Molecules and morphology suggest cryptic species diversity and an overall complex taxonomy of fish scale geckos, genus Geckolepis

The current classification of the Malagasy-Comoroan geckos of the genus Geckolepis recognizes three valid species and is based on morphological differences only. Species allocation of individuals is difficult, mainly because of the insufficiently known degree of variation and the frequent partial loss of the integument in preserved specimens. Here we study Geckolepis specimens from almost their entire known range. We combine molecular and morphological data to estimate species richness in the genus, and to assess the taxonomic validity of the morphological differences proposed. Analyses of mitochondrial (12S, ND4) and nuclear (RAG1, CMOS) gene sequences support three major clades in addition to the morphologically distinct G. polylepis (included for 12S only), and the presence of various divergent lineages within these clades, some occurring in sympatry. Among lineages, the external morphology seems to be comparatively conservative, and differences are faint in many cases. For each major lineage, we summarize the most diagnostic morphological characters that distinguish it from other lineages. The combined molecular and morphological data strongly indicate the presence of highly divergent lineages, three of which are tentatively referred to the names Geckolepis maculata, G. polylepis and G. typica, whereas three other major lineages are considered Confirmed Candidate Species. Among the remaining lineages, differentiation is shallower, and these may qualify for Deep Conspecific Lineages. In conclusion, our analysis provides evidence for an underestimation of species richness in the genus and a complex evolutionary history, not reflected by current Geckolepis species-level taxonomy.     

Interspecific variation in floral fragrances within the genus Narcissus (Amaryllidaceae)

To improve our understanding of the floral biology, pollination, and systematics of the genus Narcissus, a comparative study was made of flower volatiles from nine species native to southern Spain using headspace collection and GC-MS analysis. The species fell into three fragrance types based on the identity of their major volatiles. In all but one species the fragrances consisted mainly of monoterpene isoprenoids mixed with benzenoids: in six species trans-ß-ocimene occurred in high proportions, in two others it was lacking; the last species had a fragrance dominated by fatty acid derivatives, mixed with terpenoids. Two of the species showed marked intraspecific variation in many of their volatiles. When the volatile data matrix of all species was subjected to cluster analyses and the resulting phenetic trees compared with currently recognized taxonomic groups, there was no congruence at the subgeneric level. However, there was considerable agreement at the sectional level, although in most sections we studied only a single species. This apparent agreement was stronger when the volatiles were analyzed according to shared biosynthetic pathways rather than treated individually, pointing to the higher value of using biosynthetic pathways for uncovering and confirming phenetic, and probable evolutionary, relationships among species. In terms of possible selective pressures from pollinators in shaping fragrance chemistry, available information on the pollination of our species suggested an association between fragrance and types of pollinators. Two pollinator-fragrance groups were apparent: (1) species pollinated by insects that include butterflies and moths displayed fragrances containing volatiles typical of moth-pollinated flowers, most particularly indole combined with high amounts of esters, and (2) species visited exclusively by insects other than butterflies and moths, especially by bees and flies, had fragrances lacking this combination of volatiles. Narcissus assoanus was unusual among our species in having both fragrance chemotypes. Future pollination studies of Narcissus in the field are needed to test the reliability and predictability of the proposed fragrance-pollinator associations.