Insect Decline-Evaluation of Potential Drivers of a Complex Phenomenon

Abstract

The decline of insects is a global concern, yet identifying the factors behind it remains challenging due to the complexity of potential drivers and underlying processes, and the lack of quantitative historical data on insect populations. This study assesses 92 potential drivers of insect decline in West Germany, where significant declines have been observed. Using data from federal statistical offices and market surveys, the study traces changes in landscape structure and agricultural practices over 33 years. Over the years, the region underwent major landscape changes, including reduced cropland and grassland and increased urbanization and forest areas. Potential detected drivers of insect decline include: (1) urban expansion, reducing insect habitats as urban areas increased by 25%; (2) intensified grassland management; (3) shifts in arable land use towards bioenergy and feed crop cultivation, particularly corn, driven by dairy farming intensification and renewable energy policies. While the toxic load of pesticide application has decreased, land-use changes, most likely driven by market demands and shifts in national and EU policies, have reduced habitat availability and suitability for insects. This study highlights how these landscape and land management changes over the past 33 years align with the observed decline in insect biomass in the region.

Keywords: agricultural intensification; bioenergy crop; landscape management; permanent grassland; pesticides; renewable energy; urbanization.

Figure 1

Mean (sampled) biomass of flying insects (in grams per day, FIB) during the insect flight period (April to September) is presented based on the biomass data derived from Hallmann et al. [14] and Mühlethaler et al. [42]. Each data point represents the mean insect biomass sampled per location. The depicted graph illustrates the LOESS correlation curves for the specific insect flight period and the entire sampling duration as reported in Hallmann et al. [14] study. The data for the years 2020 and 2021 from Mühlethaler et al. [42] were sampled only during the flight period and hence do not differ between the curves.

Figure 2

Biplot with visual representation of the relationships between variables for the full biomass dataset in the PCA. The direction of an arrow indicates the correlation between the variable and the principal components. The length of an arrow represents the variable’s contribution to the principal components. Longer arrows correspond to variables that contribute more to the variance explained by the principal components. Arrows pointing in the same direction are positively correlated, while arrows pointing in opposite directions are negatively correlated. The colors visualize the cosine square values (cos2), which indicate the quality of representation of that variable on the two depicted principal components. Arrows in red (high cos2 value) indicate that a large proportion of the variable’s variance is explained by the two principal components.

Figure 3

Depiction of variable loadings on PC1 for the full biomass dataset. The length of the bar represents the magnitude of the loading, and the direction indicates the sign of the correlation between the variable and PC1. Variables with higher absolute loadings contribute more to the variance explained by PC1.

Figure 4

Comparison of the proportions of the main landscape variables, the different permanent grassland types and the main crop types between 1989 and 2021 in NRW. For 2021, the relative differences to 1989 are indicated in percent and/or in 1000 hectare.

Figure 5

The complexity and visual correlation of potential drivers behind insect decline. The time series of selected variables spanning from 1989 to 2022 (where data was available) are depicted. Included in the presentation are the annual mean flying insect biomass (FIB), derived from Hallmann et al. [14] and Mühlethaler et al. [42], and various landscape, agriculture, renewable energy, climate-related variables and the toxic load of pesticides for the federal state of North Rhine-Westphalia (NRW) in Germany. Additionally, on the right side of the presentation, the minimum and maximum values for each variable are specified as well as vertical lines for visual aid. More detailed visualizations of the variables are presented in Supplementary File S5.

Figure 6

Toxic load of foliar-applied pesticides overall (A), and of foliar-applied insecticides per crop (B).