Send

Effects of the cultivation of Diabrotica-resistant Bt maize on the maize ecosystem II

(2005 – 2008) RWTH Aachen University, Institute of Environmental Research (Biology V), Chair of Environmental Biology and Chemodynamics

Topic

The aim of this project is to investigate the effect of Bt maize that is resistant to the Western corn rootworm on various organisms in the maize ecosystem, in comparison with conventional maize varieties.

The focus in this part of the research is on effects of the Bt protein (Cry3Bb1) on arthropods in the herbaceous layer such as beetles, heteropterans and cicadas and their antagonists, such as spiders.

Further investigations will focus on the potential effect of Bt maize pollen on caterpillars that live on the associated farmland plants. The aim is to determine, among other things, the probability of caterpillars ingesting Bt pollen that falls on their host plants.

Summary

Arthropods of the herbaceous layer and their antagonists

Variety-related differences in the incidence of arthropods were found both in the maize ears and in the herbaceous layer. No Bt-specific differences were found. Fluctuations were attributed to differences in soil and weather conditions.

Z. scutellaris was the most common species of leafhopper found (65 per cent). It would therefore make a good indicator species for identifying potential effects.

Females of the mirid bug species T. caelestialium captured in Bt maize contained more Bt protein on average than the males. No mirid bugs containing Bt protein were found in the conventional maize varieties. Additional ELISA tests showed that the Bt protein in the mirid bugs had been excreted or broken down in the digestive system after just three days.

Investigating pollen feeders and arthropods in inflorescences

No varietal or Bt-specific differences were found. However, there were clear differences between the crop years. Thrips in particular fluctuated widely in terms of their occurrence and average numbers.

Exposure of caterpillars to Bt maize pollen in the agricultural landscape

For three of the butterfly and moth species mapped, over 70 per cent of the population was located in the vicinity of the maize fields. For the peacock butterfly, the figure was 100 per cent in two of the research areas.

Experiment description

A total of 32 plots were arranged on the trial field. Eight plots are planted with Bt maize (MON88017), eight with an isogenic line and eight each with two conventional varieties. The various varieties are distributed randomly over the plots but respecting two conditions: no adjacent plots with the same variety and four plots of the same variety at the edge of the field/in the middle. Bt maize is grown on the same plots in all trial years so that any accumulation effects in the soil can be identified.

The arthropods of the herbaceous layer are caught in nets.

Insects from the male flowers are shaken out into the net.

Yield from two inflorescences in August 2007: anthers, pollen and only a few arthropods.

Effects on arthropods of the herbaceous layer and their antagonists

The incidence of herbivores and their antagonists in the Bt maize compared with conventional maize varieties is measured using a range of different methods (ear samples, sticky traps, net catches). A particular focus is placed on cicadas, thrips and heteropterans and on beneficial organisms such as ladybirds and lacewings.

In order to estimate the density of parasites and parasitoids specialising in cicadas, in 2005 one yellow pan per plot was placed in the immediate vicinity of the sticky traps.

ELISA

The ELISA method is used to investigate the level of Cry3Bb1 in mirids (T. caelestialium). This makes it possible to draw conclusions about the exposure of herbivores and predators to Cry3Bb1 in the field. It is then possible to identify exposed species and determine their significance for future monitoring programmes.

Experiments were also conducted to investigate the breakdown of Cry3Bb1 in the mirid bug species T. caelestialium. Mirid bugs were captured in the maize field and kept on Bt maize in the greenhouse for a set period of time. The bugs were then either kept on isogenic maize or starved for a few days. The Bt level was then measured using the ELISA method.

Investigating pollen feeders and arthropods in inflorescences

Pollen feeders and arthropods in the inflorescences are identified using shake samples. Five male inflorescences per plot are shaken five times on two occasions during the maize flowering season (320 samples, a total of 1600 inflorescences per year).

Exposure of caterpillars to Bt maize pollen in the agricultural landscape

To determine the probability of caterpillars (e.g. small tortoiseshell, peacock, nutmeg moth) coming into contact with Bt maize pollen, both the caterpillars and their host plants, such as stinging nettles, goose-foot and giant hogweed, are to be mapped.

The caterpillar species were identified by regular shaking/examination of the host plants.

The landscape use and population data are evaluated using GIS (Geographical Information System) programs.

Since May 2007, feeding experiments in the laboratory have been investigating the effect of the Bt protein on peacock butterfly caterpillars. The caterpillars are fed leaves from their food plant that are covered with a defined quantity of Bt maize pollen. The caterpillars’ feeding and development and mortality rates are observed. Caterpillars given food covered in isogenic maize pollen are used as a control group.

Results

Arthropods of the herbaceous layer and their antagonists

Insects in the maize cobs: There were very significant differences in the numbers of arthropods in the maize cobs of the different conventional varieties. By contrast, no differences have been found to date between the transgenic and isogenic varieties.

In 2005 the most common organisms found were aphids. Thrips were much less common. The main predators found in the maize cobs were a type of heteropteran (Orius). The most common type of beetle was the Cortinicara gibbosa, a type of mould beetle. Other types of insect were found only rarely.

In 2006 there were significantly lower numbers of aphids, but higher densities of thrips than in the previous year. A possible cause of these differences are the very different weather conditions in the two trial years. There were no notable differences in the numbers of the heteropteran Orius or the mould beetle C. gibbosa between the trial years.

Density of adults and nymphs (winged larval stage) of the most common mirid (T. caelestialium) in the different maize varieties at the second sampling date (week 33/34).

Arthropods in the herbaceous layer: The species composition was largely identical in all three years. For instance, 8125 leafhoppers of 18 species were captured in 2006. Z. scutellaris was the most common species of leafhopper found (65 per cent).

There were significant differences between the conventional varieties, but not between the Bt maize and the isogenic variety. For instance, the mirid bug T. caelestialium was much more common in one of the two conventional varieties than in the other. The differences can be attributed to varietal differences. Differences in soil conditions were taken into account in the statistical evaluation.

ELISA

The mirid T. caelestialium from the Bt maize is exposed to the Bt protein throughout its development. The females contained more transgenic protein (average 16.85 nanograms of Cry3Bb1) than the males (average 3.13 nanograms). In terms of grams, the values were up to one-sixth of the amount found in fresh leaf material. No mirids with the Bt protein were found in the non-transgenic maize varieties.

Additional ELISA tests showed that the Bt protein in the mirid bugs had been excreted or broken down in the digestive system after just three days.

Investigating pollen feeders and arthropods in inflorescences

Not all the samples from 2006 have been fully processed yet. The most common arthropods so far are anthocorids of the genus Orius, mirids (Miridae; T. caelestialium), damsel bugs (Nabidae: Nabis pseudoferus) and an unusually high density of thrips (Thysanoptera).

As with the cob samples, the differences in the density of the most common arthropods from one year to another can probably be attributed to differences in weather conditions over recent years.

No varietal or Bt-specific differences were found.

Exposure of caterpillars to Bt maize pollen in the agricultural landscape

The distribution of host plant/butterfly pairs was mapped in three agricultural landscapes with different levels of use. The areas were between 150 and 230 hectares in area and contained between eight and ten maize fields, each measuring between 0.2 and 15 hectares, depending on the level of farming intensity.

The distribution and frequency of the host plants varied widely between the three areas. The evaluation of the data is not yet complete.

The distribution of the butterfly species was also very different in the three areas. In 2005 no peacock butterflies were found on the stinging nettle clumps and only one nest of small tortoiseshell butterflies. The goose-foot host plant (Chenopodium album) was usually found only in isolated incidences and contained no nutmeg moths. For this reason, in 2006 the study was widened to cover the following host plant/butterfly pairs: giant hogweed (Heracleum sphondylium) with several Eupithecia moth species and the silver Y moths (Autographa gamma, A. confusa).

Percentage of butterfly populations found at a distance of 10m or less from the edge of the maize field in the three study areas (totals for 2006 and 2007).

In order to assess the extent of the threat posed to the butterfly species, it was calculated what percentage of the butterfly populations found were located in the vicinity (10 metres or less) of maize fields during the trial years. For three of the butterfly and moth species mapped, over 70 per cent of the population/individuals was in the vicinity of the maize fields. For the peacock butterfly, the figure was 100 per cent in two of the research areas.

Feeding trials

Because of the high losses, which were attributed to the susceptibility of the larvae, and the number of larvae available, it was not possible to record sufficient data for a statistical evaluation. The feeding trial method will therefore need to be refined for future trials.