Phytotoxic Effects of Antibiotics on Terrestrial Crop Plants and Wild Plants: A Systematic Review

Abstract

This review examines the state of knowledge on the phytotoxic effects of antibiotics on terrestrial crop plants and wild (non-crop) plants with the goal of evaluating differences in their sensitivity. This is important because environmental risk assessments of antibiotics currently consider their potential effects only on crop species but not wild species. Overall, we analysed 275 datasets consisting of antibiotic-plant species-endpoint combinations for germination (mg/L) and 169 datasets for plant growth (elongation and biomass) (mg/kg). EC10 and EC50 of each parameter were compared using a quotient approach, in which the geometric mean and the 5th percentile of the crop data were divided by wild data. Quotients were > 1 for elongation growth, suggesting that wild species were more sensitive than crops, while they were < 1 for biomass growth, suggesting quite the contrary. However, < 1% of the data in each dataset came from wild species, preventing definitive conclusions. Merging crop and wild data to evaluate differences in sensitivity among classes of antibiotics and plant families, we found using a linear mixed effect model and post hoc test that plants were most sensitive to phenicol and least sensitive to macrolides and tetracyclines. Further work must be conducted to gain a better understanding of the phytotoxic effects of antibiotics on terrestrial wild plants and subsequently assess whether the current approach to environmental risk assessment of antibiotics is sufficient to protect plant biodiversity.

Fig. 1

Flow diagram of the literature screening

Fig. 2

Illustration of the ecotoxicological endpoints of seed germination and plant

Fig. 3

Endpoints (ECx) for each variable for a crop species (mg/kg soil), b wild species (mg/kg soil) and c crop species (mg/L solution)

Fig. 4

Differences in toxicity among classes of antibiotics: a elongation EC10 (mg/kg), b elongation EC50 (mg/kg), c biomass