Cytogenetic and Phylogenetic Review of the Genus Lachenalia

Riana Kleynhans

The genus Lachenalia (family Asparagaceae), endemic to southern Africa, is a horticultural diverse genus, with many species featuring in the red data list of southern Africa. The extensive morphological variation within some species complicates species delimitation and has led to taxonomic confusion. The genus is utilised in a breeding programme where cytogenetic and phylogenetic information is important for the development of breeding strategies. Chromosome numbers of 89 species have been recorded in literature, with 2n = 10 to 56 and n = 5 to 28. B-chromosomes have been described in some species. Basic chromosome numbers include x = 5, 6, 7, 8, 9, (probably 10), 11, (probably 12), 13 and (probably 15). Polyploidy was reported in 19 taxa (23%), and is most common in the x = 7 group. Molecular cytogenetic studies using 5S rDNA, 18S rDNA probes and DAPI staining, as well as molecular systematic studies using trnL-F and ITS12 were used to assess the phylogeny of the genus. All these s...

® Floriculture and Ornamental Biotechnology ©2012 Global Science Books Cytogenetic and Phylogenetic Review of the Genus Lachenalia Riana Kleynhans1,2* • Paula Spies2 • Johan J. Spies2 1 Agricultural Research Council (ARC), Roodeplaat Vegetable and Ornamental Plant Institute (VOPI), Private Bag X293, Pretoria 0001, South Africa 2 Department of Genetics (116), University of the Free State, P.O. Box 339, Bloemfontein 9300, South Africa Corresponding author: * Rkleynhans@arc.agric.za ABSTRACT The genus Lachenalia (family Asparagaceae), endemic to southern Africa, is a horticultural diverse genus, with many species featuring in the red data list of southern Africa. The extensive morphological variation within some species complicates species delimitation and has led to taxonomic confusion. The genus is utilised in a breeding programme where cytogenetic and phylogenetic information is important for the development of breeding strategies. Chromosome numbers of 89 species have been recorded in literature, with 2n = 10 to 56 and n = 5 to 28. B-chromosomes have been described in some species. Basic chromosome numbers include x = 5, 6, 7, 8, 9, (probably 10), 11, (probably 12), 13 and (probably 15). Polyploidy was reported in 19 taxa (23%), and is most common in the x = 7 group. Molecular cytogenetic studies using 5S rDNA, 18S rDNA probes and DAPI staining, as well as molecular systematic studies using trnL-F and ITS12 were used to assess the phylogeny of the genus. All these studies indicated that species with the same basic chromosome number are closely related. The one deviation is that it appears as if there are two separate groups within the x = 7 group. The cytogenetic and molecular studies are further supported by breeding studies, where improved results are generally obtained from crosses within a phylogenetic group or between closely related groups. This review of the literature reveals how different studies obtain similar results regarding the phylogenetic relationships within the genus and how these results can be utilized to improve breeding strategies. It also accentuates that further multidisciplinary studies are needed to solve the evolutionary history of the complex genus Lachenalia. _____________________________________________________________________________________________________________ Keywords: chromosome numbers, cladograms, cross-ability, phylogeny, polyploidy Abbreviations: APG, Angiosperm Phylogeny Group; atpB, ATPase beta chain; DAPI, 4',6-diamidino-2-phenylindole; FISH, Fluorescent in situ hybridization; ITS1-2, Internal transcribed spacer 1 and 2; MEGA, Molecular Evolutionary Genetics Analysis; n, gametic chromosome number; RAPD, Random amplified polymorphic DNA; rbcL ribulose bisophosphate carboxylase (large); SANBI, South African National Biodiversity Institute; trnL, leucyl-transfer RNA intron; trnF, phenylalanine-transfer RNA; VOPI, Vegetable and Ornamental Plant Institute; x, basic chromosome number; 2n, somatic chromosome number; 5S rDNA and 18S rDNA, 5S and 18S ribosomal DNA CONTENTS INTRODUCTION........................................................................................................................................................................................ 98 CYTOGENETIC STUDIES......................................................................................................................................................................... 99 Chromosome counts ................................................................................................................................................................................ 99 Chromosome morphology ..................................................................................................................................................................... 101 Basic chromosome numbers and polyploidy ......................................................................................................................................... 105 Meiotic studies....................................................................................................................................................................................... 106 PHYLOGENETIC STUDIES .................................................................................................................................................................... 106 The phylogenetic position of Lachenalia .............................................................................................................................................. 106 Phylogeny within the genus................................................................................................................................................................... 107 CROSS-ABILITY IN LACHENALIA ........................................................................................................................................................ 107 COMPARISON BETWEEN CROSS-ABILITY, CYTOGENETIC AND MOLECULAR DATA ............................................................ 108 Basic chromosome numbers and cladograms ........................................................................................................................................ 108 Basic chromosome number and cross-ability ........................................................................................................................................ 108 Evolution and relatedness of different basic chromosome numbers ...................................................................................................... 109 Existence of basic chromosome numbers .............................................................................................................................................. 112 Existence of hybrid species ................................................................................................................................................................... 113 CONCLUSION .......................................................................................................................................................................................... 113 ACKNOWLEDGEMENTS ....................................................................................................................................................................... 114 REFERENCES........................................................................................................................................................................................... 114 _____________________________________________________________________________________________________________ INTRODUCTION The genus Lachenalia Jacq. f. ex Murray, previously a member of the family Hyacinthaceae (Manning et al. 2004; Duncan and Edwards 2006, 2007), but since 2009 reclassified under the family Asparagaceae Juss. (APG III group 2009), is endemic to southern Africa. The genus now also Received: 16 December, 2010. Accepted: 20 December, 2012. includes the former genus Polyxena (Manning et al. 2004). Lachenalia is a horticultural diverse genus, with a distribution range extending from the south-western coast of Namibia, southward throughout the Northern, Western and Eastern Cape provinces of South Africa (Duncan 1998). One species extends as far inland as the south western part of the Free State Province (Duncan 1996). Of the 126 species and Invited Review Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books Fig. 1 Morphological variation in Lachenalia in the greenhouse. and appearance, collectors have recognized the horticultural potential of the genus for centuries (Duncan 1988; Du Plessis and Duncan 1989; Kleynhans 2009, 2011; Reinten et al. 2012). The huge phenotypic variation was also the most important reason for the initiation of a breeding programme at the Agricultural Research Council in South Africa. This led to the production of various hybrids and the introduction of new products to the international pot plant market (Fig. 3) (Kleynhans 2006). The variability of the genus in terms of morphology and cytogenetics, however, lead to specific challenges for the breeding of new cultivars. Both incompatibility and other isolation barriers exists (Kleynhans and Hancke 2002). A large number of inter-species crosses are unsuccessful (Kleynhans et al. 2009) and future breeding progress is dependent on information about the genetic variation in the genus. Results generated from cytogenetic and phylogenetic research has value for the breeding programme (Kleynhans et al. 2009) and can furthermore assist in the classification and delimitation of species (Crosby 1986; Spies et al. 2002). This paper reviews the current information available on cytogenetics and phylogeny for the genus Lachenalia and correlates this information to breeding results on crossability with the aim to draw some conclusions on relationships among the different species within the genus. subspecies described, 10% are endangered, 17% are vulnerable, 2% are considered to be near threatened, 6% are critically rare, 9% are rare and 2% are declining (SANBI 2009). The genus is geophytic, deciduous and is usually winter growing. The centre of diversity is in the Worcester grid (3319) in the Western Cape province of South Africa, with species diversity decreasing toward the eastern and northern parts of its range (Duncan 2005). Although Lachenalia species like L. bulbifera and L. obscura are widely distributed, a substantial number of species (e.g. L. moniliformis, L. mathewsii) have a restricted distribution, contributing to the vulnerability of these species (Duncan 1998). Lachenalia occurs in a wide range of habitats, ranging from arid to high rainfall areas. Lachenalia rubida for example always grows in deep, pure sand often very close to the sea, whilst a species like L. campanulata on the other hand is found in heavy soil at altitudes exceeding 2000 metres (Duncan 1998). Between these two extremes, there is a multitude of other habitats, including humus-rich soil on granite, mineral rich soil, barren stony flats, limestone outcrops and seasonally inundated, heavy clays (Duncan 1998). The morphological diversity within the genus is well known (Fig. 1). Variation occur in several morphological characters, such as plant size, leaf number and posture, flower-size, -colour and -orientation and flowering period (Fig. 2). The extensive morphological variation within some species complicates species delimitation and has led to considerable taxonomic confusion (Duncan 1992). Several attempts have thus been made to establish some subgeneric classification within this complex genus, starting with the work by Baker (1897), who divided the genus into five sub-genera based on morphology. The first cytogenetic work by Moffett (1936), however, already indicated that true relationships cut across the groups of Baker and this has been confirmed by various studies (Crosby 1986; Spies 2004; Hamatani et al. 2009; amongst others). Due to the extensive morphological diversity in colour CYTOGENETIC STUDIES Chromosome counts Lachenalia is unusually variable in chromosome number with the presence of different basic chromosome numbers (Moffett 1936; Crosby 1986; Johnson and Brandham 1997), polyploidy (Kleynhans and Spies 1999) and B-chromosomes (Hancke and Liebenberg 1990; Johnson and Brandham 1997). The first cytogenetic studies on the genus came from Moffett (1936). Chromosome numbers steadily increased over many years with information coming from various authors (Table 1). Currently the chromosome num99 Review of the genus Lachenalia. Kleynhans et al. A B C D E F H G Fig. 2 Morphological variation in different Lachenalia species. (A) L. aloides; (B) L. carnosa; (C) L. splendida; (D) L. bulbifera; (E) L. longibracteata; (F) L. violacea; (G) L. contaminata; (H) L. pustulata. bers of 89 species have been recorded in literature. Somatic chromosome numbers vary from 10 to 56 and gametic numbers from 5 to 28. The cytogenetics is further complicated by varying chromosome number reports for a number of species (Table 1). Deviating chromosome counts can first of all be explained by suspected wrong identification of species. In the species L. orchioides the variation could most probably be ascribed to accessions being wrongly identified. Crosby (1986) reported that he received both L. fistulosa and L. pustulata under the name of L. orchioides. Schlechter also identified an accession of L. pallida as L. orchioides (Barker 1983). Both L. pallida and L. pustulata have chromosome numbers of 2n = 16 which could explain some of the variation reported for L. orchioides. Lachenalia contaminata similarly has both 2n = 14 and 2n = 16 reported in literature (Table 1). Gouws (1965) was the first to report both these numbers. The author, however, described these two numbers in one specific bulb of L. contaminata exhibiting cells with both 2n = 14 and 2n = 16. In this case the 2n = 16 could be B-chromosomes that was not identified. Most other chromosome counts of this species, except two by Spies et al. (2008, 2009), are 2n = 16. In this species the variation is not a case of mistaken identity and further investigation is needed to explain the variation. The small size of the chromosomes (Hancke and Liebenberg 1990; Spies et al. 2000) in the genus can furthermore contribute to miscounts and possible miss-identification of B-chromosomes. The presence of B-chromosomes in Lachenalia was described by Hancke and Liebenberg (1990). According to the authors, B-chromosomes in Lachenalia do not have a specific staining pattern and are similar in size to the smallest chromosome in the normal complement. This behaviour makes them difficult to identify and therefore could explain some erroneous counts, reported in literature. B-chromosomes in Lachenalia do not occur in all cells of a specific individual and also not in all plants of a specific accession (Hancke and Liebenberg 1990). It is thus important to investigate the chromosome number of several individuals from a specific population to have accurate chromosome counts and correctly identify the presence of B-chromosomes. Counting insufficient number of cells can similarly lead to miscounts due to chromosome damage occurring during slide preparation. B-chromosomes have been reported in eight species, namely L. aloides, L. anguinea, L. bulbifera, L. carnosa, L. contaminata, L. obscura, L. reflexa and L. splendida (Crosby 1986; Hancke and Liebenberg 1990; Johnson and Brandham 1997; Kleynhans and Spies 1999; Spies et al. 2009). Hamatani et al. (1998) also reported an expected B100 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books A B D E G C F H I Fig. 3 Different Lachenalia cultivars developed at ARC - Roodeplaat VOPI. (A) ‘Rosabeth’; (B) ‘Aqua Lady’; (C) ‘Cherise’; (D) ‘Namakwa’; (E) L. bulbifera x L. rubida; (F) L. unicolor x L. splendida; (G) ‘Romaud’; (H) ‘Rainbow Bells’; (I) L. bachmannii x L. carnosa. chromosome in a 2n = 23 accession of L. zeyheri. Another example where possible B-chromosomes have not been identified, can be found in L. barkeriana where both 2n = 14 and 2n = 16 was reported (Table 1). The 2n = 16 was, however, only found in one cell (Müller-Doblies et al. 1987) of an otherwise 2n = 14 accession and could most possibly be ascribed to extra chromosomes. Chromosome morphology The chromosome morphology of Lachenalia has been described in various reports (Moffett 1936; De Wet 1957; Hamatani et al. 1998; Hancke and Liebenberg 1998; Hancke et al. 2001; Hamatani et al. 2004, 2007, 2009, 2010). Both Moffett (1936) and Hamatani et al. (1998, 2004, 2007) attempted to group the species of the genus based on 101 chromosome length and basic chromosome number. The groupings by Moffett (1936) and Hamatani et al. (1998) agreed, except for the division of the first group of Moffett into two separate groups by Hamatani et al. (1998). Further studies by Hamatani et al. (2004, 2007) added four groups based on chromosome numbers and varying numbers of larger chromosomes within specific basic chromosome numbers. Idiograms presented by De Wet (1957) do not agree with karyograms by Moffett (1936) or Hamatani et al. (1998, 2004, 2007). Neither does it agree with idiograms presented by Hancke et al. (1998, 2001) and Hamatani et al. (2009). The idiogram for L. aloides presented by Hancke et al. (2001) agrees with Moffet’s division, but differs from the karyograms of Hamatani et al. (1998, 2004, 2007) in having 6 longer chromosomes and not only 2 long chromo- Review of the genus Lachenalia. Kleynhans et al. Table 1 List of Lachenalia species with the somatic- and gametic chromosome numbers reported in literature. Number in brackets (#) indicates number of accessions for which the specific somatic or meiotic number was reported. All numbers were reported in the table under the current accepted botanical name. Aneuploidy and other abnormalities or specific detail around polyploidy are indicated with superscripts. Species Somatic no. Gametic Reference (#) no. (#) L. alba W.F. Barker ex G. D. Duncan 18 (1), 20 Johnson and Brandham 1997 (3), 20/40 (1) L. algoensis Schönland 14 (4) Crosby 1986; Hamatani et al. 2007; Spies et al. 2008, 2009 7 (1) Ornduff and Watters 1978 21 (1) Hancke 1991 L. aloides (L.f.) Engl. 14 (32)+0-1B Moffett 1936; Therman 1956; De Wet 1957; Mogford 1978; Crosby 1986; Hancke and Liebenberg 1990; Hancke 1991; Johnson and Brandham 1997; Hamatani et al. 1998, 2004, 2007; Spies et al. 2008; Hamatani et al. 2009; Spies et al. 2009 7 (6) Hancke and Liebenberg 1998; Moffett 1936 Crosby 1986 15 (1)1 Moffett 1936; Crosby 1986 21 (2)1 28 (7) Crosby 1986; Hancke and Liebenberg 1990; Hamatani et al. 1998; Spies et al. 2009; Hamatani et al. 2010 14 (1) Ornduff and Watters 1978 L. ameliae W.F. Barker 18 (2) Johnson and Brandham 1997 L. anguinea Sweet 30 (1)+2B Johnson and Brandham 1997 L. arbuthnothiae W.F. Barker 14 (6) Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 1998; Spies et al. 2008, 2009 7 (1) Spies et al. 2009 L. attenuata W.F. Barker ex G.D. 14 (1) Spies et al. 2009 Duncan L. bachmannii Baker 16 (5) De Wet 1957; Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 2004 L. barkeriana U. Müller-Doblies et 14 (3) Müller-Doblies et al. 1987 al. 16 (2) Nordenstam 1982; Müller-Doblies et al. 1987 L. bolusii W.F. Barker 18 (1) Spies et al. 2009 L. bowkeri Baker 16 (1) Dold and Philipson 1998 L. bulbifera (Cyrillo) Engl. 14 (1) Crosby 1986 28 (7) Kleynhans and Spies 1999; Spies et al. 2009 14 (1) Ornduff and Watters 1978 Moffett 1936c; Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 1998; 42 (15)+0Kleynhans and Spies 1999; Spies et al. 2008 1B1 49 (1) Kleynhans and Spies 1999 56 (5) Crosby 1986; Johnson and Brandham 1997; Kleynhans and Spies 1999 L. capensis W.F. Barker 16 (1) Hamatani et al. 1998 28 (2) Johnson and Brandham 1997; Spies et al. 2008 L. carnosa Baker 16 (26) Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 1998; Du Preez et al. 2002; Spies et al. 2008; Hamatani et al. 2009; Spies et al. 2009 8 (1)+0-2B Spies et al. 2009 L. cernua G.D. Duncan 28 (1) Spies et al. 2008 L. comptonii W.F. Barker 20 (5) Crosby 1986; Johnson and Brandham 1997; Spies et al. 2009 10 (1) Spies 2004 c26 (1) Crosby 1986 L. concordiana Schltr. Ex W.F. 14 (1) Spies et al. 2008 Barker L. congesta W.F. Barker 26, 28 (1) Johnson and Brandham 1997 L. contaminata Aiton 14 (3) Gouws 1965; Spies et al. 2008, 2009 16 (11)+1B De Wet 1957; Gouws 1965; Crosby 1986; Hancke 1991; Johnson and Brandham 1997; Hamatani et al. 2004 8 (2) Ornduff and Watters 1978 32 (1) Johnson and Brandham 1997 L. convallarioides Baker 30 (1) Johnson and Brandham 1997 L. doleritica G.D. Duncan 18 (2) Spies et al. 2008, 2009 L. duncanii W.F. Barker 18 (1) Spies et al. 2008 L. elegans W.F. Barker 14 (6) Moffett 1936; Johnson and Brandham 1997; Spies et al. 2009 28 (12) Moffett 1936; Crosby 1986; Johnson and Brandham 1997; Spies et al. 2009 14 (9) Ornduff and Watters 1978; Spies et al. 2009 42 (4) Johnson and Brandham 1997; Duncan 2001 21 (2) Spies et al. 2009 De Wet 1957 56 (1) 28 (2) Ornduff and Watters 1978 L. ensifolia (Thunb.) J.C. Manning 24 (3) Johnson and Brandham 1997 and Goldblatt 26 (2) Johnson and Brandham 1997; Hamatani et al. 2007 L. fistulosa Baker 14 (8) Johnson and Brandham 1997; Spies et al. 2002; Hamatani et al. 2004; Spies et al. 2009 7 (2) Ornduff and Watters 1978 28 (1) Spies et al. 2008 L. framesii W.F. Barker 16 (3) Du Preez et al. 2002; Spies et al. 2008 L. giessii W.F. Barker 32 (1) Spies et al. 2008 L. gillettii W.F. Barker 16 (1) Spies et al. 2008 102 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books Table 1 (Cont.) Species L. haarlemensis Fourc. L. hirta (Thunb.) Thunb. Somatic no. (#) 18 (2) Gametic no. (#) 9 (1) 22 (6) 11 (2) L. inconspicua G.D. Duncan L. isopetala Jacq. L. juncifolia Baker 24 (3) 18 (1) 30 (2) 40 (1) 22 (9) 11 (1) L. karooica W.F. Barker ex G.D. Duncan L. klinghardtiana Dinter L. kliprandensis W.F. Barker L. lactosa G.D. Duncan L. latimerae W.F. Barker 16 (1) Reference Johnson and Brandham 1997 Ornduff and Watters 1978 Johnson and Brandham 1997; Van Rooyen et al. 2002; Hamatani et al. 2004; Spies et al. 2009 Ornduff and Watters 1978 De Wet 1957; Hancke 1991; Johnson and Brandham 1997 Spies et al. 2008 Johnson and Brandham 1997 Spies et al. 2008 Johnson and Brandham 1997; Hamatani et al. 2007; Spies et al. 2008, 2009; Hamatami et al. 2010 Ornduff and Watters 1978 Duncan 1996 L. leomontana W.F. Barker L. liliflora Jacq. 14 (2) 16 (1) 14 (1) 14 (1) 18 (2) 14 (1) 16 (7) L. longibracteata Phillips 14 (4) L. longituba (A.M. van der Merwe) J.C. Manning and Goldblatt L. macgregoriorum W.F. Barker L. margaretae W.F. Barker L. marginata W.F. Barker 28 (2) Spies et al. 2008 Johnson and Brandham 1997 Spies et al. 2008 Spies et al. 2008 Hamatani et al. 2007, 2010 Spies et al. 2008 Moffett 1936; De Wet 1957; Hancke 1991; Johnson and Brandham 1997; Hamatani et al. 1998, 2009; Spies et al. 2009 Moffett 1936 Crosby 1986; Hamatani et al. 2007; Spies et al. 2008; Hamatani et al. 2009 Ornduff and Watters 1978 Hamatani et al. 2007, 2010 22 (1) 14 (1) 14 (1) 28 (3) 29 (1) 10 (1) Spies et al. 2008 Spies et al. 2008 Spies et al. 2008 Johnson and Brandham 1997 Johnson and Brandham 1997 Duncan 1996 14 (1) Spies et al. 2008 26 (1) 14 (4) 16 (1) Spies et al. 2008 Johnson and Brandham 1997; Hamatani et al. 1998; Spies et al. 2002, 2008, 2009 Spies et al. 2009 8 (1) 7 (2) L. marginata subsp. neglegta Schltr. Ex G.D. Duncan L. marlothii W.F. Barker ex G.D. Duncan L. martinae W.F. Barker L. mathewsii W.F. Barker L. maximiliani Schltr. Ex W.F. Barker L. mediana Jacq. 14 (1) 18 (2) 26 (2) 9 (2) 13 (1) L. minima W.F. Barker L. moniliformis W.F. Barker L. muirii W.F. Barker L. mutabilis Sweet 18 (1) 22 (1) 14 (3) 10 (6) 5 (2) 12 (6) 6 (2) 14 (20) 7 (5) L. namaquensis Schltr. Ex W.F. Barker 24 (1) 56 (1) 16 (11) 8 (2) L. namibiensis W.F. Barker 22 (2) L. neilii W.F. Barker ex G.D. Duncan 18 (1) L. nervosa Ker Gawll 16 (2) 8 (1) L. obscura Schltr. Ex G.D. Duncan L. orchioides (L.) Aiton 24 (2) 18 (2)+1B, 36 (2) 14 (20) 7 (19) 16 (5) 8 (1) 17 (1)1 Johnson and Brandham 1997 Spies et al. 2009 Crosby 1986; Spies et al. 2008 Spies et al. 2009 Spies et al. 2008 Spies et al. 2008 Johnson and Brandham 1997; Hamatani et al. 2007, 2009 Johnson and Brandham 1997 Ornduff and Watters 1978 Spies et al. 2000, 2009 Spies et al. 2002, 2009 De Wet 1957; Crosby 1986; Hancke and Liebenberg; 1990; Johnson and Brandham 1997; Hamatani et al. 1998; Spies et al. 2000, 2009 Hancke and Liebenberg 1998; Spies et al, 2002, 2009 Spies et al. 2000 De Wet 1957 Crosby 1986; Johnson and Brandham 1997; Du Preez et al. 2002; Hamatani et al. 2007; Spies et al. 2008; Hamatani et al. 2009; Spies et al. 2009 Spies et al. 2009 Spies et al. 2008 Spies et al. 2008 Moffett 1936; Spies et al. 2008 Moffett 1936 Johnson and Brandham 1997; Hamatani et al. 2007 Johnson and Brandham 1997 Spies et al. 2008 Crosby 1986; Hamatani et al. 2007; Spies et al. 2008, 2009 Moffett 1936; Ornduff and Watters 1978; Spies et al. 2009 Moffett 1936; De Wet 1957; Hancke 1991 Moffett 1936 Moffett 1936 103 Review of the genus Lachenalia. Kleynhans et al. Table 1 (Cont.) Species Somatic no. (#) 18 (1) 28 (13) Gametic no. (#) 14 (2) L. orthopetala Jacq. L. pallida Aiton 24 (1) 29 (1) 16 (5) 16 (7) 8 (3) L. patula Jacq. L. paucifolia (W.F. Barker) J.C. Manning and Goldblatt L. peersii Marloth ex W.F. Barker L. physocaulos W.F. Barker L. polyphylla Baker L. purpureo-caerulea Jacq. 16 (1) 26 (3) L. pusilla Jacq. 14 (8) L. pustulata Jacq. 16 (1)1 18 (1) 28 (1) 16 (24) L. reflexa Thunb. 32 (1)1 14 (5)+0-2B L. rosea Andrews 16 (1) 14 (6) L. rubida Jacq. 21 (1) 28 (2) 14 (6) L. splendida Diels. 28 (1) 16 (8)+2B 14 (3) 14 (1) 22 (1) 16 (4) 8 (2) 8 (2) 7 (1) 7 (1) 8 (2) L. stayneri W.F. Barker L. thomasiae W.F. Barker ex G. D. Duncan L. trichophylla Baker 18 (1)1 24 (1) 14 (1) 14 (2) 7 (1) L. undulata Masson ex Bak. L. unicolor Jacq. 20 (1) 16 (45) 8 (4) L. unifolia Jacq. 32 (1) 16 (1) 21 (1) 22 (24) 11 (16) L. valeriae G.D. Duncan L. variegata W.F. Barker L. ventricosa Schltr. ex W.F. Barker L. verticillata W.F. Barker L. violacea Jacq 24 (2) 26 (2) 44 (1) 16 (1) 14 (2) 12 (1) 28 (1) 14 (1) 16 (1) 14 (13) 7 (3) 15 (1) 16 (1) Reference Riley 1962 Moffett 1936; De Wet 1957; Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 2007; Spies et al. 2008; Hamatami et al. 2010 Moffett 1936; Ornduff and Watters 1978 Hancke and Liebenberg 1990 Johnson and Brandham 1997 Crosby 1986; Johnson and Brandham 1997; Spies et al. 2008, 2009 Moffett 1936; Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 1998, 2004; Spies et al. 2008, 2009 Moffett 1936; Ornduff and Watters 1978 Johnson and Brandham 1997 Johnson and Brandham 1997; Hamatani et al. 2007, 2010 Johnson and Brandham 1997; Hamatani et al. 2004; Spies et al. 2009 Spies et al. 2008 Spies et al. 2008 Moffett 1936; Johnson and Brandham 1997; Spies et al. 2009 Moffett 1936; Ornduff and Watters 1978 Crosby 1986; Müller-Doblies et al. 1987; Johnson and Brandham 1997; Hamatani et al. 1998, 2007, 2009 Nordenstam 1982 Spies et al. 2009 Hancke 1991 Moffett 1936; Crosby 1986; Johnson and Brandham 1997; Spies et al. 2000; Hamatani et al. 2004; Spies et al. 2008 Moffett 1936; Ornduff and Watters 1978 Spies et al. 2000 Crosby 1986; Hancke and Liebenberg 1990; Johnson and Brandham 1997; Hamatani et al. 1998; Spies et al. 2009 Hancke and Liebenberg 1998 De Wet 1957 Moffett 1936; Crosby 1986; Hancke 1991; Johnson and Brandham 1997; Hamatani et al. 2007; Spies et al. 2008 Crosby 1986 Spies et al. 2009 Moffett 1936; Crosby 1986; Hamatani et al. 1998, 2009; Spies et al. 2009 Moffett 1936 Crosby 1986 Crosby 1986; Johnson and Brandham 1997; Hamatani et al. 1998; Du Preez et al. 2002; Hamatani et al. 2009; Spies et al. 2009 Spies et al. 2009 Crosby 1986 Johnson and Brandham 1997 Spies et al. 2008 Johnson and Brandham 1997 Ornduff and Watters 1978 Johnson and Brandham 1997 Moffett 1936; De Wet 1957; Gouws 1965; Crosby 1986; Hancke 1991; Johnson and Brandham 1997; Hamatani et al. 1998; Spies et al. 2000; Du Preez et al. 2002; Hamatani et al. 2009 Moffett 1936; Ornduff and Watters 1978 Crosby 1986 Hancke 1991 De Wet 1957 Moffett 1936; De Wet 1957; Crosby 1986; Johnson and Brandham 1997; Van Rooyen et al. 2002; Spies et al. 2009 Moffett 1936; Ornduff and Watters 1978; Spies et al. 2009 De Wet 1957; Hamatani et al. 2004 Moffett 1936; De Wet 1957 Johnson and Brandham 1997 Spies et al. 2008 Spies et al 2008; Hamatani et al. 2009 Hamatani et al. 2004 Spies et al. 2002 Spies et al. 2008 Crosby 1986 Hancke 1991; Johnson and Brandham 1997; Hamatani et al. 1998 Ornduff and Watters 1978; Spies et al. 2009 Johnson and Brandham 1997 Crosby 1986 104 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books Table 1 (Cont.) Species L. viridiflora W.F. Barker Somatic no. (#) 14 (7) Gametic no. (#) 7 (1) L. youngii Baker L. zebrina W.F. Barker L. zeyheri Baker 16 (1) 30 (2) 22 (2) 23 (2)1 Reference Nordenstan 1982; Crosby 1986; Hancke and Liebenberg 1990; Hancke 1991; Johnson and Brandham 1997; Spies et al. 2002; Hamatani et al. 2007, 2009 Hancke and Liebenberg 1998 Spies et al. 2008 Johnson and Brandham 1997; Spies et al 2008 Johnson and Brandham 1997; Spies et al 2002 Hamatani et al. 1998, 2010 had 2n = 22 and L. unifolia as 27 out of 32 reports indicated 2n = 22 as somatic chromosome number); x basic group x = 12 (L. ensifolia as 3 out of 5 reports indicate 2n = 24 but this species can also be a possible x = 13 and L. stayneri because it formed a structural diploid based on x = 12 rather than a tetraploid based on x = 6 (Johnson and Brandham 1997); x three different basic chromosome numbers have been recorded for L. mutabilis. This is the only species in basic group x = 5, as well as basic group x = 6. The majority of reports however comes from basic group x = 7 (24 out of 38). Of the 83 taxa that could be grouped, basic x = 7 (41%) and basic x = 8 (27%) were the most common, followed by basic x = 9 (11%) and x = 11 (10%). Basic x = 10 (4%), x = 12 (2%), x = 13 (2%) and x = 15 (4%) are only present in a small number of taxa (Table 1, Fig. 4). Basic x = 5 (1%) and x = 6 (1%) were only present in L. mutabilis. Johnson and Brandham (1997) stated that x = 5 reported for L. mutabilis were derived from plants with 2n = 14 via Robertsonian fusions. Based on their observations of no constant number of long and short chromosomes in L. mutabilis, Spies et al. (2000) disagreed with Johnson and Brandham’s (1997) conclusion that the x = 5 L. mutabilis studied by them resulted from Robertsonian fusions. Spies et al. (2000) could not find any long chromosomes as a result of Robertsonian fusions linked to specific specimens or a specific basic number supporting the hypothesis of Johnson and Brandham (1997). Spies et al. (2000) thus concluded that the variation in L. mutabilis is more likely to be the result of an aneuploid series. More studies are needed to determine the actual mode of chromosome evolution in the species L. mutabilis. Dysploid series also occurs in other genera such as Prospero: x = 4, 5, 6, 7; Bernardia: x = 8, 9; Hyacinthella: x = 9, 10, 11, 12 and Stellarioides: x = 2, 3, 4, 5, 6, 7, 8 and 9. Like in Lachenalia these aneuploid/dysploid series is difficult to interpret (Pfosser and Speta 1999). Combining the chromosome counts with molecular and morphological data might aid in the interpretation of the chromosomal evolution in the genus. The presence of polyploidy was reported in 19 Lachenalia taxa (23%), excluding L. capensis and L. congesta where basic chromosome numbers could not be determined from published results. Conclusions could thus also not be drawn on polyploidy in these species (Table 1). Polyploidy are most common in the basic x = 7 group, with 12 of the 34 species (35%) containing polyploid specimens and a few species exhibiting a range of ploidy levels from triploid to octoploid (Fig. 4; Table 1). Polyploidy were also reported in basic group x = 6, 8, 9, 10 and 11, but here only tetraploids were observed. Tetraploids (present in 23% of the 83 grouped taxa) are the most common followed by octoploids (4%) hexaploids (2%), triploids (2%) and heptaploids (1%). Lachenalia bulbifera is the species with the largest number of reported polyploid accessions including 4x, 6x, 7x and 8x accessions (Table 1). The heptaploid accession of L. bulbifera originated from seed and it is thus possible that the seed could have originated from an intra-species cross between a 6x and an 8x individual (Kleynhans and Spies 1999). Specific ploidy levels in L. bulbifera were better correlated to geographic distribution than morphology (Kleynhans and Spies 1999). somes. Idiograms for L. aloides and L. splendida constructed by Hamatani et al. (2009) again correlate with that of Hancke et al. (2001). Spies et al. (2000) reported that accessions of L. mutabilis contained 4 to 8 very short chromosomes. According to the authors the number of short chromosomes can vary between different localities and even between specimens collected at the same locality. Hamatani et al. (2007) furthermore reported on varying karyotypes within the same species for a number of Lachenalia species. This reported variation and conflicting results thus indicate that karyomorphological data alone cannot be utilized successfully to construct phylogenetic relationships in the genus Lachenalia. Similar conclusions were reached by Hamatani et al. (2008), resulting in a movement towards molecular methods to determine phylogenetic relationships in the genus. Basic chromosome numbers and polyploidy Moffett (1936) identified four different basic chromosome numbers (x = 7, 8, 11 and 13) and polyploids, including 3x, 4x and 6x, in the x = 7 group. De Wet (1957) added a basic chromosome number of x = 12 and reported on an accession with 2n = 56, a possible 8x. Ornduff and Watters (1978) added x = 6, in an unidentified species as well as x = 5 and x = 9. Johnson and Brandham (1997) added x = 10 and 15. For the purpose of this review, the 89 species in Table 1 was grouped according to their basic chromosome numbers. Basic chromosome numbers of x = 5, 10 and 15 were also included as existing basic numbers for the genus and not as polyploid forms of basic group x = 5. Of the 89 species six species (L. mediana, L. latimerae, L. isopetala, L. nervosa, L. congesta and L. capensis) could not be placed into a specific basic chromosome number due to varying reports in literature indicating different basic chromosome numbers within these species. It is possible that L. mediana has two different basic chromosome numbers and that x = 9 are present in L. mediana var. mediana and x = 13 are found in L. mediana var. rogersii (Spies et al. 2008, 2009). More studies are, however, required to accurately place these six species. Other species with varying chromosome number reports were placed into specific groups according to the most commonly reported chromosome number (Table 1). These include: x basic group x = 8 (L. contaminata 14 out of 17 reports indicate 2n = 16); x basic group x = 7 (L. barkeriana 3 out of four accessions had 2n = 14, L. marginata 4 out 5 reports indicate either 2n = 14 or tetraploids of x = 7, L. orchioides – majority of reports indicate x = 7 and 2n = 16 most probably from wrongly identified species, L. pusilla as 8 out of 9 reports indicate 2n = 14, L. reflexa as 5 out of 6 reports indicate 2n = 14 and the 2n = 16 could most probably be ascribed to the presence of B-chromosomes, L. variegata as 3 out 4 reports indicate basic x = 7 and L. violaceae as 15 out of 17 reports indicate basic x = 7); x basic group x = 10 (L. alba as 4 out of 5 had 2n = 20 and Johnson and Brandham (1997) concluded that 2n = 20 forms a diploid based on x = 10 rather than a tetraploid based on x = 5); x basic group x = 11 (L. hirta as 8 out of the 12 reports 105 Review of the genus Lachenalia. Kleynhans et al. 35 30 25 2x 3x 20 4x 15 6x 7x 10 8x 5 0 x=5 x=6 x=7 x=8 x=9 x=10 x=11 x=12 x=13 x=15 Fig. 4 Basic chromosome numbers in the genus Lachenalia indicating the number of taxa for each basic number and the ploidy levels reported for these basic numbers. PHYLOGENETIC STUDIES The only other species with ploidy levels above tetraploid are L. elegans and one report of 8x in L. mutabilis (Table 1). The two triploid accessions in L. aloides and L. rosea could have resulted from intra-species crosses between diploid and tetraploid individuals in these species followed by vegetative propagation or through an unreduced gamete followed by vegetative propagation as suggested by Moffett (1936). Only a few molecular studies have been done on Lachenalia and most of these studies concentrated on the phylogenetic position of the genus. The extensive variation in the genus, and even within a species, as indicated by RAPD studies (Kleynhans and Spies 2000), complicates both the phylogeny and taxonomy. In cultivation, a number of species are easily crossed and reproduce by means of offshoots and bulb formation. The existence of possible natural hybrid species thus further complicates the phylogenetics of the genus. Meiotic studies Reports on meiotic studies within the genus are less frequent. Moffett (1936) again presented the first report on meiosis. The author found mostly normal meiosis for 2n = 14, 16 and 22 species. The only differences were reported where ploidy was present. Hancke and Liebenberg (1998) reported on the meiosis of several 2n = 14 species and hybrids. Species studied displayed normal meiosis with 7 bivalents. Four of the six hybrids studied also displayed normal meiosis with 7 bivalents indicting a close relationship between the species L. aloides, L. orchioides, L. viridiflora and L. reflexa. Two hybrids (both between L. aloides and L. mutabilis) displayed a low percentage of trivalents and quadrivalents. Hancke and Liebenberg (1998) presented evidence of structural chromosomal changes involving three chromosomes of which the acrocentric pair of chromosomes was involved in at least one interchange. This chromosome pair also seemed to be prominent in other abnormalities observed during meiosis (Hancke and Liebenberg 1998). Hancke et al. (2001) studied the chromosome associations of one interspecific dibasic hybrid between L. splendida and L. aloides and two interspecific dibasic hybrids between L. unicolor and L. aloides. Results showed that L. aloides is more closely related to both L. splendida and L. unicolor than expected with genome affinity indexes of 0.9 and above. The results of the pairing configurations observed in these hybrids revealed homoeology between two chromosomes of the x = 7 karyotype and three chromosomes of the x = 8 karyotype. This could indicate that the x = 7 plants differ from the x = 8 plants by at least two exchanges of chromosome material and involves also the loss of one centromere from the x = 8 karyotype. Hancke et al. (2001) thus suggested that the change in basic chromosome number of Lachenalia involves a reduction in number. Du Preez et al. (2002) reported on normal meiosis with 8 bivalents for the following species, as well as the hybrids between L. carnosa and L. splendida, L. splendida and L. carnosa, L. unicolor and L. carnosa and L. carnosa and L. framesii. This study indicated that these species are closely related. Hamatani et al. (2009) confirmed this relationship. The phylogenetic position of Lachenalia The genus Lachenalia was included in several studies to determine the phylogenetic position and classification of the different species, the first being the inclusion of the genus in the family Liliaceae. Lachenalia was reclassified in the family Hyacinthaceae (Perry 1985) up to 2009, where after the family Hyacinthaceae was dissolved into other families. Lachenalia now belongs to the family Asparagaceae (APG III group 2009). To find the relative position of Lachenalia in the Asparagaceae, Pfosser and Speta (1999) used sequences of the trnL-F chloroplast region. From these results the authors were able to group Lachenalia in the tribe Massonieae (which consists of all the South African genera investigated, such as Drimiopsis, Ledebouria and Polyxena). This study also presented the first evidence suggesting a close relationship between Lachenalia and Polyxena, with a bootstrap support of 100%. This was in contrast to that of MüllerDoblies and Müller-Doblies (1997), which placed Lachenalia in the subtribe Lachenaliinae and Polyxena into Massoniinae. Pfosser and Speta (1999) suggested further studies, since only a few representative species were included in their analysis. A later study (Pfosser et al. 2003) included not only more Lachenalia species, but also an additional chloroplast region (atpB), as well as data on seed morphology. Polyxena, Lachenalia and the genus Periboea formed a monophyletic clade with a bootstrap support of 100%. This study thus also supported the inclusion of Polyxena in the genus Lachenalia. Within the monophyletic clade some species of Lachenalia and Polyxena had low bootstrap support values (66% and 62%, respectively) and it was suggested that the specific delimitation may not be optimal for these clades. Another explanation was that the species are more recently derived, resulting in an insufficient number of base substitutions to resolve the taxa. The authors suggested that seed size and weight is higher in the basal genera such as Eucomis, Merwilla and Ledebouria, with Veltheimia brac- 106 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books CROSS-ABILITY IN LACHENALIA teata having seeds of 0.056 g and with a length of 6.1 mm. The smallest seeds were found in the genus Lachenalia (L. angelica: 0.0003 g; 0.9 mm long). Analysis on the seed size and weight supports the hypothesis of the authors that Lachenalia is a recently derived genus. The seed form and structure of the micropylar swelling of the seed coat in Lachenalia suggested that this genus was the most advanced in their study. The inclusion of Polyxena in the genus Lachenalia was raised again in three separate studies (Manning et al. 2004; Spies 2004; Hamatani et al. 2008) using rbcL, trnL-F and ITS1-2 sequencing data respectively. In all these studies, Lachenalia and Polyxena formed a well supported monophyletic group. The two genera were characterised from other genera in the family by their biseriate stamens with the two series inserted at different heights. The two genera can be distinguished from each other by the relative fusion of the perianth (Manning et al. 2002). Manning et al. (2004) thus included Polyxena within Lachenalia based on the paraphyletic nature of the two genera. Rev. John Nelson raised the first authenticated Lachenalia hybrid in 1878 (Moore 1905). Since then a number of claims of interspecific hybridization were published (Crosby 1978, for review of early work). None of these early hybrids became available commercially. In 1965 the genus was identified as an indigenous genus with potential for development in South Africa. A breeding programme for the development of flowering pot plants was started at the Roodeplaat Vegetable and Ornamental Plant Institute of the Agricultural Research Council and the first hybrids became available commercially in 1997/1998 (Kleynhans 2006). The extensive morphological and cytological variation in the genus Lachenalia resulted in the existence of both internal and external crossing barriers (Lubbinge 1980; Kleynhans and Hancke 2002; Kleynhans 2006). External crossing barriers like geographical separation and varying flowering periods can be overcome through the cultivation of species in controlled environments and the successful storage of pollen for a 12 month period (Kleynhans 2006). Internal crossing barriers include both post- and pre-fertilization barriers. Mechanical isolation (Lubbinge 1980) is one of the first internal pre-fertilization barriers. Flower length in Lachenalia species can vary from 5 to 30 mm (Duncan 2005). Pollen from small flowered species is thus not adapted to traverse the long distance from the stigma to the ovary of large flowered species (Stebbins 1950). The utilization of reciprocal crosses has been successful in overcoming this barrier (Lubbinge 1980; Kleynhans 2006). Other pre- and post-fertilization barriers have not been studied in detail, but the extent of these barriers become clear when the success rate of inter-species crosses are taken into account. For each crossing combinations at least 10 flowers, within two different inflorescences were pollinated to ensure that wrong conclusions were not drawn, due to specific physiological or developmental problems in the inflorescence or floret. Kleynhans et al. (2009) reported that only 33% of the inter-species crosses (1498) made over a 30 year period were successful. With additional crosses (382) made since 2005, this percentage dropped to only 18% (Table 2). Of the 82% that did not succeed, 50% was related to the absence of seed, indicating the presence of possible prefertilization barriers. A further 31% of the combinations produced abnormal or non-viable seed that could be ascribed to post-fertilization barriers. Lastly, 1% of the crossing combinations did not succeed due to seedling death shortly after germination. The reason for the death of Phylogeny within the genus Morphological studies have focused on the entire genus, and many species have, over time, been included and excluded and shifted around from one genus to another. The first of these was when the genus was split into several genera (Salisbury 1866). Later on the species in the genus were sub-divided into smaller groups by Baker (1897), Crosby (1986) and Duncan (1988, 2002). These groupings, except for that of Crosby (1986) were based on different morphological characteristics, and did not correspond with each other. Duncan et al. (2005) used morphological data of all the species in the genus to construct a cladogram. The author included 73 characters which comprised of 57 qualitative and 16 quantitative characters. This study concluded that Polyxena is paraphyletic with Lachenalia and forms the basal clade. Many of the Lachenalia species formed polytomies or unrelated groups, but there were some synapomorphies or taxa sharing some traits. Spies (2004) produced a cladogram based on chloroplast trnL-F sequencing data from 129 taxa, including four Massonia taxa as outgroup. Hamatani et al. (2008) investigated nuclear ITS1-2 sequencing data of 56 taxa, including two Massonia and one Ornithogalum as outgroup. Both authors identified specific clades within the genus Lachenalia. The topologies of the cladograms produced by these authors largely correspond. Table 2 Number of inter-species crosses made among various different Lachenalia species over a 35 year period and the results obtained from these crossing combinations. Crosses that did not succeed were linked to three different aspects namely no seed set, abnormal seeds or seedling death. Results are linked to the basic chromosome complement of the species. No of unsuccessful crosses Basic chromosome number of parents No. of successful crosses No. of crosses with no No. of crosses with No. of crosses with seed set abnormal seed seedling death 7x7 169 (27%) 274 (44%) 169 (27%) 10 (2%) 8x8 72 (46%) 44 (28%) 40 (45%) 1 (1%) 11x11 2 (67%) 1 (33%) 7x8 20 (6%) 251 (79%) 44 (14%) 3 (1%) 8x7 59 (18%) 111 (34%) 155 (47%) 6 (2%) 7x10 17 (100%) 10x7 1 (5%) 5 (25%) 13 (65%) 1 (5%) 7x11 1 (2%) 54 (86%) 8 (13%) 11x7 4 (6) 23 (33%) 39 (57%) 3 (4%) 9x8 1 (100%) 8x10 1 (33%) 2 (67%) 10x8 2 (33%) 1 (17%) 2 (33%) 1 (17%) 8x11 1 (3%) 23 (79%) 5 (17%) 11x8 1 (3%) 15 (39%) 22 (58%) 11x10 1 (100%) 15x7 2 (67%) 1 (33%) Unknown basic numbers in one or both of the parents 4 (2%) 117 (59%) 78 (39%) Total 336 (18%) 939 (50%) 580 (31%) 25 (1%) 107 Review of the genus Lachenalia. Kleynhans et al. these seedlings can not necessarily be ascribed to hybrid breakdown, as seedlings can also be affected by diseases. The genetic variability within the genus as described above has a direct influence on the cross-ability. With the additional data presented in this review the comparison between cross-ability and the cytogenetic and molecular data will be discussed in the next section. bers and phylogenetic groupings could in the future be used to confirm basic numbers for species. A single count of 2n = 32 was reported for L. giessii but based upon a close phylogenetic grouping with x = 11 (Spies 2004), it seems that this species could also be regarded as x = 11 (2n = 33) rather than x = 8 (2n = 32). In this review it was included as a tetraploid of x = 8 for the purpose of calculations, but this species should be investigated further. Similarly L. capensis groups with the x = 7 group (Spies 2004) thus supporting the chromosome counts of Johnson and Brandham (1997) and Spies et al. (2008) and suggesting that L. capensis could be a basic x = 7 rather than a basic x = 8 as reported by Hamatani et al. (1998). Further investigations and correct identification of species are, however, essential to solve the inconsistent reports in chromosome numbers in some species. COMPARISON BETWEEN CROSS-ABILITY, CYTOGENETIC AND MOLECULAR DATA The complexity in the genus, in terms of morphology, cytogenetic and genetic variation complicates the determination of the relationship within and between different species. There are questions on the existence and origin of the different basic chromosome numbers, as well as the mode of speciation. Does the different basic chromosome numbers correlate with the phylogeny of the genus? Can the phylogenetic information assist in the taxonomic grouping of some difficult species and, furthermore, can phylogenetic information shed some light on the existence of possible natural hybrids? How does the phylogeny correlate with the cross-ability between species and finally what conclusions can be drawn when the different data sets are compared. Basic chromosome number and cross-ability Kleynhans et al. (2009) presented data showing that the success rate of crossing combinations increased when crosses were made between species containing the same basic chromosome number. The information from additional crosses made in the last five years were added to this data and the number of successful crosses between species with the same basic chromosome number was substantially higher than between species from different basic chromosome numbers (Table 2). The success rate of crossing combinations dropped to below 20% when species with different basic chromosome numbers were crossed. The only exception to this is the combination of basic x = 10 crossed with basic x = 8 (Table 2). The two successful crosses resulted from a L. alba x L. unicolor and L. alba x L. pustulata combination (specific results not shown). The increased success rate reported between species with the same basic chromosome number were a confirmation of a report by Crosby (1986) who also indicated that species cross more readily within certain basic chromosome number groupings. Based on differences in the cross-ability and morphology the latter author also split the basic x = 7 group of species into two different groups. The existence of different groupings within the basic x = 7 was confirmed by Spies (2004) as discussed above. Meiotic data presented by Hancke and Liebenberg (1998), as discussed above, also indicated differences between especially the species L. mutabilis and L. aloides as illustrated by structural chromosome changes. Kleynhans et al. (2009) used the three basic clades as well as the phylogenetic groupings within the basic x = 7 group as reported by Spies (2004) and presented data that showed improved cross-ability when crosses were made between individual species within the same phylogenetic groupings. The cross-ability was at least 10 to 20% higher when crossing combinations were attempted within the groups, than between groups. The cross-ability data thus supported phylogenetic groupings as identified by Spies (2004). The close relationship illustrated in the phylogenetic trees, between species with basic x = 8 was also confirmed by the cross-ability data with a success rate of 46% (Table 2). The only success rate higher than this was that between species with basic x = 11. This data, however, only included 3 crossing combinations in comparison to the 157 combinations within the basic x = 8 group and would most probably decline with the inclusion of additional crossing combinations. The relationship among species with x = 8 was further illustrated by Du Preez et al. (2002). In this meiotic study several hybrids between different species with x = 8 were investigated and all hybrids produced 8 bivalents. Hybrids resulting from these crosses are also fertile and was successfully utilized in further crossing combinations (results not shown). Basic chromosome numbers and cladograms A comparison between the groupings from Crosby (1986) (based on chromosome numbers), Spies (2004) (chloroplast trnL-F), Duncan (2005) (morphology) and Hamatani et al. (2008) (nuclear ITS1-2) revealed that, with the exception of a few species, there is a good correlation between the basic chromosome numbers and the monophyletic groups identified in the different studies. When chromosome numbers were superimposed on the cladogram of Duncan et al. (2005) most of the x = 7 and x = 8 species fall into exclusive monophyletic groups for each chromosome number. There are only two exceptions where x = 7 species (L. congesta and L. mathewsii) grouped with x = 8. Species with x = 11 were closely related, even though they did not form a monophyletic group. The rest of the chromosome numbers form a polytomy. Although monophyletic groups linked to basic chromosome numbers were obtained the morphological cladogram is poorly resolved for many of the species. The study using trnL-F chloroplast DNA sequences (Spies 2004) of 129 taxa distinguished several well defined groups. The first group consisted of seven species with a basic number of 11. Species with x = 7 and 8 formed a monophyletic clade (the Lachenalia 1 group), suggesting a close relationship between these two basic numbers. Within this monophyletic clade, x = 8 formed a monophyletic subclade excluding only one species with a basic chromosome number of x = 8, L. verticillata, and including L. pusilla (x = 7), which was basal to this group. All species having a basic chromosome number of x = 7, were distributed in different sister subclades, of which the two largest x = 7 subclades includes 25 and 10 taxa respectively. The second largest group in the cladogram (the Lachenalia 2 group), consisted of 48 poorly resolved taxa having chromosome numbers of x = 6, 7, 8, 9, 10 and 13. This group has no consistent pattern regarding chromosome numbers. These results led the author to conclude that hybridization might have played a role in speciation and that the genus might represent a hybrid swarm. In the cladogram based on ITS1-2 sequencing data (Hamatani et al. 2008), a monophyletic group for x = 8 (supported with a bootstrap value of 83.3) as well as for x = 7 forming a polytomy was obtained. Two species, L. muirii and L. pusilla both with a basic number of 7, grouped with the x = 8 clade, but formed the base for the rest of the x = 8 species. The ITS1-2 region seemed to have more variation in the x = 8 taxa than in the x = 7 taxa, since the clade for x = 8 was better resolved. A similar observation was made by Spies (2004) with the trnL-F sequences. The good correlation between basic chromosome num108 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books Evolution and relatedness of different basic chromosome numbers in sequencing data) or they could be the product of mutation or putative hybridization between species in the same geographical distribution area. Reduction in chromosome number either by losing a chromosome or by translocation might have contributed to speciation in these two groups. Hancke et al. (2001) speculated that x = 7 evolved from x = 8 through a reduction in chromosome number based on the homoeology between two chromosomes in the x = 7 and three chromosomes in the x = 8 species studied. Five of the nine species in the x = 7 group (L. aloides var. aloides, L. aloides var. aurea, L. longibracteata, L. variegata, L. viridiflora) had very similar chromosome morphology (Hamatani et al. 2009) and seemed to be closely related. The close relationship between L. aloides and L. viridiflora can be confirmed from crossing data with a success rate of between 25 and 100% depending on the reciprocal direction (data not shown) and the production of fertile F1 hybrids with seven bivalents in meiotic analysis (Hancke and Liebenberg 1998). According to (Hamatani et al. 2009) the chromosome morphology of L. mutabilis and L. rubida were very similar, but differed from the above group, and the authors concluded that these species probably originated from a single ancestral species. For the purpose of this review a selection of ITS1-2 sequences representing only those species used in the FISH study (Hamatani et al. 2009) were obtained from Genbank and a phylogram was constructed (Fig. 5). The tree was drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The ITS phylogram yielded similar monophyletic groupings than the ITS1-2 cladogram (Hamatani et al. 2009) and included both L. mutabilis and L. rubida within the x = 7 clade. Both these species have a similar branch length that was much longer than the other species in the clade, which supported the similarity in chromosome morphology. This relationship cannot be confirmed from crossing data (success rate of only 10%), neither by the data presented by Spies (2004) or Hamatani et al. (2008). The largest number of species in Lachenalia are found within the basic x = 7 and 8 groups. Molecular data from ITS1-2 (Hamatani et al. 2009) and trnL-F (Spies 2004) sequences indicated a strong relationship between these two basic chromosome number groups and that these groups might have evolved from a common ancestor. Cross-ability data confirmed a relationship between these two basic chromosome number groups with higher success rates (18% for x = 8 crossed with x = 7), than most of the other between group success rates (Table 2). The existence of genome affinity indices of 0.9 in three interspecific dibasic hybrids (Hancke et al. 2001), as discussed above, also confirmed this relationship. Karyomorphological data presented by Hamatani et al. (2009) using FISH and DAPI staining to determine the chromosomal evolution of the x = 7 and x = 8 groups confirmed the results found from both the phylogeny and the cross-ability. The results of this study between a group of x = 7 (consisting of L. muirii, L. aloides var. aloides, L. aloides var. aurea, L. longibracteata, L. variegata, L. viridiflora, L. mutabilis, L. rubida, and L. pusilla) and x = 8 (consisting of L. carnosa, L. liliflora, L. namaquensis, L. splendida and L. unicolor) led to the conclusion, that there was little morphological chromosome variation within the x = 8 group and that this group was derived from an ancestral species followed by ongoing speciation. The x = 7 group showed much more variation, with four karyotype patterns indicating several morphological alterations of chromosomes within this group. This was in contrast with the ITS1-2 region data that seemed to have more variation in the x = 8 taxa than in the x = 7 taxa, since the clade for x = 8 was better resolved than the polytomic x = 7 clade (Hamatani et al. 2009). Hamatani et al. (2008, 2009) suggested several theories for the evolution of the x = 7 and 8 groups. Both groups might have evolved from a common ancestor (as indicated L. viridiflora L. variegata L. longibracteata L. longibracteata L. variegata 㪌㪐 L. aloides var. aurea L. viridiflora 㫏㩷㪔㩷㪎 L. muirrii L. aloides var. aloides L. mutabilis L. aloides var. aurea L. rubida 㪏㪎 L. rubida L. pusilla L. liliflora 㪉㪈 L. carnosa L. splendida 㫏㩷㪔㩷㪏㪍㪊 L. unicolor L. namaquensis L. namaquensis 㪇㪅㪇㪇㪇㪌 Fig. 5 Evolutionary relationships of 17 taxa based on the ITS1-2 region. The phylogram was constructed using the Maximum Likelihood option of MEGA 5 (Tamura et al. 2011) to compare the evolutionary development of the x = 7 and 8 groups. 109 Review of the genus Lachenalia. Kleynhans et al. L. mutabilis L. reflexa 㪈㪏 L. viridiflora L. variegata L. muirii 㪊㪌 L. aloides var. vanzyliae 㫏㩷㪔㩷㪎 L. bulbifera 㪍㪋 L. bulbifera 㪍㪈 L. carnosa (x = 8) 㪋㪉㪊㪊㪍㪌 L. rubida L. variegata L. pusilla (x = 7) 㪌㪈 L. liliflora L. rosea 㪊㪍 L. purpureo-caerulea 㪋㪉 L. namaquensis L. splendida 㪋㪈㫏㩷㪔㩷㪏 L. unicolor L. contaminata L. pallida 㪉㪈㪉㪇 L. pustulata L. doleritica (x = 9) L. latimerae (x = 9) L. alba (x = 10) L. comptonii (x = 10) L. duncanii (x = 9) L. obscura (x = 9) L. convallarioides (x = 10) L. neilii (x = 9) L. minima (x = 9) 㪏㪍 L. corymbrosa L. paucifolia L. maughanii 㪍㪍㪧㫉㪼㫍㫀㫆㫌㫊㫃㫐㩷㪧㫆㫃㫐㫏㪼㫅㪸 L. ensifolia 㪊㪋㪊㪏 L. odorata 㪊㪏 L. zeyheri L. juncifolia 㪍㪈 L. unifolia 㪐㪎㫏㩷㪔㩷㪈㪈 L. anguinea (2n = 30+2B) 㪎㪉㪈㪐 L. hirta L. isopetala L. nervosa (2n = 16/24) 㪋㪋 L. staynerii (2n = 24) 㪍㪊 L. mediana var. rogersii (x = 13) 㪍㪋㫏㩷㪔㩷㪈㪊 L. mediana var. mediana (x = 9) 㪐㪐㪍㪋 L. mediana var rogersii (x = 13) Fig. 6 Evolutionary relationships of 43 taxa based on the trnL-F region (Spies 2004), inferred using the Maximum Likelihood option of MEGA 5 (Tamura et al. 2011). The remaining two species in the x = 7 group that were investigated (Hamatani et al. 2009), L. muirii and L. pusilla, shared chromosomal characteristics with species in both the x = 7 and 8 groups. The relationship to both x = 7 and 8 of L. 110 muirii and L. pusilla was confirmed by Hamatani et al. (2008). Hamatani et al. (2009) suggested that L. pusilla might be intermediate between the x = 7 and x = 8 group. None of the crosses made with L. pusilla as either parent Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books L. unicolor L. contaminata L. unicolor Massonia depressa L. liliflora L. unicolor L. unicolor L. pallida L. bachmanii Ornithogalum umbellatum Massonia pustulata L. namaquensis L. hirta 2 1 L. hirta L. muirii L. latimerae L. mutabilis *L. latifolia L. pusilla L. pusilla L. zeyheri L. rosea L. bulbifera L. arbuthnotiae L. rubida L. reflexa L. peersii L. orchioides 3 L. juncifolia L. longituba L. unifolia L. paucifolia L. reflexa L. ensifolia L. algoensis L. variegata L. reflexa Fig. 7 Network of Lachenalia species based on ITS data using NETWORK 4.6.1.0 (Fluxus Technology, 2012). The correct current citation of L. latifolia (indicated with *) is L. nervosa. Colour codes: Red, x = 7; Yellow, x = 8; Blue, x = 11; Purple, 2n = 24/26/28; Grey, x = unknown. Node 1, L. pustulata and L. purpureo-caerulea; Node 2, L. carnosa and L. splendida; Node 3, L. aloides var. aloides, L. aloides ‘Pearsonii’, L. aloides var. luteola, L. aloides var. vanzyliae, L. aloides var. quadricolor, L. aloides var. aurea, L. viridiflora, L. orchioides var. orchioides and L. longibracteata. were successful, neither with x = 7 nor with x = 8 species. The cross-ability data available can thus not shed any light on the position of L. pusilla. There seem to be an evolutionary relationship between some of the other basic chromosome number groups and even with other genera. For better insight in the evolution of the rest of the chromosome numbers, sequences from Spies (2004) were selected to represent a broad spectrum of chromosome numbers in the genus. Sequences were selected based on the cladogram produced by Spies (2004), but all sequences forming a polytomy were excluded, and a new cladogram (Fig. 6) was constructed. Although many of the clades are not well supported, the new trnL-F cladogram (Fig. 6) supports the suggestion that the genus evolved from a common ancestor. The basic numbers x = 7 and 8 evolved from a common predecessor, even though many of the clades are not well supported, thus confirming the data presented above. The higher basic numbers (x = 9, 10, 11 and 13) form a poorly supported monophyletic clade (bootstrap value 57). It seems as if the higher numbers evolved independently from the lower numbers in at least two separate events. The basic numbers x = 9 and 10 forms a polytomy in the higher clade and seems to be the bridge from the lower to the higher numbers or vice versa (Fig. 6). Because none of the x = 9 or 10 taxa are well resolved, this group might be a recent group. The low level of variation in these two basic numbers indicates that evolution was recent and these numbers have not evolved into two definite clades. A median-joining network (Bandelt et al. 1999) was constructed from the ITS data (Hamatani et al. 2008) (Fig. 7) as well as from 43 trnL-F sequences (Spies 2004) (Fig. 8). The trnL-F network suggests that x = 11 and x = 8 have evolved independently from a common ancestor, and that x = 9 and 10 could have evolved from any one of these two numbers. The ITS network (Fig. 7) could not confirm or reject this, due to the lack of x = 10 species and the inclusion of only a single x = 9 species. Both the networks support a close relationship between the x = 7 and 8 groups. The cross-ability success rate of 33% between basic x = 10 and basic x = 8 (Table 2) could be a confirmation of the possible bridge between x = 7 and 8 and the higher numbers. The ITS network also supported the relationship between L. mutabilis and L. rubida (Fig. 5) and the trnL-F network positioned L. pusilla in an ancestral position to x = 7 and 8 thus supporting the molecular cytogenetic data. Dysploidy (through the fusion of acrocentric chromosomes at the centromere to form larger metacentric to submetacentric chromosomes) has been shown to be important in the chromosomal evolution of other plant families, e.g. the Commelinaceae (Jones 1976). If dysploidy is the mode of speciation in Lachenalia a study on the chromosome morphology of species with higher basic chromosome numbers compared to lower basic chromosome numbers could assist in confirming the hypotheses. A study of L. latimerae (x = 9 according to Hamatani et al. 2007) indicated that this species has three large chromosomes, of which two are very similar, with the third one having a satellite (Hamatani et al. 2007). The chromosome morphology thus, supports the theory of dysploidy, but it must be further investigated with chromosome banding techniques. A second hypothesis is the possibility that L. latimerae could have resulted from a hybridization event (Hamatani et al. 2007) between x = 7 and x = 11, resulting in a gametic number of n = 18. If this 111 Review of the genus Lachenalia. Kleynhans et al. L. anguinea L. hirta L. zeyheri L. juncifolia 5 L. unifolia Massonia depressa L. isopetala L. comptonii Ornithogalum umbellatum L. corymbosa L. paucifolia L. maughanii L. pusilla L. ensifolia L. odorata L. contaminata L. convallarioides L. duncanii L. variegata L. mutabilis L. rubida L. latimerae L. minima L. doleritica 1 L. obscura L. variegata L. liliflora L. mediana L. rosea L. mediana L. mediana L. muirii L. reflexa 4 L. carnosa L. pustulata 2 L. bulbifera L. bulbifera 3 L. pallida Fig. 8 Network of Lachenalia species based on trnL-F data using NETWORK 4.6.1.0 (Fluxus Technology, 2012). Colour codes: Red, x = 7; Yellow, x = 8; Blue, x = 11; Light purple, 2n = 24/26/28; Dark purple, x = 9; Orange, x = 10; Diagonal crosses, x = 9 or 13; Grey, x = unknown. Node 1, L. neilii; L. alba; Node 2, L. purpureo-caerulea; L. unicolor; Node 3, L. namaquensis; L. splendida; Node 4, L. viridiflora; L. aloides var. vanzyliae; Node 5, Massonia pustulata; M. depressa; M. echinata; M. jasminiflora. investigated to the same extend as x = 7, 8, 9 and 11. With basic chromosome numbers of 5, 6, 7, 8, 9, 10, 11, 12, 13 and 15 recorded, it is still speculated whether basic numbers of x = 5, 6, 10, 12, 13 and 15 exists. There are very few reports for n or x = 5 in Lachenalia, and usually when x = 5 has been reported for a species, it was based only on one accession. Both L. violacea and L. aloides are x = 7 species, with a single 2n = 15 reported, indicating possible miss counts in these species. Lachenalia mutabilis has chromosome counts of x = 5, 6 and 7. This is the only species where numerous counts have been recorded for all three these numbers. This species is morphologically distinct and wrong identification could not attribute to the differences in counts. All reports for x = 5 for L. mutabilis are from the same geographical distribution area (Clanwilliam in the Western Cape Province), but there are also reports of x = 7 from Clanwilliam. Other species from the Clanwilliam district include x = 7 (L. elegans var. sauveolens, L. thomasiae and L. violaceae); x = 8 (L. unicolor); x = 10 (L. marginata and L. undulata) and x = 11 (L. hirta and L. unifolia). It was suggested that the three basic numbers for L. mutabilis form an aneuploidy series (Spies et al. 2000), but there is no proof of what attributed to the chromosome diversity in this species. Based on molecular systematics, L. mutabilis specimens always group with other x = 7 species, regardless of their chromosome number (Spies 2004; Hamatani et al. 2008); are karyotypically similar to L. rubida (x = 7) and has the highest number of x = 7 counts recorded, thus supporting the theory of an aneuploid series in the species. Johnson and Brandham (1997) studied the karyotypes of x = 7-13 and 15, and reported that all the species studied formed structural diploids and thus concluded that 2n = 20 rather represents a diploid based on x = 10 than a tetraploid based on x = 5. They did state that 2n = 30 (x = 15) could be an allotetraploid derived from taxa with x = 7 and 8, following hybridization and doubling of the chromosome number. Considering this theory, it would be expected that x = 10 taxa have a phylogenetic grouping either with x = 7 or x = 8 taxa, but this have not been observed in the trnL-F cladogram (Spies 2004). The fact that the cross-ability theory is correct for other x = 9 species, one would expect at least some of the x = 9 species to group with either x = 7 or x = 11 in the chloroplast cladogram. All the x = 9 species fall between the x = 7/8 groups and the higher numbers, but because the trnL-F cladogram (Fig. 6) is not supported with high bootstrap values, neither the dysploid theory nor the hybridization theory could be proven. The trnL-F medianjoining network (Fig. 8) is inconclusive in this matter, since the evolutionary direction for x = 9 can be from either x = 11 or x = 7/8 or both (thus hybridization). The group x = 11 is very well supported with a bootstrap value of 94 in the trnL-F cladogram (Fig. 6), suggesting a strong relationship within this group. The close relationship within this group is also supported by the morphological cladogram constructed by Duncan (2005), even though these species do not form a monophyletic group. The evolution of x = 11 is not clear, but from the cladograms obtained in the different studies i.e. morphological (Duncan et al. 2005), ITS (Hamatani et al., 2008) and trnL-F (Spies 2004), x = 11 (and x = 13) is basal to the lower numbers and it seems that species with x = 11/13 is the intermediate between the outgroup species (which have higher numbers) and the lower numbers in the genus. The network drawn from the ITS sequences provides evidence of the link between the higher basic numbers in Lachenalia and outgroup species used in this study. The outgroup for the ITS network (Fig. 7) is Massonia and Ornithogalum umbellatum. The latter species has a high degree of cytogenetical variation (Czapik 1968) with numbers of 2n = 1830 and B-chromosomes reported. Hamatani et al. (2008) obtained the ITS sequences for L. hirta (x = 11) by cloning the maternal and paternal genomes. One genome was cloned in some specimens and seem to have evolved from Massonia, while the other genome have evolved from Ornithogalum this may be the reason why different specimens form two different nodes in the network. Existence of basic chromosome numbers The evolution and even existence of certain chromosome numbers (such as x = 5, 6, 12, 13 and 15) have not been 112 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books between x = 10 and 8 is relatively high could be an indication of the validity of this theory. The existence of the basic number x = 10, however, seem to be a reality, proven by the fact that some species has chromosome counts of 2n = 20, 40 (L. alba) and 2n = 30, 40 (L. isopetala – not grouped in this study) indicating the existence of polyploids. After all the evidence, it is still not clear whether x = 5 exist in any other species than L. mutabilis. Reports for six species with either x = 6 or 2n = 24 were mostly based on only one accession and differed from the majority number of counts for these species. Lachenalia nervosa has counts of n = 8 and 2n = 24, indicating that this species has a basic number of x = 8 and have a triploid somatic number. Lachenalia stayneri is also 2n = 24, and the lack of meiotic studies in this species may lead to the conclusion that this species represents a tetraploid based on x = 6 or also a triploid with x = 8. Therefore x = 6 should also be considered as a basic number. Based on trnL-F sequences, both these species indicate close relations with L. mediana (x = 9 and 13) and do not group with x = 8 (Spies 2004). Therefore, species with 2n = 24 cannot be considered as “typical” x = 8 species, and might even be considered as being miss counts based on x = 13. None of the 2n = 24 species has its own monophyletic grouping and it seems as if x = 6 does not exist except maybe in L. mutabilis. Somatic counts of 2n = 28 and 56 have been reported by several authors (Moffett 1936; de Wet 1957; Crosby 1986; Hancke and Liebenberg 1990; Johnson and Brandham 1997; Hamatani et al. 1998; Kleynhans and Spies 1999; Spies et al. 2002; Hamatani et al. 2007; Spies et al. 2008, 2009), but it has not been proven whether the basic chromosome number of x = 14 exists. Somatic numbers of 2n = 28 as sole chromosome number have been reported for L. cernua and L. longituba. Both these species were included in the basic group x = 7 for the purpose of this review, but additional accessions of these species, as well as meiosis and cytomorphological data will have to be studied to determine the actual basic chromosome number. Existence of hybrid species The question of natural hybridization in the genus has been raised several times. Both the morphological and trnL-F cladograms had monophyletic groups consisting of a mixture of chromosome numbers x = 6, 7, 8, 9, 10 and 13 and no consistent patterns regarding similar groupings. Spies (2004) concluded that hybridization might have a role in speciation, but it was not proven. Some species (L. pusilla, L. rosea and L. carnosa) do not follow the rule of grouping into monophyletic groups with similar chromosome numbers (Fig. 6). Considering the positions of these species in the networks drawn (Figs. 8, 9) the first two species is intermediate to the x = 7 and x = 8 groups in both networks. The position of L. carnosa (x = 8) fluctuate between x = 7 (Fig. 8) and x = 8 (Fig. 7). Within the trnL-F cladogram, L. carnosa, L. rubida and L. bulbifera is a sister clade with the rest of the x = 7 species. Lachenalia rubida is intermediate to x = 7 and 8 in both networks. To conclude, based on karyotypic and molecular data, some species are intermediate between x = 7 and 8, and can either be considered as predecessor species or as hybrid species. Lachenalia carnosa (x = 8) is an example of a possible hybrid species, grouping with either x = 7 or 8, depending on the type of sequencing data (nuclear or cytoplasmic). Spies (2004) reported what seemed to be B-chromosomes in the meiotic divisions if L. carnosa, which may have been unidentified univalents, also observed in cultivated Lachenalia hybrids (Hancke and Liebenberg 1998). Cross-ability data, however, strongly links L. carnosa with other members of the x = 8 group, successfully crossing with at least five different x = 8 species (data not shown), producing regular meiosis with 8 bivalents (Du Preez et al. 2002) as well as fertile hybrids. Natural hybridization may be present in the genus Lachenalia but this should be investigated 113 further. CONCLUSION This review accentuates the complex nature of the genus Lachenalia. Besides the extensive morphological variation that complicates the taxonomy of the genus, the genus is also exceptionally diverse in chromosome numbers. Lachenalia has different basic chromosome numbers (x = 5, 6, 7, 8, 9, 10, 11, 12, 13 and 15 reported in literature), contains polyploidy (ranging from triploids to octoploids), and includes B-chromosomes. Chromosome counts for the 89 species reported in literature varied from 2n = 10 to 56 and from n = 5 to 28. Polyploidy was reported in 19 taxa (23%), and is most common in the x = 7 group. The low cross-ability (only 18% successful interspecies crosses) reiterates this variation and stresses the importance of investigating the variation in order to develop breeding strategies to overcome the existing crossing barriers. Morphological and molecular phylogenetic studies confirm the complexity of the genus, but also assisted in drawing some conclusions on the relationship between species within the genus and the possible evolutionary history of the genus. Phylogenetic studies has assisted in finding the phylogenetic position of Lachenalia in relation to other genera (Pfosser and Speta 1999; Pfosser et al. 2003; Manning et al. 2004) and placed the genus within the Asparagaceae family (APG III group 2009). Morphological (Duncan et al. 2005) and phylogenetic studies within the genus (Spies 2004; Hamatani et al. 2008) supported the inclusion of Polyxena in Lachenalia, and this inclusion increased the number of recognised Lachenalia species to 126. Molecular studies on the trnL-F as well as ITS regions revealed monophyletic groupings of species containing the same basic chromosome numbers. This indicated a strong correlation between the phylogeny and basic chromosome numbers in the genus, although there were some exceptions in the larger trnL-F data set (Spies 2004). The good correlation between basic chromosome numbers and phylogenetic groupings could in the future assist to confirm basic numbers for species. The improved cross-ability when crosses were made between individual species within the same phylogenetic groupings confirms the phylogeny. Phylogenetic groupings, thus has to be taken into account when crossing combinations are planned to achieve better crossing success rates in the breeding programme. When comparing the different studies, Lachenalia might have evolved from a common ancestor and the two largest basic chromosome number groups, x = 7 and 8 have evolved from a common predecessor. The studies also indicated a close relationship between these two basic numbers, which is supported by higher success rates in cross-ability between these two groups. It seems as if the higher basic numbers (x = 9, 10, 11 and 13) evolved independently from the lower numbers and that basic numbers x = 9 and 10 could be the bridge from the lower to the higher numbers or vice versa (Fig. 6), but evidence of this is not conclusive (Figs. 7, 8). Dysploidy and hybridization might be the modes of speciation in some Lachenalia species but this could not be proven with molecular data and further studies are required to draw conclusions. The existence of some of the basic chromosome numbers reported (such as x = 5, 6, 10, 12 and 15) can been disputed. Only a few species can be linked to x = 5 and 6 and it is possible that these two basic numbers only exist as part of an aneuploid series in the species L. mutabilis. Further studies on species from these disputed basic chromosome numbers is needed to resolve the existence of all the reported numbers. This review indicates that different genetic studies on Lachenalia reveal similar results and stresses the importance of assessing the variation within complex genera to aid in decisions around breeding programme strategies. It is clear that inter-species crosses within phylogentic groups in Review of the genus Lachenalia. Kleynhans et al. the genus can improve the success rate of crossing combinations, but there are still many questions that remain unanswered. Further multidisciplinary studies are needed in the genus Lachenalia to solve the evolutionary history of this complex genus, to answer questions around species placement and the existence of basic chromosome number groups and to overcome crossing barriers. 18S rDNA probes and DAPI staining. Chromosome Botany 4, 57-63 Hamatani S, Tagashira N, Kondo K (2010) Molecular cytogenetic analysis in seven species of Lachenalia (Liliaceae) with the chromosome numbers of 2n=18, 22, 23, 26 and 28 by DAPI staining and FISH using 5S rDNA and 18S rDNA probes. Chromosome Botany 4, 57-63 Hancke FL (1991) ‘n Sitotaksonomiese ondersoek van sewe Lachenalia spesies vir gebruik in ‘n blomteeltprogram. MSc thesis, University of Pretoria, 74 pp Hancke FL, Liebenberg H (1990) B-chromosomes in some Lachenalia species and hybrids. South African Journal of Botany 56, 659-664 Hancke FL, Liebenberg H (1998) Meiotic studies of interspecific Lachenalia hybrids and their parents. South African Journal of Botany 64, 250-255 Hancke FL, Liebenberg H, Janse van Rensburg WJ (2001) Chromosome associations of three interspecific, dibasic Lachenalia hybrids. South African Journal of Botany 67, 193-198 Johnson MAT, Brandham PE (1997) New chromosome numbers in petaloid monocotyledons and in other miscellaneous angiosperms. Kew Bulletin 52, 121-138 Jones K (1976) Multiple Robertsonian fusions in the evolution of a plant genus. In: Jones K, Brandham PE (Eds) Current Chromosome Research, Elsevier North-Holland, Amsterdam, pp 220-221 Kleynhans R (2006) Lachenalia, spp. In: Anderson NO (Ed) Flower Breeding and Genetics: Issues, Challenges, and Opportunities for the 21st Century, Springer, Berlin, pp 491-516 Kleynhans R (2009) Back to basics for new crop development. Acta Horticulturae 836, 185-191 Kleynhans R (2011) Potential new lines in the Hyacinthaceae. Acta Horticulturae 886, 139-145 Kleynhans R, Hancke FL (2002) Problems and breeding strategies for the development of new Lachenalia cultivars. Acta Horticulturae 570, 233-240 Kleynhans R, Spies JJ (1999) Chromosome number and morphological variation in Lachenalia bulbifera. South African Journal of Botany 64, 357-360 Kleynhans R, Spies JJ (2000) Evaluation of genetic variation in Lachenalia bulbifera (Hyacinthaceae) using RAPDs. Euphytica 115, 141-147 Kleynhans R, Spies JJ, Spies P (2009) Cross-ability in Lachenalia. Acta Horticulturae 813, 385-392 Lubbinge J (1980) Lachenalia Breeding I. Introduction. Acta Horticulturae 109, 289-295 Manning JC, Goldblatt P, Fay MF (2004) A revised generic synopsis of Hyacinthaceae in sub-Saharan Africa, based on molecular evidence, including new combinations and the new tribe Pseudoprosperae. Edinburgh Journal of Botany 60, 533-568 Manning JC, Goldblatt P, Snijman D (2002) The Color Encyclopedia of Cape Bulbs, Timber Press, Portland, Cambridge, 486 pp Moffett AA (1936) The cytology of Lachenalia. Cytologia 7, 490-498 Mogford DJ (1978) Centromeric heterochromatin in Lachenalia tricolor (L.) Thunb. Journal of South African Botany 44, 111-117 Moore FW (1905) Lachenalia hybrids. The Gardener’s Chronicle 37, 210-211 Müller-Doblies U, Müller-Doblies D (1997) A partial revision of the tribe Massonieae (Hyacinthaceae). 1. Survey, including three novelties from Namibia: A new genus, a second species in the monotypic Whiteheadia, and a new combination in Massonia. Feddes Repertorium 108, 49-96 Müller-Doblies U, Nordenstam B, Müller-Doblies D (1987) A second species in Lachenalia subgen. Brachyscypha (Hyacinthaceae): Lachenalia barkeriana sp. Nov. from southern Little Namaqualand. South African Journal of Botany 53, 481-488 Nordenstam B (1982) Chromosome numbers of southern African plants 2. Journal of South African Botany 48, 273-275 Ornduff R, Watters PJ (1978) Chromosome numbers in Lachenalia (Liliaceae). Journal of South African Botany 44, 387-390 Perry P (1985) The restructuring of the family Liliaceae. Veld and Flora Sept., pp 66-68 Pfosser M, Speta F (1999) Phylogenetics of Hyacinthaceae based on plastid DNA sequences. Annals of the Missouri Botanical Garden 86, 852-875 Pfosser M, Wetschnig W, Ungar G, Prenner G (2003) Phylogenetic relationships among genera of Massonieae (Hyacinthaceae) inferred from plastid DNA and seed morphology Journal of Plant Research 116, 115-132 Reinten EY, Coetzee JH, Van Wyk BE (2011) The potential of South African indigenous plants for the international cut flower trade. South African Journal of Botany 77, 934-946 Riley HP (1962) Chromosome studies in some South African monocotyledons. Canadian Journal of Genetic Cytology 4, 40-55 Salisbury RA (1866) The Genera of Plants. A Fragment Containing Part of Liriogamae. J. v. Voorst, London SANBI (2009) List of SA red data listed species. Available online: http://www.sanbi.org/index.php?option=com_docman&task=documentdetails &id=43 Spies JJ, Du Preez JL, Minnaar A, Kleynhans R (2000) Hyacinthaceae: Chromosome studies on African plants. 13. Lachenalia mutabilis, L. pustulata and L. unicolor. Bothalia 30, 106-110 Spies JJ, Spies P, Reinecke SMC, Kleynhans R, Duncan GD, Edwards TJ (2008) Lachenalia. In: Marhold K (Ed) IAPT/IOPB chromosome data 5. Taxon 57, 554-555 ACKNOWLEDGEMENTS The Agricultural Research Council and the University of the Free State is thanked for the facilities and financial support during this study. REFERENCES Angiosperm Phylogeny Group III (2009) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Botanical Journal of the Linnean Society 161, 105-21 Baker JG (1897) Lachenalia Jacq. In: Thistleton-Dyer WT (Ed) Flora Capensis 6, Reeve and Co., London, pp 421-436 Bandelt H-J, Forster P, Röhl A (1999) Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16, 37-48 Barker WF (1983) A list of the Lachenalia species included in Rudolf Schlechter’s collections made in 1891-1898 on his collection trips in southern Africa, with identifications added. Journal of South African Botany 49, 45-55 Crosby TS (1978) Hybridisation in the genus Lachenalia. Veld and Flora Sept., 87-90 Crosby TS (1986) The genus Lachenalia. The Plantsman 8, 129-166 Czapik R (1968) Chromosome numbers of Ornithogalum umbellatum L. from three localities in England. Watsonia 6, 345-349 De Wet JMJ (1957) Chromosome numbers in the Scilleae. Cytologia 22, 145159 Dold AP, Phillipson PB (1998) A revision of Lachenalia (Hyacinthaceae) in the Eastern Cape, South Africa. Bothalia 28, 141-149 Du Plessis NM, Duncan GD (1989) Bulbous Plants of Southern Africa, Tafelberg Publishers, Cape Town, pp 78-82 Du Preez JL, Spies JJ, Kleynhans R (2002) A preliminary study of interspecific hybrids in Lachenalia (Hyacinthaceae). Acta Horticulturae 570, 319-326 Duncan GD (1988) The Lachenalia handbook. Annals of Kirstenbosch Botanic Gardens 17, 1-71 Duncan GD (1992) The genus Lachenalia: Its distribution, conservation status and taxonomy. Acta Horticulturae 325, 843-845 Duncan GD (1996) Four new species and one new subspecies of Lachenalia (Hyacinthaceae) from arid areas of South Africa. Bothalia 26, 1-9 Duncan GD (1998) Notes on the genus Lachenalia. Herbertia 53, 40-48 Duncan GD (2001) Lachenalia elegans var. flava (Hyacinthaceae). Curtis’s Botanical Magazine 18, 18-22 Duncan GD (2002) Lachenalia. In: Manning J, Goldblatt P, Snijman D (Eds) The Color Encyclopedia of Cape Bulbs, Timber Press, Cambridge, UK, pp 251-264 Duncan GD (2005) Character variation and a cladistic analysis of the genus Lachenalia Jacq.f. ex Murray (Hyacinthaceae: Massonieae). MSc thesis, University of KwaZulu Natal, 338 pp Duncan GD, Edwards TJ, Mitchel A (2005) Character variation and a cladistic analysis of the genus Lachenalia Jacq. f. ex Murray (Hyacinthaceae). Acta Horticulturae 673, 113-120 Duncan GD, Edwards TJ (2006) Three new species of Lachenalia (Hyacinthaceae: Massonieae) from Western and Northern Cape, South Africa. Bothalia 36, 147-155 Duncan GD, Edwards TJ (2007) A new pyrophytic Lachenalia species (Massonieae) from Western Cape, South Africa. Bothalia 37, 31-34 Fluxus Technology Ltd. (2012) NETWORK 4.6.1.0. Shareware software Available online: fluxus-engineering.com/sharenet.htm Gouws JB (1965) Cytological studies in the genus Lachenalia Jacq. Annals of the University College of the Western Cape 2, 1-7 Hamatani S, Hashimoto K, Kondo K (1998) A comparison of somatic chromosome at metaphase in Lachenalia (Liliaceae). Chromosome Science 2, 21-25 Hamatani S, Ishida G, Hashimoto K, Kondo K (2004) A chromosome study of ten species of Lachenalia (Liliaceae). Chromosome Science 8, 55-61 Hamatani S, Kondo K, Kodaira E, Ogawa H (2007) Chromosome morphology of 12 species and one variety of Lachenalia and five species of closely related, allied genera (Liliaceae). Chromosome Botany 2, 79-86 Hamatani S, Masuda Y, Kondo K, Kodaira E, Ogawa H (2008) Molecular phylogenetic relationships among Lachenalia, Massonia and Polyxena (Liliaceae) on the basis of the internal transcribed spacer (ITS) region. Chromosome Botany 3, 65-72 Hamatani S, Tagashira N, Ishida G, Kondo K (2009) Chromosome relationships among 13 species and one variety of Lachenalia (Liliaceae) with the chromosome numbers of 2n=14 and 16 detected by FISH using 5S rDNA and 114 Floriculture and Ornamental Biotechnology 6 (Special Issue 1), 98-115 ©2012 Global Science Books Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011) MEGA 5: Molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolution 28, 2731-2739 Therman E (1956) Chromocentres in the mitosis of Liliaceae. Archiwum Societatis Zoologicae Botanicae Fennicae Vanamo 11, 189-193 Van Rooyen P, Spies JJ, Kleynhans R (2002) The species delimitation of Lachenalia unifolia and L. hirta. Acta Horticulturae 570, 395-401 Spies JJ, Spies P, Reinecke SMC, Minnaar A, Du Preez JL, Kleynhans R (2009) Lachenalia. In: Marhold K (Ed) IAPT/IOPB chromosome data. Taxon 58, 1288-1289 Spies JJ, Van Rooyen P, Kleynhans R (2002) The subgeneric delimitation of Lachenalia (Hyacinthaceae). Acta Horticulturae 570, 225-232 Spies P (2004) Phylogenetic relationships of the genus Lachenalia with other related liliaceous taxa. MSc, University of the Free State, 155 pp Stebbins GL (1950) Variation and Evolution in Plants, Columbia University Press, New York, 216 pp 115

RELATED PAPERS

RELATED TOPICS

Log In

Cytogenetic and Phylogenetic Review of the Genus Lachenalia

Cytogenetic and Phylogenetic Review of the Genus Lachenalia

Related Papers

RELATED PAPERS

RELATED TOPICS