The Anthropic Pressure on the Grey Water Footprint: The Case of the Vulnerable Areas of the Emilia-Romagna Region in Italy

Abstract

Nitrogen (N) is an important component of plant development, yet its application and contamination are a global issue. Diffuse source pollution and its effects on ecosystem health are notoriously difficult to track and control. This paper answers the Sustainable development Goal 6 goal focused on reducing water pollution by improving the understanding of nitrate emissions to groundwater and their resulting water pollution level in the Emilia-Romagna region in Italy. The Grey water footprint (GWF) and water pollution level (WPL) were used as indicators and geospatial maps were constructed in order to determine if N levels exceeded groundwater quality demand from 2014 to 2020. Moreover, a selection of specific agricultural sites in the Piacenza district has been performed to better understand the potential anthropogenic nitrate pollution due to the agricultural sector. In the selected sites, the predicted nitrate pollution due to agricultural practice has been compared with the nitrate concentration measured in samples collected across the period 2015–2018. The regional results show that approximately 70% of the analysed sites resulted in a total N load exceeding the estimated agricultural load to groundwater. The analysis conducted in three selected wells in the Piacenza district shows the sporadic exceedances of the legal limit and demonstrates the presence of anthropogenic pressures of various natures insisting on the surrounding area and confirms a potential non-agricultural point or diffuse pollution source.

1. Introduction

Groundwater nitrate (NO₃⁻) contamination is a global environmental concern [1,2]. The presence of nitrate in drinking water can cause serious health problems, while its presence in the environment can lead to negative effects such as eutrophication and seasonal hypoxia [3]. Besides the natural occurrence of nitrate in water bodies, various anthropogenic drivers can be identified, among which fertilisers and animal manures, together with industrial and domestic wastewater and solid waste discharge, are the most relevant [3]. In fact, groundwater leaching is the natural consequence of diffuse/punctual spreading in soil/subsoil of large quantities of synthetic nitrogen (N) fertilisers and/or animal manure in areas with a higher agricultural vocation and industrial and civil wastewater discharge in highly urbanised areas [4]. The nitrate concentration in groundwater depends both on the extent of anthropogenic pressures and on the intrinsic vulnerability of aquifers to pollution. Among the various mineral forms of N present in the soil, nitrate is the most mobile and susceptible to leaching; losses by leaching can reach 99% of the present nitrates because they depend on the concentration of NO₃⁻ in the soil, the volume of drainage, the soil texture and structure, and climatic factors [5]. Different countries and organisations have established maximum permissible nitrate concentrations for drinking water and groundwater. This limit has been set at 50 mg/L in Europe, according to Directive 91/676/EEC [6], which has been released to protect water quality across Europe by preventing nitrates from agricultural sources from polluting groundwaters and surface waters and by promoting the use of good farming practices. In fact, it requires member states to identify nitrate vulnerable areas, implement a set of measures to reduce water pollution by nitrates, and prevent it by adopting action programs and by drawing up codes of good agricultural practice on fertilisation, soil management, and irrigation to minimise the release of nitrates. Developing effective management practices to preserve water quality and remedial plans for already polluted areas requires identifying sources and understanding the processes affecting nitrates’ local presence. However, the correspondence between nitrate concentrations in groundwater and surface water and the amount of nitrate coming from a source is difficult to be detected because of the existence of multiple nitrate sources from different areas, the overlapping of point and diffuse sources, and the coexistence of different biogeochemical processes that may vary the concentrations injected [7].

Globally, special attention was provided to the impacts on air quality of lockdowns consequent to the COVID-19 pandemic [8,9,10,11,12], while the effects on water quality were rarely considered; according to Bielefeld Academic Search Engine [13], searching the word COVID coupled with air and water quality provides 4881 and 1821 results, respectively. In water quality studies, however, an improvement in the water quality has been observed after the reduction in anthropogenic, mainly industrial, activities imposed by the lockdown [14,15,16].

The Sustainable Development Goals (SDGs) proposed in the 2030 Agenda adopted by United Nations in 2015, include a goal (SDG 6) totally devoted to water quality and sustainability via specific targets [17]. Among all targets, 6.3 is devoted to “improve water quality by reducing pollution, eliminating dumping and minimising release of hazardous chemicals and materials, halving the proportion of untreated wastewater and substantially increasing recycling and safe reuse globally”. In order to monitor progress in achieving this specific target, an indicator must be adopted that ensures a direct measure of anthropogenic activities that lead to water pollution. Different tools and approaches have been used to assess and monitor the effect of anthropogenic activities on water quality, among which the Water Footprint (WF) has been widely used [18,19,20,21,22]. In fact, WF has grown in popularity among scientists, policymakers, and the general public due to its ability to effectively communicate water issues to a wide range of audiences [23]. WF is a spatial-temporal explicit indicator of water consumption and pollution, providing a view on how human activities affect water resources [24]. The Grey WF (GWF) indicates on volumetric terms the severity of water pollution by combining loads and legal or eco-toxicological endpoints and is strictly connected to the water pollution level (WPL), which is an indicator used for evaluating if the freshwater recharge is sufficient to assimilate pollution. In general, comprehensive studies that consider the application of N-GWF at the region scale in Italy are lacking in the literature. To our knowledge, only two studies have been conducted in Italy, both in Apulia Region by Serio et al., 2018 and Miglietta et al., 2017 [19,25], where groundwater pollution has been estimated by WPL indicator calculated from actual groundwater concentrations provided by regional databases. A study by Aldaya et al., 2020 [26], conducted in the Navarra region (Spain), uses the GWF indicator to assess agricultural freshwater pollution and compare WPL with actual concentrations in water bodies. Up-to-date publications on the topic always estimate the agricultural contribution to nitrate groundwater pollution without exploring potential non-agricultural sources [27,28] and in one case investigating industrial and urban contributions, but in terms of other pollutants different from nitrates [29]. Moreover, in the studies mentioned above, the comparison and integration between the N-GWF and WPL obtained from the nitrate concentration reported in regional databases or estimated by different approaches and sources is still lacking. The WPL has been obtained by regional databases and the N-GWF has been obtained by total nitrogen surpluses using total nitrogen inputs and outputs.

This paper aims to improve the understanding of nitrate emissions to groundwater and the resulting water pollution level obtained from different approaches and sources, evaluating the relation between groundwater nitrate contamination and anthropogenic activities from 2014 to 2020 using the GWF indicator. A specific case study has been conducted focusing on the Emilia-Romagna region in Italy. The paper answers SDG 6, which is focused on the supply of high-quality drinking water and to the target of 6.3. To better understand the potential anthropogenic nitrate pollution due to the agricultural sector, a selection of specific agricultural sites in the Piacenza district has been performed. In the selected sites, the predicted nitrate pollution due to agricultural practice has been compared with the nitrate concentration measured in samples collected in the selected sites over the different years.

2. Materials and Methods

2.1. Data Collection

The groundwater nitrate concentrations of the Emilia-Romagna region were extracted from Agenzia Regionale Prevenzione Ambiente Emilia-Romagna (ARPAE) database from 2014 to 2020 [30]. Of the 447 total sites, wells with only one analysis during the year were excluded. The land use in Emilia-Romagna and site-specific annual rainfall were also collected from ARPAE [31]. The agricultural statistics, such as fertilisers, use, N content of the fertilisers, total and crop-specific cultivated area, and crop yields, were all collected from the Italian National Institute of Statistics, while the maximum N allowed in each crop was collected by regional directives [32,33].

2.2. Case Study: Area of Study, Site Selection, and Chemical Analysis

Emilia-Romagna is a northern Italy region characterised by a plain area in the north, a mountain area in the south, and a coastal area in the east with a long-lasting agricultural tradition. The climate of Emilia-Romagna is highly varied from the Adriatic Sea to the Po Valley hinterland. It is generally classified as temperate subcontinental with very hot and humid summers and cold and harsh winters, tending to be sublittoral only along the coastal strip and with low rainfall in the plains (up to 800 mm/yr), as reported in

Table 1. Type of climate, average temperature with reference to the 1961–1990 period, and annual cumulative precipitation in the Emilia-Romagna region [34].

Area	Type of Climate	Annual Cumulative Precipitation (mm/yr)	Average Temperature (°C)
Sublittoral/Coastal strip	Subcontinental	500–600	14
Po Valley	Continental	600–800	13
Apennines	Montane	800–2500 [35]	9 [35]

[34]. Higher precipitations usually occur in the Apennines where, due to the increase in altitude, the winters are colder and summers warmer [35]. The Piacenza area has a rather continental climate due to its central position in the Po Valley between the Apennines and the Po River; the climate differs from the other provinces because it has slightly colder winters and higher rainfall due to its proximity to the Ligurian watershed.

Comparing the current climate (1991–2020) with the thirty-year reference period 1061–1990 shows a very drastic climate change. ARPAE reports a considerable increase in temperatures in both the winter and summer months; worsening of the hydrological balance in the spring–summer period; reduced snow accumulation in winter; a trend towards the concentration of precipitation in narrow periods with increasingly intense phenomena; and increase in the frequency and duration of drought periods [36]. Climate change has inevitably led to changes in groundwater levels in groundwater bodies. The average annual changes in groundwater levels show, based on the frequency distribution of levels from 2002 to 2020, that groundwater bodies in the lowlands and alluvial cones have a deeper level than the neighbouring water bodies due to reduced water recharge because of the decrease in actual precipitation in the winter–spring period [36].

Three wells have been selected in the Piacenza district as reported in Figure 1, considering specific areas where, according to the cultivation register, high rates of N have been applied and there was groundwater not confined by shallow groundwater (

Table 2. Hydrogeological description of the Piacenza district [37,38].

Area	Groundwater Resources	Groundwater Availability	Subsoil Geology	Type of Water Source
Valley	Conspicuous distributed	High	Undifferentiated overflows	Artesian aquifer
Middle-hill	Fluctuating	Fluctuating	Permeable semipermeable rocks	Subterranean aquifer
Low-hill	Conspicuous distributed	Modest	Gravelly layers	Deep or artesian aquifer

). All the selected wells are protected from any point of contamination and no water treatment/filtering system is present before the sampling point.

The sampling sites PIA and PON are both located in proximity to the small city of Pontenure in a rural area with crop prevalence over the livestock and with maise, tomato, and wheat production. The sampling site GOS is instead located in proximity to the small city of Gossolengo, with a small number of livestock and a crop prevalence of maise, tomato, and wheat characterise the area.

The soil size fractions (gravel, sand, silt, and clay) have been detected through X-ray sedigraph/laser granulometer. The ultraviolet spectrophotometric screening method and the principle according to which nitrate in solutions containing sulphuric and phosphoric acids reacts with 2,6-dimethylphenol to form 4-nitro-2,6-dimethylphenol has been used to determine nitrate (Spectrophotometer DR6000—Hach Lange LCK kit nitrate test). The water samples were analysed using the analytical method “Cuvette test LCK 339 nitrate” for a predicted concentration of 1.0–60.0 mg/L of NO₃⁻ and the analytical method “Cuvette test LCK 340 nitrate” for a predicted sample concentration corresponding to 22.0–155.0 mg/L of NO₃⁻. The quantification was carried out using the external standard approach, by utilising a certified reference material (KNO₃, potassium nitrate, ACS, 99.0%, Alfa Aesar, Ward Hill, MA, USA). The validation of the method was carried out according to Document No. SANCO/12495/2011 “Validation of the method and quality control procedures for the analysis of pesticide residues in food and feed” [39].

2.3. Grey Water Footprint and Water Pollution Level

N is one of the most important fertilisers in agriculture and is usually applied in nitrate, ammonium, or organic form. Moreover, in most fertilised agrosystems, NO₃⁻ has been identified as the main form of N leached [17]. In this study, we assume that all N applied is eventually converted into nitrate form over time [40]. The GWF values have been obtained from the Water Pollution Level (WPL) (Equation (1)), which has been obtained by groundwater nitrate concentrations measured in each well (C_measured) and C_max and C_nat that represent, respectively, the maximum allowed concentration of the pollutant in groundwater and the natural concentration in absence of anthropic activities. In the context of pollution by nitrates, we considered C_max at 50 mg/L, which is the maximum allowed concentration according to national and European directives (DLgs 30/09, 2006/118/CE). In “Grey water footprint accounting: Tier 1 supporting guidelines” (hereafter referred to as guidelines) [40], a value of 0.4 mg/L for C_nat is suggested: this value is negligible in comparison to the legal limit. Thus, C_nat was assumed to be 0, as reported by [18,19,20].

$$\mathrm{WPL} =\frac{{\mathrm{C}}_{\mathrm{measured}}}{{\mathrm{C}}_{\max }-{\mathrm{C}}_{\mathrm{nat}}}$$

(1)

The WPL measures how much of a water body’s pollution-absorption capacity has been utilised. A WPL equal to 1 implies that the capacity for pollutant absorption has been completely used. A WPL greater than 1 indicates that the environmental assimilative capacity has been exceeded, resulting in a water quality violation.

The GWF (m³/ha/yr) has been obtained from the WPL and the annual water recharge (Q) following Equation (2):

$$\mathrm{GWF}=\mathrm{WPL}\times \mathrm{Q}$$

(2)

where Q is the groundwater recharge (m³/ha/yr) and is obtained by the annual cumulative rainfall, according to Lamastra et al., 2014 [22]. GWF has been used as an indicator to translate nitrate groundwater concentration into resulting water pollution.

To evaluate the non-agricultural contribution to the pollution level and considering that N leaching is positively correlated with N surplus in each well (experimental or from ARPAE), a N surplus L_sur was estimated as follows:

$${\mathrm{L}}_{\mathrm{sur}}={\mathrm{L}}_{\mathrm{tot}}-{\mathrm{L}}_{\mathrm{agr}}$$

(3)

where L_tot is the overall N load obtained by analytical data and L_agr is the estimated N load from the agricultural sector.

The agricultural load L_agr was estimated according to Equation (4) from the GWF tier 1 manual [40]:

$${\mathrm{L}}_{\mathrm{agr}}=\mathsf{\alpha }\times \mathrm{DOSE}$$

(4)

where DOSE is the amount of applied N (kg N/ha/yr) and α is the fraction of applied N that reach groundwater. Fraction α for the experimental sites has been evaluated according to the model proposed by [40], while for ARPAE sites it has been assumed to be equal to 0.1, as suggested in the guidelines for a situation where it is difficult to estimate due to the lack of data.

The total N load that reaches the groundwater has been obtained by the N-NO₃ concentration (C_N) multiplied by the volume of water that recharged the water body (Q), as follows:

$${\mathrm{L}}_{\mathrm{tot}}={\mathrm{C}}_{\mathrm{N}}\times \mathrm{Q}$$

(5)

This approach assumes that the amount of water and nitrates before the recharge is negligible.

3. Results and Discussion

3.1. Regional Nitrates Concentrations, GWF, and WPL

Table 3. Percentage of sites coupled according to nitrate concentration level ranges.

	2014	2015	2016	2017	2018	2019	2020
<25	70.77%	68.42%	68.93%	74.66%	70.56%	67.96%	61.48%
25–40	12.73%	11.90%	12.02%	10.81%	13.64%	14.56%	12.30%
40–50	3.76%	6.41%	6.80%	5.30%	6.06%	7.44%	11.48%
>50	12.73%	13.27%	12.24%	9.23%	9.74%	10.03%	14.75%
95° (mg/L)	75.05	73.16	73.50	63.60	66.96	60.00	62.45

shows the number of sites exceeding the limit concentration of 50 mg/L for nitrates in the period from 2014 and 2020 and the nitrate concentration of the 95th percentile. From 85.3% to 90.8%, the wells have nitrate concentration below 50 mg/L, with a nitrate concentration of the 95th percentile between 60 mg/L and 75.05 mg/L. Despite concentration values offering a good general view of pollution level alerts in the entire region, the WPL and GWF provide a better perspective about the actual concern about the anthropic activity. The WPL has been calculated using data collected from ARPAE related to the 2014–2020 period as the study conducted in the Apulia region (Italy) [19,25].

Table 4. Annual precipitation, WPL, GWF, and L_sur statistical analysis.

	2014	2015	2016	2017	2018	2019	2020
Average annual precipitation (mm)	949	725	739	543	764	869	629
Average annual Q (m3/ha/yr)	8197	6262	6382	4687	6600	7501	5437
Average WPL (-)	0.65	0.63	0.57	0.57	0.52	0.58	0.48
Average GWF (m3/ha/yr)	5350	4000	3632	2671	3438	4320	2696
Max GWF (m3/ha/yr)	26,236	17,965	14,932	23,993	17,423	20,609	11,728
Min GWF (m3/ha/yr)	61	44	49	38	54	59	40
Median GWF (m3/ha/yr)	3711	3155	2531	1762	2394	3835	1975
95° percentile GWF (m3/ha/yr)	15,207	11,312	9776	7309	10,318	10,742	7186
N DOSE (kg/ha/yr)	100.00	108.40	131.93	123.84	103.69	99.47	161.74
% Sites with Lsur	80%	74%	71%	64%	70%	77%	58%
Average Lsur (kg/ha/yr)	64	49	43	32	43	52	32
Max Lsur (kg/ha/yr)	286	192	155	259	186	223	116
Min Lsur (kg/ha/yr)	0	1	0	1	0	1	0
95° percentile Lsur (kg/ha/yr)	164	123	105	78	114	123	86

shows the GWF, WPL, Q, and precipitation data obtained for the years from 2014 to 2020; moreover, it shows the percentage of the number of sites for which an N surplus (L_sur) has been estimated considering the average agricultural use and the detected concentrations. From the obtained results, the effect of increased rainfall seems to be related to the intensification of N leaching: a correlation of 0.94 between the precipitations and average annual L_tot was obtained. Some authors [41,42,43] showed that rainfall intensification increased deep percolation, altering soil moisture patterns. The effects of rainfall can potentially modify N cycling and losses and microbial dynamics, including mineralisation as well as plant N dynamics, N uptake, and productivity, all of which could affect soil N availability for loss [44,45]. In 2020, despite an increase in N fertiliser application, a lower N surplus, WPL, and GWF was obtained. These lower results could be associated with the reduced contribution of non-agricultural anthropogenic activities. In fact, the agricultural sector in Italy did not stop during the lockdown period imposed by the Italian government [33]. In Figure 2, geospatial data of nitrate concentrations, WPL, GWF, and annual precipitations for the years from 2014 to 2020 are reported, obtained with an inverse distance weighting method of interpolation (WPL and GWF) and with a triangulated irregular network method (precipitations), with each map constructed with grouped areas with the same value based on distribution quintiles each represented with different colour gradations.

In the Apulia study [19], the WPL resulted high in two areas that are exclusively characterised by agricultural land use. Our results, instead, show the presence of several sites with mixed land use that present high values of WPLs. Moreover, in the Apulia region, the main crops are vineyards and olive groves and different studies confirm that perennial cropping systems are moderately fertilised [46,47,48] and are, therefore, leach less N than more heavily fertilised crops than the arable ones. Another critical factor that influences N leaching is precipitation: the average Apulia precipitation level is reported to be 460 mm/yr [49], lower in comparison to 600–800 mm of Emilia-Romagna. Finally, Apulia groundwater availability is mainly concentrated in large and deep aquifers [49], leading to lesser N leaching compared to Emilia-Romagna. All these factors, together, contribute to the WPL differences in our study, highlighting that the agricultural land use and the hydrogeological and meteorological features of the area are crucial in the assessment of environmental risks. In the Navarra study [26], as reported in our case study, the higher fertiliser contribution to N load is due to cereals (54%); moreover, it reported a high significative correlation between the WPL and actual freshwater concentrations, while in our study, the WPL is directly estimated starting from actual pollutant concentrations. Considering our case study, a good correlation between the calculated and estimated GWF is observed as shown in Section 3.2, “Case Study: chemical analysis, WPL, GWP, L_sur”. Both studies highlight that the hydrogeological and meteorological features of the interested area are crucial in the assessment of environmental risks. Hence, the GWF assessment employed in the present study creates a helpful indicator for agricultural policy planning procedures and a criteria for establishing land use management based on hydrological aquifer conditions [50].

3.2. Case Study: Chemical Analysis, WPL, GWP, and L_sur

In order to validate the method of analysis, the lower limit of quantification (LLOQ) and lower limit of detection (LLOD) were calculated as 10 and 3 times the value of the blank. The LLOQ and LLOD calculated were 0.63 mg/L and 0.21 mg/L expressed as NO₃⁻ concentrations. Furthermore, procedural recovery tests were carried out to verify the method’s performance during the analysis. An average recovery of 80–110% with a relative standard deviation of preferably < 20% was adopted as an acceptability criterion.

Table 5. Soil texture in proximity to the experimental sites located in Piacenza area.

Piez	Lat	Long	Sand (%)	Silt (%)	Clay (%)	pH	O.C. (g/kg)	D. (m)	S.S.D. (m)	C/U	Crops
PIA	44.992	9.7541	19.3	54.3	26.4	7.5	9.7	11	18	U	M, T, W
PON	45.006	9.8094	3.0	51.2	45.8	6.8	22.9	9	6	U	M, T, W
GOS	45.007	9.6315	10.8	52.4	36.9	8.4	23.3	20	0	U	M, T, W

Piez: piezometer; Lat: latitude; Long: longitude; O.C.: organic carbon; D.: depth; S.S.D.: start screen depth; C/U: confined or unconfined aquifer; M: maise; T: tomato; W: wheat.

shows the soil texture used to predict N leaching into groundwater. The soil analysis results are shown in Table 5. The Pontenure and Gossolengo sites present a silty-clay texture with comparable organic carbon contents, while the Piacenza site presents a silty-loam texture with lower organic carbon contents.

3.3. Sites Description, and Nitrate Detection

The nitrate concentrations of the 48 samples obtained from the monitored three sites in the period 2015–2018 are summarised in

Table 6. Groundwater nitrate concentration during the monitored period.

Piezometers	Period	n° Samples	Min–Max NO3− (mg/L)	Average NO3− (mg/L)	n° Samples > 50 mg/L
PIA	2015–2018	26	4.99–57.11	17.83	3
PON	2015–2018	13	1.46–99.60	41.41	5
GOS	2015–2018	9	29.77–52.54	40.91	2
Total	2015–2018	48	1.46–99.60	33.39	10

. Ten samples from all the sampling sites were above the legal limit concentration of 50 mg/L for groundwater and, in all three wells, breaches of the legal limit were found. All the wells analysed were protected from any contamination and there was no treatment/filtration before sampling.

The PIA well collects water from the first aquifer, not confined, because of the shallow depth of the same, combined with the medium–high permeability of the overlying geological layers, consisting mainly of sand or gravel. The three samples collected in 2015 were found above the limit of 50 mg/L, while in the following years, a sharp decrease in nitrate concentrations was observed, with average values below 10 mg/L mainly due to changes in production vocation through less N-intensive crops (from 310 kg N/ha of maise to 180 kg N/ha of wheat to 170 kg N/ha of tomato) [32]. The PON collects water from the first aquifer, not confined, because of the shallow depth of the same, combined with the average permeability of the overlying geological strata. This well showed critical values oscillating from year to year, probably due to crop rotation combined with changes in the intensity and type of fertilisation. The GOS well collects water from the first aquifer, not confined, because of the low depth of the same, combined with the high permeability of the geological layers above, consisting mainly of gravel. The well is situated in an area with a modest presence of farms in which maise, wheat, and tomato are mainly cultivated. However, 21% of the samples collected in this well were found above the limit of 50 mg/L.

The results concerning NO₃⁻ annual average concentration (mg/L), WPL, Q (m³/ha/yr), and GWF (m³/ha/yr) are presented in

Table 7. GWF and WPL values relative to nitrates annual average concentrations.

Piezometers	Average Annual Concentration (mg/L)	WPL	Q (m3/ha/yr)	GWF (m3/ha/yr)	GWFER 1 (m3/ha/yr)
PIA/2015	46.65	0.93	5213.87	4864.19	4838.53
PIA/2016	9.65	0.19	5066.18	977.52	3344.82
PIA/2017	7.69	0.15	4519.46	694.73	2698.52
PIA/2018	9.83	0.20	6758.17	1329.11	3840.92
PON/2015	58.76	1.18	5215.60	6129.37	3275.61
PON/2016	41.14	0.82	5358.11	4408.22	2886.94
PON/2017	41.35	0.83	4626.56	3826.16	2365.53
PON/2018	33.60	0.67	6733.12	4524.66	3245.34
GOS/2015	46.22	0.92	5431.52	5020.90	3226.40
GOS/2016	39.57	0.79	5021.27	3973.83	2882.02
GOS/2017	42.81	0.86	4411.49	3777.42	2393.93
GOS/2018	39.48	0.79	6803.08	5372.17	3561.44

¹ GWF_ER: GWF value extrapolated from maps of Figure 2.

. The WPL and GWF are calculated at a yearly level for each site. A WPL > 1 occurred in one well, PON/2015, indicating that the water quality standards have been violated. All the other samples collected presented a WPL level below 1, underlining a compliance with the quality standards (see Section 2.3, “Grey Water Footprint and Water Pollution Level”). When comparing GWF estimations from well concentrations with GWF values extrapolated from the regional interpolated map (GWF_ER in Table 7), an average Pearson correlation coefficient of 0.87 underlines a good positive correlation, highlighting the predictive potential of the geospatial interpolations obtained. This result confirms what is reported in the Navarra study [26], emphasising that the GWF and WPL indicators could be useful to roughly predict the N pollution in the absence of local information.

Because the primary crop production in the Piacenza area under study is mainly composed of maise, tomato, and wheat, it is possible to predict the worst-case scenario of agricultural N load to the total load. In fact, maise crops have a high N demand and, consequently, a high N dosage [32]. Because of the possible rotation that occurs in the area for the crops under investigation, we decided to evaluate the agricultural dosage as a weighted average, considering the average provincial cultivated area for each crop [33]. As a result, an average N dosage for the Piacenza area of 209.39 kg/ha/yr was obtained. Considering the guidelines, we can estimate the actual pollutant fraction that leaches into groundwater. Adopting N table guidelines and considering the different factors that influence the leaching, we estimated an average alpha value of 10.4% (in a range of 9.6–11.2%), perfectly in line with the value suggested in the absence of information (10%). With that in mind, we estimated the actual N leaching, obtaining 21.83 kg/ha/yr as the average L_agr. In

Table 8. Total N load (L_tot), estimated agricultural N load (L_agr), and N surplus (L_sur) in the three sites monitored from 2015 to 2018 with model parameters adopted for the calculation.

Piezometers	α	DOSE (kg N/ha/yr)	Q (m3/ha/yr)	Lagr (kg N/ha/yr)	Ltot (kg N/ha/yr)	Lsur (kg N/ha/yr)
PIA/2015	0.112	213.49	5213.87	23.87	54.92	31.05
PIA/2016	0.100	210.64	5066.18	21.05	11.04	0.00
PIA/2017	0.100	205.68	4519.46	20.56	7.84	0.00
PIA/2018	0.112	207.75	6758.17	23.23	15.01	0.00
PON/2015	0.108	213.49	5215.60	23.01	69.20	46.19
PON/2016	0.108	210.64	5358.11	22.71	49.77	27.06
PON/2017	0.096	205.68	4626.56	19.73	43.20	23.47
PON/2018	0.108	207.75	6733.12	22.40	51.08	28.69
GOS/2015	0.108	213.49	5431.52	23.01	56.69	33.67
GOS/2016	0.096	210.64	5021.27	20.20	44.87	24.66
GOS/2017	0.096	205.68	4411.49	19.73	42.65	22.92
GOS/2018	0.108	207.75	6803.08	22.40	60.65	38.26
Average	0.104	209.39	5429.87	21.83	42.24	22.99

, all the data used in the model are reported to predict the contribution of the N load in surplus to the agricultural load.

The results show that most of the wells examined exceed the worst agricultural case scenario, confirming a possible non-agricultural contribution to N groundwater pollution. This result could be explained by considering the location of the sites relatively close to the main city of Piacenza. An English review of 2005 [51] reports how the non-agricultural N contribution to Nottingham groundwater is mainly due to solid and wastewater disposal, with an estimation of 21 kg/ha/yr (considering an average of 3–17 mg/L groundwater N concentration). For the Piacenza area, all the sites present a surplus that ranges from 0.00 to 46.19 kg/ha/yr (with an average of 22.99 kg/ha/yr). The obtained results confirmed that the leaching of agricultural N showed large variations for different areas. The differences could be attributed to different soils, climates, and differences in cultivation practices. Moreover, a precise and accurate estimation of the non-agricultural contribution to groundwater N load is extremely difficult because, as reported by Lerner [52], too many contributions must be considered and several estimations are needed.

However, the data obtained confirmed that polluted wells were all characterised by a low or high plain groundwater (not confined) and a high inherent vulnerability of groundwater to pollution [53]. The vulnerability of the selected wells could be associated with the stratigraphy of the soil above the aquifer level [54]. The unsaturated zone has a texture characterised by high permeability, strong ventilation, and weak capacity to retain water. Moreover, a low level of organic matter content also decreases denitrification potential and thereby increases nitrate leaching [54] and the low depth reduces the N transit time, thus reducing the leaching effect and creating a potential self-depurating action of the unsaturated zone via precipitations [54,55].

4. Conclusions

N is a critical input in agricultural systems since it is required for crop development, yet excessive N supply may pollute groundwater [56]. In order to achieve the water sustainability goals of SDG 6.3, GWF and WPL are essential indicators that allow the monitoring of water quality: the GWF represents the volume of water to add in order to assimilate anthropogenic pollution, while the WPL is an index of the severity of the pollution entity. In the present study, the GWF and WPL were used as indicators in order to assess the nitrate pollution in groundwater and compare it with actual nitrate concentrations in groundwater aquifers in Emilia-Romagna. Specifically, GWF allowed us to obtain a regional map for N pollution levels and compare it to local wells chosen in the Piacenza district. The approach adopted for the development of regional maps was validated by selected wells samples analysis and could be applied to every region worldwide, demonstrating its strength and the potential for environmental assessment not only in terms of N but also in terms of other potential groundwater pollutants. Furthermore, a method for estimating non-agricultural groundwater N pollution was proposed to better understand the anthropogenic pressure on groundwater aquifers. The regional data collected from 2014 to 2020 show that approximately 50% of the analysed sites resulted in a total N load exceeding the estimated agricultural load to groundwater, with a high correlation between rainfall and N leached. The analysis conducted in three selected wells in the Piacenza district shows sporadic exceedances of the legal limit (>50 mg/L) and demonstrates the presence of anthropogenic pressures of various natures insisting on the surrounding area of the radius of 2 km and confirms a potential non-agricultural point or diffuse pollution source. Despite a registered increase in agricultural practice in 2020, including an increase in the agriculture N load, there was a decrease in the number of sites exceeding the N-estimated agricultural contribution into groundwater. The GWF and WPL calculated at the regional and local levels are suitable indicators of nutrient management. The calculated N surplus indicates both the variability in the site-specific N management, including site-specific conditions and the potential impact of other anthropogenic sources for which more investigations need to be performed in the future. The results from 2018 to 2020 show a decrease in the GWF and L_sur that cannot be explained by how there was lower precipitation and that higher nitrate levels were observed in most of the cases in urban areas with the highest concentrations in 2020. Is the COVID-19 pandemic in some way responsible for this result? Now, it is hard to know.