The importance of landscape composition for pest control and crop yield: A global quantitative synthesis
Data files
Oct 03, 2025 version files 1.55 MB
-
Analysis_SEM.R
36.41 KB
-
Data_SEM.csv
1.51 MB
-
README.md
4.25 KB
-
SI_3_Data_Exploration.r
6.19 KB
Abstract
Using a global structural equation model of 116 studies from 28 countries, we tested three hypotheses: The ‘natural enemy hypothesis’, that natural areas increase natural enemies and suppress pests; the ‘resource concentration hypothesis’, that simplified agriculture increases pests; and the ‘agronomic quality hypothesis’ with a structural equation model. This repository contains the dataset and R script for the analysis that examines how landscape composition directly and indirectly affects crop yield, mediated through natural enemies, biological control, and pests, across multiple studies using a structural equation model.
Dataset DOI: 10.5061/dryad.9ghx3ffwg
Description
This repository contains the dataset and R scripts used in the manuscript submitted to Ecology Letters. The analysis explores how landscape composition directly and indirectly influences crop yield mediated through natural enemies, biological control, and pests across multiple studies using a structural equation model.
Contents
- Data_SEM.csv: Excel file containing the compiled dataset from 116 studies across 28 countries. The original data from which this dataset is derived is presented in the supplementary material from Karp et al. 2018, where the data were already standardized: https://www.pnas.org/doi/abs/10.1073/pnas.1800042115#supplementary-materials
- SI_3_Data_Exploration.r : R script that performs preliminary analyses to determine needs for data transformation and selects the landscape variables that need to be included in the structural equation model. This Analysis was performed using R version 4.4.2 (2024-10-31 ucrt) -- "Pile of Leaves” with packages ‘nlme’, ‘MuMIn’ and ‘dplyr’.
- Analysis_SEM.R: R script that performs data transformation, structural equation model analysis, and analysis of spatial autocorrelation and multicollinearity. This Analysis was performed using R version 4.4.2 (2024-10-31 ucrt) -- "Pile of Leaves” with packages ‘piecewiseSEM’, ‘nlme’, ‘dplyr’, and ‘writexl’.
Dataset interpretation
Below, we will explain the content of each of the columns found in the dataset:
Nr: Unique identifier for each row of data
Study_ID: Unique identifier for each study
Site: Site name given by the authors of each study
Study_Year: The year in which the study was performed. Some studies repeat the experiment in several years, and more than one year will appear for that given study.
X: Longitude
Y: Latitude
NAT_250: Proportion of natural area at the 250 decay rate
NAT_750: Proportion of natural area at the 750 decay rate
NAT_1250: Proportion of natural area at the 1250 decay rate
AG_250: Proportion of agricultural area at the 250 decay rate
AG_750: Proportion of agricultural area at the 750 decay rate
AG_1250: Proportion of agricultural area at the 1250 decay rate
Crop: Latin or common species name of the crop, based on what was reported by the authors of each study
Poll_depend: Pollinator dependency of the crop. For details on the classification see Supplementary Information 5 of the paper. “poll” means that the crop is pollinator dependent and “no_poll” means that the crop is not pollinator dependent.
Enemy: Standardized abundance of all natural enemies (for standardization methods, refer to the paper).
Predator: Standardized abundance of all predators.
Parasitoid: Standardized abundance of all parasitoids.
Biological control: Standardized measure of biological control
pest: Standardized abundance of all pests.
specialist: Standardized abundance of all specialist pests. For more information see Supplementary Information 1 of the paper.
generalist: Standardized abundance of all generalist pests. For more information see Supplementary Information 1 of the paper.
Marketable_or_Total: Yield measure done by the authors of each study that distinguished between marketable (that can be sold in the market) and total (where there is no distinction if the produce is of marketable quality or not).
Stand_yield: Standarized yield measurement.
Cells with an ‘NA’ value mean that the data are not available, because they were not collected in the field by the research team that did this work.
Data was derived from the following source:
- Supplementary information of Karp et al. 2018, Crop pests and predators exhibit inconsistent responses to surrounding landscape composition, Proc. Natl. Acad. Sci. U.S.A. 115 (33) E7863-E7870, https://doi.org/10.1073/pnas.1800042115 (2018).
