Dynamics of silica and phytoplankton population under altered conditions of river flow in the Daugava River, Latvia

Abstract

INTRODUCTION Nitrogen and phosphorus are recognized as major nutrients because their availability governs overall algal growth (e.g. Schindler, 1977; Hecky and Kilham, 1988). At the same time, it has been discussed previously that a number of other elements might be limiting in certain ecosystems or for some organism groups. One of such elements is dissolved silica, whose availability can regulate phytoplankton species composition (Egge and Aksnes, 1992). Dissolved silica in aquatic environments originates from the weathering and breakdown of silica-containing minerals and mostly is carried to marine environments by rivers (Treguer et al., 1995). There are substantial variations in the delivery of dissolved silica from continents to the ocean due to different lithology of drainage areas, continental weathering intensity, climatic variations, and diatom production (Conley, 1997). In addition, it has been shown that anthropogenic factors, e.g. hydrological alteration of rivers as a result of construction of dams, can significantly reduce loads of dissolved silica to the sea (Garnier et al., 1999; Humborg et al., 2006), with adverse effects on the marine ecosystem (Conley et al., 1993; Humborg et al., 2000). The actual mechanism of a decrease of dissolved silica is still debated. For example, Humborg et al. (2000) argue that land-sea fluxes are smaller in the regulated than non-regulated boreal river systems as a result of the lower weathering flux of silica. It has been shown that major reservoirs built on boreal rivers can hold 30% to 70% of their annual water discharge (Dynesius and Nilsson, 1994), which can significantly decrease dissolved Si concentrations in rivers by providing preconditions for enhanced diatom growth and sedimentation of diatom frustules, and subsequent burial in sediments behind dams (Conley et al., 2000). Furthermore, the ability of small reservoirs with a short residence time of water to act as diatom traps and so affect silica land-sea flux has been discussed. For example Friedl et al. (2004) argued that the residence time characteristic for small reservoirs of lowland rivers is not sufficient for the development of massive diatom blooms, so suggesting that some other explanation for the observed silica concentration decrease in lowland rivers is needed. At the same time it was argued that dams built on lowland rivers of the south-eastern Baltic with shorter residence times than those on boreal rivers can negatively affect silica land-sea fluxes (Humborg et al., 2006). Furthermore, it was demonstrated that even a relatively small reservoir with a short water residence time slows the river flow and that low flow conditions are essential for the formation of diatom blooms in rivers (e.g. Kiss and Genkal, 1993; Mitrovic et al., 2008). Nevertheless, in our opinion there are too few direct studies on the impact of river flow alteration on land-sea silica fluxes due to alterations in diatom ecology. Therefore, we attempted to test the impact of an altered river flow on dissolved silica and the diatom population during the productive season along the heavily dammed Daugava River. We compared river areas unaffected and affected by damming. We also performed analysis of carbon, nitrogen, and silica in sediments of two reservoirs of different size to test the assumption of the dependence of the accumulation of silica of diatom origin on reservoir size. MATERIALS AND METHODS Study site The Daugava River is 1005 km long with a catchment area covering 87 900 [km.sup.2]. The catchment is composed of agricultural lands (50%), forest (40%), and lakes (2%) (Humborg et al., 2006). Consolidated sedimentary rocks dominate in the catchment, and the mean slope of the river is only 0.3%. The river flow velocity is highly variable: from 0.1 to more than 2 m [s.sup.-1] (Deksne et al., 2009). The long-term average runoff of the Daugava River is 21.1 [km.sup.3] [y.sup. …