http://www.diva-portal.org
This is the published version of a paper published in Palaeogeography, Palaeoclimatology,
Palaeoecology.
Citation for the original published paper (version of record):
Andrén, E., van Wirdum, F., Norbäck Ivarsson, L., Lönn, M., Moros, M. et al. (2020)
Medieval versus recent environmental conditions in the Baltic Proper, what was
different a thousand years ago?
Palaeogeography, Palaeoclimatology, Palaeoecology, 555: 109878
https://doi.org/10.1016/j.palaeo.2020.109878
Access to the published version may require subscription.
N.B. When citing this work, cite the original published paper.
Attribution 4.0 International (CC BY 4.0)
Permanent link to this version:
Contents lists available atScienceDirect
Palaeogeography, Palaeoclimatology, Palaeoecology
journal homepage:www.elsevier.com/locate/palaeoMedieval versus recent environmental conditions in the Baltic Proper, what
was different a thousand years ago?
Elinor Andrén
a,⁎, Falkje van Wirdum
a, Lena Norbäck Ivarsson
a, Mikael Lönn
a, Matthias Moros
b,
Thomas Andrén
aaSödertörn University, School of Natural Sciences, Technology and Environmental Studies, Huddinge, Sweden bLeibniz Institute for Baltic Sea Research Warnemünde, Rostock, Germany
A R T I C L E I N F O Keywords: Pseudosolenia calcar-avis Baltic Sea Diatoms Phytoplankton seasonality Medieval Climate Anomaly Hypoxia
A B S T R A C T
A sediment record from the western Gotland Basin, northwestern Baltic Proper, covering the last 1200 years, was investigated for past changes in climate and the environment using diatoms as a proxy. The aim is to compare the environmental conditions reconstructed during Medieval times with settings occurring the last century under influence of environmental stressors like eutrophication and climate change. The study core records more marine conditions in the western Gotland Basin surface waters during the Medieval Climate Anomaly (MCA; 950–1250C.E.), with a salinity of at least 8 psu compared to the present 6.5 psu. The higher salinity together with a strong summer-autumn stratification caused by warmer climate resulted in extensive long-lasting diatom blooms of Pseudosolenia calcar-avis, effectively enhancing the vertical export of organic carbon to the sediment and contributing to benthic hypoxia. Accordingly, our data support that a warm and dry climate induced the extensive hypoxic areas in the open Baltic Sea during the MCA. During the Little ice Age (LIA; 1400–1700C.E.), the study core records oxic bottom water conditions, decreasing salinity and less primary production. This was succeeded during the 20th century, about 1940, by environmental changes caused by human-induced eu-trophication. Impact of climate change is visible in the diatom composition data starting about 1975C.E. and becoming more pronounced 2000C.E., visible as an increase of taxa that thrived in stratified waters during autumn blooms typically due to climate warming.
1. Introduction
The present Baltic Sea ecosystem shows significant changes related to climate change, for instance an extension of the growing season due to earlier warming in spring and delayed cooling in autumn (Kahru et al., 2016). These conditions have resulted in a tripling of the net primary production and shift in the annual biomass maximum from spring to summer (Kahru et al., 2016). Hypoxia has been identified as one of the most serious threats to coastal oceans worldwide (Breitburg et al., 2018), and the Baltic Sea is considered the largest dead zone in the world (Diaz and Rosenberg, 2008). The pronounced salinity stra-tification in the Baltic Basin leads to oxygen deficiency in the bottom water and a tenfold increase in benthic hypoxia has been recorded between 1898 and 2012 (Carstensen et al., 2014a). Although hypoxia is considered a natural phenomenon in the deeper parts of the Baltic Proper, it displays an increasing trend since the 1950's in response to eutrophication caused by increased discharge of nutrients from land
and atmosphere (Carstensen and Conley, 2019; Meier et al., 2019). Further, model simulations show increased areal extent of bottom water hypoxia contributing to more sediment phosphorous release, earlier and more frequent cyanobacterial blooms and to increased supply of surface water nitrogen available for phytoplankton primary production (Andersson et al., 2015;Karlson et al., 2015). On decadal scale there is high correlation between the extent of hypoxic deep areas and fre-quency of cyanobacterial accumulation (Kahru et al., 2020). Both eu-trophication and climate warming amplify the effect of low oxygen conditions at the seafloor, which promotes phosphorous release from the sediment and stimulate cyanobacteria which further enhance eu-trophication by nitrogen fixation (Vahtera et al., 2007;Karlson et al., 2015; Carstensen and Conley, 2019). By using cyanobacterial bio-markers and geochemical proxy methods, it has been shown that cya-nobacteria and hypoxia were widespread in the Baltic Proper also during Medieval times about 550–1250C.E. (Funkey et al., 2014). Causes for the Medieval hypoxia have been discussed (Schimanke et al.,
https://doi.org/10.1016/j.palaeo.2020.109878
Received 16 March 2020; Received in revised form 15 June 2020; Accepted 15 June 2020 ⁎Corresponding author.
E-mail addresses:elinor.andren@sh.se(E. Andrén),lena.norback.ivarsson@sh.se(L. Norbäck Ivarsson),matthias.moros@io-warnemuende.de(M. Moros), thomas.andren@sh.se(T. Andrén).
Available online 17 June 2020
0031-0182/ © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/).
2012;Carstensen et al., 2014b). Different drivers have been suggested such as climate warming triggering intense cyanobacterial blooms (Kabel et al., 2012), or increased discharge of nutrients from the coastal zone caused by early human impact with changes in agricultural practice, intensified land-use, and increased population density (Zillén et al., 2008;Zillén and Conley, 2010).
Recent conditions are characterized by a typical pattern of seasonal succession in the Baltic Sea primary production, showing an annual biomass maximum during the spring bloom when light conditions im-prove, the available nutrients are high and there are limited numbers of grazers (Andersson et al., 2017). The phytoplankton spring bloom in the Baltic Proper is nitrogen limited (Granéli et al., 1990). Diatoms and dinoflagellates are the dominating phytoplankton during spring bloom and the respective share varies between years, associated primarily with weather conditions (Klais et al., 2011; Carstensen et al., 2015; Andersson et al., 2017). After cold winters with thick and long-lasting ice cover diatoms dominate the spring bloom, and after mild winters with storms and thin ice-cover dinoflagellates dominate (Klais et al., 2013). In the northwestern Baltic Proper the relative proportion of diatoms and dinoflagellates is linked to phases of the North Atlantic Oscillation (NAO), where dinoflagellates dominate spring blooms during positive NAO phases (Klais et al., 2011). Another phytoplankton biomass maximum occurs in late summer when nitrogen-fixing cyano-bacteria dominate, especially during calm and warm water conditions (Andersson et al., 2017). The cyanobacterial blooms have become more frequent and start earlier during the last decades (Kahru and Elmgren, 2014;Kahru et al., 2016). In autumn there is a second occasion with high diatom abundance in certain years, before light and temperature limits phytoplankton production in winter, except for sympagic (ice-dependent) diatoms living associated with sea-ice (Andersson et al., 2017;Thomas et al., 2017).
In this paper we use a paleoecological approach, using diatoms as a proxy, to study and reconstruct past changes in climate and the en-vironment in the northwestern Baltic Proper to extend beyond the time covered by instrumental records and environmental monitoring. A paleo perspective gives the range of natural variation and can indicate the speed and trajectories of change, which allows present-day ob-servations and ecosystem function to be viewed in a long-term per-spective (Willis and Birks, 2006;Saunders and Taffs, 2009). Learning from the past has the potential to improve anticipation of future shifts and increase the effectiveness of conservation practice (Davies and Bunting, 2010).
We have studied a core from the western Gotland Basin for changes in primary production, stratification, salinity conditions and bottom water oxygenation. The aim is to compare the environmental conditions reconstructed during Medieval times (including the warm Medieval Climate Anomaly (MCA), defined as 950–1250C.E. (Mann et al.Reimer et al., 2009)) with conditions of the last century under influence of environmental stressors like eutrophication and climate warming. Specific focus is put on Medieval conditions and the cause of the re-corded high primary productivity and extended areas of benthic hy-poxia. Changes in diatom species composition are used as a proxy for changes in salinity, nutrients, stratification and bloom seasonality. Si-liceous microfossil absolute abundances together with total carbon content are used as proxies for changes in primary productivity and lithological changes are used as an indication of oxygen status of the seafloor (laminated sediments indicate hypoxia). Above all, we here present, based on paleoecological data, an alternative model to inter-pret the causes of widespread hypoxia during the Medieval Climate Anomaly.
Fig. 1. A. Overview map and bathymetry of the Baltic Sea which consists of Baltic Proper, Gulf of Bothnia and Gulf of Finland and is delimited by the threshold areas in the Danish straits (Ds). Bathymetric data from EMODnet 2018. B. Detailed map showing the cored station MSM62–1-60 in red (gravity corer 57°58.64’N, 17°57.37′E, water depth 218.6 m and multicorer MUC 57°58.73’N, 17°57.56′E, water depth 203 m) and other stations mentioned in the text in yellow: BY31 (Landsort Deep58°35′N, 18°14′E, water depth 459 m (Höglander et al., 2004;Hjerne et al., 2019)), LL19 (Northern Gotland Basin 58.8807°N, 19.8968°E, water depth 169 m (Funkey et al., 2014)), F80 (Fårö Deep, 58.0000°, 19.8968°E, water depth 191 m (Funkey et al., 2014)). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
2. Material and methods
2.1. Site description
We use the terminology of different Baltic Sea sub-basins defined in Leppäranta and Myrberg (2009). The Baltic Sea is located in-between marine air masses from the North Atlantic in the west and Russian continental climate in the east and climate depends on the location of the Polar front and the strength of the westerlies, which makes the Baltic Sea susceptible to the North Atlantic Oscillation (NAO) (Leppäranta and Myrberg, 2009). A positive NAO index results in pre-vailing mild westerly winds and warm winters and a negative NAO index in cold northerly and easterly winds resulting in low winter temperatures (Leppäranta and Myrberg, 2009). The mean Baltic Sea salinity is strongly related to large-scale atmospheric variability and the accumulated freshwater inflow. There is a large freshwater surplus in the northern and eastern parts of the Baltic Sea which leads to estuarine salinity gradient from the more saline water in the southwestern Baltic Sea to the fresher water in the Gulf of Bothnia and the Gulf of Finland (HELCOM, 2007). The narrow and shallow thresholds at the Danish straits (Fig. 1) constrain the water exchange with the North Sea and are, together with the positive water balance due to freshwater discharge, the main reason why the Baltic Sea is brackish (HELCOM, 2007; Snoeijs-Leijonmalm and Andrén, 2017). The deep water of the Baltic Proper is replaced via episodic inflows of larger volumes of highly saline (17–25 psu) and oxygen-rich water from the North Sea which due to the strong density stratification is the only source for deep water ventilation (HELCOM, 2007;Mohrholz et al., 2015).
The western Gotland Basin is a part of the Baltic Proper delimited by the Gotska Sandö Sill at 100 m water depth in the northeast and the Hoburg-Midsjö Banks at 40 m depth in the south (Fig. 1). The western Gotland Basin has an area of 34,232 km2and contains the deepest place in the Baltic Sea, the Landsort Deep 459 m, but has a mean depth of only 71 m (Leppäranta and Myrberg, 2009). A permanent halocline at 60–80 m depth divides the water masses into an upper layer with a salinity of 6.3–7.7 psu and a lower layer with a salinity of 8.7–10.3 psu, except for the Landsort Deep where the bottom water salinity is 10–11.5 psu (Leppäranta and Myrberg, 2009). The variation in halo-cline depth is considerable and connected to variations in season with, in general, the deepest halocline in winter and spring (Meier, 2007). Data from monitoring station BY31 in the Landsort Deep based on 411 observations between 1980 and 1998 show a surface and bottom water salinity of about 6.4–6.9 psu and about 9.9–10.2 psu, respectively (Omstedt and Axell, 2003). In summer, a thermocline develops creating a warmer 10–20 m surface layer. The western Gotland Basin contains no major river outflows but the normal counter-clockwise water cir-culation in the Baltic Basin (Meier, 2007) results in a transport of e.g. salt, nutrients, contaminants and affects the dispersal of organisms from other areas (Snoeijs-Leijonmalm and Andrén, 2017). The water renewal time for the Gotland Basin is 28–34 years and in the bottom water of the western Gotland Basin over 36 years (Meier, 2007). The sea bottom area of this study in the western Gotland Basin has been hypoxic or anoxic (oxygen concentrations below 2 mgL−1or 0 respectively) during the last century reaching unprecedented areal extent in recent years (Carstensen et al., 2014a).
2.2. Sampling
A long (MSM62-1-60-4, 794 cm) and a short (MSM62-1-60-2, 42 cm) sediment core were collected March 2017 from RV Maria S. Merian using different gravity core devices to ensure a continuous se-diment record from the western Gotland Basin (position long gravity corer 57°58.64′N, 17°57.37′E, water depth 218.6 m and short multi-corer 57°58.73′N, 17°57.56′E, water depth 203 m, Fig. 1). The short multicorer (MUC) was used to ensure sampling of the uppermost soft sediments. The sediment cores were kept in a cold room and
transported to the laboratory at the Leibniz Institute for Baltic Sea Research in Warnemünde, Germany for opening and sub-sampling. The long core was opened lengthwise and subsampled for carbon analyses in 1-cm intervals and for paleoecological analyses in 3-cm contiguous intervals and subsequently freeze dried. The short multicore was sub-sampled contiguous in 0.5-cm intervals in the top 16 cm and then every cm and thereafter freeze-dried.
2.3. Methods
2.3.1. Dating and age modeling
Six bulk sediment samples were submitted for AMS14C measure-ments at the Poznań Radiocarbon Laboratory, Poland. The base-ex-tractable organic carbon fraction was radiocarbon-dated. Due to the complex geological history of the Baltic Basin (Andrén et al., 2011) the pathway for the carbon in the Baltic Sea sediments is complicated with input of resuspended organic sediments of both marine and terrestrial origin being transported from the catchment area and the shores and redeposited as “old” carbon in the deeper basins (Yemelyanov et al., 1995).Moros et al. (2020) suggest that during colder periods a re-suspension of organic carbon-rich laminated sediments deposited in shallow areas during preceding warm periods might be re-mobilized and redeposited in deeper parts of the Baltic Sea sub-basins. Given this, and the fact that no datable macrofossils were at hand in our core, we had to date the sediment bulk samples. As no robust model for dating Baltic Sea sediment bulk samples has been published, we tentatively applied the following model.
An age-model was constructed in order to estimate the age of non-dated levels using the age-modeling software CLAM version 2.2 (Blaauw, 2010) by the means of applying a smooth spline function curve with a 3rd-order smoothing between dated levels and running 10,000 iterations. To compensate for the constantly ongoing re-suspension and redeposition of older sediments, we have used a mixed calibration curve consisting of 30% IntCal 13 and 70% Marine13 (Reimer et al., 2013), i.e. 30% carbon of terrestrial origin and 70% of marine origin. A mix like this is based on the data published by Blomqvist and Larsson (1994)where they measured during a five-year period the gross sedimentation rate and the ratio between primary settling matter and resuspended sediment in Himmerfjärden, a part of southern Stockholm archipelago. They concluded that the resuspended portion commonly exceeded 50%, and periods without resuspension rarely or never occurred in the investigated area. Their investigation was, however, carried out in the coastal area but our core is from the offshore open Baltic Sea and we therefore assume a more conservative mix of 30/70% in our age-depth model construction.
In order to validate the resulting age-depth model we compared our total carbon data to published data which used an independent geo-chronology based on paleomagnetic secular variation and atmospheric lead deposition (Lougheed et al., 2012;Funkey et al., 2014). It is ob-vious that there is an offset of 50 years during the MCA, i.e. the sharp change dates 1200C.E. in our record and 1250C.E. in Funkey et al. (2014). This time lag is within the dating error in our age model but needs to be considered when comparing the two records inFig. 6.
Two additional age fixpoints have been included in the age-model in the upper part of the sediment core using the artificial radionuclides cesium-137 and americium-241. Analyses of radionuclides (137Cs and 241Am) were performed at the Leibniz Institute for Baltic Sea Research by gamma spectrometry with a Canberra Ge-well detector GCW4021-7500SL-RDC-6-ULB. The assignment of the event stratigraphic markers followed the approach for Baltic Sea sediments described inMoros et al. (2017).
2.3.2. Geochemistry and core correlation
In the present study the short multicore and the long gravity core were correlated based on geochemical measurements (total carbon and mercury content). Analyses of total carbon (TC) and mercury (Hg) were
performed at the Leibniz Institute for Baltic Sea Research. The total carbon content was determined using an EA 1110 CHN analyser from CE Instruments. In the western Gotland Basin carbonate content in the sediment is negligible and the total carbon content is assumed to re-flects the total organic carbon (Leipe et al., 2011;Norbäck Ivarsson et al., 2019). Hg was measured using a DMA-80 Analyser from MLS company (Leipe et al., 2013). The core correlation discovered a 10-cm offset between the cores (Fig. S1) where the top 10 cm is missing in the long gravity core. A complete splice sediment record was constructed by adding 10 cm to the long gravity core depth, and the two sediment cores are hereafter treated as one sequence.
2.3.3. Siliceous microfossils
Small amounts (~100–200 mg) of freeze-dried material were weighed and prepared for diatom analysis according to the method described in Battarbee (1986). To be able to calculate the siliceous microfossil concentration a known volume of microsphere solution was added to the last step after cleaning and decanting (Battarbee and Kneen, 1982). The cleaned samples together with microspheres were dried onto coverslips and mounted in Naphrax™. Quantitative diatom analyses were carried out with a light microscope Olympus BX51 using Nomarski differential interference contrast at a magnification of x1000 using oil immersion. At least 300 diatom valves were counted in each sample excluding Skeletonema marinoi and the vegetative cells and resting spores of Chaetoceros spp. The reasons for exclusions from the relative abundance base sum were mass occurrence in some levels which could result in misleading interpretation and overshadow all ongoing diatom successions in the stratigraphy. These taxa are instead reported as concentrations (absolute abundance in valves per g dry sediment). Further, Chaetoceros spp. and S. marinoi are very lightly si-licified which might result in uneven preservation, especially since the lithology shifts between laminated and homogeneous sediments in-dicating fluctuating oxic conditions at the seafloor. The floras used for diatom identification and classification into salinity requirements and life form (planktic versus benthic) areCleve-Euler (1951–1955),Hasle and Syvertsen (1997), Krammer and Lange-Bertalot (1986–1991), Snoeijs et al. (1993–1998)andWitkowski et al. (2000), and taxonomic names are updated according to AlgaeBase (www.algaebase.org/). All taxa except Chaetoceros spp. were if possible identified to species level. To make a diatom stratigraphical zonation of the diatom assem-blages, a cluster analysis was carried out on diatom taxa with a relative abundance of more than 3% in any sample using CONISS in the Tilia 2.1.1 software. CONISS uses the incremental sum of squares method which is an agglomerative algorithm that places clusters in a hierarchy (Grimm, 1987). A stratigraphically constrained analysis was applied to allow only stratigraphically adjacent clusters to be considered for merging, and a square root transformation of the data was used to in-crease the influence of rarer taxa (Grimm, 1987). To record the overall compositional change in the diatom assemblage a Detrended Corre-spondence Analysis (DCA) was calculated using the whole dataset where identification could be made to species level (except S. marinoi) on the total diatom dataset using the package rioja and vegan in R 3.6.1 (Juggins, 2015; Oksanen et al., 2015). Diatom species richness was calculated on the same data set as DCA, using rarefaction analysis in the vegan package (Birks and Line, 1992;Oksanen et al., 2015).
3. Results
3.1. Age-depth modeling
The radiocarbon dating results are presented inTable 1and used together with artificial radionuclides to construct a chronology of the splice sediment record (Fig. 2). The137Cs peak at 5.25 cm below sea-floor (bsf) is associated with the Chernobyl nuclear accident in 1986C.E. and the 241Am peak at 8.25 cm bsf is linked to the nuclear weapons testing in 1954C.E. The age-depth model constructed (Fig. 2)
cover the time span from c. 750C.E. to the year of sampling 2017. The mean sediment accumulation rate is fairly steady and varies between 1.5 mm/yr and 3 mm/yr from the top of the core down to c. 230 cm composite depth bsf.
3.2. Siliceous microfossil stratigraphy
Altogether 60 samples (13 short multicore +47 long gravity core) were counted for siliceous microfossils and 116 diatom taxa were identified together with the siliceous cysts of chrysophytes and the ebridian Ebria tripartita. Two diagrams illustrating the compositional change in the relative abundance of diatoms were constructed (Figs. 3–4), and the cut-off selection of 3% result in the display of 98–85% of the recorded taxa at different levels (Fig. 3). All other si-liceous microfossil data (absolute abundances, diatom pelagic to benthic ratio (P/B ratio), diatom species richness) is plotted inFig. 5.
A very high degree of siliceous microfossil preservation is visible in the laminated parts of the core reflected as a high abundance of the very delicate taxa S. marinoi and the vegetative valves of Chaetoceros spp. The homogeneous part (156–10.5 cm) contained slightly less preserved valves. Based on the cluster analysis (Figs. 3 and 4) and the DCA 1 sample scores (Fig. 6) the diatom stratigraphy was divided into four diatom assemblage zones (DAZ 1–4) characterized as follows:
DAZ 1 (234–186 cm; 750–1100 yr C.E.) is dominated by the marine pelagic taxon Pseudosolenia calcar-avis and brackish water pelagic taxa Pauliella taeniata, Cyclotella choctawhacheeana and Thalassiosira le-vanderi. The zone has also maximum occurrence of the marine pelagic taxon Thalassiosira eccentrica and the brackish-freshwater taxon Fragilariopsis cylindrus. Maximum absolute abundance of vegetative diatom valves and Chaetoceros spp. resting spores and other siliceous microfossils such as chrysophyte cysts and Ebria tripartita are recorded (Fig. 5). The delicate silicified vegetative valves of Chaetoceros spp. and S. marinoi display mass occurrence in this zone. The P/B ratios are high: minimum 12.7 -(mean 19.0)- maximum 39.1 and species richness low: minimum 15.1- (mean 16.6) - maximum 20.8.
DAZ 2 (186–11 cm; 1100–1940 yr C.E.) shows increases in brackish pelagic water taxa at the expense of marine pelagic taxa. The brackish taxa C. choctawhacheeana and T. levanderi attain high and fluctuating percentages in the zone. Thalassiosira hyperborea var. lacunosa increases upwards and Actinocyclus octonarius including var. crassus has a peak. Thalassiosira proschkinae occurs almost exclusively in this zone. P. tae-niata slowly decreases upwards the zone except for a peak of 68% in the topmost sample. There are sharp decreases in all siliceous microfossil absolute abundances in the lower part of the zone which correlate with a shift from laminated to homogeneous sediments (Fig. 5). The P/B ratio is much lower: 2.0 -(mean 9.7)- 27.7 and species richness show maximum levels: 12.8- (mean 22.8) -29.0.
DAZ 3 (11–2 cm; 1940–2002 yr C.E.) starts with a maximum oc-currence of C. choctawhacheeana reaching over 60% followed by peaks in T. hyperborea var. lacunosa ~ 33% and A. octonarius var. crassus 50%. There is a first occurrence of the large autumn blooming brackish water taxon Coscinodiscus granii. P. calcar-avis has more or less disappeared and there is a huge decrease in P.taeniata. An increase in the absolute abundance of Chaetoceros spp. resting spores and Ebria tripartita is visible. The P/B ratios range between 3.8 -(mean 8.8)- 17.3 which are slightly less than DAZ 2. Species richnesses are 15.6 (mean 20.5) -24.7 displaying lower values than DAZ 2 but not as low as in DAZ 1.
DAZ 4 (2–0 cm; 2002–2017 yr C.E.) is characterized by peaks in C. granii reaching 36% and A. octonarius var. crassus and var. tenellus, Thalassiosira baltica and Melosira arctica on the expense of T. hy-perborean var. lacuosa, T. levanderi, C. choctawhacheeana and P. taeniata. Increases in absolute abundances of Chaetoceros spp. resting spores, Ebria tripartita and crysophyte cysts are visible (Fig. 5). The P/B ratios are 4.8 -(mean 9.3)- 12.9 and species richness are 16.5 (mean 20.4) -25.5 which is both similar to DAZ 3.
3.3. Total carbon and lithology
The lower part of the splice record 231–164 cm displays mm-scale laminated sediments with a thickness of about 1 to 5 mm. This section has a total carbon content of 4.8- (mean 9.1) -13.3% (Fig. 5). The middle part of the stratigraphic section 164–10.5 cm shows homo-geneous sediments with a low total carbon content of 1.7- (mean 2.5) -4.7%. The uppermost sediment section 10.5–0 cm displays laminated sediments on mm-scale with a total carbon content of 3.2- (mean 7.4) -14.8%.
4. Discussion
4.1. Baltic Sea during medieval times and the transition to Little Ice Age 4.1.1. Paleosalinity
We use the relative abundance of marine diatom taxa to make in-terpretations on changes in surface water salinity in the northwestern Baltic Proper. Based on the diatom composition a higher surface water salinity than today is recorded during Medieval times. Not only Pseudosolenia calcar-avis (up to 46%) but the marine pelagic taxon Thalassiosira eccentrica shows maximum relative abundance to up to about 6% until 1100C.E. (Fig. 3). T. eccentrica is a taxon rarely recorded in the present Baltic Sea, but commonly occurring in the planktic monitoring samples from Kattegat in a salinity about 16–25 psu (Snoeijs and Potapova, 1995). Pseudosolenia calcar-avis is considered cosmopo-litan and found in warm tropical and subtropical waters, occasionally occurring also in temperate waters (Sundström, 1986). It is a marine taxon, reported as only occurring as fossil in Baltic Sea sediments (Snoeijs and Kasperovičienė, 1996), recorded as soon as brackish con-ditions were established in the basin. P. calcar-avis dominated during the Littorina Sea stage (Yemelyanov et al., 1995), c. 7500–7000 years B.P. until about 1100–1200C.E., when a sharp de-crease occurred (Andrén et al., 2000a), which agrees with our presented data (Fig. 3). Similar diatom compositional changes during Medieval times (e.g. a peak in P. calcar-avis succeeded by Thalassiosira baltica and Actinocyclus octonarius) have been recorded in the study area previously (Westman and Sohlenius, 1999), but is difficult to compare with the present study due to lack of a proper age model and low number of counted taxa. Another study from the Landsort Deep record no changes in salinity during the MCA, possibly due to a lower sample resolution (van Wirdum et al., 2019). A higher bottom water salinity is recorded in the Bornholm Basin during Medieval times evident from the benthic foraminifera record interpreted as frequent inflows of saline water 650 to 1200C.E. (Binczewska et al., 2018), but not strong enough to reach the eastern Gotland Basin (Kotilainen et al., 2014).Kotilainen et al. (2014)instead record increased marine bottom water inflows during the transition from MCA to LIA, possibly linked to unstable North Atlantic Oscillation (NAO) conditions. But scrutinizing diatom data reflecting surface water conditions from the Bornholm Basin and eastern Gotland Basin an apparent similarity with the present data is visible as a peak in P. calcar-avis and T. eccentrica indicative of higher surface water salinity about 2000–1000 years B.P. (Andrén et al., 2000a, 2000b). Further, it is evident that the more long-term trend in surface water salinity in the Baltic Proper peaked about 6000–5000 years B.P. (recorded as peaks in Chaetoceros mitra, Tha-lassionema nitzschioides, Shionodiscus oestrupii and the silicoflagellate Octactis speculum) and thereafter decreased (Andrén et al., 2000a). This maximum surface water salinity has been estimated to about 12–13 psu in the Baltic Proper (Widerlund and Andersson, 2011).
Unfortunately, we are not able to quantify our recorded salinity changes using the transfer function based on the modern pan-Baltic diatom training set used in Warnock et al. (2020), since there is no modern analogue in the Baltic Sea to our dominating marine taxon P. calcar-avis. Although considered a marine taxon, P. calcar-avis is
Table 1
Bulk sediment radiocarbon dating results from western Gotland Basin core MSM62–1–60-4.
Lab ID Depth
bsf (cm) Composite depthbsf (cm)
14C age
(yr B.P.) Error Calibrated age(yr C.E.) Min(yr C.E.) Max(yr C.E.)
Poz-104405 155,5 165,5 1115 30 1169 1075 1262 Poz-99783 170,5 180,5 1180 30 1112 1036 1187 Poz-110740 183 193 1220 30 1088 1018 1158 Poz-104,506 195 205 1235 30 1075 1001 Poz-104508 210 220 1375 30 911 823 999 Poz-104511 220 230 1490 30 789 701 876
Fig. 2. Age-depth plot based on two isotopic markers (red dots) in the top (upper left panel: increase in americium-241,241Am in blue, indicating the first nuclear tests in 1954 and the increase in cesium-137,137Cs in black, indicating the outburst from the Chernobyl accident in 1986) and six radiocarbon dates (green dots including 2 sigma). The grey shaded area indicates the uncertainty (2 sigma) in the age modeling. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 3. Diatom stratigraphy plotted on linear age scale and a depth scale and lithology shown to the left. Diatom taxa with an abun-dance greater than 3% in any sample are plotted, categorized into salinity requirements (marine, brackish-marine, brackish, brackish-freshwater and freshwater taxa) and life forms (pelagics showed in blue and benthics in red) and displayed in order of their first appearance. A cumulative summary graph shows relative abundance in salinity requirements of all counted diatom taxa.
Chaetoceros spp. and Skeletonema marinoi are excluded from the
calculation of relative abundance and plotted as absolute abun-dance to the right as well as the cluster analyse (CONISS) which divide the stratigraphy into four diatom assemblage zones (DAZ). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
reported to tolerate salinities as low as 8–10 psu (Karpinsky, 2010) and it is suggested that surface water salinity controls its occurrence in the present Baltic Sea and that paleorecords could be indicators of past marine water inflows (Kaiser et al., 2016). Accordingly, based on our diatom record, we suggest that salinity during Medieval times in the western Gotland Basin was higher, at least 8 psu compared to present-day about 6.4–6.9 psu (Omstedt and Axell, 2003). The Medieval surface water salinity corresponds to the present salinity in the Arkona Basin (Fig. 1A), where sporadic occurrence of P. calcar-avis is recorded in monitoring data today (Kaiser et al., 2016). Currently there is no plausible explanation why P. calcar-avis suddenly re-appeared in the southern Baltic Sea or where it came from (Wasmund et al., 2019). Compared with present-day conditions however, where P. calcar-avis has recently been introduced in the southern Baltic Sea, but not re-corded in the monitoring data from the western Gotland Basin, this absence is most probably due to the taxon's limitation in salinity tol-erance.
Changes in freshwater input from the drainage area has shown to be the main cause of salinity changes in the Baltic Sea during the last 8500 years (Gustafsson and Westman, 2002). To be able to visually compare our proxy records to climate, we have plotted a reconstruction of Northern Hemisphere temperature using both low- and high-re-solution proxy data (Moberg et al., 2005) and the winter temperature in Stockholm (Leijonhufvud et al., 2010) reconstructed from instrumental observations and documents (Fig. 6d). In Europe, NAO is the dominant decadal mode of climate variability (Hurrell, 1995), also affecting coastal oceans like the Baltic Sea (Lehmann et al., 2002;Börgel et al., 2018) and it has a positive correlation especially to the winter tem-perature in Stockholm (Schimanke et al., 2012). Another mode of
variability also affecting European climate is the Atlantic Multidecadal Oscillation (AMO), defined as variability in the North Atlantic sea surface temperature (Kerr, 2000). Modeling of the AMO warm and cold phases in the Baltic Sea area shows significant correlation between a positive phase (warmer) during the MCA, more westerly winds, in-creased precipitation, higher river runoff to the Baltic Sea, and less saline surface waters (Börgel et al., 2018). This is contradicted by the results from this study which show more saline surface waters during the MCA. Studies of tree-rings provide a dendroclimatic precipitation reconstruction which show very dry climate in northern Europe during the MCA (Helama et al., 2009). This has further been confirmed in studies from Finland by the construction of a temperature-precipitation relationship showing warm and dry conditions c. 300–1100C.E. using chironomids and cladocera as proxies (Luoto and Nevalainen, 2018; Fig. 6d) and a wetness reconstruction using plant macrofossils (Väliranta et al., 2007). The drought during the MCA consequently, as previously shown in our diatom data, likely resulted in higher salinity in the Baltic Proper surface waters (Gustafsson and Westman, 2002). 4.1.2. Primary production, thermal stratification and hypoxia
During Medieval times, the western Gotland Basin had high burial of organic carbon in the sediment resulting in laminated mud indicative of hypoxic/anoxic bottom water conditions (Fig. 5). The pigments zeaxanthin and echinone, biomarkers used as proxy for cyanobacterial abundance in the surface waters, are preserved in sediments and show remarkably high productivity during the MCA in the Fårö Deep and northern Gotland Basin (Funkey et al., 2014;Fig. 6c). This recorded high cyanobacterial production during the MCA would intensify ni-trogen fixation and contribute to overall available nutrient Fig. 4. A close-up of the diatom stratigraphy fromFig. 3showing the last ~150 years (relative abundance > 3%). Taxa are categorized into life form (pelagic/ benthic) and pelagic taxa have been divided into bloom seasonality (spring/ summer/ autumn). Salinity requirements for each taxon is indicated as: M marine, B brackish, BF brackish-freshwater. The cluster analysis divides the stratigraphy into diatom assemblage zones (DAZ).
concentrations, as seen in the present Baltic Sea system (Karlson et al., 2015). Once hypoxia was initiated it is a positive feedback loop of a self-supporting system which promotes leakage of phosphorous from the hypoxic seafloor (Zillén and Conley, 2010). Our data further shows that the pelagic diatom production peaked during this period, as well as other recorded primary producers such as chrysophytes (Fig. 5). All chrysophytes form a resting cyst which most likely is an adaptation strategy to survive between seasons for this planktic algae (Duff et al., 1995). Chrysophyte cysts are abundantly preserved in Baltic sediments (Sohlenius et al., 1996;Burke et al., 2002;Andersson et al., 2017) but few cyst morphotypes have been linked to the producing species (Duff et al., 1995). In the Baltic Sea the most common chysophytes is the summer blooming Dinobryon balticum and Dinobryon faculiferum (Hällfors, 2004), but we cannot draw any conclusion on which species produced our preserved cysts. Chrysophytes seem to have been favored by the conditions prevailing during Medieval times since the absolute abundance of cysts was very high (Fig. 5) following a similar pattern as diatom absolute abundance. Also, the heterotrophic protist Ebria tri-partita has been used to interpret past changes in the Baltic environment presumably indicating nutrient enrichment and salinity change (Korhola and Grönlund, 1999). Ebria tripartita feeds on spring blooming diatoms, e.g. Skeletonema and Thalassiosira (Hargraves, 2002), which might indicate that a higher abundance could be coupled to food
availability, namely high abundance of diatoms which also is visible in our data (Fig. 5). The high diatom production during the MCA is sup-ported by high biogenic silica content in the eastern Gotland Basin (Kabel et al., 2012) and western Gotland Basin (van Wirdum et al., 2019).
The pelagic to benthic (P/B) ratio is inversely correlated to species richness where we record low richness during the warmer periods with high P/B ratio, i.e. when pelagic production is high it is dominated by few taxa (Fig. 5). The P/B ratio has been used in the coastal zone to reflect changes in Sechii depth and indicate eutrophication of coastal waters (Norbäck Ivarsson et al., 2019). In the open Baltic Proper the coupling is not as straight forward since all benthic taxa are transported from their habitats in the coastal zone. The ratio still seems to be useful though, displaying a higher production of pelagic taxa during the warmer periods (Figs. 5–6).
Diatoms dominating the spring bloom during Medieval times were Chaetoceros spp., Thalassiosira levanderi as well as alternating blooms of Pauliella taeniata and Skeletonema marinoi. In our diatom assemblage record there is also a remarkable high abundance of the autumn blooming Pseudosolenia calcar-avis. This taxon has been reported as an invasive species in the Caspian Sea and resulted in a change in the energy flux from pelagic to benthic after the invasion in the 1930's, due to it is avoided by zooplankton (Karpinsky, 2010). A similar change in Fig. 5. Lithology, total carbon (A) and absolute abundances of siliceous microfossils (B-D), in number per g dry weight (no./gdw) from station MSM62–1-60 western Gotland Basin. B. Diatom data include diatom absolute abundance (excluding Skeletonema marinoi and Chaetoceros spp. vegetative taxa), Chaetoceros spp. resting spores, diatoms pelagics/benthics (P/B) ratio and diatom species richness. C. Crysophyte cysts absolute abundance. D. Ebria tripartita absolute abundance. The diatom assemblage zones (DAZ) 1–4 based on compositional change in relative diatom abundance (seeFigs. 3-4) are displayed to the right. The lithology and total carbon show bottom water conditions but is also affected by input of primary production from the surface waters. The siliceous microfossil data reflects only surface water conditions.
flux of organic matter, but the opposite, is recorded by Klais et al. (2011)in the Baltic Proper today, when dinoflagellates dominate the spring bloom and fuel the pelagic microbial food web and retain pelagic nutrients. When diatoms dominate, fresh organic matter is instead transported to the sediment and increases the consumption of oxygen at the seafloor (Klais et al., 2011). The interaction between spring phy-toplankton communities during the last decades has resulted in a dominance of dinoflagellates in relation to diatoms, which is considered an anomaly compared to patterns seen worldwide (Klais et al., 2011). The competition between phytoplankton communities recorded today can be expected to be a natural phenomenon which has occurred throughout the Baltic Sea history. Since our data only provide the si-liceous microfossil components, we can only hypothesize on how the dynamic between diatoms, dinoflagellates, cyanobacteria and other phytoplankton has varied through time. The remarkably high con-centrations of siliceous microfossils indicate however, that diatoms potentially dominated the spring bloom during the MCA. In the northwestern Baltic Proper, phytoplankton data today shows significant correlation between the spring bloom species composition, the relative abundance of diatoms versus dinoflagellates and the phase of the NAO (Klais et al., 2011). This is recorded as earlier shifts from diatoms to dinoflagellates in years with a positive NAO phase and manifested as an overall dominance of dinoflagellates (Klais et al., 2011). A model si-mulation of the Baltic Sea area shows that a positive NAO phase pre-vailed during the MCA (Schimanke et al., 2012). This means that the
coupling between dinoflagellate dominance during a positive NAO as found today (Klais et al., 2011) was not likely prevailing during the MCA.
We need to scrutinize our paleoecological data in the light of a new perspective on seasonal succession in phytoplankton suggested by Kemp and Villareal (2018). Diatoms are often treated as a functional type that thrive in turbulent, nutrient rich waters and will decrease in abundance with increased stratification (Falkowski and Oliver, 2007). Increased thermal stratification of the oceans occurs in a warmer cli-mate and models predict diatoms to be replaced by other phyto-plankton which would decrease the marine biological carbon pump (Litchman et al., 2006). This view is challenged byKemp and Villareal (2018)who show that diatoms, especially some taxa that have adapted, may thrive and bloom in stratified oligotrophic conditions (Kemp et al., 2000). This so-called shade flora has the ability to subsist stratified conditions by different adaptation strategies such as: symbiosis with nitrogen-fixing cyanobacteria, ability to grow in depth and mine nu-trients at low light conditions and finally they can regulate their buoyancy to move vertically between nutrients and the euphotic zone (Kemp et al., 2000;Kemp and Villareal, 2018). Especially rhizosolenids (including Pseudosolenia) are generally adapted to highly stratified conditions (Margalef, 1967), forming mats of high surface concentra-tions and are the main component of Mediterranean S5 sapropel (Schrader and Matherne, 1981;Kemp et al., 2000). In a sediment-trap from the U.S. west coast massive fluxes of up to 90% Pseudosolenia Fig. 6. Our western Gotland Basin proxy data (A-B), compared to published proxy and climate data (C-D) plotted versus linear age scale. The major climate anomalies are shown as light red (warm) and light blue (cold) shades comprising the warm Medieval Climate Anomaly (MCA) and Modern Warm Period (MoWP), and the cold Little ice Age (LIA) (Mann et al., 2009). A. Total carbon. B. Diatom proxy data: DCA axis 1 sample scores, diatom production (absolute abundance of all counted diatoms including Chaetoceros spp. vegetative taxa and resting spores and S. marinoi. One level was excluded as mass occurrence of S. marinoi made counting impossible), relative abundance of marine diatom taxa in %. C. Pigments zeaxanthin and echinone, biomarkers used as proxy for cyanobacterial abundance (Funkey et al., 2014) in the Fårö Deep (L80, red triangles) and the Northern Gotland Basin (LL19, blue circles), respectively. D. Climate data: combined temperature-precipitation relationship (T/P) from proxy records in Finland (bars) and with LOESS smoothing (span 0.3, line) (Luoto and Nevalainen, 2018). T/P > .5 (red shade) indicate warm and dry climate and reflect positive NAO index, and T/P < .5 (blue shade) indicate cold and wet climate and reflect negative NAO index. Winter temperature data from Stockholm combining instrumental data and historical records (Leijonhufvud et al., 2010) with 5-years running average (red line), and a proxy-based reconstruction (Moberg et al., 2005) of Northern Hemisphere temperature (plotted as anomalies with respect to 1961–1990C.E.) with 10-years running average (red line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
calcar-avis were recorded for several years during late fall and winter and accounted for about half the magnitude of the spring bloom (Sancetta et al., 1991). A microfabric study of laminated sediments deposited in the eastern Gotland Basin during the Littorina Sea identi-fied near monospecific diatom ooze laminae of P. calcar-avis from flux events (Burke et al., 2002). Since P. calcar-avis is the dominant species recorded during Medieval times in our data set we can assume that this is evidence of highly thermally stratified autumn waters making the biological carbon pump of diatoms efficient in fixing carbon in the se-diment (cf.Klais et al., 2011) regardless of oligotrophic conditions. 4.1.3. Transition to the LIA
The transition from laminated sediment (hypoxia) to homogeneous sediments (oxic) occurs subsequently about 1200 CE when production (absolute abundance) of diatoms and other siliceous microfossils de-crease, resulting in a reduction of total carbon in the sediment. This very low organic carbon content and oxic sea bottoms reflect low pri-mary production of diatoms during the Little Ice Age (LIA, defined as 1400–1700C.E. in Mann et al. (2009)) which remained until about 1940C.E. The cause of the shift from massive cyanobacterial production and benthic hypoxia at the end of the MCA to low production and oxic bottoms in LIA has been attributed to temperature (Kabel et al., 2012). It is obvious in our diatom data that salinity and primary production decreased in the Baltic Proper simultaneously as the transition to oxic conditions. This most probably occurred as a result of a shift in climate to colder and wetter (Helama et al., 2009; Luoto and Nevalainen, 2018), which caused decreased salinity, weaker stratification, less carbon flux to the seafloor and a shift in diatom composition (less au-tumn blooming taxa). The shift to oxic conditions in the open Baltic Sea has as well been associated with decreased human influence due to a halving of the population as a result of the plague (Zillén and Conley, 2010; Funkey et al., 2014). However, since there are no signs in our biostratigraphic data of human-induced eutrophication during the MCA, a connection between the oxic conditions in the open Baltic Proper and reduced population density in the drainage basin cannot be found. Further the plague struck Europe in the middle of the fourteenth century, i.e. during a twenty years period 1350–1369C.E. (Myrdal, 2011), which postdates the shift to oxic conditions by at least a century or so.
4.2. Baltic Sea during the last 100 years 4.2.1. Effects of nutrient discharge
Our data displays a significant shift in diatom composition around 1940C.E. with a huge peak in Cyclotella choctawhacheeana simulta-neously with a shift from homogeneous to laminated, hypoxic sedi-ments and a sharp increase in total carbon (Figs. 4–5). C. chocta-whacheeana is a summer blooming taxon which has increased in abundance in the Baltic Sea, both in coastal and open conditions (Andrén et al., 1999;Andrén et al., 2000a; Norbäck Ivarsson et al., 2019; Weckström et al., 2004) and in the Chesapeake Bay (Cooper, 1995) during the last century. This increase has been interpreted to occur in response to anthropogenic perturbation, in particular eu-trophication. When testing the response of C. choctawhacheeana to different environmental parameters in Finnish coastal waters it is ob-vious though that it is not responding to high nutrient conditions but to water depth and temperature (Weckström and Juggins, 2006), thus rather to altered thermal stratification. Another diatom stratigraphical study from the open Baltic Proper (eastern Gotland Basin) shows sig-nificant changes in the species composition around 1950–1960 attrib-uted to eutrophication (Andrén et al., 2000a). This timing fits with increased cyanobacterial abundance and nitrogen fixation correlated with bottom hypoxia in the Baltic Proper dated to 1950C.E. using proxy methods (Funkey et al., 2014), and to 1960C.E. using historical mon-itoring data (Finni et al., 2001). Cyanobacterial blooms are often trig-gered by nutrient enrichment in warm, calm summer months
(Andersson et al., 2017). Our summer blooming Cyclotella could benefit from the surplus nitrogen fixed by the concurrent cyanobacterial blooms.
Our recorded shift dates to 1940C.E., seemingly earlier than other studies in the area. But scrutinizing the increased carbon content and cyanobacterial production inFunkey et al. (2014)(Fig. 6c) it is obvious that the change occurred slightly earlier in the Fårö Deep, in line with our data. This timing is plausible since both nitrate and phosphorous concentrations in the Gotland Basin surface water had already started to increase (Gustafsson et al., 2012), as had the spread of bottom water hypoxia (Carstensen et al., 2014a). When comparing our results with human-induced eutrophication recorded in the Baltic Sea coastal zone it is obvious that changes are site specific but were recorded much earlier, commonly about 1800–1900C.E. (Andrén, 1999;Andrén et al., 1999; Ning et al., 2018;Norbäck Ivarsson et al., 2019), but also as late as mid-1940s to 1990s in less urban areas (Weckström, 2006).
Modeling results show that increased nutrient loads and not in-creased temperatures have caused the eutrophication and hypoxia which accelerated since the 1950's in the Baltic Sea (Meier et al., 2019). The model further shows that nitrogen fixation by cyanobacteria was important to develop hypoxia and without the increased nutrient dis-charge hypoxia would not occur in the Baltic Sea during the twentieth and twenty-first centuries (Meier et al., 2019).
4.2.2. Response to global warming
Signs of climate impact similar to those recorded during the MCA is evident in our diatom data for the last twenty years. The compositional change in our diatom data is followed by a peak in Thalassiosira hy-perborea var. lacunosa and then Actinocyclus octonarius var. crassus. These changes can be interpreted as a succession from taxa indicating enhanced availability of nutrients to taxa indicative of a long-lasting growing season in late summer-autumn during strongly stratified water conditions. The shift starts with Actinocyclus octonarius var. crassus and is followed by a very prominent peak in Coscinodiscus granii and in-crease in A. octonarius var. tenellus about year 2000C.E. (DAZ 4;Fig. 4). This change is most likely attributed to a warmer climate resulting in long-lived diatom shade flora (Kemp et al., 2000) which also would be favored by nitrogen fixed by summer cyanobacterial blooms (Vahtera et al., 2007; Karlson et al., 2015). The increase in cyanobacterial blooms recorded in the last decades, is positively correlated to surface water temperature and the amount of phosphorous and hypoxia in the bottom waters (Kahru et al., 2020). Our data also records a shift in the spring blooming diatom taxa with a decrease in Thalassiosira levanderi, T. hyperborea var. lacunosa and Pauliella taeniata and an increase in Thalassiosira baltica in the uppermost samples. This shift could most likely also be attributed to a warmer climate since T.baltica dominates the spring bloom in years with ice-free conditions together with Ske-letonema marinoi (Thomas et al., 2017). Skeletonema spp. and Chaeto-ceros spp. are normally amongst the dominating taxa in the spring bloom in marine waters worldwide (Kemp et al., 2000) and the lack in our data set is most probably a result of less preservation of their de-licate valves in the sediment, a situation recorded already in the water column during settling (Höglander et al., 2004). Thalassiosira taxa have high settling rates compared to Chaetoceros and are consequently to a higher degree fully preserved as intact cells in the sediments compared to the delicate Chaetoceros which degrade during settling (Höglander et al., 2004). At the monitoring station BY31 Landsort Deep, located in the same basin as our sampling site (Fig. 1), the vertical distribution of the spring bloom 1996 was recorded by using sedimentation traps at various water depths (Höglander et al., 2004). The recorded diatoms (Chaetoceros spp., Pauliella (previously Achnanthes) taeniata, Thalassio-sira levanderi, ThalassioThalassio-sira baltica and Skeletonema marinoi (previously costatum)) in the water column are similar to our diatom assemblage from this time, except that we also recorded Melosira arctica in the se-diment (Fig. 4). Melosira arctica is a sea ice-associated taxon, a glacial relict, which today occurs in the Baltic Sea and Arctic Ocean and as
most of sympagic diatoms it sinks rapidly after ice melt (Thomas et al., 2017). The results of our study are further confirmed by analyses of the long-term trend in phytoplankton composition in monitoring data from the Baltic Proper with a decrease in spring blooming diatoms e.g. Pauliella taeniata and an increase in autumn blooming taxa, especially Coscinodiscus granii (Wasmund and Uhlig, 2003;Wasmund et al., 2011; Andersson et al., 2017). In the southwestern Baltic Sea, the earlier phytoplankton growing season has been correlated to more sunshine in spring and the longer autumn blooms were correlated to increased surface water temperatures and stratification (Wasmund et al., 2019). In the offshore monitoring station, BY31 Landsort Deep, there is a correlation between less windy conditions, increasing light intensity and reduced mixed layer depth and the earlier total phytoplankton bloom timing (Hjerne et al., 2019).
4.3. Comparison between medieval and recent environmental conditions Our data shows that the environmental conditions in the western Gotland Basin during Medieval times were more marine, at least 8 psu in the surface water compared to the present 6.5 psu, an environment which enabled the marine diatom Pseudosolenia calcar-avis to thrive. The higher salinity could possibly be linked to a warm and dry climate which caused severe precipitation deficit in northern Europe during the MCA (Helama et al., 2009;Luoto and Nevalainen, 2018). The warmer climate induced a strong summer-autumn thermal stratification re-sulting in extensive long-lasting diatom blooms of so-called shade flora (P. calcar-avis), enhancing the vertical export of organic carbon to the sediment and contributing to the development of benthic hypoxia. A long blooming period during summer-autumn may last for months until winter mixing and breakdown of thermal stratification, which result in carbon flux comparable to or higher than the spring bloom (Kemp et al., 2000). The situation with massive P. calcar-avis flux have previously been recorded in the Baltic Sea during the Holocene Climate Maximum (Burke et al., 2002), but in particular 2000–1000 years ago in both the Bornholm Basin and the eastern Gotland Basin (Andrén et al., 2000a, 2000b). Since it is an autumn blooming species and generally prefers warm regions the re-appearance in the southern Baltic Proper recently (Kaiser et al., 2016;Wasmund et al., 2019) might be linked to warming and accompanied longer autumn blooms. Thus, both higher salinity and higher temperatures (stronger thermal stratification) during the MCA were probably favorable to the vast production of P. calcar-avis in the Baltic Proper. The less marine conditions in the area during present-day most probably prevent re-appearance and an expansion of P. calcar-avis from the southern Baltic Sea. However, other large diatom taxa that thrive in stratified waters during autumn blooms (Actinocyclus octo-narius including varieties and Coscinodiscus granii) have been increasing the last two decades, likely inferred from global warming.
A contradiction between our paleoenvironmental results and mod-eling data is found in Schimanke et al. (2012) which show fresher conditions in the Baltic Basin during the MCA and too high oxygen levels in the bottom waters to develop the extended hypoxia recorded in deep bottoms during Medieval times (Zillén et al., 2008). Large scale land-use changes in the watershed have been suggested to be the source of excess nutrient input needed to get deep water hypoxia during MCA (Zillén and Conley, 2010;Åkesson et al., 2015). There are however no evidences for increased nutrient discharge during Medieval times re-corded in the coastal zones (Jokinen et al., 2018;Ning et al., 2018; Norbäck Ivarsson et al., 2019), where one would expect to first record the effects of anthropogenic forcing. The nutrient removal in the Baltic Sea today is most efficient close to land in the coastal zone which act as an efficient nutrient filter (Edman et al., 2018;Carstensen et al., 2020). About 2500 Swedish lakes and additionally also wetlands were drained during the second half of 19th century and first half of 20th century to gain arable land (Hoffmann et al., 2000). These lakes and wetlands had a significant role as nutrient sinks and the drainage resulted in a great loss of retention capacity which might have increased the net load to
the sea more than the post-war increase in fertilizer (Hoffmann et al., 2000). A more efficient coastal filter and higher retention capacity in coastal wetlands could explain why intensified land-use during Med-ieval times did not result in higher nutrient discharge in the open Baltic Sea. Instead we suggest that massive flux of diatom shade flora pro-moted by warm and dry climate (higher salinity and thermal stratifi-cation) reinforced by cyanobacterial blooms and nitrogen fixation (Funkey et al., 2014), increased carbon sequestration to the sediments by diatoms due to extended length of the growing season, which made it possible to get high primary productivity and hypoxia during MCA despite oligotrophic conditions. This can be compared with the situa-tion during the twentieth and twenty-first centuries when hypoxia and man-made eutrophication of the Baltic Sea is attributed to increased nutrient input from land and exacerbated by the ongoing climate warming and nitrogen-fixing cyanobacteria (Meier et al., 2019; Carstensen and Conley, 2019). The modern coastal eutrophication re-corded in Swedish and Finnish waters are without parallel the last thousand years (Jokinen et al., 2018;Norbäck Ivarsson et al., 2019).
5. Conclusions
•
More marine conditions than today prevailed during Medieval times in the western Gotland Basin, reaching at least 8 psu in the surface waters compared to present-day about 6.5 psu. The higher salinity made it possible for Pseudosolenia calcar-avis, an autumn blooming warm water diatom which forms massive algal mats in stratified waters, to thrive and provide huge flux of organic matter to the seafloor contributing to form extensive areas of benthic hypoxia.•
The diatom data indicates decreased salinity, weaker stratification and low primary production during the LIA which resulted in a low organic content, homogeneous sediments and well oxygenated bottoms.•
The first signs of human-induced eutrophication are recorded about 1940 visible as a shift in diatom composition, increased primary production as well as carbon content, and benthic hypoxia.•
The last two decades a species shift indicative of taxa thriving in thermal stratified waters during autumn blooms is recorded, inter-preted as being the result of the documented global warming. Less marine conditions evident in the western Gotland Basin today probably prevent P. calcar-avis to be reintroduced from the southern Baltic Sea.Data availability
The siliceous microfossil data related to this article can be found at doi:10.17043/andren-2020, an open-source online data repository hosted by the Bolin Centre Database (https://bolin.su.se/data/).
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ-ence the work reported in this paper.
Acknowledgements
The authors wish to thank the Editor-in-Chief Professor Thomas Algeo and two anonymous reviewers for their constructive comments that contributed to improve this paper. E.A and T.A. acknowledges funding from the Foundation of Baltic and East European studies (grant 1562/3.1.1/2013 to E.A. and grant 2207/3.1.1/2014 to T.A.). The crew on RV Maria S. Merian helped with coring. We acknowledge the bathymetric data used inFig. 1from EMODnet Bathymetry Consortium (2018): EMODnet Digital Bathymetry (DTM) doi: 10.12770/18ff0d48-b203-4a65-94a9-5fd8b0ec35f6.
Appendix A. Supplementary data
Supplementary data to this article can be found online athttps:// doi.org/10.1016/j.palaeo.2020.109878.
References
Åkesson, C., Nielsen, A.B., Broström, A., Persson, T., Gaillard, M.-J., Berglund, B.E., 2015. From landscape description to quantification: a new generation of reconstructions provides new perspectives on Holocene regional landscapes of SE Sweden. The Holocene 25, 178–193.https://doi.org/10.1177/0959683614556552.
Andersson, A., Meier, H.E.M., Ripszam, M., Rowe, O., Wikner, J., Haglund, P., Eilola, K., Legrand, C., Figueroa, D., Paczkowska, J., Lindehoff, E., Tysklind, M., Elmgren, R., 2015. Projected future climate change and Baltic Sea ecosystem management. AMBIO 44, S345–S356.https://doi.org/10.1007/s13280-015-0654-8.Suppl. 3.
Andersson, A., Tamminen, T., Lehtinen, S., Jürgens, K., Labrenz, M., Viitasalo, M., 2017. The pelagic food web. In: Snoeijs-Leijonmalm, P., Schubert, H., Radziejewska, T. (Eds.), Biological Oceanography of the Baltic Sea. Springer, Dordrecht.https://doi. org/10.1007/978-94-007-0668-2_8.
Andrén, E., 1999. Changes in the Composition of the Diatom Flora during the last Century Indicate increased Eutrophication of the Oder Estuary, South-western Baltic Sea. Estuar. Coast. Shelf Sci. 48, 665–676.https://doi.org/10.1006/ecss.1999.0480. Andrén, E., Shimmield, G., Brand, T., 1999. Changes in the environment during the last
centuries on the basis of siliceous microfossil records from the southwestern Baltic Sea. The Holocene 9, 25–38.https://doi.org/10.1191/095968399676523977. 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.https://doi.org/10.1191/09596830094944. 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.https://doi.org/10.1111/j.1502-3885.2000.tb00981.x.
Andrén, T., Björck, S., Andrén, E., Conley, D., Zillén, L., Anjar, J., 2011. The Development of the Baltic Sea Basin during the last 130 ka. The Baltic Sea Basin In: Harff, J., Björck, S., Hoth, P. (Eds.), Central and Eastern European Development Studies (CEEDES). Springer-Verlag, Berlin Heidelberg, pp. 75–97.
Battarbee, R.W., 1986. Diatom analysis. In: Berglund, B.E. (Ed.), Handbook of Holocene Palaeoecology and Palaeohydrology. Wiley, Chichester, pp. 527–570.
Battarbee, R.W., Kneen, M.J., 1982. The use of electronically counted microspheres in absolute diatom analysis. Limnol. Oceanogr. 27, 184–188.https://doi.org/10.4319/ lo.1982.27.1.0184.
Binczewska, A., Moros, M., Polovodova Asteman, I., Sławińska, J., Bąk, M., 2018. Changes in the inflow of saline water into the Bornholm Basin (SW Baltic Sea) during the past 7100 years – evidence from benthic foraminifera record. Boreas 47, 297–310.https://doi.org/10.1111/bor.12267.
Birks, H.J.B., Line, J.M., 1992. The use of Rarefaction Analysis for estimating Palynological Richness from Quaternary Pollen-Analytical Data. The Holocene 2, 1–10.https://doi.org/10.1177/095968369200200101.
Blaauw, M., 2010. Methods and code for 'classical' age-modelling of radiocarbon se-quences. Quat. Geochronol. 5, 512–518.
Blomqvist, S., Larsson, U., 1994. Detrital bedrock elements as tracers of settling re-suspended particulate matter in a coastal area of the Baltic Sea. Limnol. Oceanogr. 39 (4), 880–896.https://doi.org/10.4319/lo.1994.39.4.0880.
Börgel, F., Frauen, C., Neumann, T., Schimanke, S., Meier, H.E.M., 2018. Impact of the Atlantic Multidecadal Oscillation on Baltic Sea variability. Geophys. Res. Lett. 45, 9880–9888.https://doi.org/10.1029/2018GL078943.
Breitburg, D., Levin, L.A., Oschlies, A., Grégoire, M., Chavez, F.P., Conley, D.J., Garçon, V., Gilbert, D., Gutiérrez, D., Isensee, K., Jacinto, G.S., Limburg, K.E., Montes, I., Naqvi, S.W.A., Pitcher, G.C., Rabalais, N.N., Roman, M.R., Rose, K.A., Seibel, B.A., Telszewski, M., Yasuhara, M., Zhang, J., 2018. Declining oxygen in the global ocean and coastal waters. Science 359, eaam7240.https://doi.org/10.1126/science. aam7240.
Burke, I.T., Grigorov, I., Kemp, A.E.S., 2002. Microfabric study of diatomaceous and li-thogenic deposition in laminated sediments from the Gotland deep, Baltic Sea. Mar. Geol. 183, 89–105.https://doi.org/10.1016/S0025-3227(01)00250-X.
Carstensen, J., Conley, D.J., 2019. Baltic Sea hypoxia takes many shapes and sizes. Limnol. Ocean. Bull. 28 (4), 125–129.https://doi.org/10.1002/lob.10350. Carstensen, J., Andersen, J.H., Gustafsson, B.G., Conley, D.J., 2014a. Deoxygenation of
the Baltic Sea during the last century. Proc. Natl. Acad. Sci. U. S. A. 111, 5628–5633.
https://doi.org/10.1073/pnas.1323156111.
Carstensen, J., Conley, D.J., Bonsdorff, E., Gustafsson, B.G., Hietanen, S., Janas, U., Jilbert, T., Maximov, A., Norkko, A., Norkko, J., Reed, D.C., Slomp, C.P., Timmermann, K., Voss, M., 2014b. Hypoxia in the Baltic Sea: Biogeochemical Cycles, Benthic Fauna, and Management. AMBIO 43, 26–36.https://doi.org/10.1007/ s13280-013-0474-7.
Carstensen, J., Klais, R., Cloern, J.E., 2015. Phytoplankton blooms in estuarine and coastal waters: Seasonal patterns and key species. Estuar. Coast. Shelf Sci. 162, 98–109.https://doi.org/10.1016/j.ecss.2015.05.005.
Carstensen, J., Conley, D.J., Almroth-Rosell, E., Asmala, E., Bonsdorff, E., Fleming-Lehtinen, V., Gustafsson, B.G., Gustafsson, C., Heiskanen, A.-S., Janas, U., Norkko, A., Slomp, C., Villnäs, A., Voss, M., Zilius, M., 2020. Factors regulating the coastal nu-trient filter in the Baltic Sea. Ambio 49, 1194–1210.https://doi.org/10.1007/ s13280-019-01282-y.
Cleve-Euler, A., 1951–1955. Die Diatomeen von Schweden und Finnland. Kungliga Vetenskapsakademiens Handlingar I-V. Stockholm: Almqvist & Wiksell. 4:e serien 2:1
(1951), 3:3 (1952), 4:1 (1953), 4:5 (1953), 5:4 (1955).
Cooper, S.R., 1995. Diatoms in Sediment Cores from the Mesohaline Chesapeake Bay, U.S.A. Diatom Res. 10, 39–89.https://doi.org/10.1080/0269249X.1995.9705329.
Davies, A.L., Bunting, M.J., 2010. Applications of Palaeoecology in Conservation. Open Ecol. J. 3, 54–67.
Diaz, R.J., Rosenberg, R., 2008. Spreading Dead zones and Consequences for Marine Ecosystems. Science 321 (5891), 926–929.https://doi.org/10.1126/science. 1156401.
Duff, K.E., Zeeb, B.A., Smol, J.P., 1995. Atlas of Chrysophycean Cysts. In: Developments in Hydrobiology 99. Springer, Netherlands. https://doi.org/10.1007/978-94-017-0809-8.189 pp.
Edman, M., Eilola, K., Almroth-Rosell, E., Meier, H.E.M., Wåhlström, I., Arneborg, L., 2018. Nutrient retention in the Swedish Coastal Zone. Front. Mar. Sci. 5.https://doi. org/10.3389/fmars.2018.00415.
Falkowski, P.G., Oliver, M.J., 2007. Mix and match: how climate selects phytoplankton. Nat. Rev. Microbiol. 5, 813–819.https://doi.org/10.1038/nrmicro1751. Finni, T., Kononen, K., Olsonen, R., Wallström, K., 2001. The history of Cyanobacterial
Blooms in the Baltic Sea. AMBIO 30 (4), 172–178. https://doi.org/10.1579/0044-7447-30.4.172.
Funkey, C.P., Conley, D.J., Reuss, N.S., Humborg, C., Jilbert, T., Slomp, C.P., 2014. Hypoxia Sustains Cyanobacteria Blooms in the Baltic Sea. Environ. Sci. Technol. 48 (5), 2598–2602.https://doi.org/10.1021/es404395a.
Granéli, E., Wallström, K., Larsson, U., Granéli, W., Elmgren, R., 1990. Nutrient Limitation of primary production in the Baltic Sea Area. AMBIO 19, 142–151. Grimm, E.C., 1987. CONISS: a FORTRAN 77 program for stratigraphically constrained
cluster analysis by the method of incremental sum of squares. Comput. Geosci. 13, 13–35.https://doi.org/10.1016/0098-3004(87)90022-7.
Gustafsson, B.G., Westman, P., 2002. On the causes for salinity variations in the Baltic Sea for the last 8500 years. Paleoceanography 17, 1–14.https://doi.org/10.1029/ 2000PA000572.
Gustafsson, B.G., Schenk, F., Blenckner, T., Eilola, K., Meier, H.E.M., Müller-Karulis, B., Neumann, T., Ruoho-Airola, T., Savchuk, O.P., Zorita, E., 2012. Reconstructing the Development of Baltic Sea Eutrophication 1850–2006. AMBIO 41, 534–548.https:// doi.org/10.1007/s13280-012-0318-x.
Hällfors, G., 2004. Checklist of Baltic Sea Phytoplankton Species (including some het-erotrophic protistan groups). Baltic Sea Environ. Proceed. 95, 208.
Hargraves, P.E., 2002. The ebridian flagellates Ebria and Hermesinum. Plankton Biol. Ecol. 49, 9–16.
Hasle, G.R., Syvertsen, E.E., 1997. Chapter 2 - Marine Diatoms. In: Tomas, C.R. (Ed.), Identifying Marine Phytoplankton. Academic Press, San Diego, pp. 5–385.https:// doi.org/10.1016/B978-012693018-4/50004-5.
Helama, S., Meriläinen, J., Tuomenvirta, H., 2009. Multicentennial megadrought in northern Europe coincided with a global El Niño–Southern Oscillation drought pat-tern during the medieval climate Anomaly. Geology 37 (2), 175–178.https://doi. org/10.1130/G25329A.1.
HELCOM, 2007. Climate change in the Baltic Sea Area – HELCOM thematic assessment in 2007. In: Baltic Sea Environmental Proceedings. 111. pp. 49.
Hjerne, O., Hajdu, S., Larsson, U., Downing, A.S., Winder, M., 2019. Climate driven changes in timing, composition and magnitude of the Baltic Sea phytoplankton spring bloom. Front. Mar. Sci. 6, 482.https://doi.org/10.3389/fmars.2019.00482. Hoffmann, M., Johnsson, H., Gustafson, A., Grimvall, A., 2000. Leaching of nitrogen in
Swedish agriculture — a historical perspective. Agric. Ecosyst. Environ. 80, 277–290.
https://doi.org/10.1016/S0167-8809(00)00154-7.
Höglander, H., Larsson, U., Hajdu, S., 2004. Vertical distribution and settling of spring phytoplankton in the offshore NW Baltic Sea proper. Mar. Ecol. Prog. Ser. 283, 15–27.
https://doi.org/10.3354/meps283015.
Hurrell, J.W., 1995. Decadal Trends in the North Atlantic Oscillation: Regional Temperatures and Precipitation. Science 269, 676–679.
Jokinen, S.A., Virtasalo, J.J., Jilbert, T., Kaiser, J., Dellwig, O., Arz, H.W., Hänninen, J., Arppe, L., Collander, M., Saarinen, T., 2018. A 1500-year multiproxy record of coastal hypoxia from the northern Baltic Sea indicates unprecedented deoxygenation over the 20th century. Biogeosciences 15, 3975–4001. https://doi.org/10.5194/bg-15-3975-2018.
Juggins, S., 2015. rioja: Analysis of Quaternary Science Data. Available at: https://cran.r-project.org/web/packages/rioja/index.html.
Kabel, K., Moros, M., Porsche, C., Neumann, T., Adolphi, F., Andersen, T.J., Siegel, H., Gerth, M., Leipe, T., Jansen, E., Damsté, J.S.S., 2012. Impact of climate change on the Baltic Sea ecosystem over the past 1,000 years. Nat. Clim. Chang. 2, 871–874.
https://doi.org/10.1038/nclimate1595.
Kahru, M., Elmgren, R., 2014. Multidecadal time series of satellite-detected accumula-tions of cyanobacteria in the Baltic Sea. Biogeosciences 11, 3619–3633.https://doi. org/10.5194/bg-11-3619-2014.
Kahru, M., Elmgren, R., Savchuk, O.P., 2016. Changing seasonality of the Baltic Sea. Biogeosciences 13 (4), 1009–1018.https://doi.org/10.5194/bg-13-1009-2016. Kahru, M., Elmgren, R., Kaiser, J., Wasmund, N., Savchuk, O., 2020. Cyanobacterial
blooms in the Baltic Sea: Correlations with environmental factors. Harmful Algae 92.
https://doi.org/10.1016/j.hal.2019.101739.
Kaiser, J., Belt, S.T., Tomczak, M., Brown, T.A., Wasmund, N., Arz, H.W., 2016. C25 highly branched isoprenoid alkenes in the Baltic Sea produced by the marine planktonic diatom Pseudosolenia calcar-avis. Org. Geochem. 93, 51–58.https://doi. org/10.1016/j.orggeochem.2016.01.002.
Karlson, A.M.L., Duberg, J., Motwani, N.H., Hogfors, H., Klawonn, I., Ploug, H., Barthel Svedén, J., Garbaras, A., Sundelin, B., Hajdu, S., Larsson, U., Elmgren, R., Gorokhova, E., 2015. Nitrogen fixation by cyanobacteria stimulates production in Baltic food webs. AMBIO 44, 413–426.https://doi.org/10.1007/s13280-015-0660-x. Karpinsky, M.G., 2010. Pseudosolenia calcar-avis (Bacillariophyta, Centrophyceae) in the
