Data from: Butterfly community dynamics in a monoculture-dominated agricultural landscape
Data files
Mar 14, 2026 version files 22.75 KB
-
Env.csv
5.26 KB
-
Evenness.csv
5.69 KB
-
GAM.csv
5 KB
-
NMDS.csv
3.64 KB
-
README.md
3.16 KB
Abstract
Understanding biodiversity dynamics in agricultural landscapes is essential for promoting sustainable land use and conserving wildlife. We examined butterfly communities in a monoculture-dominated landscape, comparing remnant grassland sites surrounded by agricultural land (AGR), grasslands (GR), and forests (FOR). Data were collected from 30 sites across these three categories. We used non-metric multidimensional scaling (NMDS) to assess community composition, rank abundance curves for community structure, and Hill numbers for alpha diversity. Generalized additive models (GAMs) identified key drivers of rare species richness and the abundance of the most common species, Coenonympha pamphilus. Results showed a few dominant species, with most having low abundances. Species richness was similar across all site categories, but differences in evenness and species dominance highlighted variations in community structure. GR and FOR sites had higher numbers of both rare and abundant species, indicating a more balanced distribution than AGR sites. NMDS revealed significant differences in composition, especially between FOR and AGR. The GAM results showed that the combination of agricultural land and forest cover enhances rare species richness, while agricultural cover alone had a negative impact. Even for the most common species, C. pamphilus, high agricultural cover negatively impacts its population but shifts positively when combined with greater vegetation height in adjacent grasslands. These findings highlight the critical role of forest habitats within monoculture-dominated landscapes in enhancing butterfly community diversity and stability. The adaptability of butterfly species to agricultural landscapes is limited to a small number of species and only up to a certain extent.
https://doi.org/10.5061/dryad.bvq83bkk3
Description of the data and file structure
This repository contains the datasets used for all analyses conducted in the study “Butterfly community dynamics in a monoculture‐dominated agricultural landscape.”
Rarefaction Analysis
Rarefaction analysis based on Hill numbers was performed to compare species diversity across site categories. This analysis used the Evenness.csv data file.
Community Composition Analysis
Non-metric multidimensional scaling (NMDS) followed by PERMANOVA was used to analyze butterfly community composition across different site categories. This analysis used the NMDS.csv data file.
Generalized Additive Models (GAM)
To assess the relationship between butterfly species richness and environmental variables, we fitted a generalized additive model (GAM) with individual smooth terms for each environmental predictor: agricultural cover (AGR), grassland cover (GR), forest cover (FOR), vegetation height (H), and total flower cover (TFC). This analysis used the Env.csv data file.
The same dataset (Env.csv) was also used to extend the model by incorporating a tensor product smooth to account for the potential interaction between agricultural land cover and forest cover:
te(AGR, FOR).
A third GAM was conducted to explore the relationship between the abundance of the most common species in the dataset, Coenonympha pamphilus, and environmental variables. This analysis used the GAM.csv data file. The Env.csv file was also used for analyses of rare species richness.
Description of Data Files
Env.csv
Contains environmental variables and butterfly diversity metrics for each sampling site, including:
- SiteID – site identification code
- Site category – AGR (agricultural), GR (grassland), or FOR (forest)
- SR_common – species richness of common butterfly species
- SA_common – abundance of common butterfly species
- SR_rare – species richness of rare butterfly species
- SA_rare – abundance of rare butterfly species
- SR_total – total species richness
- SA_total – total species abundance
- TFC – total flower cover
- H – vegetation height (m)
- GR – grassland cover within a 1 km buffer
- FOR – forest cover within a 1 km buffer
- AGR – agricultural land cover within a 1 km buffer
- GR_2 – grassland cover within a 2 km buffer
- FOR_2 – forest cover within a 2 km buffer
- AGR_2 – agricultural land cover within a 2 km buffer
GAM.csv
Contains variables used for the GAM analyzing Coenonympha pamphilus abundance:
- Site ID
- Site category
- Abundance of Coenonympha pamphilus
- Land cover variables
- Transect-level environmental variables
NMDS.csv
Contains species abundance data across site categories used for NMDS and PERMANOVA analyses of butterfly community composition.
Evenness.csv
Contains species abundance data across sites used for rarefaction analysis and estimation of Hill numbers to compare diversity across site categories.
Data were collected during the implementation of the project Safeguarding European wild pollinators (Safeguard, 2021-2025). Study sites were selected to represent important pollinator habitats, i.e., open semi-natural habitats, recognized as habitats of conservation value. Butterflies were observed and collected at 30 sites in the northern part of Serbia, Vojvodina province in 2022. The sampling spanned three seasons (spring, summer, and autumn) and each site was visited three times.
Community structure and diversity patterns have been analyzed, community composition, and species-environment relationships. All data analyses were conducted in R 4.3.2 (R Core Team, 2020).