in the subfamily Aurantioideae whilst the genus Merrillia (with a single species, Merrillia caloxylon) belongs to the tribe Merrilliinae, in the same subfamily. However, the taxonomy of Murraya and Merrillia is controversial. Tanaka (1929), But et al. (1986) and Li et al. (1988) suggested that Murraya should be divided into two sections, sect Bergera and sect Murraya. According to Tanaka (1929) and But et al. (1988), section Murraya is closely related to Merrillia, and the latter was considered by Swingle & Reece (1967) ‘to be a very anomalous member of the tribe, Clauseneae, that has probably evolved from a Murraya‐like ancestral form’. Based on non-coding, plastid DNA sequences, Samuel et al. (2001) suggested transferring section Murraya and Merrillia to the tribe Aurantieae (Citreae). Despite many studies of the taxonomy of Merrillia and Murraya, the systematic status of Merrillia and the relationship between Merrillia and Murraya is still unclear.

There is also considerable controversy regarding a common, ornamental form of Murraya called orange jasmine. Orange jasmine is cultivated widely in many countries where Citrus species and hybrids are also grown. Reports show that orange jasmine is a favoured host of the Asiatic citrus psyllid (Kuwayama 1931, Tsai et al. 1984, Koizumi et al. 1996), the vector of ‘Candidatus Liberibacter asiaticus’ the most important of three known ‘Candidatus Liberibacter spp.’ that cause huanglongbing (HLB). Orange jasmine is also a host of HLB (Tirtawidjaja 1981, Aubert et al. 1985, Li & Ke 2002, Lopes et al. 2005, 2006a, b, Lopes 2006, Deng et al. 2007a, Zhou et al. 2007, Manjunath et al. 2008). However, there is uncertainty about the taxonomic status of orange jasmine, with some authors (Jack 1820, de Candolle (cited in Tanaka 1929), Stone 1985) suggesting that two species, Murraya paniculata and Murraya exotica, exist, whilst Tanaka (1929) considered that these taxa are synonymous. Most of reports on the status of Murraya paniculata and Murraya exotica are based on plant morphology (see Chapter 4). Therefore, this section of my research focused on the taxonomy of Murraya and Merrillia, based on molecular phylogenetics.

Phylogenetic systematics is the biological science of studying the relationships and evolution of organisms. It aims to describe taxon diversity and to construct the hierarchy, or phylogenetic relationships, among taxa. Phylogenetic relationships among genes and organisms can be derived from different sources of data. Traditional methods used for phylogenetics are based on morphological characters; however, with the development of biotechnology, molecular information from nucleotide or amino acid sequences has been used for phylogenetics. These molecular data are now used widely to infer phylogenetic relationships (Vandamme 2009).

Molecular phylogenetics reconstructs the historical evolution of genes and organisms based on patterns, and, in the case of likelihood methods, on rates of change in the DNA. When comparing two aligned sequences, each site along the alignment can be categorised as either a match, in which case both sequences have the same nucleotide at that site, or a mismatch, in which the nucleotide in one sequence aligns with a different nucleotide or a gap in the other sequence. These differences can then be treated mathematically to infer phylogenies. Among closely related species, genes commonly differ by a limited number of point mutations, whereas genes of more distantly related species differ by a greater number of substitutions, although some genes, or parts of genes, are more highly conserved than others (Vandamme 2009).

There are different methods for inferring phylogeny and constructing phylogenetic trees from molecular data (Cracraft & Helm-Bychowski 1991, Lemey et al. 2009). Based on the data used, phylogenetic tree construction can be divided into two groups: those using a distance matrix of pairwise dissimilarities and those using discrete character states. Distance methods calculate some measure of the dissimilarity for each pair of sequences from which a pairwise distance matrix is produced; the data in the matrix are then used to infer phylogenetic relationships among the taxa (Vandamme 2009). In the character state methods, each sequence position in the aligned sequences is deemed to be a character and the nucleotides at that position are the states. Usually, all character positions are analysed independently, so that each alignment column is assumed to be an independent realisation of the evolutionary process. Character state methods use morphological or physiological characters or sequence data, and the methods used for their analysis retain the original character status of the taxa. Therefore, they can be used to reconstruct the character state of ancestral nodes (Lemey et al. 2009). In this study, the character state methods, maximum parsimony (MP) and Bayesian inference (BI),

39 were used to examine phylogenetic relationships between accessions of Murraya and Merrillia.

In order to understand plant systematics, the sequencing of chloroplast DNA (cpDNA) and nuclear DNA has been used. cpDNA has been used extensively to infer plant phylogeny at different levels (Palmer et al. 1988, Taberlet et al. 1991, Gielly & Taberlet 1994) and is valuable for studying phylogenetic relationships among closely related species (Palmer 1987, Palmer et al. 1988). The chloroplast genome consists of noncoding regions (introns and intergenic spacers) and protein coding genes (Shaw et al. 2007). The non-coding regions are considered to evolve more rapidly than protein coding genes and are used more widely in phylogenetic studies than are coding regions. With the hope for clarification of the taxonomy of Murraya and Merrillia, in particular the status of Murraya paniculata and Murraya exotica, a molecular phylogenetic approach was applied in this study through the sequencing of several regions and spacers of the maternally-inherited chloroplast genome and part of the internal transcribed spacer (ITS) region of the nuclear-encoded ribosomal RNA operon.

3.2. Materials and methods

3.2.1. Plant materials and DNA extraction

Mature leaflets from plants of the genera, Murraya, Merrillia and Bergera, were collected from the wild or from gardens, parks and other domesticated locations in Australia, Brazil, China, Indonesia, Pakistan, Taiwan, United States of America and Việt Nam. This resulted in a total of 85 accessions of Murraya, Merrillia and Bergera that were used for molecular phylogenetic analysis (Table 3.1).  DNA of samples collected from Australia (accessions 2—ANSW, 4—ANSW, 6—ANSW, 8—ANSW, 9—ANSW, 10—ANSW, 13—AQLD, 14—AQLD, 53— ANSW, 54—AQLD, 69—ANT, 70—ANT, 71—ANT, 72—AQLD, 73—AQLD, 74—AQLD, 75—AQLD, 108—ANSW, 115—AQLD), Florida, United States of America (accessions: 111—UFBG and 112—UFBG) and of samples 63—CGD, 68—CGD from China was extracted at the University of Western Sydney, New South Wales, Australia.  DNA of samples collected from Indonesia (accessions 22—IWJ, 23—IWJ, 24— IP, 25—IWJ, 27—IWJ, 28—IWJ, 30—IL, 34—IEJ, 35—IEJ, 37—IEJ, 38—IEJ, 40—IC, 42—IUCR, 44—ICJ, 45—ICJ, 46—ICJ, 47—ICJ, 48—ICJ, 51—ICJ,

113—INTT, 114—INTT) was extracted at Department of Plant Protection, Faculty of Agriculture, Gadjah Mada University, Yogyakarta, Indonesia.  DNA of samples collected from Việt Nam (accessions 57—VTG, 58—VTG, 59—VTG, 60—VTG, 61—VCP, 77—VTG, 78—VTG, 79—VTG, 80—VTG, 81—VHCM, 82—VHCM, 83—VHCM, 84—VDL, 85—VDL, 86—VTH, 87— VHN, 88—VCP, 89—VBG) was extracted at the Plant Protection Research Institute, Hanoi, and Southern Fruit Research Institute, Tien Giang.  DNA of samples from Brazil (accessions 102—BSP, 103—BSP, 104—BSP, 105—BSP, 106—BSP, 107—BSP) was extracted at Fundecitrus, São Paulo, Brazil.  DNA of samples from California, United States of America (accessions 64— UUCR, 65—UUCR and 67—UUCR), was extracted at the University of California, Riverside.  DNA of samples from Taiwan (accessions 91—T, 92—T, 93—T) was extracted at the Plant Protection Department, Chiayi Agricultural Experiment Branch, Agricultural Research Institute, Taiwan.  DNA of some samples collected from China (accessions 76—CGX, 94—CYD, 95—CYD, 96—CYD, 97—CYD, 98—CGX, 99—CH, 100—CH, 101—CGD) was extracted at South China Agricultural University, Guangzhou, Guangdong, China.

An error in labelling samples also led to a sample of DNA of an unknown Citrus species or hybrid (accession 90—VTG from Tien Giang, Việt Nam) being included in the analyses.

Total DNA from samples collected from Australia, China (accessions: 63—CGD and 68—CGD), Indonesia, Việt Nam, and Florida, United States of America (accessions: 111—UFBG and 112—UFBG) was extracted from leaf material following the modified methods of Doyle & Doyle (1990) and Warude et al. (2003). In the laboratory, leaflets were washed using running tap water to remove dirt, then 0.2 g of leaflets of each sample was ground in a mortar with 1.5 mL CTAB extraction buffer (2% CTAB, 2% 2- mercaptoethanol, 100 mM EDTA, 50 mM Tris, 1.4 M NaCl, 1% PVP, pH = 8) to make a fine paste. The paste was transferred to a 1.5 mL Eppendorf tube and incubated at 65°C for 30 min with occasional vigorous shaking; the sample was then centrifuged at 12,000 rpm for 10 min. The supernatant was transferred to a new tube and an equal

41 volume of phenol was added and mixed well for 1–3 min. The tube was centrifuged for 10 min at 12,000 rpm and the upper phase pipetted to a new Eppendorf tube. An equal volume of CIAA (chloroform plus isoamyl alcohol, 24:1) was added, mixed well and centrifuged at 12,000 rpm for 10 min. This step was repeated again, the upper layer transferred to a new tube and 2 volumes of cold, absolute ethanol and 1/10 volume of 3M sodium acetate was added and incubated at -20oC overnight or at -80°C for 1 h to precipitate the DNA. The sample was centrifuged for 15 min, at 12,000 rpm to collect the pellet and the pellet was washed twice by 70% ethanol and dissolved in 50 μL TE buffer or sterilised milliQ water.

The DNA of accessions from Brazil and Taiwan was extracted using the method of Murray & Thomson (1980) by our colleagues in those countries. The DNA of samples collected from the University of California, Riverside, and from China accessions 76— CGX, 94—CYD, 95—CYD, 96—CYD, 97—CYD, 98—CGX, 99—CH, 100—CH, 101—CGD) were extracted using the DNeasy Plant Minikit (Qiagen) and the HP Plant DNA Kit (Omega Bio-Tek) respectively following the manufacturers’ instructions with the assistance of our colleagues and posted to Australia.

3.2.2. DNA amplification

Six different regions and spacers of the maternally-inherited chloroplast genome and part of the nuclear-encoded ITS region (Table 3.2) were amplified from DNA extracts using the polymerase chain reaction (PCR). PCR was performed using: Taq DNA polymerase (5 U/L) (New England Biolabs); 10 × Thermopol buffer (New England

Biolabs, [MgSO4] = 2 mM) or Thermopol II buffer (New England Biolabs, Mg free); an equimolar mix of 10 mM dNTPs (Fisher Biotech); and acetylated bovine serum albumin (BSA, 10 mg/mL) (Promega) as an enzyme stabilizer. All primers were diluted to 10 M before use.

Table 3.1. List of accessions of Murraya, Merrillia and Bergera used for molecular phylogenetic analyses and the locations from which they were sourced. The following abbreviations were used: Murraya (M); Merrillia (Me); Bergera (B); Australia (A), New South Wales (NSW), Queensland (QLD), Northern Territory (NT); Brazil (B), São Paulo (SP); China ,(C) Yingde (YD), Guangdong (GD), Guangxi (GX), Hainan (H); Indonesia (I), West Java (WJ), Central Java (CJ), East Java (EJ), Lombok (L), Nusa Tenggara Timur (NTT), Papua (IP); Taiwan (T); United States of America (U), University of California, Riverside (UCR), Fairchild Botanic Garden (FBG); Việt Nam (V), Tien Giang (TG), Cuc Phuong National Park (CP), Ho Chi Minh city (HCM), Dac Lac (DL), Thanh Hoa (TH), Ha Noi (HN), Bac Giang (BG). Voucher numbers are given for pressed specimens lodged at the Royal Botanic Garden, Sydney.

Accession Source Latitude & longitude Name of accession based on results in this Voucher Number chapter and in Chapter 4 Number Australia 2—ANSW Richmond, NSW 33°37'S, 150°45'E M. exotica 822701 4—ANSW Richmond, NSW 33°37'S, 150°45'E M. exotica 822702 6—ANSW Windsor, NSW 33°37'S, 150°49'E M. exotica 822703 8—ANSW Royal Botanic Garden, Sydney, NSW 33°52'S, 151°13'E M. exotica 9—ANSW Government House, Sydney, NSW 33°52'S, 151°13'E M. exotica 10—ANSW Richmond, NSW 33°36'S, 150°46'E M. exotica 822704 13—AQLD Brisbane, QLD 27°28'S, 152°58'E M. exotica 822705 14—AQLD Brisbane, QLD 27°27'S, 152°59'E M. exotica 822706 53—ANSW Royal Botanic Garden, Sydney, NSW 33°52'S, 151°13'E M. exotica 822707 54—AQLD Bundaberg, QLD 24°51'S, 152°24'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 822732 66—AQLD Woongarra, QLD (ex University of California, Riverside) 24°54'S, 152°24'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 69—ANT Haddon Head Beach, Blue Mud Bay, NT 13°22'S, 135°43'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 822733 70—ANT Darwin, NT 12°27'S, 130°50'E M. exotica 71—ANT Gove, NT 12°11'S, 136°43'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 72—AQLD Mt Carbine, QLD 16°31'S, 145°09'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 822734 73—AQLD Cooktown‐Mt Webb National Park, QLD 15°04'S, 145°07'E M. ovatifoliolata var. ovatifoliolata ‘large leaflet’ 822730 74—AQLD Battle Camp, QLD 15°17'S, 144°43'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 822735 75—AQLD Cairns, QLD 16°52'S, 145°40'E M. ovatifoliolata var. ovatifoliolata ‘large leaflet’ 822731 108—ANSW Richmond, NSW 33°37'S, 150°45'E M. exotica 115—AQLD Tondoon Botanic Gardens, Gladstone, QLD (via Royal Botanic Garden, 23°53'S, 151°15'E M. ovatifoliolata var. ovatifoliolata ‘small leaflet’ 822736 43

Accession Source Latitude & longitude Name of accession based on results in this Voucher Number chapter and in Chapter 4 Number Sydney) Brazil 102—BSP Capão Bonito, SP 24°00'S, 48°20'W M. exotica 103—BSP Capão Bonito, SP 24°00'S, 48°20'W M. exotica 104—BSP Botucatu, SP 22°53'S, 48°27'W M. exotica 105—BSP Botucatu, SP 22°53'S, 48°27'W M. exotica 106—BSP Araraquata, SP 21°47'S, 48°10'W M. exotica 107—BSP Araraquata, SP 21°47'S, 48°10'W M. exotica China 62—CGD South China Agricultural University, GD 23°09'N, 113°20'E M. exotica 63—CGD South China Agricultural University, GD 23°09'N, 113°20'E M. exotica 68—CGD South China Agricultural University, GD 23°09'N, 113°20'E M. exotica 76—CGX Guangxi via South China Botanical Gardens M. asiatica 822742 94—CYD Pipashan, Yingde County, GD 24°17′N, 113°21′E M. asiatica 822739 95—CYD Pipashan, Yingde County, GD 24°18′N, 113°21′E M. asiatica 822740 96—CYD Pipashan, Yingde County, GD 24°18′N, 113°21′E M. exotica 822717 97—CYD Hengshitang, Yingde County, GD 24°24′N, 113°18′E M. exotica 822718 98—CGX Guangxi via South China Botanical Gardens in Guangzhou, GD B. kwangsiensis 822748 99—CH Hainan via South China Botanical Gardens in Guangzhou, GD B. microphylla 822749 100—CH Bawangling, Hainan, via South China Botanical Gardens in Guangzhou, GD 19°07'N, 109°04'E M. exotica 822719 101—CGD South China Agricultural University, GD 23°09'N, 113°20'E M. exotica 822720 Indonesia 22—IWJ Bogor Botanic Garden, WJ 06°36'S, 106°48'E M. paniculata 822723 23—IWJ Bogor Botanic Garden, WJ 06°36'S, 106°48'E Me. caloxylon 822747 24—IP Bogor Botanic Garden (from Pegunungan Cycloop, Papua) 02°30'S, 140°31'E M. × cycloopensis 822746 25—IWJ Bogor Botanic Garden, WJ (from Merubetiri National Park, EJ) 06°36'S, 106°48'E M. paniculata 822724 27—IWJ Bogor Botanic Garden, WJ 06°36'S, 106°48'E M. exotica 822709 44

Accession Source Latitude & longitude Name of accession based on results in this Voucher Number chapter and in Chapter 4 Number 28—IWJ Bogor Botanic Garden, WJ 06°36'S, 106°48'E M. exotica 822710 30—IL Bogor Botanic Garden, WJ (from Lombok, NTT) 06°36'S, 106°48'E M. paniculata 34—IEJ Purwodadi Botanic Garden, EJ 07°48'S, 112°44'E M. paniculata 822725 35—IEJ Purwodadi Botanic Garden, EJ 07°48'S, 112°44'E M. exotica 822711 37—IEJ Purwodadi Botanic Garden, EJ 07°48'S, 112°44'E M. exotica 822712 38—IEJ Purwodadi Botanic Garden, EJ 07°48'S, 112°44'E M. paniculata 822726 40—IC Bayan, Purworejo, CJ (from China) 07°43'S, 109°56'E M. exotica 822713 42—IUCR Bayan, Purworejo, CJ (from UCR) 07°43'S, 109°56'E M. exotica 822714 44—ICJ Bayan, Purworejo, CJ 07°43'S, 109°56'E M. exotica 822715 45—ICJ Bayan, Purworejo, CJ 07°43'S, 109°56'E M. paniculata 822727 46—ICJ Yogyakata, CJ 07°44'S, 110°25'E M. paniculata 822728 47—ICJ Universitas Gadjah Mada, Yogyakata, CJ 07°44'S, 110°25'E M. exotica 822716 48—ICJ Universitas Gadjah Mada, Yogyakata, CJ 07°44'S, 110°25'E M. paniculata 822729 51—ICJ Universitas Gadjah Mada, Yogyakata, CJ 07°44'S, 110°25'E M. exotica 113—INTT Kupang, NTT 10°12'S, 123°36'E M. ovatifoliolata var. zollingeri 822737 114—INTT Kupang, NTT 10°12'S, 123°36'E M. ovatifoliolata var. zollingeri 822738 Taiwan 91—T Orchid Island, Taiwan 22°02'N, 121°32'E M. × omphalocarpa 822743 92—T Orchid Island, Taiwan 22°02'N, 121°32'E M. × omphalocarpa 822744 93—T Orchid Island, Taiwan 22°02'N, 121°32'E M. × omphalocarpa 822745 United States of America 64—UUCR University of California, Riverside 33°58'N, 117°20'W M. exotica 65—UUCR University of California, Riverside 33°58'N, 117°20'W M. exotica 67—UUCR University of California, Riverside 33°58'N, 117°20'W M. exotica

Accession Source Latitude & longitude Name of accession based on results in this Voucher Number chapter and in Chapter 4 Number 111—UFBG Fairchild Botanic Garden, Florida 25°41'N, 80°17'W M. exotica 822721 112—UFBG Fairchild Botanic Garden, Florida 25°41'N, 80°17'W M. exotica 822722 Việt Nam 57—VTG Chau Thanh, Tien Giang 10°24'N, 106°17'E M. exotica 58—VTG Chau Thanh, Tien Giang 10°24'N, 106°17'E M. exotica 59—VTG Chau Thanh, Tien Giang 10°24'N, 106°17'E M. exotica 60—VTG Chau Thanh, Tien Giang 10°24'N, 106°17'E M. exotica 61—VCP Cuc Phuong National Park, Ninh Binh 20°15'N, 105°42'E M. asiatica 77—VTG Cai Lay, Tien Giang 10°25'N, 106°07'E M. exotica 78—VTG Cai Lay, Tien Giang 10°25'N, 106°07'E M. exotica 79—VTG Cai Lay, Tien Giang 10°21'N, 106°22'E M. exotica 80—VTG Chau Thanh, Tien Giang 10°24'N, 106°17'E M. exotica 81—VHCM Dam Sen Park, Ho Chi Minh City 10°46'N, 106°38'E M. exotica 82—VHCM Dam Sen Park, Ho Chi Minh City 10°46'N, 106°38'E M. exotica 83—VHCM Tao Dan Park, Ho Chi Minh City 10°46'N, 106°42'E M. exotica 84—VDL Buon Ma Thuot, Dac Lac 12°40'N, 108°02'E M. exotica 85—VDL Buon Ma Thuot, Dac Lac 12°40'N, 108°02'E M. exotica 86—VTH Dong Ve, Thanh Hoa 19°47'N, 105°47'E M. exotica 87—VHN Bach Thao Park, Ha Noi 21°02'N, 105°50'E M. exotica 88—VCP Cuc Phuong National Park, Ninh Binh 20°15'N, 105°42'E M. asiatica 822741 o o 89—VBG Việt Yen, Bac Giang 21 17'N, 106 05'E M. exotica

Table 3.2. List of primer sequences and references used for molecular phylogenetic analyses.

Target Forward and 5′ – 3′ primer sequence Reference sequence reverse primer names trnL‐F c CGA AAT CGG TAG ACG CTA CG Taberlet et al. (1991) f ATT TGA ACT GGT GAC ACG AG Taberlet et al. (1991) psbM‐ trnDGUCR GGG ATT GTA GYT CAA TTG GT Shaw et al. trnDGUCspacer (2005): modified from Demesure et al. (1995) psbMF AGC AAT AAA TGC RAG AAT ATT TAC TTC CAT Shaw et al. (2005) rps16 rpsF GTG GTA GAA AGC AAC GTG CGA CTT Oxelman et al. (1997) rpsR2 TCG GGA TCG AAC ATC AAT TGC AAC Oxelman et al. (1997) matK‐5′trnK matK6 TGG GTT GCT AAC TCA ATG G Johnson & Soltis spacer (1994) matK5′R GCA TAA ATA TAY TCC YGA AAR ATA AGT GG Shaw et al. (2005) trnCGCA‐ycf6 ycf6R GCC CAA GCR AGA CTT ACT ATA TCC AT Shaw et al. (2005) region trnCGCAF CCA GTT CRA ATC YGG GTG Shaw et al. (2005) rps4‐trnT trnTUGUR AGG TTA GAG CAT CGC ATT TG Shaw et al. (2005) spacer rps4R2 CTG TNA GWC CRT AAT GAA AAC G Shaw et al. (2005) ITS ITS1 TCC GTA GGT GAA CCT GCG G White et al. (1990) ITS4 TCC TCC GCT TAT TGA TAT GC White et al. (1990)

PCR conditions for trnL-F

The chloroplast trnL (UAA) 5’ exon and trnF (GAA) region was amplified from total genomic DNA using the universal primers c and f (Table. 3.2 and Figure 3.1). The expected size of PCR products is from 300 bp to 650 bp with primers c & d and 250 bp to 650 bp with primers e & f depending on the species (Taberlet et al. 1991). The PCR reaction mixture contained 2.5 μL of Thermopol buffer, 1 L dNTPs, 0.5 L Taq polymerase, 3 μL each primer, 0.5 L BSA and 50 ng of template DNA in a total reaction volume of 25 μL. The thermal cycling parameters were: an initial denaturation for 5 min at 94°C; 30 cycles of 94°C for 1 min, annealing at 55°C for 1 min, and elongation at 72°C for 2 min; followed by an elongation step of 72°C for 5 min.

trnT (UGU) trnL (UAA) 5´exon trnL (UAA) 3´exon trnF(GAA)

a c e

b d f

Figure 3.1. Positions and directions of universal primers for trnL‐F region (Taberlet et al. 1991).

PCR conditions for psbM-trnDGUC

The psbM-trnDGUC target region (Figure 3.2) was amplified using primers psbMF and trnDGUCR (Shaw et al. 2005). The concentration of reagents in each 25 μL reaction were: 2.5 L of 10 × Thermopol buffer; 1 L dNTPs; 0.5 L Taq polymerase; 1 μL each primer; 0.5 L BSA; and 50 ng of template DNA. The mixture was initially denatured at 94°C for 5 min, followed by 35 cycles of 1 min denaturation at 94°C, 1 min annealing at 55°C, 3.5 min extension at 72°C and a final elongation period at 72°C for 5 min. According to Shaw et al. (2005), the average length of this spacer is 965 bp and ranges from 506–1801 bp.

psbMR ycf6R trnDGUCR Taxodium-psbMF2 psbMF ycf6F trnCGCAF

trnD psbM ycf6 trnC

Figure 3.2. Positions and directions of primers for psbM‐trnDGUC and trnCGCA‐ycf6 regions (Shaw et al. 2005).

PCR conditions for the rps16 intron

The intron sequence of chloroplast rps16 gene was amplified with primers, rpsF and rpsR2 (Oxelman et al. 1997). Each 25 μL reaction contained: 2.5 μL of 10 × Themopol buffer; 0.25 L Taq polymerase; 0.125 L of each dNTPs, 0.5 L BSA; 0.5 L each primer; and 50 ng template DNA. The cycling parameters were: denaturation 95°C for 3 min, followed by 33 cycles of 95°C for 30 sec, 60°C for 1 min and 72°C for 2 min and ending with 5 min at 72°C. The length of the rps16 intron varies from 757–951 bp (Oxelman et al. 1997).

PCR conditions for the matK–5′trnK spacer region

The primers matK5′R (Shaw et al. 2005) and matK6 (Johnson & Soltis 1994) were used to amplify this region (Fig. 3.3). A 25 μL reaction was set up consisting of: 2.5 μL of 10 × Thermopol buffer; 0.5 L of mixed dNTPs; 0.25 L Taq polymerase; 0.5 μL BSA; and 1 L each primer. The PCR cycling parameters were: 80°C for 5 min and 35 cycles of 94°C for 1 min, 50°C for 1 min, 72°C for 1.5 min, and then an elongation at 72°C for 5 min. The expected size of this region ranges from 704–860 bp (Shaw et al. 2005).

matKAR matK5’R matK6 matK5PSIF matK5PSIR matK5

5’trnK matK

Figure 3.3. Positions and directions of primers for the matK‐5′trnK spacer (Shaw et al. 2005).

PCR conditions for the trnCGCA-ycf6 region

TrnCGCA-ycf6 region (Fig. 3.2) was amplified using primers trnCGCAF and ycf6R (Shaw et al. 2005). Each 25 μL reaction contained: 2.5 μL 10 × Thermopol buffer; 0.5 L mixed dNTPs; 0.25 L Taq polymerase; 0.5 μL BSA; 1 L of each primer; and 50 ng target DNA. The PCR cycling conditions were an initial denaturation period of 5 min at 80°C, then 35 cycles of 94°C for 1 min, 55°C for 1 min, 72°C for 3.5 min, and a final elongation at 72°C for 5 min. According to Shaw et al. (2005), the expected fragment size is between 246–1071 bp.

PCR conditions for the rps4-trnT spacer region

The amplification of the rps4-trnT intergenic spacer (Figure 3.4) was performed using primers trnTUGUR and rps4R2 (Shaw et al. 2005). The PCR reaction mixture (25 μL) consisted of: 2.5 μL of 10 × Mg-free Thermopol II buffer; 0.625 μL of 100 mM MgSO4; 0.5 L of mixed dNTPs; 0.25 L Taq polymerase; 0.5 μL of BSA; 1 L of each primer; and 50 ng of template DNA. The cycling parameters were: 92°C for 3 min, followed by 30 cycles of 92°C for 1 min, 55°C for 1 min, 72°C for 3 min, and a final elongation of 7 min at 72°C. Shaw et al. (2005) obtained PCR products with an average size of 402 bp and a range of 345–785 bp (Fig. 3.4). 49

trnTUGUR

rps4R2

trnT rps4

Figure 3.4. Positions and directions of primers for rps4‐trnT spacer (Shaw et al. 2005).

PCR conditions for ITS

The ITS region of rDNA (Figure 3.5) was amplified using primers ITS1 and ITS4 (White et al. 1990). The reaction mixture (25 μL) contained: 2.5 μL of 10 × Thermopol buffer; 0.5 L mixed dNTPs; 0.25 L Taq polymerase; 0.5 μL BSA; 1 L of each primer; and 50 ng of extracted DNA. The samples were denatured at 94°C for 90 sec, followed by 30 cycles of denaturation (95°C for 50 sec), annealing (55°C for 70 sec) and extension (72°C for 90 sec) and then by an elongation step of 3 min at 72°C.

NS5 ITS1 ITS2

5.8S 18S rRNA rRNA 28S rRNA gene gene

NS6 ITS4

ITS regions

Figure 3.5. Positions and directions of primers for ITS regions (http://www.fao.org/ DOCREP/005/ X4946E/x4946e06.gif).

3.2.3. DNA sequencing

PCR products were subjected to electrophoresis in a 1% agarose gel containing 0.5 µg/ml ethidium bromide and the gel visualised using the Gel Documentation System (Bio-Rad) to identify the success of DNA amplifications. Successful amplifications were purified using the Wizard® SV Gel and PCR Clean-Up System (Promega) following the manufacturer’s instructions. The purified PCR products were quantified using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific) and diluted to 50 50 ng/μL. Both strands of purified fragments were sequenced using the same primers as were used for amplification (Table 3.2) by automated sequencing using an Applied Biosystems 3730XL sequencer at Macrogen Inc. (908 World Meridian Venture Center, #60–24, Gasan-dong, Geumchun-gu, Seoul 153–781, Korea).

3.2.4. Phylogenetic analysis

DNA Baser software (Version V.2.91, Heracle BioSoft) was used to compile contiguous sequences (contigs) and sequence alignments were obtained using ClustalW (Thompson et al. 1994) as implemented in Bioedit (Hall 2001, Version 5.0.6); each alignment was subsequently checked by eye. Aligned datasets were then analysed using PAUP* 4.0b10 (Swofford 2002) using the maximum parsimony (MP) optimality criterion. Parsimony analysis was performed using tree-bisection-reconnection branch swapping with a heuristic search with 1000 bootstrap replicates, holding 1 tree at each step during stepwise addition and the steepest descent option not in effect. The MP analysis was performed for individual cpDNA data sets and for the ITS region then for a combination of all chloroplast regions and cpDNA+ITS. For the analysis of individual sequences, gaps were treated as missing data and branches with a minimum length of zero were collapsed. The analyses of the combined chloroplast sequences were based on two matrices, one including gaps coded only as missing characters and the other comprising this first data matrix with the addition of data from indels that were scored for presence or absence using the criteria of Simmons & Ochoterena (2000). In addition, a data matrix that consisted of the presence/absence data from the indels was also subjected to phylogenetic analysis.

The sequence data were also analysed using Bayesian inference (BI) as implemented in MrBayes v.3.1 (Ronquist et al. 2005). However, before BI analysis, an appropriate nucleotide substitution model was identified using hierarchical likelihood ratio tests (hLRTs) implemented in MrModeltest v2 (Nylander 2004) for selection of the best-fit model. The Markov chain Monte Carlo simulations (MCMC) were started with 100,000 generations and were run until the standard deviation of split frequencies was below 0.01. At this stage, the number of generations (ngen) to reach this level was recorded. A burnin value was then calculated as follows:

ngen x 0.25 Burnin = Sample frequency

For this calculation, a sample frequency of 10 was used. The program was restarted using the ‘sump’ command and the appropriate burnin value to allow the statistics associated with the substitution model to be calculated. Following this, the tree data were calculated using the ‘sumt’ command set again with the appropriate burnin value. In both BI and MP analyses, Citrus sp. (accession 90—VTG), Bergera kwangsiensis (accession 98—CGX) and Bergera microphylla (accession 98—CH) were used as outgroups.

3.2.5. Incongruence length difference (ILD) test

ILD tests were performed between chloroplast data sets that were representative of the different nucleotide substitution models determined by MrModeltest. ILD tests were also performed between the combined chloroplast data and the ITS data using the partition homogeneity test as implemented in PAUP*. A second ILD test was made on this data set; however, for this second test, accessions 24, 91, 92 and 93 were excluded. These accessions are the members of the groups that caused significant incongruences in the topology of trees derived from chloroplast and ITS data.

3.2.6. Determination of monophylogeny of Murraya

To determine whether the genus Murraya is monophyletic, the concatenated sequences of five chloroplast regions from five species within the Clauseneae, and seven members of the Aurantieae, were subjected to parsimony analysis using PAUP* together with accessions representing the taxa identified in this study within Murraya. In addition, the sequences of the ITS regions from three and eight members, respectively, from the Clauseneae and Aurantieae were also subjected to parsimony analysis together with the accessions of Murraya. The accession numbers of sequence data obtained from GenBank and the regions used for the analyses are given in Table 3.3. Outgroup selection was based on Bayer et al. (2009) with Ruta graveloens L. [Rutoideae: Ruteae] from the subfamily Rutoideae being used as the outgroup in both analyses.

Table 3.3. GenBank accession numbers for the regions used to determine the monophyly of Murraya.

Subfamily Region Tribe ITS matK‐5′trnK rps16 psbM‐trnDGUC trnL‐F rps4‐trnT Species spacer spacer spacer Rutoideae Ruteae Ruta graveolens L. FJ434146 EF138915 EU853765 EF134707 AY295275 EF134707 Aurantioideae Clauseneae Bergera koenigii L. FJ434147 EF138843 AF320261 EF134635 EF126637 EF134635 Clausena excavata Burm. f. FJ434152 EF138881 AY295258 EF134673 AY295284 EF134673 Glycosmis pentaphylla (Retz.) DC. FJ434151 Micromelum minutum (Forst. f.) Wight & Arn. EF138904 AF320266 EF134696 EF126691 EF134696 Aurantieae Aegle marmelos (L.) Corr. Serr. FJ434169 EF138836 AY295268 EF134628 AY295294 EF134628 Atalantia (Severinia) buxifolia (Poir.) Oliv. FJ434156 Atalantia monophylla (L.) DC. GQ225867 EF138841 EF126570 EF134633 EF126636 EF134633 Citrus medica L. EF138871 EF126599 EF134663 FN599494 EF134663 Pleiospermium latialatum Swingle FJ434157 EF138913 EF126628 EF134705 EF126697 EF134705 Murraya sp. FJ434153 Triphasia trifolia (Burm. f.) P. Wilson FJ434172 EF138921 AY295271 EF134713 AY295297 EF134713 Wenzelia dolichophylla (Lauterb. & K. Schum.) Tan. FJ434150

3.3. Results

In order to describe the results of the various analyses in this and the following chapter, and to avoid confusion and ambiguity, it was necessary, as previously noted above in Table 3.1, to label accessions in cladograms, and elsewhere, presented in both chapters with specific names determined on the basis of conclusions reached in the thesis:  Murraya exotica;  Murraya paniculata;  Murraya asiatica;  Murraya ovatifoliolata var. ovatifoliolata (small and large leaflet forms)  Murraya ovatifoliolata var. zollingeri;  Murraya × omphalocarpa; and  Murraya × cycloopensis.

3.3.1. Statistical summary of the sequence data

The statistical summary of all sequence data is shown in Table 3.4. The data are presented as six individual chloroplast sequence regions, the sequence of the ITS region and the combination of all chloroplast sequences as well as the combination of all chloroplast and ITS sequences together. The minimum and maximum aligned length of chloroplast sequences was from 551–954 nucleotides for the rps4-trnT spacer and trnL- F, respectively. Among the chloroplast regions, trnCGAC-ycf6 had the largest proportion of gaps (11.31%) whilst rps16 region had the smallest proportion (2.64%). The length of aligned sequences of ITS was 625 bp and the percentage of gaps for ITS was 5.93%, all chloroplast regions and combination of chloroplast and ITS is from 5–6%.

With regard to parsimony informative characters (PICs), among the chloroplast sequences, the percentage of PICs was highest for trnL-F (3.46%) and the range of PICs between other chloroplast regions was from 1.72% (matK-5’trnK) to 2.79% (psbM- trnDGUC). The PICs of combined chloroplast regions was 2.46% and that of ITS was 8.16%; the percentage of PICs in all regions was 2.43%.

In Table 3.4, the minimum and maximum sequence divergence between all taxa for each individual DNA region examined, and for various combinations, are presented. The results show that the greatest range was found for the ITS region and the least for rps4-trnT.

Table 3.4. Genetic statistics of individual sequences, all chloroplast sequences and chloroplast with ITS sequences together.

Statistics/Target regions psbM‐trnDGUC rps16 matK‐5′trnK rps4‐trnT trnL‐F trnCGCA‐ycf6 All ITS Chloroplast chloroplast & ITS sequences sequences Number of accessions 78 86 84 82 84 85 78 53 51 Aligned length of matrix 717 830 754 551 954 821 4627 625 5218 Number of constant characters 656 777 698 529 901 774 4335 519 4840 Percentage of constant characters 91.49 93.61 92.57 96.01 94.44 94.28 93.69 83.04 92.76 Number of parsimony uninformative 41 34 43 12 20 28 178 55 251 characters Percentage of parsimony uninformative 5.72 4.10 5.70 2.18 2.10 3.41 3.85 8.80 4.81 characters Number of parsimony informative characters 20 19 13 10 33 19 114 51 123 Percentage of parsimony informative 2.79 2.29 1.72 1.81 3.46 2.31 2.46 8.16 2.35 characters Range of pairwise sequence divergence among 0‐0.049 0‐0.036 0‐0.038 0‐0.020 0‐0.030 0‐0.038 0‐0.034 0‐0.065 0‐0.037 all taxa (p‐distance) Number of branches receiving > 50% bootstrap 9 6 6 8 11 8 21 10 17 support Number of branches receiving posterior 10 9 6 8 11 8 20 10 17 probability support Maximum sequence divergence between 0.033 0.028 0.028 0.015 0.03 0.02 0.023 ‐ ‐ Bergera kwangsiensis and in‐group (%) Maximum sequence divergence between 0.049 0.033 0.038 0.02 0.03 0.033 0.031 0.048 0.033 Bergera microphylla and in‐group Maximum sequence divergence between out 0.049 0.036 0.031 0.011 0.026 0.038 0.034 0.065 0.037 group, Citrus sp., and in‐group (%) Average percentage of gaps in taxa 3.64 2.64 5.34 5.09 5.29 11.31 5.5 5.93 4.96 55

3.3.2. Phylogenetic results

Phylogeny derived from trnL-F

In this study, the trnL-F region was amplified from 84 DNA accessions (80, 1, 2 and 1 accessions of Murraya, Merrillia, Bergera and Citrus, respectively), and the sequences had a length of 954 bp, of which 33 (3.46%) were parsimoniously informative (Table 3.4). The results of MP bootstrap analysis are presented in Fig. 3.6. The Murraya accessions formed a clade with a basal 6-partite polytomy of 6 clades:  a Murraya exotica clade that contained all 53 accessions of Murraya exotica collected from Australia, Indonesia, China, Việt Nam, Brazil and the United States of America, the dwarf mock orange, ‘Min-a-Min’ (accession 70—ANT), and sister to the remainder of the accessions, Murraya × cycloopensis (accession 24—IP) from Papua, and clades that comprised  a clade comprising Murraya ovatifoliolata var. ovatifoliolata large leaflet (accessions 73—AQLD, 75—AQLD) and small leaflet (accessions 54—, 66— AQLD, 69—ANT, 71—ANT, 72—, 74—AQLD, 115—AQLD) forms described by Brophy et al. (1994) from Northern Queensland and the Northern Territory of Australia, Murraya ovatifoliolata var. zollingeri from Indonesia (accessions 113—INTT, 114—INTT), and Murraya × omphalocarpa from Taiwan (accessions 91—T, 92—T, 93—T), and within whichthe accessions of Murraya × omphalocarpa formed a weakly-supported sub-clade as did three small leaflet accessions (54—AQLD, 66—AQLD, 115—AQLD) of Murraya ovatifoliolata var. ovatifoliolata.  a clade comprising Murraya paniculata accessions from Java and Lombok.  and 3 clades comprising Murraya asiatica accessions from:  Yingde in China (94—CYD, 95—CYD),  Cuc Phuong National Park in Việt Nam (61—VCP, 88—VCP), and  Guangxi in China (76—CGX).

The same set of data used in MP analysis was subjected to Bayesian analysis. However, before analysis, the sequences were analysed by MrModeltest to determine the best model of nucleotide substitution; this program selected the Felsenstein 1981 model plus Gamma (F81+G) (Felsenstein 1981, Yang 1993). The burnin value for this data set was 25,000. The phylogenetic tree derived from BI (Fig. 3.7) was identical in topology with

56 the MP tree and the posterior probabilities strongly support the existence of the 6 clades:  Murraya exotica accessions with a posterior probability of 99%:  Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa (posterior probability = 99%):  with sub-clades comprising Murraya ovatifoliolata var. ovatifoliolata and Murraya × omphalocarpa:  Murraya paniculata accessions from Indonesia (posterior probability = 99%); and  3 clades comprising Murraya asiatica accessions from mainland Asia.

As in the MP analysis, Merrillia caloxylon was found to be sister to Murraya.

Figure 3.6. 50% majority‐rule bootstrap consensus tree of the trnL‐F region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the out‐ group and bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.7. Bayesian inference tree resulting from analysis of the plastid trnL‐F region from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from psbM-trnDGUC

The aligned psbM-trnDGUC data matrix of 74 Murraya accessions, 2 Bergera accessions, 1 Citrus accession and 1 Merrillia accession had 717 characters of which 20 (2.79%) were parsimoniously informative. In the MP bootstrap analysis (Fig. 3.8), Murraya and Merrillia formed a 5-partite basal polytomy consisting of:  a clade comprising all accessions of the Murraya exotica and Murraya × cycloopensis (accession 24—IP), and that lacked further internal resolution;  a clade comprising Murraya paniculata accessions from Java and Lombok, Murraya ovatifoliolata var. ovatifoliolata (small and large leaflet forms), Murraya ovatifoliolata var. zollingeri, and Murraya × omphalocarpa;  a clade comprising Merrillia caloxylon; and  two clades comprising Murraya asiatica accessions from mainland Asia.

In the BI analysis (Fig. 3.9), the F81 model (Felsenstein 1981) was selected by MrModletest as an appropriate model for the sequences of the psbM-trnDGUC region and a burnin value of 37,500 was required for the analysis. The Murraya exotica accessions and Murraya × cycloopensis formed a basal polytomy. Separate from these was a clade that contained Merrillia caloxylon and the forms of Murraya paniculata, Murraya asiatica, Murraya ovatifoliolata var. ovatifoliolata (small and large leaflet forms) and Murraya ovatifoliolata var. zollingeri. Within this clade, the sub-clades matched the separation of the accessions in the phylogeny resulting from MP analysis.

Figure 3.8. 50% majority‐rule bootstrap consensus tree of the psbM‐trnDGUC region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup and bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.9. Bayesian inference tree resulting from analysis of psbM‐trnDGUC region from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from trnCGCA-ycf6

For the trnCGCA-ycf6 region, 85 accessions (81 Murraya, 1 Merrillia, 2 Bergera, and 1 Citrus) were successfully amplified and sequenced. The data matrix consisted of 821 characters of which 19 (2.31%) were parsimoniously informative. For the Bayesian analysis, the general time reversible model (GTR: Lanave et al. 1984, Tavaré 1986, Rodriguez et al. 1990) was selected by MrModeltest. The Markov chain Monte Carlo analysis was run for 1,500,000 generations (burnin = 37,500). The MP bootstrap and BI trees resulting from these analyses (Fig. 3.10 and 3.11) were identical. Four clades were resolved:  a clade sister to all other accessions and comprising Murraya paniculata accessions from Java and Lombok.  a clade comprising all Murraya exotica accessions, to which Murraya × cycloopensis was sister.  a clade comprising Murraya ovatifoliolata var. zollingeri, Murraya ovatifoliolata var. ovatifoliolata small and large leaflet forms, and Murraya × omphalocarpa, that did not resolve into sub-clades.  a clade comprising Murraya asiatica accessions 94—CYD and 95—CYD from Yingde, Guangdong, China, to which Murraya asiatica accessions from Cuc Phuong, Việt Nam (61—VCP 88—VCP) and 76—CGX from Guangxi, China, none of which grouped together, and the Merrillia caloxylon accession (23— IWJ), were sister.

Figure 3.10. 50% majority‐rule bootstrap consensus tree of the trnCGCA‐ycf6 region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percents from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.11. Bayesian inference tree resulting from analysis of trnCGCA‐ycf6 region from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from rps4-trnT

The aligned sequences for this region consisted of 551 characters and there were 10 (1.81%) PICs; the results of the MP bootstrap analysis are shown in Fig. 3.12. For the Bayesian analysis, F81 was proposed by MrModeltest as the most appropriate model of evolution, the analysis was run for 1,300,000 generations (burnin = 32,500) and the results of the analysis are presented in Fig. 3.13. These two analyses produced identical trees. Murraya and Merrillia form an eight-partite basal polytomy that includes the following ungrouped accessions and subclades:  a clade comprising Merrillia caloxylon;  5 separate clades comprising Murraya asiatica accessions;  a clade containing the Murraya paniculata accessions from Java and Lombok as well as Murraya ovatifoliolata var. zollingeri, Murraya ovatifoliolata var. ovatifoliolata small and large leaflet forms, and Murraya × omphalocarpa;  a clade consisting of the Murraya exotica accessions showing some internal phylogenetic resolution, with 3 internal nodes:  a basal trichotomy consisting of accessions 67—UUCR from the University of California, Riverside, and 100—CH from Hainan, and a clade that comprised all other Murraya exotica accessions,  with accession 97—CYD from Yingde, China, being sister to the rest,  and the accessions from Brazil and Australia forming a sub-clade.

Figure 3.12. 50% majority‐rule bootstrap consensus tree of the rps4‐trnT region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.13. Bayesian inference tree resulting from analysis of the rps4‐trnT region from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from matK-5′trnK

Part of the matK region was amplified and sequenced with primers matK5′R and matK6. This produced an alignment of 754 nucleotides of which 1.72% was parsimoniously informative; the results of the MP bootstrap analysis are shown in Fig. 3.14. The Bayesian analysis was performed using the F81 (Felsenstein 1981) model of nucleotide substitution and was run for 2,200,000 generations (burnin = 55,000). The result of the analysis is shown in Fig. 3.15. There were substantial differences between the topology of the trees produced by these analyses, with the MP tree being the lesser resolved of the two. Within both analyses:  Murraya exotica accessions and Murraya × cycloopensis, formed one clade and  Murraya paniculata accessions from Java and Lombok formed another.

In the Bayesian analysis (Fig 3.15):  the accessions of small and large leaflet forms of Murraya ovatifoliolata var. ovatifoliolata formed a single clade, with Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa forming sub-clades.  Murraya asiatica accessions and the accession of Merrillia caloxylon formed part of a large, basal polytomy with the accessions of Bergera microphylla (99—CH), Bergera kwangsiensis (98—CGX) and Citrus sp. (90—VTG)

In contrast, in the MP analysis, Murraya × omphalocarpa and Murraya paniculata var. zollingeri both formed separate clades, and the accessions of Murraya ovatifoliolata var. ovatifoliolata formed part of a large polytomy along with the above mentioned clades and Merrillia caloxylon.

Figure 3.14. 50% majority‐rule bootstrap consensus tree of the matK‐5′trnK region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.15. Bayesian inference tree resulting from analysis of the matK‐5′trnK region from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from rps16

The intron of rps16 was amplified with primers rpsF and rpsR2 and the rps16 alignment matrix from 86 accessions of Murraya, Merrillia, Bergera and Citrus contained 830 characters, of which 19 were PICs; the result of the MP bootstrap analysis is shown in Fig. 3.16. The Model F81 + G was used for the Bayesian analysis and the Markov chain was run for 2,000,000 generations (burnin = 50,000).

The MP analysis produced a tree with little resolution (Fig. 3.16) but with Murraya exotica, Murraya asiatica and Merrillia caloxylon forming a clade, nested within which was another clade that contained the Murraya paniculata accessions from Java and Lombok and a sub-clade containing Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa.

The Bayesian analysis (Fig. 3.17) separated a clade that contained the Murraya exotica accessions, and Murraya × cycloopensis, and within this clade, the Murraya asiatica accessions from Việt Nam were sister to the Murraya exotica accessions. A second clade contained the Murraya paniculata accessions from Java and Lombok and a sub- clade containing Murraya ovatifoliolata var. ovatifoliolata (small and large leaftelt forms), Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa. The remaining accessions (the Murraya asiatica accessions from China and Merrillia caloxylon) were separated from the accessions of Bergera microphylla, Bergera kwangsiensis and Citrus sp., but were otherwise unresolved.

Figure 3.16. 50% majority‐rule bootstrap consensus tree of the rps16 region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.17. Bayesian inference tree resulting from analysis of the rps16 region from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from 6 chloroplast genes

The phylogenetic analysis of the individual chloroplast regions did not show any major topological incongruences among these individual analyses. Before combining the chloroplast data sets for further analysis, ILD tests were performed among all pairwise combinations of the following regions: trnL-F, psbM-trnDGUC and trnCGCA-ycf6. These regions have different nucleotide substitution models determined by MrModeltest (F81+G, F81, GTR, respectively). The tests returned the P values of 1.0, 0.174, and 0.506 for trnL-F & psbM-trnDGUC, trnL-F & trnCGCA-ycf6 and psbM-trnDGUC & trnCGCA-ycf6, respectively. These results show that the sequences of chloroplast regions are homogeneous and can legitimately be combined. Therefore, the sequence data from the six different regions were combined to increase the resolution and support values. The data were from 78 accessions and the length of the combined sequences was 4627 bp of which 114 (2.46%) were PICs; the result of the MP bootstrap analysis from sequence data only is shown in Fig. 3.18. In addition, data from the coded indels were added to the combined sequence data and also subjected to MP bootstrap analysis (Fig. 3.19). The model of nucleotide substitution used for the Bayesian analysis was GTR+G and the Markov chains were run for 5,000,000 generations (burnin = 1,250,000 generations) and the tree is shown in Fig. 3.20. All three trees had similar topologies and the following features:  In all analyses, Merrillia caloxylon was sister to all Murraya accessions.  In all analyses, the Murraya exotica accessions generally separated into 5 sub- clades:  accession 24—IP, Murraya × cycloopensis, from Papua;  Murraya exotica accession 67—UUCR from the University of California, Riverside, and accession 100—CH from Hainan, China  Murraya exotica accession 97—CYD from Yingde, China  Murraya exotica accessions predominantly from Australia and Brazil including the dwarf cultivar, ‘Min-a-Min’ (70—ANT);  Murraya exotica accessions predominantly from Việt Nam and Indonesia; with  other Murraya exotica accessions from China and the USA being distributed between the sub-clades comprising accession from Australia and Brazil, and Việt Nam and Indonesia .

 In all analyses, the Murraya paniculata, Murraya asiatica, Murraya ovatifoliolata and Murraya × omphalocarpa accessions separated from the Murraya exotica accessions and Murraya × cycloopensis. In the MP analysis of the chloroplast sequence data only, these accessions formed a single clade containing three sub- clades. When the indel data were added to the data set, the Murraya asiatica accessions formed a second clade separate from the Murraya paniculata and Murraya ovatifoliolata accessions. Also, in general, the bootstrap values were higher when the indel data were included. In the analysis of the chloroplast sequence data only using BI, the Murraya asiatica accessions formed two separate clades with the accessions from China in one clade, and accessions from the Việt Nam in the second.  In all analyses, the Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa accessions formed one sub-clade that was sister to the Murraya paniculata accessions and within this sub-clade the accessions of Murraya × omphalocarpa grouped together as did the accessions of Murraya ovatifoliolata var. zollingeri. However, the accessions of Murraya ovatifoliolata var. ovatifoliolata grouped into three sets of pairs with two accessions being unresolved.

The chloroplast sequences contained 42 informative indels (psbM-trnDGUC: 10; trnL-F: 6; trnCGCA-ycf6: 9; rps16: 7; matK-5′trnK: 6; rps4-trnT: 4). The data from indels only was also phylogenetically informative and the tree derived from this analysis is shown in Fig. 3.21. The results show that all Murraya exotica accessions formed a single clade that was separate from all Murraya paniculata, Murraya asiatica, Murraya ovatifoliolata and Murraya × omphalocarpa accessions. Within this clade, two Murraya exotica accessions from China (accessions 97—CYD and 100—CH) grouped together but did not group with two other Murraya exotica accession from China (96— CYD and 101—CGD). Also, an accession 111—UFBG from Florida and 82—VHCM from Việt Nam grouped together but these accessions did not group with other Murraya exotica accessions from these countries.

Among the Murraya paniculata, Murraya asiatica, Murraya ovatifoliolata and Murraya × omphalocarpa accessions, there was little resolution and what resolution occurred was based on their geographical origin. Thus, Murraya paniculata accessions from Indonesia formed a clade, as did the Murraya × omphalocarpa accessions from

Taiwan, the Murraya ovatifoliolata var. ovatifoliolata accessions from Northern Territory, Australia, the three Chinese Murraya asiatica accessions from China (76— CGX, 94—CYD and 95—CYD), and Cuc Phuong, Việt Nam (61—VCP and 88— VCP). The other accessions formed a large polytomy at the base of Murraya plus Merrillia. Most accessions of Murraya ovatifoliolata var. ovatifoliolata formed part of this polytomy. Murraya × cycloopensis (accession 24—IP) did not group with Murraya exotica accessions.

Figure 3.18. 50% majority‐rule bootstrap consensus tree based on the combined sequences of the six chloroplast regions from accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.19. 50% majority‐rule bootstrap consensus tree based on the combined sequences of the six chloroplast regions from accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis including indels. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.20. Bayesian inference tree based on the combined sequences of the six chloroplast regions from accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.21. 50% majority‐rule bootstrap consensus tree based on the indels of six chloroplast regions from accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from ITS region

In this study, besides using chloroplast regions, the phylogenetic relationships between accessions of Murraya and Merrillia were also tested using part of the nuclear rDNA ITS region. This analysis used 53 accessions of Murraya, Merrillia, Bergera and Citrus that represented every clade and sub-clade found in the chloroplast analyses. The sequence matrix consisted of 625 nucleotides of which, 51 sites (8.6%) were PICs. The tree produced by MP bootstrap analysis is shown in Fig. 3.22. For the Bayesian inference of the ITS region, the GTR+G model was applied and the resulting phylogeny, after running 1,000,000 generations with a sample frequency of 10 and a burnin of 250,000 generations is shown in Fig. 3.23. The two trees produced by these two types of analysis were identical and consist of five clades:  a clade containing the accessions of Murraya × omphalocarpa and all Murraya exotica accessions including ‘Min-a-Min’ but not Murraya × cycloopensis (accession 24—IP) from Papua. This placement of the accessions of Murraya × omphalocarpa is strongly incongruent with their position in the cpDNA trees;  a clade containing Murraya asiatica accessions from China and Việt Nam;  a clade containing the Murraya paniculata accessions from Java, Indonesia;  a clade containing Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri, and Murraya × cycloopensis from Papua. Within this clade, the accessions of Murraya ovatifoliolata var. ovatifoliolata from the Northern Territory separated from those from Queensland. The position of Murraya × cycloopensis was strongly incongruent with its position in the cpDNA trees; and  a clade containing Merrillia caloxylon.

Figure 3.22. 50% majority‐rule bootstrap consensus tree of the ITS region of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.23. Bayesian inference tree based on the ITS sequences of accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Phylogeny derived from combination of sequences of 6 chloroplast genes and the ITS region

Before combining the sequences of chloroplast and ITS regions for phylogenetic analysis, an incongruence length difference (ILD) test was performed between the combined data from the chloroplast regions and the data from the ITS region, again using the partition homogeneity test offered by PAUP* 4.0b10 (Swofford 2002). This test was performed on 51 accessions of Murraya, Bergera, Citrus and Merrillia caloxylon: this test returned a P value of 0.001 indicating that the data sets were heterogenous and should not be combined. When accessions of Murraya × omphalocarpa from Taiwan (accessions T—91, T—92 and T—93) and Murraya × cycloopensis from Papua (24—IP) were removed from the data sets and the test repeated, it returned a P value of 0.03. Therefore, following this test, MP and BI analyses were performed on the combined chloroplast and ITS data from 47 accessions of Murraya, Merrillia, Bergera and Citrus but without Murraya × omphalocarpa and Murraya × cycloopensis (see Discussion for the acceptance of the null hypothesis at the 0.03 level). This resulted in a sequence matrix consisting of 5218 nucleotides of which, 123 sites (2.35%) were PICs. The tree produced by MP is showed in Fig. 3.24. For the Bayesian analysis, the GTR+G model was applied and the analysis run for 600,000 generations (burnin = 150,000 generations). The tree resulting from Bayesian inference is presented in Fig. 3.25. The two trees resulting from these analyses were similar:  The Murraya exotica accessions formed a clade that was clearly separated from the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions and the internal topology of this clade was identical between the two analyses.  In both trees, the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions formed part of a polytomy with the Murraya exotica clade. Using MP, 2 clades, a Murraya paniculata clade, a Murraya asiatica clade comprising the three accessions from China (76—CGX, 94—CYD and 95— CYD), and an ungrouped Murraya asiatica accession from Việt Nam (88—VCP), were formed whilst using BI, Murraya asiatica accession 88—VCP from Việt Nam grouped weakly with the clades formed by the Murraya asiatica accessions from China.

 In both analyses, a clade was formed containing the Murraya paniculata accessions from Indonesian and Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri. In the MP analysis, Murraya ovatifoliolata 85

var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri formed monophyletic sister clades, forming a clade that was, in turn, sister to a clade comprising the two Murraya paniculata accessions from Indonesia. The BI result was similar except Murraya ovatifoliolata var. ovatifoliolata was not monophyletic, forming two separate clades.

 In the MP analysis, the Murraya asiatica 88—VCP from Việt Nam did not group with any other Murraya asiatica accessions whilst using BI, this accession was sister to the Murraya asiatica accessions from China.

Figure 3.24. 50% majority‐rule bootstrap consensus tree of the ITS region combined with 6 chloroplast regions of accessions of Murraya, Merrillia, Bergera and Citrus derived from maximum parsimony analysis. Citrus sp. was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Figure 3.25. Bayesian inference tree based on the ITS region combined with 6 chloroplast regions of accessions of Murraya, Merrillia, Bergera and Citrus. Citrus sp. was used as the outgroup and posterior probabilities are shown above each branch. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

Monophyly of Murraya

The phylogenetic placement of the taxa resulting from the analysis of the combined chloroplast regions (Fig. 3.26) was not well resolved and consisted of a large polytomy. A large clade contained Merrillia caloxylon and this species was sister to all of the accessions of Murraya that formed a single subclade. The three species of Bergera and Micromelum minutum formed a second clade; however, Clausena excavata, the other species representing of the Clauseneae did not group with these species. The remaining seven species from the Aurantieae formed five clades with the two Citrus species grouping together and Aegle marmelos and Triphasia trifolia also grouping within one clade.

Within the phylogram derived from the ITS sequences (Fig. 3.27), Glycosmis pentaphylla was sister to the other species of the ingroup. Within these species, three clades were formed, the first consisting of Merrillia caloxylon, the second containing all members of Murraya and the third consisting of all remaining members of the Clauseneae and Aurantieae.

Figure 3.26. 50% majority‐rule bootstrap consensus tree derived from the combined sequences of five chloroplast regions of accessions of Murraya and members of the Clauseneae and Aurantieae. Ruta graveolens was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications.

Figure 3.27. 50% majority‐rule bootstrap consensus tree derived from the ITS regions of accessions of Murraya and members of the Clauseneae and Aurantieae. Ruta graveolens was used as the outgroup. Bootstrap values are provided as percentages from 1000 replications.

3.4. Discussion

In molecular phylogenetic analysis, there are three basic methods, maximum parsimony, maximum likelihood and distance, which have been used to construct phylogenetic trees from alignments of DNA sequences. Recently, Bayesian inference, a likelihood-based method, has also been used and has become popular in phylogenetic studies (Alfaro et al. 2003). The merits and shortcomings of these methods have been considered by several authors including Huelsenbeck (1995), Alfaro et al. (2003) and Simmons et al. (2006). In addition to these basic methods, the selection of the most suitable model of evolution can also be an important choice in the process of phylogenetic reconstructions (Talianová 2007).

Alfaro et al. (2003) suggested that Bayesian inference is useful when systematists wish to show how well data support the results of phylogenetic analyses. As a result of a simulation study performed by these authors, they claimed that BI appears to increase the sensitivity to phylogenetic signals, which may allow workers to have high confidence of a correct result with few characters. They also claimed that BI is a less biased predictor of phylogenetic accuracy than either parsimony or maximum likelihood used in combination with bootstrapping and suggested that BI provides high support values for correct topological bipartitions with fewer characters than required for nonparametric bootstrapping. Therefore, for a given data set, BI will, on average, attach high confidence to a greater number of correct internodes than does nonparametric bootstrapping. In their simulation studies, Simmons et al. (2006) also found that Bayesian analyses performed better with respect to resolution than parsimony when heterogeneous simulation parameters were used in substitution models; however, parsimony analysis was superior when heterogeneous simulation parameters were not incorporated into substitution models. Their simulation studies showed that at high rates of evolution, parsimony performed better relative to Bayesian analyses. Parsimony was also found to be more conservative than Bayesian analyses as it resolved fewer incorrect clades.

According to Talianová (2007), parsimony methods can be better than other techniques, as they are relatively free of assumptions relating to nucleotide and amino acid substitutions. Talianová (2007) also stated that MP works well when compared sequences are not too divergent, when the rates of nucleotide substitution are relatively

92 constant and the number of nucleotides examined is large. However, MP trees can suffer from inaccuracies due to long-branch attraction (Hendy & Penny 1989), a phenomenon that occurs when rapidly evolving sequences are artefactually inferred to be closely related (Felsenstein 1978). Thus, the two methods of analysis used in this study, maximum parsimony and Bayesian inference, both have strengths and weaknesses.

In this study, the results from the two phylogenetic techniques used were largely in agreement. In most analyses, Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions were separated from Murraya exotica accessions. The phylogenetic trees resulting from MP and BI were identical for three chloroplast regions (trnL-F, trnCGCA-ycf6, rps4-trnT) as well as for the ITS region. There were minor variations between MP and BI in phylogenetic trees of other regions; however, such variation mainly occurred within the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata clades.  For the psbM-trnDGUC spacer, using MP, the Murraya exotica accessions formed a single clade whilst using BI this clade was not resolved. With MP, the Murraya asiatica accessions formed separate clades, whilst with BI, they were sister to the other accessions of Murraya paniculata and Murraya ovatifoliolata that fell within one clade.  For matK-5′trnK spacer, in the MP analysis, Murraya × omphalocarpa and Murraya ovatifoliolata var. zollingeri fell in two separate clades and the accessions of Murraya ovatifoliolata var. ovatifoliolata were unresolved, whilst with BI, Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa formed a single clade. Surprisingly, using BI, the Bergera and Citrus accessions (used as outgroups) did not separate from the other accessions.  For the rps16 region, using MP, the Murraya exotica accessions were unresolved, whereas they formed a clade separate from the Murraya paniculata accessions. Using BI, the Murraya asiatica accessions from Việt Nam were sister to those of Murraya exotica whilst they were unresolved using MP.  For the combination of the six chloroplast regions, using MP the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions formed one clade whereas with BI three clades, comprising Murraya ovatifoliolata and

Murraya paniculata, Murraya asiatica from China, and Murraya asiatica from Việt Nam, respectively, were resolved.  For the combination of chloroplast and ITS sequences, greater resolution was achieved with BI analysis compared to MP, with respect to the position Murraya asiatica accession 88—VCP from Việt Nam, because it was resolved (albeit weakly) as the sister group of a clade of Murraya asiatica accessions from China, while it was left unresolved in a polytomy by MP. On the other hand, MP resolved Murraya ovatifoliolata var. ovatifoliolata as monophyletic, unlike BI, which left the two subclades of Murraya ovatifoliolata var. ovatifoliolata unresolved.

Thus, for most of the regions analysed MP produced greater resolution than did BI; however, there were no differences between the two techniques that resulted in incongruence. cpDNA is haploid, non-recombinatant and inherited maternally (Soltis et al. 1997, Hamilton et al. 2003, Zhang et al. 2005), and its rate of evolution is considered conservative (Soltis et al. 1997). However, despite its conservative mode of evolution, numerous cases of intraspecific variation have been reported (Gielly & Taberlet 1994) and cpDNA has been used extensively to infer plant phylogenies at different taxonomic levels (Gielly & Taberlet 1994) and is considered valuable for studying the genetic relationships between closely related species (Palmer 1987, Palmer et al. 1988). Variation in cpDNA can be geographically structured in some plant species (Lavin et al. 1991, Soltis et al. 1992). The noncoding regions (introns and intergenic spacers) of cpDNA have been used widely to study phylogenetic relationships at different taxonomic levels (Small et al. 1998, Morton et al. 2003, Shaw et al. 2005, 2007, Bayer et al. 2009, Morton 2009). Along with cpDNA, the internal transcribed spacer (ITS) sequence of the nuclear ribosomal RNA is considered to have sufficient variation, to be easy to sequence and to be an ideal marker for species level phylogenetic studies (den Bakker et al. 2004). It has been used extensively in phylogenetic studies at lower levels (e.g., species and interspecific relationships) (Manos 1997, Compton et al. 1998, Morton 2009). This study used 6 different non-coding regions of cpDNA (trnL-F, psbM-trnDGUC, rps16, matK-5′trnK, trnCGCA-ycf6, rps4-trnT,) and the ITS region to study the phylogenetic relationships among species of Murraya and Merrillia.

The selection of the appropriate cpDNA region for phylogenetic studies is difficult, and Shaw et al. (2005) stated that ‘phylogenetic utility of different noncoding cpDNA regions within a given taxonomic group can vary tremendously’ and these authors quoted the statement of Taberlet (1994):

‘it is not easy, for many reasons, to establish a rule for the choice of a particular region of the chloroplast genome for resolving phylogenies.’

This study used 6 chloroplast regions that have been used widely in phylogenetic studies of plants particularly in the family Rutaceae and the regions studied varied in the level of resolution of the taxa examined. Least resolution was found with rps16 where one clade that included the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions was formed. The remaining accessions, including the Murraya asiatica and Murraya × cycloopensis accessions were unresolved. Analyses of matK-5′trnK and rps4-trnT gave intermediate levels of resolution. In these analyses, all Murraya exotica accessions were separated from the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions, but the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions were poorly resolved. Analysis of the other chloroplast regions and the ITS region both separated the Murraya exotica accessions from the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions and also gave a good resolution of Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata. These results are in concordance with the statement by Shaw et al. (2005) who noted that the trnK/matK region supplies adequate information to resolve the relationships between different species but often provides little resolution at lower taxonomic levels and they suggest that this region is less suitable for infrageneric phylogenetic study than other regions. The rps16 intron is considered valuable for phylogenetic studies at the family level (Oxelman et al. 1997, Wallander & Albert. 2000). However, according to Shaw et al. (2005) usually it does not provide sufficient characters to resolve relationships below generic levels. Baker et al. (2000) and Ingram & Doyle (2003) also demonstrated that the rps16 region is not variable enough to resolve infrageneric relationships. Bayer et al. (2009), who used 9 different cpDNA sequences to study the molecular phylogeny of the Rutaceae including rps16 intron, trnD-psbM spacer, rps4-trnT, trnL-trnF regions, and 5′matK intron, also concluded that the rps16 and 5′matK introns are more suitable to determine phylogenies at family level. In this study, analysis of the matK-5′trnK region separated Murraya

95 exotica from the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions whilst, in contrast, rps16 separated the Murraya paniculata and Murraya ovatifoliolata accessions from the remainder. Given the comments of the authors above, the analysis of the matK-5′trnK and rps16 regions suggests that these two forms are reasonably distantly related.

The trnL-F region has been used widely in the phylogenetic study of the Rutaceae (de Araújo et al. 2003, Morton et al. 2003, Groppo et al. 2008, Morton 2009, Bayer et al. 2009) and the phylogenetic trees derived from trnL-F provide high bootstrap values (Groppo et al. 2008). Baker et al. (2000) (who worked on the subfamily Calamoideae (Palmae)) and Wallander & Albert (2000) (who worked on the Oleaceae) also found that the trnL-F region provides good resolution at the generic and species levels. In this current study, these genes also provided a good level of resolution between the taxa examined. However, Shaw et al. (2005) and Bayer et al. (2009) found that the rps4– trnT region had a low percentage of PICs. In contrast, in this present study, this gene provided the greatest level of variation within Murraya exotica accessions, and it also separated the Murraya paniculata and Murraya ovatifoliolata accessions from the rest of the Murraya asiatica accessions. Shaw et al. (2005) state that there is no ‘holy grail’ of non-coding regions for phylogenetic analysis as is demonstrated by the information provided by the rps4–trnT region in this study.

Phylogenetic analysis often requires the combination of data sets from different sources and certain criteria are required to test whether data sets are combinable. Bull et al. (1993) suggest that the combination of data that are not homogeneous (i.e., not sharing a single history of ancestor-descent relationships and similar frequencies of character state transformations) can lead to erroneous results. The ILD test (Farris et al. 1995) has been used to detect incongruence among individual data sets that are to be combined. This test was compared with two other tests of incongruence (Templeton test (Tempelton 1983) and the Rodrigo test (Rodrigo et al. 1993)) by Cunningham (1997) who found that the ILD test was most useful. In the ILD test, the lengths of the shortest trees for each of two data sets to be combined are obtained and the lengths added together. Following this, the data sets are combined, the data randomly partitioned into two sets equal in size to the original two data sets and the lengths of the shortest trees and their sum calculated. This partitioning is repeated to determine the distribution of

96 the sum of tree lengths. The lengths from the original data sets are then compared with this distribution (Farris et al. 1995).

In this study, as analyses of the different regions of the chloroplast genome using MrModeltest returned different models of evolution, three regions representing those with the different proposed models of evolution were subjected to ILD tests; all tests returned P values greater than 0.05 and were, therefore, suitable for combination. The chloroplast genome is normally inherited as a unit and is not subject to any kind of recombination other than inversions and translocations, both within the chloroplast genome and between it and the nuclear genome, which happen only rarely in evolutionary time. Therefore, we expected the cpDNA phylogenies to be completely congruent with one another. Non-random incongruence between different loci would indicate a systematic error in one or more assumptions of the method of analysis; however, this was not found.

The ILD test between the combined chloroplast regions and ITS for all taxa gave a P value of 0.001 indicating that the sequences were incongruent. The critical P value below which data should not be combined has been debated and Cunningham et al. (1997) suggest that it lies between 0.01 and 0.001. Therefore, in this study the chloroplast and ITS data from all accessions were not combined. However, Seelanan et al. (1997) proposed that the taxa causing the incongruence should then be removed from the data sets and the data re-examined; if, after removal of the problematic taxa, no incongruence is found then the data from the remaining taxa can be legitimately combined. This procedure was followed and when the accessions of Murraya × omphalocarpa and the Murraya × cycloopensis accession were removed from analysis, a P value of 0.03 was returned justifying the combination of the data from these remaining accessions.

3.4.1. Relationship between Murraya paniculata and Murraya exotica

As detailed in the Introduction, the taxonomic status of Murraya has been controversial for centuries and most confusion is between the two species, Murraya paniculata and Murraya exotica. With the exception of the Bayesian analysis of trnD-psbM and the MP analysis of rps16, the analyses of all other individual DNA regions and of the combination of regions indicated that Murraya exotica is a distinct species. Evidence for this separation is strong, with the posterior probabilities for the Murraya exotica

97 clade from BI being 100 for the ITS region alone, for the combined chloroplast genes and ITS plus the chloroplast genes. Bootstrap support for this separation also varied from 62 (rps4-trnT) to 100 (matK-5′trnK) from the MP analyses and with posterior probabilities from 88 (trnCGCA-ycf6) to100 (trnL-F, matK-5′trnK and rps16) for the Bayesian analyses of the individual chloroplast genes. In addition, analysis of the indel data also supported the status of Murraya exotica as a distinct species.

Within the Murraya exotica clade variation was found. Due to variation in trnL-F and trnCGCA-ycf6, the Murraya × cycloopensis accession was sister to the rest of the Murraya exotica accessions. The remainder of the structure within the Murraya exotica clade was due to variation in rps4-trnT only. Within the latter, two accessions (accession 100—CH from China and accession 67—UUCR from the USA) form a trichotomy with a clade consisting of the rest of Murraya exotica; within the latter, accession 97—CYD is sister to the rest; within the latter is a subclade of 23 Murraya exotica accessions but 21 other cultivated accessions from Australia and Brazil were unresolved at this node. Additionally, a sub-clade was formed that contains the final 23 Murraya exotica accessions. In the tree from the combined chloroplast data, the posterior probabilities from BI for the nodes within the Murraya × cycloopensis accession were between 96–100; however, the bootstrap support from MP was weak with values ranging from 60–97 (n.b., Bayesian posterior probabilities are not directly comparable with the bootstrap support from MP analysis). The sub-clade contains all individuals from Việt Nam and Indonesia. Both the unresolved group and the sub-clade also contained individuals from China. Thus, the accessions from China were dispersed throughout the Murraya exotica clade, whereas those from other countries were restricted to certain locations within the clade. This distribution of accessions would fit with a Chinese origin for Murraya exotica with individuals being dispersed from China to other parts of the world. Also, in all analyses of cpDNA sequences, as well as in the ITS sequence analyses, the commercial variety ‘Min-a-Min’ (accession 70) grouped with the Murraya exotica accessions. ‘Min-a-Min’ has very small leaves, leaflets and flowers (see Chapter 4) and, therefore, should be treated as a cultivar within the Murraya exotica group.

From the analyses of the chloroplast and ITS data, it can be seen that the Murraya exotica accessions clearly separate from the rest and these results are in agreement with those of previous studies. Using RAPD with primers OPM-16 and OPW-19,

Subandiyah et al. (2007) found that the DNA banding patterns produced support for the separation of Murraya exotica from Murraya paniculata from a limited number of accessions collected from Indonesia. Verma et al. (2009), using ISSR, DAMD and RAPD, were able to separate accessions of ‘wild’ Murraya paniculata from ‘cultivated’ Murraya paniculata collected from India; however, these authors stated that in order to confirm their result, further sequencing of chloroplast and nuclear DNA needs to be performed. The ‘wild’ and ‘cultivated’ forms studied by Verma et al. (2009) were, based on this study, most likely accessions of Murraya asiatica and Murraya exotica. This current study, using accessions sourced from geographically diverse regions, confirmed separation of Murraya exotica and Murraya paniculata (s.s.) using PCR- based genotyping methods and separation of Murraya asiatica and Murraya ovatifoliolata from Murraya paniculata.

Variation was also seen amongst the accessions that did not group with Murraya exotica. MP analysis of the chloroplast data placed these accessions within one clade. This clade subdivided into three subclades with the first comprising the Murraya asiatica accessions, the second of Murraya × omphalocarpa, Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri, and the third of the Murraya paniculata accessions. When data from the indels was included, a tripartite polytomy was formed with one clade comprising the ‘Murraya asiatica accessions, the second of Murraya paniculata, Murraya ovatifoliolata and Murraya × omphalocarpa accessions, and the third of Murraya exotica accessions. Within the second clade, Murraya ovatifoliolata var. ovatifoliolata (small and large leaflet forms), Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa accessions formed a subclade that was separate from Murraya paniculata accessions. BI of the chloroplast data resulted in the formation of a tetrapartite polytomy with Murraya asiatica accessions forming two clades, the Murraya exotica accessions a second, and Murraya paniculata, Murraya ovatifoliolata and Murraya × omphalocarpa accessions the fourth. Within the latter clade the Murraya ovatifoliolata var. ovatifoliolata (small and large leaflet forms), Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa accessions again formed a subclade separate from the Murraya paniculata accessions. Both BI and MP analysis of the ITS data resulted in the formation of two clades, with Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri forming one clade and Murraya exotica, Murraya × omphalocarpa, Murraya paniculata and Murraya asiatica accessions forming the second. This second clade was

99 formed by a tripartite polytomy with subclades, the first comprising Murraya asiatica accessions, the second Murraya paniculata accessions, and the third of the Murraya exotica and Murraya × omphalocarpa accessions. The resolution of four clades formed separately by accessions of Murraya exotica, Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata, will be discussed further in the General Discussion.

Although there was a large degree of agreement between the chloroplast data and those from the nuclear ITS region, anomalies were found. Anomalies of this sort are not uncommon. For example, Seelanan et al. (1997) found such differences among members of the Gossypieae [Malvales: Malvaceae] as well as within the genus Gossypium. Fergusson and Jansen (2002) found incongruence between restriction site data from the chloroplast genome and ITS data for species of Phlox [Ericales: Polemoniaceae], and Barber et al. (2007) showed incongruence among 23 species of Sideritis [Lamiales: Labiatae]. Further, Pfeil et al. (2002) provide evidence for speciation within the Malvaceae due to hybridisation. In this study, with respect to the chloroplast data, the Murraya × cycloopensis accession was always placed in the Murraya exotica clade, whilst using the ITS region, it grouped with Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri. The second anomaly was the placement of the three accessions of Murraya × omphalocarpa. With the chloroplast data, this variety grouped with the Murraya paniculata and Murraya ovatifoliolata accessions, whilst in the analyses using the ITS region, it grouped within the Murraya exotica clade. These anomalies are the result of well supported phylogenies from the chloroplast and ITS trees and, as such, fit into the ‘hard incongruence’ category proposed by Seelanan et al. (1997).

Wendel & Doyle (1998) suggested that phylogenetic incongruence may occur due to technical causes such insufficient data or taxon sampling. However, they also suggested that incongruence may reflect something interesting about the biology of the taxa under study and may be due to organism-level processes or gene/genome-level processes. Two reasons often thought to cause incongruence are incomplete lineage sorting, introgressive hybridisation and speciation by hybridisation. With incomplete lineage sorting, an ancestral polymorphism in a gene or haplotype that was present before a speciation event is inherited by one or both resulting lineages when speciation occurs (Galtier & Daubin 2008). The allele/haplotype causing the anomaly may have evolved independently for some time before speciation has occurred. Introgressive

100 hybridisation occurs when genetically differentiated taxa interbreed, after which extensive backcrossing occurs. The time of divergence between an incongruent allele/haplotype resulting from hybridisation and its most closely related allele/haplotype can be younger than the speciation event at which the parents of the hybrid diverged (Joly et al. 2006). Hybridisation is an important evolutionary mechanism in plants (Arnold & Hodges 1995, Arnold 1997) as interspecific hybridisation is common in plants (Raven 1980). It has been estimated that 25% of plant species hybridise (Mallet 2005) and Rieseberg et al. (1996) provide a list of ~90 species where incongruence between molecular markers is thought to be due to hybridisation and introgression. Hybridisation provides a simple explanation for the anomalous data found in this study, with the Murraya × cycloopensis accession being formed from a hybridisation event between a Murraya exotica type as the female parent and a Murraya ovatifoliolata type as the male parent whilst the accessions of Murraya × omphalocarpa resulted from hybridisation occurring between a Murraya ovatifoliolata type as the female parent and an Murraya exotica type as the male parent.

3.4.2. Relationship between Murraya and Merrillia and the monophyly of Murraya

The analysis of both the chloroplast and ITS regions of species within the Clauseneae and Aurantieae (Citreae) clearly showed that the genus Murraya is monophyletic. These analyses did not resolve the species used into clades representing these two tribes; however, the results are not incongruent with those of Bayer et al. (2009). Swingle & Reece (1967) placed the two genera, Murraya and Merrillia, of tribe Clauseneae into two different subtribes, the Clauseninae and Merrilliinae. However, they considered Merrillia to be an abnormal member of tribe Clauseneae and possibly related to ancestral forms of Murraya. However, recent studies have proposed to move Merrillia and Murraya (sect. Murraya) to the Aurantieae (Samuel et al. 2001) and further studies (de Araújo et al. 2003, Morton et al. 2003, Pfeil & Crisp 2008, Morton 2009, Bayer et al. 2009, Penjor et al. 2010) have confirmed this. The results of this study also support the division of Murraya (sl) into Murraya and Bergera as proposed by Tanaka (1929) and But et al. (1986) as the accessions of Bergera kwangsiensis, Bergera microphylla and Bergera koenigii used in this study are more closely related to each other than to the accessions of Murraya.

The results from this study also show that Merrillia caloxylon is more closely related to Murraya accessions used than to the accessions of Bergera. Merrillia caloxylon

101 grouped in the clade formed by the Murraya asiatica, Murraya paniculata and Murraya ovatifoliolata accessions in the analysis of the psbM-trnD region when analysed using BI. Using both MP and BI to analyse the trnC-ycf6 and rps4-trnT regions, Merrillia caloxylon occurred in an unresolved group together with the Murraya asiatica accessions. Merrillia caloxylon also fell in an unresolved group with the Murraya asiatica and Murraya ovatifoliolata accessions when sequences of the matK region were analysed using MP. These results suggest that Merrillia caloxylon is more closely related to Murraya asiatica than to Murraya paniculata and Murraya ovatifoliolata. Analysis of the combined chloroplast genes and the ITS regions placed Merrillia caloxylon as sister to the Murraya exotica, Murraya asiatica, Murraya paniculata and Murraya ovatifoliolata accessions. Also, when the ITS regions of Murraya and Merrillia caloxylon were analysed together with sequence data from other aurantioid species, Merrillia caloxylon was again sister to the accessions of Murraya. It could be argued that Merrillia caloxylon could be placed within the genus, Murraya, as suggested by Samuel et al. (2001). However, the morphology of the flowers and fruit (Swingle & Reece 1967) are substantially different from those of Murraya which suggests that it should remain as a single member of the genus Merrillia.

102

Chapter 4: Morphology: Derivation of Phenograms, Elliptic Fourier Descriptors and Description of Taxa

4.1. Introduction

Identification and classification of plants has until the relatively recent use of molecular techniques (Gielly & Taberlet 1994, Graham et al. 2002, Jena et al. 2009, Morton 2009) and phytochemistry (But et al. 1986, 1988, Kong et al. 1986, 1988a, b) relied on morphological and anatomical similarities and differences between plants. However, the number and nature of morphological characters used to determine differences is often inconsistent and is difficult to standardise. Thus, descriptions of Murraya (s.l.) and Merrillia have differed markedly.

Rumphius (1747) based his descriptions of Camunium vulgare and Camunium javanicum on plant habit (bush or tree), size of trunk, colour and quality of wood, colour and structure of bark, and the number and arrangement of leaflets comprising leaves, size of basal leaflets, abundance of flowers, numbers of petals, stamen morphology, and the size, shape and colour of fruits.

Linnaeus based his description of Chalcas paniculata on plant habit, brief descriptions of morphology of leaves, leaflets and flowers as described by Rumphuis (1747) and Burman (1768). He based his description of Murraea exotica on its habit as a tree, the nature of its bark, and the leaf and leaflet morphology of specimens sent to him by Koenig. Original descriptions by Rumphius, Linnaeus and the following authors were quoted in Chapter 2.

Oliver (1861), who considered Murraya paniculata to be a form of Murraya exotica, based his descriptions on the number, shape and size of leaflets, inflorescence type and flower colour.

Jack (1920), who regarded Murraya paniculata and Murraya exotica as separate species, based his descriptions on plant habit, wood, numbers of leaflets, leaflet shape, apices and margins, distances between leaflets and the rachis from which they originated, abundance, size and position of flowers, the shape and colour of fruit, numbers and seeds, and seed hairs.

Tanaka (1929), who regarded Murraya exotica as a synonym of Murraya paniculata (Chalcas paniculata) based his descriptions on plant habit, density of foliage,

103 appearance of branches, leaf type, numbers and shape of leaflets, arrangement of leaflets on rachises, flower bud size, position of inflorescences on branches, and calyx, stamen, stigma, fruit and seed morphology.

Swingle & Reece (1967), who also regarded Murraya exotica as a synonym of Murraya paniculata, used plant height, leaf and leaflet characters, numbers of leaflets per leaf, size and shape of leaflets, size and colour of flowers, inflorescence type, size of petals, numbers of stamens and locules, size and shape of fruits, and numbers of seeds in their description of the species.

Stone (1985), who regarded Murraya paniculata and Murraya exotica as separate species, included the following characters in his descriptions: plant habit; bark colour; position, and length of leaves; number, shape, size and colour of leaflets; inflorescence position; size of flowers; colour, shape and numbers of petals; numbers of stamens; anther and stigma shape; fruit shape and colour; numbers of seeds and seed coat colour.

Huang (1997) based his descriptions of Murraya paniculata and Murraya exotica on: plant habit; height; bark colour; development of leaves and leaflets; number and shape of leaflets; petiole length; inflorescence position; flower colour and numbers; flower morphology, including sepals, petals, stamen and ovary; fruit colour, shape and dimensions; nature of pulp and number and hairiness of seeds.

Zhang & Hartley (in Zhang et al. 2008), who also regarded Murraya paniculata and Murraya exotica as separate species, based their views on habit, height, bark colour, number and shape of leaflets, leaflet length, petiolule length, position of inflorescences, sepal shape and length, petal colour, shape and length, number of stamens, colour shape and length of fruit, and seed characteristics.

Full descriptions of the Murraya paniculata and Murraya exotica by the above authors were presented in Chapter 2.

The original description of Merrillia caloxylon (as Murraya caloxylon) by Ridley (1908) was based on habit, length of leaves and number of leaflets, rachis morphology, leaflet shape and size, flower colour, numbers, position, sepal size and shape, ovary characters, fruit shape, size, thickness of the pericarp, and internal structure, numbers of seeds, and the location, size, colour and lengths of seeds. Swingle (1918) based his detailed description on characteristics of the leaves and fruit. Full descriptions of Merrillia caloxylon by Ridley (1908) and Swingle (1918) were presented in Chapter 2.

In studies reported in this chapter, I focused on leaf and leaflet characters to determine their suitability for unambiguously identifying Murraya exotica and Murraya 104 paniculata accessions separated on the basis of the results of my molecular studies, as reported in Chapter 3, and the accession of Merrillia caloxylon. I analysed morphological data using analysis of variance and discriminant function analysis (DFA) to determine if differences between parameters were statistically significant, UPGMA (Unweighted Pair Group Method with Arithmetic Mean) to construct phenograms, and principal component analysis (PCA) and redundancy analysis (RDA) based on elliptic Fourier descriptors (EFD: Kuhl & Giardina 1982) to separate taxa based on leaflet shapes. I then described accessions that grouped together in clusters.

On the basis of my molecular studies in Chapter 3 I sought, in this chapter, to test the hypothesis that morphological differences should also separate the taxa.

4.2. Materials and methods

4.2.1. Plant and tissue samples

Samples of leaves and in some instances flowers and fruit were obtained from locations listed in Table 4.1, in many instances with the assistance of colleagues and collectors within Australia and overseas. These samples generally comprised pressed or preserved tissues. In some instances, measurements and descriptions were based on photographs. It was not possible to compare plants growing in the same environment: quarantine risks related to the introduction of seed-borne plant pathogens (including huanglongbing) prevented the introduction and propagation of seeds from overseas.

4.2.2. Morphological assessments

Accessions (Table 4.1) were separated into groups based on molecular results presented in Chapter 3. Most assessments for each accession were based on 10 mature leaves and their basal and terminal leaflets. Characters used to compare and describe accessions are listed in Table 4.2. Characters used to compare and describe flowers, fruit and seeds are listed in Table 4.3. Where possible, 10 flowers and 10 fruit of each accession were also examined, but flowers and fruit were difficult to obtain due to seasonal limitations and in some instances, with respect to fruit, to sterile plants. Hairiness of stems, petioles, rachises, petiolules and leaflets were compared after the derivation of the phenograms. The terminology used to describe these characters was based on Hewson (1988).

105

Table 4.1. List of 47 Murraya and 1 Merrillia accessions used for morphological studies and the countries and localities from which they were sourced. The following abbreviations were used: Murraya (M); Merrillia (Me); Australia (A), New South Wales (ANSW), Queensland (AQLD), Northern Territory (ANT); Brazil (B), São Paulo (BSP); China (C), Guangxi (GX), Yingde (YD); Indonesia (I), Papua (IP), West Java (IWJ), East Java (IEJ), Central Java (ICJ), Nusa Tenggara Timur (INTT); Taiwan (T); United States of America (U), California, Riverside (UCR), Florida (UF); Việt Nam (V), Tien Giang (VTG), Cuc Phuong National Park(VCP); Royal Botanic Garden, Sydney (RBGS), Tondoon Botanic Gardens (TBG), Bogor Botanic Garden (BBG), Purwodadi Botanic Garden (PBG), Fairchild Botanic Garden (FBG). Voucher numbers are given for pressed specimens lodged at the Royal Botanic Garden, Sydney. Measurements of specimens with no voucher number were based on fresh specimens or photographs of pressed specimens on herbarium sheets.

Accession Source Latitude Longitude Name of accession based on results in this Voucher Number chapter and in Chapter 3 Number Australia 2—ANSW Richmond, NSW 33°37'S 150°45'E M. exotica 822701 4—ANSW Richmond, NSW 33°37'S 150°45'E M. exotica 822702 6—ANSW Windsor, NSW 33°37'S 150°49'E M. exotica 822703 8—ANSW RBGS, NSW 33°52'S 151°13'E M. exotica 10—ANSW Richmond, NSW 33°36'S 150°46'E M. exotica 822704 13—AQLD Brisbane, QLD 27°28'S 152°58'E M. exotica 822705 14—AQLD Brisbane, QLD 27°27'S 152°59'E M. exotica 822706 53—ANSW RBGS, NSW 33°52'S 151°13' E M. exotica 822707 54—AQLD Bundaberg, QLD 24°51'S, 152°24'E M. ovatifoliolata var. ovatifoliolata small leaflet 69—ANT Haddon Head Beach, Blue Mud Bay, NT 13°22'S 135°43'E M. ovatifoliolata var. ovatifoliolata small leaflet 822733 70—ANT Darwin, NT 12°27'S 130°50'E M. exotica 71—ANT Gove, NT 12°11'S 136°43'E M. ovatifoliolata var. ovatifoliolata small leaflet 72—AQLD Mt Carbine, QLD 16°31'S 145°09'E M. ovatifoliolata var. ovatifoliolata small leaflet 822734 73—AQLD Cooktown‐Mt Webb National Park, QLD 15°04'S 145°07'E M. ovatifoliolata var. ovatifoliolata large leaflet 822730 74—AQLD Battle Camp, QLD 15°17'S 144°43'E M. ovatifoliolata var. ovatifoliolata small leaflet 822735 75—AQLD Cairns, QLD 16°52'S 145°40'E M. ovatifoliolata var. ovatifoliolata large leaflet 822731 108—ANSW Richmond, NSW 33°37'S 150°45'E M. exotica 110—ANSW University of Sydney, NSW 33°53'S 151°11'E M. exotica 822708 115—AQLD TBG, Gladstone, QLD (via RBGS) 23°53'S 151°15'E M. ovatifoliolata var. ovatifoliolata small leaf 822736 Brazil 102—BSP Capão Bonito, SP 24°00'S 48°20'W M. exotica 104—BSP Botucatu, SP 22°53'S 48°27'W M. exotica 106—BSP Araraquata, SP 22°53'S 48°10'W M. exotica

106

Accession Source Latitude Longitude Name of accession based on results in this Voucher Number chapter and in Chapter 3 Number China 95—CYD Pipashan, Yingde County, GD 24°18′N 113°21′E M. asiatica 822740 Indonesia 22—IWJ Bogor Botanic Garden, WJ 06°36'S 106°48'E M. paniculata 822723 23—IWJ Bogor Botanic Garden, WJ 06°36'S 106°48'E Me. caloxylon 822747 24—IP Pegunungan Cycloop, Papua (via Bogor Botanic Garden) 02°30'S 140°31'E M. × cycloopensis 822746 25—IWJ Bogor Botanic Garden, WJ 06°36'S 106°48'E M. paniculata 822724 27—IWJ Bogor Botanic Garden, WJ 06°36'S 106°48'E M. exotica 822709 28—IWJ Bogor Botanic Garden, WJ 06°36'S 106°48'E M. exotica 822710 34—IEJ Purwodadi Botanic Garden, EJ 07°48'S 112°44'E M. paniculata 822725 35—IEJ Purwodadi Botanic Garden, EJ 07°48'S 112°44'E M. exotica 822711 37—IEJ Purwodadi Botanic Garden, EJ 07°48'S 112°44'E M. exotica 822712 38—IEJ Purwodadi Botanic Garden, EJ 07°48'S 112°44'E M. paniculata 822726 40—IC Bayan, Purworejo, CJ (from China) 07°43'S 109°56'E M. exotica 822713 42—IUCR Bayan, Purworejo, CJ (from USA) 07°43'S 109°56'E M. exotica 822714 44—ICJ Bayan, Purworejo, CJ 07°43'S 109°56'E M. exotica 822715 45—ICJ Bayan, Purworejo, CJ 07°43'S 109°56'E M. paniculata 822727 46—ICJ Yogyakata, CJ 07°44'S 110°25'E M. paniculata 822728 47—ICJ Universitas Gadjah Mada, Yogyakata, CJ 07°44'S 110°25'E M. exotica 822716 48—ICJ Universitas Gadjah Mada, Yogyakata, CJ 07°46'S 110°22'E M. paniculata 822729 113—INTT Kupang, NTT 10°12'S 123°36'E M. ovatifoliolata var. zollingeri 822737 114—INTT Kupang, NTT 10°12'S 123°36'E M. ovatifoliolata var. zollingeri 822738 Taiwan 91—T Orchid Island, Taiwan 22°02'N 121°32'E M. × omphalocarpa 822743 92—T Orchid Island, Taiwan 22°02'N 121°32'E M. × omphalocarpa 822744 93—T Orchid Island, Taiwan 22°02'N 121°32'E M. × omphalocarpa 822745 United States of America 111—UFBG Fairchild Botanic Garden, Florida 25°40'N 80°16'W M. exotica 822721 112—UFBG Fairchild Botanic Garden, Florida 25°40'N 80°16'W M. exotica 822722 Việt Nam 88—VCP Cuc Phuong National Park, Ninh Binh 20°15'N 105°42'E M. asiatica 822741 107

Table 4.2. Morphological characters of leaves used to distinguish forms of Murraya, and Murraya from Merrillia. Characters highlighted in bold were used for statistical analyses: the remainder were used in descriptions of taxa.

Tissue Characters assessed Leaves length (mm) number of leaflets range of number of leaflets ratio of leaf length to number of leaflets* Petioles length (mm) hairs: pubescent or glabrous Rachis length (mm) hairs: pubescent or glabrous Rachis‐petiolule distances between rachis/petiolule junctions of basal and second leaflets (mm)* junctions distances between rachis/petiolule junctions of penultimate and terminal leaflets (mm)* Petiolules petiolule length of basal leaflet (mm)* petiolule length of terminal leaflet (mm)* hairs: pubescent or glabrous Basal leaflets shape of base: rounded, cuneate or obtuse symmetric base or asymmetric base shape of apex: acute, acuminate, obtuse or rounded shape of apex: not emarginated or emarginated shape of leaflets: ovate, obovate, or elliptic length of leaflets (mm) width of leaflets (mm) ratio between length and width* length of tip (apex) (mm)* ratio of length of leaflet tip to leaflet length* margins: entire, crenate or crenulated; ciliate angle of base (degrees)* hairs: pubescent or glabrous lamina and veins on upper and lower surfaces Terminal leaflets shape of base : rounded, cuneate or obtuse symmetric base or asymmetric base shape of apex: acute, acuminate, obtuse or rounded shape of apex: not emarginated or emarginated shape of leaflets: ovate, obovate, or elliptic length of leaflets (mm) width of leaflets (mm) ratio between length and width* length of tip (apex) (mm)* ratio of length of leaflet tip to leaflet length* margins: entire, crenate or crenulated; ciliate angle of base(degrees)* hairs: pubescent or glabrous lamina and veins on upper and lower surfaces * Characters not used in previous studies. Additionally, the extent of variation in pubescence between taxa has been previously under‐reported.

108

Table 4.3. Morphological characters used to compare flowers, fruits and seeds of Murraya paniculata and Murraya exotica. These characters were used for descriptive purposes only for the accessions for which flowers and fruit were obtained.

Tissue Characters assessed Flowers number of flowers/inflorescence length of pedicels (mm) petal length (mm) petal width (mm) shape of petal length of pistil (mm) length of stamens (mm) length of stigmas (mm) length of ovary (mm) Fruit fruit shape fruit length (mm) fruit width (mm) number of seeds seedcoat hairs seedcoat colour seed shape seed length (mm) seed width (mm)

Angles formed by leaf margins at the base of a leaflet (angle ABC) and lengths of tips of leaflets were determined as illustrated in Fig. 4.1. The length (EG) of each tip (apex) was measured as the distance between the intersection (E) of lines (DE & EF) corresponding to the leaflet margins towards the apex and the end of the tip (G) as indicated in Fig. 4.1.

Figure 4.1. Diagrammatic representation of how angles between basal margins of leaflets, the lengths of leaflet tips (in this instance an acuminate tip of accession 22) were measured.

109

4.2.3. Statistical analyses

Data obtained from morphological characters assessed for each accession were analysed using analysis of variance (Statistica, Version 9.1, StatSoft, Inc. (2010)). Data involving ratios were log transformed before analysis with the exception of ratios with zero values: ratios with zero values were not transformed. Before analysis, homogeneity of variances was assessed using Levene’s test. Where the variances were not homogenous, heteroscedasticity was removed or reduced by using square root or logarithmic transformations. Means were separated using Fisher’s LSD test at P = 0.05.

4.2.4. Phenograms

Phenograms were constructed using Unweighted Pair Group Method with Arithmetic Mean (UPGMA) (http://genomes.urv.es/UPGMA/) and PAUP* 4.0b10 (Swofford 2002) based on 12 characters of basal and terminal leaflets as follows:  lengths of basal leaflets (excluding petiolules);  widths of basal leaflets;  ratios between lengths and widths of basal leaflets;  tip (apex) lengths of basal leaflets;  ratios of lengths of leaflet tips (apices) of basal leaflets and lengths of the leaflets (excluding petiolules);  angles formed by leaf margins at bases of basal leaflets;  lengths of terminal leaflets (excluding petiolules);  widths of terminal leaflets;  ratios between lengths and widths of terminal leaflets;  tip lengths of terminal leaflets (excluding petiolules);  ratios of lengths of leaflet tips of terminal leaflets and the lengths of leaflets; and  angles formed by leaf margins at bases of terminal leaflets.

To construct the phenogram using PAUP*, the data were coded using MorphoCode (Schols et al. 2004), then the coded data were subjected to maximum parsimony (MP) analysis. MorphoCode implements a variation on Thiele's (Thiele 1993) method of differential weighting of gaps between coded states within one character.

110

4.2.4. Principal component analysis, redundancy analysis and discriminant function analysis for separating taxa on the basis of quantitative characters of leaflets

Data of six quantitative characters of basal and terminal leaflets (length, width, ratio of length to width, tip length, ratio of tip length to leaflet length, angle of base) of accessions, identified molecularly in Chapter 3 as Murraya paniculata, Murraya exotica, Murraya asiatica, Murraya ovatifoliolata var. ovatifoliolata (small and large leaftet forms), Murraya ovatifoliolata var. zollingeri, and Murraya × omphalocarpa, were subjected to principal component analysis and redundancy analysis using CANOCO Version 4.5 (ter Braak & Šmilauer 2002). Combined data of basal and terminal leaflets were also analysed by principal component analysis and reduncancy analysis and the significance of models derived from reduncancy analysis were tested with Monte Carlo tests based on 999 permutations.

Data of four quantitative characters of basal and terminal leaflets (length, width, tip length, angle of base) of each accession of Murraya paniculata, Murraya exotica, Murraya asiatica, Murraya ovatifoliolata var. ovatifoliolata (small and large leaftet forms), Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa were subjected to discriminant function analysis (Statistica, Version 9.1, StatSoft Inc. (2010)). The analysis was repeated with data from Murraya × omphalocarpa and with the data from Murraya ovatifoliolata var. ovatifoliolata (small and large leaftet forms) and Murraya ovatifoliolata var. zollingeri combined as a single taxon. Before analysis, the data were subjected to logarithmic or square root transformations to remove or reduce heteroscedasticity. For discriminant function analysis, variables were added in a forward stepwise manner with the tolerance and the ‘F to enter’ both set to 0.01. As suggested by Valcárcel and Vargas (2010), the derived leaflet characters (ratios) were excluded from this analysis.

4.2.5. Principal component analysis and redundancy analysis of elliptic Fourier descriptors of leaflet shapes

Elliptic Fourier descriptors of leaflet shapes were obtained using the image analysis (ChainCoder) and elliptic Fourier transformation (Chc2Nef) programs of the SHAPE package software (Iwata & Ukai 2002). Photographs or photocopies of five basal leaflets and five terminal leaflets of each accession of Murraya paniculata, Murraya exotica, Murraya asiatica, Murraya ovatifoliolata var. ovatifoliolata (small and large leaftet forms) and Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa, 111 were transferred to bitmap images and analysed by using ChainCoder. ChainCoder extracts the contours (margins) of the leaves from digital images and stores the relevant information as chain code (Freeman 1974). The chain code data sets were then subjected to analysis using Chc2Nef. This program calculates normalised elliptic Fourier descriptors from the chain code information. The normalised elliptic Fourier descriptors derived by Chc2Nef were then subjected to principal component analysis using PrinComp from the SHAPE package to calculate and draw the mean (± 2 std. dev) shapes of leaves. The data were also subjected to principal component analysis and redundancy analysis using CANOCO Version 4.5 (ter Braak & Šmilauer 2002). To test the relationship between the elliptical Fourier descriptors and species, the species to which an accession belonged was entered into CANOCO as dummy binary variables. For each species, the dummy variables for that species were used as the explanatory variables with the other variables entered as covariables. The significance of the variability in the response variables described by the explanatory variables was then tested using Monte Carlo tests based on 999 permutations (Lepš & Šmilauer 2003).

4.2.6. Morphological descriptions of putative taxa

Description of the plants focused mostly on leaf and basal and terminal leaflet characters listed in Table 4.2 and, in some instances, on flowers and fruits (Table 4.3). These character sets were based on characters used by authors from Rumphius (1747) to Zhang et al. (2008) (see Chapter 2) and other characters not used previously (see Table 4.2). Each quantitative character was expressed as the mean value (with the range in brackets).

4.3. Results 4.3.1. Morphological differences between Murraya accessions

On the basis of results presented in Chapter 3, the accessions were separated into groups that conformed to:  Murraya paniculata, Murraya asiatica and Murraya exotica,  Murraya ovatifoliolata var. ovatifoliolata (large leaflet) and Murraya ovatifoliolata var. ovatifoliolata (small leaflet), and  Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa.

Box and whisker plots of the leaf characteristics are presented in Figs 4.2 and 4.3.

112

Murraya paniculata and Murraya asiatica had larger leaves with fewer leaflets, almost all of which had acuminate apices, whereas Murraya exotica had smaller leaves with more leaflets that had a variety of apex shapes (Table 4.4 and 4.5). The ratio of leaf length to the number of leaflets was large (~29) for the leaves of Murraya paniculata and Murraya asiatica and about half this number (~14) for the leaves of Murraya exotica (Table 4.5).

Leaflet shapes also differed between Murraya paniculata and Murraya asiatica, and Murraya exotica (Table 4.4). Basal leaflets of Murraya paniculata and Murraya asiatica were ovate to elliptic, and leaflet blades were asymmetrically broad at their bases (obtuse or rounded base) and acuminate at the apices. The same traits were evident for Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri. In contrast, basal leaflets of Murraya exotica were obovate to elliptic, and the bases asymmetrically narrow (cuneate) and apices broad (acute). The apices of basal leaflets of Murraya paniculata and Murraya asiatica were longer and more acute or acuminate than the basal leaflet apices of Murraya exotica, which varied, being asymmetric and acute, acuminate, obtuse or rounded.

Terminal leaflets of Murraya paniculata, Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri were ovate or elliptic in contrast to the obovate or elliptic terminal leaflets of Murraya exotica. The apices of Murraya paniculata and Murraya asiatica terminal leaflets were acuminate and longer than the short acute, acuminate, obtuse or rounded apices of Murraya exotica.

Basal and terminal leaflet shapes of Murraya × omphalocarpa resembled those of Murraya exotica.

The results of the statistical analyses of morphological characters of the plants showed that Murraya paniculata and Murraya asiatica had significantly longer leaves with fewer leaflets (3–7 leaflets) than Murraya exotica (4–10 leaflets), that the ratio of these values differed markedly, and that numbers of leaflets, petiole lengths, distances between the petiolule-rachis junctions of the basal and second leaflets, and distances between the petiolule-rachis junctions of the penultimate and terminal leaflets, differed significantly between the two (Table 4.5). Leaves, petioles, petiolules and distances between petiolule-rachis junctions of Murraya paniculata and Murraya asiatica were significantly longer than those of Murraya exotica.

113

Lengths and widths of basal and terminal leaflets of Murraya paniculata and Murraya asiatica were significantly greater than those of Murraya exotica. Angles of leaflet bases and petiolule lengths of basal and terminal leaflets of Murraya paniculata and Murraya asiatica and Murraya exotica differed significantly. Lengths of tips of leaflets also differed significantly (Tables 4.6 and 4.7).

Pedicels and fruits of Murraya paniculata were significantly longer than those of Murraya exotica (Table 4.8).

There were marked differences between taxa for the presence and distribution of hairs on stems, petioles, rachises, petiolules, laminas and leaflet veins (Table 4.4), the most noticeable, particularly for distinguishing between Murraya paniculata and Murraya asiatica, and Murraya exotica, being:  upper surfaces of petioles, rachises, petiolules and leaflet midveins of Murraya paniculata and Murraya asiatica were pubescent, the lower surfaces glabrous;  upper and lower surfaces of petioles, rachises, petiolules and leaflet midveins of Murraya exotica were pubescent.

Also noticeable were the differences between the small and large leaflet forms of Murraya ovatifoliolata var. ovatifoliolata (Table 4.4).

The hairiest were the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata and the Murraya × cycloopensis from Papua (Table 4.4).

114

Figure 4.2. Box and whisker plots of the characters of basal leaflets used in the analysis of leaf morphology. Left to right in each plot: M. exotica, M. paniculata, M. asiatica, M. ovatifoliolata var. ovatifoliolata ‘sl’, M. ovatifoliolata var. ovatifoliolata ‘ll’, M. × omphalocarpa, M. ovatifoliolata var. zollingeri.

115

Figure 4.3. Box and whisker plots of the characters of terminal leaflets used in the analysis of leaf morphology. Left to right for each plot: M. exotica, M. paniculata, M. asiatica, M. ovatifoliolata var. ovatifoliolata ‘sl’, M. ovatifoliolata var. ovatifoliolata ‘ll’, M. × omphalocarpa, M. ovatifoliolata var. zollingeri.

116

Table 4.4. Comparison of some qualitative characters of basal and terminal leaflets of Murraya species, varieties and hybrids.

Taxon Shape of base of Shape of apex of basal Shape of Shape of base of Shape of apex of Shape of basal leaflet leaflet basal leaflet terminal leaflet terminal leaflet terminal leaflet

M. paniculata mostly obtuse to acute to acuminate (> 90% ovate, elliptic cuneate, symmetric acuminate (> 90% elliptic rounded, asymmetric; acuminate) or asymmetric acuminate) angle of base > 90°

young green and recently mature stems mostly glabrous, occasionally sparsely pubescent; petioles, rachises, petiolules pubescent, sometimes glabrous above; mostly glabrous, sometimes sparsely or scattered pubescent below; terminal and basal leaflet midveins and lateral veins mostly glabrous, sometimes pubescent above and below; lamina glabrous above and below; margins glabrous

M. asiatica cuneate to rounded, acute to acuminate (> 90% ovate, elliptic cuneate, symmetric acuminate (> 90% elliptic asymmetric; angle of acuminate) or asymmetric acuminate) base < 90°

young green and recently mature stems mostly glabrous, occasionally sparsely pubescent; petioles, rachises, petiolules pubescent, sometimes glabrous above; mostly glabrous, sometimes sparsely or scattered pubescent below; terminal and basal leaflet midveins and lateral veins mostly glabrous; lamina glabrous above and below; margins glabrous

M. ovatifoliolata var. obtuse, rounded, acute, acuminate, obtuse, ovate, elliptic cuneate, obtuse, acute, acuminate (~ 50% ovate, elliptic ovatifoliolata (large symmetric and rounded (~ 50% acuminate) rounded, symmetric acuminate) leaflet) asymmetric

young green and recently mature stems glabrous; petioles and rachises glabrous above and below, petiolules pubescent to scattered pubescent above and below; terminal leaflet midveins pubescent to scattered pubescent above, glabrous below, lateral veins glabrous above and below; basal leaflet midveins and lateral veins pubescent proximally, glabrous elsewhere, above and below; lamina glabrous above and below; margins glabrous except near petiolule

M. ovatifoliolata var. obtuse, rounded, acute, acuminate, obtuse, ovate, elliptic cuneate, obtuse, acute, acuminate (~ 50% ovate, elliptic, ovatifoliolata (small symmetric and rounded (~ 50% acuminate) rounded, symmetric acuminate) rarely obovate leaflet) asymmetric and asymmetric

young and recently mature stems pubescent; petioles, rachises, petiolules pubescent above and below; terminal leaflet midveins and lateral veins pubescent above, pubescent to sparsely pubescent below; basal leaflet midveins and lateral veins pubescent above and below; lamina pubescent to sparsely pubescent above and below; margins ciliate

117

Taxon Shape of base of Shape of apex of basal Shape of Shape of base of Shape of apex of Shape of basal leaflet leaflet basal leaflet terminal leaflet terminal leaflet terminal leaflet

M. ovatifoliolata var. obtuse, rounded, acute, obtuse (0% ovate, elliptic cuneate, obtuse, acute, obtuse (0% elliptic, zollingeri asymmetric acuminate) symmetric and acuminate) sometimes ovate asymmetric

young greens stems pubescent, recently matured stems sparsely pubescent; petioles, rachises and petiolules pubescent above and below; terminal leaflet midveins and lateral veins pubescent above, midveins pubescent, lateral veins glabrous, sometimes sparsely pubescent below; basal leaflet midveins pubescent, lateral veins glabrous above and below, lateral veins glabrous sometimes sparsely pubescent below; terminal and basal lamina scattered to sparsely pubescent, more towards base, above and below; margins ciliate

M. exotica cuneate, asymmetric acute, acuminate, obtuse, obovate, cuneate, symmetric acute, acuminate, obtuse, obovate, elliptic rounded (~ 30% acuminate) elliptic and asymmetric rounded (~ 30% acuminate)

young green and recently mature stems pubescent; petioles, rachises and petiolules pubescent above and below; terminal leaflet midveins pubescent above and below, lateral veins glabrous above, pubescent to sparsely pubescent below, less so proximally; basal leaflet midveins pubescent above, lateral midveins sparsely pubescent, less so proximally; terminal leaflet lamina mostly glabrous above, pubescent to sparsely pubescent below, basal leaflet lamina pubescent to sparsely pubescent above and below; margins ciliate

M. × omphalocarpa obtuse, rounded, acute, acuminate, obtuse, obovate, cuneate, symmetric acute, acuminate (~ 50% obovate, elliptic mostly asymmetric rounded (~ 40% acuminate) elliptic acuminate)

young green and recently mature stems glabrous; petioles, rachises and petiolules pubescent above, glabrous below; terminal and basal leaflet midveins and lateral veins glabrous above, midveins sometimes glabrous and lateral veins with dispersed pubescent tufts below; laminas glabrous above and below; margins, glabrous, sometimes sparsely ciliate Murraya × rounded, asymmetric acute to acuminate (> 90% ovate obtuse, seldom acute to acuminate (> elliptic cycloopensis acuminate) broadly cuneate, 90% acuminate) symmetric and asymmetric stems pubescent, tufted on residual juvenile epidermis of old stems; petioles, rachises, petiolules pubescent above and below; terminal midveins veins pubescent above and below, lateral veins glabrous to sparsely pubescent above and below; basal leaflet midveins and lateral veins pubescent, above and below; terminal and basal leaflet laminas sparsely pubescent above and below; margins ciliate

118

Table 4.5. The comparison of some quantitative characters of leaves of Murraya species, varieties and hybrids.

Taxon Average number of Leaf length Ratio of leaf Rachis Distance between Distance between Petiole length leaflets (range)** (mm) length/number length petiolule‐rachis petiolule‐rachis junctions (mm)* of leaflets* (mm) junctions of first and of penultimate and

second leaflets (mm) terminal leaflets (mm)*

M. paniculata 5.21b 145.41a 28.02a 53.56a 5.91b 15.97a 12.62bcd (4–7) (72–180) (18.0–36.0) (15–78) (0–16) (3–30) (6.0–20.0) M. asiatica 4.62b 134.18a 30.15a 44.51a 10.50a 11.56ab 17.87ab (3–6) (88–192) (22–43) (25–75) (0–25) (0–22) (4.6–30.8 M. ovatifoliolata var. ovatifoliolata 5.81ab 129.77ab 24.51b 58.86a 7.46ab 17.26a 11.24cde (large leaflet) (3–8) (88–165) (14.7–39.0) (30–88) (2.1–18) (9–29) (4–15)

M. ovatifoliolata var. ovatifoliolata 6.09ab 93.09b 15.48cd 43.7a 4.58bc 10.64b 9.17e (small leaflet) (3–9) (48–168) (9.6–26.0) (10–116) (0–13) (0–27) (3–18)

M. ovatifoliolata var. zollingeri 7.40a 127.10ab 16.77c 74.38a 4.87abc 10.87abc 19.77a (7–9) (59–250) (8.4–31.3) (50–190) (0–14) (0–27) (8.3–40) M. exotica 6.96a 97.87b 14.3d 44.44a 3.61c 7.08c 10.39de (4–10) (50–170) (9.4–21) (9–94) (0–18) (0–20) (4–21) M. × omphalocarpa 6.77a 131.96a 17.57c 58.06a 4.39abc 12.65ab 16.62abc (4–9) (75–200) (10.6–50.0) (40–85) (0–13.1) (2–24) (6–30) Murraya × cycloopensis 7 159.9 23.00 77.6 8.5 15.1 23.3 (5–8) (140–185) (20.9–28.0) (53–100) (3–20) (8–23) (20–27)

ANOVA Df 6, 31 Df 6, 31 DF 6, 301 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 (excluding Murraya × F = 3.74 F = 4.72 F = 72.76 F = 1.86 F = 4.68 F = 6.73 F = 6.83 cycloopensis) P = 0.00647 P = 0.00160 P < 0.001 P = 0.11968 P = 0.00169 P = 0.000123 P = 0.00010 * Back‐transformed from data log transformed for statistical analyses. ** Back‐transformed from data square root transformed for statistical analyses. Means followed by the same letter(s) within columns are not significantly different according to Fisher’s LSD test at α = 0.05. 119

Table 4.6. The comparison of some quantitative characters of basal leaflets of Murraya species, varieties and hybrids.

Taxon Leaflet length Leaflet width Ratio of Petiolule Tip length Proportion of Angle of base (mm)* (mm)* length/width* length (mm)* (mm)** acuminate leaflet tip to (degrees) leaflet lengths*

M. paniculata 42.08a 23a 1.8ab 3.3ab 3.4a 0.08a 103b (n=41) (25–59) (15–32) (1.4–2.0) (1.6–5.1) (0.0–7.3) (0.0–0.15) (77–135) M. asiatica (n=16) 48.29a 23.56a 2.0a 3.9a 3.6a 0.08a 83cd (28–75) (14–35) (1.4–2.5) (2.0–7.5) (1.1–6.5) (0.03–0.13) (67–95)

M. ovatifoliolata var. 31.44ab 22ab 1.5cd 2.0cd 0b 0b 120ab ovatifoliolata (large leaflet) (n=17) (14–47) (9–40) (1.2–1.8) (0.8–2.67) (90–170)

M. ovatifoliolata var. 18c 14c 1.3d 1.9d 0b 0b 123a ovatifoliolata (small leaflet) (8–40) (6–28) (1.0–2.2) (0.6–3.7) (74–150) (n=48) M. ovatifoliolata var. zollingeri 24bc 15bc 1.6c 2.3bcd 0b 0b 106ab (n=15) (13–39) (10–25) 1.3–2.1) (1.29–4.68) (90–125)

M. exotica 22b 14c 1.6c 2.3cd 0.1b 0b 72d (n=149) (13–38) (8–24) (1.0–2.4) (0.6–3.96) (0.0–3.1) (0.0–0.08) (47–97)

M. × omphalocarpa 28b 19abc 1.4cd 2.9abc 0.6b 0.01b 102bc (n=20) (12–60) (9–38) (1.2–1.7) (1.27–4.93) (0.0–3.0) (0.0–0.05) (75–142)

Murraya × cycloopensis 35 23 1.5 2.6 1.2 0.04 130 (n=10) (28–42) (18–27) (1.3–1.8) (2.0–3.1) (0.0–3.0) (0.0–0.10) (118–145) ANOVA (excluding Murraya × Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 cycloopensis) F = 9.756 F = 5.097P= F = 9.01 F = 4.434 F = 30.35 F = 66.00 F = 22.60 P < 0.0001 0.00096 P < 0.0001 P = 0.00239 P < 0.0001 P < 0.0001 P < 0.0001 * Back‐transformed from data log transformed for statistical analyses. ** Back‐transformed from data square root transformed for statistical analyses. Means followed by the same letter(s) within columns are not significantly different according to Fisher’s LSD test at α = 0.05. Values rounded to nearest whole number for values ≥ 10 and to first decimal point for values < 10 with the exception of ratio of tip/leaflet lengths measurements that were rounded to the second decimal place. 120

Table 4.7. The comparison of some quantitative characters of terminal leaflets of Murraya species, varieties and hybrids.

Taxon Leaflet length Leaflet width Ratio of Petiolule length Tip length Proportion of Angle of base (mm)* (mm) * length/width * (mm)* (mm)** acuminate leaflet tip (degrees)* to leaflet lengths* M. paniculata 76a 31a 2.5a 3.8a 6.7a 0.09a 72b (50–91) (22–43) (1.8–3.3) (1.6–6.4) (2.6–12.3) (0.04–0.14) (48–98)

M. asiatica 68ab 28a 2.4ab 3.6ab 6.9a 0.11a 65b (45–98) (20–37) (1.8–3.0) (2.4–5.7) (4–13.9) (0.06–0.15) (56–90)

M. ovatifoliolata var. 59ab 36a 1.7cd 2.1cd 0.3b 0.01b 98a ovatifoliolata (large leaflet) (39–85) (24–60) (1.3–2.2) (1–3.44) (0.0–4.4) (0.0–0.07) (62–134)

M. ovatifoliolata var. 37d 25a 1.5d 2.5bcd 0.4b 0.01b 99a ovatifoliolata (small leaflet) (22–56) (14–45) (1.0–2.4) (1.2–4.5) (0.0–2.7) (0.0–0.09) (63–130)

M. ovatifoliolata var. 46bcd 23ab 2.0bc 2.6abcd 0b 0b 81ab zollingeri (32–65) (18–33) (1.6–2.7) (2–4.16) (53–110)

M. exotica 41cd 19b 2.2b 2.0d 0.4b 0.01b 52c (27–70) (11–29) (1.5–3.1) (0.6–3.96) (0.0–5.7) (0.0–0.08) (30–75)

M.. × omphalocarpa 53bc 30a 1.9c 3.3abc 1.6b 0.02b 69b (23–95) (10–61) (1.5–2.5) (1.75–6.74) (0.0–5.8) (0.0–0.08) (53–90)

Murraya × cycloopensis 58 32 1.8 3.1 2.8 0.05 100 (54–65) (25–38) (1.6–2.2) (2.1–4.03) (0.0–6.0) (0.0–0.08) (80–117)

ANOVA result (excluding Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Df 6, 31 Murraya × cycloopensis) F = 9.740 F = 4.964 F = 9.945 F = 6.091 F = 16.822 F = 24.567 F = 15.08 P < 0.0001 P= 0.00115 P < 0.0001 P = 0.00026 P < 0.0001 P < 0.0001 P < 0.0001 * Back‐transformed from data log transformed for statistical analyses. ** Back‐transformed from data square root transformed for statistical analyses. Means followed by the same letter(s) within columns are not significantly different according to Fisher’s LSD test at α = 0.05. Values rounded to nearest whole number for values ≥ 10 and to first decimal point for values < 10 with the exception of ratio of tip/leaflet lengths measurements that were rounded to the second decimal place.

121

Table 4.8. The comparison of some quantitative characters of flowers and fruits of Murraya paniculata and Murraya exotica.

Taxon Pedicel length Petal length (mm) Petal width (mm) Fruit length Fruit width Seed length Seed width (mm) (mm) (mm) (mm) M. paniculata 9.1a 22a 5.3a 20a 9.1a 10a 4.94a

M. exotica 4.3b 19a 5.5a 13b 8.1a 9.58a 5.16a

ANOVA Df 1, 23 Df 1, 16 Df 1, 16 Df 1, 22 Df 1, 22 Df 1, 16 Df 1, 16 F = 77.71 F = 0.13 F = 1.92 F = 58.91 F = 1.67 F = 2.10 F = 0.39 P < 0.0001 P = 0.71852 P= 0.18491 P < 0.0001 P = 0.20909 P = 0.16650 P = 0.54207 Means followed by the same letter(s) within columns are not significantly different according to Fisher’s LSD test at α = 0.05. Values rounded to nearest whole number for values ≥ 10 and to first decimal point for values < 10 with the exception of ratio of tip/leaflet lengths measurements that were rounded to the second decimal place.

122

4.3.2. Dendrograms

A phenogram derived by analysing basal and terminal leaflet data sets with UPGMA clustering (http://genomes.urv.es/UPGMA/) is presented in Fig. 4.4. There was clear separation of Murraya exotica and the other taxa into two clusters, clusters A and B. Cluster A comprised Murraya paniculata, Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa. Merrillia caloxylon and Murraya × cycloopensis also fell, due to the size and shape of leaflets, into this cluster. This cluster also comprised two subclusters, A1 and A2. Subcluster A1 comprised Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri, Murraya × cycloopensis, Merrillia caloxylon and two of the three Murraya × omphalocarpa accessions. Subcluster A2 comprised Murraya paniculata accessions from Indonesia, Murraya asiatica accessions from China, and Việt Nam, and one of the three Murraya × omphalocarpa accession, in this instance with large leaflets. Cluster B comprised the Murraya exotica accessions.

A cladogram, derived from maximum parsimony (MP) analysis with PAUP* of gap- coded basal and terminal leaflet morphological data is presented in Fig. 4.5. However, the analysis did not resolve the accessions into clades that equated to species. A large, basal polytomy was formed that included the accessions of Murraya exotica, Murraya × omphalocarpa, Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri and Murraya × cycloopensis. However, the accessions of Murraya paniculata and Murraya asiatica formed a separate clade.

123

Figure 4.4. Phylogenetic tree based on morphological characters of leaflets of Murraya and Merrillia caloxylon accessions using UPGMA analysis. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

124

Figure 4.5. 50% majority‐rule consensus tree of the leaflet characters of Murraya and Merrillia caloxylon accessions derived from maximum parsimony analysis. Bootstrap values are provided as percents from 1000 replications. ‘sl’ small leaflet form, ‘ll’ large leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

125

4.3.3. Principal component analysis and redundancy analysis for separating taxa on the basis of quantitative characters of leaflets

Leaflet quantitative characters (length, width, ratio of length to width, tip length, ratio of tip length to leaflet length, angle of base) of Murraya paniculata, Murraya asiatica, Murraya exotica, Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri, and Murraya × omphalocarpa accessions were subjected to principal component and redundancy analysis. The analyses were based on a total of 304 basal leaflets and 304 terminal leaflets. The first two principal components (PC1 and PC2) accounted for 99.2% of the variance (77.2% and 22%, respectively) in the principal component analysis of basal leaflet characters (Fig. 4.6) and showed clear separation between Murraya paniculata, Murraya exotica and the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata. The redundancy analysis accounted for 53% of the variance in the data. The first two axes accounted for 33.6% and 16.8%, respectively. The most important explanatory variables were angle of base and ratio of tip length to leaflet length (Table 4.9). The redundancy analysis correlation biplot (Fig. 4.7) of these data showed large negative correlations between Murraya paniculata, Murraya exotica and the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata. In redundancy analysis correlation biplots, lines pointing in the same direction indicate that taxa have large positive correlations and lines pointing in opposite directions indicate that taxa have large negative correlations. Any two lines forming an angle of 90° have small or close to zero correlation. The length of each line is proportional to the variance.

Table 4.9. Conditional effects on the total sum of eigenvalues (variance) of adding each variable in the redunancy analysis of basal leaflet quantitative characters.

Order Explanatory variable Increase after F p including variable 1 angle of base 0.269 111.2 0.001 2 ratio of tip length to leaflet length 0.211 122.5 0.001 3 leaflet length 0.024 14.39 0.001 4 leaflet width 0.012 7.33 0.001 5 tip length 0.008 5.28 0.002 6 ratio of length to width 0.005 2.83 0.018 Total explained 0.529 Unexplained 0.471

The analyses also showed the difference between terminal leaflets of Murraya paniculata and Murraya exotica (Figs 4.8 and 4.9). PC1 and PC2 explained 98.3% of variance (61.6% and 36.7%, respectively). The redundancy analysis accounted for 52%

126 of the variance in the data. The first two axes accounted for 32.4% and 16.3%, respectively. The most important explanatory variables were angle of base and tip length (Table 4.10). Both analyses separated Murraya paniculata from Murraya exotica and the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata once again differed from Murraya paniculata and Murraya exotica. Similar results were obtained when the data of basal and terminal leaflets were combined (Figs 4.10 and 4.11) with PC1 and PC2 explaining 64.2% and 27.6% of the variance, respectively. The redundancy analysis accounted for 58.4% of the variance in the data. The first two axes accounted for 35.4% and 18.1%, respectively. The most important explanatory variables were angle of base of basal leaflet and tip length of the terminal leaflet (Table 4.11).

Table 4.10. The conditional effects on the total sum of eigenvalues (variance) of adding each variable in the redunancy analysis of terminal leaflet quantitative characters.

Order Explanatory variable Increase after F p including variable 1 angle of base 0.240 95.5 0.001 2 tip length 0.227 128.15 0.001 3 leaflet length 0.035 21.23 0.001 4 leaflet width 0.016 10.05 0.001 5 ratio of tip length to leaflet length 0.003 1.64 0.168 6 ratio of length to width 0.002 1.3 0.253 Total explained 0.523 Unexplained 0.477

Table 4.11. The conditional effects on the total sum of eigenvalues (variance) of adding each variable in the redunancy analysis of combined basal and terminal leaflet quantitative characters.

Order Explanatory variable Increase after F p including variable 1 angle of base of basal leaflet 0.269 111.2 0.001 2 tip length of terminal leaflet 0.214 124.27 0.001 3 length of terminal leaflet 0.036 22.08 0.001 4 angle of base of terminal leaflet 0.018 11.41 0.001 5 width of terminal leaflet 0.016 11.28 0.001 6 ratio of length to width of basal leaflet 0.010 7.97 0.001 7 ratio of tip to basal leaflet length 0.009 5.71 0.001 8 width of basal leaflet 0.003 1.85 0.119 9 length of basal leaflet 0.004 2.73 0.030 10 ratio of tip to terminal leaflet length 0.003 1.78 0.123 11 ratio of length to width of terminal leaflet 0.002 1.86 0.103 12 length of tip of basal leaflet 0.001 1.52 0.174 Total explained 0.584 Unexplained 0.416

127

Figure 4.6. Principal component analysis of 6 quantitative characters of basal leaflets of Murraya: M. asiatica , M. paniculata , M. exotica , M. ovatifoliolata var. ovatifoliolata (small leaflet) , M. ovatifoliolata var. ovatifoliolata (large leaflet) , M. ovatifoliolata var. zollingeri , M. × omphalocarpa . The PCI and PC2 axes accounted for 77.2% and 22% of variance in the data, respectively.

Figure 4.7. Redundancy analysis correlation biplot for 6 quantitative characters of basal leaflets of Murraya: lines clockwise from top, M. asiatica, M. paniculata, M. ovatifoliolata var. ovatifoliolata (large leaflet), M. ovatifoliolata var. ovatifoliolata (small leaflet), M. ovatifoliolata var. zollingeri, M. × omphalocarpa and M. exotica. Axes 1 and 2 accounted for 33.6% and 16.8% of variance in the data, respectively.

128

Figure 4.8. Principal component analysis of 6 quantitative characters of terminal leaflets of Murraya: M. asiatica , M. paniculata , M. exotica , M. ovatifoliolata var. ovatifoliolata (small leaflet) , M. ovatifoliolata var. ovatifoliolata (large leaflet) , M. ovatifoliolata var. zollingeri , M. ×. omphalocarpa . The PCI and PC2 axes accounted for 61.6% and 36.7% of variance in the data, respectively.

Figure 4.9. Redundancy analysis correlation biplot for 6 quantitative characters of terminal leaflets of Murraya: lines clockwise from top, M. asiatica, M. paniculata, M. ovatifoliolata var. ovatifoliolata (large leaflet), M. ovatifoliolata var. zollingeri, M. ovatifoliolata var. ovatifoliolata (small leaflet), M. × omphalocarpa and M. exotica. Axes 1 and 2 accounted for 32.4% and 16.3% of the variance in the data, respectively. 129

Figure 4.10. Principal component analysis of 6 quantitative characters of both basal and terminal leaflets of Murraya: M. asiatica , M. paniculata , M. exotica , M. ovatifoliolata var. ovatifoliolata (small leaflet) , M. ovatifoliolata var. ovatifoliolata (large leaflet) , M. ovatifoliolata var. zollingeri , M. ×omphalocarpa . The PCI and PC2 axes accounted for 64.2% of and 27.6% of variance in the data, respectively.

Figure 4.11. Redundancy analysis correlation biplot for 6 quantitative characters of both basal and terminal leaflets of Murraya: lines clockwise from top, M. asiatica, M. paniculata, M. ovatifoliolata var. ovatifoliolata (large leaflet), M. ovatifoliolata var. ovatifoliolata (small leaflet), M. ovatifoliolata var. zollingeri, M. × omphalocarpa and M. exotica. Axes 1 and 2 accounted for 35.4% and 18.1% of variance in the data, respectively. 130

The leaflet quantitative characters (length, width, tip length, angle of base) of the terminal and basal leaflets of Murraya paniculata, Murraya asiatica, Murraya exotica, Murraya ovatifoliolata var. ovatifoliolata, and Murraya ovatifoliolata var. zollingeri and Murraya × omphalocarpa, accessions were also subjected to discriminant function analysis. For this analysis, accessions of Murraya paniculata and Murraya asiatica were separated into accessions originating from Indonesia and those from Việt Nam and China. All characters contributed to the modelling of the data as indicated by the small values of Wilks’ lambda (Wilks’ lambda ranges from 1.0 (no discriminatory power) to 0.0 (perfect discriminatory power)) and the P values associated with the F values calculated from Wilks’ lambda (Table 4.12). The partial lambda values showed that relative importance of the different characters to the overall discrimination of the taxa with the length of the acuminate portion of basal leaflet (BACU, partial lambda small) contributing the most and the width of basal leaflet (BW, partial lambda high) contributing the least.

Table 4.12. Significance of the variables used for discriminant function analysis based on eight leaflet characters. BL: length basal leaflet; BW: width of basal leaflet; BACU: length of acuminate portion of basal leaflet; BAN: angle of base of basal leaflet; TL: length terminal leaflet; TW: width of terminal leaflet; TACU: length of acuminate portion of terminal leaflet; TAN: angle of base of terminal leaflet.

Character Wilks' Partial F on P lambda lambda removal SqrtBACU 0.033625 0.699136 20.79966 0.000000 LogBAN 0.031059 0.756879 15.52536 0.000000 LogTL 0.029844 0.787715 13.02556 0.000000 LogTAN 0.029290 0.802606 11.88717 0.000000 LogTW 0.027479 0.855507 8.16340 0.000000 LogTACU 0.028294 0.830859 9.83940 0.000000 LogBL 0.026298 0.893932 5.73494 0.000012 LogBW 0.025871 0.908664 4.85833 0.000095

Table 4.13 contains statistics associated with a step down test of all canonical roots with the first row representing the significance of all roots and the subsequent rows reporting the significance of the remaining roots after removal of the first and then successive roots. The χ2 values for roots 1–4 are statistically significant with these roots accounting for 0.993 of the observed variation (Table 4.14). The first root was most influenced by length of the acuminate portion of basal leaflet and the length, basal angle and width of the terminal leaflets. Root 2 was strongly influenced by the width and basal angle of the basal leaflet, root 3 by the length of the basal and terminal leaflets and root 4 by the 131 length and width of the terminal leaflets. Ordination of the data using roots 1 and 2 that account for 92% of the observed variation separated leaves from the taxa into two groups (Fig. 4.12). One group consisted of Murraya paniculata and Murraya asiatica and within this group Murraya asiatica accessions tended to separate away from the Murraya paniculata accessions. Within the second group, there was a good separation of accessions of Murraya exotica from Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri. The distribution of leaves from the large- and small-leaflet forms of Murraya ovatifoliolata var. ovatifoliolata and from Murraya ovatifoliolata var. zollingeri overlapped each other, and the distribution of accessions of Murraya × omphalocarpa overlapped those of Murraya exotica and the other varieties.

Table 4.13. Eigenvalues, Wilks’ lambda and chi square values for models using all discriminant functions (roots) (0 roots removed) and after removing each of the roots.

2 Eigenvalue Canonical R Wilks' χ Square Degrees of P Lambda freedom 0 5.804808 0.923604 0.023508 1108.245 48 0.000000 1 2.454432 0.842922 0.159969 541.585 35 0.000000 2 0.499492 0.577155 0.552602 175.266 24 0.000000 3 0.143232 0.353959 0.828622 55.551 15 0.000001 4 0.043988 0.205268 0.947307 15.996 8 0.042438 5 0.011145 0.104986 0.988978 3.275 3 0.351120

Table 4.14. Standardised coefficients for the canonical variables, eigenvalues and the proportions of the variation explained by each of the roots. BL: length basal leaflet; BW: width of basal leaflet; BACU: length of acuminate portion of basal leaflet; BAN: angle of base of basal leaflet; TL: length terminal leaflet; TW: width l of termina leaflet; TACU: length of acuminate portion of terminal leaflet; TAN: angle of base of terminal leaflet.

Root 1 Root 2 Root 3 Root 4 Root 5 Root 6 SqrtBACU 0.610726 ‐0.291229 0.32652 0.139774 0.02274 0.31492 LogBAN 0.169683 0.659542 0.03944 0.732719 ‐0.12206 0.61056 LogTL 0.687704 ‐0.139278 ‐1.23286 1.160397 0.46825 ‐1.03108 LogTAN 0.557163 0.378564 ‐0.27032 ‐0.484729 0.75149 ‐0.66785 LogTW ‐0.679638 0.374286 0.45758 ‐0.993778 ‐1.11775 0.26898 LogTACU 0.380684 ‐0.088471 0.55140 ‐0.225387 ‐0.14809 ‐0.38207 LogBL 0.266025 0.450577 ‐0.87729 ‐0.797336 0.91163 0.73385 LogBW ‐0.312502 ‐0.614613 0.63722 0.214650 ‐0.96422 0.22842 Eigen value 5.804808 2.454432 0.49949 0.143232 0.04399 0.01114 Cum.Prop 0.648068 0.922089 0.97785 0.993845 0.99876 1.00000

132

Figure 4.12. Scatterplot of the canonical scores plotted using the first two roots derived from the discriminant function analysis.

The ability of the discriminant function analysis to correctly classify leaves from the accessions studied to their taxon, as proposed from the molecular analysis, is shown in (Table 4.15). The analysis correctly assigned 87% of leaves to their correct taxon and was most successful for accessions of Murraya exotica, Murraya paniculata, Murraya asiatica, the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata with from 80 to 97% of leaves being correctly assigned. For the analysis, from 6–10 leaves of each plant were used, and for these accessions, a maximum of two incorrectly assigned leaves were found from each plant (Table 4.16). Thus, when a consensus is taken for each plant, each plant would have been correctly assigned to its taxon. Of note, if a leaf of Murraya asiatica was misclassified, it was classified (1st choice) as an accession of Murraya paniculata; however, if accessions of Murraya paniculata were misclassified, they were placed either in Murraya ovatifoliolata var. zollingeri, Murraya ovatifoliolata var. ovatifoliolata or Murraya asiatica. Less successful was the classification of Murraya ovatifoliolata var. ovatifoliolata (large leaflet form) (60% correct), Murraya ovatifoliolata var. zollingeri (60% correct) and Murraya × omphalocarpa (40% correct). The large leaflet form of Murraya ovatifoliolata var. ovatifoliolata was misclassified as either the small leaflet form or Murraya ovatifoliolata var. zollingeri. Murraya × omphalocarpa was usually misclassified as Murraya exotica and Murraya 133 ovatifoliolata var. zollingeri as either Murraya ovatifoliolata var. ovatifoliolata or Murraya × omphalocarpa.

The discriminant function analysis was repeated on a subset of leaves with those from M. ×omphalocarpa excluded and accessions of Murraya ovatifoliolata var. ovatifoliolata (large and small leaflet forms) and Murraya ovatifoliolata var. zollingeri grouped together as a single taxon designated Murraya ovatifoliolata. Thus, the ability of discriminant function analysis to discriminate between four taxa (Murraya exotica, Murraya paniculata, Murraya asiatica, and Murraya ovatifoliolata var. ovatifoliolata) was examined. As with the previous analysis, all variables contributed to the modelling of the data (Tables 4.17 and 4.18) and this resulted in three significant roots (Table 4.19). The length of the acuminate part of the basal leaflet and the length and basal angle of the terminal leaflet contributed most to root 1, the angle of the base and the length and width of the basal leaflet to root 2, and the length of the basal and terminal leaflets and the width of the terminal leaflet to root 3.

134

Table 4.15. Classification matrix giving the percentage of correct allocations of leaves to a taxon and the taxa to which leaves were incorrectly assigned.

Percent M. exotica M. paniculata M. asiatica M. ovatifoliolata M. ovatifoliolata M. ovatifoliolata M. × omphalocarpa correct var. ovatifoliolata ‘sl’ var. ovatifoliolata ‘ll’ var. zollingeri M. exotica 97.3 144 2 1 0 0 0 1 M. paniculata 92.7 0 38 1 0 1 1 0 M. asiatica 80.0 0 3 12 0 0 0 0 M. ovatifoliolata 91.7 1 0 0 44 1 1 1 var. ovatifoliolata ‘sl’ M. ovatifoliolata 58.8 0 0 0 3 10 4 0 var. ovatifoliolata ‘ll’ M. ovatifoliolata 60.0 0 0 0 3 2 9 1 var. zollingeri M. × omphalocarpa 40.0 8 1 1 1 0 1 8 Total 87.2 153 44 15 51 14 16 11

135

Table 4.16. Classification matrix for individual leaves that were misclassified according to discriminant function analysis, giving the classification according to the molecular analysis (proposed) and the 1st to 4th choices of taxon to which the leaf was assigned.

Accession Replication Taxon 1st choice 2nd choice 3rd choice 4th choice 6—ANSW 10 M. exotica M. asiatica M. exotica M. paniculata M. × omphalocarpa 6—ANSW M. exotica M. paniculata M. asiatica M. exotica M. × omphalocarpa 44—ICJ 10 M. exotica M. × omphalocarpa M. exotica M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 44—ICJ M. exotica M. paniculata M. asiatica M. × omphalocarpa M. exotica 25—IWJ 7 M. paniculata M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. paniculata ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘sl’ 34—IEJ 6 M. paniculata M. ovatifoliolata var. M. exotica M. × omphalocarpa M. ovatifoliolata var. zollingeri ovatifoliolata ‘ll’ 38—IEJ 8 M. paniculata M. asiatica M. paniculata M. × omphalocarpa M. exotica 88—VCP 9 M. asiatica M. paniculata M. asiatica M. × omphalocarpa M. exotica 95—CYD 6 M. asiatica M. paniculata M. asiatica M. × omphalocarpa M. exotica 95—CYD M. asiatica M. paniculata M. asiatica M. ovatifoliolata var. M. × omphalocarpa ovatifoliolata ‘sl’ 69—ANT 10 M. ovatifoliolata var. M. exotica M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. ovatifoliolata ‘sl’ zollingeri ovatifoliolata ‘ll’ 69—ANT M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa ovatifoliolata ‘sl’ ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘sl’ 72—AQLD 10 M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. exotica ovatifoliolata ‘sl’ zollingeri ovatifoliolata ‘sl’ 72—AQLD M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. M. exotica M. ovatifoliolata var. ovatifoliolata ‘sl’ ovatifoliolata ‘sl’ zollingeri 73—AQLD 10 M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa ovatifoliolata ‘ll’ ovatifoliolata ‘sl’ ovatifoliolata ‘ll’ zollingeri 75—AQLD 7 M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. exotica M. × omphalocarpa ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘ll’

136

75—AQLD M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. ovatifoliolata ‘ll’ ovatifoliolata ‘sl’ ovatifoliolata ‘ll’ zollingeri 75—AQLD M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘ll’ ovatifoliolata ‘sl’ 75—AQLD M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘sl’ ovatifoliolata ‘ll’ 75—AQLD M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. M. exotica ovatifoliolata ‘ll’ ovatifoliolata ‘sl’ zollingeri 75—AQLD M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. exotica ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘ll’ 91—T 10 M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. exotica zollingeri ovatifoliolata ‘sl’ 91—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 91—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. ovatifoliolata ‘sl’ zollingeri 91—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 91—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. ovatifoliolata ‘sl’ zollingeri 91—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 92—T 7 M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 92—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 92—T M. × omphalocarpa M. exotica M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ 93—T 3 M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. ovatifoliolata ‘sl’ ovatifoliolata ‘ll’ zollingeri 93—T M. × omphalocarpa M. paniculata M. asiatica M. ovatifoliolata var. M. × omphalocarpa 137

ovatifoliolata ‘sl’ 93—T M. × omphalocarpa M. asiatica M. paniculata M. × omphalocarpa M. exotica 113—INTT 10 M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa zollingeri ovatifoliolata ‘ll’ zollingeri ovatifoliolata ‘sl’ 113—INTT M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. M. exotica zollingeri ovatifoliolata ‘sl’ zollingeri 113—INTT M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. exotica zollingeri ovatifoliolata ‘ll’ zollingeri 113—INTT M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. M. exotica zollingeri zollingeri ovatifoliolata ‘sl’ 114—INTT 5 M. ovatifoliolata var. M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ zollingeri ovatifoliolata ‘ll’ 114—INTT M. ovatifoliolata var. M. ovatifoliolata var. M. × omphalocarpa M. ovatifoliolata var. M. ovatifoliolata var. zollingeri ovatifoliolata ‘sl’ zollingeri ovatifoliolata ‘ll’

138

Table 4.17. Significance of the variables used for discriminant function analysis based on eight leaflet characters. BL: length basal leaflet; BW: width of basal leaflet; BACU: length of acuminate portion of basal leaflet; BAN: angle of base of basal leaflet; TL: length terminal leaflet; TW: width of terminal leaflet; TACU: length of acuminate portion of terminal leaflet; TAN: angle of base of terminal leaflet. The data are from four taxa only.

Character Wilks' lambda Partial lambda F on removal P SqrtBACU 0.051271 0.694162 40.09333 0.000000 LogBAN 0.045601 0.780475 25.59568 0.000000 LogTACU 0.041211 0.863610 14.37164 0.000000 LogTAN 0.042446 0.838502 17.52691 0.000000 LogTL 0.038798 0.917340 8.19986 0.000030 LogTW 0.037892 0.939253 5.88547 0.000661 LogBW 0.038248 0.930513 6.79548 0.000196 LogBL 0.037803 0.941473 5.65701 0.000898

Table 4.18. Eigenvalues, Wilks’ lambda and chi square values for models using all discriminant functions (roots) (0 roots removed) and after removing each of the roots. The data are from four taxa only.

Eigenvalue Canonical R Wilks' Chi Degrees of P Lambda Square freedom 0 6.600845 0.931899 0.035591 923.9813 24 0.000000 1 2.378353 0.839046 0.270519 362.1534 14 0.000000 2 0.094201 0.293413 0.913909 24.9368 6 0.000351

Table 4.19. Standardised coefficients for the canonical variables, eigenvalues and the proportions of the variation explained by each of the roots. BL: length basal leaflet; BW: width of basal leaflet; BACU: length of acuminate portion of basal leaflet; BAN: angle of base of basal leaflet; TL: length terminal leaflet; TW: width of terminal leaflet; TACU: length of acuminate portion of terminal leaflet; TAN: angle of base of terminal leaflet. The data are from four taxa only.

Root 1 Root 2 Root 3 SqrtBACU 0.618645 ‐0.292085 0.02055

LogBAN 0.131182 0.697110 0.59526

LogTACU 0.437709 ‐0.081485 ‐0.14124 LogTAN 0.550650 0.398545 ‐0.45445 LogTL 0.559274 ‐0.021947 1.46370

LogTW ‐0.497225 0.223668 ‐0.84320

LogBW ‐0.273440 ‐0.687786 0.34255 LogBL 0.254371 0.505535 ‐1.11342 Eigenval 6.600845 2.378353 0.09420

Cum.Prop 0.727494 0.989618 1.00000

139

Ordination of the first two roots from the discriminant function analysis (Fig. 4.13) showed two groups of accessions. Within one group, a good separation of the leaves from Murraya asiatica from Murraya paniculata was found and within the second group accessions of Murraya exotica separated from most accessions of Murraya ovatifoliolata var. ovatifoliolata. Discriminant function analysis correctly classified 80– 96% of individual leaves to the correct taxon (Table 4.20). However, when a consensus was determined using the classifications from all leaves of each accession, all accessions could be correctly assigned to their taxon. Where leaves were misclassified, Murraya exotica and Murraya asiatica were mainly classified as Murraya ovatifoliolata var. ovatifoliolata. Murraya asiatica accessions were classified as Murraya paniculata and Murraya ovatifoliolata var. ovatifoliolata as Murraya exotica.

Figure 4.13. Scatterplot of the canonical scores plotted using the first two roots derived from the DFA analysis of four taxa.

140

Table 4.20. Classification matrix giving the percentage of correct allocations of leaves to a taxon and the taxa to which leaves were incorrectly assigned. The data are from four taxa only.

Percent M. exotica M. paniculata M. asiatica M. ovatifoliolata var. ovatifoliolata M. exotica 95.9 142 2 1 3 M. paniculata 92.7 0 38 1 2 M. asiatica 80.0 0 3 12 0 M. ovatifoliolata var. 96.3 3 0 0 77 ovatifoliolata Total 94.7 145 43 14 82

141

Table 4.21. Classification matrix for individual leaves that were misclassified according to discriminant function analysis, given the classification according to the molecular analysis (proposed) and the 1st to 4th choices of taxon to which the leaf was assigned. The data are from four taxa only.

Accession Replication Taxon 1st choice 2nd choice 3rd choice 4th choice 6—ANSW 10 M. exotica M. asiatica M. exotica M. paniculata M. ovatifoliolata var. ovatifoliolata 6—ANSW M. exotica M. paniculata M. asiatica M. exotica M. ovatifoliolata var. ovatifoliolata 27—IWJ 6 M. exotica M. ovatifoliolata var. M. exotica M. paniculata M. asiatica ovatifoliolata 35—IEJ 8 M. exotica M. ovatifoliolata var. M. exotica M. paniculata M. asiatica ovatifoliolata 44—ICJ 10 M. exotica M. paniculata M. asiatica M. exotica M. ovatifoliolata var. ovatifoliolata 44—ICJ M. exotica M. ovatifoliolata var. M. exotica M. paniculata M. asiatica ovatifoliolata 22—IWJ 10 M. paniculata M. ovatifoliolata var. M. paniculata M. exotica M. asiatica ovatifoliolata 34—IEJ 6 M. paniculata M. ovatifoliolata var. M. exotica M. paniculata M. asiatica ovatifoliolata 38—IEJ 8 M. paniculata M. asiatica M. paniculata M. exotica M. ovatifoliolata var. ovatifoliolata 88—VCP 9 M. asiatica M. paniculata M. asiatica M. exotica M. ovatifoliolata var. ovatifoliolata 95—CYD 6 M. asiatica M. paniculata M. asiatica M. exotica M. ovatifoliolata var. ovatifoliolata 95—CYD M. asiatica M. paniculata M. asiatica M. ovatifoliolata var. M. exotica ovatifoliolata 54—AQLD 7 M. ovatifoliolata var. M. exotica M. ovatifoliolata var. M. asiatica M. paniculata ovatifoliolata ovatifoliolata 7 2—AQLD 9 M. ovatifoliolata var. M. exotica M. ovatifoliolata var. M. paniculata M. asiatica ovatifoliolata ovatifoliolata 113—INTT 10 M. ovatifoliolata var. M. exotica M. ovatifoliolata var. M. paniculata M. asiatica ovatifoliolata ovatifoliolata

142

4.3.4. Use of elliptic Fourier descriptors for separating taxa on the basis of leaflet shapes

Elliptic Fourier transformation was undertaken with 45 harmonics by Chc2Nef, resulting in a matrix of 180 elliptic Fourier coefficients for leaflet shapes. Pictorial representations of the components of leaflet shape variation are illustrated in Figs 4.14 and 4.15. Principal components analysis separated Murraya paniculata and Murraya exotica in analyses of shapes of both basal and terminal leaflets (Figs 4.16 and 4.17). Murraya ovatifoliolata var. ovatifoliolata also differed from Murraya paniculata and Murraya exotica. The first two principal components, PC1 and PC2, of the elliptic Fourier descriptors of leaflet shape of basal leaflets explained 76.6% of the variance (60.5 and 16.1, respectively). The first two principal components based on PCA of the elliptic Fourier descriptors of leaflet shape of terminal leaflets explained 86.6% of the variance (73.0 and 13.6%, respectively). In the analyses of terminal and basal leaflet shapes, PC1 accounted for much of the variation in leaflet apex shape, base shape and leaflet width, PC2 accounted for the location of the widest point and PC3 (not shown in the principal component analysis figures) accounted for basal and apical symmetry (Figs 4.14 and 4.15).

The redundancy analysis of the elliptic Fourier descriptors of basal leaflets modelled 94.5% of the variance in the data. The first two axes accounted for 45.0% and 24.4%, respectively. In the redundancy analysis, the correlation biplot for basal leaflets (Fig. 4.18) indicated negligible correlation in the descriptors of shapes for Murraya paniculata and Murraya exotica whereas those for the small and large leaflet forms of Murraya ovatifoliolata var. ovatifoliolata were highly correlated. The descriptors for Murraya × omphalocarpa were highly negatively correlated with those for Murraya paniculata. The Murraya paniculata, Murraya exotica and Murraya ovatifoliolata var. ovatifoliolata lines in the biplot indicate that most variance was related to these taxa. Relationships in Fig. 4.19 indicate similar correlations for terminal leaflets. The redundancy analysis of the elliptic Fourier descriptors of basal leaflets accounted for 93.2% of the variance in the data. The first two axes accounted for 44.4% and 24.1%, respectively. Both sets of redundancy analysis results indicated significant differences in the leaflet shapes of Murraya paniculata, Murraya exotica and Murraya ovatifoliolata var. ovatifoliolata.

143

Figure 4.14. Pictorial representations of the components of leaflet shape variation explained by the first three principal components (PC) of elliptic Fourier descriptors of basal leaflets of Murraya. The third column from the left in the diagram represents mean shapes of each PC, the second and fourth columns the shapes at ‐2× and 2× the standard deviation of the means, and the first overlays of all of these shapes combined for each PC.

Figure 4.15. Pictorial representations of the components of leaflet shape variation explained by the first three principal components (PC) of elliptic Fourier descriptors of terminal leaflets of Murraya. The third column from the left in the diagram represents mean shapes of each PC, the second and fourth columns the shapes at ‐2× and 2× the standard deviation of the means, and the first overlays of all of these shapes combined for each PC.

144

Figure 4.16. Principal component analysis of elliptic Fourier descriptors of basal leaflet shapes of Murraya: M. asiatica , M. paniculata , M. exotica , M. ovatifoliolata var. ovatifoliolata (small leaflet) , M. ovatifoliolata var. ovatifoliolata (large leaflet) , M. ovatifoliolata var. zollingeri , M. ×. omphalocarpa . The PCI and PC2 axes accounted for 60.5% of and 16.1% of the variance in the data, respectively.

Figure 4.17. Redundancy analysis correlation biplot for elliptic Fourier descriptors of basal leaflet shapes of Murraya: lines clockwise from top, M. ovatifoliolata var. ovatifoliolata (small leaflet) M. × omphalocarpa, M. ovatifoliolata var. ovatifoliolata (large leaflet), M. ovatifoliolata var. zollingeri, M. asiatica, M. paniculata and M. exotica. Axes 1 and 2 accounted for 45% and 24.4% of variance in the data, respectively.

145

Figure 4.18. Principal component analysis of elliptic Fourier descriptors of terminal leaflet shapes of Murraya: M. asiatica , M. paniculata , M. exotica , M. ovatifoliolata var. ovatifoliolata (small leaflet) , M. ovatifoliolata var. ovatifoliolata (large leaflet) , M. ovatifoliolata var. zollingeri , M. ×omphalocarpa . The PCI and PC2 axes accounted for 73% of and 13.6% of variance in the data, respectively.

Figure 4.19. Redundancy analysis correlation biplot for elliptic Fourier descriptors of terminal leaflet shapes of Murraya: lines clockwise from top, M. exotica, M. paniculata, M ovatifoliolata var. zollingeri. M. × omphalocarpa, M. asiatica, M. ovatifoliolata var. ovatifoliolata (large leaflet) and M. ovatifoliolata var. ovatifoliolata (small leaflet). Axes 1 and 2 accounted for 44.4% and 24.1% of the variance in the data, respectively.

146

The Monte Carlo permutation tests from redundancy analysis of the data sets from the basal and terminal leaflets of the three varieties including the large and small leaflet forms of Murraya ovatifoliolata var. ovatifoliolata (Table 4.22) showed that upport for separation of the plants into four taxa.

Table 4.22. Monte Carlo permutation tests from redundancy analysis of the elliptical Fourier descriptors of terminal and basal leaflets from M. ovatifoliolata var. ovatifoliolata (small leaflet) (30 leaves from 7 plants), M. ovatifoliolata var. ovatifoliolata (large leaflet) (10 leaves from 2 plants), M. × omphalocarpa (13 leaves from 3 plants) and M. ovatifoliolata var. zollingeri (9 leaves from 2 plants).

Leaflet Origin F P Basal M. ovatifoliolata var. ovatifoliolata (small leaflet) 13.061 0.001 M. ovatifoliolata var. ovatifoliolata (large leaflet) 3.089 0.027 M. ovatifoliolata var. zollingeri 3.116 0.036 M. × omphalocarpa 5.137 0.002 Terminal M. ovatifoliolata var. ovatifoliolata (small leaflet) 14.301 0.001 M. ovatifoliolata var. ovatifoliolata (large leaflet) 1.200 0.272 M. ovatifoliolata var. zollingeri 6.616 0.005 M. × omphalocarpa 8.242 0.002

Due to the morphological similarities between the leaflets of Murraya paniculata and Murraya asiatica, the average shapes of the leaves were visualised using SHAPE. For the terminal leaflets, most variation between the two species was seen in the first principle component (Fig. 4.20), with greater variation in leaflet width occurring on the basal portions of the leaflets of Murraya paniculata than the Murraya asiatica accessions. Greater variation between the two species was seen in the basal leaflets (Fig. 4.21). In the first principle component, more variation in leaflet width occurred in the distal portion of the leaflets from the Murraya paniculata accessions whilst for the Murraya asiatica accession the greater variation in width occurred in the basal portion. Principle component 3 also showed more variation in leaflet width towards the base and apex of the Murraya paniculata leaflets than in the Murraya asiatica accessions. The Monte Carlo permutation tests from redundancy analysis of the data sets from the basal and terminal leaflets (Table 4.23) showed that greater variation occurred among the basal leaflets of the two species than among the terminal leaflets. The permutation tests from the basal leaflets also support a separation of the plants into two species.

147

Table 4.23. Monte Carlo permutation tests from the redundancy analysis of the elliptical Fourier descriptors of terminal and basal leaflets from 35 leaves taken from seven Indonesian accessions of M. paniculata and from 11 leaves taken from accessions of M. asiatica.

Leaflet Origin F P Terminal M. paniculata 2.081 0.096 M. asiatica 2.081 0.088 Basal M. paniculata 3.964 0.007 M. asiatica 3.964 0.010

148

Figure 4.20. Pictorial representations of the components of terminal leaflet shape variation explained by the first three principal components (PC) of elliptic Fourier descriptors of terminal leaflets from (top) M. paniculata and (bottom) M. asiatica. The third column from the left in the diagram represents mean shapes of each PC, the second and fourth columns the shapes at ‐2× and 2× the standard deviation of the means, and the first overlays of all of these shapes combined for each PC. Leaflet bases are on the left of the diagrams.

149

Figure 21. Pictorial representations of the components of basal leaflet shape variation explained by the first three principal components (PC) of elliptic Fourier descriptors of terminal leaflets from (top) M. paniculata and (bottom) M. asiatica. The third column from the left in the diagram represents mean shapes of each PC, the second and fourth columns the shapes at ‐2× and 2× the standard deviation of the means, and the first overlays of all of these shapes combined for each PC. Leaflet bases are on the left of the diagrams.

150

4.3.5. Morphological descriptions of putative taxa

4.3.5.1. Merrillia caloxylon

The following description of vegetative characters is based on accession 23—IWJ from a 4 m-high tree in the Bogor Botanic Garden, Indonesia (see Table 4.1):  Stems: young green and recently mature pubescent;  Leaves: 8 (7–9) foliolate, 144 (115–173) mm long, ratio of leaf length to number of leaflets 18 (13–19) (Fig. 4.22);  Rachis: 69 (52–80) mm, winged, ridged below petiolule junctions, pubescent proximally, glabrous distally above and below, wings glabrous;  Petiole: 3.5 (3–5) mm, pubescent;  Acuminate leaflets: (excluding two basal leaflets) > 90%;  Basal leaflets: opposite, 11 (5–18) × 8 (3–10) mm, rhombic or suborbicular, bases cuneate and asymmetric, angle of base 116° (100–135), apices acute, obtuse or rounded, proportion leaflet tip length to leaflet length 0%, midveins and lateral veins pubescent above and below, laminas sparsely pubescent above and below, margins crenate and glabrous, petiolules 1.2 (1–1.7) mm, pubescent above and below;  Terminal leaflets: 69 (55–86) × 32 (24–40) mm, elliptic, unevenly textured, bases cuneate and symmetric, angle of base 75° (58–90), apices acute or acuminate, proportion leaflet tip length to leaflet length 7% (0–10), midveins and lateral veins pubescent above and below, laminas sparsely pubescent, margins crenate with tufted teeth, otherwise glabrous, petiolules 1.8 (1–2.3) mm, pubescent above and below.

Observed flowers on the tree in the Bogor Botanic Gardens were born individually on the branches in an axillary position and pale yellowish-green: pedicels 3 mm; petals 5, oblance-ovate 47–51 × 11–13 mm, aestivation imbricate; stamens 10 of unequal length, longest 14–21 mm, shortest 12–18 mm; pistils 19–20 mm, ovary 7 × 2 mm; stigmas rounded, 1.8–2.3 mm diameter. Figures 4.23 and 4.24 feature flower buds and flowers, and fruit of Merrillia caloxylon in the botanic garden (Rimba Ilmu) at the University of Malaya, Kuala Lumpur. The fruit were ellipsoid, 90 mm diameter, smooth, grey-green, with 5 locules (Fig. 4.24).

151

Figure 4.22. Leaves and leaflets of Merrillia caloxylon (accession 23), Bogor Botanic Garden (Nguyen Huy Chung).

Figure 4.23. Flower buds and flowers of Merrillia caloxylon: Rimba Ilmu, University of Malaya, Kuala Lumpur (GAC Beattie). 152

Figure 4.24. Merrillia caloxylon fruit (about 90 mm diameter), Rimba Ilmu, University of Malaya, Kuala Lumpur (GAC Beattie).

153

4.3.5.2. Murraya paniculata

The following descriptions are based on accessions 22—IWJ, 25—IWJ, 34—IEJ, 38— IEJ and 48—ICJ from Indonesia (see Table 4.1) using rounded values of data in Tables 4.4–4.8. The molecular results in Chapter 3, mean character values shown in the box and whisker plots (Figs 4.2 & 4.3), and the results of elliptical Fourier descriptor analyses (Table 23, Figs. 4.16–4.19) suggested that these accessions represented a species distinct from Murraya asiatica.  Stems: young green and recently mature mostly glabrous, if sparsely pubescent to pubescent then mostly on young green stems;  Leaves: 5 (3–7) foliolate, 142 (72–192) mm long, ratio of leaf length to number of leaflets 29 (18–43) (Fig. 4.25);  Rachis: 51 (15–78) mm, mostly pubescent, sometimes glabrous above, scattered pubescent, occasionally sparse or glabrous below, distance between petiolule- rachis junctions of first and second leaflets 7 (0–25) mm, of penultimate and terminal leaflets 15 (0–30) mm;  Petiole: 14 (5–31) mm, mostly pubescent, sometimes glabrous above, scattered pubescent, occasionally sparse or glabrous below;  Acuminate leaflets: > 90%;  Basal leaflets: 51 (28–75) 42 (25–59) × 23 (15–32) mm, ratio of leaflet length to width 1.8 (1.4–2.4) mm, ovate to elliptic, bases mostly obtuse or rounded, nearly all asymmetric, angle of base 104° (77–135), apices acute or acuminate, ± emarginate, proportion leaflet tip length to leaflet length 8% (0–15), surface midveins and lateral veins mostly glabrous, sometimes pubescent above and below, sometimes in tufts (accession 38—IEJ), laminas glabrous above and below, petiolules 4 (2–8) mm, mostly pubescent, sometimes glabrous above, mostly glabrous, occasionally sparse or scattered pubescent below, margins entire or crenulated, glabrous;  Terminal leaflets: 74 (45–98) × 30 (20–43) mm, elliptic, bases cuneate, equally symmetric and asymmetric, angle of base 70° (48–98), apices acuminate, ± emarginate, proportion leaflet tip length to leaflet length 9% (4–14), midveins and lateral veins mostly glabrous, sometimes pubescent above and below, sometimes in tufts (accession 38—IEJ), laminas glabrous above and below, petiolules 4 (2–6) mm, mostly pubescent, sometimes glabrous above, mostly glabrous, occasionally sparse or scattered pubescent below, margins entire or crenulate, glabrous.

154

Fig. 4.26 features scanning electron micrographs of pubescence on rachises and terminal leaflets of accession 22—IWJ from Bogor Botanic Garden.

These observations and measurements, and those for the other Murraya accessions, are summarised in Tables 4.3–4.7.

Observed flowers (Fig. 4.27) were white, inflorescences axillary or terminal, 1–7 per inflorescence, petals 5 petals, oblanceolate 4–7 × 18–29 mm, with greenish glands visible on outer surface, apices acute, aestivation imbricate; stamens 10, 5 short (6–10 mm), 5 long (7–13 mm), alternately; carpels 9–13 mm; ovary 1–2 × 2–3 mm; stigma bilobed, 1.7–2.6 mm long; pedicels 5–14 mm long. Observed fruits orange to red, long ellipsoid 6–12 × 14–25 mm, base convex, apex mammiform or acute (Fig. 4.29); seeds 1–4, white, ovoid 3.5–6 × 8–12 mm, hairy (Fig. 4.28).

Figure 4.25. Leaves of M. paniculata (accession 22), Bogor Botanic Garden (Nguyen Huy Chung).

155

Figure 4.26. Scanning electron micrographs of pubescence on rachises and terminal leaflets of M. paniculata (accession from 22—IWJ Bogor Botanic Garden, Java, Indonesia).

156

Figure 4.27. Flowers of M. paniculata (accession 22—IWJ), Bogor Botanic Garden (Nguyen Huy Chung).

Figure 4.28. Fruit and seeds of M. paniculata (accession 22—IWJ), Bogor Botanic Garden, Indonesia (Nguyen Huy Chung).

157

4.3.5.3. Murraya asiatica

The following descriptions are based on accessions 88—VCP from Việt Nam, and 95— CYD from China (see Table 4.1) using rounded values of data in Tables 4.4–4.8.  Stems: young green and recently mature mostly glabrous, if sparsely pubescent to pubescent then mostly on young green stems;  Leaves: 5 (3–7) foliolate, 142 (72–192) mm long, ratio of leaf length to number of leaflets 29 (18–43) (Figs 4.29);  Rachis: 51 (15–78) mm, mostly pubescent, sometimes glabrous above, scattered pubescent, occasionally sparse or glabrous below, distance between petiolule- rachis junctions of first and second leaflets 7 (0–25) mm, of penultimate and terminal leaflets 15 (0–30) mm;  Petiole: 14 (5–31) mm, mostly pubescent, sometimes glabrous above, scattered pubescent, occasionally sparse or glabrous below;  Acuminate leaflets: > 90%;  Basal leaflets: 51 (28–75) × 25 (14–35) mm, ratio of leaflet length to width 2 (1.4–2.4), ovate to elliptic, bases mostly obtuse or rounded, nearly all asymmetric, angle of base 82° (67–95), apices acute or acuminate, ± emarginate, proportion leaflet tip length to leaflet length 8% (0–15), surface midveins and lateral veins mostly glabrous, laminas glabrous above and below, petiolules 4 (2–8) mm, mostly pubescent, sometimes glabrous above, mostly glabrous, occasionally sparse or scattered pubescent below, margins entire or crenulated, glabrous;  Terminal leaflets: 74 (45–98) × 30 (20–43) mm, elliptic, bases cuneate, equally symmetric and asymmetric, angle of base 70° (48–98), apices acuminate, ± emarginate, proportion leaflet tip length to leaflet length 10% (6–15), midveins and lateral veins mostly glabrous, laminas glabrous above and below, petiolules 4 (2–6) mm, mostly pubescent, sometimes glabrous above, mostly glabrous, occasionally sparse or scattered pubescent below, margins entire or crenulate, glabrous.

158

Figure 4.29. Leaves of M. asiatica; accession 95—CYD (left), Yingde County, Guangdong, China (GAC Beattie), and accession 88—VCP (right), Cuc Phuong National Park, Việt Nam (Nguyen Huy Chung).

159

4.3.5.4. Murraya ovatifoliolata var. ovatifoliolata (large leaflet form)

The following description is based on accessions 73—AQLD and 75—AQLD from Queensland (see Table 4.1) using rounded values of data in Tables 4.4–4.8:  Stems: young green and recently mature glabrous;  Leaves: 6 (3–8) foliolate, 130 (88–165) mm long, ratio of leaf length to number of leaflets 25 (15–39) (Figs 4.30–4.31);  Rachis 59 (30–88) mm, glabrous above and below, distance between petiolule- rachis junctions of first and second leaflets 8 (2–18) mm, of penultimate and terminal leaflets 17 (9–29) mm;  Petiole: 11 (4–15) mm, glabrous above and below;  Acuminate leaflets: ~ 50%;  Basal leaflets: 31 (14–47) × 22 (9–40) mm, ovate or elliptic, bases obtuse or rounded, symmetric or asymmetric, angle of base 120° (90–170), apices acute, acuminate, obtuse or rounded, proportion of leaflet tip to leaflet length 0%, midveins and lateral veins glabrous above and below, laminas glabrous above and below, petiolule 2 (1–3) mm, pubescent to scattered pubescent above and below, margins entire or crenate, glabrous except near petiolule;  Terminal leaflets: 59 (39–85) × 36 (24-60) mm, ovate or elliptic, symmetric, bases cuneate, obtuse or rounded, angle of base 98° (62–134), apices acute or acuminate, slightly emarginate, proportion of leaflet tip to leaflet length 1% (0–7), midveins scattered pubescent and lateral veins glabrous above, both glabrous below, laminas glabrous above and below, petiolule 2 (1–3) mm, pubescent to scattered pubescent above and below, margins entire or crenate, glabrous except near petiolule.

Fig. 4.32 features scanning electron micrographs of pubescence on rachises and terminal leaflets of accession from 73—AQLD Cooktown-Mt. Webb National Park, Queensland, Australia.

No flowers or fruit were observed.

160

A B

Figure 4.30. Leaves of the large leaflet form of M. ovatifoliolata var. ovatifoliolata (accession 73— AQLD) from Cooktown, Queensland (left, Nguyen Huy Chung; right, Fanie Venter).

Figure 4.31. Leaves of the large leaflet form of M. ovatifoliolata var. ovatifoliolata (accession 75— AQLD) from Cairns, Queensland (left, Nguyen Huy Chung; right, Fanie Venter).

161

Figure 4.32. Scanning electron micrographs of pubescence on rachises and terminal leaflets of the large leaflet form of M. ovatifoliolata var. ovatifoliolata (accession 73—AQLD from Cooktown‐Mt Webb National Park, Queensland, Australia).

162

4.3.5.5. Murraya ovatifoliolata var. ovatifoliolata (small leaflet form)

The following description is based on accessions 54—AQLD, 72—AQLD, 74—AQLD and 115—AQLD from Queensland and 69—ANT and 71—ANT from the Northern Territory (see Table 4.1) using rounded values of data in Tables 4.4–4.8:  Stems: young green and recently mature stems pubescent;  Leaves: 6 (3–9) foliolate, 93 (48–168) mm long, ratio of leaf length to number of leaflets 16 (10–26) (Fig. 4.33);  Rachis: 44 (10–116) mm long, pubescent above and below, distance between petiolule-rachis junctions of first and second leaflets 5 (0–13) mm, of penultimate and terminal leaflets 11 (0–27) mm;  Petiole: 9 (3–18) mm, pubescent above and below;  Acuminate leaflets: ~ 50%;  Basal leaflets: 18 (8–40) × 14 (6–28) mm, ovate or widely elliptic, bases obtuse or rounded, asymmetric or symmetric, angle of base 123° (74–150), apices acute, acuminate, obtuse to rounded, emarginate or not emarginate, proportion of leaflet tip to leaflet length 0%, midveins and lateral veins pubescent (midveins less distally) above, midveins and lateral veins pubescent below, laminas pubescent to sparsely pubescent above and below, petiolules 2 (1–4) mm, pubescent above and below, margins entire or crenate, ciliate;  Terminal leaflets: 37 (22–56) × 25 (14–45) mm, ovate or elliptic, rarely obovate, bases cuneate, obtuse or rounded, asymmetric or symmetric, angle of base 99° (63–130), apices acute or acuminate, emarginate or not emarginate, proportion of leaflet tip to leaflet length 1% (0–9), midveins and lateral veins pubescent above, midveins and lateral veins pubescent below, lateral veins sparsely, laminas pubescent above and below, petiolules 3 (1–5) mm, pubescent above and below, broad margins entire or crenate, ciliate.

Fig. 4.34 features scanning electron micrographs of pubescence on rachises and terminal leaflets of accession 69—ANT from Haddon Head Beach, Blue Mud Bay, Northern Territory, Australia.

No flowers or fruit were observed.

163

Figure 4.33. Small leaflet form of M. ovatifoliolata var. ovatifoliolata (Nguyen Huy Chung) from (A) Blue Mud Bay, Northern Territory (accession 69—ANT), (B) Battle Camp, Queensland (accession 74—AQLD), (C) Gove, Northern Territory (accession 71—ANT), and (D) Tondoon Botanic Garden, Gladstone, Queensland (via SRBG) (accession 115—AQLD).

164

Figure 4.34. Scanning electron micrographs of pubescence on rachises and terminal leaflets of the small leaflet form of M. ovatifoliolata var. ovatifoliolata (accession 69—ANT from Haddon Head Beach, Blue Mud Bay, Northern Territory, Australia).

165

4.3.5.6. Murraya ovatifoliolata var. zollingeri

The following description is based on accessions 113—INTT and 114—INTT from Kupang, Nusa Tenggara Timur, Indonesia (see Table 4.1) using rounded values of data in Tables 4.4–4.828:  Stems: young greens stems pubescent, recently matured stems sparsely pubescent;  Leaves: 7 (7–9) foliolate, 127 (59–250) mm long, ratio of leaf length to number of leaflets 17 (8–31) (Fig. 35);  Rachis: 74 (50–190) mm long, pubescent above and below, distance between petiolule-rachis junctions of first and second leaflets 5 (0–14) mm, of penultimate and terminal leaflets 11 (0–27) mm;  Petiole: 20 (8–40) mm, pubescent above and below;  Acuminate leaflets: 0%;  Basal leaflets: 24 (13–39) × 15 (10–25) mm, ovate or elliptic, base obtuse or rounded, asymmetric, angle of base 106° (90–125), apices acute, blunt and unemarginate, margins undulate, proportion of leaflet tip to leaflet length 0%, midveins pubescent, lateral veins glabrous above and below, laminas scattered to sparsely pubescent, more towards base, above and below, petiolules 2 (1–5) mm, pubescent above and below, margins ciliate;  Terminal leaflets: 46 (32–65) × 23 (18–33) mm, elliptic, sometimes ovate; base cuneate or obtuse, symmetric or asymmetric, angle of base 81° (53–110), apices acute, blunt, not emarginate, proportion of leaflet tip to leaflet length 0%, midveins and lateral veins pubescent above, midveins pubescent, lateral veins glabrous rarely sparse, below, laminas scattered to sparsely pubescent above and below, more towards base, petiolules 3 (2–4) mm, pubescent above and below, margins slightly crenate, ciliate.

Fig. 4.36 features scanning electron microscope photographs of pubescence on rachises and terminal leaflets of accession 113—INTT from Kupang, Nusa Tenggara Timur, Indonesia.

Fruits red, turbinate, the longitudinal cross section ovate, base convex, apex acuminate; seeds ovoid, hairy, light brown.

28 Tanaka (1929) described the hairs of this variety as tomentose. 166

Figure 4.35. Leaves of M. ovatifoliolata var. zollingeri accession 113—INTT from Nusa Tenggara Timur, Indonesia (Inggit Puji Astuti).

167

Figure 4.36. Scanning electron micrographs of pubescence on rachises and terminal leaflets of M. ovatifoliolata var. zollingeri (accession 113—INTT from Kupang, Nusa Tenggara Timur, Indonesia).

168

4.3.5.7. Murraya exotica

The following description is based on accessions 2—ANSW, 6—ANSW, 8—ANSW, 13—AQLD, 53—ANSW, 108—ANSW and 110—ANSW from Australia), 102—BSP, 104—BSP and 106—BSP from Brazil, 27—IWJ, 28—IWJ, 35—IEJ, 37—IEJ, 44—ICJ and 47—ICJ from Indonesia, and 111—UFBG and 112—UFBG from the United States of America (see Table 4.1) using rounded values of data in Tables 4.4–4.8:  Stems: young green and recently mature stems pubescent;  Leaves: 7 (4–10) foliolate, 98 (50–170) mm long, ratio of leaf length to number of leaflets 14 (9–22) (Fig. 4.37);  Rachis: 44 (9–94) mm long, pubescent above and below, distance between petiolule-rachis junctions of first and second leaflets 4 (0–18) mm, of penultimate and terminal leaflets 7 (0–20) mm;  Petiole: 10 (4–21) mm, pubescent above and below;  Acuminate leaflets: ~ 30%;  Basal leaflets: 22 (13–38) × 14 (8–24) mm, obovate or elliptic, base cuneate, mostly asymmetric, angle of base 72° (47–97), apices acute, acuminate, obtuse or rounded, emarginate or entire, proportion of leaflet tip to leaflet length 0% (1–8), midveins pubescent above and below, lateral midveins sparsely pubescent, less so proximally, above and below, lamina pubescent to sparsely pubescent above and below, petiolules 2 (1–4) mm, pubescent; margins entire or crenulate, ciliate;  Terminal leaflets: 41 (27–70) × 19 (11–29) mm, elliptic or obovate, base cuneate, symmetric, rarely asymmetric, angle of base 52° (30–75), apices acute, acuminate, obtuse or rounded with a blunt or short tip, entire or emarginate, proportion of leaflet tip to leaflet length 1% (0–8), midveins pubescent above and below, lateral veins glabrous above, pubescent to sparsely pubescent below, less so proximally above and below, laminas mostly glabrous above, pubescent to sparsely pubescent below above and below, petiolules 2 (1–4) mm, pubescent; margins mostly entire but sometimes undulate, ciliate.

Fig. 4.41 features scanning electron microscope photographs of pubescence on rachises and terminal leaflets of accession 6 from Windsor, New South Wales, Australia.

Flowers terminal or axillary, pedicel 3-8 mm, usually in clusters near the end of branches, 1–8 per inflorescence (Fig. 4.42), white, fragrant, petals 5, stamens10, 169 alternately long (7–11 mm) and short (5–8 mm); pistil 6–9 mm, ovary 1–2 × 2–3 mm; stigma yellow and bilobed, 1–2 mm long, petals 3–7 × 10–26 mm, oblance-ovate or oblanceolate, apex blunt, greenish oil glands on the outer surface.

Fruit 6–12 × 8–17 mm, ellipsoid, orange to red; base of fruits convex, apex mammiform, sometimes oblique mammiform or acute, 1 or 2 seeds, rarely 3, 3–7 × 6– 13 mm, light brown, hairy and spheroid or ovoid, rounded around the middle if one seed per fruit, flattened on one side if more than one.

170

Figure 4.37. Leaves of M. exotica from Fairchild Botanic Garden, Florida (accession 111—UFBG: upper left), Brisbane, Queensland (accession 13—AQLD: upper right), (Windsor, New South Wales (accession 6: lower left), and Bogor Botanic Garden, West Java (accession 26—IWJ: lower right) (Nguyen Huy Chung).

171

Figure 4.38. Scanning electron micrographs of pubescence on rachises and terminal leaflets of M. exotica (accession 6—ANSW from Windsor, New South Wales, Australia).

172

Figure 4.39. Flowers and petals of M. exotica from (A and B) Mulgrave (accession 109—ANSW), New South Wales, (C) Bogor Botanic Garden, West Java (accession 27—IWJ), and (D) Richmond, New South Wales (accession 2—ANSW), (Nguyen Huy Chung).

173

4.3.5.8. Murraya × omphalocarpa

The following description is based on accessions 91—T, 92—T and 93—T from Orchid Island, Taiwan (see Table 4.1) using rounded values of data in Tables 4.4–4.8:  Stems: young green and recently mature stems glabrous;  Leaves 7 (4–9) foliolate, 132 (75–200) mm long, ratio of leaf length to number of leaflets 18 (11–50) (Fig. 40);  Rachis: 58 (40–85) mm long, ridged, pubescent above, glabrous below, distance between petiolule-rachis junctions of first and second leaflets 4 (0–13) mm, of penultimate and terminal leaflets 13 (2–24) mm;  Petiole 17 (6–30) mm, pubescent above, glabrous below;  Acuminate leaflets: ~ 40%;  Basal leaflets: 28 (12–60) × 19 (9–38) mm, obovate or elliptic, base obtuse or rounded, asymmetric, seldom symmetric, angle of base 102° (75–142), apices acute, acuminate, obtuse or rounded, not emarginate, proportion of leaflet tip to leaflet length 1% (0–5), midveins and lateral veins glabrous above, midveins with dispersed pubescent tufts, lateral veins glabrous below, laminas glabrous above and below, petiolules 3 (1–5) mm, pubescent above and below, margins entire or sometimes slightly crenate, glabrous;  Terminal leaflets: 53 (23–95) × 30 (10–61) mm, obovate or elliptic; base cuneate, symmetric, angle of base 69° (53–90), apices acute or acuminate, proportion of leaflet tip to leaflet length 2% (0–8), midveins and lateral veins glabrous above, midveins with dispersed pubescent tufts, lateral veins glabrous below, laminas glabrous above and below, petiolules 3 (2–7) mm, pubescent above and below, margins entire or crenate, glabrous.

Fig. 4.41 features scanning electron microscope photographs of pubescence on rachises and terminal leaflets of accession 91 from Orchid Island, Taiwan.

Inflorescence terminal or axillary, flowers white, with greenish oil glands in the outer surface of petals (Fig. 4.42), petals oblance-ovate. Fruit red, turbinate shape (broad at base, gradually narrowing to the apex, the longitudinal cross section lance-ovate, base convex, apex acute (Fig. 4.42).

174

Figure 4.40. Leaves of M. × omphalocarpa accessions 93—T (left) and 92—T (right) from Orchid Island, Taiwan: upper, Nguyen Huy Chung, lower, Hung Shih‐Cheng.

175

Figure 4.41. Scanning electron micrographs of pubescence on rachises and terminal leaflets of the small leaflet form of M. × omphalocarpa (accession 91—T from Orchid Island, Taiwan).

176

Figure 4.42. Flowers and fruits of M. × omphalocarpa from Orchid Island, Taiwan (Hung Shih‐Cheng).

177

4.3.5.9. Murraya × cycloopensis

The following description is based on accession 24—IP, Bogor Botanic Gardens from Cycloop, Papua, Indonesia (see Table 4.1) using rounded values of data in Tables 4.4– 4.8:  Stems: young green stems densely pubescent recently mature stems with tufts of pubescent hairs, latter on juvenile epidermis on older stems;  Leaves: 7 (5–8) foliolate, 160 (140–185) mm long, ratio of leaf length to number of leaflets 23 (21–28) (Fig. 43);  Rachis: 78 (53–100) mm long, pubescent, distance between petiolule-rachis junctions of first and second leaflets 9 (3–20) mm, of penultimate and terminal leaflets 15 (8–23) mm;  Petiole: 23 (20–27) mm, densely pubescent;  Acuminate leaflets: > 90%;  Basal leaflets: 35 (28–42) × 23 (18–27) mm, ovate, base rounded and asymmetric, angle of base 130° (118–145), apices acuminate and not emarginate, proportion of leaflet tip to leaflet length 4% (0–10), midveins densely pubescent, lateral veins pubescent, above and below, laminas sparsely pubescent above and below, petiolules 3 (2–3) mm, densely pubescent above and below, margins entire or crenulate, ciliate;  Terminal leaflets: 58 (54–65) × 32 (25–38) mm, elliptic; bases broad, usually obtuse, and very seldom broadly cuneate, symmetric or asymmetric, angle of base 100° (80–117), apices acuminate, not emarginate, proportion of leaflet tip to leaflet length 5% (0–8), midveins and lateral veins pubescent above and below, laminas sparsely pubescent above and below, petiolules 3 (2–4) mm, pubescent above and below (Fig. 4.44), margins crenulate, ciliate.

Fig. 45 features scanning electron microscope photographs of pubescence on rachises and terminal leaflets of accession 24.

Flowers terminal or auxillary on the branches, 1–6 per inflorescence, white, pedicel 3-6 mm, petals 5, stamens 10 of unequal length, shortest 5–6 mm, longest 7–8 mm, pistil 8– 9 mm, ovary 1 × 3 mm, petals 3–4 × 18–19 mm, narrow elliptic, imbricate and greenish on the outer surface; stigma bilobed, 1.5–1.60 mm long. Fruits short ellipsoid, base convex, apex acute, red, resembling fruit of the Murraya exotica clade.

178

Figure 4.43. Murraya × cycloopensis at Bogor Botanic Garden, West Java, cultivated from accession from Cycloop, Papua (accession 24—IP) (Nguyen Huy Chung).

Figure 4.44. Pubescent petiolule and lower mid vein of Murraya × cycloopensis. (accession 24—IP), Bogor Botanic Garden, West Java: cultivated from Cycloop, Papua (Nguyen Huy Chung).

179

Figure 4.45. Scanning electron micrographs of pubescence on rachises and terminal leaflets of the small leaflet form of Murraya × cycloopensis from Papua (accession 24—IP Bogor Botanic Garden ex. Pegunungan Cycloop, Papua, Indonesia).

180

4.3.5.10. Murraya ‘Min-A-Min’

The following description is based on accession 70—ANT from the Northern Territory (see Table 4.1):  Stems: young green pubescent, recently mature stems glabrous;  Leaves: 7 (7–8) foliolate, 55 (47–62) mm long, ratio of leaf length to number of leaflets 7 (7–8);  Rachis: 30 mm (23–35) long, winged, pubescent above and below;  Petiole: 7 (5-9) mm, pubescent above and below;  Acuminate leaflets: 0%;  Basal leaflets: 7 (5–8) × 11 (8–14) mm, elliptic, obtuse and asymmetric base, apices rounded, no emargination, angle of base 52° (38–57), apices rounded, no emargination, proportion of leaflet tip to leaflet length 0%, midvein above pubescent near base, lateral veins glabrous, midveins and lateral veins pubescent below, laminas glabrous above and below, petiolules 1.4 mm (1–1.8), pubescent above and below, margins entire to crenulate, ciliate, sparser distally;  Terminal leaflets: 8 (7–9) × 19 (18–23) mm, obovate, cuneate and symmetric base, angle of base 33° (30–36), apices rounded, unemarginate, proportion of leaflet tip to leaflet length 0%, midvein pubescent, sparser distally above, lateral and minor veins glabrous, similar below, laminas glabrous above and below, petiolules 1.7 mm (1.6–1.9), pubescent, margins entire to crenulate, ciliate, sparser distally.

Flowers terminal or axillary, 4-6 per inflorescence, white and greenish on the outer surface of petals, stamens 10, petals 5; stigma 1 mm long, bilobed, petals 4–5 × 13–14 mm, oblance-ovate or elliptic.

181

4.4. Discussion and conclusions

The results presented in this chapter supported the results of my molecular studies as reported in Chapter 3, corroborating support for Murraya paniculata and Murraya exotica to be regarded as separate species and Merrillia caloxylon as distinct from Murraya. Clear evidence for these separations was found in the analysis of quantitative characters and of the elliptic Fourier descriptors. The results in both chapters also led to identification of a new species, Murraya asiatica, raising to specific rank of Murraya ovatifoliolata var. ovatifoliolata (small and large leaflet forms) and Murraya ovatifoliolata var. zollingeri. Two hybrids were also identified, Murraya × omphalocarpa (formerly Murraya paniculata var. omphalocarpa) and Murraya × cycloopensis.

As in the molecular studies, trees, based in this instance on phenetic analyses of leaflet characters of accessions regarded as Murraya paniculata and Murraya exotica fell into two separate clusters. Merrillia caloxylon, Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri, Murraya × omphalocarpa Murraya × cycloopensis grouped within a clade comprising two sub-clades. In contrast, all Murraya exotica accessions fell into a single clade.

However, the cladogram derived from maximum parsimony was poorly resolved and Murraya paniculata and Murraya asiatica accessions were sister to accessions of Murraya exotica, Murraya ovatifoliolata var. ovatifoliolata, Murraya ovatifoliolata var. zollingeri, Murraya × omphalocarpa, Murraya × cycloopensis, and Merrillia caloxylon.

Discriminant function analysis, based on leaflet characters, also supported my molecular studies. My molecular studies suggested that four clades occur within Murraya:  Murraya exotica  Murraya asiatica;  Murraya paniculata; and  Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri.

When leaflet data from all taxa studied were examined, the analysis correctly assigned 87% of leaves to their taxon and when a consensus was taken for each accession of Murraya exotica, Murraya asiatica and Murraya paniculata these were correctly

182 assigned to their taxon, thus supporting existence of the first three of the clades listed above. Less successful was the assignment of accessions of Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri to their named taxa. However, when the analysis was repeated with these accessions all being classified as a single taxon, 96% of leaflets and all plants were correctly assigned to this taxon. Also, the second choice classification from the discriminant function analysis placed misclassified leaves within the single taxon. These data all support the molecular evidence for Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri forming a clade.

The ordination diagram shows the distribution of the leaves of Murraya × omphalocarpa straddling the distributions of Murraya exotica and the other varieties. Also, when the leaves of Murraya × omphalocarpa were misclassified, they were classified as Murraya exotica. The molecular data suggested that Murraya × omphalocarpa may be a hybrid between Murraya exotica and one of the varieties of Murraya ovatifoliolata and the discriminant function analysis analysis supported this assertion.

Principal component analysis of the elliptic Fourier descriptors of basal and terminal leaflet shapes separated the accessions of Murraya paniculata, Murraya asiatica and Murraya exotica, and also Murraya ovatifoliolata var. ovatifoliolata, particularly the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata from Murraya paniculata, Murraya asiatica and Murraya exotica. Principal component analysis of elliptic Fourier descriptors has been used widely in many biological fields, including plant taxonomy (Iwata et al. 1998, 2002, Yoshioka et al. 2004, Väliranta & Weckström 2007, Andrade et al. 2008). The methods can accurately detect small shape variations that are difficult to detect by human vision (Yoshioka et al. 2004). They have been used to study variation in leaves of Begonia spp. [Curcurbitales: Begoniaceae] (McLellan 1993), roots of radish (Raphanus sativus L. [Brassicales: Cruciferae]) (Iwata et al. 1998), citrus leaves (Iwata et al. 2002), pods of soybean (Glycine max (L.) Merr. [Fabales: Leguminosae]) (Truong et al. 2005) and fruit of wingless birch (Betula spp. [Fagales: Betulaceae]) (Väliranta & Weckström 2007). This type of analysis has been used to evaluate leaf shape in a number of taxonomic studies (e.g. Olsson et al. 2000, Jensen et al. 2002, Viscosi et al. 2009). Fritsch et al. (2009) included principal components analysis of elliptic Fourier descriptors to show that leaf morphology was

183 not a reliable character to distinguish between varieties of eastern redbud (Cercis canadensis L. [Fabales: Leguminosae]) in North America as had been proposed in earlier work (Hopkins 1942, Isely 1975). However, in this study, the analysis of elliptical Fourier descriptors has proved to be a useful method for the separation of the taxa.

Unfortunately, elliptic Fourier descriptors cannot be used for the identification of plants in the field. However, the analysis of leaflet quantitative characters showed similar separation of Murraya paniculata and Murraya asiatica from Murraya exotica, although not as clear as that provided by the elliptic Fourier descriptors. In the redundancy analysis of leaflet quantitative characters, the two most important explanatory variables were the angle of the base of the basal leaflet and the length of the tip of the terminal leaflet (Table 4.11). In the discriminant function analysis of leaflet quantitative characters, the length of the acuminate portion of the basal leaflet and the angle of the base of the basal leaflet were identified as the two most important contributing characters to the classification of the taxa. These characters are readily observable in the field. Basal leaflets of Murraya paniculata and Murraya asiatica have obtuse or rounded bases and acuminate tips and terminal leaflets have acuminate tips; whereas, basal leaflets of Murraya exotica have cuneate bases and rounded to acuminate tips and terminal leaflets have acute to acuminate tips.

My results showed, for the first time, that vegetative characters alone can be used to identify species and varieties of Murraya. This will be welcomed by botanists, entomologists, plant pathologists, chemists and others, as it will allow them to readily identify plants in the absence of flowers and fruit that only occur seasonally or after rains. The most important leaf characters were the ratio of leaf length to the number of leaflets and the proportion of leaflets with or without acuminate tips. Leaves of Murraya paniculata and Murraya asiatica were longer with fewer leaflets and with most leaflets consistently having acuminate apices than were the leaves of Murraya exotica, which were shorter, with more leaflets which had variable apex shapes. The additional characters, the length of acuminate apices of basal and terminal leaflets, and the hairiness of stems, petioles, rachises, petiolules and leaflets, could also be used to aid in the identification of the species and varieties.

My descriptions of Murraya paniculata and Murraya exotica are similar to those of authors, particularly Stone (1985), who regarded these plants as species. 184

Of the leaflets of Murraya paniculata, Jack (1820) stated for Murraya paniculata ‘the number of leaflet is general 5, ovate terminating in a long acumen with the slightly emarginate at the point, the terminal one is the largest’, regarding it as abundantly distinct species from

Murraya exotica the leaflets of which he described as ‘numerous, and closer, obovate, blunt,

and of much firmer thicker substance’.

Stone (1985) described the leaves and leaflets of Murraya paniculata as ‘Leaves alternate, imparipinnate, with, usually 3–5 (rarely 7) leaflets, rarely unifoliolate, to 10–17 cm long: leaflets petiolulate, the petiolules mostly 2–6 mm long: leaflets mostly 3–7 cm long, ovate or ovate‐elliptic, acuminate, at base cuneate to rounded, thinly coriaceous, densely glandular, glossy and darker above, the midrib slightly depressed above, slightly raised beneath, the main lateral nerves about 5–8 pairs, evident on both surfaces as are the reticulations; margins entire or obscurely crenate.’

Jones (1995) described its leaves and leaflets as ‘Leaves pinnate, to 17 cm long; leaflets 3–7, rarely unifoliolate, ovate or ovate-elliptic to rhomboid, 3–7 × 2–3.5 cm, chartaceous or thinly leathery, glossy and darker above, glabrous; base cuneate to rounded, margin entire or faintly crenulate, apex acuminate; lateral veins 5–8 pairs; petiolules 2–6 mm long.’

Huang (1997) described its leaves and leaflets as ‘mature leaves bearing 3–5 leaflets, rarely with 7 leaflets; leaflets dark green, shiny on upper surface, ovate or ovate-lanceolate, 3–9 cm long, 1.5–4 cm wide, the tip narrowly acuminate, rarely mucronate, the base mucronate, the two sides symmetrical or slanted on one side, margins entire, undulating, lateral veins 4–8 on each side; petiolules less than 1 cm long.’

Zhang et al. (2008) described its leaves and leaflets as ‘Leaves 2–5-foliolate; petiolules less than 1 cm; leaflet blades mostly suborbicular to ovate to elliptic, 2–9 × 1.5–6 cm, margin entire or crenulate, apex rounded to acuminate.’

The only descriptions that with any certainty pertain to Murraya paniculata are those of Jack (1820), based on specimens from Sumatra, and Jones (1995), based on wild plants in Sarawak. Descriptions and distributions given by Stone (1985), Huang (1997) and Zhang et al. (2008) relate to Murraya paniculata, Murraya asiatica, Murraya ovatifoliolata and/or Murraya × omphalocarpa. However, descriptions of Murraya exotica by Stone (1985), Huang (1997) and Zhang (2008) were similar to my description of the species.

Stone (1985) described the leaves and leaflets of Murraya exotica as ‘Leaves to 9 cm long, leaflets mostly 3–7 (below inflorescences sometimes 1), alternate or the lowest pair opposite, petioles

185

4–12 mm long, petiolules 1–3 mm long; blades mostly obovate to subelliptic, obtuse to bluntly acuminate with the apex minutely notched, base cuneate; mostly 1–3.5 cm long, 0.9–1.8 cm wide, subentire or the distal margins obscurely crenate; upper surface darker, glossy, lower surface medium semi‐glossy, both glandular punctate (pellucid by transmitted light).’

Huang (1997) described its leaves and leaflets as ‘Leaflets 3–5(–7), obovate or obovate elliptic, asymmetrical, 1–6 cm long, 0.5–3 cm wide, the apex rounded or obtuse, sometimes retuse, the base mucronate, slightly oblique on one side, the margins entire, flat, the petiolules very short.’

Zhang et al. (2008) described its leaves and leaflets as ‘Leaves 3–7-foliolate; petiolules rather short; leaflet blades elliptic-obovate or obovate, 1–6 × 0.5–3 cm, margin entire, apex rounded or obtuse.’

Mention of the nature of hairs by these authors was however scant. For Murraya paniculata and Murraya exotica, Stone (1985) mentioned them as being ‘leaflets

petiolulate’ and ‘short rather ephemeral puberulence of simple curved hairs 0.05–0.1 mm long, on the

stems, petioles, leaf axes, petiolules and midribs’, respectively. Huang (1997) and Zhang et al. (2008) did not mention them, but Jones (1995) noted that leaflets of Murraya paniculata were glabrous, as I observed in this study.

Tanaka (1929) mentioned that the leaflets of Murraya ovatifoliolata var. zollingeri (as

Murraya paniculata var. zollingeri) were ‘puberulent on both sides, ends obtuse, surface somewhat buckled, margin considerably reflexed; rachis very thin and declined, minutely soft tomentose.’

Swingle & Reece (1967) cited Bailey’s (1909) description of Murraya ovatifoliolata var. ovatifoliolata (as Murraya paniculata var. ovatifoliolata) as ‘This, our indigenous form, is of a more straggling habit with more numerous and larger oil‐dots, and is often decidedly hirsute and tomentose, thus very distinct from the two Indian ones of our gardens. The leaves are 3–9‐foliolate; the twigs, calyx, petals, and ovary hirsute.’ This description and the accompanying illustration (Bailey 1909) (Fig. 4.46) suggest that the plant was the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata, not the large leaflet form (see Table 4.4). I assume the ‘two Indian ones’ were the plant identified as Murraya exotica in my studies. There are no records of Murraya paniculata, or of Murraya asiatica, as identified in my studies, in Australia. Frederick Manson Bailey lived in Brisbane, the southeast end of the natural distribution of the small-leaflet form of Murraya ovatifoliolata var. ovatifoliolata.

186

Figure. 4.46. ‘Murraya exotica var. ovatifoliolata’ as illustrated in Bailey (1909).

My observations are the first to be made of an accession of Murraya × cycloopensis. Its leaves were the hairiest of all the Murraya accessions I studied. Leaflet shapes were similar to those of Murraya paniculata and Murraya asiatica, being ovate or elliptic, broad at base and narrow at apex, and larger than those of Murraya exotica. This accession nested with Murraya ovatifoliolata clade in the phenogram constructed using morphological characters and the phylogenetic tree produced from my ITS alignment (Chapter 3). However, it grouped with the Murraya exotica clade in my cpDNA analysis (Chapter 3). The differences could be due to hybridisation between Murraya ovatifoliolata (as the female parent) and Murraya exotica (as the male parent). This interesting plant warrants closer study and further sampling from Papua.

187

The ‘Min-A-Min’ Murraya accession (accession 70) I examined had very small leaves, leaflets and flowers. Hockings (1998) described ‘Min-A-Min’ Murraya as erect, compact, densely branches shrub; leaflets: length short, width narrow, lamina elliptic- obovate, apex acute, base cuneate, undulation margin very weak, fruit absent. It was, according to Hockings (1998) selected from Murraya ovatifoliolata var. ovatifoliolata (as Murraya paniculata var. ovatifoliolata) by propagation of cuttings through 3 generations. However, the image of ‘Murraya paniculata var. ovatifoliolata’ (Fig. 38, Plant Varieties Journal, 1998, Vol 11, No 3) from which ‘Min-A-Min’ Murraya selected resembles Murraya exotica. Moreover, in my molecular phylogeny studies (Chapter 3), ‘Min-A-Min’ fell in the Murraya exotica clade. The leaflet shapes resemble those of Murraya exotica. Therefore, it is a form of Murraya exotica, not Murraya ovatifoliolata var. ovatifoliolata.

Swingle & Reece (1967) suggested that Merrillia caloxylon originated from a Murraya- like ancestor. But et al. (1988) in chemotaxonomic studies confirmed the close relationship between Merrillia and Murraya (sect. Murraya). With the exception of basal leaflets of Merrillia caloxylon, the size and leaflets of Merrillia caloxylon and Murraya paniculata and Murraya asiatica were similar. In my phylogenetic analyses based on 12 leaflet characters Merrillia caloxylon nested with Murraya paniculata and Murraya asiatica accessions in the UPGMA phenogram and with Murraya exotica in my maximum parsimony analysis. However, in my molecular studies Merrillia caloxylon formed a single lineage, sister to Murraya, and Murraya was monophyletic. Merrillia caloxylon inflorescences are structurally simpler and differently positioned, and its flowers and fruit are also larger and morphologically different to those of Murraya. Therefore, Merrillia caloxylon is most probably best treated as a monotypic genus as circumscribed by Swingle & Reece (1967).

188

Chapter 5. Phytochemistry of Murraya paniculata, Murraya exotica and Merrillia

5.1. Introduction

The presence or absence of compounds belonging to a specific class of secondary metabolites, either at the level of family, genus, or species can serve as chemotaxonomic markers indicating that particular biosynthetic pathways have been conserved within a taxon, or alternatively, have arisen two or more times within a taxon through evolutionary convergence (Crockett & Robson 2011). Chemosystematic studies of Aurantioideae have compared secondary metabolites of species of Bergera (s.s.), Clausena, Glycosmis and Micromelum in the Clauseneae and Murraya (s.s.) and Merrillia in the Aurantieae (But et al. 1986, 1988, Kong et al. 1988a, b, Li et al. 1988, Brophy et al. 1994, Samuel et al. 2001). But et al. (1986) reported that roots of Murraya alata, Murraya exotica, Murraya asiatica (cited as Murraya paniculata) and Murraya × omphalocarpa (cited as Murraya paniculata var. omphalocarpa) contained yuehchukene and 8-prenylated coumarins but no carbazoles. In contrast, roots of 4 species of Bergera contained carbazoles but not yuehchukene and 8-prenylated coumarins. Li et al. (1988) reported that the essential oils of three species of Bergera were predominantly monoterpenes, whereas, the oils in leaves of Murraya alata, Murraya exotica and Murraya asiatica (cited as Murraya paniculata) were mostly or entirely sesquiterpenes. Li et al. (1988) also showed the major oil component in Murraya exotica is t-caryophyllene whereas the main oil in Murraya asiatica (cited as Murraya paniculata) is γ-elemene.

But et al. (1988) and Kong et al. (1988a) reported that the roots and bark of Merrillia caloxylon contained the anti-implantation indole alkaloid, yuehchukene and the 8- prenylated coumarins, sibiricin and phebalosin, as well as 3-(3-methyl-buta-1,3-diene) indole and eupatorin. Samuel et al. (2001) reported the presence of 8-phenylated coumarins (including sibricin and phebalosin) and polyoxygenated flavonoids (including euqatorin and nobiletin) in Murraya and Merrillia, and the absence of carbazoles in these genera. In addition to Bergera, they also recorded carbazoles in Clausena, Glycosmis and Micromelum (Samuel et al. 2001). These studies confirmed the close relationship between Merrillia and Murraya, as both contain yuehchukene and the 8-prenylated coumarins, but no carbazole alkaloids.

189

Brophy et al. (1994) reported significant differences between leaf volatiles compounds of the small and large leaflet forms of Murraya ovatifoliolata var. ovatifoliolata (cited as Murraya paniculata var. ovatifoliolata). The oil yield of the small leaflet form was about fivefold that of the large leaflet form. Germacrene-D and E-nerolidol were also more prevalent in the small leaflet form.

In this chapter, I used gas chromatography (GC) to determine differences between accessions identified in my molecular (Chapter 3) and morphological (Chapter 4) studies as belonging to Murraya paniculata, Murraya exotica and Merrillia caloxylon. All accessions of Murraya paniculata used in this analysis were from Indonesia. Peaks with areas ≥ 0.001 were identified and a subset consisting of the major peaks derived and used for statistical analysis. I did not attempt to identify compounds.

5.2. Materials and methods 5.2.1. Gas chromatography

For each of 19 accessions, 20 g of leaflets and 10 g bark were collected and washed under running tap water to remove dirt and then separately immersed in 50 mL of 80% (v/v) ethanol. The leaflets and bark were allowed to remain in the ethanol at room temperature for several months. The ethanol was then decanted from the leaflets, filtered through Whatman No. 1 filter paper and used for GC analysis. The leaflets and bark were then dried at room temperature and immersed in 50 mL n-hexane for 72 h at room temperature. The n-hexane extracts were again filtered and used for GC analysis.

The nineteen accessions comprised:  7 Murraya paniculata accessions 22—IWJ, 25—IWJ, 34—IEJ, 38—IEJ, 45— ICJ, 46—ICJ and 48—ICJ;  11 Murraya exotica accessions 26—IWJ, 27—IWJ, 35—IEJ, 37—IEJ, 40—IC, 41—IUCR, 42—IUCR, 43—IC, 44—ICJ, 47—ICJ and 53—ANSW; and  1 Merrillia caloxylon accession 23—IWJ.

Murraya exotica accessions 26—IWJ, 41—IUCR and 43—IC were not used for molecular or morphological studies reported in Chapter 3 and Chapter 4, respectively.

GC was performed using a HP 5890 chromatograph (Hewlett Packard, USA) fitted with a flame ionisation detector. Each sample (2 µL of extract) was injected onto the

190 capillary column using a microsyringe with hydrogen as the carrier gas. The oven temperature was programmed as follows:  for the ethanol extracts: isothermal at 80ºC for 2 min, then rising to 300ºC at 10ºC per min with an isothermal period (300ºC) for 15 min; and  for the n-hexane extracts: isothermal at 70ºC for 2 min, then rising to 300ºC at 10ºC per min with an isothermal period (300ºC) for 15 min.

The chromatograms, including the retention time (RT) and peak areas (PA), were recorded and printed by an integrating recorder. To select the major compounds found amongst the accessions, the following procedure was performed. For each compound giving a peak area ≥ 0.001, the peak areas were summed for each accession following which the total areas for each compound were also summed (total peak area). From this data, the percentage of the total peak area contributed by each compound was calculated. Compounds contributing > 0.5% were then used for analysis by redundancy analysis using Canoco and for multidimensional scaling and cluster analysis using Primer 6.

5.2.2. Statistical analyses

Both sets of data (all peaks ≥ 0.001 and the major peaks) from the GC were subjected to three types of statistical analysis: redundancy analysis (RDA), MANOVA and multidimensional scaling (MDS).

For redundancy analysis, peak areas were treated as ‘species relative abundance’ and used to form the ‘species’ matrices used by CANOCO for Windows (Version 4.5, ter Braak & Šmilauer 2002). When necessary, eigenvectors were corrected for negative values using Euclidean distance and then the redundancy analysis axes were exported to CANOCO and treated as ‘species’ data. To test the effect of taxon (Murraya paniculata, Murraya exotica and Merrillia caloxylon), data coding the taxa were entered as dummy binary variables to create the ‘environmental’ matrices used by CANOCO. Relationships between chemical composition and taxa were determined by redundancy analysis and the significance of such models was tested with Monte Carlo test based on 999 permutations.

191

For MDS, data sets were analysed using Primer 6 (PRIMER-E Ltd., Ivybridge, United Kingdom). After entry, peak areas were converted to presence/absence data and Bray Curtis distances calculated. The resulting similarity matrices were then used for MDS.

MANOVA was performed using Statistica (StatSoft Inc. Version 9.1).

5.3. Results

Numbers of peaks with peak areas ≥ 0.001 on chromatograms of bark and leaflet extracts of accessions of Murraya paniculata, Murraya exotica and Merrillia caloxylon are shown in Table 5.1. Numbers of peaks that were common or unique amongst these three taxa are shown in Table 5.2. Chromatograms of typical bark extracts of the taxa made with either ethanol or n-hexane are shown in Figs 5.1–5.3 and Figs 5.4–5.6, respectively. For the accessions of Murraya paniculata and Murraya exotica, the bark extracts contained fewer compounds than comparable extracts from leaflets of the same taxa, as did the extracts from Merrillia caloxylon made with n-hexane. However, the ethanol extracts of Merrillia caloxylon bark and leaflets contained approximately the same number of peaks.

Amongst the accessions of Murraya paniculata and Murraya exotica, there was considerable variation in total numbers of peaks and peaks with peak area values ≥ 0.001 in both the ethanol and n-hexane bark extracts. In particular, the ethanol bark extracts of Murraya paniculata accession 22—IWJ and Murraya exotica accessions 26—IWJ and 27—IWJ from West Java contained few compounds. Furthermore, for the Murraya exotica, the Javanese accessions tended to have fewer compounds in these extracts than accessions from the United States of America, China and Australia. Amongst the three taxa, the Murraya exotica accessions had a higher total number of compounds than did the Murraya paniculata and Merrillia caloxylon accessions. The accession of Merrillia caloxylon appeared to have a similar number of compounds to the Murraya exotica accessions. The comparison of the distribution of compounds (Table 5.2) within the three taxa showed that Murraya exotica is more chemically diverse with 39 and 49 of the unique compounds identified in the ethanol and n-hexane bark extracts, respectively, and 29 and 23 identified in the ethanol and n-hexane leaflet extracts, respectively. In contrast, ethanol and n-hexane bark and leaflet extracts of the Murraya paniculata accessions contained 3 and 3, and 21 and 5 unique compounds,

192 respectively. The Merrillia caloxylon accession yielded only 11 unique bark compounds and 5 leaflet compounds.

Chromatograms of typical leaflet extracts of the taxa made with either ethanol or n- hexane are shown in Figs 5.7–5.9 and Figs 5.10–5.12, respectively. Variation the numbers of peaks with peak area values ≥ 0.001 in bark extracts was more apparent than in leaflet extracts (Table 5.1). Numbers of peaks in the four extracts from Merrillia caloxylon were within the ranges seen in the other two taxa. The relatively lower number of compounds extracted from the bark of West Javanese accessions of Murraya paniculata and Murraya exotica compared to accessions of these taxa from elsewhere was not evident in the leaflet results. Indeed, the ethanol extract of the West Javanese Murraya paniculata accession (22—IWJ) shared the highest number (146) of peaks with Murraya paniculata accession 46—ICJ from Central Java (Table 5.1). Within the Murraya exotica accessions, there was a tendency for less chemical complexity in the ethanol and n-hexane bark and leaflet extracts from Java accessions than in extracts from accessions from other locations.

Both the complete data sets from the four extracts and the data sets representing the major peaks were subjected to MANOVA; no significant were found amongst the three taxa in any analysis. The ordination diagrams from the redundancy analysis (Figs 5.13 and 5.14) suggested that chemical composition can be used to separate the taxa. However, the Monte Carlo permutation tests (Tables 5.3 and 5.4) gave little support for the separation, with only the peaks from the ethanol extract of the leaflets separating Murraya exotica from Murraya paniculata. When data were converted from peak areas to presence/absence data, both the ethanol (Fig. 5.15) and n-hexane (Fig. 5.16) extracts subjected to multidimensional scaling showed a clear separation of all three taxa.

193

Table 5.1. Number of peaks on chromatograms of bark and leaflet extracts of accessions of M. paniculata, M. exotica and Me. caloxylon.

Accession Taxon Bark ethanol extract Bark n‐hexane extract Leaf ethanol extract Leaf n‐hexane extract Number Number of Number Number of Number Number of Number of Number of of peaks peaks with area of peaks peaks with area of peaks peaks with area peaks peak with area value ≥ 0.001 value ≥ 0.001 value ≥ 0.001 value ≥ 0.001 22—IWJ M. paniculata 20 5 12 10 146 102 78 72 25—IWJ M. paniculata 51 23 11 10 86 49 67 63 34—IEJ M. paniculata 35 18 7 7 122 82 74 64 38—IEJ M. paniculata 33 16 7 7 128 83 89 83 45—ICJ M. paniculata 40 18 3 3 95 44 21 20 46—ICJ M. paniculata 74 36 6 6 146 94 133 119 48—ICJ M. paniculata 43 22 7 7 106 67 91 87 Average 42.3 19.7 7.6 7.1 118.4 74.4 79.0 72.6 26—IWJ M. exotica 12 5 16 14 128 78 90 83 27—IWJ M. exotica 14 7 10 8 123 73 ‐‐ 35—IEJ M. exotica 36 19 12 10 107 68 58 49 37—IEJ M. exotica 32 12 12 9 105 72 80 68 40—IC M. exotica 104 69 40 31 133 84 122 108 41—IUCR M. exotica 103 76 12 12 123 80 88 71 42—IUCR M. exotica 77 47 3 3 133 90 130 118 43—IC M. exotica 90 60 13 13 132 93 117 109 44—ICJ M. exotica 85 51 15 15 101 47 93 85 47—ICJ M. exotica 44 18 3 3 97 58 100 88 53—ANSW M. exotica 63 32 48 43 114 80 125 113 Average 60.0 36.0 16.7 14.6 117.8 74.8 100.3 89.2 23—IWJ Me. caloxylon 53 52 14 14 88 48 40 36

194

Table 5.2. Number of peaks with peak areas ≥ 0.001 in the chromatograms of bark and leaflet extracts of accessions of M. paniculata and M. exotica and Me. caloxylon shared between all three forms, between two forms or unique to a particular taxon.

Extract Total Present in Present in Present in Present in Present in Present in Present in number all 3 taxa M. paniculata M. paniculata and M. exotica and Me. M. exotica M. paniculata Me. caloxylon and M. exotica Me. caloxylon caloxylon Bark 139 19 45 2 22 39 3 9 ethanol Bark n‐ 74 6 8 1 5 49 3 2 hexane Leaf ethanol 201 32 103 9 4 29 21 3 Leaf n‐ 192 29 128 1 4 23 5 2 hexane

195

Figure 5.1. Chromatograms of ethanol extracts of bark from M. paniculata accessions 25—IWJ, 34—IEJ and 48—ICJ.

196

Figure 5.2. Chromatograms of ethanol extracts of bark from M. exotica accessions 26—IWJ, 37—IEJ and 47—ICJ.

197

Figure 5.3. Chromatogram of ethanol extracts of bark from Me. caloxylon accession 23—IWJ.

198

Figure 5.4. Chromatograms of n‐hexane extracts of bark from M. paniculata accessions 25—IWJ, 34—IEJ and 38—ICJ.

199

Figure 5.5. Chromatograms of n‐hexane extracts of bark from M. exotica accessions 27—IWJ, 37—IEJ and 43—IC.

200

Figure 5.6. Chromatogram of an n‐hexane bark extract from Me. caloxylon accession 23—IWJ.

201

Figure 5.7. Chromatograms of ethanol extracts of leaflets from M. paniculata accessions 25—IWJ, 34—IEJ and 48—ICJ.

202

Figure 5.8. Chromatograms of ethanol extracts of leaflets from M. exotica accessions 26—IWJ, 37—IEJ and 47—ICJ. 203

Figure 5.9. Chromatogram of an ethanol extracts of leaflets of Me. caloxylon accession 23—IWJ.

204

Figure 5.10. Chromatograms of n‐hexane extracts of leaflets of M. paniculata accessions 25—IWJ, 34—IEJ and 48—ICJ.

205

Figure 5.11. Chromatograms of n‐hexane extracts of leaflets of M. exotica accessions 26—IWJ, 37—IEJ and 47—ICJ. 206

Figure 5.12. Chromatogram of n‐hexane extracts of leaflets of Me. caloxylon accession 23—IWJ.

207

Figure 5.13. Ordination diagrams from redundancy analysis of peak areas of compounds in ethanol (left) and n‐hexane (right) extracts from bark of M. paniculata, M. exotica and Me. caloxylon accessions. The analyses were based on data from all peaks ≥ 0.001 in area.

208

Figure 5.14. Ordination diagrams from redundancy analysis of peak areas of compounds in ethanol (left) and n‐hexane (right) extracts from leaflets of M. paniculata, M. exotica and Me. caloxylon accessions. The analyses were based on data from all peaks ≥ 0.001 in area.

209

Table 5.3. Monte Carlo permutation tests from the redundancy analysis based on peak areas for all compounds with peak areas ≥ 0.001. Figures in parenthesis are the number of compounds identified in each combination of tissue source and solvent.

Taxon Source/Solvent Bark/ethanol Bark/hexane Leaflet/ethanol Leaflet/hexane (n = 139) (n = 74) (n = 201) (n = 192) F P F P F P F P Merrillia 16999 0.058 2.073 0.109 1.070 0.257 0.153 0.957 caloxylon Murraya exotica 1.405 0.433 1.536 0.221 3.357 0.001 0.438 0.914 Murraya 0.570 0.833 2.105 0.089 2.634 0.011 0.387 0.935 paniculata

Table 5.4. Monte Carlo permutation tests from the redundancy analysis based on the peak areas for the major compounds. Figures in parenthesis are the number of compounds identified in each combination of tissue source and solvent.

Taxon Source/Solvent Bark/ethanol Bark/hexane Leaflet/ethanol Leaflet/hexane (n = 27) (n = 37) (n = 47) (n = 33) F P F P F P F P Merrillia 16999 0.050 2.111 0.109 1.005 0.257 0.148 0.957 caloxylon Murraya exotica 1.405 0.433 1.552 0.220 3.488 0.001 0.397 0.931 Murraya 0.570 0.830 2.146 0.087 2.619 0.012 0.352 0.947 paniculata

210

Figure 5.15. Relationships determined by multidimensional scaling among the 19 accessions of M. exotica, M. paniculata and Me. caloxylon based on leaflet compounds extracted with ethanol: all peaks ≥ 0.001 (above) major peaks (below).

211

Figure 5.16. Relationships determined by multidimensional scaling among the 19 accessions of M. exotica, M. paniculata and Me. caloxylon based on leaflet compounds extracted with n‐hexane: all peaks ≥ 0.001 (above) major peaks (below).

212

5.4. Discussion

As discussed in Chapters 1 and 2, the systematics of the genus Murraya is controversial with the most problematic issue being the taxonomic status of Murraya paniculata and Murraya exotica. In this section of my research I used GC to analyse extracts of stem bark and leaflets of 7 Murraya paniculata accessions from Indonesia, 11 Murraya exotica accessions from Australia, China, Indonesia and the United States of America29, and a single Merrillia caloxylon accession from Indonesia to determine differences and similarities in their chemical composition in order to provide further evidence for the taxonomic status of the taxa. Plants of closely related taxa may have similar compounds; therefore, the presence or absence of particular compounds may be used in chemotaxonomy. Wink (2003) studied the profiles of secondary metabolites in Labiatae, Leguminosae and Solanaceae and concluded that in some instances, almost all members of a monophyletic clade share a particular chemical characteristic; on the other hand, particular secondary metabolites may occur in several unrelated clades or plant families. Nyman & Julkunen-Tiitto (2005) stated that the taxonomic distribution of individual secondary compounds varies considerably and some compounds can be found in many distantly related plant groups whereas others may occur only in a few closely related taxa or a single species. Wink (2003) and Nyman & Julkunen-Tiitto (2005) both concluded that chemical data are poorly suited for phylogenetic studies and should be used cautiously when inferring phylogenetic relationship among plant taxa. Waterman (1990) noted that although secondary metabolite expression in plants is genetically controlled, it is subject to the same phenotypic variation as any other character and he saw no justification for weighting these chemical data any more than the more traditional characters used in plant taxonomy. In a review of the status of chemical systematics, Waterman (2007) concluded that chemical systematics provided useful insights into plant phylogeny.

Identical retention times for two compounds separated by GC indicate a possibility that they are the same substance. This in not necessarily so, but in this analysis, peaks of different samples extracted in the same solvent and run using the same GC parameters were considered as the same compound. In my study, 74 to 201 compounds were found in the four different extracts. The Murraya exotica accessions were more chemically diverse than those of Murraya paniculata, with larger average numbers of compounds

29 The accessions from China, Indonesia and the United States of America were all growing in the field site in Java, 213 in the four extracts. Moreover, 23 to 39 unique compounds were detected in the extracts from the Murraya exotica accessions compared to 3 to 21 unique compounds found in the extracts from the Murraya paniculata accessions. Identification of these compounds, which was beyond the scope of my study, may permit the detection of one or more compounds to be used as a means of distinguishing the two taxa, and distinguishing them from Merrillia caloxylon. Unfortunately, only one accession of Merrillia caloxylon was available for study making comparisons with the other taxa difficult. However, the data suggest that the bark of Merrillia caloxylon has similar or greater chemical complexity than the accessions of Murraya, but the composition of the leaves seems less complex. Li et al. (1988) examined the compounds in distilled oils from Bergera and Murraya. They found 50 compounds in the three species of Murraya (Murraya alata, Murraya exotica and Murraya asiatica, the latter cited as Murraya paniculata), 19 of which were found in Murraya exotica and 22 in Murraya asiatica. Raina et al. (2006) studied the composition of the essential oil from leaves and flowers of Murraya exotica growing in India and found that there were 56 and 72 constituents, respectively, in the leaf and flowers oil of which 40 compounds were common to both oils. Olawore et al. (2005) found 42 compounds in leaf and fruit oils of ‘Murraya paniculata’ but presumably Murraya exotica. Thus, the number of compounds found in this study was similar to those reported in other studies.

Ge et al. (2008) studied the chemotaxonomy of Taxus [Pinales: Taxaceae] species and found that the variability of metabolites in Taxus was caused by many non-genetic factors such as development stage of the plants and the climate, elevation, slope and exposure of the location where they were grown. In this study, accessions 40, 41, 42, 43 and 44 of Murraya exotica and 45 of Murraya paniculata were taken from a field trial conducted in Central Java. These plants were all grown from seed at the same time and were grown in a randomised block design within the trial site. As a consequence, the plants were all exposed to similar edaphic and climatic conditions. Therefore, the variation in chemical composition seen between the Murraya exotica and Murraya paniculata accessions and amongst the Murraya exotica accessions is most likely due to genetic differences.

Although I was not able to identify compounds I extracted from plants I used in this study, statistical analyses of the data provided further evidence for the separation of the three taxa studied. However, no one compound was able to clearly separate Murraya

214 exotica from Murraya paniculata, as no single compound was found in all accession of one taxon that was absent from all accessions of the second. The bark extracts provided no differentiation between Murraya exotica from Murraya paniculata; this may be due to the relative chemical simplicity compared to the extracts from the leaflets. Statistical analyses of the leaflet extracts, particularly the ethanol extract, using either all peaks ≥ 0.001 or only the major peaks separated Murraya exotica from Murraya paniculata, as well as separating Merrillia caloxylon from these taxa. Although the evidence from these analyses is not strong, differences in phytochemistry support the results of my molecular (Chapter 3) and morphological (Chapter 4) studies in separating Murraya exotica from Murraya paniculata.

215

Chapter 6: Susceptibility of taxa to huanglongbing 6.1. Introduction

In Chapters 1 and 2, I briefly reviewed the causes of HLB, vectors of the pathogens that cause the disease, and impacts and threats of the disease on citriculture and citrus germplasm. In these chapters, I also reviewed taxonomic confusion surrounding the status of Murraya exotica (sensu Huang (1997) and Zhang & Hartley (2008)) as a species, and mentioned that the three pathogens that cause the disease, ‘Candidatus Liberibacter asiaticus’, ‘Candidatus Liberibacter americanus’ and ‘Candidatus Liberibacter africanus’ (Bové 2006), have only been detected in Murraya exotica.

In this chapter, I used PCR to determine the presence of ‘Candidatus Liberibacter asiaticus’ in Murraya accessions, that in my molecular (Chapter 3) and morphological studies (Chapter 4) fell into two major clades, one comprising Murraya exotica accessions, the other Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions. All Murraya exotica accessions resembled plants widely grown as ornamentals. With the exception of Murraya paniculata accessions from Java, all Murraya paniculata accessions were from plants growing in the wild. I also tested Merrillia caloxylon (accession 23—IWJ from the Bogor Botanic Garden, Java, Indonesia) for presence of ‘Candidatus Liberibacter asiaticus’.

My original objective was to grow and expose accessions to HLB under field conditions in Việt Nam. However, shortly after commencing my studies reports of possible transmission of ‘Candidatus Liberibacter asiaticus’ in seeds were published (Benyon et al. 2008, Graham et al. 2008, Hartung et al. 2008, Shatters 2008). This led me to curtail this part of my studies owing to quarantine concerns about possibility of transporting different strains of HLB to Việt Nam. Therefore, I restricted my assessments of the susceptibility of Murraya and Merrillia to HLB to the accessions I used in my thesis.

6.2. Materials and methods 6.2.1. Plant materials

Leaflets from 88 accessions were tested. Based on results reported in Chapters 3 and 4 these comprised:

216

 1 accession of Merrillia caloxylon from Indonesia;  8 accessions of Murraya paniculata from Indonesia  5 accessions of Murraya asiatica (3 from China and 2 from Việt Nam);  3 accessions of Murraya × omphalocarpa from Taiwan;  1 accession of Murraya ovatifoliolata var. ovatifoliolata (University of California, Riverside ex Bundaberg, Queensland); and  70 accessions of Murraya exotica—20 from Brazil, 10 from China, 21 from Indonesia, 3 from the United States of America (University of California, Riverside) and 16 from Việt Nam.

6.2.2. DNA extraction

Total genomic DNA was extracted using the modified methods of Doyle & Doyle (1990) and Warude et al. (2003), Murray & Thomson (1980), or using the DNeasy Plant Minikit (Qiagen) and the HP Plant DNA Kit (Omega Bio-Tek) (see Chapter 3).

6.2.3. HLB detection

Nested PCR was performed to detect the presence of ‘Candidatus Liberibacter asiaticus’ using: primers F1, R1 & F2, R2 (Table 6.1); Taq DNA polymerase (New

England Biolabs); Thermopol buffer (New England Biolabs, [MgSO4] = 2 mM); an equimolar mix of dNTPs (Fisher Biotech); and acetylated bovine serum albumin (BSA) (Promega) as an enzyme stabilizer. The test comprised 2 rounds of PCR:  External round: The PCR reaction mixture for single step PCR contained 2.5 μL of 10 × Thermopol buffer, 10 mM dNTPs, 2.5 U Taq polymerase, 10 μM each primer F1 (forward) and R1 (reverse), 5 μg BSA and 50 ng of template DNA in a total reaction volume of 25 μL. The thermal cycling parameters were: an initial denaturation for 3 min at 95°C; 35 cycles of 94 °C for 0.5 min, annealing at 53°C for 0.5 min, and elongation at 72°C for 1 min; followed by an elongation step of 72°C for 10 min.  Internal round: The nested PCR reaction mixture was made up in a volume of 25 μL, containing 2.5 μL of 10 × Thermopol buffer, 10 mM dNTPs, 2.5 U Taq polymerase, 10 μM each primer F2 (forward) and R2 (reverse), and 1 μL of the first step PCR product. The thermal cycling parameters were: an initial denaturation for 3 min at 95°C; 35 cycles of 94°C for 0.5 min, annealing at 55°C

217

for 0.5 min, and elongation at 72°C for 1 min; followed by an elongation step of 72°C for 10 min. The expected size of PCR products is 400 bp (Ding et al. 2005).

DNA from a HLB-infected ‘Siem’ mandarin (Citrus reticulata Blanco syn. Citrus suhuiensis hort ex. Tanaka) tree from Purworejo (07°43' S, 109°56' E), Central Java, Indonesia was used as a positive control and water as a negative control.

Table 6.1. Primer sequences (5’ to 3’) used for nested PCR (Ding et al. (2005).

Primers Sequences

F1 TGAATTCTTCGAGGTTGGTGAGC R1 AGAATTCGACTTAATCCCCACCT F2 GCGTTCATGTAGAAGTTGTG R2 CCTACAGGTGGCTGACTCAT

6.3. Results

Nested PCR successfully amplified a fragment of 400 bp from the positive control (Fig. 6.1). No amplicons were obtained from the negative control. Two of the Murraya exotica accessions tested positive for ‘Candidatus Liberibacter asiaticus’: accession 102—BSP from Capão Bonito, São Paulo, Brazil, and accession 62—CGD from South China Agricultural University, Guangzhou, Guangdong, China. None of the Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata accessions tested positive. The single Merrillia caloxylon accession also tested negative.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Figure 6.1. External round of nested PCR; lane 1: 100bp DNA ladder; lane 2: negative control; lane 3: positive control; lanes 4–11 and 13–20: Murraya samples; lane 12: infected Murraya sample.

218

The 400 bp amplicons produced by nested PCR from the ‘Candidatus Liberibacter asiaticus’-positive accessions were sequenced and analysed using BLAST searches at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/). The results showed 99% of nucleotides were identical with the trmU-tufB-secE-nusG- rplKAJL-rpoB region of ‘Candidatus Liberibacter asiaticus’ from Genbank accession AB480155 (Tomimura et al. 2009). This confirmed that the pathogen detected in accessions 62 and 102 was ‘Candidatus Liberibacter asiaticus’.

6.4. Discussion

My results confirmed that Murraya exotica can harbour ‘Candidatus Liberibacter asiaticus’ but only 2 of 67 accessions (2.98%) from locations in Brazil, China, Indonesia and Việt Nam where HLB is prevalent tested positive for the pathogen. Only one of these accessions (62—CGD from South China Agricultural University: see Fig. 2.6) was symptomatic for the disease. The low percentage of positive tests was probably related to the incidence of ‘Candidatus Liberibacter asiaticus’-infected Diaphorina citri in the immediate vicinity of the plants and persistence of ‘Candidatus Liberibacter asiaticus’ in plants after transmission of the pathogen by Diaphorina citri. Studies reported by Lopes et al. (2010) indicated that Murraya exotica does not favour liberibacter multiplication as much as citrus. Lopes et al’s (2010) studies also suggested that ‘Candidatus Liberibacter americanus’ and ‘Candidatus Liberibacter asiaticus’ interact with orange jasmine as opportunistic pathogens. Furthermore, observations (Silvio Lopes, Fundecitrus, pers. comm. with Andrew Beattie, December 2006, April 2010; Deng Xiaoling, South China Agricultural University, pers. comm. with Andrew Beattie, July 2007) suggest that Murraya exotica is a transient, asymptomatic and symptomatic host of citrus liberibacters. As I mentioned in Chapter 2, all reports (including Li & Ke 2002, Damsteegt et al. 2010, Gasparoto et al. 2010) of species of Murraya harbouring liberibacters pertain to Murraya exotica. I did not detect the pathogen in Murraya paniculata, Murraya asiatica and Murraya ovatifoliolata, or in Merrillia caloxylon. However, I only evaluated 8 Murraya paniculata and 5 Murraya asiatica accessions, 1 Murraya ovatifoliolata accession, and one Merrillia caloxylon accession. Further studies need to be undertaken regarding the host status of Murraya and Merrillia.

219

Chapter 7: General Discussion

My specific objectives were to:  characterise and delimit species and forms of Murraya, focusing on Murraya paniculata, and to reconstruct phylogenetic relationships among taxa, based on chloroplast and ITS sequences (Chapter 3);  characterise and delimit species and forms of Murraya, focusing on Murraya paniculata, and to reconstruct phylogenetic relationships among taxa, based on morphology (Chapter 4);  test whether Murraya paniculata and Murraya exotica are differentiated on the basis of phytochemistry (Chapter 5); and  evaluate the susceptibility of worldwide-collected accessions of Murraya species and Merrillia caloxylon to HLB (Chapter 6).

I hypothesised that Murraya paniculata and Murraya exotica are species and that Murraya exotica is not a variety of Murraya paniculata (Hypothesis 1). I also hypothesised that Merrillia caloxylon may be a species of Murraya (ss) (Hypothesis 2).

7. 1. Hypothesis 1

In my molecular studies (Chapter 3), my accessions of Murraya paniculata sensu Swingle & Reece (1967) that resembled Murraya exotica sensu Jones (1985) and Huang (1997) were clearly separated from accessions that resembled Murraya paniculata sensu Jones (1985) and Huang (1997). This was true for the phylogenetic trees from 6 maternally-inherited chloroplast genes (trnL-F, psbM-trnDGUC, trnCGCA- ycf6, rps4-trnT, matK-5′trnK and rps16) and for pooled data, and for the internal transcribed spacer (ITS) region of the nuclear-encoded ribosomal RNA operon.

The phylogenetic trees derived from analysis of key leaflet characters (Chapter 4) also separated accessions that resembled Murraya exotica sensu Jones (1985) and Huang (1997) from those that resembled Murraya paniculata sensu Jones (1985) and Huang (1997). The accessions grouped into two distinct clades, as they did in the molecular studies. Principal component analysis (PCA) and redundancy analysis (RDA) of elliptic Fourier descriptors and discriminate function analysis (DFA) of leaflet dimensions also separated the taxa. Moreover, I determined leaflet characters for readily distinguishing

220 the taxa, including the Murraya paniculata varieties sensu Swingle & Reece (1967) and Murraya × cycloopensis (accession 24—IP) from Papua in New Guinea.

Further evidence of significant differences between Murraya paniculata (sensu Jack (1820)) and Murraya exotica was found in the chromatograms of ethanol and n-hexane extracts of leaflets and bark, and analysis of the presence or absence of these compounds as revealed by multidimensional scaling (MDS) (Chapter 5).

Murraya exotica accessions from China were dispersed throughout the Murraya exotica clades derived from the cpDNA and the ITS analyses, whereas those from other countries were restricted to certain locations within the clades. The distributions of these accessions would fit with a Chinese origin for Murraya exotica. Although my results showed variation within Murraya exotica, this variation was insufficient to support any further subdivision of this taxon.

I obtained no evidence to support the classification of Murraya exotica as Murraya paniculata, Murraya asiatica, or as a subspecies or variety of either Murraya paniculata or Murraya asiatica. Thus, my results support the views of Jones (1985), Huang (1997), Samuel et al. (2001), Subandiyah et al. (2007), Zhang et al. (2007) and Verma et al. (2009) who considered Murraya exotica to be a distinct species. My results overwhelmingly supported Hypothesis 1. However, my results showed considerable variation among the accessions of Murraya paniculata. This is discussed below.

7.2. Hypothesis 2

Analysis of the combined chloroplast genes and the ITS region (Chapter 3) placed Merrillia caloxylon as sister to the Murraya paniculata and Murraya exotica accessions. Although there were similarities in leaf and leaflet characters (Chapter 4), the morphology of the flowers and fruit are substantially different from those of Murraya (s.s) (Swingle & Reece 1967). This suggests that Merrillia caloxylon should remain as a single member of the genus Merrillia. My results did not support Hypothesis 2.

221

7.3. Placement of other Murraya taxa

My molecular (Chapter 3) and morphological (Chapter 4) studies showed that Murraya asiatica accessions from mainland Asia (Việt Nam and China) and Murraya paniculata accessions from the Indonesian archipelago represent distinct species, the former not previously recognised and possibly representing more than one taxon. The molecular and morphological data also showed that Murraya paniculata var. zollingeri and Murraya paniculata var. ovatifoliolata (small and large leaflet forms) sensu Swingle & Reece (1967) are varieties of Murraya ovatifoliolata (Engl.) Domin., viz., Murraya ovatifoliolata var ovatifoliolata (small and large leaflet forms) and Murraya ovatifoliolata var. zollingeri. Analysis of the sequence data from the ITS region suggested that Murraya ovatifoliolata may be more distantly related to Murraya exotica than to Murraya asiatica. The phenogram derived from the chloroplast sequences showed that the accessions of Murraya ovatifoliolata var. ovatifoliolata were grouped based on location, with accessions from the Bundaberg region grouping together, accessions from near Cairns grouping together, and accessions from the Northern Territory grouping together. However, the ITS region also suggested that Murraya ovatifoliolata var. ovatifoliolata from the Northern Territory may be more closely related to Murraya ovatifoliolata var. zollingeri than to Murraya ovatifoliolata var. ovatifoliolata from Queensland. Amongst the accessions of Murraya ovatifoliolata var. ovatifoliolata were the large and small leaflet forms; these forms grouped with other accessions from their place of origin. Thus, this variation in leaflet form is not of use taxonomically. It is of interest to note here that Mabberley (1998) raised the possibility of a relationship between the small leaflet form of Murraya ovatifoliolata var. ovatifoliolata and Murraya ovatifoliolata var. zollingeri. Clearly, the exact relationship amongst plants from Australia and the Indonesian archipelago needs further examination.

Murraya × omphalocarpa fell in the Murraya ovatifoliolata clade in the analysis of chloroplast DNA regions and in the Murraya exotica clade in the analysis of the ITS region. Conversely, Murraya × cycloopensis from Papua fell in the Murraya exotica clade in the chloroplast analysis and the Murraya ovatifoliolata clade in the ITS analysis. These results suggest that Murraya × omphalocarpa may be a hybrid with Murraya ovatifoliolata as the female parent and Murraya exotica as the male parent, and conversely, that Murraya × cycloopensis from Papua may be a hybrid with 222

Murraya ovatifoliolata as the male parent and Murraya exotica as the female parent. Further work with a wider range of accessions would be necessary to determine such relationships. In addition, other nuclear genes need to be sequenced to confirm the phylogeny derived from ITS sequences.

7.4. Conclusions

Species concepts and species delimitation are controversial (Levin 1979, Weins & Servedio 2000, Rieseberg & Burke 2001, Agapow et al. 2004, Rieseberg et al. 2006, de Queiroz 2007, Duminil & Di Michele 2009, Hausdorf 2011, Ley & Hardy 2010). Until the advent of molecular techniques, classification of plants relied initially on perceived morphological differences (Duminil & Di Michele 2009) and then more rigorous numerical taxonomy based on quantitative analyses of as many characters as possible (Rieseberg et al. 2006). However, variable morphological characters can reflect local adaptations to an environment (Duminil & Di Michele 2009, Fritsch et al. 2009) while in others, similarities can fail to separate taxonomic entities, thereby concealing cryptic species (Duminil & Di Michele 2009). Molecular phylogenetics has widened the scope for delimiting species of plants, but conclusions reached may not concur with those made by other means (Agapow et al. 2004). A literature survey by Agapow et al. (2004) showed a 48% increase in species delimited by phylogenetic studies compared to those delimited in morphological studies. New insights from Mendelian, molecular and biochemical genetics have revealed that speciation is not a one step phenomenon but an ongoing process, and morphological characters alone are not sufficient anymore to properly describe the results of this process (Ley & Hardy 2010). Furthermore, morphological characters might be dependent on external environmental factors without being genetically fixed and processes of genetic divergence of populations might occur without leaving traces in morphological characters (Ley & Hardy 2010).

My studies have compared morphological characters (leaf and leaflet dimensions and ratios), 6 maternally inherited chloroplast genes, sequences of the ITS region from the nuclear genome, and biochemical properties of Murraya accessions. The data from these studies are in concordance and most of the accessions I used were sourced from a wide range of environments, although some were growing in the same environment. Thus, there is strong support for the taxa I propose. However, additional support should be sought through the examination of flower and fruit characters. In addition, other

223 authors have proposed several taxa/species within Murraya (or the genera to which plants of this genus were previously assigned) and this is discussed below.

The history of the circumscription of Murraya paniculata and Murraya exotica is confused and confusing and has been so since Rumphius collected specimens of Camunium vulgare and Camunium japonense in the late 1600s. Many descriptions have been brief, using too few characters to accurately identify taxa. Rumphius’ description of Camunium vulgare and his (inaccurate) illustration of it in ‘Herbarium Ambionense’ (Rumphius 1747), appears to represent Murraya paniculata (Jack 1820). His description of Camunium japonense (javanicum) resembles Murraya exotica. Burman (1768) described a plant from Java as Chalcas camuneng in his Flora Indica. Rumphius’ plants were subsequently described by Linnaeus (1771) as Chalcas paniculata and Murraya exotica, respectively.

However, according to Merrill (1917), Linnaeus’ description of Chalcas paniculata (Linnaeus 1771) was based on Rumphius’ description of Camunium vulgare and Burman’s (1768) description of specimens that were primarily from cultivated plants growing in Java. Merrill (1917) concluded that these specimens were probably Murraya exotica L. Thus, Linnaeus’ description of Chalcas paniculata (Linnaeus 1771) was flawed, as it was based on two taxa, Murraya paniculata and Murraya exotica. In addition, Linneaus also described a second species, ‘Murraea’ exotica, based on material from König and from descriptions by Burman. Merrill (1917) also noted that Lamarck (1797) reduced Camunium japonense to Murraya exotica L.

Jack (1820) was emphatic in his views of the status of Murraya paniculata and Murraya exotica, regarding them as distinct, and my descriptions of them are in broad agreement with his. Merrill (1952) stated that Jack’s descriptions of Murraya paniculata (L.) Jack was apparently based on material from Penang and Singapore, and noted that Jack considered Chalcas paniculata Lour. (Chalcas paniculata L.) and Camunium vulgare Rumph. as synonyms of the species. Other records suggest that Jack’s specimens of Murraya paniculata may have been collected in Sumatra, not Penang and Singapore, as he travelled to Sumatra in 181930. Moreover, when William Hunter visited Penang (Prince of Wales Island) in 1802, he (Hunter 1909) reported that there was only one Murraya exotica tree on the island, a young tree, yet to ripen seeds,

30 (http://www.nationaalherbarium.nl/fmcollectors/j/JackW.htm) 224 in the garden of Lieut. Col. Polhill who, according to Hodson (1910), was the commanding officer of the settlement. Hunter (1909) also noted that the tree, with fragrant flowers, was common in the neighbourhood of Queda (Kedah) on the adjacent region of the Malay Peninsula.

Roxburgh (1832) placed Chalcas paniculata L., Camunium japonense (Lour.) and Marsana buxifolia Sonn. under Murraya exotica, noting that it was a cultivated plant introduced to India from China ‘many years ago’. He regarded this cultivated plant as distinct from a Sumatran species he described as Murraya sumatrana based on plants growing in the botanic garden in Calcutta and sent to India by Dr C Campbell labelled as Chalcas.

Wight & Walter-Arnott (1834) recognised the following taxa as occurring on the eastern peninsular of India and commented on them:  Murraya exotica (L.) [Chalcas japonense Lour., Marsana buxifolia Sonn.], cultivated;  Murraya paniculata (Herb. Sm.!31) [Chalcas paniculata Linn.], not uncommon; and  Murraya sumatrana (Roxb.) [Murraya paniculata Jack.], most probably not a species from peninsular India.

They (Wight & Walter-Arnott (1834)) noted that Murraya paniculata (Herb. Sm.!) [Chalcas paniculata Linn.] and Murraya sumatrana (Roxb.) [Murraya paniculata

Jack.] were closely allied and that the descriptions of ‘Rumphius, Jack, Blume and Roxburgh’ of the plant from the Indonesian islands, agreed with the above characters for Murraya paniculata (Herb. Sm.!) [Chalcas paniculata Linn.].

Voigt (1845) also recognised three taxa occurring in India, and also commented on them:  Murraya exotica (Marsana buxifolia and Chalcas japonense), much cultivated and growing at the foot of the Himalaya;  Murraya paniculata (Chalcas paniculata), growing on tablelands at Mahableshwur in Maharashtra and on the Ghauts (Ghats) and at Goalpara in Assam; and

31 The Smith Herbarium 225

 Murraya sumatrana from China, Cochin-China (southern Việt Nam) and the Moluccas.

Oliver (1861), who regarded Murraya paniculata and Murraya sumatrana as diverse forms of Murraya exotica, noted that Murraya elongata A. De Candolle from Burma appeared to be allied to the Murraya paniculata forms. Hooker (1875) considered Murraya elongata as a very different looking plant from any of the forms of Murraya exotica, including Murraya paniculata. He also noted that Murraya gleniei Thwaites ex Oliv., which occurs in Sri Lanka (Swingle & Reece 1967), closely resembled Murraya exotica. Hooker (1875) regarded Murraya gleniei as variety of Murraya exotica whilst Swingle & Reece (1967) regarded it as a species. Further studies are required to resolve these issues: I was not able to acquire specimens of Murraya gleniei for my research. It will also be important to sample plants from Burma to determine the status of Murraya elongata32.

Heyne (1917) regarded Murraya exotica as distinct from Murraya paniculata Jack, stating, with respect to the latter and in the context of plants in Indonesia, that ‘By preserving this last name, some acknowledge our plant as a separate species’. He also stated that, according to Koorders & Valeton (1894), Murraya exotica var. sumatrana Hook. and Murraya sumatrana Roxb. were synonyms of Murraya paniculata Jack. He regarded Camunium japonense (javanicum) Rumph. as a synonym of Murraya exotica. Heyne’s (1917) text is not clear with respect to Rumphius’ descriptions, but it suggests that he thought that Murraya exotica was only planted for its ornamental value, that it was imported from Japan, and that it also occurred in China. Rumphius’ (1747) name for the plant, Camunium japonense, suggests that he considered it to be species introduced to Java from Japan. It may have also been introduced to Indonesia and Malaysia as an ornamental. Hunter’s (1909) observations support this possibility. Moreover, other records, as noted in Chapter 2, indicate that Murraya exotica was introduced from China to India before 1771 (Edwards & Lindley 1819–1820, Roxburgh 1832) and that the type specimen for the genus Murraya, LINN 539.1 in Herbarium Linnaeus (Murraea exotica: http://www.linnean-online.org/5864/) (Fig. 2.3), collected by Johann Gerhard König, probably in Tharangambadi in Tamil Nadu in southern India, is a specimen of Murraya exotica, not Murraya paniculata.

32 Talapatra et al. (1973) based their studies on the phytochemistry of Murraya elongata on shrubs growing in gardens in West Bengal. 226

These records and the results of my studies suggest that Murraya paniculata as circumscribed by Swingle & Reece (1967) comprises four putative species and two hybrids, with the validity of names of new taxa potentially requiring amendment following a revision of the genus:  Murraya exotica L.  It appears to have originated in southern China.  Synonyms include Camunium japonense (javanicum) Rumph. and Marsana buxifolia Sonn.  It was introduced to the Malay Peninsula and the Indonesian Archipelago before 1700, to India before 1760, to England in 1771 and subsequently elsewhere.  It is the widely cultivated ornamental known as orange jasmine.  Murraya paniculata (L.) Jack  Most probably described from specimens collected in Sumatra.  Synonyms appear to include Camunium vulgare Rumph. and Murraya sumatrana Roxb.  It may only occur in the Indonesian Archipelago.  It is not the widely cultivated ornamental known as orange jasmine.  Murraya asiatica n. sp. ineditus  Representing a previously undescribed species in mainland Asia, and represented in this study by accessions from China, Việt Nam and Pakistan;  Murraya elongata A. de Candolle from Myanmar (Burma) may be representative of these forms but extant descriptions (Hooker 1875) and imperfect specimens deposited in the Wallichian (Oliver 1861) (Royal Botanic Gardens, Kew) may render valid comparisions impractical.  Further studies are required to determine taxonomic implications of variation in the species.  Phytochemical studies reported by Li et al. (1988).were based on Murraya asiatica accessions from Guangdong, not accessions of Murraya paniculata  Murraya ovatifoliolata (Engl.) Domin. ineditus:  An Australasian species comprising:  Murraya ovatifoliolata var. ovatifoliolata comb. nov (small and large leaflet forms), formerly Murraya paniculata var. ovatifoliolata Engl. (Swingle & Reece 1967); and 227

 Murraya ovatifoliolata var. zollingeri comb. nov., formerly Murraya paniculata var. zollingeri Tan. (Swingle & Reece 1967).  Murraya × omphalocarpa (Hay.) Tan. ineditus from near Taiwan, with Murraya exotica the male parent and Murraya ovatifoliolata the female parent.  Murraya × cycloopensis nom. nov. ineditus from New Guinea, with Murraya exotica the female parent and Murraya ovatifoliolata the male parent..

My accessions from Asia and Australasia grouped into biogeographical regions that conformed to historical records for the taxa that suggest natural allopatric distributions with limited overlap. Although wider collection of material is required, including Murraya gleniei (Fig 7.1) from Sri Lanka to confirm distribution boundaries, my results and historical records related to the origin of Murraya exotica (Edwards & Lindley 1819-1820, Roxburgh 1832) and the soil types on which Murraya exotica, Murraya asiatica33 (Kong et al. 1986, Huang 1997), Murraya paniculata (Jones 1995), and Murraya ovatifoliolata occur (Brophy et al. 1994) suggest that the natural distribution of:  Murraya exotica is on low altitude maritime and riparian sites on acid red soils in a region incorporating southern China (including Hainan) and northern Việt Nam;  Murraya asiatica is a mainland Asia species associated with limestone hills from Pakistan through India, Myanmar, Thailand, peninsular Malaysia, Lao, Việt Nam to southern China, with some overlap with Murraya exotica in southern China and northern Việt Nam;  Murraya paniculata comprises the western half of the Indonesian archipelago, possibly including Borneo, usually in association with rocky soils or limestone; and  Murraya ovatifoliolata is Australasian, with the small and large leaflet forms of Murraya ovatifoliolata var. ovatifoliolata north and northeastern Australia, and Murraya ovatifoliolata var. zollingeri in the Nusa Tenggara islands of the Indonesian archipelago.

33 cited as Murraya paniculata by Kong et al. (1986) and Huang (1997). 228

Figure 7.1. Specimen of Murraya gleniei, sensu Swingle & Reece (1967): courtesy of Dr Brett Jestrow, Montgomery Botanical Center, Miami, Florida.

My analysis of leaf and leaflet characters suggests that the following characters could be used to develop the following key to separate the four species, and varieties and forms of Murraya ovatifoliolata: length of the basal leaflet, the ratio of the length to the width of basal leaflet, angle of the base of the basal and terminal leaflets and the ratio of the acuminate part of the terminal leaflet. This needs to be confirmed from a larger number of accessions from throughout the natural ranges of these taxa. Additionally, the

229 validity of the key will need to be evaluated by others, and additional specimens, particularly, Murraya asiatica from mainland Asia (from the Western Ghats of India to southeast China) and Murraya × cycloopensis from the eastern Indonesian archipelago and New Guinea.

1 Leaves with almost all (> 90%) leaflets acuminate. 2 1* Leaves with leaflets with variable apex shape, obtuse to 3 rounded to acute to acuminate. 2 base of basal leaflet obtuse to rounded, angle > 90°; distance Murraya paniculata between petiolule‐rachis junction of first and second leaflets averaging  6 mm. base of basal leaflet obtuse to rounded, angle < 90°; distance Murraya asiatica between petiolule‐rachis junction of first and second leaflets averaging  10 mm. 3 Base of basal leaflet asymmetric cuneate. Murraya exotica 3* Base of basal leaflet obtuse to rounded. 4 4 Petiole <15 mm long. 5 4* Petiole >15mm long. 6 5 Ratio of leaf length to number of leaflets 25 (15–39), basal Murraya ovatifoliolata var. leaflets ovate to elliptic 31 (14–47) mm long * 22 (9–40) mm ovatifoliolata ‘large leaflet’ wide, terminal leaflets ovate to elliptic 59 (39–85) mm * 36 (24–60) mm. 5* Ratio of leaf length to number of leaflets 15 (10–26), leaflets Murraya ovatifoliolata var. ovate to elliptic 18 (8–40) mm long * 14 (6–28) mm wide, ovatifoliolata ‘small leaflet’ terminal leaflets ovate to elliptic 37 (22–56) * 25 (14–45) mm. 6 Leaflets sparsely pubescent on rachis, petiolules, upper Murraya × omphalocarpa surface of midvein, lower surface of midvein with tufted hairs. 6* Leaflets pubescent. 7 7 Basal leaflet ovate to elliptic, 24 (13–39) mm long * 15 (10– Murraya ovatifoliolata var. 25) mm wide. zollingeri 7* Basal leaflet ovate, 35 (28–42) mm long * 23 (18–27) mm Murraya × cycloopensis wide.

230

References

Aiton WT. 1811. Hortus Kewensis: A Catalogue of Plants Cultivated in the Royal Botanic Garden at Kew. Volume III. London: Longman, Hurst, Rees, Orme and Brown. Agapow P-M, Bininda-Emonds ORP, Crandall KA, Gittleman JL, Mace GM, Marshall JC, Purvis A. 2004. The Quarterly Review of Biology 79: 161–179. Alfaro ME, Zoller S, Lutzoni F. 2003. Bayes or bootstrap? A simulation study comparing the performance of Bayesian Markov chain Monte Carlo sampling and bootstrapping in assessing phylogenetic confidence. Molecular Biology and Evolution 20: 255–266. Andrade M, Mayo S, Kirkup D, van den Berg C. 2008. Comparative morphology of populations of Monstera Adans. (Araceae) from natural forest fragments in Northeast Brazil using elliptic Fourier analysis of leaf outlines. Kew Bulletin 63: 193–211. Arnold ML, Hodges SA. 1995. Are natural hybrids fit or unfit relative to their parents? Tree 10: 67–71. Arnold ML. 1997. Natural Hybridization and Evolution. New York: Oxford University Press. Aubert B, Garnier M, Guilliaumin D, Herbagyandono B, Setiobudi L, Nurhadi F. 1985. Greening, a serious threat for the citrus productions of the Indonesian archipelago. Future prospects of integrated control. Fruits 40: 549–563. Aubert B. 1988. Le greening une maladie infectieuse des agrumes, d’origine bactérienne, transmise par des Homopteres psyllidés. Stratégie de lutte devéloppée à l’île de la Réunion. Circonstances épidémiologiques en Afrique/Asie et modalités d’intervention. INFA/CIRAD – B.P. 180 – 97455 Saint Pierre Cedex. 185 pp. (PhD thesis). Aubert B. 1990. Prospects for citriculture in Southeast Asia by the year 2000. FAO Plant Protection Bulletin 38: 151–173. Bailey FM. 1909. Comprehensive Catalogue of Queensland Plants, Both Indigenous and Naturalised. Brisbane: Government Printer. Baker WJ, Hedderson TA, Dransfield J. 2000. Molecular phylogenetics of subfamily Calamoideae (Palmae) based on nrDNA ITS and cpDNA rps16 Intron sequence data. Molecular Phylogenetics and Evolution 14: 195–217.

231

Barber JC, Finch CC, Francisco-Ortega J, Santos-Guerra A, Jansen RK. 2007. Hybridization in Macaronesian Sideritis (Lamiaceae): evidence from incongruence of multiple independent nuclear and chloroplast sequence datasets. Taxon 56: 74–88. Bayer R, Mabberley DJ, Morton C, Miller CH, Shama IK, Pfeil BE, Rich S, Hitchcock R, Sykes S. 2009. A molecular phylogeny of the orange subfamily (Rutaceae: Aurantioideae) using nine cpDNA sequences. American Journal of Botany 96: 668–685. Beattie GAC, Holford P, Mabberley DJ, Haigh AM, Bayer R, Broadbent P. 2006. Aspects and insights of Australia-Asia collaborative research on huanglongbing. In: Proceedings of an International Workshop for the Prevention of Citrus Greening Disease in Severely Infected Areas, Ishigaki, Japan, 6–7 December 2006. Tokyo: Multilateral Research Network for Food and Agricultural Safety. Japanese Ministry of Agriculture, Forestry and Fisheries, Tokyo, Japan. pp. 47– 64. Beattie GAC, Holford P, Mabberley DJ, Haigh AM, Broadbent P. 2008. Australia and Huanglongbing. In: Ku TY, Pham THH (eds), Proceedings of FFTC-PPRI-NIFTS Joint Workshop on Management of Citrus Greening and Virus Diseases for the Rehabilitation of Citrus Industry in the ASPAC, Plant Protection Research Institute, Hà Nội, Việt Nam, 8–12 September 2008. http://www.agnet.org/library/eb/607/. pp. 75–100. Beattie GAC, Barkley P. 2009. Huanglongbing and its vectors: A pest–specific contingency plan for the citrus and nursery and garden industries (Version 2), February 2009. Horticulture Australia Ltd., Sydney. 272 pp. Benyon LS, Zhou LJ, Duan YP, Powell CA, Gottwald TR. 2008. Improved detection of low-titer, non-lethal, seed transmitted Candidatus Liberibacter asiaticus in citrus, periwinkle and dodder using nested PCR. Proceedings of the International Research Conference on Huanglongbing, Orlando, Florida, 2–5 December 2008. http://www.doacs.state.fl.us/pi/hlb_conference/Proceedings.pdf. Bové JM. 2006. Huanglongbing: A destructive, newly–emerging, century–old disease of citrus. Journal of Plant Pathology 88: 7–37. Brophy JJ, Forster PI, Goldsack RJ. 1994. Diversity in Australian populations of Murraya paniculata (Rutaceae): New evidence from volatile leaf oils. Australian Systematic Botany 7: 409–418.

232

Bull JJ, Huelsenbeck JP, Cunningham CW, Swofford DL, Waddell PJ. 1993. Partitioning and combining data in phylogenetic analysis. Systematic Biology 42: 384–397. Burman NL. 1768. Flora Indica: cui accedit series zoophytorum indecorum, nec non Prodromus Florae Capensis. Leiden: Cornelius Haak; Amsterdam: Johannes Schreuder. But PPH, Kong YC, Ng KH, Chang HT, Li Q, Yu SX, Waterman PG. 1986. A chemotaxonomic study of Murraya (Rutaceae) in China. Acta Phytotaxonomica Sinica 24: 186–192. But PPH, Kong YC, Li Q, Chang HT, Chang KL, Wong KM, Gray AI, Waterman PG. 1988. Chemotaxonomic relationship between Murraya and Merrillia (Rutaceae). Acta Phytotaxonomica Sinica 26: 205–210. Cheema SS, Kapur SP. 1975. Murraya paniculata Linn.―A new host for Diaphorina citri Kuwayama. Current Science 44: 249. Chohan SN, Qamar R, Sadiq I, Azam M, Holford P, Beattie A. 2007. Molecular evidence for the presence of huanglongbing in Pakistan. Australasian Plant Disease Notes 2: 37–38. Clausen CP. 1933. The citrus insects of tropical Asia. Circular 266. Washington: United States Department of Agriculture. 35 pp. Compton JA, Culham A, Jury SL. 1998. Reclassification of Actaea to include Cimicifuga and Souliea (Ranunculaceae): Phylogeny inferred from morphology, nrDNA ITS, and cpDNA trnL-F sequence variation. Taxon 47: 593–634. Conant P, Hirayama C, Kumashiro BR, Heu RA. 2007. Asian citrus psyllid Diaphorina citri Kuwayama (Hemiptera: Psyllidae). New Pest Advisory 06–01. Honolulu: Hawaii Department of Agriculture. 2 pp. http://www.hawaiiag.org/hdoa/npa/npa06–01–ACP.pdf. Cracraft J, Helm-Bychowski K. 1991. Parsimony and phylogenetic inference using DNA Sequences: Some methodological strategies. In: Miyamoto MM, Cracraft J (eds), Phylogenetic Analysis of DNA Sequences. New York: Oxford University Press. Crockett SL, Robson NK. 2011. Taxonomy and chemotaxonomy of the genus Hypericum. Medicinal and Aromatic Plant Science and Biotechnology 5: 1–13. Cunningham CW. 1997. Can three incongruence tests predict when data should be combined? Molecular Biology and Evolution 14: 733–740.

233 da Graça JV. 1991. Citrus greening disease. Annual Review of Phytopathology 29: 109–136. da Graça JV, Korsten L. 2004. Citrus huanglongbing: review, present status and future strategies. In: Naqvi SAMH (ed.), Diseases of Fruits and Vegetables, Volume I. Dordrecht: Kluwer Academic Press. pp. 229–245. Dai K, Ikeshiri T, Matsuura T, Kimura S, Hamagami A, Fujiwara Y, Kobashigawa Y, Miyakuni S. 2005. Investigation of host range of Candidutus Liberobacter asiaticum — Is Murraya paniculata a host plant of Candidutus L. asiaticum? Research Bulletin of the Plant Protection Service, Japan 41: 53–57. Damsteegt V, Postnikova E, Stone A, Kuhlmann M, Wilson C, Sechler A, Schaad N, Brlansky R, Schneider W. 2010. Murraya paniculata and related species as potential hosts and inoculum reservoirs of ‘Candidatus Liberibacter asiaticus’, causal agent of huanglongbing. Plant Disease 94: 528–533. Darderes CSG. 2009. Distribución geográfica de Diaphorina citri Kuwayama. Ministerio de la Producción. Secretaría de Agricultura, Ganadería, Pesca y Alimentación. Buenos Aires, Argentina. de Araújo EF, de Queiroz LP, Machado MA. 2003. What is Citrus? Taxonomic implications from a study of cpDNA evolution in the tribe Citreae (Rutaceae subfamily Aurantioideae). Organisms Diversity & Evolution 3: 55–62. de Queiroz K. 2007. Species concepts and species delimitation. Systematic Biology 56: 879–886. de Loureiro J. 1790. Flora Cochinchinensis. Volume 1. Ulyssipone (Lisbon): Typis, et expensis academicis. Demesure B, Sodzi N, Petit RJ .1995. A set of universal primers for amplification of polymorphic non–coding regions of mitochondrial and chloroplast DNA in plants. Molecular Ecology 4: 129–131. den Bakker HC, Gravendeel B, Kuyper TW. 2004. An ITS phylogeny of Leccinum and an analysis of the evolution of minisatellite-like sequences within ITS1. Mycologia 96: 102–118. Deng X, Zhou G, Li H. 2007a. Nested–PCR detection and sequence confirmation of ‘Candidatus Liberibacter asiaticus’ from Murraya paniculata in Guangdong, China. Plant Disease 91: 1051. Deng X, Zhou G, Li H, Chen J, Civerolo EL. 2007b. Detection of Candidatus Liberibacter asiaticus from wampee (Clausena lansium Skeels) by nested PCR. Plant Health Progress. doi: 10.1094/PHP–2007–0419–01–BR. 234

Ding F, Yi G, Wang G. 2004. Research on the PCR and nested–PCR detection of citrus huanglongbing pathogen. Acta Horticulturae Sinica 31: 803–806. Ding F, Wang G, Yi G, Zhong Y, Zeng J, Zhou B. 2005. Infection of wampee and lemon by the citrus huanglongbing pathogen (Candidatus Liberibacter asiaticus) in China. Journal of Plant Pathology 87: 207–212. Doyle JJ, Doyle LL .1990. Isolation of plant DNA from fresh tissue. Focus 12: 13–15. Duminil J, Di Michele M. 2009. Plant species delimitaion: A comparison of morphological and molecular markers. Plant Biosystems 143: 528–542. Edwards ST, Lindley J. 1819–20. The Botanical Register: Consisting of coloured figures of exotic plants, cultivated in British gardens; with their history and mode of treatment. Vol V. London: James Ridgeway. [Murraya exotica 434]. Faghihi MM, Salehi M, Bagheri A, Izadpanah K. 2009. First report of citrus huanglongbing disease on orange in Iran. Plant Pathology 58: 793. Fan GC, Cai ZJ, Weng QY, Ke C, Liu B, Zhou LJ, Duan YP. 2011. First report of a new host (Pithecellobium lucidum Benth) of the citrus huanglongbing bacterium, ‘Candidatus Liberibacter asiaticus’. Abstract 8.17, Second International Research Conference on Huanglongbing, Orlando, Florida, January 10–14, 2011. http://irchlb.org/files/54ce6e13-3130-4fa9-b.pdf. Farris JS, Källersjö M, Kluge AG, Bult C. 1995. Testing significance of incongruence. Cladistics 10: 315–319. Felsenstein J. 1978. Cases in which parsimony or compatibility methods will be positively misleading. Systematic Zoology 27: 401–410. Felsenstein J. 1981. Evolutionary trees from DNA sequences: A maximum likelihood approach. Journal of Molecular Evolution 17: 368–376. Fletcher TB (ed.). 1917. Report of the Proceedings of the Second Entomological Meeting, Pusa, India, 5–12 February 1917. Calcutta: Superintendent Government Printing. pp. 213, 215–216. Fletcher TB (ed.). 1919. Report of the Proceedings of the Third Entomological Meeting, Pusa, India, 3–15 February 1919. Calcutta: Superintendent Government Printing.1: 276. Ferguson C, Jansen RK. 2002. A chloroplast DNA phylogeny of Eastern Phlox (Polemoniaceae): Implications of congruence and incongruence with the ITS phylogeny. American Journal of Botany 89: 1324–1335. Forman LL. 1989. The illustrations to William Hunter's 'Plants of Prince of Wales Island' Kew Bulletin 44: 151–161. 235

Forster PI, Smith MW. 2010. Citrus wakonai P.I.Forst & M.W.Sm. (Rutaceae), a new species from Goodenough Island, Papua New Guinea. Austrobaileya 8(2): 133– 138. Freeman H. 1974. Computer processing of line-drawing images. Computing Surveys 6: 57–97 Fritsch PW, Schiller AM, Larson KW. 2009. Taxonomic implications of morphological variation in Cercis canadensis (Fabaceae) from Mexico and adjacent parts of Texas. Systematic Botany 34: 510–520. Galtier N, Daubin V. 2008. Dealing with incongruence in phylogenomic analyses. Philosophical Transactions of the Royal Society B 363: 4023–4029. Garnier M, Bové JM. 1993. Citrus greening disease and greening bacterium. In: Moreno P, da Graça JV, Timmer LW (eds), Proceedings of the Twelfth Conference of the International Organization of Citrus Virologists, New Delhi, India, 23–27 November 1992. International Organization of Citrus Virologists, University of California: Riverside. pp. 212–219. Garnier M, Jaqoueix-Eveillard S, Cronje PR, Le Roux HF, Bové MJ. 2000. Genomic characterization of a liberibacter present in an ornamental rutaceous tree, Calodendrum capense, in the Western Cape province of South Africa. Proposal of ‘Candidatus Liberibacter africanus subsp. capensis’. International Journal of Systematic and Evolutionary Microbiology 50: 2119–2125. Gasparoto M, Bassanezi R, Amorim L, Montesino L, Lourenço S, Wulff N, Teixeira D, Mariano A, Martins E, Leite A, Filho A. 2010. First report of ‘Candidatus Liberibacter americanus’ transmission from Murraya paniculata to sweet orange by Diaphorina citri. Journal of Plant Pathology 92: 546. Ge GB, Zhang YY, Hao DC, Hu Y, Luan HW, Liu XB, He YQ, Wang ZT, Yang L. 2008. Chemotaxonomic study of medicinal Taxus species with fingerprint and multivariate analysis. Planta Medica 74: 773–779. Gielly L, Taberlet P. 1994. The use of chloroplast DNA to resolve plant phylogenies: Noncoding versus rbcL sequences. Molecular Biology and Evolution 11: 769– 777. Gottwald TR. 2010. Current epidemiological understanding of citrus huanglongbing. Annual Review of Phytopathology 48: 119–139. DOI: 10.1146/annurev-phyto- 073009-114418. Graham JH, Irey MS, Dawson WO, Hall D, Duan Y. 2008. Assessment of transmission of Liberibacter asiaticus from seed to seedlings of ‘Pineapple’ sweet orange and 236

‘Carrizo’ citrange. Proceedings of the International Research Conference on Huanglongbing. Orlando, Florida, 2–5 December 2008. http://www.doacs.state.fl.us/pi/hlb_conference/Proceedings.pdf. Graham SW, Olmstead RG, Barrett SCH. 2002. Rooting phylogenetic trees with distant outgroups: A case study from the commelinoid monocots. Molecular Biology and Evolution 19: 1769–1781. Groppo M, Pirani JR, Salatino MLF, Blanco SR, Kallunki JA. 2008. Phylogeny of Rutaceae based on two noncoding regions from cpDNA. American Journal of Botany 95: 985–1005. Guo WW, Deng XX. 1998. Somatic hybrid plantlets regeneration between Citrus and its wild relative, Murraya paniculata via protoplast electrofusion. Plant Cell Reports 18: 297–300. Halbert SE, Manjunath KL. 2004. Asian citrus psyllids (Sternorrhyncha: Psyllidae) and greening disease of citrus: A literature review and assessment of risk in Florida. Florida Entomologist 87: 330–353. Halbert SE, Núñez CA. 2004. Distribution of the Asian citrus psyllid, Diaphorina citri Kuwayama (Rhynchota: Psyllidae) in the Caribbean Basin. Florida Entomologist 87: 401–402. Hall T. 2001. Bioedit Version 5.0.6. North Carolina State University, Department of Microbiology. Hamilton MB, Braverman JM, Soria-Hernanz DF. 2003. Patterns and relative rates of nucleotide and insertion/deletion evolution at six chloroplast intergenic regions in New World species of the Lecythidaceae. Molecular Biology and Evolution 20: 1710–1721. Hartung JS, Halbert S, Shatters R. 2008. Can Ca. Liberibacter asiaticus be transmitted through citrus seed? Proceedings of the International Research Conference on Huanglongbing. Orlando, Florida, 2–5 December 2008. http://www.doacs.state.fl.us/pi/hlb_conference/Proceedings.pdf. Hausdorf B. 2011. Progress towards a general species concept. Evolution 65: 923–931. He FM (Hoffmann WE), Zhou (Djou) YW. 1935. Notes on citrus pests. Lingnan Agricultural Journal 2 (1): 165–218. (in Chinese with English summary by WE Hoffmann). Hendy MD, Penny D. 1989. A framework for the quantitative study of evolutionary trees. Systematic Zoology 38: 297–309.

237

Hennig W. 1966. Phylogenetic systematics. Urbana, Illinois: University of Illinois Press. Hewson HJ. 1988. Plant Indumentum: A Handbook of Terminology. Australia Flora and Fauna Series. Number 9. Canberra: Australian Government Publishing Service. Heyne K. 1917. De nuttige planten van Nederlandsch-Indië. Part 3. Batavia: Gedrukt bij Ruygrok & Co. Hockings D. 1998. Mock orange Murraya paniculata var. ovatifoliolata 'Min–A–Min'. Plant Varieties Journal 11: 27. Hocquellet A, Toorawa P, Bové JM, Garnier M. 1999. Detection and identification of the two Candidatus Liberobacter species associated with citrus huanglongbing by PCR amplification of ribosomal protein genes of the β operon. Molecular and Cellular Probes 13: 373–379. Hodkinson ID. 1986. The psyllids (Homoptera: Psylloidea) of the Oriental zoogeographical region: an annotated check-list. Journal of Natural History. 20: 299–357. Hodson VCP. 1910. Historical Records of the Governor-General’s Body Guard. London: W Thacker & Co. Hoffmann WE. 1936. Diaphorina citri Kuw. (Homoptera: Chermidae), a citrus pest in Kwangtung. Lingnan Science Journal 15: 127–132. Hollis D. 1984. Afrotropical jumping plant lice of the family Triozidae (Homoptera: Psylloidea). Bulletin of the British Museum Natural History (Entomology) 49: 1– 102. Hollis D. 1985. Parapsylla, a Gondwanan element in the psyllid fauna of southern Africa (Homoptera). Zoological Journal of the Linnean Society 83: 325–342. Hollis D. 1987. A new citrus-feeding psyllid from the Comoro islands, with a review of the Diaphorina amoena species group (Homoptera). Systematic Entomology 12: 47–61. Hooker JD. 1875. The Flora of British India. Volume 1. London: L Reeve. pp. 502–503. Hopkins M. 1942. Cercis in North America. Rhodora 44: 192–211. Huang CC. 1959. Preliminary study on Chinese Rutaceae. Acta Phytotaxonomica Sinica 8: 69–124. Huang CC. 1997. Flora Reipublicae Popularis Sinicae. Tomus 43 (2). Huelsenbeck JP. 1995. Performance of phylogenetic methods in simulation. Sytematic Biology 44: 17–48. 238

Hung TH, Hung SC, Chen CN, Hsu MH, Su HJ. 2004. Detection by PCR of Candidatus Liberibacter asiaticus, the bacterium causing citrus huanglongbing in vector psyllids: application to the study of vector–pathogen relationships. Plant Pathology 53: 96–102. Hung TH, Wu ML, Su HJ. 2000. Identification of alternative hosts of the fastidious bacterium causing citrus greening disease. Phytopathology 148: 321–326. Hung TH, Wu ML, Su HJ. 2001. Identification of the Chinese box orange (Severinia buxifolia) as an alternative host of the bacterium causing citrus huanglongbing. European Journal of Plant Pathology 107: 183–189. Husain MA, Nath D. 1927. The citrus psylla (Diaphorina citri, Kuw.) [Psyllidae: Homoptera]. Memoirs of the Department of Agriculture India, Entomology Series 10 (2): 5–27. 1 plate. Hunter W. 1909. Plants of Prince of Wales Island. Journal of the Straits Branch of the Royal Asiatic Society 53: 49-127. Ingram AL, Doyle JJ. 2003. The origin and evolution of Eragrostis tef (Poaceae) and related Polyploids: Evidence from nuclear waxy and plastid rps16. American Journal of Botany 90: 116–122. Isely D. 1975. Leguminosae of the United States: II. Subfamily Caesalpinioideae. Memoirs of the New York Botanical Garden 25: 1–228. Iwata H, Nesumi H, Ninomiya S, Takano Y, Ukai Y. 2002. Diallel analysis of leaf shape variations of citrus varieties based on elliptic Fourier descriptors. Breeding Science 52: 89–94. Iwata H, Niikura S, Matsuura S, Takano Y, Ukai Y. 1998. Evaluation of variation of root shape of Japanese radish (Raphanus sativus L.) based on image analysis using elliptic Fourier descriptors. Euphytica 102: 143–149. Iwata H, Ukai Y. 2002. SHAPE: A computer program package for quantitative evaluation of biological shapes based on elliptic Fourier descriptors. Journal of Heredity 93: 384–385. Jack W. 1820. Descriptions of Malayan Plants. In: Malayan Miscellanies. Volume 1. Bencoolen: Mission Press. pp. 31–33. Jagoueix S, Bové JM, Garnier M. 1994a. Candidatus Liberibacter. In: Garrity GM (ed.), Bergey's Manual of Systematic Bacteriology, volume 2. The Proteobacteria. New York: Springer-Verlag. pp. 400–402.

239

Jagoueix S, Bové J, Garnier M. 1994b. The phloem–limited bacterium of greening disease of citrus is a member of the alpha subdivision of the Proteobacteria. International Journal of Systematic Bacteriology 44: 379–386. Jagoueix S, Bové JM, Garnier M. 1996. PCR detection of the two 'Candidatus' Liberobacter species associated with greening disease of citrus. Molecular and Cellular Probes 10: 43–50. Jena SN, Kumar S, Nair NK. 2009. Molecular phylogeny in Indian Citrus L. (Rutaceae) inferred through PCR–RFLP and trnL–trnF sequence data of chloroplast DNA. Scientia Horticulturae 119: 403–416. Jensen RJ, Ciofani KM, Miramontes LC. 2002. Lines, outlines, and landmarks: morphometric analyses of leaves of Acer rubrum, Acer saccharinum (Aceraceae) and their hybrid. Taxon 51: 475–492. Johnson LA, Soltis DE. 1994. matK DNA sequences and phylogenetic reconstruction in Saxifragaceae s. str. Systematic Botany 19: 143–156. Joly S, Starr JR, H LW, Bruneau A. 2006. Polyploid and hybrid evolution in roses east of the Rocky Mountains. American Journal of Botany 93: 412–425. Jones DT. 1995. Rutaceae. In: Soepadmo E, Wong K (eds.), Tree Flora of Sabah and Sarawak, Volume 1. Forest Research Institute Malaysia, Sabah Forestry Department and Sarawak Forestry Department. pp. 351–419. Koizumi M, Prommintara M, Ohtsu Y. 1996. Wood apple, Limonia acidissima: A new host for the huanglongbing (greening) vector, Diaphorina citri. In: da Graça JV, Moreno P, Yokomi RK (eds), Proceedings of the Thirteenth Conference of the International Organization of Citrus Virologists, Fuzhou, Fujian, China, 16–23 November 1995. International Organization of Citrus Virologists, University of California: Riverside. pp. 271–275. Kong YC, Ng GKH, Wat CKH, But PPH. 1986. Pharmacognostic differentation between Murraya paniculata (L.) Jack and Murraya koenigii (L.) Spreng. International Journal of Crude Drug Research 24: 167–170. Kong YC, But PPH, Nguyen KH, Cheng KF, Chang KL, Wong KM, Gray AI, Waterman PG. 1988a. The biochemical systematics of Merrillia; in relationship to Murraya, the Clauseneae and the Aurantioideae. Biochemical Systematics and Ecology 16: 47–50. Kong YC, But PPH, Ng KH, Li Q, Cheng KF, Waterman PG. 1988b. Micromelum: a key genus in the chemosystematics of the Clauseneae. Biochemical Systematics and Ecology 16: 485–489. 240

Koorders SH, Th. Valeton. 1894. Bijdragen tot de kennis der Boomsoorten op Java. Volumes 4. Batavia and Gravengage: G. Kolff & Co. Korsten F, Jagoueix S, Bové JM, Garnier M. 1996. Huanglongbing (greening) detection in South Africa. In: da Graça JV, Moreno P, Yokomi RK (eds.), Proceedings of the Thirteenth Conference of the International Organization of Citrus Virologists, Fuzhou, Fujian, China, 16–23 November 1995. International Organization of Citrus Virologists, University of California: Riverside. pp. 395–398. Kuhl FP, Giardina CR. 1982. Elliptic Fourier analysis of a closed contour. Computer Graphics and Image Processing 18:259–278. Kuwayama S. 1931. A revision of the Psyllidae of Taiwan. Insecta Matsumarana 5: 117–133. Kuwayama Shigeru. 1908. Die psylliden Japans. I. Transactions of the Sopporo Natural History Society 2: 149-189. (D. citri: p. 160-161, Plate III, Fig. 16). Lal MM. 1920a. Report of the Assistant Professor of Entomology (for year ended 30 June 1917). Review of Applied Entomology 20: 109. Lal MM. 1920b. Report of the Assistant Professor of Entomology (for year ended 30 June 1918). Review of Applied Entomology 20: 109. Lamarck JB. 1797. Encyclopédie Méthodique Botanique (Dictionnaire Encyclopeique Botanique). Volume 4. Paris: H. Agasse. Lanave C, Preparata G, Saccone C, Serio G. 1984. A new method for calculating evolutionary substitution rates. Journal of Molecular Evolution 20: 86–93. Lavin M, Mathews S, Hughes C. 1991. Chloroplast AND variation in Gliricidia sepium (Leguminosae): Intraspecific phylogeny and tokogeny. American Journal of Botany 78: 1576–1585. Lemey P, Salemi M, Vandamme A-M (eds.) .2009. The Phylogenetic Handbook 2nd Edition. Cambridge, UK: Cambridge University Press Lepš J, Šmilauer P. 2003. Multivariate Analysis of Ecological Data using CANOCO. Cambridge, UK: Cambridge University Press. Levin DA. 1979. The nature of plant species. Science 204: 381–384. Ley AC, Hardy OJ. 2010. Species delimitation in the Central African herbs Haumania (Marantaceae) using georeferenced nuclear and chloroplastic DNA sequences. Molecular Phylogenetics and Evolution 57: 859–867. Li Q, Zhu LF, But PPH, Kong YC, Chang HT, Waterman PG. 1988. Monoterpene and sesquiterpene rich oils from the leaves of Murraya species: Chemotaxonomic significance. Biochemical Systematics and Ecology 16: 491–494. 241

Li T, Ke C. 2002. Detection of the bearing rate of Liberobacter asiaticum in the citrus psylla and its host plant Murraya paniculata by nested PCR. Acta Phytophylacica Sinica 29: 31–35. Li W, Hartung JS, Levy L. 2006. Quantitative real–time PCR for detection and identification of Candidatus Liberibacter species associated with citrus huanglongbing. Journal of Microbiological Methods 66: 104–115. Liefting LW, Sutherland PW, Ward LI, Paice KL, Weir BS, Clover GRG. 2009a. A new ‘Candidatus Liberibacter’ species associated with diseases of solanaceous crops. Plant Disease 93: 208–214. Liefting LW, Weir BS, Pennycook SR, Clover GRG. 2009b. ‘Candidatus Liberibacter solanacearum’, associated with plants in the family Solanaceae. International Journal of Systematic and Evolutionary Microbiology 59: 2274–2276. Lim WH, Shamsudin OM, Ko WW. 1990. Citrus greening disease in Malaysia: status report. In: Aubert B, Tontyaporn S, Buangsuwon D (eds), Proceedings of the Fourth International Asia Pacific Conference on Citrus Rehabilitation, Chiang Mai, Thailand, 4–10 February 1990. Rome: FAO UNDP. pp. 100–105. Lin KH. 1956. Observations on yellow shoot of citrus: aetiological studies of yellow shoot of citrus. Acta Phytopathologica Sinica 2: 1–42. Linnaeus C (Linné C). 1767. Mantissa Plantarum. Generum editionis VI. et Specierum editionis II. Holmiae (Stockholm): Laurentii Salvii. Linnaeus C (Linné C). 1771. Mantissa Plantarum. Altera generum editionis VI. et Specierum editionis II. Regni animalis appendix. Holmiae (Stockholm): Laurentii Salvii. Lopes SA. 2006. Huanglongbing in Brazil. In: Proceedings of an International Workshop for the Prevention of Citrus Greening Disease in Severely Infected Areas, Ishigaki, Japan, 6–7 December 2006. Tokyo: Multilateral Research Network for Food and Agricultural Safety. Japanese Ministry of Agriculture, Forestry and Fisheries, Tokyo, Japan. pp. 11–19. Lopes SA, Frare GF, Camargo NA, Wulff DC, Teixeira DC, Bassanezi RB, Beattie GAC, Ayres AJ. 2010. Liberibacters associated with orange jasmine in Brazil: incidence in urban areas and relatedness to citrus Liberibacters. Plant Pathology

59: 1044–1053. Lopes SA, Martins EC, Frare GF. 2006a. Detecção de Candidatus Liberibacter americanus em Murraya paniculata (Detection of Ca. L. americanus in M. paniculata). Fitopatologia Brasileira 31: 303. 242

Lopes SA, Frare GF, Martins EC. 2006b. Hosts of Liberibacter in Brazil. S11. Proceedings of the Huanglongbing–Greening International Workshop, Ribeirão Preto, Sao Paulo, Brazil, 16-20 July 2006. p. 25 Lopes SA, Martins EC, Frare GF. 2005. Detecção de Candidatus Liberibacter americanus em Murraya paniculata. Summa Phytopathologica 31: 48-49. Mabberley DJ. 1997. A classification for edible Citrus (Rutaceae). Telopea 7.2: 167– 172. Mabberley DJ. 1998. Australian Citreae with notes on other Aurantioideae (Rutaceae). Telopea 7: 333–344. Mabberley DJ. 2004. Citrus (Rutaceae): A review of recent advances in etymology, systematics and medical applications. Blumea 49: 481–498. Mabberley DJ. 2010. The species of Citrus (Rutaceae) with pinnate leaves. Blumea 55: 73–74. Maki M. 1915. Namiki oyobi Kanshôyô–Shokubutsu no Jûyô Gaichu ni kwansura Chôsa (Investigations on the principal insect pests of avenue and ornamental plants). Ringyô Shienjô Tokubetsu Hôkoku (Special Report of the Forest Experiment Station, Government of Formosa), 1: 112 + 29 pp., 18 pls (reference to Diaphorina citri and/or Murraya paniculata on pp. 36–38, Pl. VIII). Malik MN. 1973. A new concept in Citrus classification. Pakistan Journal of Scientific Research 25: 268–271. Mallet J. 2005. Hybridization as an invasion of the genome. Trends in Ecology and Evolution 20: 229–237. Manjunath KL, Halbert SE, Ramadugu C, Webb S, Lee RF. 2008. Detection of ‘Candidatus Liberibacter asiaticus’ in Diaphorina citri and its importance in the management of citrus huanglongbing in Florida. Bacteriology 98: 387–396. Manos PS. 1997. Systematic of Nothofagus (Nothofagaceae) based on rDNA spacer sequence (ITS): Taxonomic congruence with morphology and plastid sequences. American Journal of Botany 84: 1137–1155. Martínez Y, Llauger R, Batista L, Luis M, Iglesia A, Collazo C, Peña I, Casín JC, Cueto J, Tablada LM. 2008. First report of Candidatus ‘Liberibacter asiaticus’ associated with huanglongbing in Cuba. British Society for Plant Pathology, New Disease Reports 18. http://www.bspp.org.uk/ndr/jan2009/2008–50.asp. Matos L, Hilf ME, Camejo J. 2009. First report of ‘Candidatus Liberibacter asiaticus’ associated with citrus huanglongbing in the Dominican Republic. Plant Disease (Disease Notes) 93: p. 668. 243

McClean APD, Oberholzer PCJ. 1965. Citrus psylla, a vector of the greening disease of sweet orange. South African Journal of Agricultural Science 8: 297–298. McLellan T. 1993. The roles of heterochrony and heteroblasty in the diversification of leaf shapes in Begonia dregei (Begoniaceae). American Journal of Botany 80: 796–804. Merrill ED. 1917. An interpretation of Rumphius's herbarium amboinense. Manila, Philippines: Bureau of Printing. Merrill ED. 1935. A commentary on Loureiro's “Flora Cochinchinensis”. Transactions of the American Philosophical Society (New Series) 24: l–445. (page 222). Merrill ED. 1952. William Jack's genera and species of Malaysian plants. Journal of the Arnold Arboretum 33: 199–251. Miyakawa T. 1980. Experimentally-induced symptoms and host range of citrus likubin (greening disease). Annals of the Phytopathological Society of Japan 46: 224– 230. Miyatake BY. 1965. Notes on Psyllidae from the Ryukyu Islands (Hemiptera: Homoptera). Kontyủ 33: 171–189. Mondal KS, Ray B, Ghosal PK, Teleman A, Vuorinen T. 2001. Structural features of a water soluble gum polysaccharide from Murraya paniculata fruits. International Journal of Biological Macromolecules 29: 169–174. Moreno P, da Graça JV, Yokomi RK (eds.). 1996. Preface. In: da Graça JV, Moreno P, Yokomi RK (eds), Proceedings of the Thirteenth Conference of the International Organization of Citrus Virologists, Fuzhou, Fujian, China, 16–23 November 1995. Riverside: International Organization of Citrus Virologists, University of California: Riverside. pp. v–vi. Morton CM, Grant M, Blackmore S. 2003. Phylogenetic relationships of the Aurantioideae inferred from chloroplast DNA sequence data. American Journal of Botany 90(10): 1463–1469. Morton CM. 2009. Phylogenetic relationships of the Aurantioideae (Rutaceae) based on the nuclear ribosomal DNA ITS region and three noncoding chloroplast DNA regions, atpB rbcL spacer, rps16, and trnL-trnF. Organisms, Diversity & Evolution 9: 52–68. Mou FJ. 2009. Systematics of Clauseninae (Rutaceae). PhD thesis. Graduate School of the Chinese Academy of Sciences. Mou FJ, Zhang DX. 2009. Pollen morphology supports the reinstatement of Bergera (Rutaceae). Nordic Journal of Botany 27: 298–304. 244

Munyaneza JE, Fisher TW, Sengoda VG, Garczynski SF, Nissinen A, Lemmetty A. 2010. First report of “Candidatus Liberibacter solanacearum” associated with psyllid-affected carrots in Europe. Plant Disease 94: 639. Murray RGE, Schleifer KH. 1994. Taxonomic notes: a proposal for recording the properties of putative taxa of procaryotes. International Journal of Systematic Bacteriology 44: 174–176. Murray MG, Thompson WF. 1980. Rapid isolation of high molecular weight plant DNA. Nucleic Acids Research 8: 4321–4325. Nylander JAA .2004. MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University Nyman T, Julkunen-Tiitto R. 2005. Chemical variation within and among six northern willow species. Phytochemistry 66: 2836–2843. Oberholzer PCJ, Von Standen DFA, Basson WJ. 1965. Greening disease of sweet orange in South Africa. In: Price WC (ed.), Proceedings of the Third Conference of the International Organization of Citrus Virologists, Campinas, Sao Paulo, Brazil, 16–18 September 1963. Gainesville: University of Florida Press. pp. 213– 219. OEPP/EPPO. 2005a. EPPO Standards PM 7/57. Diagnostic protocol for Diaphorina citri. OEPP/EPPO Bulletin 35: 331–333. OEPP/EPPO. 2005b. EPPO Standards PM 7/57. Diagnostic protocol for Trioza erytreae. OEPP/EPPO Bulletin 35: 357–360. Ohtsu Y, Prommintara M, Okuda S, Goto T, Kano T, Nakashima K, Koizumi M, Imadai J, Kawashima K. 2002. Partial purification of Thai isolate of citrus huanglongbing (greening) bacterium and antiserum production for serological diagnosis. Journal of General Plant Pathology 68: 372–377. Olawore NO, Ogunwande IA, Ekundayo O, Adeleke KA. 2005. Chemical composition of the leaf and fruit essential oils of Murraya paniculata (L.) Jack. (Syn. Murraya exotica Linn.). Flavour and Fragrance Journal 20: 54–56. Oliver D. 1861. The natural order Aurantiaceae, with a synopsis of the India species. Journal of the Linnean Society Botany 2: 1–45. Olsson A, Nybom H, Prentice HC. 2000. Relationships between Nordic dogroses (Rosa L. sect. Caninae, Rosaceae) assessed by RAPDs and elliptic Fourier analysis of leaflet shape. Systematic Botany 25: 511–521. Oxelman B, Lidén M, Berglund D.1997. Chloroplast rps16 intron phylogeny of the tribe Sileneae (Caryophyllaceae). Plant Systematics and Evolution 206: 393–410. 245

Palmer JD. 1987. Chloroplast DNA evolution and biosystematic uses of chloroplast DNA variation. The American Naturalist 130: S6–S29. Palmer JD, Jansen RK, Michaels HJ, Chase MW, Manhart JR. 1988. Chloroplast DNA variation and plant phylogeny. Annals of the Missouri Botanical Garden 75: 1180–1206. Penjor T, Anai T, Nagano Y, Matsumoto R, Yamamoto M. 2010. Phylogenetic relationships of Citrus and its relatives based on rbcL gene sequences. Tree Genetics & Genomes 6: 931–939. Pfeil BE, Brubaker CL, Craven LA, Crisp MD. 2002. Phylogeny of Hibiscus and the tribe Hibisceae (Malvaceae) using chloroplast DNA sequences of ndhF and the rpl16 intron. Systematic Botany 27: 333–350. Pfeil BE, Crisp MD. 2008. The age and biogeography of Citrus and the orange subfamily (Rutaceae: Aurantioideae) in Australia and New Caledonia. American Journal of Botany 95: 1621–1631. Phahladira MNB. 2010. Identification of alternative hosts to citrus of ‘Candidatus Liberibacter africanus’ amongst indigenous Rutaceae of South Africa MSc. Dissertation, University of Pretoria, South Africa. Pietersen G, Arrebola E, Breytenbach JHJ, Korsten L, le Roux HF, la Grange H, Lopes SA, Meyer JB, Pretorius MC, Schwerdtfeger M, van Vuuren SP, Yamamoto P. 2010. A survey for ‘Candidatus Liberibacter’ species in South Africa confirms the presence of only ‘Ca. L. africanus’ in commercial citrus. Plant Disease 94: 244– 249. Planet P, Jagoueix S, Bové JM, Garnier M. 1995. Detection and characterization of the African citrus greening Liberobacter by amplification, cloning, and sequencing of the rp/KAJL–rpoBC Operon. Current Microbiology 30: 137–141. Poe S. 2007. Movement of regulated articles from citrus greening and Asian citrus psyllid quarantine zones: Environmental assessment October 2007. Poe SR, Shea K. 2007. Citrus greening and Asian citrus psyllid; availability of an environmental assessment. United States Department of Agriculture Federal Register 72(212): 62204–62205. Rahman AU, Shabbir M, Sultani SZ, Jabbar A, Choudhary MI. 1997. Cinnamates and coumarins from the leaves of Murraya paniculata. Phytochemistry 44: 683–685. Raina V, Verma S, Dhawan S, Khan M, Ramesh S, Singh S, Yadav A, Srivastava S. 2006. Essential oil composition of Murraya exotica from the plains of northern India. Flavour and Fragrance Journal 21: 140–142. 246

Ranade SA, Rana TS, Srivastava AP, Nair KN. 2006. Molecular differentiation in Murraya Koenig ex L. species in India inferred through ITS, RAPD and DAMD analysis. Current Science 90: 1253–1258. Raven P. 1980. Hybridization and the nature of species in higher plants. Canadian Botanical Association Bulletin 13: 3–10. Ridley HN. 1908. New or rare Malayan plants, series IV. Journal of the Straits Branch of the Royal Asiatic Society 50: 111–114. Rieseberg LH, Burke JM. 2001. The biological reality of species: gene flow, selection, and collective evolution. Taxon 50: 47–67 Rieseberg LH, Whitton J, Linder CR. 1996. Molecular marker incongruence in plant hybrid zones and phylogenetic trees. Acita Botanica Neerlandica 45: 243–262. Rieseberg LH, Wood TE, Baack EJ. 2006. The nature of plant species. Nature 440: 524–527. Rodrigo AG, Kelly–Borges M, Bergquist PR, Bergquist PL. 1993. A randomisation test of the null hypothesis that two cladograms are sample estimates of a parametric phylogenetic tree. New Zealand Journal of Botany 31: 257–268. Rodriguez F, Oliver J, Marin A, Medina J. 1990. The general stochastic model of nucleotide substitution. Journal of Theoretical Biology 142: 485–501. Ronquist F, Huelsenbeck JP, van der Mark P. 2005. MrBayes v.3, Distributed under the GNU General Public License. Roxburgh W. 1832. Flora Indica: Descriptions of Indian Plants. Vol. II. Serampore: W Thacker and Co. pp. 374–375. Plate 48. Rumphius GE (Georg Eberhard Rumph). 1747. Herbarium Amboinense. Volume 5. Amsterdam. [Camunium: Liber VII. cap, XV, pp. 26–29]. Samuel R, Ehrendorfer F, Chase MW, Greger H. 2001. Phylogenetic analyses of Aurantioideae (Rutaceae) based on non–coding plastid DNA sequences and phytochemical features. Plant Biology 3: 77–87. Schols P, D’hondt C, Geuten K, Merckx V, Janssens S, Smets E. 2004. MorphoCode: coding quantitative data for phylogenetic analysis. Phyloinformatics 4: 1–4. Schwarz RE, Knorr LC, Prommintara M. 1973. Greening―cause of a recent decline of citrus in Thailand. Plant Protection Service Technical Bulletin 20. Department of Agriculture, Ministry of Agriculture & Co–operatives, Bangkok, Thailand & UNDP 9/FAO THA 68/526. pp. 1–17. Scora RW. 1975. On the history and origin of citrus. Bulletin of the Torrey Botanical Club 102: 369–375. 247

Sealy JR.1956. The Roxburgh Flora Indica drawings at Kew. Kew Bulletin 11(3): 349– 399. Seelanan T, Schnabel A, Wendel JF. 1997. Congruence and consensus in the cotton tribe (Malvaceae). Systematic Botany 22: 259–290. Shafee SA, Alam SM, Agarwal MM. 1975. Taxonomic survey of encyrtid parasites (Hymenoptera: Encyrtidae) in India. Aligarh Muslim University (Zoological Series) on Indian Insect Types 10 (i–iii): 1–125. Shatters Jr RG. 2008. Detection of Candidatus Liberibacter asiaticus in citrus seedlings germinated from Florida seed. Proceedings of the International Research Conference on Huanglongbing, Orlando, Florida, 2–5 December 2008. http://www.doacs.state.fl.us/pi/hlb_conference/Proceedings.pdf. Shaw J, Lickey EB, Miller J, Siripun KC, Winder CT, Schilling EE, Small R. 2005. The tortoise and the hare II: Relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92: 142–166. Shaw J, Lickey EB, Schilling EE, Small RL. 2007. Comparison of whole chloroplast genome sequences to choose noncoding regions for phylogenetic studies in Angiosperms: The tortoise and the hare III. American Journal of Botany 94: 275– 288. Simmons MP, Ochoterena H. 2000. Gaps as characters in sequence-based phylogenetic analyses. Systematic Biology 49: 369–381. Simmons MP, Zhang LB, Webb CT, Reeves A, Miller JA. 2006. The relative performance of Bayesian and parsimony approaches when sampling characters evolving under homogeneous and heterogeneous sets of parameters. Cladistics 22: 171–185. Small RL, Ryburn JA, Cronn RC, Seelanan T, Wendel JF. 1998. The tortoise and the hare: choosing between noncoding plastome and nuclear Adh sequences for phylogeny reconstruction in a recently diverged plant group. American Journal of Botany 85: 1301–1315. Soltis DE, Soltis DS, Kuzoff RK, Tucker TL. 1992. Geographic structuring of chloroplast DNA genotypes in Tiarella trifoliata (Saxifragaceae). Plant Systematics and Evolution 181: 203–216. Soltis DE, Gitzendanner MA, Strenge DD, Soltis PS. 1997. Chloroplast DNA intraspecific phylogeography of plants from the Pacific Northwest of North America. Plant Systematics and Evolution 206: 353–373.

248

Stone BC, Jones DT. 1988. New and noteworthy Rutaceae–Aurantioideae from Northern Borneo. Studies in Malesian Rutaceae, V. Proceedings of the Academy of Natural Sciences of Philadelphia 140: 267–274. Stone BC. 1985. Rutaceae In: Dassanayake MD, Fosberg FR (eds), A Revised Handbook to the Flora of Ceylon. New Delhi: Amerind Publishing Co. Pvt. Ltd. pp. 406–465. Su HJ, Huang AL. 1990. The nature of likubin organism, life cycle morphology and possible strains In: Aubert B, Tontyaporn S, Buangsuwon D (eds.), Proceedings of the Fourth International Asia Pacific Conference on Citrus Rehabilitation, Chiang Mai, Thailand, 4–10 February 1990. Rome: FAO UNDP. pp. 106–110. Su HJ, Hung TH, Lim WH. 1995. Infection and spreading of citrus greening. In: Abstracts of the International Symposium on Integrated Management Insects– borne Virus Diseases of Tropical Fruit. FFTC/ASPAC, Taipei, Taiwan. p. 29. Subandiyah S, Himawan A, Astuti IP, Poerwanto ME, Horford P, Beattie A. 2007. Genetic resources of Rutaceous plants against the psyllid vector preference and pathogen transmission of huanglongbing. International Training Workshop on the Conservation and Utilisation of Tropical/Subtropical Plant Genetic Resources, Taiwan Agricultural Research Institute, Taichung, Taiwan. pp. 43–50. Subandiyah S, Himawan A, Joko T, Astuti IP, Holford P, Beattie GAC, Krugger R. 2008. Colonisation of Asiatic citrus psyllid and huanglongbing development on Citrus and Citrus Relatives in Indonesia. Proceedings of the International Research Conference on Huanglongbing, Orlando, Florida, 2–5 December 2008. http://www.doacs.state.fl.us/pi/hlb_conference/Proceedings.pdf. Supriyanto A, Whittle AM. 1991. Citrus rehabilitation in Indonesia. In: Brlansky RH, Lee RF, Timmer LW (eds), Proceedings of the Eleventh Conference of the International Organization of Citrus Virologists, Orlando, Florida, 6–10 November 1989. Riverside: International Organization of Citrus Virologists, University of California: Riverside. pp. 409–413. Swingle WT. 1918. Merrillia, a new Rutaceous genus of the tribe Citreae from the Malay Peninsula. The Philippine Journal of Science 13: 335–343. Swingle WT. 1943. The botany of Citrus and its wild relatives of the orange subfamily (Family Rutaceae, Subfamily Aurantioideae). In: Webber HJ, Batchelor LD (eds), The Citrus Industry, Volume 1. Berkeley & Los Angeles: University of California Press. pp. 129–474.

249

Swingle WT, Reece RC. 1967. The botany of Citrus and its wild relatives. In: Reuther W, Webber HJ, Batchelor LD (eds), The Citrus Industry, Volume I, History, World Distribution Botany, and Varieties. Berkeley: Division of Agricultural Sciences, University of California. pp. 190–430. Swofford DL. 2002. PAUP*. Phylogenetic analysis using parsimony and other methods. Version 4. Sinauer Associates, Sunderland, Massachusetts. Taberlet P, Gielly L, Pauton G, Bouvet J. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105– 1109. Taberlet P, Gielly L, Pauton G. Bouvet J. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105– 1109. Talapatra SK, Dutta LN, Talapatra B. 1973. The structure and stereochemistry of murrangtin a new monomeric coumarin from Murraya elongata Aplf. DC. Tetrahedron 29: 2811–2814. Talianová M. 2007. Survey of molecular phylogenetics. Plant Soil and Environment 53: 413–416. Tanaka T. 1929. Chalcas, a Linnean genus which includes many new types of Asiatic plants. Journal of the Society for Tropical Agriculture 1: 23–44. Tanaka T. 1937. Further revision of Rutaceae–Aurantioideae of India and Ceylon (Revisio aurantiacearum VIII). Journal of the Indian Botanical Society 16: 227– 40. Tanaka T. 1977. Fundamental discussion of Citrus classification. Studia Citrologica, Osaka 14: 1–6. Tavaré S. 1986. Some probabilistic and statistical problems in the analysis of DNA sequences. Lectures on Mathematics in the Life Sciences 17: 57–86. Teixeira DC, Ayres J, Danet L, Jagoueix–Eveillard S, Saillard C, Bové JM. 2005a. First report of a huanglongbing-like disease of citrus in São Paulo State, Brazil and association of a new Liberibacter species, ‘Candidatus Liberibacter americanus’, with the disease. Plant Disease 89: 107. Teixeira DC, Danet JL, Eveillard S, Martins EC, Junior WCDJ, Yamamoto PT, Lopes SA, Bassanezi RB, Ayres AJ, Saillard C, Bové JM. 2005b. Citrus huanglongbing in Saõ Paulo State, Brazil: PCR detection of the ‘Candidatus’ Liberibacter species associated with the disease. Molecular and Cellular Probes 19: 173–179.

250

Templeton AR. 1983. Phylogenetic inference from restriction endonuclease cleavage site maps with particular reference to the evolution of humans and the apes. Evolution 37: 221–244. ter Braak,CJF, Šmilauer P. 2002. CANOCO Reference Manual and CanoDraw for Windows User's Guide: Software for Canonical Community Ordination (version 4.5). Ithaca, New York: Microcomputer Power Thiele K. 1993. The holy grail of the perfect character: the cladistic treatment of morphometric data. Cladistics 9: 275–304. Thompson JD, Higgins DG, Gibson TJ. 1994. Clustal W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position- specific gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673–4680. Tirtawidjaja S. 1981. Insect, dodder and seed transmissions of citrus vein phloem degeneration (CVPD). In: Matsumoto K (ed.), Proceedings of the Fourth International Society Citriculture Congress, Tokyo, Japan, 9–12 November 1981. Riverside: International Society of Citriculture 1: 469–471. Tomimura K, Miyata S, Furuya N, Kubota K, Okuda M, Subandiyah S, Hung TH, Su HJ, Iwanami T. 2009. Evaluation of genetic diversity among ‘Candidatus Liberibacter asiaticus’ isolates collected in Southeast Asia. Phytopathology 99: 1062–1069 Truong NT, Gwag JG, Park YJ, Lee SH. 2005. Genetic diversity of soybean pod shape based on elliptic Fourier descriptors. Korean Journal of Crop Science 50: 60–66 Tsai PY, Hwang MT, Wang HC. 1984. Diaphorina citri on Murraya paniculata. Plant Protection Bulletin (Taiwan, Republic of China) 26: 285–287. Valcárcel V, Vargas P. 2010. Quantitative morphology and species delimitation under the general lineage concept: optimization for Hedera (Araliaceae). American Journal of Botany 97: 1555–1573. Väliranta M, Weckström J. 2007. Applying principal components analysis (PCA) for separating wingless birch fruits—a palaeoecological case study from northern Norway. Annales Botanici Fennici 44: 213–218. van den Berg MA, van Vuuren SP, Deacon VE. 1991–1992. Studies on greening disease transmission by the citrus psylla, Trioza erytreae (Hemiptera: Triozidae). Israel Journal of Entomology 25–26: 51–56.

251 van den Berg MA, van Vuuren SP, Deacon VE. 1992. Vergroeningsietke in n inheemse plant in die sitrusfamilie (Presence of greening in an indigenous tree). Institute for Tropical and Subtropical Crops: Agricultural Research Council 241: 1–3. van Vuuren SP. 1996. Huanglongbing the official name for greening disease of citrus. Inligtingsbulletin – Instituut vir Tropiese en Subtropiese Gewasse 287: 5–6. Vandamme A-M. 2009. Basic concepts of molecular evolution. In: Lemey P, Salemi M, Vandamme A-M (eds), The Phylogenetic Handbook. 2nd edition. Cambridge, UK: Cambridge University Press. pp. 3–29. Verma S, Rana TS, Ranade SA. 2009. Genetic variation and clustering in Murraya paniculata complex as revealed by single primer amplification reaction methods. Current Science 96: 1210–1216. Villalobos W, Hollis D, Gody C, Rivera C. 2005. First report of Diaphorina citri (Hemiptera: Psyllidae) in Costa Rica. Insecta Mundi 19: 191–192. Villechanoux S, Garnier M, Renaudin J, Bové JM. 1992. Detection of several strains of the bacterium-like organism of citrus greening disease by DNA probes. Current Microbiology 24: 89–95. Villechanoux S, Garnier, Laigret F, Renaudin J, Bové JM. 1993. The genome of the non–cultured, bacterial-like organism associated with citrus greening disease contains the nusG–rplKAJL–rpoBC gene cluster and the gene for a bacteriophage type DNA polymerase. Current Microbiology 26: 161–166. Viscosi V, Fortini P, Slice D, Loy A, Blasi C. 2009. Geometric morphometric analyses of leaf variation in four oak species of the subgenus Quercus (Fagaceae). Plant Biosystems 143: 575–587. Voigt JO. 1845. Hortus Suburbanus Calcuttensis. Calcutta: Bishop’s College Press. Wallander E, Albert VA. 2000. Phylogeny and classification of Oleaceae based on rps16 and trnL-F sequence data. American Journal of Botany 87: 1827–1841. Wang Z, Yin Y, Hu H, Yuan Q, Peng G, Xia Y. 2006. Development and application of molecular–based diagnosis for ‘Candidatus Liberibacter asiaticus’, the causal pathogen of citrus huanglongbing. Plant Pathology 55: 630–638. Warude D, Chavan P, Joshi K, Patwardhan B. 2003. DNA isolation from fresh and dry plant samples with highly acidic tissue extracts. Plant Molecular Biology Reporter 21: 467. Waterman PG. 1990. Chemosystematics of the Rutaceae: comments on the interpretation of DA SILVA & al. Plant Systematics and Evolution 173: 39–48.

252

Waterman PG. 2007. The current status of chemical systematics. Phytochemistry 68: 2896–2903. Waterston J. 1922. On the chalcidoid parasites of psyllids (Hemiptera: Homoptera). Bulletin of Entomological Research 13: 41–58. Weinert MP, Jacobson SC, Grimshaw JF, Bellis GA, Stephens PM, Gunua TG, Kame MF, David RI. 2004. Detection of huanglongbing (citrus greening disease) in Timor-Leste (East Timor) and in Papua New Guinea. Australasian Plant Pathology 33: 135–136. Weins JJ, Servedio MR. 2000. Species delimitation in systematics: inferring diagnostic differences between species. Proceedings of the Royal Society of London B: Biological Sciences 267: 631–636. Wendel JF, Doyle J. 1998. Phylogenetic incongruence: Window into genome history and molecular evolution. In: Soltis DE, Soltis PS, Doyle JJ (eds), Molecular Systematics of Plants. Boston: Kluwer Academic Publishers. White IM, Hodkinson ID. 1985. Nymphal taxonomy and systematics of the Psylloidea (Homoptera). Bulletin of the British Museum of Natural History (Entomology) 50: 153–301. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ (eds.), PCR Protocols: A Guide to Methods and Applications. New York: Academic Press. pp. 315–322. Wight R, Walker-Arnott GA. 1834. Prodromus florae peninsulae Indiae Orientalis: containing abridged descriptions of the plants found in the peninsula of British India arranged according to the natural system. Volume 1. London: Parbury, Allen & Co. Wink M. 2003. Evolution of secondary metabolites from an ecological and molecular phylogenetic perspective. Phytochemistry 64: 3–19. Yang YP, Huang MD, Beattie GAC, Xia YL, Ouyang GC, Xiong JJ. 2006. Distribution, biology, ecology and control of the psyllid Diaphorina citri Kuwayama, a major pest of citrus: A status report for China. International Journal of Pest Management 52: 343–352. Yang Z. 1993. Maximum–likelihood estimation of phylogeny from DNA sequences when substitution rates differ over sites. Molecular Biology and Evolution 10: 1396–1401.

253

Yoshioka Y, Iwata H, Ohsawa R, Ninomiya S. 2004. Analysis of petal shape variation of Primula sieboldii by elliptic Fourier descriptors and principal component analysis. Annals of Botany 94: 657–664. Zhang DX, Hartley TG, Mabberley DJ. 2008. Rutaceae. In: Wu ZY, Raven PH, Hong DY (eds.), Flora of China, Vol. 11 (Oxalidaceae through Aceraceae). Beijing: Science Press, St. Louis: Missouri Botanical Garden Press. Zhang Q, Chiang TY, George M, Liu JQ, Abbott RJ. 2005. Phylogeography of the Qinghai-Tibetan Plateau endemic Juniperus przewalskii (Cupressaceae) inferred from chloroplast DNA sequence variation. Molecular Ecology 14: 3513–3524. Zhou LJ, Gabriel DW, Duan YP, Halbert SE, Dixon WN. 2007. First report of dodder transmission of huanglongbing from naturally infected Murraya paniculata to Citrus. Plant Disease 91: 227.

254