Assessing the most sensitive and reliable endpoints in plant growth tests to improve arsenic risk assessment

Arsenic (As) is toxic to humans and the environment. Its toxicity has been assessed in many ways, including plant growth tests integrated in As risk assessments (RA). The tiered approach used in RA schemes assumes that lower tiers are more conservative than higher tiers. Plant growth tests may comprise lower tier of a RA and include the measurement of several endpoints. However, only few of these endpoints are highly sensitive and reliable, which makes them more appropriate to comprise lower tiers. Therefore, the selection of those endpoints is needed. The present study aimed to evaluate the most appropriate endpoints of plant growth tests to use in lower tier As RA schemes. This selection of endpoints was based on their sensitivity and reliability, using different tropical soils and plant species. In order to achieve this objective, six plant species were exposed to eight levels of As contamination (0; 8; 14.5; 26; 46.5; 84; 150; 270 mg kg^-1), in three different tropical soils (Oxisol, Inceptisol, and tropical artificial soil). The endpoints measured were: first germination count (FrC), plant height (PH), relative leaf area (RLA), stem diameter (SD), total germination (TG), germination speed index (GSI), dry mass (DM), number of completely expanded leaves (CEL), plant survival (PS), soil plant analysis development chlorophyll level (SPAD), and the final germination count (FnC). Toxic values for 50% of effect were estimated for each endpoint within each species and test soil, to rank them according to their sensitivity and reliability. The most sensitive endpoints were: FrC, RLA, DM, GSI, PH, and FnC, while the most reliable endpoints were: FrC, DM, GSI, DM, PH, FnC, and TG. Our findings suggest that FrC, DM, GSI, PH, and FnC are the most adequate endpoints to be used in plant growth tests as lower tiers of As RA in tropical regions.