Kurzfassung

Currently used risk assessment schemes targeting genetically modified organisms (GM) organisms (GMO) are primarily developed for GM plants and their potential impact in terrestrial ecosystems. However, the transportation of GM plant material and related transgenic products into aquatic ecosystems via leaching from GM plant material, from root exudates or soil erosion in combination with their implications in aquatic life has only been addressed in isolated cases. As a consequence, there are...Currently used risk assessment schemes targeting genetically modified organisms (GM) organisms (GMO) are primarily developed for GM plants and their potential impact in terrestrial ecosystems. However, the transportation of GM plant material and related transgenic products into aquatic ecosystems via leaching from GM plant material, from root exudates or soil erosion in combination with their implications in aquatic life has only been addressed in isolated cases. As a consequence, there are no coherent or agreed risk assessment concepts for aquatic ecosystems. The aim of the project was to develop a selection procedure allowing for a priorization of ecological entities by using Bt-maize as a model. Thereby, the procedure should reflect the variety of potential ecological effects by keeping it at the same time practicably feasible in the context of the environmental risk assessment.

As the first part, we developed in the present project a priorization procedure for the stream types in Germany according to their potential to receive crop plant residues, while maize served as a model crop. Using geographic information systems and geostatistical procedures, we blended the length of each stream type – defined by the EU Water Framework Directive – with the maize cultivation density. The resulting exposure factor served as starting point for the calculation of exposure related indices and was supplemented by stream type inherent factors such as their sensitivity. Following this developed numerical, transparent and thus reproducible selection matrix, the final priorization of stream types was realised. Further criteria for the verification of the stream types have been the exposition of potentially sensitive sections of these stream types as well as the representation of all ecoregions relevant for Germany. Finally, eleven stream types have been prioritized by this selection procedure.

In the second part, based on a selection procedure for non-target-species in terrestrial systems, a procedure for aquatic systems was developed, supporting the identification of ecologically relevant non-target-species in aquatic systems that are potentially at risk to be adversely affected by GM plants or its transgene products. The tools that were developed for the operationalization of the selection procedure in terrestrial systems were also modified and adapted to the conditions of aquatic systems. These new, adapted procedure and its tools were tested in an expert workshop.

The modified selection procedure for nontarget testing organisms is implemented through several matrices. The initial step are species list that are characteristic for the selected aquatic system (identified in Part 1 of the project). During the workshop, this basic information compiled from published literature was checked for completeness and updated where needed, meaning, species were eliminated, replaced or added. During a second step the temporal coincidence of the GM plant material and the nontarget species as well as the species ecological functions within the stream ecosystem have to be verified. These data were subsequently amended by information on the species’ potential protection status and the feasibility to use each of the species during laboratory testing. Following this step-by-step approach, the number of potential aquatic test species has been substantially reduced to numbers that may actually be feasible for an empirical study. The presented ecological function focussed and transparent procedure was unanimously judged as a promising approach in the context of GMO risk assessment for aquatic systems.

In this second part of the project, also the most important existing ecotoxicological standard test guidelines for aquatic systems were evaluated regarding their suitability for risk assessment of GM crop plants. By assuming that microbially produced Bt-toxins have the same ecotoxicological potential as Bt-toxins produced within the GM crop, all test guidelines seem to reflect an exposure via the water phase and the sediment. The exposure via food, although a vital and the potentially most important exposure pathway for aquatic non-target species, is not formally considered in most of the test guidelines but can be included after some rather minor modifications. Considering, however, the high selectivity at least of some of the toxins to caused adverse effects, this calls for a new orientation of the ecotoxicological testing strategy, while the developed selection procedure for non-target-species provides one potential solution. Furthermore, a secondary exposure via the food chain, which would mainly be relevant for predators, is not easily implementable in the current guidelines and would involve substantial development efforts. This, however, seems questionable due to ethical considerations.

Within the third project part, shredders from different orders were exposed towards GM maize plant residues showing adverse effects relative to and its isogenic control at variable levels: Gammarus, Hyalella, Asellus, for instance, showed a modified food (=leaf) consumption, while the three tested trichopteran species (i.e., Sericostoma spp., Chaetopteryx spp., Halesus spp.) had a lower lipid content if fed with Bt-maize relative to conventional maize. The latter observation, however, could not be verified during semi-field stream microcosm experiments. This deviation in the effects between the laboratory and semi-field approach may be explained by the potential availability of alternative food sources of higher nutritious quality (such as periphyton) during the stream microcosm experiments. This in turn suggests that GMO mediated effect under field relevant conditions are – especially if alternative food sources are available – less pronounces relative to test systems established under laboratory conditions.

Finally, the degradation of Bt-toxins (Cry1Ab) in aged GM corn plant material was evaluated over 56 days in aquatic systems uncovering that the median Bt-toxin concentration remained nearly constant over the first 14 days. After 28 days, however, its concentration dropped substantially just to rise again by a factor of two after another four weeks. In accordance with these Bt-toxin concentrations, the toxicity of the maize plant residues for the European corn borer was reduced with increasing aging time until day 28 but increased again after 56 days of submergence in the experimental systems prior to testing. These data are thus challenging the general assumption that Bt-toxins are linearly degraded in GM plant material over time. At the same time the toxins seem to conserve their biological activity over at least 56 days.

In summary, the following key-insights have been generated over the course of the present research and developmental project:
- A transparent priorization procedure for the identification of potentially highly exposed stream types was developed
- Current ecotoxicological standard test guidelines are not unconditionally suitable for their use in the risk assessment of GMO
- A transparent procedure for the identification of potential test species is introduced
- GM crop plant residues may induce negative effects in aquatic invertebrates under laboratory conditions, but these effects seem to be much less pronounced under field relevant conditions
- GM maize plant material contains even after an aging duration of 56 days detectable levels of Bt-toxins, which remain biologically active.
» weiterlesen» einklappen