Assessment of phosphorus fractions in streams draining different land use and development of new monitoring method

Abstract

This study was an integrated part of the interdisciplinary research project EUTROPIA, funded by Research Council of Norway (190028/S30). The goal of the nature science part of the EUTROPIA project is to improve the understanding of hydro-biogeochemical processes governing the P-mobility to the MORSA catchment, with the Western Vansjø Lake as a main study area. This study has examined different phosphorus fractions from the nine tributaries to the lake draining the different contingency of agriculture and forested land. In addition, this work has also examined the novel monitoring method Diffusive Gradients in Thin Films (DGT) for simultaneous time averaged sampling of bioavailable free-PO4 (orthophosphate) and organic P compounds. The levels of three to four different phosphorus (P) fractions were determined in the stream water samples collected from nine inlet streams to the Western Vansjø Lake over a period of several months. DGT samplers deployed at three of the nine monitoring streams to study the performance of this new sampling method and DGT samples were analysed for total-P and free orthophosphate. Temporal and spatial variations in the stream water chemistry, with emphasis of P fractions, were assessed in order to distinguish the effect of environmental pressures, such as climate change, that may counteract the mitigations measures. The nine studied streams draining from the Western Vansjø catchment were selected in order to capture gradient in land use from agriculture to increasing proportion of forested area. The concentration and composition of P fractions were greatly influenced by land use distribution and runoff. The organic bound P constituted from 30 to 60% of the total P transport (~ 6-12 µg P/L) in the runoff from the forested area, due to high transport of DOC. In the agricultural areas the concentration of total-P were mainly constituted by particle bound P. Results from the four hydrological years (2006-2010) showed that autumn rainstorms, winter mild-spells and spring snowmelt were important periods of P transfer from land to water. Inter-annual variation in winter temperature is found to be important factor that may govern the variation in inter-annual flux of P. A higher flux of total-P in the streams was found during the years 2006/07 and 2007/08 mainly due to higher winter temperatures than normal causing snowmelt, accompanied with higher precipitation, generating several runoff episodes during the winter season. During the hydrological years 2008/09 and 2009/10, the main peak of total-P release occurred during the normal spring snowmelt period. Elevated concentrations of free-PO4 and dissolved organic bound P (DOM-P) were found during the episodes of flood and snow smelt. Increased runoff and milder winter temperature may lead to increased leaching of free-PO4 bound to the soils through Fe (III) from water logged soils due to reduction to Fe (II) and trapping as FeS. This may be a governing factor for the strong negative effect of the flood in the year 2000, still 11 years after flood. Furthermore, the drainage pipes in agriculture soils in the Western Vansjø catchments are installed in apatite containing clay, which are eroded by flow to and through tiles. This is found to be an important governing factor for the total-P loading and a significant flow-path of total-P flux to the streams. The DGT samplers with two adsorbents were studied, the new Metsorb (Titanium dioxide) adsorbent and the more established Ferrihydrite (Iron hydroxide) adsorbent. Laboratory studies of the two DGT samplers showed time linear uptake of the low molecular weight organic P compounds e.g. AMP and IP6 (Adenosine Monophosphate, Inositol 6-Phosphate). DGTs based diffusion-coefficients for AMP was ranged between 3 to 3.5 E-06 cm2s-1 which is close to the theoretical value (3.9E-06) while the value of 1.5E-06 for IP6 was lower than the theoretical value. The elution efficiencies for ferrihydrite DGT and Metsorb DGT were found to be 100% and 85%, respectively. Metsorb DGT method performs similarly to ferrihydrite DGT method for orthophosphate ions, but collects organic P compound at ~ 25% higher efficiency than ferrihydrite DGT method. The concentrations of free-PO4 in the stream water sample are found to be lower when obtained by the DGT samplers than when determined in grab samples. This is probably due to that the conventional field sampling method for free-PO4 includes P in small colloids and surface adsorbed P, while DGT samplers exclude these fractions. Moreover, low molecular weight organic fractions collected by the DGT samplers showed large variation from agriculture to forested site. DGTs thus offers a completely new sampling method for this fraction of P in water which often have been missing in transport budgets, and thus add considerable new value to such studies.