The varved succession of Crawford Lake, Milton, Ontario, Canada as a candidate Global boundary Stratotype Section and Point for the Anthropocene series

Abstract

An annually laminated succession in Crawford Lake, Ontario, Canada is proposed for the Global boundary Stratotype Section and Point (GSSP) to define the Anthropocene as a series/epoch with a base dated at 1950 CE. Varve couplets of organic matter capped by calcite precipitated each summer in alkaline surface waters reflect environmental change at global to local scales. Spheroidal carbonaceous particles and nitrogen isotopes record an increase in fossil fuel combustion in the early 1950s, coinciding with early fallout from nuclear and thermonuclear testing - ^239+240Pu and ¹⁴C:¹²C, the latter more than compensating for the effects of old carbon in this dolomitic basin. Rapid industrial expansion in the North American Great Lakes region led to enhanced leaching of terrigenous elements by acid precipitation during the Great Acceleration, and calcite precipitation was reduced, producing thin calcite laminae around the GSSP that is marked by a sharp decline in elm pollen (Dutch Elm disease). The lack of bioturbation in well-oxygenated bottom waters, supported by the absence of fossil pigments from obligately anaerobic purple sulfur bacteria, is attributed to elevated salinities and high alkalinity below the chemocline. This aerobic depositional environment, highly unusual in a meromictic lake, inhibits the mobilization of Pu, the proposed primary stratigraphic guide for the Anthropocene.

Keywords: Dutch Elm disease; Great Acceleration; SCPs; acid precipitation; combustion; fossil plankton; radionuclides; stable isotopes; varves; µXRF.

Figure 1.

Diatoms and the rotifers that consume them were abundant in Crawford Lake during intervals of human impact (shading), recorded by the pollen of cultigens (e.g., Zea – maize and Helianthus – sunflower) and spores of their pathogens (e.g., Ustilago maydis – corn smut), and nonarboreal pollen (e.g., ragweed – Ambrosia). ×10 = ten times exaggeration of rare but significant elements in palynological preparations. Palynomorph data from Turton and McAndrews (2006), diatom data modified from Ekdahl et al. (2004) with updated taxonomic nomenclature – Lindavia intermedia = Cyclotella bodanica lemanica and Lindavia michiganiana = Cyclotella michiganiana. Diatoms identified as Synedra nana by Ekdahl et al. (2004) are likely to have been various large Fragilariaspp., including the common Fragilaria andreseniana (compare Figure 16). Reproduced in color in online version.

Figure 2.

(a) Crawford Lake is in the Lower Great Lakes region of North America. (b) Location of the study site relative to the Niagara Escarpment and Milton Outlier (darker brown highlighting higher elevation), northeast of the industrial city of Hamilton. (c) Groundwater infiltrates hydraulically conductive units in the Silurian bedrock that transect the karstic basin of Crawford Lake at Nassagaweya Canyon (see Supplemental Figure S3). Reproduced in color in online version.

Figure 3.

Bathymetry (from Boyko, 1973) and physical limnology of Crawford Lake. (a) The 15m isobath identifies the area below which seasonally laminated sediments accumulate and cross-sections (b, c) illustrate the small volume of the monimolimnion (darkest shading; see Figure S4 for water mass key). Measured physicochemical variables demark the water column zonation, with a permanent chemocline (~15.5 m) isolating the dense, electrically conductive monimolimnion (d) from the overlying mixolimnion. (e) Dissolved oxygen concentrations were highest in the metalimnion (lightest grey) but always exceeded the 2mg/L threshold for hypoxia in freshwater ecosystems (vertical line). Modified from Llew-Williams (2022). Reproduced in color in online version.

Figure 4.

(a) Range of water temperature and pH required for calcite to precipitate in Crawford Lake, based on analysis of sediment traps and Langelier Saturation Index calculations. (b, c) These conditions were only met during summer and early fall in the upper 6 m of the water column. Crystals grow when they encounter Ca²⁺ and CO₃²- saturated water along the chemocline where the density contrast slows their descent until they become large enough to overcome Stokes’ Law and sink through the dense and slightly acidic monimolimnion to form the light-coloured summer varve couplet. Modified from Llew-Williams (2022). Reproduced in color in online version.

Figure 5.

Freeze cores CRA19-2FT-1FR-A and CRA19-2FT-B2 collected from the deep basin of Crawford Lake in February 2019; see Supplemental Figure S6 for logs and images of cores CRA19-2FT-B1 and CRA19-2FT-D1. Dashed lines illustrate the ease with which varve-dated couplets can be correlated across freeze cores and the thin calcite lamina deposited in the summer of 1950 CE is indicated by the white arrow. Note that the loose sediments at the top of the cores were not recovered due to the short residence time of freeze corers in the substrate during the 2019 expedition, so varve ages were not determined by counting from the top but always with reference to the thickest white lamina assigned to 1935 CE, the warmest and driest year of the Dust Bowl interval; see Supplemental Table S1, Figure 7, and Supplemental Information for details of varve counting using high-resolution varve imaging (Supplemental Figure S9). Photo credit: K. Lafond. Reproduced in color in online version.

Figure 6.

Composite image of a section (side-view) along core CRA19-2FT-B2 through the Canadian Zone (post-1867 CE); note curvature of horizontal varves as sediments froze onto the sampler as it sank slowly into the lakebed. The age model assumes that the thickest white lamina formed during the exceptionally warm, dry summer of 1935 CE (Lafond et al., 2022, Table S1). At this resolution, transition to and from the white “summer” layer includes substantial organic matter, whereas the center of the white layer consists almost entirely of calcite crystals. The first calcite lamina deposited during the proposed Anthropocene (1950 CE) is near the base of the lowermost of two dark intervals separated by a distinct triplet of varves dated to 1956, 1957, and 1958 CE (see Figure S9). Reproduced in color in online version.

Figure 7.

Composite image (side-view) of sections along the entire length of core CRA22-1FR-3, showing the position of the proposed GSSP. The uppermost enlargement allows the faint white lamina capping the 1950 CE varve year, at the base of the dark layer with almost imperceptible calcite laminae below the distinct triplet (varve ages 1956, 1957, and 1958 CE). The middle image shows the transition from reddish brown to very dark brown organic matter in the mid-19^th century, when logging began. The lower enlargement illustrates the most prominent calcite laminae, deposited when an agricultural settlement was established during the Medieval Warm Interval (Llew-Williams et al., in revision). Reproduced in color in online version.

Figure 8.

Section of core CL-2011, photographed during subsampling for radiocarbon analysis in 2018. This core was analyzed using micro-XRF and for additional palynological information including SCPs in “pollen slides,” with age model based on varve imagery (see Supplemental Information). Photo credit: F. McCarthy. Reproduced in color in online version.

Figure 9.

Activities of total ²¹⁰Pb and ¹³⁷Cs and associated counting error measured approximately every 1 cm from the CL-19 gravity core measured by gamma spectroscopy (Supplemental Table S2). The modelled ages are based on a constant rate of supply model on the estimated unsupported ²¹⁰Pb activity constrained to a tie point at 23 cm (the Ambrosia rise, that was set at 1830 CE ± 20 years). Background (supported) ²¹⁰Pb activities were estimated from the mean of the bottommost 7 samples for total ²¹⁰Pb activities in the CL-19 core (14.8 Bq/Kg ± 3.2) which was similar to the tail of the total ²¹⁰Pb activity in this core (10.8 Bq/Kg ± 10.7, n=7). The correspondence of the ¹³⁷Cs peak in the early 1960s CE supports the age estimates of the CRS model (Supplemental Table S3). Reproduced in color in online version.

Figure 10.

The rapid decline in Ulmus, marking the effects of Dutch elm disease, regionally identifies the base of the proposed Anthropocene series/ epoch (compare with varve-age dated pollen spectra over the last millennium, Supplemental Figure S11). The pollen zonation reveals the much lower rates of sediment accumulation between the intervals of human impact on the Crawford Lake catchment indicated by stippling. An anomalous peak in pine pollen (Pinus) followed by a peak in early successional taxa such as birch (Betula) suggests disturbance during the Little Ice Age, probably by fire. This anomaly was also identified in a freeze core record by McAndrews and Turton (2010) in sediments dating to the early 18th century. Reproduced in color in online version.

Figure 11.

Anthropogenic radiogenic signatures in varved sediments from Crawford Lake, calendar year using varve chronology (Table S1, Figure S9) with horizontal line identifying 1950 CE. (a) Values of F¹⁴C in bulk sediment from core CRA19-2FT-1FR-A and macrofossils from core CL-2011 (Table 2) compare well with atmospheric ¹⁴C concentration (Hua et al., 2022). In addition to old carbon found in DOC and POC throughout the water column (Table S5), bulk sediment samples from core CL-2011 probably retain some dolomitic old carbon because of incomplete digestion. Nonetheless, values of F¹⁴C > 1 were measured through the post-bomb era, identifying the “bomb pulse” of artificially created radiocarbon. (b) Activity of ^239+240Pu (with counting error) in sediments subsampled by varve age from core CRA19-2FT-D1, supplemented by Pu material from core CRA19-2FT-B2 mirrors the global signature of atmospheric deposition of plutonium from above-ground nuclear testing (United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), 2000). Reproduced in color in online version.

Figure 12.

SCPs in core CRA19-2FT-B2. (a) Concentrations increase sharply between 1948 and 1954 CE. (b) Particles larger than 10 µm occur throughout, while smaller particles gradually increase in abundance after 1945 CE. The decline in SCPs larger than 25 µm coincides with closure of the lumber mill at Crawford Lake in 1900 CE. The proposed base of the Anthropocene is shown at 1950 as a solid horizontal line. Reproduced in color in online version.

Figure 13.

Organic proxies in sediments subsampled by varve age from core CRA19-2FT-D1, supplemented by material from core CRA19-2FT-B2. Sharply depleted values of δ¹⁵N mark the proposed base of the Anthropocene epoch (1950 – horizontal line). High values of TN and TOC are in dark calcite-poor sediments that characterise the 1950s and 1960s (see Figures 5, 14), attributed to acidic precipitation. Reproduced in color in online version.

Figure 14.

Laminae visible in the original core photo (Figure 6) superimposed on the ITRAX-generated RBG scan allow sediment geochemistry determined using µXRF analysis of freeze core CL-2011 to be related directly to the varved sediments. Peaks in activity attributed to Cu and Pb peak around the proposed base of the Anthropocene where SPCs rich in these metals are abundant (see Figure 12). Peaks in activity attributed to Ca align with prominent light-coloured laminae (with prominent laminae deposited during the summers of 1956, 1957 and 1958 CE identified across the diagram by dashed lines) whereas peaks in terrigenous elements, such as Fe, K and Ti in dark layers record enhanced leaching of the catchment. Compare the µXRF record in the upper 20 cm of gravity core CL-19 (Figure S21). Reproduced in color in online version.

Figure 15.

Summary fossil pigment concentrations in core CL-19 – note the absence of okenone above 38 cm, recording the unusual absence of anoxia in this meromictic lake since the early 16^th century. The ratio of chlorophyll a (not plotted individually) to pheophytin a (Chla : pheoa ratio) suggests poor preservation in the sample from 35 cm where an anomalous peak in resistant and readily transported pine pollen was identified (see Figure 10). Gray shading identifies intervals of human impact, Indigenous and colonial-Canadian. Reproduced in color in online version.

Figure 16.

Summary of analysis of siliceous microfossils at annual resolution from core CRA19-2FT-D1. Stratigraphically constrained cluster analysis identified three zones, with Zone B defined by a decline in the planktonic diatoms Lindavia michiganiana and Fragilaria crotonensis that dominated Zone C and sharp increases in non-planktonic taxa such as Achnanthidium spp. Significant dissimilarity was also noted around 1950 CE (solid line), primarily in the scaled chrysophyte record, with several species of Synura dominating assemblages in the upper part of zone B. Modified from Marshall et al. (2023). Reproduced in color in online version.

Figure 17.

Summary of key Anthropocene markers and the synchronous ecosystem response recorded by siliceous microfossils around 1950 CE in varve-age dated samples from freeze cores from the deep basin of Crawford Lake. Abundant SCPs and depleted δ¹⁵N values record increased fossil fuel combustion following the Second World War, when plutonium fallout first appears in varved sediment. The increase in non-planktonic diatoms and the scaled chrysophyte genus Synura across the proposed Anthropocene GSSP (marked by horizontal line) begins in the 1940s and declines through the 1970s, as do the physical Anthropocene markers. Reproduced in color in online version.

Figure 18.

Summary of key Anthropocene markers in freeze core CL-2011. The proposed base of the Anthropocene (1950 CE, shown by horizontal line) is marked by high XRF activity attributed to terrigenous elements (e.g., titanium) in calcite-poor sediments and a sharp increase in concentration of SCPs in palynological preparations. The radiocarbon ‘bomb pulse’ coincides with the sharp decline in Ulmus pollen resulting from Dutch elm disease. Reforestation also helps identify the late 20^th century, as do several non-pollen palynomorphs, such as loricae of the chrysophyte Dinobryon divergens and the rotifer Keratella quadrata, almost unknown from earlier sediments from Crawford Lake. Reproduced in color in online version.

Figure 19.

Summary of key Anthropocene markers plotted against depth in gravity core CL-19. a) Increases in terrigenous elements and non-arboreal pollen (NAP) identify two distinct intervals of human impact on Crawford Lake, highlighted by stippling. The decline in elm pollen (Ulmus) and increase in ¹³⁷Cs activity identify the mid-20^th century, when increased fossil fuel combustion produced a sharp increase in SCP concentration in palynological preparations. b) Fossil pigments and algal palynomorphs record an increase in green algae and the chrysophyte Dinobryon divergenssince the mid-20^th century, recording a slight increase in overall primary productivity. The absence of the fossil pigment okenone confirms the well-oxygenated water column in this unusual meromictic lake since the early 16^th century. Reproduced in color in online version.