Pollination ecology of Roscoea species in Nepal Himalaya

Abstract

Himalayan region endemic alpine genus Roscoea (Zingiberaceae), evolved in the Himalayas ~50 Ma subsequent to the tectonic collision between Indian and Eurasian plate, currently comprises two distinct clades (Himalayan and North-Indochinese clades) that diverged at ~23 (13-38) Ma synchronously with the third uplift of the Himalayas and lateral extrusion of Indochina. The genus Roscoea exhibits the suites of floral traits that would fit pollination by long tongued insects, but previous studies in North-Indochinese clades of Roscoea did not find any pollinators with long tongue that matches with the corolla tube of Roscoea. In this study, I hypothesized that selection by long tongued insect is one of the major driving forces for the evolution of long corolla tubed Roscoea species from relatively shorter corolla tubed lineage. Thus, I anticipated that some Himalayan Roscoea species at their evolutionary centre are effectively pollinated by long tongued insects. To test this hypothesis, I explored the reproductive biology of five Himalayan Roscoea species (Roscoea alpina, R. auriculata, R. capitata, R. purpurea and R. tumjensis) that represent diverse floral morphology and phenology together with broad altitudinal distribution range within Himalayan clades of Roscoea. To determine the natural breeding system of these five Roscoea species, I studied their floral biology and conducted six types of pollination experiments (natural, autonomous selfing, emasculated, hand-selfing, hand-outcrossing and supplemented pollination) in their natural populations for two consecutive years. Additionally, I observed the pollinators’ behaviour, visitation frequency and calculated the pollinator importance of all insects that visited the flowers of five Himalayan Roscoea species. Moreover, based on the preliminary evidence of a long tongued fly as a specialized pollinator of R. purpurea, I made observational and experimental studies of natural selection and coevolution between R. purpurea and a long tongued fly (Philoliche longirostris) across the landscape in Nepal Himalayas. All Himalayan Roscoea species studied here are self-compatible but exhibit pollinator dependent breeding system. Except R. alpina which achieves autonomous selfing by the gradual shrinkage of style, rest four Roscoea species lack autonomous selfing. R. purpurea is pollinated by a specialized pollinator, Philoliche longirostris and thus exhibits specialized pollination system and avoids non specialized pollinator by hiding nectar deep inside the corolla tube so that only a long tongued fly can access the secure nectar. While rest of the Roscoea species depend upon generalized pollinators such as bumblebees (Bombus flavescens and B. haemorrhoidalis), a moth (Macroglossum nycteris), butterflies and even a beetle in the absence of a specialized pollinator. To ensure the reproductive success in the absence of specialized pollinator, R. auriculata, R. capitata and R. tumjensis secrete more nectar that wells up to the top of corolla tube so that it is accessible to all types of pollinators. Autonomous selfing coupled with the passive pollination by a beetle, a pollinator of limited mobility, provides reproductive assurance in R. alpina, in the absence of a specialized pollinator. Floral traits of R. purpurea (inflorescence height and corolla tube length) showed significant correlation with female fitness (fruit set per flower and seed set per flower) and thus these results indicate that evolution of inflorescence height and corolla length in R. purpurea is under the influence of pollinator mediated directional selection. My result also indicates that corolla length of R. purpurea and tongue length of P. longirostris show significant correlation and well-matched mechanical fit across landscape, in a manner consistent with the Darwin’s hypothesis of reciprocal selection. Results of reciprocal selection experiment at two of my study sites demonstrate that corolla tube length of R. purpurea and tongue length of P. longirostris act as agent and target of selection upon each other. This study provides some novel information on pollination biology of Roscoea species by discovering the yet unknown effective pollinators. Presence of highly specialized pollinator in R. purpurea (one of the ancestral Roscoea species) may indicate that Roscoea species are originally pollinated by long tongued flies. Moreover, finding of species specific reciprocal selection between corolla length of R. purpurea and proboscis length of P. longirostris suggests that selection by long tongued fly is one of the major driving forces for the evolution of long corolla tubed Roscoea species in the Himalayas. Temporal mismatch between the seasonal prevalence of specialized pollinator and the blooming period of plant may have caused the evolution of generalized pollination system in R. alpina, R. auriculata, and R. tumjensis while why R. capitata evolved to generalized pollination system despite synchronized blooming period with the winged stage of long tongued fly, P. longirostris, is a question of further research. The findings of effective pollinators of Roscoea, including a long tongued fly, bumblebees, beetle and a moth in Nepal Himalayas, but their absence in Southwest China, may indicate that those effective pollinators of Roscoea (long tongued fly) have been extinct, or have switched their foraging preference (bumblebees and moths) in the Chinese Himalayas due to the environmental change associated with the Himalayan topographic movement. This finding contributes to explain how temporal and spatial mismatch between plants and their specialized pollinator causes the evolution of different mating system in closely related plant species. Moreover, this research for the first time discovered the African originated and/or confined unique plant-animal mutualism (long tongued-long tubed fly/flower mutualism) out of Africa, in a different continent and climatic region, the Himalayas. Thus, this finding enhances the biogeographic relationship between Africa and South Asia in the origin and diversification of biodiversity via Indian collision with Eurasian at ~50 Ma. Overall, this study, besides enhancing the current understanding of pollination biology of Roscoea, provides important insights for the evolution and existence of Roscoea in the Himalayas and opens up an opportunity for future research if Himalayan Roscoea species are switching from specialized to generalized pollination system.