Caloplaca marina and C. rosei, Two Difficult Species in North America

Caloplaca rosei Hasse and C. marina (Wedd.) Zahlbr. both grow on seashore rocks and are widely distributed along the North American west coast. Caloplaca marina occurs in Europe as well, but not on the North American east coast. Due to their similarities to one another and the lack of correctly determined specimens in herbaria, the two species have been very much misunderstood in North America. Thepresent study, based on my own material and on herbarium material, was carried out to clarify their relationship and to establish whether or not the North American populations of C. marina are conspecific with those in Europe. A large set of morphological and anatomical characters was studied and analyzed using a computer-assisted principal components analysis. The results show that C. rosei and C. marina, even though closely related, are well separated by morphological traits such as continuity, areolation, and surface of the thallus, together with the shape, disk, and thalline margin of the apothecia. Within C. marina two subgroups could be recognized, corresponding well with their disjunct distribution on the two continents. The American population is designated C. marina subsp. americana Arup, differing from subsp. marina mainly in ecology but also having narrower spores and spore septum and a thicker hymenium. Caloplaca rosei Hasse and C. marina (Wedd.) Zahlbr. are two closely related lichen species within the large family Teloschistaceae. The two species are characterized by their yellow to orange, crustose, areolate thalli, often distinct hypothallus, and abundant lecanorine apothecia. Both species grow mainly in the supralittoral zone on seashore rocks and are part of the "orange belt" just above the black Verrucaria zone. Sometimes, however, they also occur in adjacent zones. Caloplaca rosei is rather common along most of the American west coast, but does not occur on the American east coast. Caloplaca marina is a more widespread species, occurring not only in western North America but also in Europe, from where it originally was described. It has also been reported several times from the American east coast (e.g., Degelius 1940; Lamb 1954; Taylor 1974) but all these records seem to be erroneous. The taxonomy of C. marina in Europe has been treated by Nordin (1972), but the relationship of C. rosei and C. marina has never been scrutinized. The two species have been misunderstood in North America for several reasons. In addition to being similar to one another in many respects, they have been confused with other species as well. The fact that they often grow together (Fig. 1), and even upon each other, may be part of the reason there are hardly any correctly determined specimens in herbaria. Most reports of one or the other species refer to a mixture of both (e.g., Benton et al. 1977; Noble 1982; Otto 1968; Ryan 1988). The purpose of this study was to clarify the taxonomic relationship between C. marina and C. rosei, and to obtain a good picture of their geographical distribution and ecology in North America. I also wanted to study geographical variation within the species and to find out if C. marina in Europe and North America are conspecific. The study is part of my doctoral thesis project on "The Littoral Species of Caloplaca in North America." MATERIALS AND METHODS The study is based primarily on materials collected on a field trip to North America in 1989. I spent eight weeks on the west coast where I collected from the Los Angeles area up to Vancouver Island in British Columbia, and three weeks on the east coast, from Massachusetts to Quebec. In addition to my own material I have studied herbarium material from the following institutions: ASU, Bratt (priv. herb.), CANL, COLO, FH, LAM, LD, LISU, MIN, S, SFSU, TUR, UBC, UPS, WTU, and WWB. The collections of C. marina from Europe included in the study were selected to give a good representation from throughout the European range of the species. A total of almost 230 specimens of both species are included in the study. My own collections are deposited at the Botanical museum in Lund (LD). The specimens were examined by interference and light microscopy. Sections of the apothecia were made both by hand and with a freezing microtome and embedded in water or in phenol-cottonblue. All measurements of anatomical features were carried out on sections mounted in water. 0007-2745/92/148-160$1.45/0 This content downloaded from 157.55.39.35 on Thu, 01 Sep 2016 05:37:25 UTC All use subject to http://about.jstor.org/terms 1992] ARUP: CALOPLACA MARINA AND C. ROSEI 149 The chemistry was examined by high performance thinlayer chromatography (HPTLC) as described by Arup et al. (1992), a somewhat different procedure than the standardized method developed by Culberson (1972, 1974). The plates used were glass HPTLC-Fertigplatten Merck no. 5629, silica-gel 60 F254, 0.2 mm thick. The silicon particles on these plates are smaller and of more uniform size than on the normal TLC plates, resulting in an ability to detect also low concentrations of the substances applied. The chamber used was a Camag Horizontal Developing Chamber, 10 x 10 cm. This method differs from the normal in that two fronts of the same solvent move toward each other, meeting in the middle on a horizontal plate, and the amount of substance applied on the plate is much smaller, ca. 3 01. The solvents used were: A, toluene: dioxan: acetic acid (18:6:0.8 ml); B, cyclo-hexane: methyl tert.-butyl ether : formic acid (13:10:2 ml); and C, toluene: acetic acid (20:3 ml). A large set of morphological and anatomical characters was studied, and 19 of these were scored on a total of 118 specimens (42 specimens of C. rosei, 39 American and 37 European specimens of C. marina) and analyzed using principal components analysis (PCA), run on a Macintosh SE 30 computer. A PCA simplifies multivariate data by assigning most of the variation to a smaller number of "principal components" (Dunn & Everitt 1982). All calculations were performed with SYSTAT (Wilkinson 1986). Only specimens for which all 19 traits could be scored were used in the analysis. The characters used in the PCA are listed in Tables 1 and 2 and discussed in the following section.