Effect of pore water Fe(II) on stream oxygen uptake rates
Our results clearly demonstrate that Fe(II) supply from the subsurface can have a substantial effect on the oxygen budget of streams, particularly within their hyporheic zone, which provides an important habitat for many species 24–27. The results show that the advective supply of Fe(II) reduces pore water and stream DO concentrations relative to the control. The effect is larger in stream beds of lower porosity.
In order to quantify these effects, we have compared the overall oxygen uptake of the streambeds in the control experiments with that of treated cells by use of a steady-state mass balance approach 12 with the oxygen concentration in the controls as the reference state:
$$\frac{d{c(O}_{2})}{dt}={k}_{{O}_{2}} . \left({c\left({O}_{2}\right)}_{sat}-{c\left({O}_{2}\right)}_{flume}\right)+GPP-UR$$
1
With,
c(O2)sat: Calculated oxygen concentration at saturation at the measured flume temperature, which was 11.3 g m− 3 (T = 12°C) as derived from Henry´s law.
c(O2)flume: mean oxygen concentration measured in surface water of the flumes treatments or the control cells [g m− 3],(Table 1)
Table 1: Uptake rates of oxygen in the flume cells with and without Fe(II) addition and the increase in uptake rate induced by the presence of Fe(II).
kO2: Oxygen transfer coefficient [d− 1]
GPP: gross primary production rate [g O2 m− 3 d− 1]
UR: Flume O2 uptake rate [g O2 m− 3 d− 1]
The GPP rate at the time of the experiment (November) can be assumed to be negligible and was set to zero. With this the flume-cell O2 uptake rate accounts for all processes that consume dissolved O2, i.e., respiration but also oxidation of Fe(II). The oxygen transfer coefficient kO2 (d− 1) accounts for the rate at which O2 is transferred between the air and surface water, and which largely depends on mean flume-flow velocity and mean flume depth. Values for kO2 were calculated by use of two empirical equations formulated from re-aeration experiments by O’connor and Dobbins (1958, Eq. (2) and Negulescu and Rojanski (1969, Eq. (3))28,29
$${k}_{1,O2}=3.904\frac{{\stackrel{-}{u}}^{0.5}}{{\stackrel{-}{d}}^{1.5}}$$
2
and
$${k}_{2,O2}=10.92{\left(\frac{\stackrel{-}{v}}{\stackrel{-}{d}}\right)}^{0.85}$$
3
with
\(\stackrel{-}{u}\) : mean flow velocity, 5200 m d− 1, and 870 m d− 1, respectively, for the moderate and low flume flow rate
\(\stackrel{-}{d}\) : mean flume depth, 0.1 m)
These two expressions provide a range of oxygen transfer coefficients that can be used to further compare oxygen uptake rates with other reports on ecosystem respiration. Assuming steady-state conditions, estimates for area normalized molar oxygen uptake rates UR* in the control and treated flume cells can be calculated as:
$${UR}^{*}={k}_{i,O2} \bullet \stackrel{-}{d}\bullet \left(c{\left({O}_{2}\right)}_{sat}- c{\left({O}_{2}\right)}_{flume}\right)$$
4
with units of mg O2 m− 2 d− 1.
Uptake rates in the control flumes ranged, accounting for both sediment types, between ~ 1.65 g O2 m− 2 d− 1 (Eq. 3) and ~ 7.35 g O2 m− 2 d− 1 (Eq. 2.) for the mean flow rate and ~ 0.36 g O2 m− 2 d− 1 and ~ 3.0 g O2 m− 2 d− 1 respectively, for the low flow rate (Table 1) which is in the order of rates reported for natural freshwater streams 12,30–32. In the treated flumes, the uptake rates increased to values between ~ 1.93 g O2 m− 2 d− 1 (Eq. (3)) and 8.26 g O2 m− 2 d− 1 (Eq. (2)) at mean flow rate and between 0.44 g O2 m− 2 d− 1, and 3.62 g O2 m− 2 d− 1 at low flow rate, respectively. These values correspond to mean increases in uptake rates by between 12 and 17 % in th mean-flow flume cells and 21 % in th low-flow flume cells implying a substantial effect of subsurface Fe(II) input on the overall DO budget in the flumes.
Sensitivity of Stream Ecosystem respiration to Fe(II) fluxes
Fe concentrations in stream waters appear to be on the rise. In a survey on temporal trends of Fe concentrations across 340 water bodies in northern Europe and North America20 demonstrated that Fe concentrations have significantly increased in 28% of monitored sites, and decreased in 4%, with the most positive trends located in northern Europe. They conclude that the phenomenon of increasing Fe concentrations is widespread, especially in northern Europe, with potentially significant implications for wider ecosystem biogeochemistry.
In their study, Björnerås et al. (2017)20 was not able to distinguish between Fe(II) and Fe(III). However, there is evidence that a substantial fraction of the input of iron, particularly from reducing peat riparian soils, is Fe(II) 33–36. It has been suggested that rising temperatures may even stimulate the mobilization of Fe35 due to an increase in reducing conditions. Input of Fe(II) through drains and ditches in agricultural soils has been observed in wide areas of central European lowlands 37,38. In these studies, the focus was put on the coupling between Fe(II) oxidation, formation of ferric colloids, and the retention of phosphate 39–41. Surprisingly, the role of Fe(II) oxidation for the oxygen budget in the streams has not been addressed by now.
In their survey, Björnerås et al., (2017)20 emphasized that regions with rising Fe concentrations tend to coincide with those with increases in organic carbon (OC) concentration. Increasing Fe concentrations were paralleled by increasing OC contents in 77% of the waters studied. Organic carbon is well known to stabilize Fe(II) against oxidation via complexation and hinders subsequent precipitation as ferric hydroxides42–44 which may also explain the coupled increase both in DOC and in Fe(II) concentrations observed in the flume surface waters of our experiments.
Collectively, our results indicate a strong coupling between stream-water DO concentrations, oxygen uptake-rates, and input of Fe(II), which needs to be accounted for in future freshwater ecological studies. Against the background of rising groundwater contribution to discharge during drought affected low-flow conditions in headwater streams 45these systems may be particularly sensitive to Fe(II) related oxygen uptake.