Alpine regions are at serious risk of disappearing as warming temperatures drive species towards higher altitudes. Plant distributions are expected to drastically change in alpine areas; however, it is not clear what effect this will have on associated insect communities. The purpose of our study was to document the current plant-pollinator interactions in the highly restricted alpine zone of Kosciuszko National Park, Australia. We used two different methods to describe the plant-pollinator community: direct field observations and pollen DNA metabarcoding from the insects. During timed observation periods, two groups recorded each insect that visited the fertile plants of a flower along transects. One of the groups also collected observed insects for pollen metabarcoding. We sequenced two commonly used plant metabarcoding markers (ITS2 and trnL) from the pollen samples. We then built separate plant-pollinator networks from observations and metabarcoding, and compared community indices including species diversity, and network metrics including nestedness and connectance.
Results/Conclusions
The pollen metabarcoding results revealed over three times more interactions between plants and insects than the field observations. They also indicated that insects were interacting with more than twice the number of plant species we observed. Herbaceous plants, such as Celmisia (Asteraceae) and Craspedia (Asteraceae), are key genera in the Australian alpine community, visited by the highest number of insect species. Flies (Diptera) were the most abundant pollinators and had the most interactions with plants in both networks. Additionally, pollen metabarcoding detected that some native bees, such as Austrevyleaus (Halictidae), interact with significantly more plant species than previously thought. Our results suggest that incorporating pollen metabarcoding to pollinator surveys can provide greater resolution of plant-pollinator interactions, leading to better informed conservation management decisions.