In situ investigation of the effects of Vision® and Release® silvicultural herbicides on plankton and larval amphibians
The effects of silvicultural herbicides Vision® and Release® on native larval amphibians (' Rana pipiens' and 'Rana clamitans'), natural zooplankton and phytoplankton populations, and periphytic algae were investigated using ' in situ' aquatic mesocosms or limnocorrals. Amphibians were assessed for mortality, avoidance response, and growth. Growth was examined through a unique protocol using digital image analysis. Zooplankton were assessed for declines in abundance among major taxonomic groups (Copepoda, Cladocera, Rotifera). Phytoplankton abundance was quantified using previously established flow cytometric methods, and taxonomic identification of species was carried out at three sampling times. Herbicide effects generally showed variation in temporal trends, severity of impact, and time to recovery between experimental sites, likely due to differing physical and chemical characteristics of the water therein (e.g. pH). These herbicide formulations did not cause reductions in phytoplankton abundance or periphytic biomass; rather, stimulation was seen in many instances. Shifts in algal dominance occurred in treated enclosures independent of the temporal changes seen in untreated controls. Zooplankton were the most sensitive group of organisms. Concentrations causing 50% declines in the abundance of major zooplankton taxa (EC50) ranged from 0.054 to 0.21 mg a.e. L-1 Release®, and 0.17 to 6.31 mg a.e. L-1 Vision®, overlapping previously detected aquatic concentrations. Depressions in zooplankton abundance likely released phytoplankton from herbivory, resulting in increased abundance as an indirect effect. Concentrations causing 50% mortality of larval amphibians (LC50) ranged from 2.79 to 3.29 mg a.e. L-1 Release ®, and 4.25 to 11.47 mg a.e. L-1 Vision ®. Herbicide treatment did not reduce larval amphibian growth, likely due to the lack of herbicide impact on phytoplankton and periphyton. Single-species laboratory-based toxicity tests would neither have been able to discern variations between different wetlands, nor detect indirect effects. Given the use pattern and expected aquatic herbicide concentration, neither herbicide is expected to cause significant effects in forest wetlands. ' In situ' aquatic mesocosms proved to be a valuable tool in this aquatic toxicology study because of their ability to address the shortcomings of laboratory-based tests.