Modelling ionisation-dependent fungicide penetration into grape berries using a Level IV fugacity model with compound-specific cuticular permeability

Fungicide residues in table grapes are a critical food-safety concern because the fruit is consumed fresh, often with the skin, and grapes rank among the commodities with the highest frequency of Maximum Residue Limit (MRL) exceedances in European Union and Brazilian monitoring programmes. We present a Level IV fugacity model for the three-compartment grape berry (boundary air layer, lipidic skin, aqueous pulp) that extends the periodic fugacity control-system framework to incorporate (i) ionisation correction via the Henderson–Hasselbalch equation, (ii) compound-specific cuticular permeability following the Riederer–Schreiber model, (iii) logistic berry growth, and (iv) Arrhenius-dependent degradation rates derived from experimental D T 50 values. The model is applied to 11 fungicides used in Brazilian viticulture. Reference degradation rates are derived from experimental D T 50 values per compound and compartment, correcting a systematic underestimation by a factor of 15 to 154. The model reproduces the field observation that degradation accounts for 52%–100% of the initial mass within 240 h when calibrated with experimental D T 50 values. The Spearman correlation between D T 50 , pulp and residual skin fraction at 240 h is ρ = 0.96 ( p < 0.001 ): persistence governs long-term skin residue more strongly than lipophilicity or cuticular permeability. Cymoxanil ( f n ⋅ K ow = 0.15 < 1 ) is the only compound with a thermodynamic driving force towards the pulp; rapid degradation prevents this from producing measurable pulp penetration. A Monte Carlo analysis ( N = 500 per compound) with Partial Rank Correlation Coefficient sensitivity assessment reveals that mean temperature and D T 50 are the dominant parameters ( | PRCC | = 0.88–0.95), temperature acting simultaneously on degradation, volatilisation, and air diffusivity. A 10 °C regional temperature range shifts the predicted skin fraction of persistent compounds by 29–33 percentage points, providing a quantitative basis for region-specific pre-harvest interval recommendations.