Overcoming water quality effects in biological monitoring: a case study of amphipod in situ exposures in Ontario agricultural streams

Bioindicators add valuable understanding of biological impacts to contaminant monitoring programs. However, attributing effects (e.g., mortality and growth impairment) to contaminant exposures is challenging because of potential confounding by environmental variables. We assessed the influence of four water quality variables (temperature, dissolved oxygen, pH, and conductivity) on contaminant effects assessments during in situ exposures of the amphipod Hyalella azteca in six agricultural and urban watersheds in southern Ontario, Canada (2005-06, 2008-10). We further tested whether sampling in specific months of the growing season would minimize confounding effects. While high toxicity from organophosphate and carbamate pesticides increased mortality and reduced growth in caged amphipods, warmer stream temperatures also affected endpoints, increasing mortality and improving growth. Seasonal patterns indicated early summer (June) as optimal for detecting impacts when: (1) stream pesticide concentrations were highest, (2) acetylcholinesterase (AChE) inhibition (a biomarker of organophosphate/carbamate exposure) was highest, and (3) stream temperature was below its seasonal peak. Specifically, higher correlations among organophosphate pesticide concentrations, AChE inhibition, and mortality indicated better attribution of cause in this month (r = 0.53-0.76, p < 0.05). Ability to discriminate between pesticide-impacted sites and reference sites was also greater than other months (June = 100% correct assignment as high-impact or reference site). Considering sampling times of stream bioindicators that maximize pesticide responses and limit confounding effects of water quality may improve the accuracy and resource-efficiency of biological monitoring programs.