In the changing Arctic, predictions of organic carbon and nitrogen fluxes in marine food-webs are critical to understanding future ecosystem productivity and the efficiency of energy transfer to higher trophic levels (Leu et al., 2011; Tamelander et al., 2013). Concentrations and ratios of particulate organic carbon (POC) and nitrogen (PON) in both sea ice and the water column are important parameters for Arctic ecosystem modelling at a time when patterns of Arctic productivity and the transitional period from ice algae to phytoplankton are changing (Arrigo, 2013). Biogeochemical modelling (e.g., process and regional models), using appropriate measured parameters, is an important tool for the integration of Arctic climate and ecological system variability, enabling the description and prediction of current and future states of Arctic ecosystems (Tedesco et al., 2012).
The fundamental relationship between carbon and nitrogen ratios (carbon:nitrogen, C:N) sets the stoichiometric basis for modelling biogeochemical processes in aquatic systems. The classical Redfield ratio (carbon106:nitrogen16:phosporus1; Redfield et al., 1963) provided the theoretical and stoichiometric foundation for elemental relationships between particulate and dissolved components of marine and freshwater systems. Although the Redfield ratio is widely used in modelling and nutrient-based production estimates (Mills and Arrigo, 2010; Sambrotto, 2014), departures from the classical Redfield proportions and plasticity in C:N ratios are well-documented for both particulate and dissolved fractions (Sterner et al., 2008; Martiny et al., 2013; Letscher and Moore, 2015). Variability in the elemental ratios of seston has been linked to several factors including inter-species differences, storage products and impacts of varying nutrient conditions (Geider and La Roche, 2002; Arrigo, 2005; Frigstad et al., 2011).
Reliance on the assumption of a fixed C:N ratio may limit or misdirect the modelling of biogeochemical cycles in large or small scale models (Vancoppenolle et al., 2013). In this regard, recent reviews have considered specificity and scale-dependence of stoichiometric ratios (e.g., C:N) for marine and freshwater ecosystems (Sterner et al., 2008) as well as for waters of the Arctic Ocean and pan-Arctic shelves (Frigstad et al., 2014). These recent reviews propose that the classical Redfield model be reassessed to better reflect regionally meaningful relationships for POC:PON ratios. Improved parameterization of seston POC:PON ratios is therefore required to enhance the modelling of carbon cycling including autotrophic production estimates, air-sea CO2 flux and carbon sequestration.
Similar to water column modelling, sea-ice models have also relied on POC:PON ratios for the parameterization of ice-associated biogeochemical cycles (Lavoie et al., 2009; Deal et al., 2011; Tedesco et al., 2012). However, local and regional descriptions of sea-ice stoichiometry are largely lacking for POC, PON, dissolved organic carbon and dissolved nitrogen components. Ice-associated autotrophic production contributes seasonally significant amounts of organic biomass during the bloom period (Riedel et al., 2008; Gradinger, 2009; Różańska et al., 2009), and directly contributes to organic carbon fluxes impacting biochemical processes in both the water column (Michel et al., 2002; Pineault et al., 2013; Yamamoto et al., 2014) and sediments (Renaud et al., 2007). The objective of this paper is to provide an overview of POC:PON ratios in first-year sea ice on Canadian Arctic shelves, investigating both seasonal and regional variability. Data coverage includes the productive spring period as well as the dark winter period. Regional comparisons are made between first-year sea ice on the Beaufort Sea shelf (BSS) and first-year sea ice from a productive region of the Canadian Arctic Archipelago (CAA). We hypothesized that sea-ice POC:PON ratios will decrease with increasing sea-ice biomass, as reported for seston stoichiometric ratios in temperate and Arctic waters (Sterner et al., 2008; Frigstad et al., 2014).
First-year sea ice was sampled in two distinct areas of the Canadian Arctic between 2001 and 2013. In the Canadian Arctic Archipelago (CAA), spring to early summer sampling was conducted around Cornwallis Island during first-year landfast seasonal studies (2001–2003) based from ice camps in southern McDougall Sound. Spatial sea-ice studies were conducted between 2010 and 2013 with 78 stations, primarily pack ice, sampled in Barrow Strait, Resolute Passage, McDougall Sound and Wellington Channel. On the interior Beaufort Sea shelf (BSS) a seasonal sea-ice study (Canadian Arctic Shelf Exchange Study, CASES) was conducted in Franklin Bay between February and June 2004. A spatial survey of winter sea ice was also conducted in the Amundsen Gulf as part of the Circumpolar Flaw Lead (CFL) system study in January 2008. The locations of the sampling stations are shown in Figure 1 and the sampling programs are summarized in Table 1.
|Arctic Shelf||Season||Sampling dates||Study year(s)||Study type||Ice type||n||POC:PON (mol:mol)|
|CAA||Spring-Summer||29 April–1 July||2001–2003||Seasonal||Landfast||100||8.3||4.8–15|
|CAA||Spring||1–18 May||2010–2013||Spatial||Landfast, pack ice||78||9.8||6.4–17|
|BSS||Spring||24 Feb–20 June||2004||Seasonal||Landfast||44||7.8||3.0–24|
|BSS||Winter||7–27 January||2008||Spatial||Pack ice||35||23||12–46|
In total, 257 samples of bottom sea ice were analyzed. The bottom 3-cm section of the ice core was routinely retained, targeting the expected majority of the biomass in first-year sea ice of the BSS and CAA. At each station basic sea-ice thickness and snow depth measurements were taken. All sea-ice samples for the determination of chlorophyll a (chl a), particulate organic carbon (POC) and nitrogen (PON) were collected with a manual ice corer (Mark II coring system, 9-cm internal diameter, Kovacs Enterprise). On each sampling day, three to five bottom core sections were collected, pooled together, and melted in the dark with the addition of filtered (0.2 µm) surface water collected at the time of sampling (Garrison and Buck, 1986). Particulate concentrations were corrected for dilution by the added surface water.
Chlorophyll a subsamples were filtered onto 25-mm Whatman GF/F filters and extracted for 24 h in 90% acetone at 4°C in the dark, for fluorometric determination (10AU Turner Designs), according to Parsons et al. (1984). POC and PON were analyzed from the same pre-combusted 21-mm GF/F filter, without prior acidification. Duplicate filters were either dried at 60°C and stored in sealed sterile Petri dishes at room temperature or frozen at −80°C until analysis on a Perkin Elmer elemental analyzer. Blank filters were also stored, as above, on each sampling day; average blank values were subtracted from measured sample values. A minimum weight of 20 µg of N per filter was targeted to ensure POC and PON values were well above detection limits. POC:PON is presented as molar ratios unless otherwise stated.
Inorganic carbon contributions were likely negligible in the study areas, as there was no evidence of shells or terrestrial material that would have contributed particulate inorganic carbon to the sea ice (e.g., the Prymnesiophycaea spp. reported in Riedel et al., 2003, and Różańska et al., 2009, were not coccolithophores), and the methods used here are inadequate to retain ikaite crystals (Dieckmann et al., 2010). Inorganic nitrogen contributions were also likely negligible, as the presence of nitrogen bound to inorganic particles (e.g., ammonium-bound clay minerals) or nitrogen adsorbed to particles has not been documented in Arctic sea ice to our knowledge. To test these assumptions, additional POC/PON samples were acidified during the 2004 BSS and 2011 CAA sampling campaigns (n = 33). The acidified POC and PON samples made up, on average, 93% (± 12%, range of 59–118%) and 98% (± 20%, range of 60–161%) of the values of non-acidified samples, respectively (for internal consistency, only the non-acidified values have been included in this study). Therefore the POC and PON values presented herein are considered reasonable estimates of the particulate organic fraction within the sea ice of both shelves.
To test the hypothesis that POC:PON ratios change with varying concentrations of sea-ice biomass, Model II (standardized major axis [SMA], 95% confidence intervals) linear regressions of log-transformed data were estimated with R statistical software. Model II linear regressions (Sokal and Rohlf, 1995) were used, as both POC and PON are affected by biological sources of stochasticity and analytical error. A linear relationship between POC and PON assumes that the POC:PON ratio remains constant over the range of POC and PON measurements (e.g., POC = 6.6*PON for Redfield). This assumption was tested by determining if the slopes of the Model II regressions were significantly different than 1 in log-log space (T-test, α = 0.05).
If POC:PON ratios are found to vary, a power function relationship following the arguments of Sterner et al. (2008) and Frigstad et al. (2014), could be used to describe the POC:PON ratios over the range of observed POC and PON concentrations, as in equations 1-3.(1)
POC is the concentration of POC (µmol L−1), PON is the concentration of PON (µmol L−1), a is a proportionality constant with log(a) corresponding to the y-intercept at PON = 1 µmol L−1, and b is the slope of the regression. A slope (b), if significantly different than 1, indicates that the POC:PON ratio varies over the range of measurements. When b < 1, the POC:PON ratio decreases with increasing POC and PON concentrations; when b > 1, the POC:PON ratio increases with increasing POC and PON concentrations.
To identify relationships between POC:PON ratios and selected measured variables (e.g., ice thickness), a standard linear model was used since the significance of the relationship cannot be tested with a SMA model regression. To test for significant differences in measured POC:PON ratios between the CAA and BSS regions, a non-parametric Mann-Whitney test was used on untransformed data. A Wilcoxon Signed Rank test was also used to compare measured POC:PON ratios with the Redfield ratio. Statistical tests were performed with JMP (SAS).
Sea-ice thickness averaged 157 cm (range of 77–214 cm, n = 224) during the spring sampling period and 75 cm (range of 1–112 cm, n = 35) in January during the dark winter. Measured snow depths averaged 11.8 cm (range of 1.1–42 cm, n = 224) and 4.2 cm (range of 0–12 cm, n = 35) in the spring and dark winter, respectively.
Autotrophic sea-ice biomass was consistently concentrated at the ice-water interface of all ice cores. For example, in full cores collected in the CAA in 2011 and 2013 over 97% of the chl a biomass was located in the bottom 3 cm of the entire ice core. Therefore the POC:PON ratios presented herein are representative of the bulk of the autotrophic particulate organic matter within ice cores of the study regions. However, POC:PON ratios should not be considered uniform throughout the ice thickness. The ratios presented here are not readily applicable to upper ice sections of first-year ice cores, nor multi-year sea ice. In addition bottom ice concentrations of particulate material, including chl a and POC, should not be integrated over the entire thickness of the ice as this would grossly overestimate total sea-ice biomass concentrations.
In the productive CAA, chl a concentrations in bottom first-year sea ice were on average 4.4 times higher than concentrations in the BSS during spring (Figure 2). Maximum chl a concentrations occurred during May on both shelves with generally gradual declines to concentrations around or below 100 µg L−1 by the end of June. During the spring period which included the onset of growth and peak biomass (i.e., bloom) of ice algae, chl a concentrations were on average 1022 and 233 µg L−1 in the CAA and BSS, respectively, with maximum values reaching 4560 µg L−1 in the CAA and 870 µg L−1 in the BSS. In January during the dark winter, chl a concentrations averaged 0.08 µg L−1 with a maximum of only 0.19 µg L−1. This study provides a unique dataset covering sea-ice chl a biomass spanning four orders of magnitude.
Average and range of POC:PON ratios for measured POC and PON concentrations are summarized in Table 1 for the CAA and BSS shelves during the spring and dark winter periods. The POC:PON ratios on the BSS shelf during spring displayed higher variability (Figure 3a) and were significantly lower (Mann-Whitney, p < 0.01) than spring ratios for the CAA (Table 1). POC:PON ratios generally increased during the spring period and tended to increase or remain stable in association with the onset of the melt period (Figure 2). Monthly variations in POC:PON ratios were evident for both the CAA and BSS. January POC:PON ratios displayed a large range of variability and were significantly higher (Mann-Whitney, p < 0.01) than spring ratios (Table 1, Figure 3a). Spatial variability in POC:PON ratios was best displayed by the CAA spatial surveys conducted between 2010 and 2013 (Table 1). Figure 4 shows the POC:PON distribution for the CAA region surrounding Cornwallis Island. The lowest ratio (6.1) was observed at the western end of the Barrow Strait sampling area with the nearshore region of Resolute Passage also showing consistently low ratios. High ratios (14–17) were measured at the northern extent of the sampling region in both McDougall Sound and Wellington Channel.
POC:chl a ratios (g:g) showed similar monthly trends to those observed for POC:PON ratios (Figure 3). During the spring period, POC:chl a ratios ranged between 17 and 560 (average 81) and 10 and 870 (average 52) in the BSS and CAA, respectively. The ratios were not significantly different between the BSS and CAA study regions (Mann-Whitney, p = 0.06). There were a few very high POC:chl a ratios (> 1000) at the beginning (e.g., February 24th) of the growth period in the BSS, but similar high ratios were observed at the end of the sampling period (e.g., June 29th) in the CAA (Figure 3a). The very low chl a content of the dark winter sea ice was evident from the extremely high POC:chl a ratios (average 8400, range 1478–23800) in January.
Combining the dark winter and spring periods resulted in a weak but significant seasonal decrease in POC:PON ratios with increasing chl a concentrations (log-transformed, r2 = 0.24, p < 0.01). When considering only the spring period, a weak but significant increase in POC:PON ratios occurred with increasing chl a concentrations (log-transformed, r2 = 0.11, p < 0.01). There was no significant relationship between POC:PON ratios and ice thickness; however, POC:PON ratios did show a weak significant decreasing trend with increasing snow thickness (r2 = 0.12, p < 0.01). When compared to the Redfield ratio, the measured POC:PON ratios were significantly higher for the dark winter period (Wilcoxon, p < 0.01) and for the spring period in the CAA alone and the CAA and BSS measurements combined (Wilcoxon, p < 0.01). The spring sea-ice POC:PON ratios from the BSS were not significantly different than the Redfield ratio (Wilcoxon, p = 0.1; Table 1).
Model II linear regressions between log-transformed POC and PON values for the spring and dark winter are presented in Figure 5. The log-transformed regression lines were log(POC) = 0.634 + 1.12(log[PON]) (r2 = 0.94) and log(POC) = 1.39 + 0.793(log[PON]) (r2 = 0.66) for the spring (CAA and BSS combined) and dark winter (January) data, respectively. The SMA slopes for the spring (CAA and BSS combined and separate) and dark winter data were all significantly different than 1 (p < 0.05). The anti-log of the y-intercept estimates the POC:PON ratio when PON is equal to 1 µmol L−1. A summary of POC:PON ratios from the Model II SMA regression, scaled to average PON concentrations for each study period, is provided in Table 2. The POC:PON ranges in Table 2 were calculated from the 95% confidence intervals of the SMA regressions.
|Arctic Shelf||Season||POC:PON (mol:mol)||CI 2.5%||CI 97.5%||r2|
|CAA + BSS||Spring||9.8||6.3||15||0.94|
Given that the POC:PON ratios changed with varying concentrations of sea-ice biomass (i.e., SMA slopes significantly different than 1, p < 0.05), a power function (equation 1) could be used to describe the relationship between sea-ice POC and PON concentrations. Power function equations derived to estimate POC:PON ratios were POC:PON = 4.31 PON0.12 and POC:PON = 24 PON−0.21 for the spring (CAA and BSS combined) and dark winter (January) periods, respectively. In Figure 6, POC:PON ratios estimated using the power function equations are shown with the range of PON concentrations measured in this study.
First-year sea ice on Canadian Arctic shelves is characterized by highly concentrated biomass at the ice-water interface. Microorganisms and protists begin colonizing the ice at the time of ice formation (Riedel et al., 2007; Różańska et al., 2008) and autotrophic production can commence early in the spring (e.g., in March; Riedel et al., 2006) when light intensities are sufficient (Różańska et al., 2009). With the onset of autotrophic production there is a rapid shift in species composition with heterotrophic groups becoming overshadowed by the biomass of autotrophic species, primarily pennate diatoms during the algal bloom in first-year sea ice of the study regions (Riedel et al., 2008; Różańska et al., 2009; Niemi et al., 2011; Piwosz et al., 2013). The dynamics of growth, decline and loss of the sea-ice community are complex. The availability of light and inorganic nutrients and physical-chemical interactions between the atmosphere, ice and ocean can have community level impacts (Frigstad et al., 2011; Martiny et al., 2013) potentially altering the stoichiometric balance of sea-ice particulate organic matter.
Despite the ecological likelihood of stoichiometric plasticity for carbon and nitrogen ratios, the fixed Redfield ratio (6.6 for C:N; Redfield et al., 1963) continues to be used for parameterization in sea-ice modelling studies (e.g., Lavoie et al., 2009). In this study, average POC:PON ratios for first-year ice on Canadian Arctic shelves during the spring were generally higher than the Redfield ratio (Tables 1 and 2), similar to POC:PON ratios from landfast ice in the Chukchi Sea (8.9 mol:mol; Jin et al., 2006) as well as in Antarctic sea ice (9.4 mol:mol; Fransson et al., 2011). The POC:PON ratios for the first-year sea ice in spring were generally indicative of a growing, non-degraded autotrophic community (Pineault et al., 2013).
The spring POC:PON ratios from the BSS were not significantly different (p = 0.1) than the Redfield ratio whereas the POC:PON ratios from the CAA were significantly higher (p < 0.01), indicating that on the more productive CAA shelf (Michel et al., 2006, 2015), sea-ice particulate organic matter would be more carbon-rich, or slightly nitrogen-depleted, than is assumed based on the classical Redfield proportions. This carbon to nitrogen imbalance has consequences for nitrogen-based models; for example, for estimates of production based on nutrient drawdown. For bottom sea ice in the CAA, using Redfield instead of the average regression POC:PON ratio from the CAA in this study (12 for the CAA; Table 2) would result in an underestimation of organic carbon production by a factor of 1.8. Upward scaling, to represent regional estimates would carry this factor, suggesting that underestimates of production by a factor of two are not unlikely based on POC:PON ratios alone. For the spring BSS and CAA values combined, the relationship between POC and PON follows the Redfield ratio (Figure 5), and the 95% confidence interval (6.3–15; Table 2) does overlap with the Redfield ratio in the lower range of observations, suggesting that the Redfield ratio could provide a reasonable approximation of spring sea-ice POC:PON. However, the combined spring POC:PON ratios were significantly higher than the Redfield ratio (p < 0.01). Incidentally, estimating sea-ice PON concentrations for which the Redfield ratio applies using the SMA spring regression generates PON concentrations between 22.9 and 52.8 µmol L−1 (logPON 1.4–1.7) for a 5% confidence interval on Redfield values, and between 14.6 and 77.8 µmol L−1 (logPON 1.2–1.9) for a 10% confidence interval. These PON concentrations respectively represent only 5.8% and 12% of spring observations (Figure 5), thereby demonstrating the need to use variable POC:PON ratios to provide a realistic representation of the actual range of observations in first-year sea ice. Therefore, the selection of a single ratio to represent sea-ice POC:PON requires careful consideration especially in highly productive areas (e.g., CAA) and throughout an annual cycle that includes a wide range of sea-ice POC and PON concentrations.
Figure 5 clearly shows that the Redfield ratio is not applicable to bottom sea-ice POC:PON during the dark winter period (Figure 5, Table 2). During the dark winter period, bacteria and both autotrophic and heterotrophic protists continue to be incorporated into the growing sea ice (Niemi et al., 2011) and nil or very low light levels preclude photosynthetically-derived autotrophic production. POC is largely decoupled from chl a concentrations during the dark period (Figure 3b) with significant contribution of organic carbon coming from non-pigmented or low-pigmented biomass in a system that is driven by heterotrophic/mixotrophic production. Sea-ice POC:PON ratios during that time decrease with increasing biomass, opposite the spring trend (Figure 5).
POC:PON ratios in the bottom sea ice during spring were similar to average seston POC:PON from temperate (8.3; Sterner et al., 2008) and Arctic (7.4; Frigstad et al., 2014) waters. The variability of spring sea-ice POC:PON in this study (Figure 3a) was also comparable to the range of variability for seston POC:PON ratios in the Arctic Ocean, Chukchi Sea and polynya areas (e.g., the North Water; Frigstad et al., 2014). Future detailed comparisons between sea-ice and under-ice seston POC:PON ratios would be useful for organic carbon and nitrogen flux estimates during the spring and especially the melt period.
Temporal variability in POC:PON ratios was evident at a seasonal (spring; Figure 2) and monthly (January–July; Figure 3a) scale. The general increase in bottom ice POC:PON ratios during the spring period in both the CAA and BSS (Figure 2) is consistent with observations of higher POC:PON ratios in phytoplankton toward the end of the productive period in the North Water polynya (Mei et al., 2005). Seasonal increases in POC:PON ratios have been attributed to carbon overconsumption later in the season when nutrients are drawn down or depleted. Earlier in the spring, ratios can be lower due to the production of nitrogen-rich organic material in the absence of nutrient limitation (Bates et al., 2005; Mei et al., 2005). With trends towards a shorter Arctic ice season and earlier melt, sea-ice biomass may be released prior to the onset of seasonal conditions that elevate sea-ice POC:PON ratios. Consequently, the organic carbon contribution of sea-ice fluxes during melt would be reduced relative to nitrogen, altering carbon budgets for predictive ice-ocean coupled biogeochemical models.
High variability in POC:PON ratios persisted for each month regardless of whether sea-ice biomass was low (February–March) or at its peak (May–June). The range of monthly variability was also very high for POC:chl a (Figure 3b) ratios. The high POC:chl a ratios (> 500) from the BSS sea ice in February and March were likely due to the numerical dominance of non-autotrophic species in the sea-ice community (Riedel et al., 2008; Różańska et al., 2009; Frigstad et al., 2011). In contrast, the high POC:chl a measured towards the end of the season (e.g., > 2000 in June from CAA sea ice) may have been driven by the production of carbon-rich particulate exopolymers that can be retained within the sea ice even when chl a biomass is lost during melt (Riedel et al., 2008; Juhl et al., 2011). These findings suggest that drivers of biochemical ratios in first-year sea ice vary seasonally over an annual cycle.
The dark winter period is clearly distinct with respect to elemental concentrations and ratios in the sea ice. The transition from the dark period to the spring growth period in first-year sea ice represents a drastic stoichiometric shift in POC:PON ratios (Figure 6) occurring with the early onset and increase in irradiance. Dark period sea-ice POC and PON were collected in the Amundsen Gulf where terrigenous organic carbon from Mackenzie River waters could impact stoichiometric ratios in the ice and water. However, during the winter period river water is largely contained near shore, away from the dark winter sampling area, by the formation of the stamukhi ice field (Carmack et al., 2002). The dark winter bottom ice community is comprised of a similar number of taxa as the spring community (Niemi et al., 2011) indicating that the diversity of the sea-ice community is largely established as cells are incorporated into the ice prior to the onset of the ice algal growth period. Therefore, models which attempt to capture baseline conditions for first-year sea-ice communities should reflect the, on average, order of magnitude decline in POC:PON ratios (i.e., from 29 to 9.8; Table 2) that occurs with the transition from the dark to light period.
Spatial variability in POC:PON ratios was evident at local (Figure 4) and regional scales (i.e., BSS versus CAA; Tables 1 and 2). In the productive CAA, POC:PON ratios varied 2.7 fold within the vicinity of Cornwallis Island during a limited time period (early May; Table 1, Figure 4) covering the ice algal bloom (Figure 2; Smith et al., 1993; Michel et al., 2006). At the larger regional scale, POC:PON ratios in first-year sea ice were significantly lower on the BSS as compared to the CAA (Table 2). From an ecological perspective, the higher nitrogenous content per unit of carbon biomass in the BSS sea ice may benefit the growth of herbivorous grazers (Malzahn et al., 2010; Frigstad et al., 2011) that consume ice algae when phytoplankton biomass is low under ice-covered conditions.
The observed regional variability with higher POC:PON ratios in the CAA is consistent with the weak significant positive relationship between bottom ice POC:PON ratios and chl a concentrations during spring. The drivers of spatial variability in seston or sea-ice POC:PON ratios at local or regional scales are not directly evident, and their analysis is beyond the scope of this paper. Chlorophyll a is a proxy that integrates multiple processes that could alter POC:PON ratios over a range of spatial and temporal scales. The observed relationship between sea-ice POC:PON ratios and chl a concentrations implies that several factors could be impacting the POC:PON ratios, including bloom dynamics or stages of algal growth, cell viability and physiology, nutrient availability and light acclimation responses (Frigstad et al., 2011; Ayata et al., 2014; Talmy et al., 2014). Snow cover was also weakly negatively related to first-year ice POC:PON ratios indicating that factors altering irradiance at the bottom of the ice as well as ice growth/melt processes (Campbell et al., 2015) may also influence stoichiometric relationships within the bottom ice. Detailed analyses of sea-ice and surface water variables and further process studies are required to identify which variable or combination of variables determine stoichiometric relationships within first-year sea ice.
Seston POC:PON ratios are reported to decrease with increasing seston concentrations (Sterner et al., 2008; Frigstad et al., 2014). Similar trends have also been observed for exported organic material (Tamelander et al., 2013). For the linear relationships between POC and PON (Figure 5) a log-log SMA slope < 1 implies that POC:PON ratios decrease with increasing concentrations of POC and PON. Based on the previous observations of wide-scale trends for water column POC:PON, we originally hypothesized that sea-ice POC:PON ratios would also decrease with increasing sea-ice biomass. This hypothesis could not be validated for first-year sea ice during spring, as the slopes were significantly higher than 1 and POC:PON ratios significantly increased with increasing PON concentrations (Figure 6). Both temporal (Figure 2) and spatial (Figure 5, Table 2) increases in POC:PON ratios with increasing biomass in spring contribute to the rejection of the working hypothesis for the productive period. The hypothesized relationship between POC:PON ratios and sea-ice biomass was only upheld during the dark winter period when POC:PON ratios decreased with increasing PON concentrations (Figure 6), or when considering the full seasonal range from the dark winter to spring.
The decrease in POC:PON ratios at increasing seston concentrations has been linked to shifting nutrient use efficiencies (e.g., physiological status of autotrophs; Sterner and Elser, 2002) as well as shift in seston composition (Frigstad et al., 2011). Autotrophic material, with a generally lower POC:PON than non-autotrophic protists (Frigstad et al., 2011), would drive the seston POC:PON ratio downwards as chlorophyll biomass increases and autotrophic material comprises a greater proportion of the seston. A similar increase in autotrophic material, with respect to both biomass and abundance, occurs in bottom first-year sea ice during the growth period (Michel et al., 2003; Riedel et al., 2008), yet the sea-ice POC:PON ratios do not appear to decline accordingly (Figures 2 and 3a). Therefore, the proportional composition of the sea-ice community and the presence of key ice-algal species (i.e., diatoms) do not appear to be the strongest drivers of biomass-ratio stoichiometric relationships within the bottom ice. Complex biological (e.g., shift in species composition, as in Michel et al., 1996; Riedel et al., 2008) and physiological (e.g., pigment composition, as in Mundy et al., 2011; protein/lipid cellular allocation, as in Cota and Smith, 1991; Mock and Gradinger, 2000) interactions, together with environmental forcings that influence nutrient availability and uptake (e.g., nitrification, as in Fripiat et al., 2014; nutrient status, as in Granskog et al., 2003) are to be considered to explain POC:PON ratios in multifaceted sea-ice communities. In addition, the relative importance of environmental forcings for POC:PON ratios will likely shift under conditions of Arctic change, thereby impacting predictive modelling of sea-ice biogeochemical cycling. For example, the relative influence of nutrient availability on bottom ice POC:PON ratios may shift from external (e.g., nutrient concentrations at the ice-water interface) to internal (e.g., sea-ice remineralisation processes) drivers as sea-ice melt and other environmental forcings (e.g., stratification, Arrigo, 2013; Tremblay et al., 2015) alter nutrient cycling in Arctic surface waters.
Bottom sea-ice differs from the pelagic environment in that it contains a highly concentrated layer of biomass, similar to that of a biofilm, rich in carbon and abundant in extracellular organic carbon (Krembs et al., 2002; Meiners et al., 2003; Riedel et al., 2006; Juhl et al., 2011). Sea-ice exopolymer carbon production is closely linked to ice algae (Underwood et al., 2013) such that the drivers favoring both increased (i.e., carbon-rich polymers) and decreased (i.e., autotrophic material) POC:PON ratios could offset expected biomass-ratio trends. Physiological changes in cell status and changes in community composition are likely to be as important in affecting POC:PON ratios within sea ice as for the water column. However, the distinct habitat and growth conditions in sea ice (e.g., low temperatures, self-shading, melt processes) need to be considered when interpreting POC:PON ratio trends.
Using the Redfield ratio for POC:PON conversion would provide reasonable estimates only over a limited range of sea-ice biomass concentrations observed in this study. Therefore, our results argue in favour of using variable POC:PON stoichiometry in sea-ice biogeochemical models, supported by the wide range of biomass concentrations in first-year sea ice and the evidence of variability in sea-ice POC:PON ratios at the regional (CAA), shelf (CAA versus BSS), and seasonal (dark winter versus spring) scales. Incorporating this variability into analytical and predictive modelling efforts is essential to achieve a better understanding of the role of first-year sea ice in regional food webs and global biogeochemical cycles. The use of the power function model presented here is recommended, as it reflects that POC:PON ratios do not remain constant over the observed sea-ice POC and PON concentrations. For sea-ice biogeochemical modellers, recommendations include: 1) parameterization using variable POC:PON ratios rather than consistent Redfield or average values, 2) the inclusion of distinct power functions to parameterize dark winter versus spring POC:PON ratios and, 3) Arctic-wide or regionally-based models based on areal (i.e., m2) estimates should not apply bottom ice stoichiometry to the upper sections of first-year sea ice.
The distinct dynamics of POC:PON ratios during the dark winter and spring periods (Figure 6) clearly show that spring parameters should not be extrapolated to winter, for annual budget estimates or simulations. Surprisingly, our original hypothesis that POC:PON ratios would decrease with increasing biomass, as observed in Arctic waters (Frigstad et al., 2014), was valid only during the dark winter period. Further to these findings, we hypothesize that the relationship between POC:PON ratios and bottom sea-ice biomass during the dark winter period is applicable to stoichiometric relationships within upper ice sections, including in multi-year sea ice, and possibly in first-year sea ice replacing multi-year ice in new areas of the high Arctic where ice biomass may remain low due to light limitation.
All data will be publically available from BioChem: database of biological and chemical oceanographic data. Fisheries and Oceans, Canada. http://www.meds-sdmm.dfo-mpo.gc.ca/biochem/biochem-eng.htm.
© 2015 Niemi and Michel. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Contributed to conception and design: AN, CM
Contributed to acquisition of data: AN, CM
Contributed to analysis and interpretation of data: AN, CM
Drafted and/or revised the article: AN, CM
Approved the submitted version for publication: AN, CM
The authors do not have any competing interests.
NSERC (Natural Sciences and Engineering Research Council of Canada) Research Network Grant, and Discovery Grants to CMDFO Strategic Research Fund, International Governance Strategy and NCAARE (National Centre for Arctic Aquatic Research Excellence) to CM Canada Foundation for Innovation and Fonds québécois de la recherche sur la nature et les technologies (FQRNT) OERD (Office of Energy Research and Development), NRCan (Natural Resources Canada) to CM PCSP (Polar Continental Shelf Program) logistical and in-kind support NSTP (Northern Science Training Program).
We sincerely thank the officers and crew of the CCGS Amundsen for their support of the Beaufort Sea studies and the Polar Continental Shelf Program (PCSP) for many years of excellent logistical support in the Canadian Arctic Archipelago. We also thank the Resolute Bay Hunters and Trappers Association for their support of the scientific research. We gratefully acknowledge the many chief scientists and expedition coordinators of the CASES (Canadian Arctic Shelf Exchange Study) and CFL (Circumpolar Flaw Lead System study) programs as well as numerous collaborators and team members who assisted with field sampling and laboratory support over the different years of study. Specifically, we thank the staff of the Institut des Sciences de la Mer de Rimouski (ISMER, Université du Québec à Rimouski) and the Stable Isotopes in Nature Laboratory (SINLAB, University of New Brunswick) who analyzed the POC and PON samples. We also thank J. Deming and two anonymous reviewers for insightful suggestions that helped improve the manuscript.
Arrigo KR. 2005. Marine microorganisms and global nutrient cycles. Nature 437: 349–355. doi: 10.1038/nature04159.
Arrigo KR. 2013. The changing Arctic Ocean. Elem Sci Anth 1: 000010. doi: 10.12952/journal.elementa.000010.
Ayata SD , Levy M , Aumont O , Resplandy L , Tagliabue A , et al. 2014. Phytoplankton plasticity drives large variability in carbon fixation efficiency. Geophys Res Lett 41(24): 8994–9000. doi: 10.1002/2014gl062249.
Bates NR, Hansell DA, Moran SB, Codispoti LA. 2005. Seasonal and spatial distribution of particulate organic matter (POM) in the Chukchi and Beaufort Seas. Deep-Sea Res Pt II 52(24–26): 3324–3343. doi: 10.1016/j.dsr2.2005.10.003.
Campbell K, Mundy CJ, Barber D, Gosselin M. 2015. Characterizing the sea ice algae chlorophyll a–snow depth relationship over Arctic spring melt using transmitted irradiance. J Mar Syst 147: 76–84. doi: 10.1016/j.jmarsys.2014.01.008.
Carmack EC, MacDonald RW. 2002. Oceanography of the Canadian shelf of the Beaufort Sea: A setting for marine life. Arctic 55: 29–45. doi: 10.14430/arctic733.
Cota GF, Smith REH. 1991. Ecology of bottom ice algae. III. Comparative physiology. J Mar Syst 2: 297–315. doi: 10.1016/0924-7963(91)90038-V.
Deal C , Jin MB , Elliott S , Hunke E , Maltrud M , et al. 2011. Large-scale modeling of primary production and ice algal biomass within arctic sea ice in 1992. J Geophys Res 116: C07004. doi: 10.1029/2010jc006409.
Dieckmann GS , Nehrke G , Uhlig C , Gottlicher J , Gerland S , et al. 2010. Ikaite (CaCO3·6H2O) discovered in Arctic sea ice. Cryosphere 4(2): 227–230. doi: 10.5194/tc-4-227-2010.
Fransson A, Chierici M, Yager PL, Smith WO. 2011. Antarctic sea ice carbon dioxide system and controls. J Geophys Res 116: C12035. doi: 10.1029/2010jc006844.
Frigstad H, Andersen T, Bellerby RGJ, Silyakova A, Hessen DO. 2014. Variation in the seston C:N ratio of the Arctic Ocean and pan-Arctic shelves. J Mar Syst 129: 214–223. doi: 10.1016/j.jmarsys.2013.06.004.
Frigstad H , Andersen T , Hessen DO , Naustvoll LJ , Johnsen TM , et al. 2011. Seasonal variation in marine C:N:P stoichiometry: Can the composition of seston explain stable Redfield ratios? Biogeosciences 8(10): 2917–2933. doi: 10.5194/bg-8-2917-2011.
Fripiat F , Sigman DM , Fawcett SE , Rafter PA , Weigand MA , et al. 2014. New insights into sea ice nitrogen biogeochemical dynamics from the nitrogen isotopes. Global Biogeochem Cy 28(2): 115–130. doi: 10.1002/2013GB004729.
Garrison DL, Buck KR. 1986. Organism losses during ice melting – a serious bias in sea ice community studies. Polar Biol 6(4): 237–239. doi: 10.1007/bf00443401.
Geider RJ, La Roche J. 2002. Redfield revisited: variability of C:N:P in marine microalgae and its biochemical basis. Eur J Phycol 37(1): 1–17. doi: 10.1017/s0967026201003456.
Gradinger R. 2009. Sea-ice algae: Major contributors to primary production and algal biomass in the Chukchi and Beaufort Seas during May/June 2002. Deep-Sea Res Pt II 56(17): 1201–1212. doi: 10.1016/j.dsr2.2008.10.016.
Granskog MA, Kaartokallio H, Shirasawa K. 2003. Nutrient status of Baltic Sea ice: Evidence for control by snow-ice formation, ice permeability, and ice algae. J Geophys Res 108(C8): 3253. doi: 10.1029/2002jc001386.
Jin MB , Deal C , Wang J , Shin K-H , Tanaka N , et al. 2006. Controls of the landfast ice-ocean ecosystem offshore Barrow, Alaska. Ann Glaciol 44: 63–72. doi: 10.3189/172756406781811709.
Juhl AR, Krembs C, Meiners KM. 2011. Seasonal development and differential retention of ice algae and other organic fractions in first-year Arctic sea ice. Mar Ecol-Prog Ser 436: 1–16. doi: 10.3354/meps09277.
Krembs C, Deming JW, Junge K, Eicken H. 2002. High concentrations of exopolymeric substances in wintertime sea ice: Implications for the polar ocean carbon cycle and cryoprotection of diatoms. Deep-Sea Res Pt I 49: 2163–2181.
Lavoie D, Macdonald RW, Denman KL. 2009. Primary productivity and export fluxes on the Canadian shelf of the Beaufort Sea: A modelling study. J Mar Syst 75(1–2): 17–32. doi: 10.1016/j.jmarsys.2008.07.007.
Letscher RT, Moore JK. 2015. Preferential remineralization of dissolved organic phosphorus and non-Redfield DOM dynamics in the global ocean: Impacts on marine productivity, nitrogen fixation, and carbon export. Global Biogeochem Cy 29(3): 325–340. doi: 10.1002/2014gb004904.
Leu E, Søreide JE, Hessen DO, Falk-Petersen S, Berge J. 2011. Consequences of changing sea-ice cover for primary and secondary producers in the European Arctic shelf seas: Timing, quantity, and quality. Prog Oceanogr 90 (1–4): 18–32. doi: 10.1016/j.pocean.2011.02.004.
Malzahn AM, Hantzsche F, Schoo KL, Boersma M, Aberle N. 2010. Differential effects of nutrient-limited primary production on primary, secondary or tertiary consumers. Oecologia 162(1): 35–48. doi: 10.1007/s00442-009-1458-y.
Martiny AC, Vrugt JA, Primeau FW, Lomas MW. 2013. Regional variation in the particulate organic carbon to nitrogen ratio in the surface ocean. Global Biogeochem Cy 27(3): 723–731. doi: 10.1002/gbc.20061.
Mei ZP , Legendre L , Tremblay JE , Miller LA , Gratton Y , et al. 2005. Carbon to nitrogen (C:N) stoichiometry of the spring-summer phytoplankton bloom in the North Water Polynya (NOW). Deep-Sea Res Pt I 52(12): 2301–2314. doi: 10.1016/j.dsr.2005.07.001.
Meiners K, Gradinger R, Fehling J, Civitarese G, Spindler M. 2003. Vertical distribution of exopolymer particles in sea ice of the Fram Strait (Arctic) during autumn. Mar Ecol-Prog Ser 248: 1–13. doi: 10.3354/meps248001.
Michel C , Hamilton J , Hansen E , Barber D , Reigstad M , et al. 2015. Arctic Ocean outflow shelves in the changing Arctic: A review and perspectives. Prog Oceanogr . doi: 10.1016/j.pocean.2015.08.007.
Michel C, Ingram RG, Harris LR. 2006. Variability in oceanographic and ecological processes in the Canadian Arctic Archipelago. Prog Oceanogr 71(2–4): 379–401. doi: 10.1016/j.pocean.2006.09.006.
Michel C, Legendre L, Ingram RG, Gosselin M, Levasseur M. 1996. Carbon budget of sea-ice algae in spring: Evidence of a significant transfer to zooplankton grazers. J Geophys Res 101(C8): 18345–18360. doi: 10.1029/96jc00045.
Michel C, Nielsen TG, Nozais C, Gosselin M. 2002. Significance of sedimentation and grazing by ice micro- and meiofauna for carbon cycling in annual sea ice (northern Baffin Bay). Aquat Microb Ecol 30(1): 57–68. doi: 10.3354/ame030057.
Michel C, Riedel A, Mundy CJ. 2003. Biological Investigation of First-Year Sea Ice near Resolute Bay, Nunavut, Spring to Early Summer 2001. Can Data Rep Hydrogr Ocean Sci 160: vi + 29 p.
Mills MM, Arrigo KR. 2010. Magnitude of oceanic nitrogen fixation influenced by the nutrient uptake ratio of phytoplankton. Nat Geosci 3: 412–416. doi: 10.1038/ngeo856.
Mock T, Gradinger R. 2000. Changes in photosynthetic carbon allocation in algal assemblages of Arctic sea ice with decreasing nutrient concentrations and irradiance. Mar Ecol Prog Ser 202: 1–11. doi: 10.3354/meps202001.
Mundy CJ , Gosselin M , Ehn JK , Belzile C , Poulin M , et al. 2011. Characteristics of two distinct high-light acclimated algal communities during advanced stages of sea ice melt. Polar Biol 34(12): 1869–1886. doi: 10.1007/s00300-011-0998-x.
Niemi A, Michel C, Hille K, Poulin M. 2011. Protist assemblages in winter sea ice: Setting the stage for the spring ice algal bloom. Polar Biol 34(12): 1803–1817. doi: 10.1007/s00300-011-1059-1.
Parsons TR, Maita Y, Lalli CM. 1984. A Manual of Chemical and Biological Methods for Seawater Analysis . Oxford: Pergamon Press.
Pineault S, Tremblay JE, Gosselin M, Thomas H, Shadwick E. 2013. The isotopic signature of particulate organic C and N in bottom ice: Key influencing factors and applications for tracing the fate of ice-algae in the Arctic Ocean. J Geophys Res 118(1): 1–14. doi: 10.1029/2012jc008331.
Piwosz K, Wiktor JM, Niemi A, Tatarek A, Michel C. 2013. Mesoscale distribution and functional diversity of picoeukaryotes in the first-year sea ice of the Canadian Arctic. ISME J 7(8): 1461–1471. doi: 10.1038/ismej.2013.39.
Redfield AC, Ketchum B, Richards F. 1963. The influence of organisms on the composition of sea-water, in, Hill M, ed., The Sea . New York, US: Interscience.
Renaud PE , Riedel A , Michel C , Morata N , Gosselin M , et al. 2007. Seasonal variation in benthic community oxygen demand: A response to an ice algal bloom in the Beaufort Sea, Canadian Arctic? J Mar Syst 67(1–2): 1–12. doi: 10.1016/j.jmarsys.2006.07.006.
Riedel A, Michel C, Gosselin M. 2006. Seasonal study of sea-ice exopolymeric substances on the Mackenzie shelf: Implications for transport of sea-ice bacteria and algae. Aquat Microb Ecol 45(2): 195–206. doi: 10.3354/ame045195.
Riedel A, Michel C, Gosselin M, LeBlanc B. 2007. Enrichment of nutrients, exopolymeric substances and microorganisms in newly formed sea ice on the Mackenzie shelf. Mar Ecol-Prog Ser 342: 55–67. doi: 10.3354/meps342055.
Riedel A, Michel C, Gosselin M, LeBlanc B. 2008. Winter-spring dynamics in sea-ice carbon cycling in the coastal Arctic Ocean. J Mar Syst 74(3–4): 918–932. doi: 10.1016/j.jmarsys.2008.01.003.
Riedel A, Michel C, Poulin M, Lessard S. 2003. Taxonomy and abundance of microalgae and protists at a first-year sea ice station near Resolute Bay, Nunavut, spring to early summer 2001. Can Data Rep Hydrogr Ocean Sci 159: vi + 53 p.
Różańska M, Gosselin M, Poulin M, Wiktor J, Michel C. 2009. Influence of environmental factors on the development of bottom ice protist communities during the winter-spring transition. Mar Ecol Prog Ser 386: 43–59. doi: 10.3354/meps08092.
Różańska M, Poulin M, Gosselin M. 2008. Protist entrapment in newly formed sea ice in the Coastal Arctic Ocean. J Mar Syst 74(3–4): 887–901. doi: 10.1016/j.jmarsys.2007.11.009.
Sambrotto RN. 2014. Ocean chemistry: Biogeochemical regimes in focus. Nat Geosci 7: 862–863. doi: 10.1038/ngeo2309.
Smith REH, Cavaletto JF, Eadie BJ, Gardner WS. 1993. Growth and lipid composition of high Arctic ice algae during the spring bloom at Resolute, Northwest Territories, Canada. Mar Ecol Prog Ser 97(1): 19–29. doi: 10.3354/meps097019.
Sokal RR, Rohlf FJ. 1995. Biometry: The principles and practice of statistics in biological research , 3rd edn. New York: WH Freeman.
Sterner RW , Andersen T , Elser JJ , Hessen DO , Hood JM , et al. 2008. Scale-dependent carbon : nitrogen : phosphorus seston stoichiometry in marine and freshwaters. Limnol Oceanogr 53(3): 1169–1180. doi: 10.4319/lo.2008.53.3.1169.
Sterner RW, Elser JJ. 2002. Ecological stoichiometry: The biology of elements from molecules to the biosphere . Princeton, New Jersey: Princeton University Press.
Talmy D , Blackford J , Hardman-Mountford NJ , Polimene L , Follows MJ , et al. 2014. Flexible C:N ratio enhances metabolism of large phytoplankton when resource supply is intermittent. Biogeosciences 11(17): 4881–4895. doi: 10.5194/bg-11-4881-2014.
Tamelander T, Reigstad M, Olli K, Slagstad D, Wassmann P. 2013. New production regulates export stoichiometry in the ocean. PLoS One 8(1): e54027. doi: 10.1371/journal.pone.0054027.
Tedesco L, Vichi M, Thomas DN. 2012. Process studies on the ecological coupling between sea ice algae and phytoplankton. Ecol Model 226: 120–138. doi: 10.1016/j.ecolmodel.2011.11.011.
Tremblay J-É , Anderson LG , Matrai P , Coupel P , Bélanger S , et al.. 2015. Global and regional drivers of nutrient supply, primary production and CO2 drawdown in the changing Arctic Ocean. Prog Oceanogr (2015) . doi: 10.1016/j.pocean.2015.08.009.
Underwood GJC , Aslam SN , Michel C , Niemi A , Norman L , et al. 2013. Broad-scale predictability of carbohydrates and exopolymers in Antarctic and Arctic sea ice. Proc Natl Acad Sci USA 110(39): 15734–15739. doi: 10.1073/pnas.1302870110.
Vancoppenolle M , Meiners KM , Michel C , Bopp L , Brabant F , et al. 2013. Role of sea ice in global biogeochemical cycles: emerging views and challenges. Quat Sci Rev 79: 207–230. doi: 10.1016/j.quascirev.2013.04.011.
Yamamoto S , Michel C , Gosselin M , Demers S , Fukuchi M , et al. 2014. Photosynthetic characteristics of sinking microalgae under the sea ice. Polar Sci 8(4): 385–396. doi: 10.1016/j.polar.2014.07.007