Bekæmpelsesmiddelforskning fra Miljøstyrelsen, 92 – Effekten af sprøjtemiddelafdrift på buske og træer i læhegn - Bær som indikator for biodiversitetsforandringer

Effects of Herbicide Spray Drift on Trees and Shrubs in Hedgerows

Summary

In this report, effects of simulated herbicide spray drift on hedgerows consisting of hawthorn were studied. Observations were made both within the spraying season, as well as in the following year.

Background

It is known that the density and diversity of weeds, insects and birds in the field are affected by the use of pesticides. It is furthermore expected that herbicide spray drift might affect trees and shrubs forming the property line or hedgerows between fields. However, there is only limited knowledge of the unintended effects of pesticides on trees and shrubs, both nationally and internationally. This includes direct toxic effects as well as indirect effects on the organisms associated with the trees and shrubs such as insects and birds. Experiments, which simulated herbicide spray drift, all found negative effects from sulfonylurea-herbicides and from glyphosate on the growth and fruit set of bird cherry (Prunus avium) (Al-Khatib et al. 1992, Bhatti et al. 1995 and Fletcher et al. 1996). Negative effects were observed down to dosages equivalent to 1% of the maximum dosage given on the label. According to our knowledge similar studies of herbicide spray drift have not been conducted in Denmark or in the rest of Europe. However, herbicide spray drift is most likely to be important for food chains associated with flowering, fruit set and seed dispersal.

Aim

The primary aim of the project was to provide knowledge of potential effects of herbicides on trees and shrubs in the year of exposure as well as in the year after. Potential effects on these species were exemplified with the herbicide metsulfuron-methyl and with hawthorn (Crataegus monogyna).

The project sought to fulfil the aim by answering the following questions:

Is hawthorn affected at dosages of metsulfuron that correspond to levels of spray drift expected when spraying with the maximal dosage recommended on the label under normal spraying conditions?
Is timing of herbicide applications important?
Which end-points respond the most to spray exposure: Leaf growth, bud formation, flowering or fruit set?
Is any effect traceable the year after spraying?
What is the effect of herbivorous insects on the fruit set?

Design

The project was designed on the basis of a pilot project. The pilot project estimated the variability of biomass and number of leaves, buds, flowers, green and mature fruits (Kjær et al. 2002). Based on statistical power analyses and logistic constrains, the design included eight hedgerows dominated by hawthorn. The experimental hedgerows were geographically distributed so that five of the hedgerows were positioned in Jutland and three on Zealand. Each hedgerow was sprayed with four dosages on two occasions in order to study the importance of spray timing. The nominal dosages were 5%, 10%, 20% and 40% of maximum label rate, and a control. For each spray event, 13 trees were sprayed per dosage per hedgerow. The control trees were not sprayed.

Activities

Spraying

The early (spring) spraying was carried out in the middle of May 2002 under variable weather conditions. The late (summer) spraying was done in the beginning of June under fairly stable weather conditions. Glycine was added to the spray solution in order to estimate the spray deposition onto leaves sampled immediately after the actual spraying. This measure was used to convert nominal spray dosages to spray solution deposited on the leaves.

Sampling

Year one

Leaves, buds, flowers, green berries and mature berries were sampled from trees subjected to the early spraying. Sampling from trees subjected to the late spraying included leaves, green and mature berries. For all samples except flowers, the number and weight of the single end-point were measured. Flowers were only counted.

The fall-down of buds, flowers, early and mature fruits were sampled in buckets placed under the experimental trees in order to measure the variability over the season, and to determine the life stage most susceptible to loss due to herbivory and other causes.

Second year

All end-points were sampled in the second year, i.e. leaves, buds, flowers green berries and mature berries.

In the year after herbicide treatment, untreated trees in a single hedgerow were randomly selected and repeatedly sprayed with the insecticide cypermethrin in order to estimate the effect of insect herbivory on the fruit set. This was done by counting the number of buds, flowers and berries on selected shoots. The amount of berries was also counted as present within an area of 35 x 35 cm².

Results

Pesticide effects in the year of exposure:

Significant effects on the number and weight of leaves were observed as a result of the early spraying. They were, however, not biologically significant. Furthermore, they were only significant when all hedgerows were pooled in the analyses. There were no significant effects of the summer spraying on the leaves.
The number of flowers was significantly affected as a consequence of the early spraying. The reduction was approximately 10% at the highest dosage. We assess this to be biologically insignificant.
The number and weight of berries was significantly reduced both in the case of an early and a late spraying. The early spraying was most effective showing a reduction of between 40 and 99% at levels realistic for spray drift from field spraying with the maximum label rate, i.e. a spray deposition equal to 7% of the label rate. The amount of berries was, at the highest dosage, reduced by more than 99%. The susceptibility of hawthorn was not as high for the late spraying as for the early spraying (Figure 1).

Pesticide effects the year after exposure:

Metsulfuron causes a reduction in growth (leaves) and the fruit set (buds, flowers and berries) the year after herbicide exposure.
The effects on leaves, buds and flowers were statistically significant in the year after spraying, and larger than in the year of spraying. However, with effects ranging between 25 and 50% there is still a considerable amount of leaves, buds and flowers left, especially if the unexposed parts of the trees are included.
The year after spraying the effect on the berries was of the same order as in the first year after the summer spraying, but lower than after the early spraying (Figure 1).

Figure 1. Comparison of percentage effects on the amount of berries in the year of exposure and the year after. The figure is based on the number of berries on side shoots the first year and the number counted in 0.1225m² frames in the second year

Biological observations:

The amount of fruit increases with height, at least on the south/western side of the hedgerow, nevertheless a large proportion of the fruit production occurs in the lower three meters, where spray drift is most likely
The fall-down of reproductive units is largest shortly after flowering when buds are attacked by insects, flowers are aborted due to insufficient pollination, and fruits attacked by insects fall to the ground
Insects reduced a successful fruit set by approximately 40% in hedgerow number five.

Relevance of the data in connection with pesticide regulation

The pesticide regulation does not, at present, account for any of the effects observed or made probable in the present study. This applies for the circumstance that perennial plants can be affected longer than the year of exposure. Furthermore, the standard test methods used for ecotoxicity testing do not measure effects on reproduction. The results suggest that effects on reproduction and tests that involve perennial species are recommended to consider ecosystem protection.