
Mycologia, 103(2), 2011, pp. 361–378. DOI: 10.3852/10-120 # 2011 by The Mycological Society of America, Lawrence, KS 66044-8897 Align or not to align? Resolving species complexes within the Caloplaca saxicola group as a case study Ester Gaya1 unambiguously aligned ITS sites or when the entire Department of Plant Biology (Botany Unit), Facultat de ITS1 and 2 regions were not aligned a priori and Biologia, Universitat de Barcelona, Av. Diagonal 645, included as an integral component of a Bayesian 08028 Barcelona, Spain analysis (BAli-Phy). The C. arnoldii subgroup includes Benjamin D. Redelings C. arnoldii, comprising four subspecies, and the C. National Evolutionary Synthesis Center (NESCent), saxicola subgroup encompasses seven species. Con- Durham, North Carolina 27705 trary to the C. saxicola subgroup, monophyly of taxa included within the C. arnoldii subgroup and their Pere Navarro-Rosine´s relationships could not be resolved with combined Xavier Llimona ITS and morphological data. Unequivocal morpho- Department of Plant Biology (Botany Unit), Facultat de Biologia, Universitat de Barcelona, Av. Diagonal 645, logical synapomorphies for all species except C. 08028 Barcelona, Spain arnoldii and C. pusilla are recognized and presented. Key words: Caloplaca, Gasparrinia, lichens, mo- Miquel De Ca´ceres lecular phylogeny, morphological phylogeny, mor- Biodiversity and Landscape Ecology Lab, Centre phospecies Tecnolo`gic Forestal de Catalunya, Ctra. St. Llorenc¸de Morunys km 2, 25280 Solsona, Spain Franc¸ois Lutzoni INTRODUCTION Department of Biology, Duke University, Durham, North Carolina 27708-0338 The widespread lichen-forming genus Caloplaca Th. Fr. is the largest of the Teloschistaceae, with about 500 known species (Kirk et al. 2008). It includes Abstract: The Caloplaca saxicola group is the main mostly crustose species that often have anthraqui- group of saxicolous, lobed-effigurate species within nones in both thalli and apothecia and colorless genus Caloplaca (Teloschistaceae, lichen-forming polarilocular ascospores. The delimitation of Calo- Ascomycota). A recent monographic revision by the placa and its relationship to other genera of Tel- first author detected a wide range of morphological oschistaceae is controversial and has been the variation. To confront the phenotypically based objective of several phylogenetic studies (Arup and Grube 1999; Gaya et al. 2003, 2008; Søchting and circumscription of these taxa and to resolve their Lutzoni 2003). Furthermore, problems of delimita- relationships morphological and ITS rDNA data were tion are not restricted to genus but also occur among obtained for 56 individuals representing eight Calo- species groups and taxa within Caloplaca. Specifically, placa species belonging to the C. saxicola group. We the lobed Caloplaca species constitute a difficult tested the monophyly of these eight morphospecies group encompassing extraordinary phenotypic varia- by performing maximum parsimony, maximum like- tion. Clauzade and Roux (1985) referred to these lihood and two different types of Bayesian analyses lobed Caloplaca as subgenus Gasparrinia (Torn.) Th. (with and without a priori alignments). Restricting Fr. and defined within it five groups: C. aurantia, C. phylogenetic analyses to unambiguously aligned aurea, C. carphinea, C. persica and C. saxicola. The C. portions of ITS was sufficient to resolve, with high saxicola group, which represents the core of the bootstrap support, five of the eight previously saxicolous lobed-effigurate Caloplaca species, is recognized species within the C. saxicola group. among the most taxonomically controversial. Howev- However, phylogenetic resolution of all or most of er not all authors have followed Clauzade’s and the eight species currently included as two distinct Roux’s classification and often considered the lobed subgroups within the C. saxicola group was possible Caloplaca species as a whole, without segregating only by combining morphological characters and them into groups (e.g. Verseghy 1970, 1971, 1972; signal from ambiguously aligned regions with the Nordin 1972; Ka¨rnefelt 1989; Wetmore and Ka¨rnefelt 1998; see also Gaya et al. 2008). Submitted 21 Apr 2010; accepted for publication 27 Aug 2010. Adding to the taxonomic complexity, some of the 1 Corresponding author. Current address: Department of Biology, Duke University, Durham, North Carolina 27708-0338. E-mail: taxa treated here often have been included within a [email protected] broader concept of a single taxon, C. saxicola 361 362 MYCOLOGIA (Hoffm.) Nordin. Thus Poelt (1954) distinguished and molecular data for members of the C. saxicola what he called C. arnoldii s.l., including C. arnoldii- group and to assess the value of morphological confusa Gaya & Nav.-Ros., from C. murorum (Hoffm.) characters in delimiting species within species com- Th. Fr. (5 C. saxicola), but included within C. plexes. We conducted a detailed reassessment of murorum what is currently recognized as C. arnoldii morphological and anatomical features and se- ssp. obliterata (Pers.) Gaya and C. rouxii Gaya, Nav.- quenced the nuclear ribosomal internal transcribed Ros. & Llimona. However, he took into account the spacer (ITS) region of 56 specimens representing 11 variability of C. murorum and informally used several taxa within the C. saxicola group, which were the names to refer to the diversity of morphologies subject of a taxonomic revision by the first author associated with different ecologies and suggested the (Gaya 2009). Because a high rate of indels is integral to possibility of detecting truly genetically based forms the evolution of the ITS, resulting in the fastest (independent from environmental factors) in future evolving sites being not alignable, much of the signal studies (Gaya 2009). Nordin (1972) later synony- from ITS1 and 2 must be excluded from conventional mized C. murorum f. miniata (Hoffm.) Ozenda & phylogenetic analyses, even when the phylogenetic Clauzade (currently C. rouxii), some of the morpho- signal is not saturated. Yet to resolve species complexes types now included within C. arnoldii ssp. obliterata it is most often these fast-evolving sites that are the (e.g. ‘‘obliterata,’’ ‘‘obliterascens’’) and C. pusilla (A. most phylogenetically informative. It is possible to Massal.) Zahlbr. under C. saxicola. Apart from C. capture this signal by adopting a coding procedure decipiens (Arnold) Blomb. & Forssell, Wetmore and that does not require multiple sequence alignments or Ka¨rnefelt (1998) only recognized C. saxicola and by using methods in which the alignment is one of the described three morphotypes within it. The conflicts variables optimized during the phylogenetic search. between these taxonomic treatments are understand- Hence a second objective of this study was to compare able because, according to Gaya (2009), a wide range the abilities of two methods to resolve species of morphological variation at both the intra- and complexes with ITS data. A novel Bayesian method interspecific levels is present in this group, causing that does not require sequences to be aligned or coded major problems for species delimitation and identifi- was compared with a second approach in which the cation, especially for taxa within the C. arnoldii signal from nonalignable regions is coded and added complex (part of the C. saxicola group), where to the matrix of alignable sites before conducting an intermediate phenotypes are frequently observed unequally maximum parsimony analysis. within a population or even within a single thallus. Phylogenetic studies on the Teloschistaceae (Arup MATERIALS AND METHODS and Grube 1999; Gaya et al. 2003, 2008; Søchting and Lutzoni 2003) have confirmed that genus Caloplaca Taxon sampling.—This study was carried out on collections and subgenus Gasparrinia are both polyphyletic. from the first author and on material from selected herbaria Furthermore, the C. saxicola group sensu Clauzade (BCN, E, GZU, LD, MARSSJ, MIN, VAB-LICH and the U. and Roux (1985) includes monophyletic groups Arup personal herbarium). For the phylogenetic analysis resulting from several independent origins (Gaya et based on morphological data we selected a total of 56 specimens that were chosen to represent as much of the al. 2003, 2008). Gaya (2009) restricted the C. saxicola range of morphological variation of each taxon as possible. group to the largest clade, which includes C. arnoldii, We attempted to keep the number of specimens per taxon C. arnoldiiconfusa, C. biatorina (A. Massal.) J. Steiner, proportional to the morphological variation of each taxon, C. decipiens, C. pusilla, C. rouxii, C. saxicola s. str. and and when possible we included fertile specimens. Never- C. schistidii (Anzi) Zahlbr., and defines the focal theless, due to the scarcity of available material, some taxa taxonomic group of this study. All previous taxonom- are represented only by single specimens. Thus we included ic studies of the C. saxicola group were based six specimens of Caloplaca arnoldii ssp. arnoldii,one exclusively on phenotypic characters. In general, only specimen of C. arnoldii ssp. clauzadeana Gaya, one a few molecular investigations have focused on specimen of C. arnoldii ssp. nana Gaya, 11 specimens of species complexes within Caloplaca; Arup (2006) C. arnoldii ssp. obliterata, nine specimens of C. biatorina, and Vondra´k et al. (2009) focused on the C. citrina two specimens of C. decipiens, six specimens of C. complex, Arup (2009) on the C. holocarpa complex, arnoldiiconfusa, six specimens
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages18 Page
-
File Size-