Data from: Pollinator community composition and pollen resource use in calcareous grasslands under different landscape contexts across Europe
Data files
Jun 26, 2026 version files 260.24 KB
-
Landscape_variables_JAE-2025-01022.xlsx
9.49 KB
-
Matrix_connected_BE_JAE-2025-01022.csv
6.46 KB
-
Matrix_connected_CZ_JAE-2025-01022.csv
54.41 KB
-
Matrix_connected_DE_JAE-2025-01022.csv
13.23 KB
-
Matrix_connected_EE_JAE-2025-01022.csv
28.24 KB
-
Matrix_connected_SE_JAE-2025-01022.csv
13.17 KB
-
Matrix_fragmented_BE_JAE-2025-01022.csv
11.84 KB
-
Matrix_fragmented_CZ_JAE-2025-01022.csv
39.65 KB
-
Matrix_fragmented_DE_JAE-2025-01022.csv
8.38 KB
-
Matrix_fragmented_EE_JAE-2025-01022.csv
52.74 KB
-
Matrix_fragmented_SE_JAE-2025-01022.csv
18.28 KB
-
README.md
4.36 KB
Abstract
Plants and pollinators interact in complex ways to facilitate plant reproduction and increase insect fitness. While some plant and insect species show high levels of specialization, others engage in more generalized interaction patterns. However, interactions between plants and pollinators are dynamic and can change in response to altered environmental conditions.
Land-use change is one of the main factors that may affect the interactions between plants and insects, as the associated shifts in land cover, habitat availability, and habitat connectivity are expected not only to directly impact pollinator and plant diversity, but also to influence the foraging behavior and resource use of pollinators.
In this study, we compared pollinator community composition, pollinator resource use and the structure of the network of plant-pollinator interactions among calcareous grasslands embedded in agricultural landscapes differing in grassland amount, connectivity and surrounding land use across five European countries. Pollinator surveys were conducted to assess variation in their communities, while pollen metabarcoding was conducted to assess the interactions between plants and pollinators. Network analyses were used to investigate whether the overall structure of the network of interactions differed between landscapes.
Our results show substantial differences in pollinator communities and pollen composition between grasslands, leading to strong interaction turnover. However, despite strong species turnover, dominant families such as Asteraceae (plants) and Syrphidae (insects) remained consistently common across sites, suggesting considerable functional redundancy. In addition, pollinators consistently foraged on plant species in the wider landscape, regardless of variations in landscape composition and habitat connectivity. This likely contributed to the limited differences in overall network structure, with similar levels of connectance, nestedness and modularity across landscapes.
Synthesis. Overall, these findings indicate that pronounced differences in pollinator communities and opportunistic foraging strategies resulted into large interaction turnover, but these shifts did not translate into major changes in overall network structure, likely due to functional redundancy. Future long-term monitoring studies are needed to assess whether these network properties remain stable through time under continued landscape change.
Data description
Plant-pollinator interaction matrices and landscape variables from calcareous grasslands sampled across five European countries (Estonia, Sweden, the Czech Republic, Belgium, and Germany). In each country, one grassland in two contrasting landscapes (referred to as Connected and Fragmented) was sampled in summer 2023. Pollinators were collected in the field and identified morphologically. Plants visited by each pollinator were identified through DNA metabarcoding of pollen carried on their bodies.
The dataset includes ten plant-pollinator interaction matrices (one per country × landscape), each providing presence/absence data on detected interactions between plant and pollinator species. A separate file contains landscape variables quantifying the connectivity and surrounding habitat composition of each sampled grassland.
These data enable comparison of plant-pollinator interactions and network structure between grasslands differing in grassland amount, connectivity, and surrounding land use.
Corresponding author
Olivia Bernhardsson
Plant Conservation and Population Biology Group
Katholieke Universiteit Leuven
Kasteelpark Arenberg 31, 3001 Heverlee, Belgium
olivia.bernhardsson@kuleuven.be
File and data information
Plant-pollinator interaction matrices
The datasets contain binary plant-pollinator interaction matrices for grasslands located in connected and fragmented landscapes in five European countries.
File list:
- Matrix_connected_EE_JAE-2025-01022.csv- Binary plant-pollinator interaction matrix from the connected landscape in Estonia
- Matrix_fragmented_EE_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the fragmented landscape in Estonia
- Matrix_connected_SE_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the connected landscape in Sweden
- Matrix_fragmented_SE_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the fragmented landscape in Sweden
- Matrix_connected_CZ_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the connected landscape in the Czech Republic
- Matrix_fragmented_CZ_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the fragmented landscape in the Czech Republic
- Matrix_connected_BE_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the connected landscape in Belgium
- Matrix_fragmented_BE_JAE-2025-01022.csv - Binary plant-pollinator matrix from the fragmented landscape in Belgium
- Matrix_connected_DE_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the connected landscape in Germany
- Matrix_fragmented_DE_JAE-2025-01022.csv - Binary plant-pollinator interaction matrix from the fragmented landscape in Germany
-
Matrix structure:
Column names: insect species -
Row names: plant species
-
Cell values: 1 = interaction, 0 = no interaction
Landscape variables
The dataset contains landscape metrics quantified within a 1 km radius circular buffer centered on each focal grassland site
File:
Landscape_variables_JAE-2025-01022.xlsx - quantified landscape measures in a 1 km circular buffer centered on each focal grassland
- Country: The country in which the focal grassland is located
- Landscape: The landscape category of the focal grassland (Connected or Fragmented)
- Year: Year in which the landscape variables were measured
- Hanski's connectivity index: Connectivity of each grassland calculated using Hanski’s connectivity index (α = 0.5), based on the size and distance of neighboring grassland fragments
- Forest (m): Total amount of forest within the 1 km buffer
- Arable (m): Total amount of arable land within the 1 km buffer
- Water/Wetland (m): Total amount of water bodies within the 1 km buffer
- Human (m): Total amount of human settlement within the 1 km buffer
- Semi-natural grassland (m): Total amount of semi-natural grassland within the 1 km buffer
- Midfield islets (m): Total amount of midfield isletswithin the 1 km buffer
- Improved grassland (m): Total amount of improved semi-natural grassland within the 1 km buffer
