# A cohort study of sustainable cultivation methods in mandarin orange orchards across Japan

## TL;DR

This study uses real-world data from mandarin orange orchards in Japan to evaluate how pesticide and fertilizer methods affect fruit quality and soil health.

## Contribution

The study introduces a nationwide cohort approach with inverse probability weighting to analyze sustainable cultivation methods in real-world agricultural settings.

## Key findings

- Cultivation methods significantly affect plant pathogens, nutritional quality, and soil properties.
- Nationwide cohort analysis detected more significant effects than local-scale comparisons due to larger sample size.
- Real-world data provide valuable feedback for improving sustainable agricultural practices.

## Abstract

Variability in environmental conditions and farming practices often leads to discrepancies between experimental results and outcomes in farmers’ fields. This gap poses a challenge for understanding the effects of agricultural inputs and methods under real-world conditions, particularly in fruit cultivation systems, where large-scale experimental data are limited. In this study, we applied a cohort study approach leveraging data from farmers’ fields to investigate the effects of pesticide and fertilizer application methods on fruit quality and soil properties in mandarin orange orchards. Biases arising from differences in covariates among cultivation methods were controlled using the inverse probability weighting (IPW) based on propensity scores. Consequently, compared to local-scale analysis between adjacent fields, the nationwide cohort analysis detected a greater number of significant effects of cultivation methods by utilizing its larger sample size. Through this analysis, we found important insights into the effects of pesticide and fertilizer application methods on plant pathogens, nutritional quality, and soil properties in sustainable cultivation systems of mandarin orange. This study demonstrates that cohort analyses using real-world data have great potential to advance agricultural biotechnology by providing effective feedback from farmers’ fields and bridging the gap between scientific research and real-world agriculture.