Annual-scale assessment of mid-20th century anthropogenic impacts on the algal ecology of Crawford Lake, Ontario, Canada

Study area

Crawford Lake (43°28′06″N, 79°56′55″W), situated on dolomitic bedrock within the Crawford Lake Conservation Area in Milton, Ontario, features a maximum water depth of ∼24 m, surface area of approximately 2.4ha, and is at an elevation of 286 m above sea level. The deflection of wind from nearby forests comprised of Thuja, Betula, Ulmus, Quercus, and Pinus (McAndrews & Boyko-Diakonow, 1989) and proximity to the Niagara Escarpment with highly alkaline groundwater inputs (Llew-Williams, 2022), as well as the general shape and depth of this karst basin, provides a sheltered environment conducive to meromixis. This environment, with non-mixing but well-oxygenated bottom waters below a sharp chemocline at ∼15.5 m water depth with distinctive pH and conductance conditions, impedes lake bottom bioturbation (Heyde, 2021), allowing seasonal varves to accumulate undisturbed on the lakebed. Light-colored layers formed of calcite precipited in the epilimnion during the summer and predominantly authigenic organic matter accumulates the rest of the year, although primarily during fall turnover (Dickman, 1979; Dickman, 1985; Llew-Williams, 2022) forming the dark laminae. Previous research shows the lake to be alkaline, with high specific conductance but pH lower than 7 in the monimolimnion (Ekdahl et al., 2007; Gushulak et al., 2021; Llew-Williams, 2022).

The climate in the Halton region of southern Ontario is humid continental, with average yearly air temperature ∼7.3 °C (median 7.6) and mean annual precipitation of ∼650 mm (median 632 mm) between the years 1937 and 1990 (Toronto Airport, Ontario: Environment Canada, 2020). Toronto Airport is 35 km east of Crawford Lake. There were 12 years with elevated total summer rainfall >400 mm and one year (1945) with >500 mm. Average annual snow fall from November to March 1937–1990 was 127 mm (median 125) with two years of heavy snow >200 mm, 1950 and 1972.

Sample collection

A two-faced freeze core was collected (CRA19-2FT-B2), 83 cm in length, in February of 2019 by RTP and associates, using ethanol and dry ice to freeze sediment along the length of a flat faced metal prism. The collection method was similar to that described in Glew, Smol & Last (2001). The core was collected at a depth of ∼22 m, with the sampling location shown in Fig. 1, and transported frozen to Carleton University for subsequent processing and analysis. A scalpel was used to isolate and subsample adjacent light and dark bands varve couplets deposited each calendar year. Each varve-year was matched to the corresponding Gregorian calendar year using known climate phenomena within the area, such as thick varves representative of the Dust Bowl, and were then counted forward and backward temporally using distict varves which are well-preserved within the slabs of frozen sediment. This varve age model was confirmed by analysis of radionuclides whose distribution in the varved sediment mirrors nuclear fallout (McCarthy et al., 2023). Sixty-one subsamples, corresponding to the years 1930 to 1990 CE were prepared for analysis. Diatom analysis was restricted between 1930 to 1980 CE.

Siliceous algae analysis

Each subsample was processed for siliceous algae following a protocol similar to Battarbee et al. (2001). For this methodology, ≤0.1 g of sample from each varve-identified year was treated with 10% HCl to remove carbonates, and then washed with distilled water numerous times (∼10 washes) until the supernatant reached the pH of the distilled water, to reduce the concentration of dissolved calcium, and neutralize the pH. These samples were then treated with 30% H₂O₂ to remove organic matter, and subsequently put through a series of washes until reaching the pH of the distilled water used for washes. These siliceous slurries were then pipetted, following a series of $\frac{1}{2}$ dilutions until an appropriate siliceous microfossil density was reached, onto coverslips and allowed to dry for ∼12–18 h on a warming tray before being affixed onto microscope slides using Cargille Meltmount™ 1.704 (Cargille, Wayzata, MS, USA). The slides were examined under an Olympus BX51 microscope at 1,000x oil-immersion magnification (NA 1.25), to identify chrysophyte scales and diatom valves to species level, with statistically significant populations of ∼400 scales and ∼600 valves counted for each (Patterson & Fishbein, 1989).

The siliceous remains of certain species that did not present easily-observable, taxon-defining ornamentation under standard light microscope conditions were verified using scanning electron microscopy (SEM). For SEM analysis, subsamples from previously prepared sample slurries were pipetted onto aluminum foil or glass coverslips, dried and mounted onto scanning electron microscope stubs using double-sided carbon tape. The sample stubs were then coated with gold-palladium using a Denton DESKII sputter coater. The sample stubs were examined using either the Tesca VegaII XMU SEM at the Nano Imaging Facility at Carleton University or the FEI Apreo SEM at the Canadian Museum of Nature (Ottawa, Canada) using images created by both secondary and back-scatter electrons. SEM working distances ranged from 3–10 mm under 5KV. Scales and frustules were identified using images and descriptions found within the taxonomic literature and the SEM image collection at the Canadian Museum of Nature (Siver, 1991; Krammer & Lange-Bertalot, 1988; Houk, Klee & Tanaka, 2010; Siver & Hamilton, 2011; Alexson et al., 2022; Guiry & Guiry, 2022). The chronology of examined samples was varve-year inferred by counting from known events such as the thick layers associated with the Dust Bowl. Samples and slides are deposited at the national phycology collection at the Canadian Museum of Nature (CANA 129343-129391).

Statistical analysis

Species abundance of the statistically significant taxa were plotted using Tilia v. 2.0.2 (Grimm, 1987). A stratigraphically constrained cluster analysis (CONISS) was performed on the combined chrysophyte scale and diatom valve relative abundance data. This analysis provides clusters based on an incremental sum of squares method and Euclidean distance, considering only stratigraphically adjacent samples to better represent groups through the history of the sediment record (Grimm, 1987). Separate zones marked through CONISS were validated with a broken-stick model using the rioja package (Juggins, 2020) within R studio (R Core Team, 2020). Non-metric multidimensional scaling (NMDS) was performed on non-transformed and square-root transformed species abundance data, using Bray–Curtis Dissimilarity as distance within R studio using the vegan package (Oksanen et al., 2020), to view assemblage and species data with reduced dimensions, and allow for visualization of data within a two-dimensional graphical interpretation; as such, the NMDS bi-plot was a second metric along with CONISS to identify and visualize groups in a reduced dimensional space.

Chrysophyte-delineated zones

The 1930–1952 Pre-Anthropocene Zone was characterized by a large proportion of Mallomonas pseudocoronata Prescott; Mallomonas tonsurata Teiling being the second most dominant species of the interval. Combined, these two species generally provide more than 90% of the scales counted in this zone.

Anthro-Transitionary Zone 1 was defined by a large and continuous increase of Synura species, Synura curtispina Asmund, Synura echinulata Korshikov and Synura petersenii Korshikov, changing from <10% abundance in most years within the Pre-Anthropocene Zone to becoming the species present in highest abundance. This zone also captured the large increase in SCP concentration within the sediments of Crawford Lake (Figs. 2 and 3).

Anthro-Transitionary Zone 2 (1970–1978) was defined by the spike of Mallomonas acaroides Zacharias sensu stricto, followed by a community shift to M. pseudocoronata with a decrease of Synura species compared to the previous zone. Between the shift from M. acaroides to M. pseudocoronata, there was a short notable increase in S. curtispina and S. petersenii (Fig. 3). Within this interval, SCP concentration began to decline from the high present within the previous zone and had the lowest relative abundance of Mallomonas striata Harris and Bradley sensu stricto, which was present in all zones.

The Post-Transitional Zone, characterizing Crawford Lake from 1979 to 1990 had a total dominance of M. acaroides (>70% relative abundance) with minor populations of M. striata, M. pseudocoronata and M. tonsurata. Synura species made up less than 5% scale abundance through this time interval.

Non-metric multidimensional scaling (NMDS) analysis showed the Pre-Transitional Zone plotting in the same space as M. pseudocoronata and M. tonsurata, Anthro-Transitionary Zone1 plotted with S. echinulata, S. curtispina and S. petersenii. The zone Anthro-Transitionary 2 formed a broad scatter having no tangible species alignments with the exception of a distant association of S. uvella with one assemblage (Fig. 4). The Post-Transitional Zone assemblages plotted together showing a clear association with M. acaroides.

Diatom-delineated zones

From 1930–1941, the Pre-Transitional zone comprised Fragilaria crotonensis Kitton, Lindavia michiganiana (Skvortzov) T.Nakov et al., and Discostella stelligera (Cleve & Grunow) Houk & Klee at their highest relative abundances within the observed record, and generally made up the majority of observed valves (Fig. 5).

The Anthro-Transitionary Zone (1942–1968) featured a decrease in the relative abundance of F. crotonensis, a large increase in the relative abundance of Cyclotella distinguenda Hustedt and Achnanthidium minutissimum (Kützing) Czarnecki, as well as the appearance of Fragilaria vaucheriae (Kützing) J.B.Petersen in quantities above 5% relative abundance (Fig. 5). This Zone also showed increases of other tychoplanktic and benthic taxa such as Meridion circulare (Greville) C.Agardh, Navicula radiosa Kützing, and Gomphonema variostriatum K.E. Camburn & D.F. Charles. A subdivision of this zone was noted at 1962 with a subsequent general decline of F. andreseniana Alexson et al., and increases in C. distinguenda and A. minutissimum.

The communities within the Post-Transitional Zone (1969–1978) were mostly of the planktic and tychoplanktic species Fragilaria andreseniana, Lindavia affinis (Grunow) Nakov, Guillory, M.L.Julius, E.C.Theriot & A.J.Alverson and C. distinguenda. Through this period there was a shift from A. minutissimum, L. affinis and C. distinguenda to F. andreseniana and then back to A. minutissimum and C. distinguenda (Fig. 5).

NMDS of the diatom communities showed the Pre-Transitional samples were within the same space as F. crotonensis, Fragilaria spp, D. stelligera, and L. michiganiana (Fig. 6). Assemblages from the Anthro-Transitionary Zone plotted toward Post-Transitional Zone but were still distinct and associated with many diatom species. Pre-1960, the assemblages contained small araphids (Staurosira spp, Staurosirella spp and Pseudostaurosira brevistriata (Grunow) D.M.Williams & Round), along with other benthic forms including Encyonema spp, Adlafia spp. and Nitzschia spp. Post 1960, the Anthro-Transitionary Zone plotted in the same space as the monoraphids (Achnanthidium spp), as well as the biraphid taxa; Gomphonema variostriatum and Brachysira microcephalum. Assemblages belonging to the Post-Transitional Zone were scattered along the second the second NMDS axis,, but are associated with species such as L. affinis, F. andreseniana, other Fragilaria spp, and Nitzschia taxa (Fig. 6)

Initial anthropogenic transition (1942-1968)

Through the 1930s prior to the transition, the scaled chrysophytes showed little to no alteration in the community assemblage, while the planktic dominant diatom community fluctuates between centric diatoms and the pennate Fragilaria crotonensis, associated with changes in productivity (Hutchinson, 1967). Although Synura species are found lower within the water column comparatively to those of the Mallomonas genus, all these species require sunlight to maintain photosynthetic activities needed for survival. As such, the observed increases, and eventual decreases in Synura taxa, is concomitant with the massive changes in SCP deposition, and reduced calcite precipitation, which are most likely associated with changes in light-attenuation regimes within the waters of Crawford Lake brought about through a regional weather and a reduction of humic substances within the water in response to acidification related to industrial emissions (Monteith et al., 2007; Hruška et al., 2009; Llew-Williams, 2022). Through this period into the early 1950s elevated summer rainfall (5 years) possibly enhanced local terrestrial loading with four years of rainfall >450 mm. Dilution of surface waters at this time is evidenced by very thin light-coloured laminae, since warm summer temperatures and high pH (high concentrations of Ca⁺² and CO₃⁻² ions) are required for precipitation and accumulation of calcite crystals (Llew-Williams, 2022; McCarthy et al., 2023).

Diatom community changes occurred in the early 1940s with a shift from planktic species to tychoplanktic and benthic; this shift to deeper-dwelling species associated with these waters increasing in clarity (Ekdahl et al., 2004; Gushulak et al., 2021). From 1941 to 1942, there was a dramatic increase in summer storms (262 mm to 479 mm of accumulation) in the region creating local impacts to the lake. Regional storms likely affected terrestrial loading to the lake with the observed switch from planktic to tychoplanktic and benthic species. Although not linked directly to peak carbon particle concentrations, changes in the diatom community occur with the first stepwise increase of SCPs also suggesting the lake environment was changing. SCP terrestrial deposition from previous years could be impacting SCP deposition rates in the lake with the more frequent summer storms increasing loading through the 1940s, which may further impact algal communities with an increased availability of heavy metals related to the sources of SCP deposition (McCarthy et al., 2023). The diatom community was further characterized by two sub-assemblages during the transition period, pre- and post-1962 (Fig. 6). Diatom assemblages in a constrained CONISS analysis shows a species shift after 1962 (Fig. 5). Community assemblages are changing through time, but the significance of these changes is unclear, although average monthly summer air temperatures (May-September) were elevated (>18 C) from 1959 to 1961 with 1959 temperatures reaching 19.8 C, the highest levels during the Great Acceleration in the region. Achnanthidium minutissimum and C. distinguenda are the primary species in the pre- and post- 1962 shift. Likewise, Synura species are increasing and then declining during the same period. Markers of global anthropogenic signatures on Crawford Lake would include plutonium (Pu) from the peak of atomic testing during the transition period. McCarthy et al. (2023) show Pu^239,240 concentrations in the lake sediments increase through the early 1950s, peak in 1963 and then decline. Although Pu and algal assemblages are not causally correlated, the metrics both show a similar trend that can be linked to a time in which the lake was documenting and also impacted by global anthropogenic stressors, and climate.

The probable driver for the shift in chrysophyte assemblages to favoring Synura over Mallomonas as well as the changes within diatoms to favor a higher relative abundance of benthic species is believed to be weather patterns with increased industrial emissions and acidic deposition on the lake catchment. These changes result in reduced transport of humic substances into the water column, and increasing light penetration within the waters. This is brought about by WWII and post-WWII industrialization, including increased ore smelting and steel production within the Toronto-Hamilton regions and further across eastern North America (Arnold, 2007; Arnold, 2012). Associated with more industry is increased human settlement, urban expansion, and higher motor vehicle usage, all of which would be associated with the observed increase in atmospheric SCPs and deposition within the area (Fig. 2; Steffen et al., 2011; Rose, 2015; McCarthy et al., 2023). The size fractionation of SCPs through the study show the prominence of 10–25 µm carbon particles throughout the ∼150 years of analysis (Fig. 2). However, during the Great Acceleration (1940–2010) smaller particles (<10 µm) appear and persist, comprising <20% of the contamination, reflecting longer distance transport. The large particles that are relatively common (albeit in low total concentrations) earlier are probably derived from local burning of coal. Accumulating SCP and smelting gases (e.g., sulfur) create the potential for lake acidification. Crawford Lake is an alkaline buffered system that should withstand atmospheric acidification events. Despite this, like many other bodies of water with well-buffered catchment areas (Hruška et al., 2009; Meyer-Jacob et al., 2020), Crawford Lake shows changes implying acid impacts.

Although Crawford Lake is a highly buffered system overlaying a dolomitic limestone base, it is most probably regional weather and impacts of acid deposition on the catchment of the lake that are probably the most significant drivers behind the inferred changes to the water clarity and sediment/particulate loading. Acidification impacts on the lake may have been amplified through chemical interactions and the binding of organic humic substances which would contribute to the reduction of dissolved organic carbon (DOC) reaching the lacustrine system (Monteith et al., 2007; Hruška et al., 2009). The peak of acid deposition within this region occurred by 1970 and was due primarily to large emissions of sulfur and nitrogen oxides from industrial emitters both locally in Ontario and from upwind sources in the industrial heartland of the US; this trend is mirrored in the observed deposition of SCPs within Crawford Lake (National Air Pollution Surveillance Program, 2020; National Pollutant Release Inventory, 2022).

Biologically, studies have found similar trends in Synura species abundance changes related to DOC content within the water column (Hadley, 2012). Like scaled chrysophytes, the diatom shifts from planktic to littoral and benthic taxa, along with the recognition of disturbance tolerant species like A. minutissimum (Jüttner et al., 2022), indicate water clarity, along with less obvious factors played a role in the 1940’s anthropogenic shift. During the Anthro-Transitionary period when weather and anthropogenic acid deposition was impacting Crawford Lake, reforestation was also occurring in the immediate area, which may have contributed to increasing water-clarity through reduction in erosion and sediment export into connected waters (Keller & Fox, 2019).

Post-transition (1968-1990)

The later shifts in assemblages to post-Transitional are hypothesized to have been caused by the opposite effect; the lake waters becoming less clear due to local stressors and a reduction in acid deposition within the region caused primarily by emissions legislation changes. During the 1968–1969 transition period, regional summer rain levels dropped from 408.9 to 293.3 mm indicating fewer storms and possibly higher lake productivity. Also, at this time the acquisition of the Crawford Lake area into lands protected by Conservation Halton was completed with educational related site construction activities (Meyer-Jacob et al., 2019; Meyer-Jacob et al., 2020; National Air Pollution Surveillance Program, 2020). The change from this Anthro-Transitionary period of clear water termed by some as “water re-browning”, has been reported from other well-buffered, as well as chronically acidic lakes, that suffered from similar anthropogenic impacts (e.g., Meyer-Jacob et al., 2019; Meyer-Jacob et al., 2020).

The large assemblage shift that occurred in the 1970’s coincides with the re-emergence of Mallomonas species in high abundances within the chrysophytes and an increase in centric planktic species within the diatom community. Although this change was in part brought about by water re-browning after recovery from acid impacts, it may also have been impacted by Crawford Lake’s incorporation into the Crawford Lake Conservation Area with boardwalk construction around the lake initiated in the late 1960s along with land alteration near the park in 1969 as well as a fire that destroyed the adjacent Crawford family home around the same interval (Muller, 1982; B. Bartley, Halton Conservation, 2020, pers. comm.) In addition, the region experienced a clear decline in rainfall between 1968 and 1969 (Supplement 3, Environment Canada, 2022). The diatom assemblage shift, at this time, was the second most important during the Great Acceleration (Fig. 5). In light of the continued occurrence of A. minutissimum at the beginning of the Post-Transition, then decline over the 1970s, as well as the appearance of L. affinis, we suggest that regional weather and boardwalk construction (local anthropogenic impact) in the late 1960s assisted in the observed transition. This transition was associated with increasing water turbidity and amplifying the re-browning of lake waters, which in turn reduces the amount of light benthic taxa such as A. minutissimum have for photosynthetic activities. The shift to M. acaroides at the Anthro-Transitionary 2, followed by a sharp shift to M. pseudocoronata, then shift back to favouring M. acaroides, also supports the observation of weather and a local anthropogenic disturbance.

The Post-Transitional change in scaled chrysophytes back to a dominance of M. acaroides during the 1980s is not aligned with annual rainfall or seasonal air temperatures, but is associated with Conservation Halton’s second phase of construction activities in the early 1980s, in which a boardwalk was completely constructed around the lake (B. Bartley, Conservation Halton, 2020, pers. comm.). The planktic diatom Asterionella formosa Hassall appeared for the first time during the Post-Transition showing changes in lake eutrophication status associated with disturbance (e.g., Reynolds, 1984). The dominance of M. acaroides throughout the 1980s clearly shows a consistent longer-term change in the lake community. It is important to note, that changes in abundance of Synura species, and replacement of M. pseudocoronata with M. acaroides have been observed within lakes elsewhere in Canada, some of which are considered minimally impacted, and may be resultant from changes in CO₂ regimes brought about by climate change (Mushet et al., 2017; Mushet et al., 2018; Wolfe & Siver, 2013).

The documentation of weather and anthropogenic stressors at the local, regional and likely global scale show the sensitivity of, and impact on Crawford Lake. Climate change and water chemistry shifts brought about by anthropogenic impacts, as well as similar trends being seen in algal assemblages elsewhere, indicate the preserved communities within the sediments of Crawford Lake are representative of regional (eastern North America) weather and industrialization trends as well as the associated anthropogenic impacts on the global environment. The historical sediment record documenting and assessing past anthropogenic impacts by First Nations cultures further supports the sensitivity of the Crawford Lake ecosystem to external stress across long term spatial and temporal scales.