Abstract:
Frullania Raddi is worldwide in distribution and is a large and complex liverwort genus with over 1400 validly published binomials. Over the last five decades, there have been abundant changes and modifications to the subgeneric classification of Frullania, in part due to the establishment of supposedly allied genera, and the contractions and expansions of generic,subgeneric and sectional concepts; consequently, this has led to a complex and confusing classification.This study revises the Frullania in New Zealand, which is represented by 28 species and 1 variety, including the new record of Frullania chevalieri (Schust.) Schust. This species represents a critical group of taxa that had previously (in 1972) been placed into an autonomous genus, Schusterella - a placement that has remained controversial ever since.This group of species were also revised along with the apparently allied species and formed the basis of a worldwide revision of Frullania subg. Microfrullania, where the species of Schusterella are aligned. In addition, these revisions, together with the study of subsidiary species, which included some 90 species, were examined in various levels of detail to pioneer a worldwide revision of the genus as a whole. This also provided an opportunity to clarify the identification and the presence of a suite of species for the Australasian region. During the course of the current investigation six taxa new to science are described, five from New Zealand and one from Tasmania, and a significant number of validly published binomials have been lectotypified or nomenclaturally realigned, and 10 species are concluded to be conspecific and placed into synonymy.The taxonomic conclusions of this study were based on a multi-faceted approach, including a wide range of data obtained from morphology, anatomy, histochemistry, chemistry, ecology,reproductive biology, and growth studies. A checklist of significant criteria is provided to assist in future monographic and regional studies. Despite the vast number of taxonomic publications on the genus, many critical morphological and anatomical features have constantly been neglected or overlooked. Salient morphological features include those associated with the initial branching appendages, cell wall anatomy, oil bodies, position of gynoecia, ornamentation of the spore surface, and capsule wall anatomy. The use of initial branching appendages had almost routinely been ignored or overlooked in previous Frullaniastudies, and these were assessed in detail for 100 taxa representing 9 subgenera. Speculation is also made regarding possible phylogenetic relationships between taxa based on characters associated with the initial branching appendages and other data sources. Anew technique is reported for examining anatomical features of the different layers of the capsule wall in hepatics, and the potential taxonomic value of capsule wall characters for some New Zealand Frullania species are also discussed. The possible chemical components present in the cell-wall of the thickenings of the capsule wall layers are also discussed.The volatile components of 24 taxa from New Zealand, Australia and South America were analysed by GC-MS. This almost doubled the previous number of Frullania species chemically examined. Previously, the genus had been chemically divided into six major types.The presence of triterpenoids in F. fugax, which is the first instance of such a large amount in liverworts, required the establishment of a seventh chemical type. Justification for an eighth chemical type was based on the identification of 2-alkanones in F. pentapleura, which is the first record in the genus.It is evident that ecological data may offer a valuable insight into possible taxonomic relationships between Frullania species. Also reported for the first time are 11 species of Frullania growing as epiphylls in New Zealand, New Caledonia, and Colombia. The highest diversity of epiphyllous Frullania species is in the floristic regions of New Zealand, New Caledonia, Macronesia, and Madagascar.It is predicted that the current nomenclature of Frullania is over inflated and that many currently recognised species will be found to be conspecific, especially as we increase our understanding of the biological diversity of the genus. In this context, species concepts as applied to the genus Frullania is also discussed.