Deviation From the Regression of Yield on Nitrogen Fertiliser Rate as a Tool for Detecting Fraud in Organic Banana Production

ABSTRACT Background and aims Bananas are demanding in nitrogen (N) input; therefore, there is a temptation for organic farmers for using synthetic N fertilisers, which are not allowed under organic standards. The aim of our study was to develop a tool that identifies high banana yields obtained with suspiciously low organic N input. Methods We systematically reviewed literature from experimental studies on N fertilisation in bananas from all over the world. We also developed a simplified N balance model for organic bananas. Furthermore, N fertilisation and banana yield data from organic and conventional farmers in different countries were collected. From these, a subset of trustworthy organic farms was identified, as a reference concerning plausible ratios of yield versus fertilisation. A model was developed to estimate the deviation from the regression of trustworthy farms and thus identify suspicious cases. Results Neither literature nor the N balance led to a meaningful benchmark for differentiating plausible from non‐plausible yields. The regression of yield on N fertiliser rate from the trustworthy organic farmers, however, turned out to be a helpful reference, and the deviation from this regression helps to achieve our aim. Depending on the alert limit, that is, the probability of obtaining false positive results, 4, 6, or 9 out of 157 data‐pairs from organic farmers turned out to be suspicious. Conclusion Measuring deviation from the regression of the trustworthy farms is a useful tool for identifying organic banana farmers suspected to be using synthetic N fertilisers but is not in itself a proof of fraud. The model will improve as more data becomes available.