aspabir.biogeochemistry.ro © 2012-2014

The tools for EIA and SEA developed would include an aspect of cross-scale effects, i.e. would go beyond the local scale (the scale of the project in itself) towards the scale of the landscape (applying the biogeochemical concept of hazard assessment, Jianu et al. 2012), and would take into consideration services produced by the biodiversity located both within inside and outside protected areas. The SEA concept will allow an explicit assessment of the consequences of land cover/habitat/vegetation types change as a result of management plans and policies on the biogeochemical services supported by the landscape. The integration between SEA and EIA will be made possible by a tool linking plants SPU distribution in space with the large scale plant objects distribution via soil and hydro-geomorphological variables.

University of Bucharest will be represented by three research centers: one specialised in biogeochemistry and ecological services (CESEC, www.cesec.ro, project coordination), one in mineralogy and geochemistry (Lythos, directly involved in objective 1), and one in hydrogeology, modeling and GIS databases (Hydro, directly involved in objective 3). An institute of biology (IBB) is responsible for the plant ecology part, an institute of applied mathematics (ISMMA) is responsible for the mathematical part of modelling and for software development, and an institute of soil science (ICPA) is responsible for the soil analyses and setting up method for the fast mapping of trace elements distribution. A national authority on biodiversity is the end-users of the results who already expressed its interest in the project.

 We operationally consider the SPUs of grasses as the populations of grasses in a 50x50 m grid cell of land. The discretization unit for erosion and sediment transport model (DU) is a 100 x 100 m grid cell of land. The large scale vegetation object (LSVO) are land cover objects, or habitats (as in Habitat Directive), or plant associations units from vegetation maps.

Our approach will start firstly from the real SPUs as identified in the field to the DUs for erosion and sediment transport models, and from the LSVOs to the same discretization units. In the first case one needs mostly to upscale the information on SPU using a model of the heterogeneity inside a DU. In the second case one needs downscaling models to characterize the heterogeneity of plant SPUs inside large scale vegetation objects using remote sensing images analyses. In the floodplain we will evaluate indicators of habitat quality for mammals and birds as related to the structure of the vegetation cover (citation) by comparison with indicators of parameters relevant for the retention of elements in the floodplain.

In order to model the impact of metals in soil on plant development and parameters relevant for soil erosion control and metals immobilization vegetation mapping in the sites will be done with quantitative methods (ten 2 m2 quadrates by 10 m x 10 m plot) and will lead to the identification of the distribution of dominant species in 50x50 m plots in terms of plant cover. The sampling in the 2x2 m plots randomized after the stratification of the plots will be done two times a growing season during two years and will lead to abundances in terms of aboveground plant biomass. Belowground parts will be sampled only for 3 dominant species for investigating bioaccumulation. A field-XRF (Niton) will be used for mapping in situ the metals and corrected maps will be generated using regression equations between the concentration determined in the field and concentrations determined in the lab by ICP-MS on a sub-set of samples. XRF measurement will be done also in the 2x2 plots. At all points of XRF measurement soil will be sampled for lab analyses of total and extractable N and P, pH, EC. At the points located in the 2x2 plots extra measured parameters will be soil granulometry, sequential extractions, total metals, K, P (by ICP-AES and ICP-MS) and loss on ignition (LOI). Plant samples (above and belowground parts) will be analysed for total N and P, total metals, and K.

For mineralogy and soil magnetic properties used for building the up-scaling models soil and alluvial sediments sampling will be implemented by UB-Lythos in the same 50x50 m with a resolution of 10x10 m, and in a randomly selected 10 x 10m sub-site with a resolution at 2x2 m. An extra sampling with 10x10 m resolution will be done in the surrounding surface included in the 100 x 100 m site. Also an extra grid sampling of the rest of core sites and extended sites at resolution 10x10 m will be performed. At the moment of sampling field XRF measurement will be made. The magnetic study will be carried out on the collected samples (mass magnetic susceptibility and thermal and alternating field demagnetizations analyses). The data will feed the information system and for the core site (50x50) will be interpreted in correlation with the vegetation cover of the core and extended site, and for the rest of the site will be used for up-scaling models to DU scale. The statistical relationships developed based on the data characterizing the core sites will be checked for reproducibility with the data characterizing the extended sites.

Models for the use of lab XRF and field XRF in the mapping of metals in Romanian contaminated soils at multiple scales will be built using the sample base of ICPA (who is in charge with the soil monitoring in Romania). From 20 Romanian areas we will select at least 50 samples and will analyze them by field and lab XRF. Statistical correlations will be performed with metals concentrations already analyzed by classical methods.

For the nested characterisation of soil parameters at the scale of the DU of catchment erosion models in the upland and in the floodplain UB will sample quadrats of 50x50 m distributed in the Ampoi catchements, in Ampoi floodplain and in Arieș floodplain. The samples will be analyzed by ICPA for metals, total N and P, pH, loss on ignition, orgamic matter and granulometry. For the characterization of soil parameters corresponding to the large scale vegetation objects UB will sample soil from UD located in 10 Corine land cover objects corresponding to grasslands, and ICPA will perform the analyses as above. Soil sampling will be correlated with plant SPUs characterization and sampling by IBB, and the analyses of plant samples analyzed at UB-CESEC. The parameters of the SPUs will be directly related to the vegetation parameters used as input for describing the role of vegetation in erosion models (e.g. contact cover, surface roughness, soil penetration depth by roots etc), and with those relevant for the immobilization of metals in biomass. The UD sampled in the top-down approach will cover the internal heterogeneity of grassland vegetation cover of each LSVO as characterized by the processing of Landsat images and field visits for calibration (IBB partner).

The information system will be organized in Arc-Info. Beside usual interpolation methods and spatial analysis at the investigated scales, experimental and field data sets will be used for the production of bioaccumulation models. In order to set up the nonlinear correlation models, the software packages MATLAB and MARS will be used. The models developed based on the experimental data will be done by code programming (EDN - ESRI development network). The optimisation of the scientific computing code of CASESAR-TRACER will be done for some parts of it for parallel computers in order to improve its speed. It will target both multicore CPUs and general computing capable graphic cards (Nvidia Fermi architecture), and will use the OpenCL industrial standard implemented on both CPUs and graphic cards (GPUs). For the verification of the model by distribution patterns of isotopic ratios of Pb we will analyse selected soil and sediment samples from the floodplain  and potential sources of Pb in the catchments (the tailing pond, contaminated soils, background soil, etc)  using multi-collector and mono-collector (for Pb) ICP-MS. This will allow the major sources identification and the result will be compared with those from the simulation obtained using the new sediment transport model.

The models resulted from DU scale studies integrated in GIS platform with the sediment transport model will represent a tool for improving the EIA concept and the SEA as cumulative impact assessment and will allow a direct evaluation of cross-scale effects of the plant SPU services of erosion control and metals immobilization. The information about the internal heterogeneity of grassland dominated UDs in the LSVO with respect to the erosion control and metals immobilization will be used for constructing a tool for use in SEA of plans involving landscape structure changes.

References

Jianu Denisa, Iordache Virgil, Barbara Soara, Lucian Petrescu, Aurora Neagoe, Cezar Iacob, Răzvan Orza, 2012, The Role of Mineralogy and Geochemistry in Hazard Potential Assessment of Mining Areas, in Kothe E, Varma A (Eds) Biogeointeractions in contaminated soils, Springer, http://www.cesec.ro/pdf/Jianu_etal_2012.pdf

Download Gant diagram of the project here: […]