Genet Resour Crop Evol (2010) 57:501–514 DOI 10.1007/s10722-009-9486-3 RESEARCH ARTICLE Are Lotus creticus and Lotus cytisoides (Leguminosae) closely related species? Evidence from nuclear ribosomal ITS sequence data Graeme Sandral • Galina V. Degtjareva • Tatiana E. Kramina • Dmitry D. Sokoloff • Tahir H. Samigullin • Steve Hughes • Carmen M. Valiejo-Roman Received: 28 April 2009 / Accepted: 7 September 2009 / Published online: 4 October 2009 Ó Springer Science+Business Media B.V. 2009 Abstract Lotus creticus is a potentially important Recent morphological studies suggested that L. cret- perennial legume for soil management in the Mediter- icus and L. cytisoides are taxonomically distinct at ranean climate. This plant is in focus of experimental specific level and not even closely related to each other. research in different countries. The so-called Lotus Two molecular phylogenetic studies published so far creticus group is taxonomically problematic. Although and based on one accession of L. creticus each gave some authors consider all members of the group as a conflicting results on relationships of this species. In single variable species, others segregate several spe- this paper, sequences of nuclear ribosomal ITS (nrITS) cies distinct from L. creticus, especially L. cytisoides. are produced from multiple accessions of both L. creticus and L. cytisoides. Taxonomic identity of each G. Sandral accession has been tested using morphology. The new NSW Department of Primary Industries, Wagga Wagga nrITS sequences are inserted in a data matrix contain- Agricultural Research Institute, Private Mail Bag, Pine ing a representative set of Lotus species covering all the Gully Road, Wagga Wagga, NSW 2650, Australia geographical range and all major taxonomic groups of G. Sandral Lotus. Phylogenetic analysis of nrITS sequence data Faculty of Natural and Agricultural Science, University of showed that L. creticus is related to L. assakensis and Western Australia, Crawley, WA, Australia other species of the section Pedrosia while L. cytiso- ides is related to L. longisiliquosus and other species of G. V. Degtjareva (&) Botanical Garden of Moscow University, Biological the section Lotea. These results are important for Faculty, Moscow State University, 119991 Moscow, establishing future germplasm collection, breeding and Russia selection programs of perennial Lotus species. e-mail: [email protected] T. E. Kramina Á D. D. Sokoloff Keywords Loteae Á Lotus creticus Á Department of Higher Plants, Biological Faculty, Lotus cytisoides Á Mediterranean Á Molecular Moscow State University, 119991 Moscow, Russia phylogeny Á nrITS Á Perennial legumes Á Taxonomy T. H. Samigullin Á C. M. Valiejo-Roman Department of Evolutionary Biochemistry, A.N. Belozersky Institute of Physicochemical Biology, Moscow State University, 119991 Moscow, Russia Introduction S. Hughes South Australian Research and Development Institute, The genus Lotus L. (Leguminosae Juss.: Loteae DC.) GPO Box 397, Adelaide, SA 5001, Australia contains about 120–130 species of perennials and 123 502 Genet Resour Crop Evol (2010) 57:501–514 annuals native to Eurasia, Africa, Australia and report on nodulating bacteria in wild L. creticus adjacent islands, with the main centre of diversity in from Tunisia (Zakhia et al. 2004) might belong to the Mediterranean region (reviewed in Degtjareva L. cytisoides. et al. 2006). Several of these species are of significant The question whether L. cytisoides is distinct from agricultural importance including the perennials L. creticus at a specific level or should it be accepted L. corniculatus L., L. pedunculatus Cav. s.l. and as a subspecies was mainly of pure taxonomic interest L. glaber Mill. as well as the annual (to biennial or when the Lotus creticus group was considered as a perennial—Kramina 2006) L. subbiflorus Lag. The natural unit of perennial species within a section genus also includes the model legume L. japonicus Lotea (Medik.) DC. of the genus Lotus (e.g., Ball and and ornamentals such as L. berthelotii Masf. and Chrtkova´-Zˇ ertova´ 1968). Things became more com- L. maculatus Breitf. Another potentially important plicated when Kramina and Sokoloff (1999) sug- perennial Lotus is L. creticus L. because of its capacity gested that L. creticus and L. cytisoides may not be to stabilise sand dunes (Tsuriell 1974), provide forage even closely related to each other. They argued that for livestock (Le Houerou 1979) and withstand low L. cytisoides, together with L. polyphyllus, L. longi- humidity and increased salinity (Sanchez-Blanco et al. siliquosus and some other species belong to the 1998; Morales et al. 2000; Save´ et al. 2000; Franco section Lotea, while L. creticus belongs to the section et al. 2002; Vignolio et al. 2002, 2005; Ban˜on et al. (or subgenus) Pedrosia (Lowe) R.P. Murray. The 2004; Zakhia et al. 2004; Rejili et al. 2007). section Pedrosia includes about 30 species occurring The so-called Lotus creticus group (Heyn and mostly in Canary Islands, Madeira, Cape Verde, the Herrnstadt 1968) is taxonomically problematic Azores, and Morocco (Sandral et al. 2006). Morpho- because of technical difficulties that sometimes occur logically, the section (or subgenus) Pedrosia differs in the identification of its species. Some authors from the section Lotea in the presence of a distinct accepted a single variable species Lotus creticus s.l. tooth on ventral side of a style (e.g., Monod 1980; within this group (e.g. Quezel and Santa 1962), while Mader and Podlech 1989; Kramina and Sokoloff others recognized more than one species. A detailed 1999; Sandral et al. 2006). Kramina and Sokoloff taxonomic revision of the Lotus creticus group, based (1999) found a very short and often scarcely recog- on morphology, was published by Heyn and Herrns- nizable tooth on style in L. creticus. Since they tadt (1968). These authors recognized four different considered presence of a tooth quite an important mainly Mediterranean species, namely L. polyphyllus feature, Kramina and Sokoloff (1999) concluded that Clarke, L. creticus L. s.str. (= Lotus commutatus L. creticus should be included in the section Pedrosia Guss.), L. cytisoides L. and L. collinus (Boiss.) Heldr. rather than in the section Lotea. This taxonomic re- Currently, individuality of these four species is arrangement extends considerably the geographical accepted in most taxonomic publications with a range of the section Pedrosia, because L. creticus, change that the name L. collinus is corrected to though occurring in the Azores and Morocco is L. longisiliquosus de Roem. (Greuter et al. 1989; distributed in various Mediterranean countries up to Kramina and Sokoloff 1999; Valde´s 2000). The wide Israel in the eastern Mediterranean, i.e., in the regions taxonomic concept is still accepted in some experi- where other members of the section Pedrosia are mental studies, where L. cytisoides is treated as a totally absent. Valde´s(2000), like Kramina and subspecies of L. creticus (e.g., Save´ et al. 2000;Franco Sokoloff (1999), illustrated presence of a tooth on et al. 2002). Rejili et al. (2007) compared effect of style of L. creticus. Nevertheless, he placed the NaCl on the growth and the ionic balance K?/Na? of species near L. cytisoides within the section Lotea. two populations of Lotus creticus in Tunisia. However, Sandral et al. (2006), in the context of a detailed L. creticus s.str. is very rare in this part of Africa and comparative morphological study of all species of Sandral et al. (2006) were unable to trace any the section Pedrosia, agreed with placement of herbarium specimen from Tunisia. Since L. cytisoides L. creticus in this section. Kramina and Sokoloff is common in Tunisia and many herbarium collections (1999) and Sandral et al. (2006) noted close mor- of this species are misidentified as ‘‘L. creticus’’ we phological similarity between L. creticus and a local could hypothesize that Rejili et al. (2007) studied endemic of SW Morocco, L. pseudocreticus Maire, L. cytisoides rather than L. creticus s.str. Also, the Weiller et Wilczek. The main distinction between the 123 Genet Resour Crop Evol (2010) 57:501–514 503 two species is the presence of a well developed style specimen has been established using morphological tooth in L. pseudocreticus, while in most other characters listed in Heyn and Herrnstadt (1968), Ball characters the two species are alike. Sandral et al. and Chrtkova´-Zˇ ertova´ (1968), Kramina and Sokoloff (2006) found that, morphologically, L. pseudocreti- (1999) and Valde´s(2000). Out of 30 accessions cus is intermediate between L. creticus and cultivated, 14 were identified as L. creticus and 15 as another Moroccan member of the section Pedrosia, L. cytisoides. One accession (PI287858), which is L. assakensis Brand. Since all three species occupy listed in the seed bank collection under L. creticus, has similar habitats in SW Morocco, Sandral et al. been morphologically identified in this study as L. (2006) suggested hybrid origin of L. pseudocreticus arenarius Brot. This accession has the same voucher (L. assakensis 9 L. creticus). details as the accession of L. creticus used by Allan Molecular phylogenetic data based on nuclear et al. (2003). DNA was extracted from each accession. ribosomal ITS sequences supported the idea that the Voucher herbarium specimens are pressed from the section Lotea is distinct from the section Pedrosia. same plant individuals from which DNA was Indeed, according to nrITS data, the section Lotea is extracted; they are deposited at Herbaria of Royal closer to the section (formerly genus) Tetragonolobus Botanic Gardens, Kew (K) and Moscow State Univer- (Scop.) Benth. et Hook. f. than to Pedrosia (Allan et al. sity (MW). In addition, DNA was extracted from one 2003, 2004; Degtjareva et al. 2006). However, place- herbarium specimen of L. creticus, one herbarium ment of L. creticus was conflicting in phylogenies of specimen of L. cytisoides, three herbarium specimens Allan et al. (2003, 2004) and Degtjareva et al. (2006). of L. pseudocreticus and two herbarium specimens of According to Allan et al.