Earth-Science Reviews 91 (2008) 77–92 Contents lists available at ScienceDirect Earth-Science Reviews j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e a r s c i r ev Past occurrences of hypoxia in the Baltic Sea and the role of climate variability, environmental change and human impact Lovisa Zillén a,⁎, Daniel J. Conley a, Thomas Andrén b, Elinor Andrén b, Svante Björck a a b Geobiosphere Science Centre, Department of Quaternary Geology, Lund University, Sölvegatan 12, SE-223 62 Lund, Sweden School of Life Sciences, Södertörn University College, SE-141 89 Huddinge, Sweden a r t i c l e a b s t r a c t i n f o The hypoxic zone in the Baltic Sea has increased in area about four times since 1960 and widespread oxygen deficiency has severely reduced macro benthic communities below the halocline in the Baltic Proper and the Gulf of Finland, which in turn has affected food chain dynamics, fish habitats and fisheries in the entire Baltic Sea. The cause of increased hypoxia is believed to be enhanced eutrophication through increased anthropogenic input of nutrients, such as nitrogen and phosphorus. However, the spatial variability of hypoxia on long time-scales is poorly known: and so are the driving mechanisms. We review the occurrence of hypoxia in modern time (last c. 50 years), modern historical time (AD 1950–1800) and during the more distant past (the last c. 10 000 years) and explore the role of climate variability, environmental change and human impact. We present a compilation of proxy records of hypoxia (laminated sediments) based on long sediment cores from the Baltic Sea. The cumulated results show that the deeper depressions of the Baltic Sea have experienced intermittent hypoxia during most of the Holocene and that regular laminations started to form c. 8500–7800 cal. yr BP ago, in association with the formation of a permanent halocline at the transition between the Early Littorina Sea and the Littorina Sea s. str. Laminated sediments were deposited during three main periods (i.e. between c. 8000–4000, 2000–800 cal. yr BP and subsequent to AD 1800) which overlap the Holocene Thermal Maximum (c. 9000–5000 cal. yr BP), the Medieval Warm Period (c. AD 750–1200) and the modern historical period (AD 1800 to present) and coincide with intervals of high surface salinity (at least during the Littorina s. str.) and high total organic carbon content. This study implies that there may be a correlation between climate variability in the past and the state of the marine environment, where milder and dryer periods with less freshwater run-off correspond to increased salinities and higher accumulation of organic carbon resulting in amplified hypoxia and enlarged distribution of laminated sediments. We suggest that hydrology changes in the drainage area on long time-scales have, as well as the inflow of saltier North Sea waters, controlled the deep oxic conditions in the Baltic Sea and that such changes have followed the general Holocene climate development in Northwest Europe. Increased hypoxia during the Medieval Warm Period also correlates with large-scale changes in land use that occurred in much of the Baltic Sea watershed during the early-medieval expansion. We suggest that hypoxia during this period in the Baltic Sea was not only caused by climate, but increased human impact was most likely an additional trigger. Large areas of the Baltic Sea have experienced intermittent hypoxic from at least AD 1900 with laminated sediments present in the Gotland Basin in the Baltic Proper since then and up to present time. This period coincides with the industrial revolution in Northwestern Europe which started around AD 1850, when population grew, cutting of drainage ditches intensified, and agricultural and forest industry expanded extensively. © 2008 Elsevier B.V. All rights reserved. Article history: Received 7 December 2007 Accepted 13 October 2008 Available online 11 November 2008 Keywords: Baltic Sea Littorina Sea hypoxia Holocene Medieval Warm Period modern historical period climate change and human impact Contents 1. 2. 3. 4. Introduction . . . . . . . . . . Baltic Sea characteristics . . . . History of the Baltic Sea . . . . Evidence for hypoxia in the past 4.1. Laminated sediments . . 4.2. Geochemistry . . . . . . 4.3. Chronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ⁎ Corresponding author. Tel.: +46 46 222 7805; fax: +46 46 222 4830. E-mail address: lovisa.zillen@geol.lu.se (L. Zillén). 0012-8252/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.earscirev.2008.10.001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 79 80 81 81 81 81 78 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 5. Occurrence of hypoxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. During modern time (last 50 years) . . . . . . . . . . . . . . . . . . . . 5.2. During modern historical time (AD 1950–AD 1800) . . . . . . . . . . . . 5.3. On geological time-scales (AD 1800–10 000 cal. yr BP) . . . . . . . . . . . 5.3.1. Bothnian Sea, Bothnian Bay, Gulf of Finland and Archipelago Sea . . 5.3.2. Gotland Basin . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3. NW and NC Baltic Proper . . . . . . . . . . . . . . . . . . . . 5.3.4. Landsort Deep and W Gotland Basin . . . . . . . . . . . . . . . 5.3.5. Southern Baltic Sea . . . . . . . . . . . . . . . . . . . . . . . 6. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Hypoxia in time and space . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1. Bothnian Bay, Bothnian Sea, Gulf of Finland and the Archipelago Sea 6.1.2. Baltic Proper . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Hypoxia and driving mechanisms . . . . . . . . . . . . . . . . . . . . . 7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction Hypoxia, defined as b2 mg/l dissolved oxygen, occurs in aquatic environments when dissolved oxygen becomes reduced in concentration to a point harmful to aquatic organisms living in the environment. Hypoxia is a globally significant problem with over 400 reported sites suffering from its effects (Diaz and Rosenberg, 2008). Hypoxia not only causes severe ecosystem disturbances (Diaz and Rosenberg, 1995) but alters nutrient biogeochemical cycles (Vahtera et al., 2007) and forms hydrogen sulfide which is hazardous to numerous fauna and flora communities (Diaz and Rosenberg, 1995). Widespread oxygen deficiency has for that reason severely reduced macrobenthic communities below the halocline in the Baltic Sea over the past decades (Laine, 2003) and produced benthic “ecological deserts” that annually cover over 30% of the seafloor (Karlson et al., 2002). Hypoxia also affects food chain dynamics, fish habitats and fisheries (Karlson et al., 2002; Bonsdorff, 2006). The hypoxic zone in the Baltic Sea has increased about four times since the 1960s (Jonsson et al., 1990), and currently covers an area averaging 41000 km2 annually (e.g. Conley et al., 2002). Hypoxia has for that reason developed into a serve environmental problem for the Baltic Sea and its dependents. The increasing trend in hypoxia is thought to be caused by enhanced eutrophication due to excess load of waterborne and airborne nutrients (nitrogen and phosphorus) to the sea from anthropogenic sources (Wulff et al., 2007). Eutrophication has been of great concern for the countries in the Baltic region at least since the 1980s, with ministerial level commitments to reduce nutrients and improve water quality (Johansson et al., 2007). The debate about improving the present state of the Baltic Sea through implementation of the European Water Framework Directive often refers to conditions prior to the turn of the last century (1900) as an environmental reference status, when the Baltic is suggested to have been an oligothrophic clear-water body with oxygenated deep waters (Österblom et al., 2007). However, geological records show that the Baltic Sea is a dynamic ecosystem that has undergone many environmental changes over the last c. 16 000 years (Björck, 1995; Andrén et al., 2000a,b; Berglund et al., 2005). Studies of the more recent past, for instance, reveal that hypoxia has been present in some basins for at least the last 100 years (Jonsson et al., 1990). Furthermore, analyses of long sediment cores suggest that hypoxia in the Baltic Sea has occurred intermittently in deep basins in the Baltic Proper over thousands of years (Andrén et al., 2000a,b; Sohlenius et al., 2001; Emeis et al., 2003) and that cyanobacterial blooms have occurred during the last c. 7000 years (Bianchi et al., 2000; Kunzendorf et al., 2001; Poutanen and Nikkilä, 2001). Various investigations also imply that there may be a correlation between climate variability in the past . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 81 82 83 84 84 84 85 85 85 85 85 85 88 89 90 90 and the state of the marine environment, where warmer periods correspond to increased primary production and higher salinities resulting in amplified hypoxia and enlarged distribution of benthic mortality and laminated sediments (e.g. Thorsen et al., 1995; Fjellså and Nordberg, 1996; Andrén et al., 2000a; Bianchi et al., 2000; Nordberg et al., 2000; Emeis et al., 2003). In addition, the occurrence of hypoxia in the deeper basins today need not necessarily be attributed to human activity but could be naturally driven by oceanographic, environmental and climate forcing (Filipsson and Nordberg, 2004a,b). In addition, both climate variability and human impact have the potential to greatly affect the environment in the semi-enclosed Baltic Sea and its catchment. However, the long-term spatial and temporal extent of hypoxia and its possible connections to these parameters are poorly known. Anthropogenic forcing in the drainage area, such as, changes in land use and population density, could indirectly have affected the marine/brackish environment already in the Late Holocene. It is known from numerous long-term studies of lake sediments in Northwest Europe that population growth and agricultural development have impacted lakes for thousands of years and that cultural eutrophication of lakes has a history longer than just decades or centuries (e.g. Fritz, 1989; Renberg et al., 2001; Bradshaw et al., 2005). Furthermore, the International Panel on Climate Change (IPCC) has recognized that hypoxia is a problem of growing concern with projected climate change (www.ipcc.ch) and recent studies predict that cyanobacteria blooms will magnify with global warming (Pearl and Huisman, 2008). It is thus essential to improve our understanding about the timing, extent and mechanism(s) causing hypoxia on millennial time-scales in order to understand the full range of the natural variability and to put forward realistic measures to improve the future environment of the Baltic Sea. It is also important to put the recent human impact in a time-perspective in order to understand the modern environmental issues. This paper aims to review and synthesize the current knowledge in the Baltic Sea about the appearance of hypoxia in modern time (last 50 years), modern historical time (50–200 years ago) and in the geological past (last c. 10 000 years), based on previous publications. We examined a large number of papers about Baltic Sea sediments, but only present here papers which report or address hypoxia. We present a compilation of several long sediment records covering most of the Holocene (i.e. the last c. 10 000 years) and explore possible connections with the presence of past hypoxia and the role of climate and environmental variability and human impact. With the purpose to spatially and temporally reconstruct the occurrence of hypoxia our main objective is to answer three basic questions: where, when and why was the Baltic Sea hypoxic in the past? L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 2. Baltic Sea characteristics The Baltic Sea (area of 412 560 km2; volume of 21631 km3; average depth of 52 m; maximum depth of 459 m; Seifert and Kayser, 1995) is a semi-enclosed brackish water body consisting of a series of basins where the Bothnian Bay, the Bothnian Sea, the Gulfs of Finland and Riga and the Baltic Proper are the main water masses (Fig. 1). We will not address hypoxia in the transition zone between the Baltic Sea and the Skagerrak, nor in coastal areas. The modern record of hypoxia in the Danish straits has been previously reported by Conley et al. (2007). The Baltic Sea watershed has a population over 85 million. The drainage area is c. 4 times larger than the area of the sea itself, which yields substantial freshwater input to the basin. Inflow of denser saline bottom waters takes place through shallow and narrow connections (i.e. the Danish belts and Swedish sound) from the Skagerrak/Kattegat and brackish surface water is transported out over the same sills. Inflow of saline water in the south and the occurrence of large rivers in the north yield a south-north salinity gradient where the surface salinity is c. 8–10 in the southern Baltic, c. 7–8 in the Baltic Proper 79 and c. 3–5 in the Gulf of Finland, Bothnian Sea and Bothnian Bay (Matthäus, 2006). The water column of the Baltic Proper is permanently stratified, consisting of two water masses, an upper layer of brackish water with salinities of c. 7–8 and more saline deep waters of c. 11–13. At the transition zone between these water masses a strong permanent halocline is formed which prevents vertical mixing of the water column and ventilation of more oxygen rich waters to the bottom (Matthäus and Schinke, 1999). The halocline is formed at depths varying between c. 30 m in the Arkona Basin, 60 m in the Bornholm Basin, c. 80 m in the Gotland Basin and Landsort Deep (Matthäus, 2006). The Bothnian Sea and, in particular, the Bothnian Bay have weaker haloclines and a better ventilation due to less variable inflow conditions (Stigebrandt, 2001). In combination with a lower productivity this makes these basins mostly oxic. Large inflows (100–250 km3) of higher salinity and oxygen-rich water from the North Sea represent an important mechanism by which the Baltic deep water is displaced and renewed to a significant degree. All inflows are substantially mixed with surface water, on Fig. 1. Map of the Baltic Sea showing the major basins in black (BB = Bothnian Bay, BS = Bothnian Sea, GF = Gulf of Finland, GR = Gulf of Riga, BP = Baltic Proper, WB = West Baltic, K = Kattegat and S = Skagerrak) and sub-basins in white (AS = Archipelago Sea, NWBP = Northwest Baltic Proper, NCBP = North Central Baltic Proper, LD = Landsort Deep, GB = Gotland Basin, BB = Bornholm Basin, AB = Arkona Basin, MB = Mecklenburgian Bay, and GD = Gdansk Deep) mentioned in the text and coring sites for the sediment records studied in this paper. Black and white circles represent localities containing laminated sediments (geological records of hypoxia) while white circles show the position of sediment cores with homogeneous (oxygenated and bioturbated) sediments. Also, the locations of the lake sediment sequences referred to in the text (black circles). 80 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 average 3 times (Stigebrandt and Gustafsson, 2003). However, frequent but small inflows (10–20 km3) are generally insufficient to displace the deep bottom water because they have a much less density and therefore do not penetrate into the deepest parts of the Baltic Sea (Stigebrandt, 2001; Matthäus, 2006). In summary, the physical setting of the Baltic Sea, the narrow and shallow connections to the Skagerrak/Kattegat and relative large river runoff influence the salinity and oxygen distribution in the basin resulting in low salinity, long residence time, strong density stratification of the water mass and limited water exchange (Wulff et al., 1990; Gustafsson and Westman, 2002). This estuarine circulation retains nutrients and organic matter within the basin and leads to high nutrient availability. 3. History of the Baltic Sea The present Baltic Sea formed in the course of several events that accompanied the onset of the last deglaciation of Scandinavia at c. 17 000–15 000 cal. yr. BP (e.g. Björck, 1995). The postglacial history of the Baltic Sea and its drainage area is characterized by regionally confined transgressions and regressions caused by the interaction of the differential subsidence and uplift of land, deglaciation, ponding and drainage of fresh-water stages and phases of eustatic sea-level rise. By tradition, the late Quaternary history of the Baltic Sea is divided into four main stages, i.e. the Baltic Ice Lake, the Yoldia Sea, the Ancylus Lake and the Littorina Sea, representing time periods with either brackish condition with connections to the Skagerrak/Kattegat or isolated and up-dammed freshwater lake conditions (Björck, 1995). The Littorina Sea can further be divided into three sub stages i.e. the Early (or Initial) Littorina Sea (c. 10 000–8500 cal. yr BP) the Littorina Sea sensu stricto (s. str.; c. 8500–3000 cal. yr BP) and Late Littorina Sea (c. 3000 cal. yr BP to present; Berglund et al., 2005). We will apply the latter terminology for the Littorina Sea, which is based on studies of near shore sediment cores, because it is supported by high-resolution chronologies established by 14C-dating of terrestrial plant macrofossils and environmental reconstructions constructed from multistratigraphical analyses (Berglund and Sandgren, 1996; Berglund et al., 2005; Yu et al., 2007). The Baltic Ice Lake began to form c. 16 000 cal. yr BP and was filled by the accumulation of glacial waters from the waning ice sheet. It existed until c. 11600 cal. yr BP when the dam broke open at Mount Billingen in south central Sweden (Björck and Digerfeldt, 1984; Strömberg, 1992; Andrén et al., 2002). This resulted in a considerable fall of the Baltic lake-level (c. 25 m; Svensson, 1991; Björck, 1995). The deep basins of the freshwater Baltic Ice Lake were characterized by oxic conditions (due to seasonal overturn of the water column) and low organic productivity (Andrén et al., 2002). During this period glacial varved and non-varved clays were deposited. After the final drainage of the Baltic Ice Lake the Yoldia Sea formed (11600–10 700 cal. yr BP) and its level was determined by that of the global ocean. The water of the Yoldia Sea was partly brackish for 200– 300 yr around 11200 cal. yr BP (Wastegård et al., 1995; Andrén et al., 2002) and linked to Lake Vänern through a broad connection over the central part of southern Sweden and to Skagerrak/Kattegat through narrow fjords west of Lake Vänern. The brackish phase of the Yoldia Sea stage had its highest salinities in depression areas between Lake Vänern and Stockholm in central Sweden (Schoning, 2001). However, it has been suggested that brackish bottom-waters also penetrated as far south as the Bornholm Basin and Hanö Bay in southern Baltic (Björck et al., 1990; Andrén et al., 2000b; Andrén et al., 2007). The dominant sedimentary deposits from the Yoldia Sea stage consist of organic-poor (silty) clays, often varved in and north of the Gotland Basin (Andrén et al., 2002). Through the continuous land-uplift, the fjord connections to the Skagerrak/Kattegat shallowed and forced Baltic level to rise above sea level (Björck, 1995). The Baltic changed once more to lake conditions, the Ancylus Lake. During the Ancylus Lake stage (c. 10 700–10 000 cal. yr BP) the relatively faster land uplift in the northern parts of the basin caused it to be tilted towards the south and at c. 10 000 cal. yr. BP the Ancylus Lake attained the level of the global ocean (which was also rising) through a fluvial-lacustrine system in the lowlands of the Fehmarn-Great Belt region (Björck, 2008). This connection to Kattegatt was subsequently broadened and later the Öresund Strait began to function as an important inlet of saltwater around c. 8500– 7800 cal. yr BP (e.g. Andrén et al., 2000a,b; Sohlenius et al., 2001; Emeis et al., 2003). However, due to the lack of a robust geochronology, the timing of the first clear brackish/marine Littorina intrusion is much debated. In the southern Baltic it has been dated to c. 6500 cal. BP with the Optically Stimulated Luminescence (OSL) dating method (Moros et al., 2002; Rößler, 2006; Kortekaas, 2007) and it is presumed that a transitional phase (i.e. the Early Littorina Sea Stage or the Mastogloia Sea) existed with episodic marine influxes starting at c. 9800 cal. yr BP based on traditional 14C-dating (Andrén et al., 2000b; Yu et al., 2003; Berglund et al., 2005; Witkowski et al., 2005). The saltwater intrusion during the Early Littorina Sea/Littorina Sea s. str. transition (often referred to as the transition phase between the Ancylus Lake and the Littorina Sea or the Mastogloia Sea) initiated the establishment of a permanent halocline in the Baltic Sea and a more marine flora and fauna (Andrén et al., 2000a; Sohlenius et al., 2001). During this time, the inlet transect area in the Öresund Strait was about twice its size compared to today (Gustafsson and Westman, 2002) and the surface water of the Baltic was more saline, i.e. 10–15 in the Baltic Proper, than at present, i.e. 7–8, (Gustafsson and Westman, 2002; Emeis et al., 2003; Berglund et al., 2005). However a more conservative estimate, c. 4 units higher then present, is given by Donner et al. (1999) for the Gulf of Finland and Gulf of Bothnia based on the stable isotope composition in bivalve shells. Furthermore, surface salinity values of c. 7.3–10.3 and 4.8–10.3 (at present c. 3–5) were reported by Widerlund and Andersson, (2006) between 6770 and 3070 cal. yr BP in the Bothnian Sea and Bothnian Bay, respectively. The transition is also associated with an abrupt increase in organic matter content and a clear change from clay to clay-gyttja (often laminated) in the sediment record (Fig. 2). Due to the continued melting of the North American and Antarctic ice sheets the sea-level rise continued until c. 6500–6000 cal. yr BP. Based on studies of marginal lake basins along the coast of Blekinge, southeastern Sweden, Yu et al. (2007) demonstrate that the Baltic Sea level rose steadily from 8600 to 6500 cal. yr BP, with a rapid sea-level rise at c. 7600 cal. yr BP. After this stage the land-uplift was greater than the almost stable global sea-level and a more continuous regression Fig. 2. Photo of sediments from core 211660-1 recovered from a depth of 241 m in the Gotland Deep showing laminated Littorina Sea s. str. clay-gyttja. L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 started. The connections via Öresund and the Danish Straits became shallower, salinity decreased and present conditions in the Baltic Sea gradually developed. 4. Evidence for hypoxia in the past 4.1. Laminated sediments One of the environmental prerequisites for the formation and preservation of laminated sediments in aquatic environments is, apart from cyclic sedimentation, the absence of a relatively large benthic fauna that bioturbate (vertically mix) the uppermost sediments (e.g. Zillén et al., 2003). Laminated sediments, therefore, tend to be found in relatively deep basins where vertical mixing of the water column is limited and hypoxic conditions prevail, which restrict the presence of bottom fauna (Zillén et al., 2003). Laminated sediments form in sedimentation basins of the open Baltic Sea below the permanent halocline, where the vertical mixing of the water mass is weak (Stigebrandt, 2001; Matthäus, 2006). The composition of individual laminae in Littorina Sea sediments1 (in the deep basins of the Baltic Sea is either depositional or authigenic in origin (Neumann et al., 1997; Burk et al., 2002). Depositional laminae couplets (or sometimes triplets and quadruplets) can occur and consist of alternating lithogenic (clay-rich) and diatomaceous (diatom-rich) mud. The diatom component forms mainly during the spring bloom, while the lithogenic sediment input dominates in autumn and winter when biogenic production is minimal and storms redistribute sediment to the deeper basins. Distinctly laminated authigenic Ca-rich rhodochrosite (Mn,Ca)CO3 laminae occur regularly throughout the Littorina Sea sediments of the Gotland Deep (Neumann et al., 1997; Burk et al., 2002; Burke and Kemp, 2004). The precipitation of Mn-carbonates takes place close to the sediment–water interface and is a result of changing redoxconditions due to episodic oxic water inflows to the generally hypoxic basin. For a more detailed description about the formation and composition of laminations in Baltic Sea sediments see e.g. Sohlenius et al. (1996), Neumann et al. (1997), Sternbeck and Sohlenius (1997), Sohlenius and Westman (1998) and Burk et al. (2002). Given that the majority of relatively large benthic faunas cannot live under hypoxic conditions, the occurrence of laminae in sediment records is a good indicator of past bottom water hypoxia in aquatic environments. However, the preservation of laminations can be affected by other post-depositional disturbances, such as erosion by water movements created by currents (Larsen et al., 1998). Sediment records displaying alternating laminated and homogeneous sequences could also be subjected to bioturbation at the transition zone. When bottom environments change from hypoxic to more oxic conditions benthic fauna communities are reestablished, which can homogenize the upper layers of the earlier deposited laminated interval. Therefore, the age approximation of the boundary between laminated and homogeneous sediments (i.e. the change from hypoxic to more oxic conditions) is a maximum estimate. In this review, hypoxic conditions will be ascribed to the formation and preservation of laminations in sediment records. 4.2. Geochemistry The effect of redox conditions on the distribution of trace elements (e.g. Mn, Mo, U, V, Cu and Zn) has been discussed in a number of studies (e.g. Sohlenius and Westman, 1998; Sohlenius et al., 2001). It has been demonstrated that trace elements accumulate in sediments deposited under reducing conditions and several of these have been 1 From now on and throughout this paper the term Littorina Sea sediments refers to sediment deposited after the Early Littorina Sea/Littorina Sea s. str. transition (i.e. after c. 8500-8000 cal. yr BP based on traditional 14C-dating) if not stated otherwise. 81 reported to be enriched in laminated sediments of Littorina Sea age in the Baltic Sea (Sternbeck et al., 2000). The presence of iron sulfides (e.g. pyrite and greigite) in sediments is also thought to reflect reduced or hypoxic bottom conditions (Snowball, 1993). However, the primary distribution of both trace elements and Fe-sulfides may be altered by diagenetic changes occurring after sediment deposition (Snowball and Thompson, 1990). Studies show that changes in redox conditions can lead to sulfide diffusion and pyritization of earlier deposited sediments (Thompson et al., 1995) and that the primary signals can be altered by diagenetic changes occurring several thousands of years after sediment deposition (Sternbeck et al., 2000). There are also a number of different mechanisms by which trace elements may be enriched in sediments. Thus, the absence or presence of trace elements in sediments should be used with caution when interpreting ancient sedimentary environments. Therefore, we have only used the presence of laminae in sediments as a proxy for hypoxic/anoxic conditions since it has been shown to be the most prominent/ undisputable evidence of oxygen depletion (Sohlenius et al., 1996). Studies of recent marine sediments have not shown any clear-cut and systematic relationships between bottom water oxygen concentrations and the sediment organic carbon content (Burdige, 2007). While enhanced preservation of organic material may occur due to changes in macrofaunal abundance or changes in reducing equivalents, oxygen per se has no demonstrated effect on carbon preservation or remineralization (Cowie et al., 1995). 4.3. Chronology Good chronological control is necessary to understand past environmental and climate changes (Zillén et al., 2003). On millennial time-scales, sediments from the Baltic sub-basins mostly rely on timescales based on measurements of the 14C content in organic material. However, this dating method can be biased by unidentified marine/ brackish 14C reservoir effects (Köningsson and Possnert, 1988), redeposition of older carbon from shallower areas to deep accumulation bottoms, and uncertainties arising from inferred ages derived from interpolation between individually dated levels. The sediments in the Baltic Sea are characterized by low organic carbon contents (usually b10%) and deficiency in macrofossil remains (especially in the deeper basins), which further makes it a delicate task to construct an accurate chronology based on 14C-dating. In the majority of studies dating is performed on bulk sediment samples, but should if possible be performed on macrofossils, preferably terrestrial, since bulk sediment samples are known to show too old ages, and poor temporal resolution (at best c. 5–10 14C-dates in sediment sequences covering the last c. 10 000 years) often resulting in great uncertainties in the estimation of sedimentation rates. The 14C-ages presented here are calibrated (1σ) by using the Oxcal calibration program v4.0 (Bronk Ramsey, 2001) and expressed as cal. yr BP (calendar years before 1950). In order to simplify the comparison of the different marine records the ages are not corrected for marine reservoir effects since they are inconsistent and poorly known in the Baltic Sea. For example, Andrén et al. (2000a) corrected the calibrated ages for a marine reservoir effect ranging from 100 to 300 years and used a correction age of 400 years to compensate the contribution of old resuspended carbon in the Gotland Basin, while Hedenström and Possnert (2001) estimated a total reservoir effect of 700–1000 14C yr from analyses of coastal Littorina Sea s. str sediments. 5. Occurrence of hypoxia 5.1. During modern time (last 50 years) From water quality monitoring records, the bottom surface area covered by hypoxic water averaged c. 41000 km2 over the time period 1970–2000 (Conley et al., 2002). The smallest hypoxic area occurred in 82 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 Fig. 3. Modern variability of hypoxia in the Baltic Proper (after Conley et al., 2002). 1993 during the peak of the “stagnation period” when major salt water inflows were at their minimum (Stigebrandt and Gustafsson, 2007) with only 12 000 km2 of bottom covered by hypoxic waters. The peak in hypoxic area was in 1971 following large salt water inflows in previous years (Conley et al., 2002). While major salt water inflows are well known to displace and renew bottom water oxygen supplies (Matthäus, 2006), it is less well appreciated that they also create large areas of stratification where oxygen can be depleted (Gerlach, 1994; Conley et al., 2002). In modern times it has been typical that most of the deep basins are continuously hypoxic including the Gotland Deep, Landsort Deep, NW Baltic Proper, and Gdansk Deep (Fig. 3) with permanent anoxia in the deepest parts of the Gotland Deep (Conley et al., 2002). During years with large areas of hypoxia, low oxygen zones migrate higher up in the water column and the different basins become connected to form one large hypoxic area (Fig. 3). In the Gulf of Finland salt water inflows create stratification, and together with the small volume of water under the halocline and coupled with high productivity, oxygen is rapidly depleted below the halocline (Laine et al., 2007). Trends in hypoxia in modern times in the Gulf of Finland are related to variations in salt water inflows, with less stratification and less hypoxia during “stagnation periods” when salt water inflows are reduced. Fig. 4. Figure showing the expansion of laminated bottoms in the Baltic Proper during the last 100 years (after Jonsson et al., 1990). Laminations deposited before 1940 are considered as natural while post-1960 laminae formation is regarded as a result of human impact. The study was based on results from 29 coring sites and ages of the sediments were obtained by counting the individual laminae. 5.2. During modern historical time (AD 1950–AD 1800) Historical proxy records of hypoxia, defined here as the sediment archives deposited during the last 50–200 years in the Baltic Sea, show that laminations are prevalent in the sediments (Gripenberg, 1934; Ignatius et al., 1968; Jonsson et al., 1990; Hille, 2005). Jonsson et al. (1990) reported laminations in 8 out of 29 coring sites in the Gotland Basin, Landsort Deep and Bornholm Basin with the upper 40 cm of sediment consisting of 100–200 continuously and annually deposited laminae. Based on the distribution of laminations in surficial sediments, Jonsson et al. (1990) summarized the expansion of laminated bottoms in the Baltic Proper with rapid increases in the presence of laminated sediments following 1960 (Fig. 4), the time period when eutrophication rapidly increased in the Baltic Sea. Hille (2005) sampled the upper 1 m of sediments at 53 stations at depths deeper than 150 m in the Gotland Basin and mapped the spatial distribution of laminated sediments. At almost all stations the uppermost sediments were distinctly laminated, and had accumulated over homogeneous non-laminated sediments. Fig. 5. Bathymetric map over the Gotland Deep showing the extension of laminated sediments deposited 100 years ago to the present (after Hille, 2005). The result is based on samples of the upper sediments at 53 stations. The age estimates are based on the 210 Pb dating technique. L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 Based on 210Pb dating, Hille (2005) observed that the laminated sediment structures started to form, basin wide, at depths greater than 150 m, 100 years ago basin-wide in the Gotland Basin (Fig. 5). Laminated sediments were reported from a few stations at shallower depths less than 150 m, which suggest that the area with lamina formation could be even greater than shown in Fig. 5. Gripenberg (1934) recovered about 80 short (b30 cm) sediment cores during the years 1924–1928. She describes sediments with stratifications (some rich in hydrogen sulfide) in clayey muds in both coastal and deeper basins within the Bothnian Bay, Bothnian Sea, Gulf of Finland and N Baltic Proper. Although the ages of these laminated sediment cores have not been estimated, they should represent the period from the turn of the century to the time of collection in 1924–1928 to be consistent with Hille (2005) and Jonsson et al. (1990). 5.3. On geological time-scales (AD 1800–10 000 cal. yr BP) A compilation of the geological records reporting the occurrence of Holocene laminated sediments (i.e. the proxy used for occurrence 83 of past hypoxia in this paper) in the Baltic Sea during the last c. 10 000 cal. yr BP is presented in Fig. 6. Owing to the fact that accumulation basins have the potential to provide long undisturbed sediment sequences, the majority of the sediment cores were therefore recovered from the deeper basins in the Baltic Sea (see Fig. 1.) The core lengths covering the last c. 10 000 years vary between c. 4 and 10 m. In most studies, data from the uppermost sediments (i.e. the last c. 100–200 years) are missing due to sediment lost during the coring operation. All studies suggest that laminations started to form in the deeper depressions of the Baltic Sea (i.e. Gotland Deep, Landsort Deep, NW Baltic Proper, North Central Baltic Proper, W Gotland Basin and Gdansk Deep) c. 8500–7800 cal. yr BP ago, in association with the formation of a permanent halocline at the transition between the Early Littorina Sea and the Littorina Sea s. str. (e.g. Witkowski, 1994; Sohlenius et al., 1996; Lepland and Stevens, 1998; Andrén et al., 2000a; Harff et al., 2001; Emeis et al., 2003). The deeper basins have experienced intermittent hypoxia during most of the Holocene except at depths N250 m in the Landsort Deep where sediment cores show Fig. 6. Figure showing the occurrence of laminated sediment plotted on time scales based on calibrated uncorrected 14C-ages (1σ; Oxcal calibration program v4.0, Bronk Ramsey, 2001), predominantly preformed on bulk-sediment samples. In a few studies, the ages of the sediment are based on a combination of calibrated 14C-dates and palaeomagnetic secular variations or biological/physical correlations. Note that the calibrated uncorrected 14C-ages presented here, may vary 500–1000 years in relation to terrestrial dates due to reservoir effects (e.g. Andrén et al., 2000a; Hedenström and Possnert, 2001). In cases of questions regarding lithological descriptions a question mark is shown in order to illustrate the uncertainties in the stratigraphy. The records are presented in order of water depth at the coring site with greater water depths to the right. The data are obtained from various research papers, which are presented in Table 1. Most of the long sediment records are missing the uppermost sediments during the last 200–100 years. 84 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 more continuous laminated structure (Fig. 6). A more detailed description divided by basin is presented below. 5.3.1. Bothnian Sea, Bothnian Bay, Gulf of Finland and Archipelago Sea Studies of the Bothnian Sea and Bothnian Bay reveal that laminated sediments have been deposited at various depths within these basins at least during the beginning of Littorina Sea s. str. (i.e. from c. 6900 cal. yr BP). The upper limit of the laminated sequences is not dated, but the thickness of the deposit is about 1 m, which, considering the estimated sedimentation rates of 1.76–2 mm yr− 1 (Jerbo, 1961; Ignatius et al., 1968) corresponds to about 500–600 years. In the Bothninan Sea continuous laminated sequences have been reported in coastal areas in sediments of the age of c. 8000–1000 cal. yr BP (Ignatius et al., 1968; Fig. 6.). However, the substandard sediment description (white question mark) and the dating (based on correlations to pollen zones) make the interpretation of that record uncertain. In both the Bothnian Sea and Bothnian Bay, sulfide banding is frequently recorded in both Early Littorina clays and in younger sediments (Jerbo, 1961; Fig. 6). In parts of the Gulf of Finland, laminated sediments were deposited from the Early Littorina Sea-Littorina Sea s. str. transition at c. 8000 cal. yr BP until at least c. 5000 cal. yr BP (Åker et al., 1988; Fig. 6). In this area, sulfide bands are reported in sediments older than Littorina Sea s. str. age. Periods of laminated sediments have been recorded in the Archipelago Sea during the majority of the Littorina Sea s. str. stage (i.e. c. 7600–2800 cal. yr BP; Virtasalo et al. 2005, 2006; Fig. 6). It should be noted that some of the above mentioned sediment cores were recovered from areas situated in relative shallow and sheltered coastal areas (see Fig. 1). 5.3.2. Gotland Basin We show six records from the Gotland Basin, which in most of the cases are a compilation of several sediment cores (see Table 1.). The majority of the sediment sequences display periods of alternating laminae and homogeneous sediment deposition. Laminae deposition started to form between 8500 and 7800 cal. yr BP. Records 2, 3, 5 and 6 show three periods of laminae formation and preservation during the last c. 8500 cal. yr BP (Fig. 6; Table 1). These periods are centered around c. 8000–7000, 6000–4000 and 2000–800 cal. yr BP. Records 1 and 4, display continuous laminae formation between 8300–5800 and 8500–4600 cal. yr BP, respectively (younger sediments deposited above the laminated sequences were not sampled). However, a substandard sediment description makes it difficult to determine whether the latter sediment sequences contain homogeneous sediments between 6000 and 7000 cal. yr BP or not (Fig. 6). Four of the records (i.e. 2–5) display iron sulfide banding in sediments pre-dating Littorina Sea s. str. age. 5.3.3. NW and NC Baltic Proper The five records from the NW and NC Baltic Proper show similar alternating periods of laminated and homogeneous sediment Table 1 Number, core label, water depth (m), lithological description, dating method, analyses preformed and references presented in Fig. 6 No Core label Water depth Lithological (m) description Dating Bothnian Bay (BB), Bothninan Sea (BS), Gulf of Finland (GF) and Archipelago Sea (AS) 1 J5, J1, 19, 20, 22, 27 50–100 2 n/a n/a Graphical Inferred from other source 3 AS6-VH1; AS6-VH2 66;64 Text, graphical PM, 14C 14 C 4 Core C 40 Table, B/W Photo Gotland Basin 1 GD0101 2 G94-5; G95-3 3 20048-1; 20048-4; 20007-1 4 n/a 5 6 211660-1; 211660-2 211660-5; 211660-6 Gdansk Deep 1 96 200 206 240 238 Table, graphical Table, graphical Digital core scan, graphical Table, graphical 241 241–249 Table, graphical Graphical 101 Table NW Baltic Proper (NWBP) and NC Baltic Proper (NCBP) 1 9106 73 Graphical 2 9208 83 Graphical 14 Analyses References X-ray diffraction MS, LOI, X-ray, fossils Diatoms, LOI, pollen Jerbo, 1961; Jerbo and Hall, 1961 (BB, BS) Ignatius et al., 1968 (BS) Virtasalo et al., 2005;2006 (AS) Åker et al., 1988 (GF) 14 C C, 14 C 210 Diatoms, pigments, vce Diatoms, TOC, d13C Diatoms, TOC, vce Borgendahl and Westman, 2007 Andrén et al., 2000a Sohlenius et al., 2001; Emeis et al., 2003; Miltner et al., 2005 14 C Diatoms, TOC, vce Sternbeck et al., 2000; Sternbeck and Sohlenius 1997; Sohlenius et al., 1996 14 210 C, Pb Diatoms, TOC, vce Andrén et al., 2000a PM, and correlation to 211660-1 Fossils, chlorophyll, TOC, Harff et al., 2001; Dippner and Voss 2004; Kunzendorf vce, geo statistic and Larsen, 2002; Alvi and Winterhalter, 2001; Brenner, 2001a; Voss et al., 2001; Kotilainen et al., 2000; Nytoft and Larsen, 2001 14 210 Diatoms,vce C, Pb Pb 14 Diatoms, TOC, vce Diatoms, pigments, TOC, vce 14 C Diatoms, TOC, vce No direct datings, PM, 14C, both Fossils, vce inferred from other cores MS PM, 14C Westman and Sohlenius 1999 (NWBP) Bianchi et al., 2000; Westman and Sohlenius 1999 (NWBP) Westman and Sohlenius 1999 (NWBP) Brenner 2001b; Nytoft and Larsen 2001 (NCBP) Landsort Deep (LD) and W Gotland Basin (WGB) 9205 106 Graphical 14 Bianchi et al., 2000; Westman and Sohlenius 1999 (LD) 2 Psh-2537 155 Table, graphical 14 3 n/a 221 Text, graphical 14 4 9302 250 Graphical 14 Text, graphical Text, graphical 14 3 4 9102 211670-4; 211670-7 111 175 Graphical Graphical 5 GC-4 175.8 m Graphical 5 6 14754-4 n/a 252 459 C C Witkowski, 1994 14 C C C C C C 14 Diatoms, pigments, TOC, vce Diatoms, pollen, Grain size, C. org, Vce TOC, vce Diatoms, pigments, TOC, vce Diatoms, pollen TOC, vce Kotilainen et al., 2001 (NWBP) Elmelyanov et al., 2001 (WGB) Lepland and Stevens 1998; Böttcher and Lepland, 2000 (LD) Bianchi et al., 2000; Westman and Sohlenius 1999 (LD) Thulin et al., 1992 (LD) Lepland and Stevens 1998 (LD) PM = Palaeomagnetic dating; MS = Magnetic susceptibility; LOI = Loss on ignition; TOC = Total organic carbon; vce = various chemical elements. L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 deposition as those in the Gotland Basin (Fig. 6). However, two of the records (i.e. 1 and 3) are missing data for the last c. 6000 and 4300 cal. yr BP, respectively, due to non-deposition or erosion of sediments. Laminations begin to form between 8400 and 8000 cal. yr BP. A short period of homogeneous sedimentation is evident between 8000 and 7000 cal. yr BP in record 1 and 3, while records 4 and 5 from the NC Baltic Proper display longer periods of homogeneous sedimentation during the early Littorina Sea s. str. i.e. between c. 8000 and 6000 cal. yr BP. In the three continuous sediment sequences homogeneous sediment deposition dominates after c. 5000 cal. yr BP until c. 2600 cal. yr BP (record 2) and c. 1400 cal. yr BP (records 4 and 5), where after a period of c. 500 year of laminated sediments are deposited. Record 4 also displays laminae formation during the last c. 100 years. 5.3.4. Landsort Deep and W Gotland Basin Three sediment records (i.e. 1, 3 and 4) from the Landsort Deep are characterized by alternating laminated and homogeneous sedimentation as in the Gotland Basin and NW and NC Baltic Proper, while the remaining records (i.e. 2, 5 and 6) display only one long interval of laminae formation extending the majority of the sediment sequences analysed (Fig. 6). Laminations start to form in all complete records between 8400 and 7800 cal. yr BP. In records 1 and 3, laminations dominate the sediment structure from that period until c. 5000 cal. yr BP, while record 4 is very similar to 4 in the NC Baltic Proper and displays a dominant homogeneous sedimentation during the early Littorina Sea s. str. Homogeneous sedimentation dominates after c. 5000 cal. yr BP, where after a period of Late Littorina Sea laminae formation occurs between c. 2000 and 1600, 3300 and 1600, 1800 and 1000 cal. yr BP in records 1, 3 and 4, respectively. Record 2, from the West Gotland Basin (south of Landsort Deep) may imply that laminated sediments were deposited there persistently, at a depth of 155 m, between c. 7800 and 2000 cal. yr BP, but sediment descriptions are not sufficient enough to rule out the presence of non-laminated phases. Records from the deepest part of the Landsort Deep (no. 5 and 6) show continuous laminae deposition from 6800 and 5200 cal. yr BP to present, respectively, although these shorter cores do not have the complete Holocene record. In core 6 two Mn(II)-precipitate rich sapropel intervals 1390–1990 cal. yr BP are found indicating an expansion of bottom water anoxic conditions. One record from these areas report the occurrence of sulfide banded Early Littorina clay (i.e. 3). 5.3.5. Southern Baltic Sea Investigations from the shallower southern Baltic suggest that the Bornholm and Arkona basins were frequently oxygenated during the Holocene and that sediments from that region were bioturbated, although geochemical records imply more reduced conditions after the Early Littorina–Littorina s. str transition (Sohlenius et al., 2001; Moros et al., 2002). Andrén et al. (2000b) reports frequent occurrence of Macoma shell throughout a sediment core from the deepest part of the Bornholm Basin since c. 6500 cal. yr BP. Frequent shell debris of Macoma baltica is also reported in a long sediment sequence from the Gdansk Deep at a water depth of 88 m (Witkowski, 1994). However, in the deeper Gdansk Deep, at a depth of 101.5 m Witkowski (1994) describes continuous laminated sediments between c. 9000 and 3300 cal. yr BP (Fig. 6). No long laminated sediment sequences have been reported from Mecklenburgian Bay or Gulf of Riga (e.g. Kowalewska, 2001; Rößler, 2006). 6. Discussion In the very deepest areas of the Baltic Sea, below 250 m in the Landsort Deep, and in some coastal areas, hypoxia has been present more or less continuously during the last c. 8500–7800 cal. yr BP. At shallower water depths (above 250 m), but below the permanent 85 halocline, there are 3 major periods of hypoxia: during the early Littorina Sea s. str. (c. 8000–4000 cal yr. BP), during the middle Late Littorina Sea (c. 2000–800 cal. yr BP) and during the last c. 100 years, although an intensification of hypoxia after 1950 has occurred. Three periods of homogeneous sedimentation are prominent in most of the records: between c. 7000–6000, c. 4000–2000 and c. AD 1200–1900. In the following sections we will discuss the spatial and temporal variability in hypoxia within the Baltic Sea and explore possible driving factor(s) for hypoxia especially focusing on the role of climate and anthropogenic pressures. 6.1. Hypoxia in time and space 6.1.1. Bothnian Bay, Bothnian Sea, Gulf of Finland and the Archipelago Sea Short laminated sequences have been registered (between c. 500 and 600 yr in duration) in early Littorina s. str. sediments in the Bothnian Sea and Bothninan Bay by Jerbo (1961). High relative sealevel (c. 100–150 m above present) together with limited freshwater input (Gustafsson and Westman, 2002) lead to higher salinities, increased stratification (reduced mixing) enhancing the conditions needed for the occurrence of hypoxia and the deposition of laminated sediments in the northern basins during the early development of the Baltic Sea. The salinity gradually decreased throughout the Littorina Sea and the halocline in the northern Baltic diminished. To our knowledge, no laminated sediments from the open sea of younger age have been reported in this region except for those of modern historical age by Gripenberg (1934). The Gulf of Finland has experienced periods of severe hypoxia in modern times (Laine et al., 2007). The geological records from Gulf of Finland and the Archipelago Sea demonstrate that these regions were affected by hypoxia during most part of the more saline Littorina Sea s. str (Åker et al., 1988; Virtasalo et al., 2005, 2006). However, well described long sediment records from these regions are sparse (Heinsalu et al., 2000) and those presented here are recovered from shallower coastal areas, which may reflect hypoxic conditions influenced by local environmental factors, such as topography and water currents, rather than regional signals of past hypoxia. 6.1.2. Baltic Proper The Baltic Proper is the main basin affected by hypoxia during the Holocene. However, due to the large dating uncertainties, it is difficult to determine precisely if the separate periods with lamina formation in this area occurred synchronously (Fig. 6). Better chronological control is needed for sediment cores from the deeper basins, as most studies are based on 14C dating on bulk sediment samples with the associated problem of contamination from old reworked carbon and unknown brackish/marine reservoir ages and their temporal variability. It can, however, be argued that the alternating periods of laminae and homogeneous sediment deposition in the Gotland Basin, NW and NC Baltic Proper and Landsort Deep coincide within the error estimates associated of the individual chronologies and that such deposition has occurred basin wide at depths varying between 73 and 250 m in the Baltic Proper during the last c. 8500–7800 cal. yr BP (Fig. 6). In order to generalize the discussion about the Baltic Proper and possible trigging mechanism for open sea hypoxia, we show data from only two sediment cores (211660-1 and 20048-1), covering the last c. 10 000 years, recovered from the Gotland Deep (e.g. Andrén et al., 2000a; Emeis et al., 2003). These two sediment cores (Fig. 7) are representative of conditions in the Gotland Deep at a water depth of c. 240 m (they show almost identical stratigraphy in comparison with at least 6 other cores; see Table 1) and were selected because they provide (i) well established chronologies, (ii) detailed sediment descriptions, and (iii) data from various analyses, which can be used in the analysis for the driving mechanisms for hypoxia. Terrestrial 86 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 palaeoclimate data obtained from lake sediments in the Baltic Sea drainage area (see Fig. 1) were compiled for comparison to the hypoxia records (Fig. 7). 6.1.2.1. Littorina Sea s. str. (8500–4000 cal. yr BP). From c. 8500–7800 to c. 4000 cal. yr BP laminated sediments were formed in all the deeper basins (except for a shorter period between 7000 and 6000 cal. yr BP) in the Baltic Proper (i.e. Gotland Deep, Landsort Deep, NW and NC Baltic Proper, W Gotland Basin and Gdansk Deep; Fig. 6), which suggest that these areas were affected by hypoxia during most of the Littorina Sea s. str. At this time, the relative salinity in the Baltic Proper was high (10–15; Gustafsson and Westman, 2002; Emeis et al., 2003), most likely influenced by high sea-levels between c. 8000 and 6500 cal yr BP (Yu et al., 2007), as well as dry conditions (Snowball et al., 2004) minimizing the freshwater discharge to the basin, and the upper limit of the hypoxic zone was probably determined by a strong stratification. Laminae formation also corresponds to high TOC-values (6–8%; reflecting high organic accumulation) in the Baltic, and less negative δO18 values (reflecting dry conditions and low net precipitation), low lake-levels, high annual mean temperatures and low χhf values (indicating low precipitation in the form of snow and associated reduced catchment erosion rates during spring snowmelt; Fig. 7) in the terrestrial surroundings. The time period between 9000 and 5000 cal. yr BP is known as the Holocene Thermal Maximum (HTM), which in Northwest Europe is characterized by relatively high summer insolation, high atmospheric temperatures and dry conditions (Snowball et al., 2004). The HTM caused the possible disappearance of glaciers in the Scandinavian mountains (Dahl and Nesje, 1994; Snowball and Sandgren, 1996; Nesje et al., 2001), high sea surface temperatures in the North Atlantic (Koç et al., 1993), a pine tree-limit c. 300–400 m higher than today (Kullman, 1999; Barnekow, 2000), lower lake-levels and decreased humidity in southern Sweden (Digerfeldt, 1988; Hammarlund et al., 2003) and increased summer atmospheric temperatures as recorded by pollen analyses (Seppä and Birks, 2001; Seppä et al., 2005) and chironomid reconstructions (Korhola et al., 2002). Fig. 7. Figure showing the occurrence of laminated sediments (Andrén et al., 2000a), total organic carbon content (TOC; Andrén et al., 2000a) and salinity estimates (Emeis et al., 2003) derived from 13C/12C ratio of organic carbon, plotted against a calendar year time scale, which is corrected following Andrén et al. (2000a) i.e. we have applied a marine reservoir effect of 100–300 yr in sediments older than 8000 cal. yr BP, a marine reservoir effect of 300 yr in sediments younger then 8000 cal. yr BP and due to the present of old resuspended carbon in the Baltic Sea an additional correction age of 400 yr is applied throughout the sequence. Also shown are modelled salinity estimates (shaded curve) after Gustafsson and Westman, (2002) based on a compilation of records. The chronology of the latter curve is based on 14C dates of various materials, including macrofossil remains, and is adjusted for a marine reservoir effect of 300 years. The sediment records suggest that during the last c. 8500–7800 cal. yr BP hypoxia has occurred intermittently in sub-basins in the Gotland Deep. Note that periods of laminated sediments coincide with intervals of high TOC content and high salinity estimates. Marked are also the timing of the Holocene Thermal Maximum (HTM) and the Medieval Warm Period (MWP). Also, terrestrial palaeoclimate data obtained from Swedish lake sediments i.e. (1) oxygen isotopes reflecting effective humidity/net precipitation (Hammarlund et al., 2003), (2) regional reconstruction of lake-level fluctuations reflecting past hydrology changes (Digerfeldt, 1988), (3) pollen-based annual mean temperature reconstructions (Seppä et al., 2005), and (4) high-field magnetic susceptibility (χhf) which is a signal of catchment erosion (Snowball et al., 2002). The timescales for Lake Igelsjön, Bysjön and Flarken sediment sequences are based on 14C dating of terrestrial macrofossil remains while the chronology for Lake Sarsjön sediment sequence, which is annually laminated, is established by varve counting. L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 The interval between 7000 and 6000 cal. yr BP is characterized by homogeneous sedimentation and corresponds to a decline in the salinity estimates (Emeis et al., 2003), lower TOC-values, increased effective humidity (more negative δO18-values and rising lake-levels), enhanced catchment erosion (a peak in the χhf record) and a c. 0.5 °C drop in temperature (Fig. 7). This event has also been recorded in varved lake sediments in west central Sweden (Zillén, 2003), where it was interpreted as a response to colder climate conditions with increased winter snow accumulation. It also coincides with one of the major ice-rafting episodes in the North Atlantic (Bond et al., 1997) and glacier advances in northern Sweden (Karlén and Kuylenstierna, 1996). 6.1.2.2. Late Littorina Sea (4000 cal. yr BP–AD 1800). A period of homogeneous sediment accumulation is evident in most records between c. 4000 and 2000 cal. yr BP (Fig. 6), and corresponds to lower salinity estimates (c. 4–10; Gustafsson and Westman, 2002; Emeis et al., 2003) and lower TOC concentrations (c. 3–4%; Andrén et al., 2000a; Fig. 7). The terrestrial data-sets show a marked change around 4000–3700 cal. yr BP and coincide with the Late Holocene climate development in Northwestern Europe, which was characterized by decreasing temperatures and more moist conditions (e.g. Hammarlund et al., 2003; Snowball et al., 2004; Seppä et al., 2005; Fig. 7). In Scandinavia the start of this period (4000–3500 cal. yr BP) was marked by tree-line decline in the Scandinavian mountains (Barnekow, 2000) and glacier advances (Nesje and Kvamme, 1991; Snowball, 1996). During this period, bottom water in the Baltic Proper became oxic promoting the re-establishment of a benthic fauna and bioturbation of sediments. From c. 2000 to 800 cal. yr BP laminated sediments were again deposited, basin wide, in the Baltic Proper (Fig. 7). The laminae deposition corresponds to maximum TOC concentrations (8.4%) around 900–800 cal. yr BP in the Gotland Basin, which was interpreted by Andrén et al. (2000a) as a result of high primary production. The lake records show less negative δO18-values, a lowering of lake-levels, increased temperatures and low χhf values between c. 1500 and 800 cal. yr BP (Fig. 7). The laminae deposition and the marine and terrestrial paleoclimate and palaeoenvironmental changes presented here, overlap with the timing of the climate anomaly known as the MWP (or Medieval Climate Optimum-MCO) when temperatures were probably c. 0.5–0.8 °C higher than today (e.g. Briffa et al., 1990). However, the salinity reconstructions by Gustafsson and Westman (2002) and by Emeis et al. (2003) are inconsistent during this time period and onwards. The estimates by Emeis et al. (2003) show a steady increase in salinity from about 3000 cal. yr BP, although higher salinities during this period are not supported by diatoms. The reconstruction of salinity by Gustafsson and Westman (2002), which are based on modeling of the physical process of saltwater input from an expanded opening of the Danish Straits and physical mixing of water masses, imply low and relatively stable values from c. 2000 cal. yr BP to the present and agrees better with the diatom records. In the absence of multiple high-resolution salinity reconstructions, it is therefore difficult to determine if the salinity increased during this time interval or not, even though the terrestrial palaeoclimate records suggest decreased freshwater discharges. The period of hypoxia overlapping the MWP also correlates with population growth and large-scale changes in land use (see Fig. 8) that occurred in much of the Baltic Sea watershed. Watershed studies show large increases in nutrient loading with cutting of forests (Likens et al., 1970). Reconstructed phosphorus concentrations and associated eutrophication of lakes (Fig. 8) began during the Bronze and Iron Ages (1700 BC–AD 1050) with rapid increases in nutrient loading associated with major changes in agriculture during the Medieval period (Bradshaw et al., 2005). From high diatom-inferred total phosphorous values, Renberg et al. (2001) concluded that Lake Mälaren, Sweden, was culturally eutrophicated already from Medieval 87 time and that even more nutrient-rich conditions developed after c. AD 1850. The early-medieval expansion was followed by a period of stagnation and population decline in the late 14th and early 15th centuries (Fig. 8) mainly due to the Black Death. In Sweden the population decreased from 1100 000 to 347 000 from AD 1300 to AD 1413 (Andersson Palm, 2001; Fig. 8). This decline is known as the latemedieval crisis and was characterized by decreased total production and abundance of farms (Lagerås, 2006). It affected most parts of Europe and had an impact on all levels of society (Lagerås, 2006). During this period (after c. AD 1200), the bottom waters in the Baltic Sea became more oxic and homogeneous sediments were deposited until the beginning of the 19th century (Fig. 7). In Northwestern Europe, this period is characterized by a climate deterioration with the onset of the Little Ice Age (LIA) in the 14th century when glaciers advanced in the Scandinavian mountains and it was c. 1 °C cooler than today (Matthews and Briffa, 2005). This alteration is also visible in the Baltic Proper as a change in the diatom composition dated to c. 650–750 cal. yr BP (Andrén et al., 2000a). 6.1.2.3. Modern historical time period to present (AD 1800 to present). Laminae of historical age have been reported from the Baltic Proper with appearance of laminae around 200–100 years ago with a modern expansion from 1940 to the present (Jonsson et al., 1990). However, the age assumptions were based on counting individual laminae in short sediment cores. Applying that technique requires identification of an annual sedimentation cycle that explains the seasonally deposited laminae and their composition (Zillén et al., 2003); no such sedimentological model is presented in their paper and their age estimates should therefore be considered with caution. However, Hille (2005) showed that continuous laminae formation are present basin wide at depths N150 m in the Gotland Basin beginning 100 years ago based on 210Pb dating. Interestingly, Gripenberg (1934) reported laminated sediments and hydrogen sulfide formation of approximately the same age (i.e. 100 years) in the Bothninan Bay, Bothnian Sea, Gulf of Finland and N Baltic Proper. In addition, the historical sediment records imply that large areas of the Baltic Sea were hypoxic at least from AD 1900 (Jonsson et al., 1990) and that the deeper bottom waters of the Baltic Proper have been depleted of oxygen since then and up to the present time. The start of the industrial revolution in Northwestern Europe around AD 1850 dramatically increased human impact in the Baltic Sea drainage area. The Swedish population grew about 0.8% per year from AD 1801–1900 (i.e. from 2.35 to 5.14 million people; Andersson Palm, 2001; Fig. 8) and the agricultural production increased with 0.5% per year and capita between AD 1750 and AD 1850 (Larsson and Olsson, 1992). Around AD 1850 digging of drainage ditches intensified in Fennoscandia (often recorded as a massive clay layer in individual lake sediments; Petterson, 1999; Ojala, 2001; Zillén et al., 2003) and the forest industry rapidly expanded. Between AD 1850 and1880 Swedish export of wood increased about 4 times (from 667000 to 2763000 m3 per year; Larsson and Olsson, 1992). In Finland the same number increased 3 times between AD 1830 and 1870 (Schybergson, 1973). The forests were heavily exploited and numerous saw mills were built along the main rivers in Fennoscandia. For example, in Finland, the number of saw mills doubled between AD 1840 and 1870 (Schybergson, 1974). Maps show that during the 19th century and the beginning of the 20th century, large areas in southern Sweden were sparsely vegetated. Grazing, cultivation and deforestation were very intensive. It is known from historical sources that the most treeless period in the area occurred during this period (Emanuelsson, 1993). As an example, at the end of the 18th century oak stands had been destroyed in many areas and oaks were therefore protected by a Royal Ordinance. Studies of Swedish lake sediments show that after the reorganization of the agrarian system during 19th century soil erosion and accumulation of allochtonous matter in the lake increased substantially and subsequent bottom water anoxia developed around AD 1900 (Olsson et al., 88 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 Fig. 8. (A) The Swedish population AD 1300–2000 as estimates by Andersson Palm, (2001). (B) The reconstructed P profiles on a calendar year time-scale BC/AD from Dallund Sø, Denmark after Bradshaw et al. (2005). (C) Vegetation/land use changes in Southern Sweden on a BC/AD time-scale illustrated as approximate proportions of three main land use categories based on pollen diagrams after Berglund et al. (1991). Coast refers to the coastal landscape and inland to the inner landscape. 1997). Significant changes in diatom assemblages attributed to eutrophication dated to c. AD 1850–1900 have been observed in estuaries of the Baltic (Clarke et al., 2006) and in the southern Baltic Sea (Witkowski and Pempkowiak, 1995; Andrén et al., 1999). 6.2. Hypoxia and driving mechanisms From the discussion in Section 6.1., it is clear that climate and anthropogenic pressures both have played a role as drivers of hypoxia through time in the Baltic Sea. Climate can be a short-term driver influencing hypoxia on the scale of years through variations in deep water inflows (Conley et al., 2002) and on longer time-scales with influencing freshwater inflow and the overall salt balance of the Baltic (Gustafasson and Westman, 2002). The occurrence of hypoxia in the Baltic basin on geological timescales (prior to more intense human impact) as reconstructed in this study, seems to be related to two primary forcings: increased salinity and increased productivity in a warmer climate. In contrast, homogeneous sedimentation corresponds to more oxic conditions and weakening of those two driving factors. The salinity in the Baltic Sea varies due to changes in both saltwater inflow and freshwater input (Stigebrandt, 2001; Gustafsson and Westman, 2002), where the former is controlled by the depth of the salinity surface in the inlet areas (Stigebrandt, 2001) and the latter by net precipitation (precipitation-evaporation) in the drainage area. Unfortunately, there are no data of saltwater inflow variability, beyond the time of instrumental records, which makes the interpretation of such changes and their relation to hypoxia on long time-scales difficult. However, freshwater variability should be reflected in the terrestrial sediment proxies of effective humidity/net precipitation, lake-level and catchment erosion (Fig. 7). Due to the good coherence between the stratigraphy in the sediment records from the open Baltic Sea and the reconstructed hydrology changes in the catchment area (Fig. 7) it is our hypothesis that freshwater variability has been one of the factors, in addition to saltwater inflow from the Kattegatt/ Skagerrak and sea-level changes, controlling the deep oxic conditions in the basin and that such changes have followed the general Holocene climate development in Northwest Europe. It is noteworthy, that the period of oxic bottom conditions between c. 7000 and 6000 cal. yr BP occurs during a phase of generally high salinity and high sea-levels (Gustafsson and Westman, 2002). The evidence of such a condition suggest that climate driven hydrological changes in the drainage area (Fig. 7) managed to turn a hypoxic system into a more oxic state for a period of c. 200–1000 years. Run-off variability in the catchment area could be explained by oceanographic and atmospheric changes over the North Atlantic, L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 where periods of strong westerlies across northern Europe in association with more positive North Atlantic Oscillation (NAO) index would increase the precipitation and wind stress over the Baltic Sea (Hurrell, 1995). Such conditions would cause a greater outflow and a smaller and fresher inflow due to lower surface salinity in the inlet areas (Stigebrandt, 2001) and subsequently leading to a freshening of the basin (Samuelsson, 1996). This freshening, in combination with increased wind stress (wind stress is one of the major physical parameters that controls the vertical mixing of the upper layers in the Baltic Sea; Stigebrandt, 2001), over the Baltic Sea would result in a weakened halocline and enhanced vertical mixing (Stigebrandt, 2001) allowing more efficient exchange of oxygen across the halocline (Conley et al., 2002). This scenario would promote more oxic bottom water conditions and the deposition of homogeneous sediments. Changes in the Baltic Sea and the connection to climate and environmental variability have previously been explored by Emeis et al. (2003) who postulate that high salinity resulted in an increased density contrast and a stronger stratification of the water column, which promoted hypoxia in the bottom waters with the subsequent deposition of laminated sediments. The changes in salinity were hypothesized to be linked to climatic fluctuations over the North Atlantic, where high salinity phases coincided with warmer and drier periods of low river runoff and prominent atmospheric high-pressures over Scandinavia and northeastern Europe. Similar relationship between NAO indices and the salinity has been recognized in marine waters in fjords on the Swedish west-coast (Nordberg et al., 2000, 2001). Based on comparison between instrumental measurements and sediment records, the latter studies showed that there was a close connection between higher bottom-water salinity during negative NAO index and the deposition of laminated sediments. Also, Zorita and Laine (2000) analyzed the relationship between annually average salinity and oxygen concentrations during the last 30 years in the Baltic Sea and changes in NAO indices. They show that on such time-scales the salinity and oxygen concentrations were negatively correlated within each layer, i.e. low salinity corresponded to higher than normal oxygen contents and vice versa. Stronger than normal westerly winds were related to lower than normal salinity in the upper and lower layers in almost all areas in the Baltic Sea together with higher than normal oxygen concentrations. They suggest that negative salinity anomalies may be caused by increased precipitation in the Baltic Sea catchment during positive NAO indices and that this process may be more important at longer time-scales than the inflow of saltier North Sea waters as suggested by Matthäus (2006). This is supported by Gustafsson and Westman (2002) who demonstrate that changes in net freshwater input from the catchment area explains the major part of salinity variability in the Baltic Sea during the last 8500 years and that periods with larger than normal freshwater run-off in the drainage area result in decreased salinities. It is well established that recent anthropogenic enhanced eutrophication during the last c. 50 years has caused expansion of hypoxia in bottom waters of the Baltic Sea (Jonsson et al., 1990; Conley et al., 2002). However, it is our hypothesis that hypoxia occurring as early as the MWP are also associated with the first large-scale impact of man in the Baltic Sea watershed. Large land use changes, as those described in Sections 6.1.2.2. and 6.1.2.3., were probably common in most of the countries around the Baltic Sea. Such changes (i.e. agricultural development and population growth) would most likely have caused increased erosion and nutrient input to the Baltic Sea and promoted eutrophication and, in combination with the warmer climate, triggered hypoxia (TOC concentrations display maximum values during this time period). The implication for this scenario would be that the historical records of laminated sediments in the Baltic Sea may not only be of natural origin, but also partly anthropogenic, and that the modern 89 expansion of laminated sediments is a result of further human impact. Vahtera et al. (2007) have hypothesized that there are internal feedbacks within the Baltic Sea that act to sustain hypoxia. These include enhanced P release (Conley et al., 2002) and increases in denitrification potential with hypoxia (Vahtera et al., 2007), leading to an acceleration of eutrophication and fueling of ever larger cyanobacteria blooms. In addition, once the system becomes hypoxic it modifies the benthic communities, reducing the abundance of large animals. These benthic organisms both mix organic material downward and irrigate the sediments, and with hypoxia there is a subsequent loss of reducing equivalents that can oxidize organic matter (Karlson et al., 2005). Therefore, repeated hypoxic events can lead to an increase in the susceptibility to eutrophication increasing the vulnerability to further hypoxia perpetuating hypoxia. Once hypoxia occurs, reoccurrence is likely and may be difficult to reverse demonstrating the sensitivity of the Baltic Sea to anthropogenic alterations. We hypothesize that minor increases of nutrients from forest clearance and agricultural development, together with climate warming during e.g. the MWP created hypoxic conditions. Once the Baltic Sea has switched into a hypoxic mode it is difficult to reoxidize because of internal feedbacks in the system. An important question is what factors are responsible for terminating hypoxic conditions in the Baltic Sea? Why did hypoxia end after nearly 4000 years of near continuous hypoxia (c. 8000– 4000 cal. yr BP) during the Holocene Thermal Maximum? Was the cooling that followed the HTM together with decreasing salinity enough to end hypoxia? Why did the Baltic become oxic 200– 1000 years during that time period? The Little Ice Age with really cold winters, less nutrients coming in from less agriculture (in addition to the plague reducing population and impact; Fig. 8), is believed to be the reason for hypoxia ending. The Baltic was oxic again until c. 100 years ago when man's influence was large enough again during the industrial revolution and development of the forest industry, in combination with a warmer climate, to cause widespread hypoxia. Will proposed reductions in nutrient inputs (Johansson et al., 2007) be sufficient to switch the Baltic back to an oxic situation in the face of a substantial global warming? In summary, environmental changes in the Baltic Sea catchment, either triggered by climate change or human impact, may have been of more significance for the development of hypoxia in the past than previously thought. Land–sea interactions are thus not just modern phenomenon but have been important on long time-scales. More research is needed to determine the relative importance of the main driving forces, i.e. climate change, human impact, and internal feedback mechanisms. In the light of global warming and increased anthropogenic pressures in the Baltic Sea region, it is essential to include these issues into the discussion of the present state and the future of the Baltic Sea. 7. Conclusions During the early and more saline development of the Littorina Sea s. str. (c. 8000–6000 cal. yr BP), laminated sediments were deposited in the northern Baltic Sea (Bothnian Sea and Bothnian Bay) where hypoxic bottom conditions prevailed. Laminae deposition has occurred basin wide, at depths varying between 73 and 250 m during three major periods in the Baltic Proper the last c. 8500–7800 years: i.e. between c. 8000–4000, 2000–800 cal. yr BP and subsequent to AD 1800. These periods overlap the HTM (c. 9000–5000 cal. yr BP), the MWP (c. AD 750–1200) and the modern historical period (AD 1800 to present). Laminae formation coincides with increases in salinity (at least during the HTM) and organic carbon accumulation. Hypoxia appears to have been more or less continuous at greater water depths (N250 m). Oxic bottom conditions were common in the Baltic Proper between c. 7000–6000, c. 4000–2000 and c. 800–200 cal. yr BP and 90 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 correspond to estimates of low salinity, low organic carbon accumulation and to Holocene climate development in Northwestern Europe characterized by decreasing temperatures and more moist conditions. The southern Baltic Sea (e.g. the Bornholm Basin and Arkona Basin) was frequently oxygenated during the Holocene, except for the deepest part in Gdansk Deep. Large hydrological changes in the Baltic Sea catchment as a response to climate changes, have most probably affected the environmental conditions in the basin. Long-term freshwater discharge variability may have been an important factor controlling the stratification of the water column and the deposition of laminated sediments in the Baltic Sea during the last c. 8000 cal. yr BP. Laminated sediments spanning the MWP correlate with population growth and large-scale changes in land use that occurred in the Baltic Sea watershed during the early-medieval expansion. This implies that the change from homogeneous to laminated sediment deposition in the Baltic Sea was probably caused by a combination of climate changes and human impact. Large areas of the Baltic Sea (i.e. the Bothnian Sea, Bothnian Bay, Gulf of Finland and Baltic Proper) were hypoxic around AD 1900, which coincides with the beginning of the industrial revolution at AD1850 in Northwestern Europe when population grew, agricultural production increased, cutting of drainage ditches intensified and forest industry expanded explosively, as well as with the beginning of the on-going global warming. The deep waters of the Baltic Proper have been depleted of oxygen since then and up to present time. Once widespread areas of the Baltic Sea become hypoxic there are processes, such as enhanced P flux and reduced denitrification, which act to sustain and continue hypoxic conditions. This demonstrates the sensitively of this large enclosed sea to anthropogenic perturbations. Acknowledgments This work was partially supported by a grant from Baltic Sea 2020 and by a Marie Curie Chair to DJC (MEXC-CT-2006-042718). We thank Bo Gustafsson and Maren Voss for comments on earlier versions of the manuscript. We also acknowledge the valuable reviews from Erik Bonsdorff and Anders Stigebrandt. References Alvi, K., Winterhalter, B., 2001. Authigenic mineralisation in the temporally anoxic Gotland Deep, the Baltic Sea. Baltica 14, 74–83. Andersson Palm, L., 2001. Livet, kärleken och döden. L Palm. Historiska institutionen, Gothenburg, Sweden. 203 pp. Andrén, E., Shimmield, G., Brand, T., 1999. Environmental changes of the last three centuries indicated by siliceous microfossil records from the southwestern Baltic Sea. The Holocene 9, 25–38. Andrén, E., Andrén, T., Kunzendorf, H., 2000a. Holocene history of the Baltic Sea as a background for assessing records of human impact in the sediments of the Gotland Basin. The Holocene 10, 687–702. Andrén, E., Andrén, T., Sohlenius, G., 2000b. The Holocene history of the southwestern Baltic Sea as reflected in a sediment core from the Bornholm Basin. Boreas 29, 233–250. Andrén, T., Lindeberg, G., Andrén, E., 2002. Evidence of the final drainage of the Baltic Ice Lake and the brackish phase of the Yoldia Sea in glacial varves from the Baltic Sea. Boreas 31, 226–238. Andrén, T., Andrén, E., Berglund, B.E., Yu, S., 2007. New insights on the Yoldia Sea low stand in the Blekinge archipelago, southern Baltic Sea. GFF 129, 273–281. Barnekow, L., 2000. Holocene regional and local vegetation history and lake-level changes in the Torneträsk area, northern Sweden. Journal of Paleolimnology 23, 399–420. Berglund, B.E., Sandgren, P., 1996. The early Littorina Sea environment in Blekinge — chronology, transgressions, salinity and shore vegetation. Geologiska Föreningen i Stockholms Förhandlingar 118, A64–A65. Berglund, B.E., Larsson, L., Lewan, N., Olsson, G.A., Skansjö, S., 1991. Ecological and social factors behind the landscape changes. In: Berglund, B.E. (Ed.), The cultural landscape during the 6000 years in southern Sweden – the Ystad Project. Ecological Bulletins, vol. 41, pp. 425–445. Berglund, B.E., Sandgren, P., Barnekow, L., Hannon, G., Jiang, H., Skog, G., Yu, S.Y., 2005. Early Holocene history of the Baltic Sea, as reflected in coastal sediments in Blekinge, southeastern Sweden. Quaternary International 130, 111–139. Bianchi, T.S., Engelhaupt, E., Westman, P., Andrén, T., Rolff, C., Elmgren, R., 2000. Cyanobacterial blooms in the Baltic Sea: natural or human-induced? Limnology and Oceanography 45, 716–726. Björck, S., 1995. A Review of the History of the Baltic Sea, 13.0–8.0 ka BP. Quaternary International 27, 19–40. Björck, S., 2008. The late Quaternary development of the Baltic Sea basin. In: The BACC author Team (Ed.), Assessment of climate change for the Baltic Sea Basin. SpringerVerlag, Berlin, Heidelberg, pp. 398–407. Björck, S., Digerfeldt, G., 1984. Climatic changes at Pleistocene/Holocene boundary in the Middle Swedish endmoraine zone, mainly inferred from stratigraphic indications. In: Mörner, N.-A., Karlén, W. (Eds.), Climatic Changes on a Yearly to Millennial Basis. Reidel, Dordrecht, pp. 37–56. Björck, S., Dennegård, B., Sandgren, P., 1990. The marine stratigraphy of the Hanö Bay, SE Sweden, based on different sediment stratigraphic methods. GFF 112, 265 280. Bond, G., Showers, W., Cheseby, M., Lotti, R., Almasi, P., DeMenocal, P., Priore, P., Cullen, H., Hajdas, I., Bonani, G., 1997. A pervasive millennial-scale cycle in North Atlantic Holocene and glacial climates. Science 278, 1257–1266. Bonsdorff, E., 2006. Zoobenthic diversity-gradients in the Baltic Sea: Continuous postglacial succession in a stressed ecosystem. Journal of Experimental Marine Biology and Ecology 330, 383–391. Borgendahl, J., Westman, P., 2007. Cyanobacteria as a trigger for increased primary production during sapropel formation in the Baltic Sea — a study of the Ancylus/ Litorina transition. Journal of Paleolimnology 38, 1–12. Bradshaw, E.G., Rasmussen, P., Nielsen, H., Andersen, N.J., 2005. Mid- to Late-Holocene land change and lake development at Dallund Sø, Denmark: trends in lake primary production as reflected by algal and macrophyte remains. The Holocene 15, 1130–1142. Brenner, W.W., 2001a. Distribution of organic walled microfossils single lamina from the Gotland basin, and their environmental evidence. Baltica 14, 34–39. Brenner, W.W., 2001b. Organic walled microfossils from the central Baltic Sea, indicators of environmental change and base for ecostratigraphic correlation. Baltica 14, 40–51. Briffa, K.R., Bartholin, T., Eckstien, D., Jones, P.D., Schweingruber, F.H., Zetterberg, P., 1990. A 1400-year tree-ring record of summer temperatures in Fennoscandia. Nature 306, 434–439. Bronk Ramsey, C., 2001. Development of the radiocarbon calibration program. Radiocarbon 43, 355–363. Burdige, D.J., 2007. Preservation of organic matter in marine sediments: Controls, mechanisms, and an imbalance in sediment carbon budgets? Chemical Reviews 107, 467–485. Burk, I.T., Grigorov, I., Kemp, A.S., 2002. Microfabric study of diatomaceous and lithogenic deposition in laminated sediments from the Gotland Deep, Baltic Sea. Marine Geology 183, 89–105. Burke, I.T., Kemp, A.S., 2004. A mid-Holocene geochemical record of saline inflow to the Gotland Deep, Baltic Sea. Holocene 14, 94–948. Böttcher, M.E., Lepland, A., 2000. Biogeochemistry of sulphur in a sediment core from the west-central Baltic Sea: Evidence from stable isotopes and pyrite textures. Journal of Marine Systems 25, 299–312. Clarke, A.L., Weckström, K., Conley, D.J., Adser, F., Anderson, N.J., Andrén, E., de Jonge, V., Ellegaard, M., Juggins, S., Kauppila, K., Korhola, A., Reuss, N., Telford, R.J., Vaalgamaa, S., 2006. Long-term trends in eutrophication and nutrients in the coastal zone of northwestern Europe. Limnology and Oceanography 51, 385–397. Conley, D.J., Humborg, C., Rahm, L., Savchuk, O.P., Wulff, F., 2002. Hypoxia in the Baltic Sea and Basin-Scale changes in phosphorous and biogeochemistry. Environmental Science and Technology 36, 5315–5320. Conley, D.J., Carstensen, J., Ærtebjerg, G., Christensen, P.B., Dalsgaard, T., Hansen, J.L.S., Josefson, A.B., 2007. Long-term changes and impacts of hypoxia in Danish coastal waters. Ecological Applications 17, S165–S184. Cowie, G.L., Hedges, J.I., Prahl, F.G., de Lance, G.J., 1995. Elemental and major biochemical changes across an oxidation front in a relict turbidite: an oxygen effect. Geochimica Cosmochimica et Acta 59, 33–46. Dahl, S.O., Nesje, A., 1994. Holocene glacier fluctuations at Hardangerjøkulen, centralsouthern Norway: a high-resolution composite chronology from lacustrine and terrestrial deposits. The Holocene 4, 269–277. Diaz, R.J., Rosenberg, R., 1995. Marine benthic hypoxia: a review of its ecological effects and behavioural responses of marine macrofauna. Oceanography and Marine Biology Annual Reviews 33, 245–303. Diaz, R.J., Rosenberg, R., 2008. Spreading dead zones and consequences for marine ecosystems. Science 321, 926–929. Digerfeldt, G., 1988. Reconstruction and regional correlation of Holocene lake-level fluctuations in Lake Bysjön, South Sweden. Boreas 17, 165–182. Dippner, J.W., Voss, M., 2004. Climate reconstruction of the MWP in the Baltic Sea area based on biogeochemical proxies from a sediment record. Baltica 17, 5–16. Donner, J., Knakainen, T., Karhu, A., 1999. Radiocarbon ages and stable isotope composition of Holocene shells in Finland. Quaternaria A 31–38. Elmel´yanov, E.M., Trimonis, E.S., Bostrom, K., Yuspina, L.F., Vaikutene, G., Lei, G., 2001. Sedimentation in West Gotland Basin, Baltic Sea (from the data of core Psh-2537). Oceanology 41, 873–885. Emanuelsson, U., 1993. Vegetationshistoria. In: Johansson, K.R. (Ed.), Stenhuvud nationalparken på Österlen. Statens naturvårdsverk, Stockholm, Sweden. 136 pp. Emeis, K.-C., Struck, U., Blanz, T., Kohly, A., Voß, M., 2003. Salinity changes in the Baltic Sea (NW Europe) over the last 10 000 years. Holocene 13, 411–421. Filipsson, H.L., Nordberg, K., 2004a. Climate variations, an overlooked factor influencing the recent marine environment. An example from Gullmar Fjord, 27. Estuaries, Sweden, pp. 867–880. Filipsson, H.L., Nordberg, K., 2004b. A 200 year environmental record of a low oxygen fjord, Sweden, elucidated by benthic foraminifera, sediment characteristics and hydrographic data. Journal of Foraminiferal Research 34, 277–293. L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 Fjellså, A., Nordberg, K., 1996. Toxic dinoflagellate “blooms” in the Kattegat, North Sea, during the Holocene. Palaeogeography, Palaeoclimatology, Palaeoecology 124, 87–105. Fritz, C.S., 1989. Lake development and limnological response to prehistoric and historic land-use in Diss, Norfolk, UK. Journal of Ecology 77, 182–202. Gerlach, S.A., 1994. Oxygen conditions improve when the salinity in the Baltic decreases. Marine Pollution Bulletin 28, 413–416. Gripenberg, S., 1934. A study of the sediments of the North Baltic and adjoining Seas. Fennia 60, 1–231. Gustafsson, B.G., Westman, P., 2002. On the causes of salinity variations in the Baltic Sea during the last 8500 years. Paleocenography 17, 1–14. Hammarlund, D., Björck, S., Buchardt, B., Israelson, C., Thomsen, C., 2003. Rapid hydrological changes during the Holocene revealed by stable isotope records of lacustrine carbonates from Lake Igelsjön, southern Sweden. Quaternary Science Reviews 22, 195–212. Harff, J., Bohling, G., Davis, J.C., Endler, R., Kunzendorf, H., Olea, A., Schwarzacher, W., Voss, M., 2001. Physico-chemical stratigraphy of Gotland Basin Holocene sediments, the Baltic Sea. Baltica 14, 58–66. Hedenström, A., Possnert, G., 2001. Reservoir ages in Baltic Sea sediment — a case study of an isolation sequence from the Litorina Sea stage. Quaternary Science Reviews 20, 1779–1785. Heinsalu, A., Kohonen, T., Winterhalter, B., 2000. Early postglacial environmental changes in the western Gulf of Finland based on diatom and lithostratigraphy of sediment core B-51. Baltica 13, 51–60. Hille, S., 2005. New aspects of sediment accumulation and reflux of nutrients in the Eastern Gotland Basin (Baltic Sea) and its impact on nutrient cycling. Ph.D. Thesis, der Mathematisch-Naturwissenschaftlichen fakultät, Rostock Univ. Rostock, Germany. Hurrell, J.W., 1995. Decadal trends in the North Atlantic Oscillation: regional temperatures and precipitation. Science 269, 676–679. Ignatius, H., Kukkonen, E., Winterhalter, B., 1968. Notes on a pyretic zone in upper Ancylus sediments from the Bothnian Sea. Bulletin of the Geological Society of Finland 40, 131–134. Jerbo, A., 1961. Den gyttjebandade leran i bottniska sediment: Några geologiska och geotekniska undersökningsresultat. Geologiska Föreningen i Stockholms Förhandlingar 83, 303–311. Jerbo, A., Hall, F., 1961. Några synpunkter på högsensitiva bottniska sediment. Geologiska Föreningen i Stockholms Förhandlingar 83, 312–315. Johansson, S., Wulff, F., Bonsdorff, E., 2007. The MARE Research Program 1999–2006: Reflections on program management. Ambio 36, 119–1222. Jonsson, P., Carman, R., Wulff, F., 1990. Laminated sediments in the Baltic — a tool for evaluating nutrient mass balance. Ambio 19, 152–158. Karlén, W., Kuylenstierna, J., 1996. On the solar forcing of Holocene climate: evidence from Scandinavia. Holocene 6, 359–365. Karlson, K., Rosenberg, R., Bonsdorff, E., 2002. Temporal and spatial large-scale effects of eutrophication and oxygen deficiency on benthic fauna in Scandinavian and Baltic waters — a review. Oceanographic Marine Biology Annual Review 40, 427–489. Karlson, K., Hulth, S., Ringdahl, K., Rosenberg, R., 2005. Experimental recolonization of Baltic Sea reduced sediments: survival of benthic macrofauna and effects on nutrient cycling. Marine Ecology Progress Series 294, 35–49. Koç, N., Jansen, E., Haflidason, H., 1993. Paleoceanographic reconstructions of surface ocean conditions in the Greenland, Iceland and Norwegian seas through the last 14 ka based on diatoms. Quaternary Science Reviews, 12, 115–140. Korhola, A., Vasko, K., Toivonen, H.T.T., Olander, H., 2002. Holocene temperature changes in northern Fennoscandia reconstructed from chironomids using Bayesian modelling. Quaternary Science Reviews 21, 1841–1860. Kortekaas, M., 2007. Post-glacial history of sea-level and environmental change in the southern Baltic Sea. LUNDQUA Thesis 57, Lund University, Sweden. Kotilainen, A.T., Saarinen, T., Winterhalter, B., 2000. High-resolution paleomagnetic dating of sediments deposited in the central Baltic Sea during the last 3000 years. Marine Geology 166, 51–64. Kotilainen, A., Kankainen, T., Ojala, A., Winterhalter, B., 2001. Paleomagnetic dating of a Late Holocene sediment core from the North Central Basin, the Baltic Sea. Baltica 14, 67–73. Kowalewska, G., 2001. Algal pigments in Baltic sediments as markers of ecosystem and climate changes. Climate Research 18, 80–96. Kullman, L., 1999. Early Holocene tree growth at a high elevation site in the northernmost Scandes of Sweden (Lappland). A palaeobiogeographical case study based on megafossil evidence, vol 81. Geografiska Annaler, pp. 63–74. Kunzendorf, H., Voss, M., Brenner, W., 2001. Molbydenum in sediments of the Baltic Sea as an indicator for algal blooms. Baltica 14, 123–130. Kunzendorf, H., Larsen, B., 2002. A 200–300 year cyclicity in sediment deposition in the Gotland Basin, Baltic Sea, as deduced from geochemical evidence. Applied Geochemistry 17, 29–38. Köningsson, L.-K., Possnert, G., 1988. Ancylus fauna studied by accelerator 14C dating of single small shells. In: Winterhalter, B. (Ed.), The Baltic Sea, Geological Survey of Finland Special Paper, vol. 6, pp. 137–145. Lagerås, P., 2006. The ecology of expansion and abandonment. Medieval and postmedieval land-use and settlement dynamics in a landscape perspective. Riksantikvarieämbetet, Sweden. 256 pp. Laine, A., 2003. Distribution of soft-bottom macrofauna in the deep open Baltic Sea in relation to environmental variability. Estuarine Coastal and Shelf Science 57, 87–97. Laine, A., Andersin, A., Leiniö, S., Zuur, A.F., 2007. Stratification-induced hypoxia as a structuring inducing factor in the open Gulf of Finland (Baltic Sea). Journal of Sea Research 57, 65–77. Larsen, C.P.S., Pienitz, R., Smol, J.P., Moser, K.A., Cumming, B.F., Blais, J.M., Macdonald, G.M., Hall, R.I., 1998. Relations between lake morphometry and presence of 91 laminated lake sediments: a re-examination of Larsen and Macdonald (1993). Quaternary Science Reviews 17, 711–717. Larsson, M., Olsson, U., 1992. Industrialiseringens sekel. In: Sveriges ndustriförbund (Ed.), Sveriges industri. Gotab, Stockholm, pp. 17–43. Lepland, A., Stevens, R.L., 1998. Manganese authigenesis in the Landsort Deep, Baltic Sea. Marine Geology 151, 1–25. Likens, G.E., Bormann, F.H., Johnson, N.M., Fisher, D.W., Pierce, R.S., 1970. Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-ecosystem. Ecological Monographs 40, 23–47. Matthews, J.A., Briffa, K.R., 2005. The “Little Ice Age”: re-evaluation of an evolving concept. Geografiska Annaler 87A, 17–36. Matthäus, W., Schinke, H., 1999. The influence of river runoff on deep water conditions in the Baltic Sea. Hydrobiologia 393, 1–10. Matthäus, W., 2006. The history of investigation of salt water inflow into the Baltic Sea — from the early beginning to recent results. Marine Science Reports vol. 65, 1–74. Miltner, A., Emeis, K.C., Struck, U., Leipe, T., Voss, M., 2005. Terrigenous organic matter in Holocene sediments from the central Baltic Sea, NW Europe. Chemical Geology 216, 313–328. Moros, M., Lemke, W., Kuijpers, A., Endler, R., Jensen, J.B., Bennike, O., Gingele, F., 2002. Regressions and transgressions of the Baltic basin reflected by a new highresolution deglacial and postglacial lithostratigraphy for Arkona Basin sediments (western Baltic Sea). Boreas 31, 151–162. Nesje, A., Kvamme, M., 1991. Holocene glacier and climate variations in western Norway: evidence for early Holocene glacier demise and multiple Neoglacial events. Geology 19, 610–612. Nesje, A., Matthews, J.A., Dahl, S.O., Berrisford, M.S., Andersson, C., 2001. Holocene glacier fluctuations of Flatebreen and winter-precipitation changes in the Jostedalsbreen region, western Norway, based on glaciolacustrine sediment records. Holocene 11, 267–280. Neumann, T., Christiansen, C., Clasen, S., Emeis, K.C., Kunzendorf, H., 1997. Geochemical records of salt-water inflow into the deep basins of the Baltic Sea. Continental Shelf Research 17, 95–115. Nordberg, K., Gustafsson, M., Krantz, A.-L., 2000. Decreasing oxygen concentrations in the Gullmar Fjord, Sweden, as confirmed by benthic foraminifera, and the possible association with NAO. Journal of Marine Systems 23, 303–316. Nordberg, K., Filipsson, H., Gustafsson, M., Harland, R., Roos, P., 2001. Climate, hydrographic variations and marine benthic hypoxia in Koljö Fjord, Sweden. Journal of Sea Research 46, 187–200. Nytoft, H.P., Larsen, B., 2001. Triterpenoids and other organic compounds as markers of depositional conditions in the Baltic Sea deep basins during the Holocene. Baltica 14, 95–107. Ojala, A.E.K., 2001. Varved lake sediments in southern and central Finland: long varve chronologies as a basis for Holocene palaeoenvironmental reconstructions. Ph.D. Thesis, Geological Survey of Finland. Espoo, Finland. Olsson, S., Regnéll, J., Persson, A., Sandgren, P., 1997. Sediment-chemistry response to land-use change and pollutant loading in a hypertrophic lake, southern Sweden. Journal of Palaeolimnology 17, 275–294. Petterson, G., 1999. Image analysis, varved lake sediments and climate reconstruction. PhD Thesis, Umeå University, Sweden. Pearl, H.W., Huisman, J., 2008. Blooms like it hot. Science 320, 57–58. Poutanen, E.L., Nikkilä, K., 2001. Carotenoid pigments as tracers of cyanobacterial blooms in recent and post-glacial sediments of the Baltic Sea. Ambio 30, 179–183. Renberg, I., Bindler, R., Bradshaw, E., Emteryd, O., McGowan, S., 2001. Sediment evidence of early eutrophication and heavy metal pollution in Lake Mälaren, Central Sweden. Ambio 30, 496–502. Rößler, D., 2006: Reconstruction of the Littorina Transgression in the Western Baltic Sea. PhD thesis, Ernst-Moritz-Arndt-University Greifswald, Germany. Samuelsson, M., 1996. Interannual salinity variations in the Baltic Sea during the period 1954–1990. Continental Shelf Research 16, 1463–1477. Schoning, K., 2001. The brackich Baltic Sea Yoldia Stage — palaeoenvironmental implications from marine benthic fauna and stable oxygen isotopes. Boreas 30, 290–298. Schybergson, P., 1973. Hantverk och fabriker I. Finlands konsumtionsvaruindustri 18151870: Helhetsutveckling. Societas Scientiarum Fennica, Helsinki. 205 pp. Schybergson, P., 1974. Hantverk och fabriker III. Finland's consumer goods industry, 1815-1870: Statistics. Societas Scientiarum Fennica, Helsinki. 125 pp. Seifert, T., Kayser, B., 1995. A high resolution spherical grid topography of the Baltic Sea. Meereswissenschaftliche Berichte 9, 72–88. Seppä, H., Birks, H.J.B., 2001. July mean temperature and annual precipitation trends during the Holocene in the Fennoscandian tree-line area: pollen based climate reconstructions. The Holocene 11, 527–529. Seppä, H., Hammarlund, D., Antonsson, K., 2005. Low-frequency and high-frequency changes in temperature and effective humidity during the Holocene in South central Sweden: implications for atmospheric and oceanic forcings of climate. Climate Dynamics 25, 285–297. Snowball, I.F., 1993. Geochemical control of magnetite dissolution in sub-artic lake sediments and the implications for environmental magnetism. Journal of Quaternary Science 8, 339–346. Snowball, I.F., 1996. Holocene environmental change in the Abisko region of northern Sweden recorded by the mineral magnetic stratigraphy of lake sediments. GFF 118, 9–17. Snowball, I.F., Thompson, R., 1990. A mineral magnetic study of Holocene sedimentation in Lough Catherine, Northern Ireland. Boreas 19, 127–146. Snowball, I.F., Sandgren, P., 1996. Lake sediment studies of Holocene glacial activity in the Kårsa valley, northern Sweden: contrasting opinions. The Holocene 6, 367–372. 92 L. Zillén et al. / Earth-Science Reviews 91 (2008) 77–92 Snowball, I., Zillén, L., Gaillard, M.J., 2002. Rapid early-Holocene environmental changes in northern Sweden based on studies of two varved lake-sediment sequences. The Holocene 12, 7–16. Snowball, I.F., Korhola, A., Briffa, K.R., Koç, N., 2004. Holocene climate dynamics in Fennoscandia and the North Atlantic. In: Battarbee, R.W., Gasse, F., Stickley, C.E. (Eds.), Past climate variability through Europe and Africa. Springer, Dordrecht, the Netherlands, pp. 465–494. Sohlenius, G., Westman, P., 1998. Salinity and redox alterations in the northwestern Baltic Proper during the late Holocene. Boreas 27, 101–114. Sohlenius, G., Stenbeck, J., Andrén, E., Westman, P., 1996. Holocene history of the Baltic Sea as recorded in a sediment core from the Gotland Deep. Marine Geology 134, 183–201. Sohlenius, G., Emies, K.-C., Andrén, E., Andrén, T., Kohly, A., 2001. Development of anoxia during the Holocene fresh — brackish water transition in the Baltic Sea. Marine Geology 177, 221–242. Sternbeck, J., Sohlenius, G., 1997. Authigenic sulfide and carbonate mineral formation in Holocene sediments of the Baltic Sea. Chemical Geology 135, 55–73. Sternbeck, J., Sohlenius, G., Hallberg, R.O., 2000. Sedimentary trace elements as proxies to depositional changes induced by a Holocene fresh-brackish transition. Aquatic Geochemistry 6, 325–345. Stigebrandt, A., 2001. Physical oceanography of the Baltic Sea. In: Wulff, F., Rahm, L., Larsson, P. (Eds.), A systems analysis of the Baltic Sea. Springer Verlag, pp. 19–74. Stigebrandt, A., Gustafsson, B., 2003. The response of the Baltic Sea to climate change — theory and observations. Journal of Sea Research 49, 243–256. Stigebrandt, A., Gustafsson, B., 2007. Improvement of Baltic Proper water quality using large-scale ecological engineering. Ambio 36, 280–286. Strömberg, B., 1992. The final stage of the Baltic Ice Lake. In: Robertsson, A.-M., Ringberg, B., Miller, U., Brunnberg, L. (Eds.), Late Quaternary stratigraphy, glacial morphology and environmental changes, 81. Sveriges Geologiska Undersökning Ca, pp. 347–354. Svensson, N.-O., 1991. Late Weichselian and early Holocene shore displacement in the central Baltic Sea. Quaternary International 9, 7–26. Thompson, J., Higgs, N.C., Wilson, T.R.S., Croudace, I.W., de Lange, G.J., van Santvoort, P.J.M., 1995. Redistribution of geochemical behavior of redox-sensitive elements S1, the most recent eastern Mediterranean sapropel. Geochimica et Cosmochimica Acta 59, 3487–3501. Thorsen, T., Dale, B., Nordberg, K., 1995. ‘Blooms’ of the toxic dinoflagellate Gymnodinium catenatum as evidence of climatic fluctuations in the Late Holocene of southwestern Scandinavia. The Holocene 5, 435–446. Thulin, B., Possnert, G., Vuorela, I., 1992. Stratigraphy and age of two postglacial sediment cores from the Baltic Sea. GFF 114, 165–179. Vahtera, E., Conley, D.J., Gustaffson, B., Kuosa, H., Pitkanen, H., Savchuk, O., Tamminen, T., Wasmund, N., Viitasalo, M., Voss, M., Wulff, F., 2007. Internal ecosystem feedbacks enhance nitrogen-fixing cyanobacteria blooms and complicate management in the Baltic Sea. Ambio 36, 186–194. Virtasalo, J.J., Kotilainen, A.T., Räsänen, M.E., 2005. Holocene stratigraphy of the Archipelago Sea, northern Baltic Sea; the definitions and descriptions of the Dragsfjärd, Korppoo and Nauvo Alloformations. Baltica 18, 83–97. Virtasalo, J.J., Kotilainen, A.T., Gingras, M.K., 2006. Trace fossils as indicator of environmental change in Holocene sediments of the Archipelago Sea, northern Baltic Sea. Palaeoceanography, Palaeoclimatology, Palaeoecology 240, 453–467. Voss, M., Kowalewska, G., Brenner, W., 2001. Microfossil and biogeochemical indicators of environmental changes in the Gotland Deep during the last 10 000 years. Baltica 14, 131–140. Wastegård, S., Andrén, T., Sohlenius, G., Sandgren, P., 1995. Different phases of the Yoldia Sea in the Northwestern Baltic proper. Quaternary International 27, 121–129. Westman, P., Sohlenius, G., 1999. Diatom stratigraphy in five offshore sediment cores from the northwestern Baltic proper implying large scale circulation changes during the last 8500 years. Journal of Paleolimnology 22, 53–69. Widerlund, A., Andersson, P.S., 2006. Strontium isotopic composition of modern and Holocene mollusc shells as a palaeosalinity indicator for the Baltic Sea. Chemical Geology 232, 54–66. Witkowski, A., 1994. Recent and fossil diatom flora of the Gulf of Gdansk, Southern Baltic Sea. Origin, composition and changes of diatom assemblages during the Holocene. Bibliotheca Diatomologica, vol. 28. J. Cramer, Berlin, Stuttgart. 312 pp. Witkowski, A., Pempkowiak, J., 1995. Reconstructing the development of human impact from diatoms and 210Pb sediment dating (the Gulf of Gdansk-southern Baltic Sea). Geographia Polonica 65, 63–78. Witkowski, A., Broszinski, A., Bennike, O., Janczak-Kostecka, B., Jensen, J.B., Lemke, W., Endler, R., Kuijpers, A., 2005. Darss Sill as a biological border in the fossil record of the Baltic Sea: evidence from diatoms. Quaternary International 130, 97–109. Wulff, F., Stigebrandt, A., Rahm, L., 1990. Nutrient dynamics of the Baltic Sea. Ambio 19, 126–133. Wulff, F., Savchuk, O.P., Sokolov, A., Humborg, C., Mörth, C.-M., 2007. Management options and effects on a marine ecosystem: assessing the future of the Baltic. Ambio 36, 243–249. Yu, S.Y., Andrén, E., Barnekow, L., Berglund, B.E., Sandgren, P., 2003. Holocene palaeoecology and shoreline displacement on the Biskopsmala Peninsula, southeastern Sweden. Boreas 32, 578–589. Yu, S.Y., Berglund, B.E., Sandgren, P., Lambeck, K., 2007. Evidence for a rapid sea-level rise 7600 yr ago. Geology 35, 891–894. Zillén, L., 2003. Setting the Holocene clock using varved lake sediments in Sweden. LUNDQUA Thesis 50, Lund University, Sweden. Zillén, L., Snowball, I., Sandgren, P., Stanton, T., 2003. Occurrence of varved lake sediment sequences in Värmland, west central Sweden: lake characteristics, varve chronology and AMS radiocarbon dating. Boreas 4, 612–626. Zorita, E., Laine, A., 2000. Dependence of salinity and oxygen concentrations in the Baltic Sea on large-scale atmospheric circulation. Climate Research 14, 25–41. Åker, K., Eriksson, B., Grönlund, T., Kankainen, T., 1988. The Baltic Sea. In: Winterhalter, B. (Ed.), Sediment stratigraphy in the northern Gulf of Finland. Geological Survey of Finland Special Paper 6, pp. 101–117. Österblom, H., Hansson, S., Larsson, U., Hjerne, O., Wulff, F., Elmgren, R., Folke, C., 2007. Human-induced thropic cascades and ecological regime shifts in the Baltic Sea. Ecosystems 2007, 1–13.