Solar parks provide heterogeneous habitats for winter-active ground-dwelling predatory arthropods
Abstract
Winter-active ground-dwelling spiders and ground beetles mainly inhabit non-crop habitats (e.g., grasslands and forests) with complex-structured vegetation and stable microclimates. In spring, they migrate from non-crop habitats into crop fields, contributing to pest control. Nowadays, there is an increasing number of solar parks in agricultural landscapes. However, the role of solar parks as habitats for winter-active ground-dwelling spiders and ground beetles, and thus their potential contribution to pest control, has not been studied yet. We investigated how different habitat types (i.e., forest, grassland, habitat between and under solar panels, and abandoned farmland) are associated with variation in microclimatic conditions (i.e., air temperature and humidity) and with the diversity of ground-dwelling spiders and ground beetles across 50 sites in western Hungary. Using pitfall traps, we collected 957 ground-dwelling spiders belonging to 69 species, and 327 ground beetles belonging to 40 species. We recorded microclimatic conditions using data loggers simultaneously with arthropod sampling. We showed that patterns in arthropod assemblages likely reflect differences in microclimatic conditions across habitat types. Solar parks hosted species of different habitats (e.g., forest, grassland, wetland) with a relatively strong preference for humidity. Solar parks also supported a high abundance of agrobiont ground-dwelling spiders (i.e., species dominant in agroecosystems). In contrast, grasslands and abandoned farmlands exhibited the most extreme microclimatic conditions, supporting mainly dry-habitat species. Our results demonstrate for the first time that solar parks can serve heterogeneous habitats for diverse assemblages of winter-active ground-dwelling spiders and ground beetles and may positively contribute to biocontrol and biodiversity conservation.
Dataset DOI: 10.5061/dryad.rbnzs7hr8
Description of the data and file structure
This dataset provides the data used to evaluate the association between different habitat types (i.e., forest, grassland, habitats between and under solar panels, and abandoned farmland) and variations in microclimatic conditions (i.e., air temperature and humidity), as well as the diversity of ground-dwelling spiders and ground beetles.
Fifty sites in western Hungary were selected. The study sites were located around five settlements: Bicske, Csurgó, Nagyatád, Nyirád, and Óhíd. Near each settlement, we established two independent sampling sites in each of the following habitat types: forest, grassland, and abandoned farmland. In solar parks, two sampling sites were established within each park, placed 100 meters apart. Ground-dwelling predatory arthropods were collected using pitfall traps. Traps were open for two 14-day periods: at the beginning of November (2023) and at the end of February (2024). The two pitfall traps from each sampling site were pooled across both sampling periods, resulting in 50 samples (5 settlements × 5 habitats × 2 sites).
Microclimatic conditions were recorded using data loggers (Optin TH4-32U) every 20 minutes for 11 days (from February 23 to March 4, during the second arthropod sampling period). Near each settlement, one data logger was placed in each habitat type, located at one of the two sampling sites for ground-dwelling predatory arthropods.
Files and variables
File: data.xlsx
Description: The dataset contains five sheets: (1) microclimate, (2) spiders, (3) spiders_compo, (4) beetles, and (5) beetles_compo. Missing values are marked as ‘NA’.
settlement: Names of the settlements.
habitat: Habitat types – forest, grassland, habitat between solar panels (‘between’), habitat under solar panels (‘under’), and abandoned farmland (‘abandoned’).
site_ID: Site number within the given settlement, followed by the first (or first two) letters of the settlement name and the first letter of the habitat type.
1) Microclimate
Description: Contains values of microclimatic variables to analyse the effect of habitat type on microclimatic conditions. Three sites where data loggers were lost were excluded from the dataset.
date: Date of measurement.
day: Order of the day of measurement.
datalogger_ID: ID of the data logger.
humidity (%), min_temp (°C), max_temp (°C), mean_temp (°C), temp_variation (°C): Values of environmental variables (temp = temperature) for each habitat type and settlement on each date.
2) Spiders
Description: Contains variables used to analyse the effect of habitat type on ground-dwelling spiders.
richness: Number of spider species per site.
agrobiont: Abundance of agrobiont spiders per site.
humidity, shading: Community-weighted means (CWM) of humidity preference and shading tolerance per site.
3) Spiders_compo
Description: Contains a community matrix of pooled spider data from the two sampling periods, used to compare assemblage composition between habitat types. One site with zero individuals was excluded from the dataset. Each entry represents the Hellinger-transformed abundance of a given spider species at a site.
4) Beetles
Description: Contains variables used to analyse the effect of habitat type on ground beetles.
richness: Number of ground beetle species recorded per site.
humidity, shading: Community-weighted means (CWM) of humidity preference and shading tolerance per site.
5) Beetles_compo
Description: Contains a community matrix of pooled ground beetle data from the two sampling periods, used to compare assemblage composition between habitat types. Ten sites with zero individuals were excluded from the dataset. Each entry represents the Hellinger-transformed abundance of a given ground beetle species at a site.
Code/software
Analyses of the effect of habitat type on microclimatic conditions and ground-dwelling predatory arthropods were conducted using R (R Development Core Team, 2024), while comparisons of assemblage composition between habitat types were performed using CANOCO 5 (ter Braak and Šmilauer, 2012).
We tested the effect of habitat type on microclimatic conditions by generalized linear mixed models (GLMMs) with a Gaussian error structure using the ‘glmmTMB’ function in the ‘glmmTMB’ package (Brooks et al., 2017). We added an autoregressive correlation structure to each model: ar1 (as.factor(day) + 0 | datalogger ID). To account for repeated observations within the same settlement and datalogger, settlement and datalogger ID (unique for each datalogger) were included as random effects in each model.
We compared the species richness of both predatory groups and the abundance of agrobiont species of spiders between habitat types using generalized linear mixed-effects models (GLMMs) with a Poisson error structure (GLMM-p) using the ‘glmer’ function in the ‘lme4’ package (Bates et al., 2024). We used GLMMs with a negative binomial error structure (GLMMs-nb) if we detected overdispersion (Pekár and Brabec, 2009).
The community-weighted means (CWMs) were calculated within the ‘functcomp’ function in the ‘FD’ package (Laliberté et al., 2024). We evaluated the effect of habitat type on trait composition (CWM) by linear mixed-effects models (LMMs) using the ‘lmer’ function in the ‘lme4’ package (Bates et al., 2024). We added settlement as a random effect to each model.
Post hoc comparisons were made using the ‘glht’ function in the ‘multcomp’ package (Hothorn, 2024).
We compared the composition of ground-dwelling predatory arthropod assemblages between habitat types by canonical correspondence analysis (CCA). Only species with occurrence in more than three samples were selected for the analysis. The Hellinger transformation was applied to the species data before the calculation. The significance of the effects was tested using Monte Carlo permutation tests with 999 permutations within the blocks represented by the settlements.
References:
Bates, D., Maechler, M., Bolker, B. & Walker, S. (2024) lme4: Linear Mixed-Effects Models Using ‘Eigen’ and S4. R package version 1.1‑35.5.
Brooks, M.E.J., Kristensen, K., van Benthem, K., Magnusson, A., Berg, C.W., Nielsen, A., Skaug, H.J., Mächler, M. & Bolker, B.M. (2017) glmmTMB balances speed and flexibility among packages for zero‑inflated generalized linear mixed modeling. R Journal, 9, 378–400. https://doi.org/10.32614/RJ-2017-066
Hothorn, T., Bretz, F. & Westfall, P. (2024) multcomp: Simultaneous Inference in General Parametric Models. R package version 1.4-26.
Laliberté, E., Legendre, P. & Shipley, B. (2024) FD: Measuring Functional Diversity (FD) from Multiple Traits, and Other Tools for Functional Ecology. R package version 1.0-12.3.
Pekár, S. & Brabec, M. (2009) Moderní analýza biologických dat. 1. vydání. Zobecněné lineární modely v prostředí R. Scientia, Praha. [in Czech].
R Development Core Team (2024) R: A Language and Environment for Statistical Computing. Version 4.4.0. R Foundation for Statistical Computing, Vienna [software]. Available at: http://www.Rproject.org/
ter Braak, C.J.F. & Šmilauer, P. (2012) Canoco 5. Software for Multivariate Data Exploration, Testing, and Summarization. Netherlands [software].