Pharms on farms: Impacts of emerging pharmaceutical contaminants on arbuscular mycorrhizal symbioses and microbial communities

Background
Agricultural soils are commonly amended with manure and biosolids, introducing veterinary and human-use pharmaceuticals. Consequences for soil microbes and arbuscular mycorrhizal (AM) fungal symbioses, where AM fungi exchange phosphorus (P) and nitrogen (N) for plant-fixed carbon (C), are poorly understood. This thesis investigates the impacts of antibiotics (doxycycline, sulfamethoxazole), an anthelmintic (flubendazole), and an antifungal mixture (clotrimazole, miconazole nitrate, fluconazole) at environmentally relevant concentrations on AM fungi and soil microbes across multiple plant systems.

Methods
Using isotope tracing approaches (33P, 15N, 14C), microscopy, DNA sequencing, and qPCR, I evaluated pharmaceuticals effects on AM fungal structures, nutrient exchange, and broader microbial community diversity and abundance. Additionally, I assessed impacts on plant biomass and shoot P and N content.

Results
Sulfamethoxazole had minimal effects, while flubendazole impaired AM root colonisation and C allocation to AM fungi. Conversely, manure-borne doxycycline enhanced AM hyphal density and 33P transfer to dwarf runner beans, while reducing bacterial diversity. Lower doxycycline concentrations typical of biosolids negatively impacted AM structures and 33P transfer to barley. Meanwhile, antifungals negatively impacted AM symbioses with spring onion, lettuce, and barley. Dual exposure to both doxycycline and antifungals caused similar negative impacts on AM-barley symbioses. If not specifically mentioned, other metrics were unaffected.

Conclusions
This work demonstrates that pharmaceuticals can significantly impact AM fungal symbioses and microbial communities, with effects likely depending on chemical class and plant identity. However, AM-mediated 33P transfer appeared particularly sensitive to pharmaceuticals and occurred alongside changes to AM structures. Because AM fungal responses to doxycycline in ‎Chapter 3 were likely mediated indirectly by changes to bacterial communities, broader microbial responses should be evaluated. Despite these specific impacts, many metrics were unaffected by pharmaceutical exposure. Together, my findings emphasise that the consequences of soil pharmaceutical contamination are unlikely to be predicted from single-crop and single-pharmaceutical assessments.