|
4. Moving from the lab to the landscape: Extending the role of ecology in risk assessment (TU4/5) Landscape related community classification of macroinvertebrates in the region of Braunschweig, Germany. Hommen, Udo1, Schaefers, Christoph1, Biggs, Jeremy2, Liess, Matthias3, Ohe, Peter3, Ball, Maria4, Holmes, Chris4, Gonzalez-Valero, Juan5, Hendly, Paul5, Maund, Steve6, 1 Fraunhofer-Institute for Molecular Biology and Applied Ecology, Schmallenberg, Germany2 The Ponds Conservation Trust, Oxford, UK3 Centre for Environmental Research, Leipzig-Halle, Germany4 Waterborne Environmental Inc., Leesburg, VA, USA5 Syngenta Crop Protection Inc., Greensboro, NC, USA6 Syngenta Crop Protection Inc., Basel, Switzerland ABSTRACT- In order to attribute differences in macroinvertebrate communities to agrochemical exposure, the ecological influence factors have to be accounted for. Macroinvertebrate communities of small streams, ditches and ponds in the area of Braunschweig, Lower Saxony, Germany, were analysed for the possibility to differentiate groups of communities related to topography, land use of surrounding area, ecosystem morphology, water and sediment parameters. Topographical and landscape related parameters were available via a GIS system using different digitalized maps of the region. Three macroinvertebrate data sets of different quantity and quality were available: For 30 ponds abundance data for macroinvertebrates as well as many environmental variables were collected once (Summer 2001 or 2002). 66 locations at small streams and ditches had been sampled intensively in the last years ('High density data'), including abundance data as well as water chemistry and morphological data. The third data set consists of abundance classes of macroinvertebrates for 880 locations, sampled over different years ('Low density data'). Due to the different sampling intensity of the sites, the different years and seasons of the samples, first classification by TWINSPAN revealed clear differences between the three data sets and the following analyses were restricted to the pond data set and subsets of the high and low density data sets with respect to sampling seasons, years and sampling area. Locations with samples for at least one sample available for spring, summer and autumn were used to classify the macroinvertebrate communities. Environmental factors were related to the communities by the means of discriminant analysis (applied to the TWINSPAN groups) and ordination techniques. Key words: Ecological monitoring, Macroinvertebrates, Community classification, Ordination |
