Bekæmpelsesmiddelforskning fra Miljøstyrelsen, 89 – Pesticiders påvirkning af planter og alger i vandmiljøet

The Effect of Pesticide Residues on Aquatic Plants and Algae

Summary and conclusions

The aim of this study has been to investigate the effect of pesticide residues on aquatic plants and algae in the Danish environment. Our primary focus has been on herbicides, as they are products designed to kill plants and therefore are believed to be the group of pesticides with the largest potential effect on the aquatic flora. To get a broad view of the aquatic toxicity of herbicides we did a database study on approximately 150 different herbicides, representing 15 different modes of action. The aquatic toxicity, expressed as the EC₅₀ of algae, was compared to the terrestrial toxicity, estimated as the average recommended field dose. Generally, herbicides were very inefficient in water. For more than half the herbicides included in this study, a full field dose sprayed on a 30 cm water column could not reduce algal growth by more than 50%. The low aquatic toxicity was verified in laboratory experiments on both Lemna minor (Duckweed) and the algae Pseudokirchneriela subcapitata with 10 selected herbicides. Two of the potentially most toxic herbicides sold and used frequently in Denmark were chosen for further investigation. These were the sulfonylurea herbicide metsulfuron-methyl and the s-triazine terbuthylazine. For these two herbicides, species specific sensitivity was investigated on 10-12 aquatic plant species and a naturally evolving epiphyte community under two growth conditions, and Species Sensitivity Distributions (SSD's) were constructed. The results revealed a relatively small variance in sensitivity between species, varying by a factor of 56 and 34 between the most and the least sensitive species for metsulfuron-methyl and terbuthylazine, respectively. This should be compared with the selectivity between terrestrial species, which can vary by a factor of up to 1000. The experiments also showed that the standard Lemna and algae tests represented the sensitivity of the aquatic plants and the epiphyte community well, despite the rather different growth conditions in the different test systems. If a safety factor of 10 was used between the maximal allowable environmental concentration and No Observable Effects Concentration (NOEC), defined as EC₁₀, or a factor of 100 between the maximal allowable environmental concentration and EC₅₀ on the Lemna test, more than 95% of the species would be protected. The toxicity of metsulfuron-methyl was independent of the growth conditions in these experiments, while the toxicity of terbuthylazine was enhanced as irradiance increased.

As the highest herbicide concentrations in the environment are expected to occur in pulses associated with spray and/or precipitation events, pulse exposure with different herbicides was investigated on Lemna minor. The six herbicides selected represented three modes of action and had different K_ow within the same mode of action. The experiments showed that a three-hour pulse, with concentrations that were lethal at chronic exposure, had no effect on Lemna growth measured over 4 days when the herbicides were photosynthetic inhibitors, diquat and terbuthylazine. For the ALS-inhibitors, imazamox and metsulfuron-methyl, and the inhibitors of microtubule assembly, propyzamid and pendimethalin, a three-hour pulse did affect growth. The EC₅₀ of the pulse-exposed plants was approximately a factor of 10 higher than the EC₅₀ of the plants experiencing chronic exposure. There was no detectable difference between herbicides with the same mode of action, but different K_ow. The reason why a pulse exposure with photosynthetic inhibitors was so relatively inefficient in reducing plant growth compared to a pulse exposure with the other herbicides could be that damage caused by reactive oxygen species, which is the immediate effect of photosynthetic inhibiting herbicides, is a common phenomenon in all plants. Hence, an efficient repair system already exists to ameliorate this kind of damage. It seems to be harder to recover from direct or indirect inhibition of cell division, and for metsulfuron-methyl it was shown that it took about four days to recover from an almost lethal pulse exposure.

The general conclusion of all experiments was that herbicides in the concentrations measured in Danish surface waters (< 10 μg l^-1) would generelly neither reduce growth nor change the composition of submerged species significantly,- at least not if the effect of the herbicides was assessed for each herbicide individually. On the contrary, a stimulation of plant growth at low herbicide concentrations was observed in several experiments, a phenomenon called hormesis. Some fungicides have been reported to stimulate plant growth in the field, and we saw a similar effect on Lemna grown under sub-optimal light conditions when exposed to low doses of prochloraz (10 and 20 μg l^-1). As soon as growth conditions were optimized, however, the growth stimulating effect disappeared. The presence of prochloraz (<100 μg l^-1) did not affect the toxicity of either metsulfuron-methyl or terbuthylazine.

Even though the presence of herbicides in the concentrations measured in Denmark is unlikely to have a significant effect on the aquatic flora if present alone, it is more uncertain what the effect of several simultaneously occurring herbicides would have. Here, we present the Additive Dose Model, also called Concentration Addition, which can be used to assess the effect of mixtures of herbicides, if comparable toxicity data for all the single substances are available. We show an example where several photosynthetic inhibitors have been measured in a pulse event associated with high precipitation. The example shows, that the cumulative effect of the three herbicides could potentially affect the growth of algae. We consider the Additive Dose Model to be a good tool to identify critical situations that would need further investigation. A mixture experiment with metsulfuron-methyl and terbuthylazine performed on Lemna and algae showed that the cumulative effect of the mixture was less than expected from the Additive Dose Model. This proved to be a common pattern and supports the use of the Additive Dose Model as conservative estimates of a worst-case situation. As one of the goals of the present project was to focus on low concentrations and determine No Effect Concentrations, we have discussed the definition of "no effect". We chose to use EC₁₀ and in general we prefer the use of EC_x values instead of the often used NOEC and Lowest Observable Effect Concentration (LOEC), as EC_x values are more consistent across test-systems. Also, the variance on the EC_x gives an impression of the experimental quality that is not reflected in either NOEC or LOEC.

It is our opinion that the largest uncertainty about the effect of pesticides on the aquatic flora is to be found in the situations where pesticides occur in mixtures. More extensive knowledge of the composition, concentrations and time courses of mixtures and their effect on the flora could help to define places and situations where potentially harmful mixtures could occur. We also believe that the mechanisms of pesticide uptake by aquatic plants are not very well described as there are indications that these are very different from uptake mechanisms in the terrestrial environment, especially for aerial herbicides. A better understanding of the mechanisms of pesticide uptake could facilitate better predictions of the aquatic toxicity of pesticides and other chemical compounds. The general picture, however, is that the individual pesticide concentrations measured in Danish surface waters will not affect the growth and species composition of the aquatic flora significantly.