Irrigation return flow causing a nitrate hot spot and denitrification imprints in groundwater at Tinwald, New Zealand 225

This paper looks at the impact of irrigation return waters in areas of intensive agriculture on nitrate concentrations and considers more broadly how we track the sources of nitrate via dual isotopes. The study itself is scientiﬁcally sound although I struggled in places with the text. Some of the sections are not very clearly written and the structure could be improved. introduction Abstract. Nitrate concentrations in groundwater have been historically high (N ≥ 11.3 mg/L) in an area 230 surrounding Tinwald, Ashburton since at least the mid-1980s. The local community are interested in methods to remediate the high nitrate in groundwater. To do this they need to know where the nitrate is coming from. Tinwald groundwater exhibits two features stemming from irrigation with local groundwater (i.e. irrigation return flow). The first feature is increased concentrations of nitrate (and other chemicals and stable isotopes) in a ‘hot spot’ around Tinwald. The chemical concentrations of the groundwater are increased by recirculation of water already 235 relatively high in chemicals. The irrigation return flow coefficient C (irrigation return flow/irrigation flow) is found to be consistent with the chemical enrichments. The stable isotopes of the groundwater show a similar pattern of enrichment by irrigation return flow of up to 40% and are also enriched by evaporation (causing loss of about 205% of the original water mass). Management implications are that irrigation return flow needs to be taken into account in modelling of nitrate transport through soil/groundwater systems and in avoiding overuse of nitrate 240 fertilizer leading to greater leaching of nitrate to the groundwater and unnecessary economic cost. The second feature is the presence of ‘denitrification imprints’ (shown by enrichment of the δ 15 N and δ 18 O NO3 values of nitrate) in even relatively oxic groundwaters. The denitrification imprints can be clearly seen because (apart from denitrification) the nitrate has a blended isotopic composition due to irrigation return flow and N being retained in the soil-plant system as organic-N. The nitrate concentration and isotopic compositions of nitrate are found to 245 be correlated with dissolved oxygen concentration. This denitrification imprint is attributed to localised denitrification in fine pores or small-scale physical heterogeneity where conditions are reducing. The implication is that denitrification could be occurring where it is not expected because groundwater DO concentrations are not low.

Tinwald, Ashburton since at least the mid-1980s. The local community are interested in methods to remediate the high nitrate in groundwater. To do this they need to know where the nitrate is coming from." R2: Introduction The introduction sets the scene for the study. For a paper in an international journal such as HESS, it would be appropriate to add a few comments about how New Zealand compares to other intensiv e agricultural areas globally in terms of the scale of the problem. High nutrient loads are of global interest and this 100 research will have broader interest, so some more comments here are warranted. The structure could be improved as it alternates between general and area-specific statements. Try to group these more. Some of the description of the issues around Tinwald could be in section 2.
Authors: We have referred to OECD reports (2013 and 2017), which place New Zealand's intensive agricultural area management in the context of those of OECD member countries: "Eutrophication causing hypoxia and algal blooms, due primarily to agricultural runoff of excess nutrients, is considered the most prevalent water quality problem globally (OECD, 2017). In New Zealand the N balance worsened (i.e. became more positive) more than in any other OECD member country between 1998 and 2009, almost entirely because of expansion and intensification of farming in New Zealand (OECD, 2013). (The N balance is the difference between N inputs to farming systems (fertiliser and livestock manure) and N outputs R2: Background As mentioned above, some of the detail from the introduction (eg the high nitrate, which is also covered in section 2.4) would be better here. This section is also long for the information it contains and probably it could be written more succinctly.
Authors: We have removed some of the detail from the introduction because it is given here.
R2: Lines 91-95 (and elsewhere). It would be preferable to use "residence times" not "ages". Also, here and throughout suggest referring to the groundwater not the well (the well's age is when it was installed, which is not what you mean).

130
Authors: These changes have been made.
R2: Line 125. MAV not defined (I think that it is the maximum WHO limit for nitrate in drinking water?). Is this the best value to use of should the lower NZ limit be used?
Authors: MAV is now defined in the introduction. We prefer to use MAV, but either limit could be used.
R2: Line 126-136. Adding values to the text would get the message across better (rather than the reader having to 135 find them for their self).
Authors: Values have been added to the text. R2: Methods Section 3.1 could use a few more details. a) Somewhere, the screened intervals should be notes as geochemical data from long screened production wells conveys different information to that from short-screened monitoring wells. b) In Table 1, is the depth the mid-screen? c) Were the wells purged or was sampling done from 140 flowing wells? d) The Comments on groundwater levels and river flows is not very clear.
Authors: a) Screened intervals and mid-screen depths have been added to Table 1. b) The depths given were total depths, but we now use mid-screen depths instead where there were screens, and total depths where there were no screens. c) Field measurements had stabilised before sampling for all wells. 25 wells were purged of at least three well casing volumes before sampling, the 8 remaining wells were sampled by low flow methods (pumps were Authors: A mistake in this calculation has been corrected, Errors in measured quantities (δ) and estimated quantities (δb, h, temperature) do not change the fraction evaporated much when the fraction is small (5%) because of the form of the equation.
R2: Section 4.3. Again, this description is long in places and could be rationalised. The Raleigh equation (line 261) is reported later in the section -it might be better to report how the calculations were done in the methods which would let this section focus on the results (at the very least try to group the descriptions of the calculations and the outcomes more).

175
R2: Section 5.1. Lines 330-364. Do you need all these calculations? It seems though the Cl mass balance together with the isotopic enrichment defines the recharge % well enough (it is the basis of an often -used recharge rate calculation after all). Could you start off with that and then report the results of the water mass balance as support? Also, the infiltration data look to be from a lysimeter, which may be less than total recharge (given that it is probably above the water table). Since you are interested in the chemistry of the recharging water it would be 180 simpler to relate it to a recharge estimate based on the chemistry.
Authors: This has been rewritten.
R2: Section 5.2. The other aspect that is often ignored is that the source has to be there (regardless o f the isotopic composition). The points made on lines 398-404 are correct but there are a fair few studies where the isotopic compositions point to stores that are not locally present. This more general discussion would be better in the 185 conclusions.
Authors: Lines 398-404 are needed here for the logic of this section. A summary of this is already in the Conclusions (Lines 470-474).
R2: Lines 440-450. It is not clear whether these are general points or some things that may be related to your studies. If it is the latter, is there any evidence that they may apply? You make the point again in the next 190 paragraph, so something to back it up would be good.
Authors: Lines 440-450 include one piece of evidence directly relevant to the study (Koba et al., 1997) and two pieces of evidence drawn from the study itself. As pointed out by Referee #1, other studies have shown the presence of at least two pore sizes in Canterbury gravels (Dann et al., 2009). This reference has been added to the paper.   Stewart (m.stewart@gns.cri.nz) Abstract. Nitrate concentrations in groundwater have been historically high (N ≥ 11.3 mg/L) in an area 230 surrounding Tinwald, Ashburton since at least the mid-1980s. The local community are interested in methods to remediate the high nitrate in groundwater. To do this they need to know where the nitrate is coming from. Tinwald groundwater exhibits two features stemming from irrigation with local groundwater (i.e. ir rigation return flow).
The first feature is increased concentrations of nitrate (and other chemicals and stable isotopes) in a 'hot spot' around Tinwald. The chemical concentrations of the groundwater are increased by recirculation of water already 235 relatively high in chemicals. The irrigation return flow coefficient C (irrigation return flow/irrigation flow) is found to be consistent with the chemical enrichments. The stable isotopes of the groundwater show a similar pattern of enrichment by irrigation return flow of up to 40% and are also enriched by evaporation (causing loss of about 205% of the original water mass). Management implications are that irrigation return flow needs to be taken into account in modelling of nitrate transport through soil/groundwater systems and in avoiding overuse of nitrate 240 fertilizer leading to greater leaching of nitrate to the groundwater and unnecessary economic cost. The second feature is the presence of 'denitrification imprints' (shown by enrichment of the δ 15 N and δ 18 ONO3 values of nitrate) in even relatively oxic groundwaters. The denitrification imprints can be clearly seen because (apart from denitrification) the nitrate has a blended isotopic composition due to irrigation return flow and N being retained in the soil-plant system as organic-N. The nitrate concentration and isotopic compositions of nitrate are found to 245 be correlated with dissolved oxygen concentration. This denitrification imprint is attributed to localised denitrification in fine pores or small-scale physical heterogeneity where conditions are reducing. The implication is that denitrification could be occurring where it is not expected because groundwater DO concentrations are not low.

250
Excessive nitrate concentrations in groundwater are of great concern for human health and for the environment. New Zealand drinking-water standards set a Maximum Acceptable Value (MAV) for nitrate at 50 mg/L (equivalent to nitrate-nitrogen of 11.3 mg/L), based on the risk to bottlefed babies (Ministry of Health, 2008), in line with the WHO guidelines (WHO, 2016(WHO, , 2017. Hereafter in this paper "nitrate' is quantified as concentrations of nitrate-N in mg/L. Concerning health, the Canterbury Plains in New Zealand (Fig. 1a) has identified several (OECD, 2017). In New Zealand the N balance worsened (i.e. became more positive) more than in any other OECD member country between 1998 and 2009, almost entirely because of expansion and intensification of farming 260 (OECD, 2013). The N balance is the difference between N inputs to farming systems (fertiliser and livestock manure) and N outputs (crop and pasture production) -a positive N balance indicates a build-up of N and increased potential for N pollution of soil, water and air. Such concentrations result from land use activities, including pastoral and arable farming, market gardening, application of nitrogenous fertilizers and industrial and sewage waste disposal (Laegreid et al., 1999). Pastoral farming has increased rapidly in recent years world-wide and especially in New Zealand. For example, dairy farming acreage on the Canterbury Plains in New Zealand ( Fig.   1a) increased from 20, 000 to 190,000 ha between 1990(Pangborn and Woodford, 2011. Arable farming has also increased and intensified. Health concerns of nitrate concentrations in drinking water relate to infant methaemoglobinaemia (WHO, 2016). The WHO guideline for maximum allowable nitrate concentration in freshwater is 50 mg/L or 11.3 mg/L of nitrate-N (WHO, 2017). Because Nnitrate can also be toxic for aquatic life 270 in groundwater-fed streams at lower levels than MAV, the New Zealand Government has set a maximum median for nitrate-N of 30.5 mg/L or 6.9 mg/L of nitrate-N for river systems. Hereafter in this paper 'nitrate' is quantified as concentrations of nitrate-N in mg/L.
Nitrate concentrations in groundwater in the Tinwald area have historically been high, commonly greater than 11.3 mg/L within an area approximately 3 km wide and 11 km long (Fig. 1b). This area of high nitrate was first 275 identified by Hanson (2002), but nitrate was already elevated in the area in the 1980s (Aitchison -Earl, 2019). The high values are due to the history of land use in the overall area, but the Tinwald values are accentuated because the area is irrigated with local groundwater which has relatively high nitrate concentrations (N ≥ 11.3 mg/L), whereas surrounding areas are irrigated with alpine river water with low nitrate concentrations (N < 1 mg/L). The terms 'irrigation return flow' (e.g. Chakraborty et al., 2015) and 'groundwater recirculation' (Brown et al., 2011) 280 are often applied to situations where irrigation is from water that has been pumped from the underlying aquifer.
This situation is common around the world sometimes with unrecognised effects on chemical concentrations (Sánchez Pérez et al., 2003, Park et al., 2018. An important and well-recognised example of the effects of irrigation return flow is non-point sourced arsenic pollution in the groundwater of the Bengal basin, regarded as one of the largest public health concerns in human history (Edmunds et al., 2015).

285
Irrigation return flow has important implications for water resources management as regards understanding and modelling of nitrate transport in groundwater systems. Much effort is being expended to model the effects of nitrate produced by farming practices in order to substantiate the introduction of appropriate controls on farming to protect the water supplies of downstream communities (e.g. Environmental Canterbury website, 2020).
Irrigation return flow can seriously distort such modelling by extending the time scale of nitrate transport by 290 abstracting water from groundwater downstream and applying it upstream, and also by adding nitrate on a second pass through the soil. This work examines the chemical and isotopic compositions of Tinwald groundwater to look for signatures attributable to irrigation return flow and how it contributes to the nitrate hot spot at Tinwald. Similar effects are expected to be important for many other locations in agricultural areas throughout the world.
Irrigation return flow also appears to contribute to an enhanced 'denitrification Imprint' imprint' in groundwater at Tinwald., where Ddenitrification imprints are discernible in even reasonably oxic groundwaters. The stable isotopes of nitrate ( 15 N and 18 O) have often been used to investigate both the sources of the nitrate and its natural attenuation via denitrification (i.e. microbial reduction of nitrate) (e.g. Mariotti et al., 1988, Böhlke and Denver, 1995, Kendall, 1998, Wexler et al., 2014, Park et al., 2018, Spalding et al., 2019. Understanding the sources of nitrate is important for remediation of excessive nitrate concentrations as at Tinwald (Aitchison-Earl, 2019).

300
Natural attenuation of nitrate via denitrification is a vital eco-service to the environment, and comparison of estimates of nitrate loss by leaching from the bottom of the root zone in catchments compared with the outflow of nitrate from streams shows that considerable attenuation of nitrate occurs in the vadose zone-groundwater continuum. However, little is known about the detailed processes affecting nitrate transport and fate in this region (Clague et al., 2015, Wells et al., 2016, Stenger et al., 2018, Burbery, 2018.

305
In summary, the objectives of this paper are to investigate the role of irrigation return flow in: 1. Accentuating the nitrate hot spot at Tinwald, and 2. Producing denitrification imprints in relatively oxic groundwaters.

310
The study area centres around the small town of Tinwald (population 3000) situated on the south bank of the Hakatere/Ashburton River and located on the large coalescing alluvial plain known as the 'Canterbury Plains' (Fig. 1a). The Canterbury Plains were built up by rivers fed by glaciers over several million years. Deposition in the Tinwald area ( Fig. 1b) was mainly by the South Branch Hakatere/Ashburton River and its ancestors (Barrell et. al, 1996). The alluvial deposits are poorly stratified greywacke gravel dominated with silts and sands which 315 become finer towards the coast. Oil well exploration drilling and seismic surveys of the Ashburton -Hinds areas indicate thicknesses of over 1000 m of alluvial gravels overlying marine sediments (Jongens et. al, 2012).
Existing wells Iin the Tinwald study area are almost all less than 100 m, and over half are less than 40 m deep (Aitchison-Earl, 2019). Wells are generally screened within post-glacial (Holocene) or last glacial (Late Quaternary) deposits. Shallow wells and springs are common close to the river, within the Holocene age deposits.

320
There is little geological impedance for movement of groundwater between shallower and deeper screened wells.
The regional groundwater flow direction is parallel to the Hakatere/Ashburton River. State Highway 1 runs through the study area ( Fig. 1b) and was originally built to take advantage of drier conditions at the inland point of the old 'Hinds swamp'. The swamp has been largely drained but influences soil types, with deeper, poorly drained organic soils with less leaching and greater denitrification potential located coastwards of the highway.

325
Soils are lighter and more freely draining with greater nitrate leaching risk inland of the highway and adjacent to the Hakatere/Ashburton River (Landcare Research, 2015).
In the Tinwald study area, groundwater in two wells with depths less than 50 m had mean ages residence times of 12 and 63 years based on CFC and tritium measurements (Stewart et al, 2002;van der Raaij, 2013). Ages Groundwater residence times generally increase with depth in other wells in the greater Ashburton area. A trial 330 site for managed aquifer recharge (MAR) has been operating since 2016 just outside of the study are a (Fig. 1b).

Hydrology
The closest long-term rainfall site is part of a climate station at Ashburton Council (Fig. 1b). Annual average rainfall at Ashburton Council is around 730 mm (measured between 1909-2017), ranging from 382 to 1147 mm.
There is little seasonality in rainfall, which averages 61 mm a month. Groundwater recharge was reported by Thorpe and Scott (1999) based on lysimeter measurements of soil drainage at Winchmore (10 km north of Ashburton, Fig. 2). In the 10-year period (1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971), average recharge was 293.5 mm/year with average rainfall of 730 mm/year and PET of 765 mm/year. Average monthly recharge was much higher in winter months (April to September). Winchmore soil is described as Lismore Stony Silt Loam characteristic of that at Tinwald west of Highway 1, and much of the Canterbury Plains.

340
The Hakatere/Ashburton River has a north and south branch sourced from the Canterbury Ranges which converge at the north of the study area. The Hakatere/Ashburton River interacts with local groundwater, losing and gaining water along its length. Flow is lost to groundwater from the South Branch, and gains towards the confluence with the North Branch.
Springs and wetlands indicate areas where the water table is naturally close to the surface and groundwater 345 discharge is occurring. Many springs are found in the Hakatere/Ashburton River catchment, and often occur in relict river channels (Aitchison-Earl, 2000). In the study area, Carters Creek and Laghmor Creek are both sourced from springs, and there are springs above Lake Hood that flow into the lake (Fig. 1b).

Land and groundwater use
Cropping has been a major land use in the Tinwald area since at least the early 1940s ( Fig. 1b, Engelbrecht, 2005).

350
Most of the area is not part of any of the major surface water irrigation schemes, so irrigation was developed from groundwater sources within the area from the 1980s. Cultivation and fertilizer practice in cropping has an impact on the amount of nitrate that is leached from the soil to the groundwater. Winter is the most likely time for leaching to occur due to saturated soils and less nitrogen being used by crops. Nitrogen-fixing clover crops have been used  Information on the wells is given in Table 1. Screened intervals and mid-screen depths are given for screened wells, total well depths are given where the wells have no screens. 52% of the wells had short screens (average 2m length), 21% had long screens (average 10m length) and 27% had no screens. Field measurements had 385 stabilised for all wells before sampling. 25 wells were purged of at least three well casing volumes before sampling, the 8 remaining wells were sampled by low flow methods (pumps were lowered into the wells and water was sampled after the pipes had been purged of three pipe volumes).

Chemical measurements
Samples were analysed for Environment Canterbury's standard suite of major ions through Hills Laboratories 390 (Aitchison-Earl, 2019). Field measurements included dissolved oxygen (DO), pH, conductivity, temperature and depth to groundwater. A selection of the field quantity and ion concentration results are given in Tables 1 and 2. The samples have been ordered from lowest to highest DO concentrations, and four groups of samples (A to D) are identified to aid discussion. Groups A and B have low DO values (<4 mg/L) with A having high δ 15 N (15‰) and B moderate δ 15 N (7-9‰). Groups C and D have high DO (>8.2 mg/L) with C having the highest and D the 395 lowest Cl and SO4 concentrations.

Water Isotopes (δ 18 O, δ 2 H)
Water samples were analysed on an Isoprime mass spectrometer; for δ 18 O by water equilibration at 25°C using an Aquaprep device, for δ 2 H by reduction at 1100 °C using a Eurovector Chrome HD elemental analyser. Results are reported with respect to VSMOW2. The analytical precision for this instrument is 0.2‰ for δ 18 O and 2.0‰ 400 for δ 2 H. Results are given in Table 1.

Nitrate isotopes (δ 15 N, δ 18 ONO3)
Nitrate samples (NO3) were converted to nitrite (NO2) using cadmium, then to nitrous oxide (N2O) using sodium azide in an acetic acid buffer. The N2O was then extracted from the water sample, passed through a series of chemical traps to remove H2O and CO2, and cryogenically trapped under liquid nitrogen. After being cryofocused signature of nitrogen and oxygen. Our method is modified from McIlvin and Altabet (2005)

Groundwater chemistry
DO concentrations in the Tinwald groundwaters range from 0.18 to 11.8 mg/L, although the majority are high and indicate relatively oxic groundwater. As noted above, the data in Tables 1 and 2 are ordered from lowest to highest DO values.

415
Chloride concentrations are useful to distinguish recharge sources. Chloride concentrations are highest in rainfall originating over the sea and near the coast, and generally decrease with distance inland. In particular, alpine rivers (with chloride values of 0-5 mg/L) and coastal rainfall-derived infiltration (with chloride ranging from 10-20 mg/L) can be clearly distinguished (Hayward, 2002;Stewart et al., 2002). However, chloride concentrations in the Tinwald area (Fig. 4a) are greater than expected even for coastal rainfall (most are >15 mg/L). The values are 420 lower (0-10 mg/L) to the northeast side of the study area near the Hakatere/Ashburton River and to the southwest.
Sulphate occurs naturally in groundwater and is present in fertilizers and fungicides, and so can be an indicator of human influence when concentrations are in excess of background levels as here (see Fig. 4b). As with chloride, levels in alpine river and low-altitude rainfall infiltration are very different, but in the case of sulphate the difference is caused by the nature of additions to the soils in the respective catchments rather than the 425 concentrations in rainfall. Concentrations are lower on the northeast and southwest boundaries of the study area.
Nitrate concentrations are shown in Fig. 4c. Nitrate concentration exceeds MAV (11.3 mg/L) in 17 of 33 wells sampled in the study area and is over 20 mg/L in four wells. The highest nitrate concentrations cluster inland of SH1 to the west and northwest of Tinwald and underlie an area of dominant cropping land use (Fig. 1b). Nitrate is lowest on the northeast boundary of the study area (near the Hakatere/Ashburton River) where it is generally 430 below ½ MAV, 5.65 mg/L, and lower but still over ½ MAV on the southwest boundary.
To investigate possible irrigation return flow effects, we compare the concentrations of different solutes and isotopes and include the effect of evaporation as indicated by the stable water isotopes (Figs. 5a, b, c). Fig. 5a shows water δ 18 O versus the chloride. Higher δ 18 O correlates with higher chloride, but this is not due to evaporation (because the evaporation vector is not parallel to the trend). Instead the main influence is the source 435 of the recharge because both chloride and δ 18 O are higher in land surface recharge (e.g. Group C samples) and lower in alpine river recharge (Group D samples). There is no effect due to DO. Sample 21 shows an extra evaporation effect.  probably within the soil (see below). Evaporation has an almost negligible effect.
The clear message from these results is that nitrate, sulphate and chloride concentrations are increased in areas irrigated by local groundwater compared to those irrigated by alpine river water. (1)

465
On a global scale, d (called the deuterium excess) is equal to +10‰ on average, and the linear relationship is known as the Global Meteoric Water Line (Craig, 1961). A New Zealand Meteoric Water line with deuterium excess, d = +13‰ was established by Stewart and Taylor (1981), based on westerly -dominated zones of New Zealand (North Island and western parts of the South Island). However, Scott (2014) and previous workers found that data from Canterbury fitted local meteoric lines with d = +10‰ for the Waimakariri Zone and d = +11‰ for 470 the Selwyn-Te Waihora Zone. Scott (2014) noted a lack of paired δ 18 O and δ 2 H data in the Ashburton area, but aAvailable evidence supports a local meteoric water line (LMWL) for Canterbury with intercept d = +10‰ (Stewart and Taylor, 1981;Taylor et al., 1989;Stewart and Morgenstern, 2001;Scott, 2014;Stewart et al., 2018).
i.e. the LMWL is This is taken as the LMWL below.
The slope of less than 8 for this line suggests indicates that the waters have been affected by evaporation. A ratio of about 5 in the 2 H and 18 O enrichments is expected for evaporation at ambient temperatures (Stewart, 1975). It is expected likely that the isotopic compositions of the water would have been enriched by evaporation and/or evapotranspiration during the irrigation return flow process. Estimates of the isotopic enrichments required to explain the displacement of the average isotopic compositions of the group from the LMWL are shown by the red arrow with slope of 5 in Fig. 7. The average isotopic composition of the samples except sample 21 (Table 1) Fig. 7). This degree of isotopic enrichment is obtained by evaporation of approximately 205% of the water according to both isotopes (see the calculation in the aAppendix 1based on Stewart, 1975). Uncertainty in this fraction evaporated is small because the isotopic composition of the remaining water changes rapidly with the degree of evaporation due to the form of the equation, so considerable changes of isotopic composition do not 490 change the fraction evaporated much.Recharge would be a mix of rainfall and irrigation water which would be evaporated. The effects would be expected to be variable from well to well as observed.
In addition, the difference in the δ 18 O and δ 2 H values of Groups C and D is attributed to their different irrigation sources (local groundwater or alpine river water) as observed for the chemical compositions. Assuming that both groups are affected by evaporation to the same extent, the difference between the groups compared to the 495 difference between the irrigation sources will give an approximate measure of the irrigation input. The δ 18 O difference between Groups C and D is 0.63‰ (Table 4) and that between the sources is 1.63‰, giving irrigation input of 39%. For δ 2 H it is 4.1‰ compared to 10.2‰ giving 40% irrigation input. These are likely to be overestimates because Group C waters may be more affected by evaporation than Group D waters. (The compositions of the irrigation sources are taken as local rainfall (-8.17, -58.7) and alpine river (-9.80, -68.9)).

Nitrate isotopes δ 15 N and δ 18 ONO3
The nitrate isotope results are given in Table 1

505
The first feature is the linear relationship between the δ 15 N and δ 18 ONO3 values of the nitrate (except sample 06 and to a smaller extent samples 02 and 03). Denitrification causes increased δ values of nitrate, along with decrease of nitrate concentration. The slope of the isotopic enrichments caused by denitrification (i.e. enrichment in δ 18 ONO3/enrichment in δ 15 N) has been reported to be in the range 0.48 -0.77 (Kendall, 1998;Burns et al., 2011;Kaushal et al., 2011;Zhang et al., 2019). The line shown in Fig. 8a has a slope of 0.68 and was calculated to   Canterbury (Trevis, 2012)., Bbut we think its contribution is not large, because the abundance of clover has decreased over the years as fertilizer use (particularly urea) has increased.
The approximate Rayleigh formula used is where δ is the δ 15 N or δ 18 O value of the nitrate remaining after the microbes have catalysed partial denitrification, and δo is the initial composition of the nitrate. ε is the enrichment factor for the reaction and f is the fraction of nitrate remaining after the reaction. Results of the calculation are given in Table 3. gave an ε( 15 N) range from -5 to -8‰. Other authors (Kendall, 1998, and references therein) gave much larger negative values. Mariotti et al. (1988) suggested that low values may occur if denitrification occurs in dead -end pores causing a non-fractionating sink for nitrate by diffusion. Stenger et al. (2018) considered that small -scale 550 physical heterogeneity, including the localised distribution of resident electron donors and the effect of lateral flows, was a more likely cause with their coarse-textured ignimbrite materials.

Irrigation return flow effects on chemical and isotopic concentrations
Fertilizers have been applied to much of the area between the Ashburton and Hinds Rivers not just to the Tinwald study area, and rainfall applies to the whole area with contours of the δ values in rainfall decreasing inland from the coast (Stewart et al., 2002). Yet the Tinwald area shows elevated nitrate concentrations (and chloride, sulphate, 585 etc.) compared to the surrounding areas (see Fig. 2). The difference is that the Tinwald study area is irrigated by groundwater from shallow local wells with high solute concentrations, whereas much of the rest of the area is irrigated by alpine river water with low concentrations of solutes. This has affected the chemical and water isotope concentrations.
The irrigation return flow process is illustrated schematically in Fig. 9. Dewandel et al. (2007) defined an irrigation 590 return flow coefficient C equal to the recharge from irrigation (i.e. irrigation return flow, IRF , divided by the irrigation flow itself, I, so that C = IRF/I). C is equal to the overall recharge rate for rainfall and irrigation in our system (we omit surface runoff and lateral seepage in this treatment because both are expected to be small). (Note that drains in the area are fed mainly by groundwater.) C is used to quantify the effect of irrigation return flow on the water balance.

595
The average chemical concentrations of Groups C and D are taken as representative of the Tinwald hot spot and outside groundwaters respectively (average values and standard deviations for each chemical are given in Table   4). The table also gives the enrichment factors and irrigation return flow coefficients between Groups C and D derived from each chemical; the enrichment factor is equal to 1/C assuming that chemicals put in via rainfall and irrigation are concentrated into the recharge fraction, i.e. are concentrated by the loss of AET . Cl mass balance 600 has often been used to estimate recharge because it is conservative, here the enrichment factor is 3.2 and coefficient C obtained is 31%. The SO4 enrichment factor is very large (7.0) suggesting greater fertilizer input into the Group C area soil than into the Group D area soil. The HCO3 factor is small (1.2) probably because of chemical re-equilibration as water passes through the soil in both areas. The average enrichment factor is 2.4 and coefficient C is 42% for all of the chemicals except SO4 and HCO3. The coefficient C can be compared with 605 lysimeter measurements of recharge fraction from the nearby research station of Winchmore (Thorpe and Scott, 1999, Fig. 2); the recharge fraction is recharge divided by input (i.e. rainfall plus irrigation). Thorpe and Scott found that the recharge fraction at Winchmore was 43% for the average irrigation input of 300 mm/year, in approximate agreement with the chemical results above.
Recharge to groundwater (RGW) is given by where R is rainfall, I is irrigation input and AET is actual evapotranspiration (all in mm/year). Surface runoff and lateral seepage are neglected in this treatment as both are expected to be small. (Note that drains in the area are fed mainly by groundwater.) Thorpe and Scott (1999) reported on many years of lysimeter measurements of recharge at the nearby research station of Winchmore. They found that 615 = 0.80( + ) − 380 (6) for pasture, while for bare soil it was = 0.77( + ) − 244 (7) The recharge fraction (RGW%) is the ratio of recharge to incident rainfall plus irrigation applied, i.e. if the recharge rate is assumed to be the same for both rainfall and irrigation. C is used to quantify the effect of irrigation return flow on the water balance.

625
Equations (6) and (7) can be used to determine the coefficient C for Tinwald. Assuming that I = 200 mm, C is 364/930 = 39% for pasture, and 51% for bare soil. These give predicted chemical enrichment factors (1/C) of 2.6 for pasture and 2.0 for bare soil (if chemicals put in via rainfall and irrigation are concentrated into the recharge fraction, i.e. are concentrated by the loss of AET). The pasture value may be the most appropriate to compare with the observed values in Table 4 for the different chemicals. The observed values are very scattered, but are similar on average to the predicted value of 2.6. Chloride is the most conservative element and its enrichment factor (3.2) is a little higher than the predicted value. The sulphate enrichment factor is very large suggesting greater fertilizer input into the Group C area soil than into the Group D area soil. The bicarbonate factor is small perhaps because of chemical re-equilibration as groundwater passes through the soil in the Group C area. Another effect of irrigation return flow is distortion of tracer age dating results. Tritium concentrations will not be reset by interaction with the atmosphere when irrigation water is applied to the soil, so the tritium ages residence times of groundwater affected by irrigation return flow will appear to be older than they really are. In contrast , CFC/SF6 ages residence times will be reset to zero in the soil and groundwater ages residence times will reflect 650 time since recharge. This appears to be the case for data in the Tinwald area, although data is scarce (Stewart et al., 2002).
A practical consideration is that if irrigation water already contains nitrate then too much fertilizer could be applied leading to unnecessary economic cost and greater nitrate leaching potential, if the nitrate in the groundwater is not accounted for by nutrient budgeting (e.g. Flintoft, 2015).

Nitrate source identification
Nitrate isotope results that have not been affected by denitrification (i.e. usually the oxic samples) potentially give information on the nitrate sources and also on the starting points for denitrification vectors. Numerous studi es of the δ 15 N values produced by different nitrate sources have identified ranges which have differed under local 660 conditions (e.g. Kendall, 1998;Fogg et al., 1998;Stewart et al., 2011, Fig. 8b). Results for oxic samples from recent New Zealand studies are given in Fig. 10. The rectangles show source signature fields resulting from urea fertilizer/soil N/ruminant excreta at Toenepi Catchment (Clague et al., 2015), urine/urea/soil N at Harts Creek (Wells et al., 2016), low intensity animal grazing (soil N/manure) at Waihora wellfield (Stenger et al., 2018), two sources (inorganic fertilisers/manure and piggery effluent) at Waimea Plains , and inorganic 665 fertiliser/urea/manure at Tinwald (Groups C and D, this work). Despite the variety of nitrate sou rces, the δ 15 N values generally show overlapping ranges as illustrated in Fig. 10 (except for the Waimea Plains piggery effluent source).
Use of δ 18 ONO3 in combination with δ 15 N to identify nitrate sources has not been very successful, as illustrated in Fig. 10 where the δ 18 ONO3 values overlap each other. On the other hand, the combination has proven to be effective 670 for detecting the occurrence of processes in the nitrogen cycle, such as nitrification and denitrification (Aravena and Robertson, 1998). The only distinctive source δ 18 ONO3 values are those expected for nitrate fertilizer (see 'nitrate fertilizers' box in Fig. 10, Xue et al., 2009, Wells et al., 2015. Many researchers have looked for such δ 18 ONO3 values and generally failed to find them (Kloppman et al., 2018). Instead the values observed in groundwaters are usually characteristic of soil nitrate or effluent (as illustrated in Fig. 10).
The probable answer to this failure to observe the expected high δ 18 ONO3 values in groundwater is that inorganic fertilizer-derived nitrate is not directly and rapidly transferred to groundwater but is retained in the soil -plant system as organic-N, and only later mineralised and re-oxidised thereby becoming available for leaching to the groundwater (Somers and Savard, 2009, Wells et al., 2015, Kloppmann et al., 2018. The process of mineralisation and re-oxidation resets the δ 18 ONO3 and also changes the δ 15 N. The time delays in this process can be considerable 680 (as much as several decades, Sebilo et al., 2013). This means that there will be a legacy of organic -N built up in the Tinwald soil from past applications of fertilizer in addition to past soil management practices. This time delay is in addition to the time delay due to the mean residence time of the groundwater. Others have previously identified the importance of organic-N in the soil (variously known as soil organic matter (SOM, Somers and Savard, 2009) or soil organic nitrogen (SON, Wells et al., 2015)) as the pool of nitrogen within the soil controlling 685 the rate and timing of nitrate releases to groundwater. The transfer to organic -N is most efficient at times of high microbial activity (spring/summer growth) and much less in low microbial activity (winter), when increased nitrate leaching to the groundwater is likely (Mengis et al., 2001;Somers and Savard, 2009).
The nitrate isotopes (Fig. 8a) show an unexpected blending of the isotopic compositions of the nitrate in the groundwater (and therefore the soil/vadose zone). This blending is considered to be due to irrigation return flow 690 in conjunction with the action of organic-N in mediating and retaining N in the soil. This has allowed the denitrification process to be identified and explored in this study, and the enrichment factors for denitrification to be determined.

Denitrification imprint in oxic groundwater
The nitrate isotopes show clearly that denitrification is important in the Tinwald soil and vadose zonesgroundwater 695 (Fig. 8). Firstly, the nitrate isotopes show that the nitrate sources are blended within the soil and that inorganic

710
Several factors suggest that the denitrification imprint arises from localised denitrification in fine pores where conditions are reducing. 1. Koba et al. (1997) showed that denitrification can occur in anaerobic pockets within otherwise oxic sediments or water bodies. 2. The low values of ε( 15 N) and ε( 18 O) observed here indicate that denitrification occurs in fine pores or small-scale physical heterogeneity. 3. The occurrence of the denitrification imprint in moderately oxic waters (in which denitrification could not have occurred) means that the denitrification must have occurred in parts of the system which were much more reducing. Logically these are fine pores or inhomogeneities containing electron donors with heterotrophic bacteria.
The Tinwald study area is not in an area where the groundwater is generally reducing (Close et al., 2016), but nevertheless groundwater from some wells show the denitrification imprints. It would appear that denitrification imprints in moderately oxic groundwater should be common, but many more nitrate isotope measurements would 720 be required to show this.
As a final comment, there appear to be two types of pore space in the gravels at Tinwald, i.e. larger pores with mobile water and finer pores with almost stagnant water, that communicate by diffusion (e.g. Dann et al., 2009).
This is likely to cause slowing of nitrate transport and decrease of nitrate within through the system as nitrate is retained is transferred to the finer pores (together withand partially denitrified some denitrification).
and δb is the isotopic composition of the atmospheric vapour relative to the initial composition of the water, αp and αk are the equilibrium and kinetic fractionation factors respectively between water and vapour, and h is the 760 relative humidity.

765
Chemical enrichment due to evaporation is given by where Co and C are the initial and final concentrations of the chemical.

Appendix 2: Isotopic effect of denitrification on nitrate
The approximate Rayleigh formula (Kendall, 1998) was used to calculate the isotopic effects of denitrification on 770 nitrate. This formula is where δ is the δ 15 N or δ 18 ONO3 value of the nitrate remaining after the microbes have catalysed partial denitrification, and δo is the initial isotopic composition of the nitrate. ε is the enrichment factor for the reaction and f is the fraction of nitrate remaining after the reaction. Results of the calculation are given in Table 3.

775
Appendix 3: Isotopic effect of mixing of two sources of nitrate A mixing curve between two nitrate source end members (soil nitrate and fertilizer/effluent mixture nitrate) has been fitted to the solid green points (not plotted in the figures). The equation of the curve (Kendall, 1998) is: where δ is the δ 15 N or δ 18 ONO3 value of the sample nitrate, b is the δ 15 N or δ 18 ONO3 value of the fertilizer/effluent