Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Over the past decades the prevalence of pollen allergy has increased. This trend is expected to continue due to ongoing urbanization, climate change and increasing air pollution. In order to mitigate urban heat and improve air quality in the city, urban green space has been promoted. Urban green spaces have been associated with numerous health benefits. Contact with the natural environment strengthens the human microbiome and immune system. Green spaces promote an active lifestyle and provide mental health benefits. By improving the air quality, green spaces can also contribute to better respiratory health. However, urban vegetation is also a source of airborne pollen.The literature provides contradicting results regarding the effects of green space on the health of pollen allergy patients. These contradictions stem from the complex interactions between environmental factors, as well as from a heterogeneity in approaches when it comes to exposure studies. To properly study the health effects from exposure to green space on the health of people with pollen allergy we need detailed spatial data on green space and the allergenic trees within these greenspaces. In addition, we need insights in the effects of local vegetation on local pollen composition. Until now, the majority of exposure studies relied on exposure within various buffers around the residence. Nevertheless, a significant share of exposure takes place outside the residential area and pollen and air pollutant concentrations vary in time and space. In order to define exposure more realistically there is need for a method that can account for the spatiotemporal aspect of personal exposure.Nevertheless, the first step towards an exposure analysis is the generation of a distribution model of potential allergenic trees. For Flanders a database with validated observations of vascular plants at a 1 km resolution is available (Florabank). The observations can be combined with environmental covariates (soil characteristics, land use, habitat type) in a species distribution model. We modelled habitat suitability for 13 wind-pollinated tree genera. Genus-specific thresholds were used to obtain presence-absence maps. By summing the 13 presence-absence maps we obtained a tree diversity map at genus-level. We find that summing binary maps does not result in an overestimation of diversity when the study area is urbanized (Flanders) and the spatial resolution is coarse (1 km). The obtained diversity map can be used to determine exposure to biodiversity in health studies.In a second step, we studied the effects of local vegetation on the pollen composition. Although pollen can travel long distances, local vegetation contributes to local peak concentrations of pollen. Standardized pollen monitoring takes place at roof top level and measures a background level of pollen. Possible local peaks remain undetected, yet contribute to human pollen exposure. By mounting passive samplers at 2 m above the ground in 13 locations in Flanders we aimed to measure local pollen compositions during the tree pollen season (February-May) of 2017. The passive samplers successfully measured local pollen compositions characterized by 12 taxa. We used Non-metric Multidimensional Scaling (NMDS) to characterize each sampling site by its pollen composition. Then we use an indirect gradient analysis to associate the pollen composition with the land cover within a 20, 200, 500, 1000, 2000 and 5000 m radius around the sampling site. We found an urban-rural and wet-dry gradient associated with the first and second NMDS axis. Most of the associations were found for the land cover in a 1000-5000 m radius. This local scale effect is of importance for urban green management: urban forests at the edge of the city contribute to the pollen concentration within the city. In addition, we understand that environmental health studies require sufficiently large exposure radii.Thus, in the residential exposure study we determined garden cover, grassland cover and forest cover within a 1, 2 and 5 km radius around the residence of 157 adults with a tree pollen allergy. The density of allergenic trees (alder, hazel and birch) was derived from the regional forest inventories. We used a generalized linear models with a Poisson distribution to associate residential exposure to mental well-being (standardized questionnaires) and respiratory health (average symptom severity reported in a smartphone application). All green space types were protective for mental well-being, yet a risk factor for symptom severity. The density of allergenic trees in forests was a risk factor for mental well-being as well as symptom severity. We found that green space effects on health became smaller as the exposure radius increased.In an attempt to determine personal exposure more realistically, we used GPS data gathered by a smartphone application. We compared exposure between case-days with severe allergy symptoms and control-days without symptoms. We determined exposure by extracting green space cover, allergenic tree density, birch pollen levels and pollutant levels at the GPS point locations. For each day we could determine the average exposure taking into account spatiotemporal variability. Grassland cover, forest cover, alder density and hazel density protected against severe allergy symptoms. Birch density and birch pollen as well as pollutants (nitrogen dioxide (NO2), ozone (O3), and particulate matter < 10 µm (PM10)) were risk factors for severe allergy symptoms. For GPS tracks that were entirely within Flanders, we could calculate exposure to the tree diversity model at genus level. We found no associations between severe allergy symptoms and alpha-diversity of tree genera.This manuscript shows that adults with a tree pollen allergy experience mental and respiratory health benefits from exposure to green space, given that density of allergenic trees (birch) is low. GPS tracking allowed for a more realistic approximation of personal exposure. Although we did not identify a diversity effect, we do promote biodiverse urban green spaces in order to prevent domination of allergenic vegetation.