Urbanisation is associated with reduced Nosema sp. infection, higher colony strength and higher richness of foraged pollen in honeybees

Abstract

Bees are vital pollinators, but are faced with numerous threats that include loss of floral resources and emerging parasites amongst others. Urbanisation is a rapidly expanding driver of land-use change that may interact with these two major threats to bees. Here we investigated effects of urbanisation on food store quality and colony health in honeybees (Apis mellifera) by sampling 51 hives in four different land-use categories: urban, suburban, rural open and rural wooded during two seasons (spring and autumn). We found positive effects of urban land use on colony strength and richness of stored pollen morphotypes, alongside lower late-season Nosema sp. infection in urban and suburban colonies. Our results reveal that honeybees exhibit lower colony performance in strength in rural areas, adding to the growing evidence that modern agricultural landscapes can constitute poor habitat for insect pollinators.

Keywords: Nosema spp.; Varroa destructor; colony strength; pollen foraging; urbanisation.

Figure 1.

a Location of 51 apiary sites in SE England. The Greater London region is identified by dark grey shading, and inset circles show GIS land-use mapping for a representative site from each of the four land-use types (urban, suburban, rural open and rural wooded). Land-use map colours indicate 29 land classes; in summary, grey colours represent urban land classes (darker with increasing building density); green colours represent vegetated land including arable, pasture, woodland and urban parkland; orange colours represent sport and recreational open spaces; and blue colours represent water. b Cluster analysis (Ward’s method) of land-use types of 51 sites located in SE England; branch terminals show land-use maps of individual sites generated from GIS classification.

Figure 2.

Means and standard errors for raw data for a colony strength, b Varroa mite count and c Nosema spore count across four land-use types in two periods, spring and autumn. Raw data are displayed as green (spring) and orange (autumn) points. Important significant pairwise differences with large effect sizes (see Table S3) are highlighted in grey boxes and variables included in the final model or model set are inset. LU land category.

Figure 3.

Means and standard errors for a proportion woody pollen, b pollen morphotype richness and c pollen Shannon diversity across four land-use types in two periods, spring and autumn. Raw data are displayed as green (spring) and orange (autumn) points. Important significant pairwise differences with large effect sizes (see Table S3) are highlighted in grey boxes and variables included in the final model or model set are inset.

Figure 4.

Contribution of different pollen types to pollen samples in each of 51 sites across SE England collected in a spring and b autumn. c Five most important pollen types (greatest contribution) in each of four land-use types in spring and autumn. d Colour legend for five most important pollen types in each land-use type. Additional pollen types and their colours are shown in the supplementary material (Online Resource 1).

Figure 5.

Principal coordinate analysis (PCoA) plots to visualise differences in pollen species composition in four land-use types in a spring and b autumn using Bray-Curtis distances (Beals 1984). Ellipses represent one standard deviation from the centroid and p values from PERMANOVA analyses are inset. Points represent individual sites: urban (black diamonds), suburban (red triangles), rural wooded (green diamonds), rural open (blue squares).