Aug 17, 2009

Research Projects

Send

Research into the impact of Bt maize (Cry 3Bb1) on non-target organisms living in the soil

(2005 – 2008) Biological Research Centre for Agriculture and Forestry (BBA) (since 2008 Julius Kühn Institute (JKI)), Institute for Plant Protection in Field Crops and Grassland; Braunschweig

Topic

In Bt maize that is resistant to the Western corn rootworm (Diabrotica v. virgifera) the Bt toxin (Cry 3Bb1) is produced primarily in the roots, since this is where the Western corn rootworm attacks the maize. The aim of this project was to investigate the effects of the Cry3Bb1 Bt protein on soil-dwelling insects, particularly those living in the rhizosphere.

One focus of the project was on flies and midges (two-winged flies, dipterans), in particular sciarid flies, which are directly involved in the decomposition of plant material in the soil.

The other focus was on predatory beetles and their larvae (coleopterans) that feed on the primary decomposers (e.g. dipteran larvae, springtails, mites, earthworms). Since the Western corn rootworm, which the Cry3Bb1 toxin is designed to combat, is also a beetle, it is suspected that predatory beetles, particularly their larvae that hunt in the soil, will also be affected by ingesting prey containing Bt.

Summary

Cry3Bb1 Bt protein in maize plant litter does not affect the development of sciarid fly larvae in tested amounts of up to 16.45 micrograms per gram. MON88017 Bt maize litter does not take longer to decompose in the field than maize litter from non-Bt varieties. Neither is the diversity of two-winged flies (dipterans) affected.

Sciarid fly larvae and pupae ingest the Cry3Bb1 protein (up to 263.92 ng/g detected), their predators (rove beetle larvae) ingest up to 27.7 nanograms per gram. The Cry3Bb1 protein is passed on up the food chain.

A significant reduction in offspring production was observed in the rove beetle Atheta coriaria after it had consumed sciarid fly larvae that had been reared on MON88017 Bt maize litter. However, the maize variety did not affect the grazing rates or lifespan of this beetle species.

The total number of newly emerged ground beetles found in the Bt maize variant was between 50 and 60 per cent of that found in the non-Bt maize variants.

A significantly higher incidence of frit flies, particularly Oscinella frit, and gall midges that specialise in aphids (Aphidoletes aphidimyza) were recorded in the Bt maize. This suggests that Bt maize benefits secondary plant pests (frit flies and aphids) – although further investigations are needed to validate this observation.

Experiment description

Feeding sciarid fly larvae with plant material from the field trial (see below)

Larvae from one species of sciarid fly (Lycoriella castanescens) were fed on litter (BBCH80), roots (BBCH19 and 51) and pollen from the four maize varieties grown in the field trial (transgenic, isogenic and two conventional varieties). Their pupation and hatching rates, mortality, larval development time and pupal dormancy were compared.

Feeding predatory beetles with prey (predation experiments)

In the laboratory, the dominant ground beetle on the maize field (Calathus fuscipes), the larvae of another ground beetle (Poecilus cupreus) and the L3 larvae of a rove beetle (Atheta coriaria) were fed on sciarid fly larvae (L. castanescens) that had eaten plant parts (e.g. root pieces, leaf or stem pieces) from the four different maize varieties. For four days they were monitored every day to see how many of the sciarid fly larvae had been eaten.

Parameters like survival rate, feeding activity, length of time to pupation and to the emergence of the adult insects, pupation/hatching rate, weight gain/loss and behavioural changes were measured.

A long-term experiment was also conducted in which rove beetles (A. coriaria) were fed daily for a six-month period with a sciarid fly larva (L. castanescens) that had itself been fed on maize roots from the various Bt and non-Bt maize varieties.

Toxin analysis

In addition, the ELISA method was used to investigate the toxin levels of the transgenic plant parts used in the feeding experiments, the sciarid fly larvae and predatory beetles.

The Cry3Bb1 concentrations in the insects caught in the trail plots were also measured.

Measuring decomposition activity

Mini containers in the field: Four mini container cylinders (see diagram) were placed in each of the 32 trial plots and removed one by one at monthly intervals. Before being installed, each container was primed with 0.2 grams of maize litter. The container cylinders were placed in the soil horizontally at a depth of approximately five centimetres. The decomposition of the litter was calculated by measuring the dry weight as a percentage of the starting weight.

Mini containers under standardised conditions: In this experiment, the decomposition activity was calculated specifically with regard to sciarid fly larvae. Soil samples were taken from each of the trial plots using a sampling frame (diameter 16 cm; sampling depth 20 cm) and heated to remove any living organisms. In the greenhouse, mini containers containing maize litter from the different variants were placed in the soil samples. Six were placed on the ground and six at a depth of five centimetres below the surface. Around 50 sciarid fly larvae were placed in each soil sample. One mini container was removed each month and the decomposition activity is determined by measuring the dry weight of the plant material.

Field surveys of dipterans and predatory beetles

Ground photoeclectors: To calculate the density, hatching rate and species spectrum of the dipterans and predatory beetles under field cultivation conditions, ground photoeclectors were placed on the plots. These trap creatures emerging from the soil and moving towards the light. They were used to determine whether there were any differences between Bt and non-Bt maize varieties.

The eclectors were placed on the plots at the end of June 2005. They were emptied once a fortnight from May to October and moved once a month. Between the harvest and the sowing of the next crop (November to April) the eclectors were emptied every four weeks and are not moved.

The dipterans were identified down to family level, the adult beetles generally down to species level.

Live traps: The live traps were used to trap ground beetles for laboratory feeding experiments. They consist of trapping containers buried with their tops level with the surface of the soil. The bottom of the container was covered with expanded clay pellets. The live traps were set up at the start of the growing season or shortly before the feeding experiments were due to start.

Pit traps: These record the activity of invertebrates in the soil. Two underground pit traps were lowered into the ground on each plot in the field trial. The traps were installed after the maize plants have emerged close to the plants or in the plant row in order to investigate arthropod activity in the root zone. The traps were monitored fortnightly during the growing period and monthly during the rest of the year.

Soil sample extraction: Soil sample extraction can be used to estimate the density of larval stages of predatory beetles living in the soil. Every six weeks a sampling spear (diameter 16cm) was used to take two soil samples (0-5 cm and 5-10 cm depth) from each plot. In the laboratory a gentle extraction process (lasting two to three weeks) was used to flush the soil creatures out and identify them.

Results

Feeding experiments under laboratory conditions

Feeding sciarid fly larvae with plant material from the field trial: Feeding sciarid fly larvae with different plant parts from Bt and non-Bt maize varieties did not reduce their pupation or hatching rates. The duration of pupal dormancy was virtually identical in all maize variants.

Pupal mortality of larvae fed on the Bt maize litter was somewhat higher – though not statistically significant – than with those fed on the isogenic maize litter. There were no significant differences in the development time of the decomposer larvae in any of the four test series.

The numbers of pupae and adult sciarid flies did not differ significantly when fed with different varieties of maize pollen (Bt maize with Bt concentration of 4.04 μg/g). The larval development time and pupal dormancy were also statistically equivalent.

Feeding predatory beetles with prey (predation experiments): The dominant ground beetle on the maize field (Calathus fuscipes) and the larvae of Poecilus cupreus showed significantly lower acceptance of prey larvae raised on Bt maize litter compared with sciarid fly larvae raised in isogenic or conventional maize litter. This effect was primarily a result of the lower acceptance of the prey fed on Bt material in the first two days of the trial. Acceptance rose after this. The reasons for this phenomenon have not yet been ascertained.

With the larvae of another rove beetle (Atheta coriaria), the average consumption of Bt-fed fly larvae was significantly lower than for other maize variants. In addition, with these larvae, the larval development time to pupation was longer. It must be noted though, that these findings became significant only in comparison with the isogenic maize variety and one of two conventional varieties. Varietal effects cannot therefore be excluded.

Whilst prey capture (feeding rates) and lifespan of female predatory rove beetles in the Bt and non-Bt maize variants did not differ, the fertility of females (offspring production) in the Bt maize group was approximately 50 per cent lower than in the other three groups. The majority of offspring in the Bt maize group were produced at the start of the trial. Offspring production subsequently declined very rapidly. By contrast, in the non-Bt maize variants, offspring production slowly increased until the end of the trial.

The results indicate a reduction in the fertility of Atheta-coriaria females after ingesting prey contaminated with Bt toxin.

Toxin analysis: Toxin analysis: Assuming a Cry3Bb1 toxin content in pollen (404.0 ng/g) and in roots (1645 ng/g), approximately ten per cent (= 42.4 ng/g, based on pollen) and approx. 16 per cent (= 263.9 ng/g, based on roots) of the toxin content was detected in the decomposer larvae after they had been fed for a period of six days. Bt toxin was also detected in the pupae after feeding on Bt pollen (78.8 ng/g). Pupae that had eaten Bt maize roots were found to contain 39.5 nanograms of Cry3Bb1 per gram. This indicates that the toxin is not completely eliminated in the faeces. 0.33 nanograms of Cry3Bb1 were detected in the faeces of 60 sciarid fly larvae. This means that predatory soil-dwelling organisms like rove beetles which eat the dipteran larvae and pupae ingest Cry3Bb1 as well (larvae: 27.7 ng/g; imagines: 12.2 ng/g). The Cry3Bb1 toxin is therefore passed on up the food chain.

Measuring decomposition activity

Mini containers in the field: Mini containers in the field: In the first winter (2005/2006) the maize litter in the Bt maize plots was 70 per cent decomposed after four months. The decomposition rate of one of the conventional varieties was always significantly lower than that of the other varieties.

In winter 2006/07 there were no significant differences in the decomposition of the maize litter across all varieties. In November 2006 the Bt maize litter contained on average 7.918 micrograms of Cry3Bb1 per gram of litter.

Mini containers under standardised conditions: In all maize variants, the decomposition rate rose steeply in the first month and then levelled off. After four and a half months, the litter had been about 70% decomposed by the sciarid fly larvae. Bt maize litter is decomposed by sciarid fly larvae to the same extent as litter from the isogenic and conventional varieties.

Ground photoeclectors: In the third growing season an average of 1632 dipterans hatched per square meter. This was significantly more than in 2006 (1078 insects/m²) and 2005 (535 insects/m²). Evaluation showed that the differences in hatching rates overall and for individual dipteran families were not statistically significant. In the winter months the number of hatched dipterans was at least twice as high as in the growing season.

Dipteran families, such as frit flies (Chloropidae), some species of which are plant pests, were recorded in significantly higher abundances in the MON88017 Bt maize, particularly Oscinella frit.

Significantly higher hatching rates of the gall midge Aphidoletes aphidimyza were also recorded in the Bt maize compared with the non-Bt maize variants. The larvae of the species feed exclusively on aphids and are used in greenhouses as a form of biological pest control. The results indicate that Bt maize may possibly benefit secondary pests such as aphids and frit flies. However, further investigations in this area are needed.

No differences between the maize variants were found in terms of species numbers, diversity or evenness.

Over 30 groups of predatory beetles were assessed, but only seven groups were suitable for statistical analysis of hatching rates. Significant differences were found between the different maize varieties for individual ground beetle groups, but on average there were no significant differences between the Bt maize and conventional varieties.

Pit traps: Only the numbers of ground beetles (Carabidae) were demonstrably affected by the maize varieties. They were significantly less abundant in one of the conventional varieties and in the Bt maize compared with the other two varieties (including the isogenic variety). Of the soil-dwelling larvae, only around half as many ground beetles hatched in the Bt maize variant compared with the other variants in the third trial year.

Soil sample extraction: The number of beetle larvae extracted from the soil samples was too low for statistical evaluation. Higher densities of springtails, annelids and millipedes were extracted from the soil samples, but these were not investigated in this project.