Abstract
Abundance and diversity of airborne pollen are important to human health and biodiversity. The UK operational network collects airborne pollen from 8 flowering trees, grasses and three weeds using Hirst traps and microscopic identification from urban areas. Knowledge of total pollen diversity and differences between rural and urban zones is limited. We collect environmental DNA (eDNA) from air during summer and autumn over 3 years with mini cyclones from one urban and one rural site. Data are analysed using next generation sequencing and metabarcoding. We find the most common genus, Urtica (57%), is also identified by the national network. The grasses Lolium (10%), Agrostis (2%) and Holcus (1%) are in the national network grouped at family level, while Brassica (2%), Chenopodium (1%), Impatiens (2%), Plantago (4%) and Tilia (7%) are not part of the UK operational network. DNA from 138 genera was identified, where 2% of the sample could not be associated with specific genera. 40% of the sample was classified better using eDNA methods at the genus level, than by optical methods. We calculate Bray-Curtis dissimilarity for the rural and urban zones and find a systematic difference in biodiversity. Overall, this shows airborne DNA reveals more information than methods based on morphological differences. The results also suggest data from sites located in large urban areas will be less representative for less populated rural areas. This presents a dilemma in balancing a network and the associated costs delivering health relevant information to the most populated areas vs. a nation-wide approach.