Alkalinity and nitrate concentrations in calcareous watersheds : Are they linked , and is there an upper limit to alkalinity ?

Data from aquifers in calcareous watersheds in Switzerland demonstrate that alkalinity initially increases approximately in proportion to nitrate (NO3) concentration in the groundwater and eventually approaches an apparent maximum of approximately 8 mmol L at high NO3 concentrations. This close positive relationship between alkalinity and 10 NO3 concentration appears to be predominantly a result of three processes: (i) mineralization of organic nitrogen (N) fertilizer, (ii) exchange of OH and H during the uptake of NO3 or ammonium (NH4), and (iii) CO2 released by roots as a result of fertilizer-stimulated plant growth. Atmospheric deposition of N and strong acids (H2SO4 and HNO3) play a minor role. We suggest that the asymptotic approach to a maximum groundwater alkalinity at NO3 concentrations exceeding 0.25 mmol L may be caused by (i) a maximum possible areal crop production at excessive N fertilization and (ii) an increasing 15 CO2 loss to the atmosphere due to the increasing CO2 production in the soil. Thus, we estimate that the fertilizer-intensive agriculture of Switzerland generates an annual flux from the soil to the atmosphere of at least 0.26 Mt CO2 a. This analysis provides a general understanding and quantitative prediction of steady-state groundwater NO3 concentration; equilibrium groundwater alkalinity, pH, and pCO2; and soil CO2 emissions to the atmosphere based on quantitative and qualitative information on the supply of N and acidity to the soil by atmospheric deposition and N fertilization. The positive correlation 20 between alkalinity and NO3 concentration in groundwaters persists in rivers and lakes. However, due to the diffusive loss of CO2 to the atmosphere, subsequent precipitation of calcite, dilution with surface water, input of wastewater discharges and NO3 consumption by aquatic photoautotrophs, the correlation is less distinct.

3 ] up to 0.25 mmol L −1 alkalinity increases linearly with nitrate, for higher concentrations is levels off never exceeding 8 mmol L −1 . The authors try to explain these variations and the existence of a maximum alkalinity concentration (at 8 mmol L −1 ) by various processes.
In a paper where alkalinity is a central concept (and the first word in the title) I would expect a definition or at least a reference to the definition (I suggest citing Dickson, 1981, who gave the most precise definition) and a description or reference how alkalinity (better total alkalinity, TA) was measured or estimated (for example, Dickson et al., 2007).
Based on the TA definition by Dickson (1981) (the other terms are probably small or roughly cancel each other). This expression for TA shows that addition of nitrate would decrease TA. The observed increase of TA with increasing nitrate is, therefore, not a direct effect of nitrate addition (concentration too small and wrong sign in the TA expression), but is rather a proxy for other processes, namely CaCO 3 dissolution (enhanced weathering) caused by agriculture. Although TA varies almost linearly with nitrate concentration for nitrate concentrations up to 0.25 mmol L −1 , the relation become nonlinear (levels off, saturates) for higher concentrations. This also speaks against a direct impact of nitrate, but suggests that nitrate could be a proxy for other processes.  The authors have compiled a lot of data that are somewhat hidden in various archives. It would be great if these data could be made publicly/more easily available at, for example, the Carbon Dioxide Information Analysis Center, https://cdiac.ess-dive.lbl.gov.
Minor points: p.2 L9-12: "Therefore, in calcareous soils, in-soil production of CO 2 (e.g., due to root respiration and heterotrophic mineralization of organic matter) ... result in an increased alkalinity concentration. In contrast, in the absence of carbonate minerals, alkalinity is expected to decrease in proportion to the amount ... in-soil CO 2 production (Perrin et al., 2008)." This can be misleading or is wrong. Addition of CO 2 does not change total alkalinity. However, addition of CO 2 will decrease pH and may lead to dissolution of CaCO 3 resulting in the increase of total alkalinity. p.2 L38-39: I suggest changing '10 −3.5 to 10 −3.4 atm' to '316 to 398 µatm' (or after 'rounding': 300 to 400 µatm) p.2 L39-41 "... in the absence of acids other than H 2 CO 3 , an alkalinity concentration of 1.42 mmol L −1 and a pH of 8.24 would be expected in water draining a hypothetically N-free (and thus sterile) calcareous soil, assuming a groundwater temperature of 8 • C." Which assumptions have been made? (I guess Ω calcite = 1, equilibrium of CO 2 partial pressures, alkalinity ≈ [HCO 3 ] + 2 [CO 2+ 3 ] ≈ Ca) Can you give a reference? Which equilibrium constants did you use? p.3 L12-13: "We calculated the CO 2 saturation index of water as Ω CO 2 = CO 2 (aq)/CO 2 (atm), where CO 2 (aq) is the partial pressure of CO 2 in the water (in atm) and CO 2 (atm) is the partial pressure of CO 2 in the atmosphere ..." I suggest to use the notation pCO 2 for partial pressures. p. 6: "As pCO 2 increases and as pH concurrently decreases, the extent to which H 2 CO 3 dissociates into HCO − 3 and CO 2− 3 decreases (because increasingly greater percentages of the DIC remain as H 2 CO 3 as the H + concentration increases)." This could be quantified and illustrated by a Bjerrum plot.

Supplement:
S-4: "p" is the probability value of the slope or intercept being equal to 0 (i.e., p < 0.05 indicates significant difference from zero) No! p is the probability to obtain an estimateβ or more extreme values for the slope, β, under the null hypothesis H 0 : 'slope β is zero'. The null hypothesis is rejected if p < α where α is the level of significance (commonly chosen as α = 0.05). Same for the intercept.