Uptake of Trace Elements on Suspended Matter
in the Kalix River

Johan Ingri Division of Applied Geology, Luleå University of Technology, S-971 87 Luleå, Sweden

Johan.Ingri@sb.luth.se

Anders Widerlund Division of Applied Geology, Luleå University of Technology, S-971 87 Luleå, Sweden

Introduction

The Kalix River is a relatively large river (catchment area 23 600 km2, average water discharge 300 m3s-1) situated within the boreal zone, in the northern part of Sweden. The water discharge is almost constant at approximately 50 m3s-1 during winter baseflow (December-April), when the river is ice-covered. In mid May most of the snowpack melts in the lower parts of the catchment and the discharge increases to a maximum around 1500 m3s-1. Although the river has its source in the area among Swedens highest mountains the suspended load is low compared with "average river water" suspended load. Suspended particulate matter (>0.45 µm) rarely exceeds 20 mgl-1 at the river mouth. The physical weathering is low and chemical precipitates and organic matter strongly influence the composition of the suspended phase.

Methods

The suspended phase was sampled once a week (twice a week during high discharge) during 18 months. The samples were filtered in-situ, using battery operated peristaltic pumps and large (142 mm) membrane filters. The trace element samples were dissolved in nitric acid in a teflon bomb in a microwave oven, and analysed with ICP-MS. Major elements were determined by ICP-AES on ashed samples fused with lithium metaborate and dissolved in nitric acid.

Results

The average total concentration of suspended matter (TSM) during the period was 1.5 mgl-1 and approximately 50% of TSM was organic matter. The average ashed suspended load (ASL, fraction larger than 0.45 µm) had the following major element composition; SiO2 32.0%, Al2O3 6.5%, Fe2O3 45.1%, MnO 1.1%, MgO 1.8%, CaO 4.5%, Na2O 2.2%, K2O 2.3%, TiO2 0.3%, P2O5 1.2% (sum of oxides 97.0). The composition of ASL is strongly influenced by authigenic Fe and Mn phases. In winter, up to 70% of the ASL consists of Fe2O3 and during summer Mn-rich particles form in the river (Pontér et al., 1992) increasing the MnO concentration up to 2.2%. At high water discharge, in spring, most of the major elements (except Fe and P) are hosted in detrital particles (physically weathered rock-forming minerals) but in summer and winter between 40% and 80% of the major element concentrations are in a non-detrital form. The uptake of major elements from the dissolved phase can be fully explained by non-detrital iron (Ingri and Widerlund, 1994).

The average concentrations of Ba, Co, Cu, La, Mo, Rb and Sr in the ashed suspended phase, corresponding average fraction in non-detrital particles and discharge weighted suspended concentrations (based on total concentration) were 543 ppm (42% non-detrital) (0.9 µgl-1), 27 ppm (78% non-detrital) (0.04 µgl-1), 66 ppm (76%) (0.09 µgl-1), 52 ppm (73%) (0.07 µgl-1), 10 ppm (95%) (0.01 µgl-1), 58 ppm (47%) (0.09 µgl-1) and 178 ppm (42%) (0.26 µgl-1) respectively. The total transport of these trace elements was dominated by the dissolved phase, although a significant non-detrital enrichment was found for each element. Of the total load approximately 12% of Ba, 41% of Co, 15% of Cu, 35% of La, 4% of Mo, 8% of Rb and 2% of Sr were transported in the fraction larger than 0.45 µm. For Al, Fe and Mn the corresponding numbers were 77%, 43% and 48% respectively. The relatively low fraction transported in the suspended phase is related to the very low detrital fraction in the suspended load compared with "average river water". The KD values (total concentration in suspended phase, mg/kg, divided by the dissolved concentration mg/l) changes with season and discharge, but the average coefficients are Ba 6.8x104, Co 4.5x105, Cu 2.4x105, La 4.7x105, Mo 2.9x104, Rb 5.3x104 and Sr 1.0x104 respectively.

Barium was scavenged by both Fe and Mn, Co by Mn, Cu, La, Mo, Rb and Sr by Fe. All elements showed a linear uptake on the Fe and/or Mn-rich phases, more than 80% of the uptake could be explained by these elements. Hence, although approximately 50% of the TSM was organic matter, variables like the total organic C concentration or the suspended C/N ratio did not influence the regression analysis much. It is not known to what extent the organic matter was associated with the Fe-Mn rich particles.

References

Ingri, J. & Widerlund, A., Geochim. Cosmochim. Acta 58, 5433-5442 (1994).

Pontér, C., Ingri, J. & Boström, K., Geochim. Cosmochim. Acta 56, 1485-1494 (1992).