Summary

Effekter af pyrethroidet lambda-cyhalothrin på biologisk struktur, funktion og rekolonisering i vandløb

Background and aim

Pesticides are transported to the aquatic environment in varying number and amounts. The main sources are runoff from sprayed fields, spray drift and point sources such as runoff from sites where spraying equipment is being cleaned. A considerable number of different pesticides, primarily herbicides and fungicides, have been found in Danish streams. Insecticides, including pyrethroids which are the most important group, have also been found on several occasions. Pyrethroids are relatively insoluble in water and may easily adsorb to surfaces of different minerals and organic matter. Consequently, they will adhere to soil particles and can be transported in this form to the aquatic environment by runoff. However, even in relatively small amounts, they are extremely toxic and efficient against target pest insects as well as many non-target invertebrates. Therefore there is every reason to focus on these substances and their effects in e.g. streams. Previous projects carried out under The Danish Environmental Protection Agency’s ”Pesticides Research Programme” have provided a considerable amount of information on the effects of pyrethroids on mortality, activity, feeding rate and reproduction of selected stream macroinvertebrates. However, most of this research was done in the laboratory. Therefore there is a great need for studies of the effects of transient pulses of pyrethroids on natural assemblages of macroinvertebrates, especially how fast these assemblages and their function in the turnover of organic matter are re-established, and whether pyrethroid contamination of the habitats and food of the macroinvertebrates may prolong the process of re-establishment. The aim of the present study is to fill this gap in our knowledge if possible.

The study

We carried out a number of experiments on the effects of a selected pyrethroid, lambda-cyhalothrin, on stream macroinvertebrates. These experiments were performed on varying scales of space and time. However, in all cases a pulse exposure of 90 minutes was used, either exposing the macroinvertebrates themselves or their habitat and food.

In laboratory tanks and small artificial streams (2 metres long plastic gutters) survival, recovery, mobility, and feeding rate were studied over a period of one to two weeks in four species of macroinvertebrates: the crustacean Gammarus pulex, the mayfly Heptagenia sulphurea, and the caddisflies Chaetopteryx villosa and Sericostoma personatum. Nominal concentrations of 0.005-5.0 µg l^-¹ were used.

On a larger scale the acute effects on macroinvertebrate assemblages, including their ability to process organic matter and to recolonize, were studied in six large artificial stream channels for a period of 30 days. Each of these channels were 12 metres long (width 0.6 metres), fed with 4-5 litres s^-¹, and equipped with riffle and pool sequences containing stones/coarse gravel and fine gravel/sand, respectively. We also added mesh bags with conditioned beech leaves. The channels were uniformly stocked with natural macroinvertebrate assemblages from the nearby stream that also fed them. The downstream halves of three of the six channels were exposed to 5.0 µg l^-¹ of lambda-cyhalothrin, whereas the remaining channels acted as controls.

Furthermore, effects of lambda-cyhalothrin on macroinvertebrate assemblages in artificial substrates were studied in nine Mid-Jutland streams (including three size classes) and one Funen stream, all being relatively unaffected by human activities. In the Mid-Jutland streams colonized gravel boxes and leaf packs were exposed to 0, 0.5 and 5.0 µg l^-¹ in special tanks out of the streams and afterwards reintroduced to these. Hereafter, recolonization and turnover of organic matter (benthic algae and leaf loss) was followed over a period of 30 days. Only macroinvertebrate data from three of the streams were analysed, as they seemed of relatively little informative value due to an inappropriate experimental design (the current velocity in the tanks was so low that it affected the macroinvertebrates even in the control experiments). Instead, we made an additional experiment introducing beech leaf packs that where either treated with 5.0 µg l^-¹ or untreated. The colonization and weight loss of these packs were then studied for a period of 60 days.

Finally, we carried out a study of the dispersal and colonization potential of stream macroinvertebrates on a much larger scale in time and space. Thus, we analysed data from yearly monitoring of approximately 900 stream sites in Funen over a period of 20 years. This time series was especially suitable as many pollution sensitive species only occurred very locally at the beginning of this period due to the previously poor water quality. The water quality improved considerably after about 1990 due to better treatment of waste water and a stop for illegal discharges of silage juice and liquid manure from husbandries. This made it possible to study the dispersal of macroinvertebrates and their recolonization of physically suitable stream habitats, including their over-land colonization from one stream catchment to another.

Main conclusions

Lambda-cyhalothrin concentrations of 5.0 µg l^-¹ may kill the majority of stream macroinvertebrates (mostly crustaceans and insects), and the breakdown of dead organic matter and grazing of benthic algae is thereby markedly reduced. Concentrations of 0.5 µg l^-¹ may increase mortality among crustaceans and insects, and the surviving individuals may be less active and their processing of food is reduced. Concentrations in this order of magnitude are likely to occur in small streams due to leaching from newly sprayed fields following heavy rain incidents. Moreover, contamination of habitats and food (benthic algae, beech leaves) due to a pulse of 0.5 µg l^-¹ induced increased mortality or made the food less attractive. This may reduce decomposition of organic matter and delay macroinvertebrate recolonization of an impacted stream reach.

From the present studies, supported by previous studies, it may be concluded that the macroinvertebrate community may be successfully re-established following pulse contamination with a pyrethroid that affects a stream reach. If the stream reach in question is not located at the most upstream part of the respective stream system, recolonization will primarily take place form upstream reaches by so-called drift. The recolonization will be especially fast during summer and early autumn where many stream insects deposit their eggs, but even then complete reestablishment may take one to two years. If, however, the insecticide affected reach did contain species that, due to their rareness or very local distribution, were not present elsewhere in the respective catchment, recolonization must occur over land from streams in other catchments. Dependent on the distance to the nearest populations of the species, favourable whether conditions during the flight periods of the adults, and successful reproduction when finally reaching a new habitat, recolonization may take years, decades or even longer.

Results from the project

The experiments that were carried out under controlled conditions in laboratory tanks, small stream mesocosms, and in artificial as well as natural streams rendered probable that a pyrethroid pulse of a magnitude of 5 µg l^-¹ not only dramatically changed the macroinvertebrate assemblages. It also contaminated potential habitats and food sources like beech leaves and benthic algae, thereby increasing mortality, activity and feeding rates. Even pulse axposure at concentrations 10-1000-fold lower may have effects (increased mortality, reduced activity, or hyperactivity) on some macroinvertebrate species under laboratory conditions. It is therefore possible that some effects may also be expected in natural streams at these concentrations. However, it seems unlikely that contamination of habitats and food alone will affect the macroinvertebrate assemblages and their turnover of organic matter.

Pulse concentration of lambda-cyhalothrin	Increased mortality	Within 2 weeks after pulse exposure
Pulse concentration of lambda-cyhalothrin	Increased mortality	Reduced activity	Reduced consumption of food
0.05 µgl^-¹l	None in laboratory studies with freshwater shrimps and case-bearing caddisflies	None in laboratory studies with freshwater shrimps and case-bearing caddisflies	None in laboratory studies with freshwater shrimps and case-bearing caddisflies
0.5 µgl^-¹	Freshwater shrimps and case-bearing caddisflies in laboratory tanks and stream microcosms	Freshwater shrimps and case-bearing caddisflies in laboratory tanks and stream microcosms	Freshwater shrimps, mayflies and case-bearing caddisflies in laboratory tanks and/or stream microcosms
5.0 µgl^-¹	Freshwater shrimps and case-bearing caddisflies (almost 100%) in laboratory studies, and in macroinvertebrate assemblages in artificial streams	Freshwater shrimps and case-bearing caddisflies in laboratory studies	Freshwater shrimps, mayflies and case-bearing caddisflies in laboratory studies
Contaminated beech leaves - 5,0 µgl^-¹	Freshwater shrimps and case-bearing caddisflies in laboratory studies, and reduced macroinvertebrate abundance and biodiversity in leaf packs in natural streams	Freshwater shrimps and case-bearing caddisflies in laboratory studies	Freshwater shrimps and case-bearing caddisflies in laboratory studies
Contaminated algae - 5,0 µgl^-¹	Mayflies in laboratory studies		Mayflies in laboratory studies

The results from the laboratory and field experiments are summarized in the table.

The experiments in natural streams clearly showed that macroinvertebrates colonized artificial substrates very fast (i.e. within a few days), as they only had to overcome small distances. However, recolonization of pesticide contaminated artificial stream reaches of several metres was not at all complete after 30 days, and may take considerably longer time. Our study did not permit conclusions on recolonization over larger distances within stream systems and to identify important dispersal and colonization routes. However it is well known that dispersal may occur as passive drift, active swimming or crawling, as well as downstream and upstream flight of females carrying eggs, all these being of importance.

On the other hand, the study of monitoring data from Funen streams clearly demonstrated that the adults of many stream insects were able to colonize over land from one stream to another, often crossing catchment boundaries. Recolonization could take place over a distance of 8 – 16 km for several species of stoneflies, mayflies and caddisflies, although there were examples of even longer dispersal distances. Several species showed a positive relationship between the distance over which they colonized and the time until colonization was successful. We also found evidence that weather condition during the flight period of insects influenced the dispersal and colonization. However, it was not possible to show any effect of landscape structure on over land colonization.