Exposure assessment of anticoagulant rodenticides in the liver of red foxes (Vulpes vulpes) in Slovenia

TL;DR

In 148 red fox livers from Slovenia (2019–2022), at least one anticoagulant rodenticide was detected in 77.7%, with SGARs dominating (bromadiolone 75.0%, brodifacoum 51.4%, difenacoum 18.9%). Pooled AR concentrations were high (range 1.5–2866.5 ng/g; mean 601.4, median 350.2 ng/g) and exceeded levels in comparable studies, indicating serious toxicological risk for red foxes in Slovenia and the Western Balkans.

Key findings

• At least one AR residue was detected in 115/148 livers (77.7%); 22.3% contained one substance, 35.8% two, 16.9% three, 2.0% four, and 0.7% five. SGARs were found in 77.7% and FGARs in 9.5%; bromadiolone was most frequent (75.0%), followed by brodifacoum (51.4%), difenacoum (18.9%), coumatetralyl (8.1%), flocoumafen (1.4%), and warfarin (1.4%).
• Pooled AR concentrations ranged from 1.5 to 2866.5 ng/g (mean 601.4, median 350.2). Bromadiolone had mean/median 390.7/206.5 ng/g and maximum 2638.8 ng/g; brodifacoum had mean/median 331.4/210.6 ng/g and maximum 2642.3 ng/g. FGAR residues were mostly below 10 ng/g, with one coumatetralyl exception at 55 ng/g.
• Bromadiolone exceeded 200 ng/g in 37.8% of samples, 800 ng/g in 10.8%, and 2000 ng/g in 1.4% (two samples); brodifacoum exceeded 200 ng/g in 26.4%, 800 ng/g in 10.1%, and 2000 ng/g in 0.7% (one sample). These concentrations were reported as much higher than in comparable monitoring studies.
• Residues were detected in all monitored statistical regions (prevalence 56–100%), with higher concentrations in eastern Slovenia. Professional AR use was 1.677 kg in 2020, mainly bromadiolone (1.054 kg), difenacoum (0.252 kg), and brodifacoum (0.186 kg); brodifacoum professional use increased from 90 g to 201 g during 2016–2019 (223%).

Methods

• Liver tissue from 148 red foxes was collected in 2019–2022: 89 males and 59 females, age 0.4–5.7 years (mean 1.6), body mass 3–9.5 kg (mean 6), from 55 hunting districts and five special-purpose state hunting grounds across 10 of 12 Slovenian statistical regions.
• Approximately 1 g wet liver was extracted with methanol/water (2:1, v/v), cleaned up by solid-supported liquid-liquid extraction, and measured by LC-ESI-MS/MS; reporting limits were 0.5–5.0 ng/g, all samples were measured twice, and concentrations were not surrogate- or recovery-corrected.
• Analysis used descriptive statistics, nonparametric tests (Shapiro-Wilk, Spearman/Kendall, Wilcoxon/Mann-Whitney, Kruskal-Wallis, chi-square/Fisher), and generalised linear models (logistic for presence, Gaussian for log concentration) in R 4.2.3; professional AR use was assessed by questionnaire to DDD providers (33/44 DDD; 44/46 veterinary concessionaires) for 2016–2020.

Limitations

• Stated: urban foxes were not included because urban areas in Slovenia are non-hunting areas; private/retail AR use could not be surveyed for confidentiality, so use data reflect professional DDD use only; no toxicological data for brodifacoum were evaluated for foxes; synergistic/additive effects of multiple AR residues were not investigated; limited sample size led to pairwise statistical relationships.
• Stated: interpretation of AR residues and toxicity is challenging because many variables affect toxicity thresholds, and only bromadiolone had toxicological thresholds applied for foxes.
• Evident from the text: carcasses came from regular harvest or deaths from vehicle/firearm injuries rather than a random population sample; exclusions for decomposition (n=6) or severe organ injury (n=7 vehicle, n=7 firearms) may bias the sample; regional sample sizes were uneven (Coastal Karst n=1 was excluded from some tests); two sampling periods were treated as a single interval; no macroscopic lesions indicating poisoning were observed.