Geochemical mapping is a technique that has been used all over the world for a multitude of purposes. The biogeochemical mapping programme conducted by the Geological Survey of Sweden (SGU) is mainly based on stream plant chemistry, which, opposed to mapping based on inorganic stream sediments or soils, is a method that is very uncommon. The stream plants are roots of Carex species, roots of Filipendula ulmaria, and whole plant samples of Fontinalis antipyretica. Samples are collected at an average density of one sample per 7 km2. In urban areas, where access to naturally growing stream plants often is difficult or impossible, transplants of F. antipyretica are used. This thesis aims to increase understanding of some fundamental properties of the biogeochemical data set, and to describe some applications. The focus of the applications described is related to environmental issues, but these can to some degree be extended towards mineral exploration as well. Three important properties studied and described are 1) analytical quality of the data, 2) effects of expressing contents on a dry weight basis instead of as residuals, and 3) species dependent responses. It is shown that the analytical quality for most elements used are quite good, at least for the parameters further studied in the thesis. Species dependent responses do exist, but only for a few of the parameters is the influence strong enough to cause problems during interpretation. The most pronounced effect is a significant enrichment of Mn and Co in F. antipyretica as compared to the other species. Some effects can also be seen on Ba and Zn, but not to the same extent. Incidentally, the problem is somewhat less pronounced if the residual method of expressing concentrations is used. This residual method is based on stepwise regression, where Mn is one of the independent variables. For all other elements, expressing contents relative to dry weight could not be determined to cause any negative effects. As the dry weight basis is very commonly used for expressing concentrations, for most elements it can and should be used instead of the residuals historically used. One problem, that has proved to be difficult to solve, is the presence of variable amounts of minerogenic matter in the samples. However, as there are excellent correlations between several of the parameters (e g Ti, Cr, Al, Si, Zr) and loss on ignition, the other parameters can be used to aid interpretation. Chromium is one example, where unusually high Cr:Ti ratios occur at several locations with metallurgical or other industrial activities. Active acid sulphate (AS) soils is another example of anthropogenic contamination of the environment. The AS soils are quite common in areas below the highest shoreline in Finland and Sweden. It is however quite difficult to determine exactly where they are located. The thesis demonstrates that biogeochemical data provides information that can be used to localize such areas, where the stream water regularly is acidic and metal rich. Here again, a ratio (Y:Pb or Ni:Pb) can be used in order to further amplify the signature. Yttrium and Ni have been shown to occur in elevated levels in stream water affected by active AS soils, whereas Pb-concentrations even tend not to increase, or even decrease. In a selected drainage basin, it is shown that transplants of F. antipyretica responds fairly well to the stream water typical for AS soils. The F. antipyretica transplants as well as the other biological media studied, stair step moss (Hylocomium splendens), also contribute to demonstrating that the urban areas are geochemically complex, there is no straightforward relationship to the other sampling media studied. As in the other examples demonstrated above, baseline information from till and other soils provide a valuable key to understanding the biogeochemical data set.
Luleå: Luleå tekniska universitet, 2009. , 29 p.
Godkänd; 2009; 20091117 (mivu); DISPUTATION Ämnesområde: Tillämpad geologi/Applied Geology Opponent: Professor Reijo Salminen, Geological Survey of Finland Ordförande: Professor Björn Öhlander, Luleå tekniska universitet Tid: Fredag den 18 december 2009, kl 10.00 Plats: F 531, Luleå tekniska universitet