Potential feed-to-milk transfer of tridemorph-fungicide detected by untargeted LC-HRMS-based metabolomics screening.

Abstract

The transfer of agrochemical residues into dairy products is a potential food safety concern, particularly in regions where residue monitoring remains limited. This study applied untargeted metabolomic LC-HRMS screening to investigate chemical features in dairy feed and corresponding milk samples collected from smallholder farms in East Java, Indonesia. A tridemorph-related feature was detected in both matrices, supported by accurate mass measurements (≤ ±5 ppm), consistent retention times (ΔRT = 0.025 min), stable mass errors within ±5 ppm, and MS2 spectral matching, corresponding to Schymanski Level 2 confidence. Semi-quantitative analysis based on relative peak areas showed significantly higher values in milk than in feed (p < 0.0001), with a median milk-to-feed ratio of 2.32. However, no significant farm-level correlation was observed between matrices (ρ = -0.19, p = 0.523), suggesting that transfer dynamics may be complex rather than directly proportional. Although absolute quantification was not performed, the co-occurrence of consistent tridemorph-related signals in feed and milk is compatible with the possibility of feed-to-milk transfer. These findings highlight the value of high-resolution untargeted screening approaches for proactive surveillance of agrochemical residues in dairy production systems.