Abigail S. L. Lewis1, Madeline E. Schreiber2, B. R. Niederlehner1, Arpita Das1, Nicholas W. Hammond2, Mary E. Lofton1, Heather L. Wander1, Cayelan C. Carey1
1Department of Biological Sciences, Virginia Tech, Blacksburg, Virginia, USA
2Department of Geosciences, Virginia Tech, Blacksburg, Virginia, USA
Corresponding author: Abigail S. L. Lewis (aslewis@vt.edu)
Key Points:
Abstract
Freshwater lakes and reservoirs play a disproportionate role in the global organic carbon (OC) budget, as active sites for carbon processing and burial. Associations between OC and iron (Fe) are hypothesized to contribute substantially to the stabilization of OC in sediment, but the magnitude of freshwater Fe-OC complexation remains unresolved. Moreover, global declines in bottom-water oxygen concentrations have the potential to alter OC and Fe cycles in multiple ways, and the net effects of low-oxygen (hypoxic) conditions on OC and Fe are poorly characterized. Here, we measured the pool of Fe-bound OC (Fe-OC) in surficial sediments from two eutrophic reservoirs, and we paired whole-ecosystem experiments with sediment incubations to determine the effects of hypoxia on OC and Fe cycling over multiple timescales. Our experiments demonstrated that short (2–4 week) periods of hypoxia can increase aqueous Fe and OC concentrations while decreasing OC and Fe-OC in surficial sediment by 30%. However, exposure to seasonal hypoxia over multiple years was associated with a 57% increase in sediment OC and no change in sediment Fe-OC. These results suggest that the large sediment Fe-OC pool (~30% of sediment OC in both reservoirs) contains both oxygen-sensitive and oxygen-insensitive fractions, and over multiannual timescales OC respiration rates may play a more important role in in determining the effect of hypoxia on sediment OC than Fe-OC dissociation. Consequently, we anticipate that global declines in oxygen concentrations will alter OC and Fe cycling, with the direction and magnitude of effects dependent upon the duration of hypoxia.
Plain Language Summary
Freshwater lakes and reservoirs (hereafter: lakes) play a remarkably important role in the global carbon cycle, as important sites for both carbon sequestration and greenhouse gas emissions. The extent to which lakes sequester carbon vs. release greenhouse gases depends upon many factors, including associations between carbon and iron that can help to preserve carbon in sediment. However, global declines in freshwater oxygen concentrations have the potential to affect these chemical complexes. Here, we added oxygen to the bottom waters of a lake to test how changes in oxygen concentration affect carbon and iron cycling. We found that over short timescales (weeks), low oxygen conditions decreased the amount of carbon in sediment by breaking apart associations between iron and carbon that help retain carbon in sediment. However, over long timescales (years), low oxygen conditions appeared to increase carbon burial by decreasing the rate at which carbon inputs were decomposed. These results suggest that declining oxygen concentrations in lakes around the world may have important effects on global carbon cycling, with the direction and magnitude of the impact depending on the duration of low oxygen conditions.
1. Introduction
Freshwater lakes and reservoirs are increasingly recognized as hotspots in the global carbon cycle (Bastviken et al., 2011; Battin et al., 2009; Carey, Hanson, et al., 2022; Raymond et al., 2013; Tranvik et al., 2018). Due to high organic carbon (OC) loading from the surrounding watershed, lakes and reservoirs likely bury more OC than coastal sediments each year (Dean & Gorham, 1998; Knoll et al., 2013; Mendonça et al., 2017; Pacheco et al., 2014; USGCRP 2018). Much of this OC remains sequestered in the sediments, especially in reservoirs, which may bury organic carbon at over six times higher rates than natural lakes (per unit area; Mendonça et al. 2017). However, OC inputs can also be respired to carbon dioxide and methane, making lakes and reservoirs a source of greenhouse gas emissions equivalent to 20% of the global emissions from fossil fuels (Deemer et al., 2016; DelSontro et al., 2018). The balance between carbon burial and emission in freshwater systems is controlled by numerous factors, notably including associations between OC and iron (Fe). To refine global carbon budgets and manage water resources in a changing world, it is critically important to quantify the role of Fe in OC cycling in lakes and reservoirs against the backdrop of rapidly-changing environmental conditions.
Recent research suggests that associations between OC and iron (Fe) may play a critical role in OC sequestration, though the importance of Fe in freshwater OC dynamics is currently unresolved. Fe can promote OC stability through multiple mechanisms, including occlusion of OC in aggregates, which can result in physical inaccessibility to microbial degradation and subsequent burial of OC in deeper soil or sediment horizons (Kleber et al., 2015 and references therein). Consequently, protection of OC through complexation with Fe may facilitate OC sequestration over decades to millennia (Kleber et al., 2015; Lalonde et al., 2012 and references therein). To date, most research on OC-Fe cycling has focused on terrestrial and marine environments, where it has been shown that Fe-OC complexation can serve as an important mechanism for stabilizing OC, the so-called “rusty carbon sink” (e.g., Barber et al. 2017; Hemingway et al., 2019; Kramer & Chadwick, 2018; Lalonde et al., 2012). In contrast to terrestrial and marine ecosystems, only a few studies have explicitly examined Fe-OC in freshwater lakes and reservoirs. Peter and Sobek (2018) analyzed Fe-OC in surficial sediment from five boreal lakes that spanned a gradient of oxygen conditions and found that less than 11% of sediment OC was bound to Fe, in comparison with ~20% across a range of primarily marine sediments (Lalonde et al., 2012). Furthermore, Peter and Sobek (2018) found no association between Fe-OC content in sediment and oxygen in overlying water when comparing across lakes. However, it should be noted that the lakes in that study were particularly high in dissolved OC (DOC) concentrations (9–42 mg/L DOC), and may not be representative of all freshwater ecosystems. Bai et al. (2021) studied Fe-OC along a salinity gradient in a subtropical tidal wetland and similarly found that freshwater areas had lower levels of sediment Fe-OC (18% of sediment OC in freshwater and 29% in saltwater), but these results were attributed primarily to wetland plant characteristics, which may not be relevant in the bottom waters of lakes and reservoirs. Differences in sediment Fe-OC between freshwater and saltwater environments may be expected based on water column characteristics, as increasing ionic strength can increase aggregation and flocculation of Fe, with differential effects depending on the quantity and quality of organic matter (Heerah and Reader, 2022; Beauvois et al. 2021; Herzog et al. 2020).
Despite limited research on Fe-OC in freshwater sediments, there are multiple reasons to expect that Fe may play an important role in OC sequestration in some freshwater ecosystems. Concentrations of Fe and DOC are strongly correlated in many freshwaters (Björnerås et al., 2017; von Wachenfeldt et al., 2008; Weyhenmeyer et al., 2014), and aqueous Fe concentrations are correlated with sediment OC accumulation in boreal lakes (Einola et al., 2011). Moreover, it is well-documented that DOC can be released from lake sediments under low-oxygen conditions, and this sediment flux is often attributed to reductive dissolution of Fe (Brothers et al., 2014; Yang et al. 2014; Kim & Kim, 2020; Lau & del Giorgio, 2020; Peter et al., 2017). Still, few studies have examined whether reactions involving Fe-OC complexes are the driving force for observed correlations between dissolved Fe and OC (but see Peter et al. 2018). Furthermore, it remains unknown how the Fe-OC cycling occurring on sub-annual time scales may affect OC sequestration on the multi-annual timescales relevant for global carbon budgets.
Over short timescales (days to weeks), Fe-bound OC (Fe-OC) complexes are sensitive to the redox conditions of the surrounding environment (Figure 1). Fe-OC complexes form under oxic conditions (Riedel et al., 2013), as Fe(III) is more effective at complexing with organic matter than Fe(II) (Nierop et al., 2002). Under hypoxic conditions (low oxygen, defined here as < 2 mg/L following, e.g., Yang et al. 2014), OC can be released from Fe-OC complexes through Fe(III) reduction and dissolution (e.g., Pan et al., 2016; Patzner et al., 2020; Skoog & Arias-Esquivel, 2009), which can either result directly from hypoxia or through resultant increases in pH that promote OC release (Kirk, 2004; Thompson et al., 2006). Given these conflicting patterns—i.e., that Fe-OC complexes can be preserved over decades to millennia and yet may be unstable under the reducing conditions which commonly occur on day to month timescales in aquatic sediments—it remains unclear how changing oxygen dynamics will affect coupled OC and Fe cycling in freshwater ecosystems.
Currently, the duration of bottom-water hypoxia is increasing in many lakes and reservoirs around the world (Bartosiewicz et al., 2019; Jane et al., 2021; Jenny et al., 2016; Williamson et al., 2015), which could have varying consequences for OC sequestration (Figure 1). In many dimictic lakes and reservoirs, bottom-water hypoxia is interrupted by oxic conditions during spring mixing and fall turnover, resulting in dynamic oxygen conditions on the week to month scale. Combined, these short-term patterns sum to determine the net role of lakes and reservoirs in the global carbon cycle over multiannual timescales. Periods of hypoxia have the potential to decrease OC sequestration through reductive dissolution of Fe(III) in Fe-OC complexes (Chen et al., 2020; Huang et al., 2021; Patzner et al., 2020). However, hypoxia also has the potential to increase OC sequestration by decreasing the rate of OC respiration (Carey et al., 2018; Carey, Hanson, et al., 2022; Hargrave, 1969; Peter et al., 2017; Sobek et al., 2009; Walker & Snodgrass, 1986), particularly if Fe-OC complexes are resistant to, or protected from, changes in oxygen concentrations in overlying water. Decreased OC respiration rates under hypoxic conditions is thought to occur primarily because respiration is less thermodynamically favorable in the absence of oxygen (e.g., Arndt et al., 2013; LaRowe & Van Cappellen, 2011). Because reductive dissolution of Fe(III) in Fe-OC complexes and decreased OC respiration under hypoxic conditions would have divergent effects on total OC sequestration, understanding the relative importance of these two processes across multiple timescales is critical for predicting the effect of hypoxia on OC sequestration in the bottom waters of lakes and reservoirs (Figure 1).
Analyzing the complex effects of oxygen on coupled OC and Fe cycling requires multiple experimental approaches. Field surveys have been effective at identifying correlations between OC and Fe (Björnerås et al., 2017; von Wachenfeldt et al., 2008; Weyhenmeyer et al., 2014). However, these observational approaches have limited capacity for identifying causal relationships. Whole-ecosystem experiments may be highly effective at identifying real-world impacts of freshwater oxygen on Fe and OC dynamics, while allowing for important ecosystem-scale processes such as turbulence and external loading (Carpenter, 1996; Dzialowski et al., 2014; Schindler, 1998). However, high levels of variability on a whole-ecosystem scale may limit the detection of subtle changes in OC and Fe processing. Small-scale incubations may be particularly useful for identifying changes that result from hypoxia (i.e., increased DOC and Fe release from sediment, decreased levels of Fe-OC, changes in sediment OC). However, small-scale incubations are limited by fouling and changes in microbial communities, among other microcosm effects, and do not reflect the full suite of processes that interact to control OC and Fe cycling in lakes and reservoirs. Consequently, integrating multiple approaches can provide complementary information on Fe-OC dynamics across spatial and temporal scales and overcome the limitations of single-approach studies.
To analyze how hypoxia impacts OC and Fe cycling over multiple scales, this study paired whole-ecosystem oxygen manipulations with laboratory incubations. We had two objectives: (1) characterize Fe-OC (operationally defined as dithionite-extractable OC) levels in sediment of two iron-rich reservoirs, and (2) analyze how hypoxia affects coupled OC and Fe cycling over both short-term (2–4 week) and multiannual timescales. Through this work, we aimed to provide insight on how increasing prevalence and duration of hypoxia in lakes and reservoirs may affect the critical role of these ecosystems in the global carbon cycle.