Umeå University's logo

Organic carbon (OC) burial rates in northern lakes are estimated to have increased by 2–3 fold over the past 150 years. However, assessing OC burial efficiency is challenging because (a) long-term (decadal) process are difficult to study in situ, and (b) sediment organic matter (OM) consists of thousands of different compounds from both terrestrial and aquatic sources, which are subject to different degrees of degradation, transformation, or preservation. Here, we used pyrolysis–gas chromatography/mass spectrometry to track changes in the organic molecular composition of individual varve years in a series of sediment freeze cores collected during 1979–2010, allowing us to assess diagenetic changes over ≤31 years (or 12.5 cm depth). As predicted from previous work, the greatest losses over time/depth (18–19 years; 8.5 cm) are for compounds indicative of fresh OM, both terrestrial (e.g., levosugars with 58%–77% lost) and particularly aquatic origin (e.g., phytadiene and phytene amongst chlorophylls with 40%–82% lost). This high variability in degradation of specific compounds has implications for interpreting past changes in C and N. Although OM composition changes only slightly beyond 20 years (8.5 cm), the chlorophyll:lignin ratio (fresh vs. degraded compounds) continues to decline to 31 years (12.5 cm) and is predicted to continue up to 100 years (37 cm depth). In most northern lakes, indications of OM degradation to these depths correspond to sediment ages of 50 to >150 years, suggesting that much of the recent increase in OC burial in northern lakes does not represent permanent sequestration of C.

Accelerations of ecosystem transformation raise concerns, to the extent that high rates of ecological change may be regarded amongst the most important ongoing imbalances in the Earth system. Here, we used high-resolution pollen and diatom assemblages and associated ecological indicators (the sum of tree and shrub pollen and diatom-inferred total phosphorus concentrations as proxies for tree cover and lake-water eutrophication, respectively) spanning the past 150 years to emphasize that rate-of-change records based on compositional data may document transformations having substantially different causes and outcomes. To characterize rates of change also in terms of other key ecosystem features, we quantified for both ecological indicators: (i) the percentage of change per unit time, (ii) the percentage of change relative to a reference level, and (iii) the rate of percentage change per unit time relative to a reference period, taking into account the irregular spacing of palaeoecological data. These measures document how quickly specific facets of nature changed, their trajectory, as well as their status in terms of palaeoecological indicators. Ultimately, some past accelerations of community transformation may document the potential of ecosystems to rapidly recover important ecological attributes and functions. In this context, insights from palaeoecological records may be useful to accelerate ecosystem restoration.

The release of toxic metals from local mining activities often represents a severe environmental hazard for nearby lake ecosystems. Previous studies on the impact of mining have primarily focused on single lakes, with less emphasis on spatial and temporal recovery patterns of multiple lakes within the same catchment, but with different hydrological connection and distance to the pollutant source. This knowledge gap prevents us from assessing the real environmental risk of abandoned mines and understanding ecosystem recovery. This study explores the intensity and spatial patterns of sediment contamination and the potential for ecosystem recovery in three lakes in close vicinity of a lead (Pb) and zinc (Zn) mine in Sweden that has been inoperative for >20 years. Dated (210Pb and 137Cs) sediment cores from each lake were used to reconstruct temporal patterns in trace element deposition and relate those with past mining activities. Results show that all lakes were affected by mining, indicated by increasing Pb and Zn concentrations and decreasing organic matter content, at the onset of mining. However, the extent and timing of mining impact differed between lakes, which was partly ascribed to differences in the historical use of tailings and settling ponds. Assessment of toxicity levels in sediments, based on normalized Probable Effect Concentration Quotient (PEC-Q) to organic matter content, provided more consistent results with the historical mining than conventional methods, showing a decreasing impact in lakes once the operations ceased. Still, sediment Pb concentrations were > 10 times higher than pre-mining values, evidencing the urgent need for remediation actions in the study lakes. This study highlights the importance of considering spatial heterogeneity in metal deposition, sediment organic matter content, and hydrological connectivity with tailings when risk assessments are performed in mining-impacted lakes. The use of normalized PEC-Q in toxic assessments is also recommended.

The use of lake sedimentary DNA to track the long-term changes in both terrestrial and aquatic biota is a rapidly advancing field in paleoecological research. Although largely applied nowadays, knowledge gaps remain in this field and there is therefore still research to be conducted to ensure the reliability of the sedimentary DNA signal. Building on the most recent literature and seven original case studies, we synthesize the state-of-the-art analytical procedures for effective sampling, extraction, amplification, quantification and/or generation of DNA inventories from sedimentary ancient DNA (sedaDNA) via high-throughput sequencing technologies. We provide recommendations based on current knowledge and best practises.

On the annual and interannual scales, lake microbial communities are known to be heavily influenced by environmental conditions both in the lake and in its terrestrial surroundings. How-ever, the influence of landscape setting and environmental change on shaping these communities over a longer (millennial) timescale is rarely studied. Here, we applied an 18S metabarcoding approach to DNA preserved in Holocene sediment records from two pairs of co‐located Swedish mountain lakes. Our data revealed that the microbial eukaryotic communities were strongly influenced by catchment characteristics rather than location. More precisely, the microbial communities from the two bedrock lakes were largely dominated by unclassified Alveolata, while the peatland lakes showed a more diverse microbial community, with Ciliophora, Chlorophyta and Chytrids among the more predominant groups. Furthermore, for the two bedrock‐dominated lakes—where the oldest DNA samples are dated to only a few hundred years after the lake formation—certain Alveolata, Chlorophytes, Stramenopiles and Rhizaria taxa were found prevalent throughout all the sediment profiles. Our work highlights the importance of species sorting due to landscape setting and the persistence of microbial eukaryotic diversity over millennial timescales in shaping modern lake microbial communities.

The history of mining and smelting and the associated pollution have been documented using lake sediments for decades, but the broader ecological implications are not well studied. We analyzed sediment profiles covering the past similar to 10,000 years from three lakes associated with an iron blast furnace in central Sweden, as an example of the many small-scale furnaces with historical roots in the medieval period. With a focus on long-term lake-water quality, we analyzed multiple proxies including geochemistry, pollen and charcoal, diatom composition and inferred pH, biogenic silica (bSi), visible near-infrared spectroscopy (VNIRS)-inferred lake-water total organic carbon (LW-TOC), and VNIRS-inferred sediment chlorophyll (sed-Chl). All three lakes had stable conditions during the middle Holocene (similar to 5000 BCE to 1110 CE) typical of oligo-dystrophic lakes: pH 5.4-5.6, LW-TOC 15-18 mg L-1. The most important diatom taxa include, for example, Aulacoseira scalaris, Brachysira neoexilis, and Frustulia saxonica. From similar to 1150 CE, decreases in LW-TOC, bSi, and sed-Chl in all three lakes coincide with a suite of proxies indicating disturbance associated with local, small-scale agriculture, and the more widespread use of the landscape in the past (e.g. forest grazing, charcoal production). Most important was a decline in LW-TOC by 30-50% in the three lakes prior to the 20th century. In addition, the one lake (Fickeln) downstream of the smelter and main areas of cultivation experienced a shift in diatom composition (mainly increasing Asterionella formosa) and a 0.6 pH increase coinciding with increasing cereal pollen and signs of blast furnace activity. The pH did not change in the other two lakes in response to disturbance; however, these lakes show a slight increase (0.3-0.5 pH units) because of modern liming. LW-TOC has returned to background levels in the downstream lake and remains lower in the other two.

To assess the sensitivity of lakes to anthropogenically-driven environmental changes (e.g., nutrient supply, climate change), it is necessary to first isolate the effects of between-year variability in weather conditions. This variability can strongly impact a lake's biological community especially in boreal and arctic areas where snow phenology play an important role in controlling the input of terrestrial matter to the lake. Identifying the importance of this inherent variability is difficult without time series that span at least several decades. Here, we applied a molecular approach (metabarcoding on eukaryotic 18S rRNA genes and qPCR on cyanobacterial 16S rRNA genes) to sedimentary DNA (sed-DNA) to unravel the annual variability of microbial community in 40 years' sediment record from the boreal lake Nylandssjon which preserve annually-laminated sediments. Our comparison between seasonal meteorological data, sediment inorganic geochemistry (X-ray fluorescence analyses) and organic biomarkers (pyrolysis-gas chromatography/mass spectrometry analyses), demonstrated that inter-annual variability strongly influence the sediment composition in Nylandssjon. Spring temperature, snow and ice phenology (e.g., the percentage of snow loss in spring, the timing of lake ice-off) were identified as important drivers for the inputs of terrestrial material to the lake, and were therefore also important for shaping the aquatic biological community. Main changes were detected in the late-80s/mid-90s and mid-2000s associated with increases in algal productivity, in total richness of the protistan community and in relative abundances of Chlorophyta, Dinophyceae as well as Cyanobacteria abundance. These changes could be linked to a decline in terrestrial inputs to the lake during the snow melt and run-off period, which in turn was driven by warmer winter temperatures. Even if our data shows that meteorological factors do affect the sediment composition and microbial communities, they only explain part of the variability. This is most likely a consequence of the high inter-annual variability in abiotic and biotic parameters highlighting the difficulty to draw firm conclusions concerning drivers of biological changes at an annual or sub-annual resolution even with the 40-year varved sediment record from Nylandssjon. Hence, it is necessary to have an even longer time perspective in order to reveal the full implications of climate change.

Disentangling the process information contained in a diatom sediment signature is crucial for reliable future predictions based on paleolimnological records. In this study, we combine limnological and paleolimnological monitoring to address the fundamental question: Which environmental information is contained in a diatom sediment signal? We compared annual diatom sequential sediment trap records with the diatom record of the annually varved lake sediment of Nylandssjon (northern Sweden) from three meteorologically different years (2012-2014). The seasonal patterns in diatom sedimentation were strikingly different in varve years 2012 and 2014 compared to varve year 2013. In 2012 and 2014, up to 70% of the annual flux occurred in a single spring month and was dominated by Cyclotella glomerata. In contrast, in 2013, peak fluxes were much lower and more annually integrated. Next, we compared the full-year diatom trap results with year round in-lake physical, chemical, and biological monitored parameters, as well as meteorological variables. Annual averages of environmental conditions did not explain the interannual variability in diatom sedimentation. Instead, the seasonality of diatom sedimentation was determined by the timing of the spring diatom bloom relative to lake over-turn in winters with warm vs. cold air temperature. With our combined limnological and paleolimnological monitoring approach, we find that an annual diatom signal can either contain primarily seasonal climate information from a short time period or be annually integrated. We synthesize our results in a novel conceptual model, which describes the response of sediment diatom signals to two distinct sequences of late-winter conditions.

Unglaciated forelands and summits protruding from ice-sheets are commonly portrayed as areas where plants first establish at the end of glacial cycles. But is this prevailing view of ice-free refugia too simplistic? Here, we present findings suggesting that surface debris supported plant communities far beyond the rim of the Late Weichselian Ice-sheet during Greenland interstadial 1 (GI-1 or Bolling-Altered interstadial). We base our interpretations upon findings from terrigenous sediments largely resembling 'plant-trash' deposits in North America (known to form as vegetation established on stagnant ice became buried along with glacial debris during the deglaciation). In our studied deposit, we found macrofossils (N = 10) overlapping with the deglaciation period of the area (9.5-10 cal kyr BP) as well as samples (N = 2) with ages ranging between 12.9 and 13.3 cal kyr BP. The latter ages indicate growth of at least graminoids during the GI-1 interstadial when the site was near the geographic center of the degrading ice-sheet. We suggest that exposure of englacial material during GI-1 created patches of supraglacial debris capable of supporting vascular plants three millennia before deglaciation. The composition and resilience of this early plant community remain uncertain. Yet, the younger group of macrofossils, in combination with pollen and ancient DNA analyses of inclusions, imply that shrubs (Salix sp., Betula sp. and Ericaceae sp) and even tree species (Larix) were present in the debris during the final deglaciation stage. 

Aim: We provide the first European-scale geospatial training set relating the charcoal signal in surface lake sediments to fire parameters (number, intensity and area) recorded by satellite moderate resolution imaging spectroradiometer (MODIS) sensors. Our calibration is intended for quantitative reconstructions of key fire-regime parameters by using sediment sequences of microscopic (MIC from pollen slides, particles 10-500 mu m) and macroscopic charcoal (MAC from sieves, particles > 100 mu m). Location: North-south and east-west transects across Europe, covering the mediterranean, temperate, alpine, boreal and steppe biomes. Time period: Lake sediments and MODIS active fire and burned area products were collected for the years 2012-2015. Methods: Cylinder sediment traps were installed in lakes to annually collect charcoal particles in sediments. We quantitatively assessed the relationships between MIC and MAC influx (particles/cm(2)/year) and the MODIS-derived products to identify source areas of charcoal and the extent to which lake-sediment charcoal is linked to fire parameters across the continent. Results: Source area of sedimentary charcoal was estimated to a 40-km radius around sites for both MIC and MAC particles. Fires occurred in grasslands and in forests, with grass morphotypes of MAC accurately reflecting the burned fuel-type. Despite the lack of local fires around the sites, MAC influx levels reached those reported for local fires. Both MIC and MAC showed strong and highly significant relationships with the MODIS-derived fire parameters, as well as with climatic variation along a latitudinal temperature gradient. Main conclusions: MIC and MAC are suited to quantitatively reconstructing fire number and fire intensity on a regional scale. However, burned area may only be estimated using MAC. Local fires may be identified by using several lines of evidence, e.g. analysis of large particles (> 600 mu m), magnetic susceptibility and sedimentological data. Our results offer new insights and applications to quantitatively reconstruct fires and to interpret available sedimentary charcoal records.