Environmental implications of salinity and flooding induced changes in porewater nitrogen and phosphorous dynamics

ABSTRACT

This study aimed to investigate the effects of fresh water and seawater flooding on phosphorous (P) and nitrogen (N) dynamics in two calcareous soils (Biscayne and Krome) of South Florida using a 28-day greenhouse flooding experiment. Porewater samples were collected at three depths over four periods (1, 7, 14, and 28 days) of flooding and analyzed for selected water quality parameters. Results showed that seawater-flooded Biscayne soil had a 4.4 fold higher porewater soluble reactive phosphorus (SRP) concentration  than freshwater. Additionally, porewater SRP concentration and total phosphorus (TP) in Biscayne soil increased with prolonged flooding duration and column depth. Furthermore, both freshwater and seawater flooding increased porewater NH₄⁺ - N concentration; however, the increase was more pronounced in seawater flooded samples. NH₄⁺ - N concentration also increased with flooding duration. Overall, the study highlights the risks of nutrient leaching from seawater flooded coastal soils that could contribute to the deterioration of surface and groundwater quality.

KEYWORDS:

• Climate change
• sea level rise
• coastal agriculture
• water quality
• nutrients

1. Introduction

Flooding from rising sea levels and saltwater intrusion increase salinity in coastal areas [1]. Global warming induced thermal expansion of seawater as well as the melting of glaciers and ice caps tend to accelerate the rate of sea level rise [2]. The prediction by the Intergovernmental Panel on Climate Change (IPCC) shows that global sea levels would rise to 87 cm in 2100 [3], but some studies have estimated that sea level rise could reach 100 cm [4]. Approximately 424 million hectares of soil are currently salinized [5].

Saltwater intrusion increases salinity and affects the hydraulic and chemical properties of soils in coastal regions, leading to changes in nutrient dynamics and biotic population [6,7], cycles of carbon (C), nitrogen (N), and phosphorus (P), as well as the dynamics of calcium (Ca²⁺), potassium (K⁺), magnesium (Mg²⁺), sodium (Na⁺)^, and sulfate (SO₄^2-) in soils, are all notably affected by saltwater intrusion. Cation exchange is one of the well-known phenomena affected during saltwater intrusion. The cations in the sediment and/or soil (such as Ca²⁺, K⁺, and Mg²⁺) could be potentially released into the soil solution due to an exchange with Na⁺ from saltwater [8]. All the processes in the cycle of N such as nitrification, denitrification, and dissimilatory nitrate reduction are affected by saltwater intrusion in one way or another. The reason for this is the shift in the soil’s aerobic-anaerobic condition (limiting oxygen availability) and the change in the dynamics of the soil microbial community. Furthermore, saltwater induced exchange of cations plays a key role, especially in the dynamics of NH₄⁺ - N [9]. When soil salinity exceeds 16.5 dS m⁻¹, the process of nitrification, which converts ammonium to nitrate under aerobic conditions is inhibited [10]. This inhibition is caused by sulfide ions in saltwater, which compute with ammonia for the active sites on the ammonia monooxygenase enzyme responsible for oxidizing ammonia [11]. This suppresses nitrate production, the primary form of N assimilated by plants. The inhibition of nitrification ultimately affects the process of denitrification due to the decline in the production of nitrate, which is the substrate for the process of denitrification and changing microbial community [12]. On the other hand, the concentration of NH₄⁺ could increase due to the process of cation exchange [13]. The released NH₄⁺ due to saltwater intrusion will probably wash into our freshwater systems and/or change into easily leachable nitrate. Additionally, seawater intrusion could release significant P to the soil solution. The increment in the concentration of PO₄^3- due to saltwater intrusion induced desorption from sediment is evident [14]. This makes soil P accessible for leaching to groundwater and/or washable by seawater flooding or runoff to water bodies. Based on a summary of Reddy et al. [15], there is a loading of 170 metric tons of P per year to Lake Okeechobee from agricultural areas of Everglade. Therefore, plants grown in seawater intruded soil will not only be affected by salt stress but also from the change in the nutrient status and chemistry of the soil [7]. Based on the study by Zhang et al. [16], after Hurricane Katrina, the concentration of nitrate and P in Biscayne Bay, Florida, increased by 5.2 and 2 folds, respectively. The increment of nitrate and soluble reactive phosphate was much higher at the water quality monitoring station, accounting for up to 7 and 10-fold higher, respectively. Thus, the effect of saltwater intrusion is not limited to soil and plants; it is extended to the water system where the flood flows back to water systems and intensifies the eutrophication problem. Furthermore, the released nutrients are prone to leaching toward the groundwater systems. Hence, the effect of saltwater intrusion is worse in aquifers like Biscayne, where the surface is made up of highly permeable limestone [17]. This, along with the unconfined nature of the Biscayne aquifer, creates a water table that reacts quickly for recharge, evapotranspiration, and pumping from supply wells. Additionally, a direct hydraulic link between the drainage canals and the aquifer allows quick water exchange [18,19]. Therefore, it is crucial to understand how saltwater intrusion affects nutrients like N and P in soils typical of the Biscayne aquifer to decide on proper management practices hindering nutrients like N and P from ending in the groundwater system 20,21]. Therefore, this study was conducted to examine the effects of soil flooding with fresh and seawater on P and N dynamics, pH, and electrical conductivity (EC) of calcareous soils and their implication to the environment.

2. Materials and methods

2.1. Study sites and soil collection

The experiment was conducted in a ventilated greenhouse system at the University of Florida, Tropical Research and Education Center (TREC), Homestead, FL, U.S.A.. The two common calcareous soils, Krome and Biscayne, found in South Florida, were used for the study. The Krome gravelly loam soil, with more than 50% gravel, was collected from TREC’s research farm. The marl loamy carbonatitic Biscayne soil was collected from a commercial farm in Homestead, FL. Both Krome and Biscayne soils were collected from agricultural fields that are mainly used to produce fall vegetables, such as snap beans and sweet corn. Krome soil is highly drained due to high gravel content, while Biscayne soil is known for its poor drainage. Soils were air-dried by spreading on a tarp inside a greenhouse for several days. Air-dried soils were then mixed thoroughly and sieved using a 2 mm mesh sieve. Soil samples were analyzed for selected parameters to establish baseline information (Table 1).

Table 1. Properties of water and soil used in the study.

Properties	Biscayne	Krome	Freshwater	Seawater^‡
pH	8.1 ± 0.0	8.4 ± 0.1	7.2	7.5
Organic matter* (g kg − 1)	62.2 ± 1.2	71.1 ± 1.1	-	-
EC (µS cm − 1)	457	131	524	56267
Ca	40911 ± 6963	16590 ± 541	39.7	313
P	32.1 ± 8.18	107 ± 4.09	0.004	0.008
TKN	2298 ± 42.3	932 ± 19.2	-	-
NH4+ - N	1.61 ± 0.23	1.29 ± 0.05	0.015	0.005
NO3+ - N	79.8 ± 2.90	9.90 ± 0.30	0.44	0.038

The concentration units for nutrients are mg kg⁻¹ for soils and mg L⁻¹ for freshwater and seawater.

*.Krome soil contains more than 50% gravel, which was not included in calculating the organic matter content since the organic matter was measured after sieving the soil with a 2 mm sieve. Organic matter was measured using Loss on ignition method.

‡.TKN: Total Kjeldahl Nitrogen (Total organic nitrogen N + ammonium).

2.2. Experimental setup

Six polyvinyl chloride (PVC) columns (61 cm long and 15.24 cm diameter) were prepared (Figure 1). The bottom end of each column was closed using a PVC end cap. End caps were drilled to create 9 holes (1 cm diameter) to allow the flow of water to and out of the columns. An ultra fine nylon net was put inside the columns before packing to ensure soil is retained inside the columns. In addition, three holes (3 cm diameter) were made at three depths (15 cm increment). Suction lysimeters connected to a tygon tube were installed horizontally at the three depths. Glue was applied to fully seal holes after installing the lysimeters. Similar holes were created for the installation of soil moisture, temperature, and EC sensors on the opposite side of the lysimeter holes. Finally, columns were packed (1 g cm⁻³ bulk density), leaving a 15 cm head space to avoid overflowing of water.

Figure 1. Design depicting the experiment while soil columns are inside IBC tanks filled with freshwater and seawater.

2.3. Flooding experiment and pore water sampling

Two 1040 L top open intermediate bulk containers (IBC) were used for flooding the soil columns. A two-tier wooden table with holes on each tier was placed inside each tank. The hole on the top tier was big enough to slide the columns, while the hole in the bottom tier was smaller than the column diameter. The table was used to keep the columns straight and suspended to allow water to flow through the bottom holes. One tank was filled with freshwater from groundwater wells at TREC. The second container was filled with seawater collected from Key Largo, Florida. Before starting the experiment, the salinity of seawater was checked to ensure that the EC is at least 50 dS m⁻¹. In addition, freshwater and seawater samples were collected before the start of the experiment for lab analysis. Three columns filled with soil and carefully placed inside each tank filled with water (Figure 1). To ensure fully flooded conditions, the water level in each tank kept about 3 cm above the soil surface inside the columns. Air stones were placed in each tank to avoid algal growth and light effect, and tanks were covered using black plastic tarps. Porewater samples were collected using 50 mL surgical syringes connected to Tygon tubes. Before each sampling, a vacuum was created by pulling the plunger until the syringes were filled with porewater. From each column, three replicates of porewater samples were collected from three depths after 1, 7, 14, and 28 days of flooding. Porewater samples were then kept in a refrigerator at 4°C.

Concentrations of soluble reactive P and Ca in porewater were measured directly using an inductively coupled plasma–optical emission spectrometer (ICP – OES) (Thermo Jarrell Ash ICAP 61E, Franklin, MA), and ICP measured concentration in soil after extraction of soil with Mehlich-3 extractant (1:10 w/v) following the standard procedures by the Environmental Protection Agency (EPA). TKN, NO₃⁻ - N, and NH₄⁺ - N in soil porewater were directly measured using a segmented flow analyzer (AAII Technicon, SEAL Analytical, UK) following EPA standard procedure Method 353.2 and EPA Method 350, respectively. The concentration of NO₃⁻ - N, and NH₄⁺ - N in soil were measured by a segmented flow analyzer after extraction of soil with KCl (1:10 w/v). Additionally, soil pH was measured by EPA Method 150.1 and EC by EPA Method 120.1 after extraction with deionized water using Accumet XL50 meter (Fisher Scientific, MA, U.S.A.). Finally, the loss on ignition method was used for the estimation of soil organic matter (OM).

2.4. Data analysis

Data was checked for outliers and outlier values were replaced by the mean from the same treatment group and sampling date. Data analysis and figures were done using the R programming language and SAS software. Since repeated porewater samples were collected, which violates the independence assumption of parametric tests, a generalized linear mixed (GLIMMIX) model was applied to assess the effect of water source, soil type, sampling depth, sampling event, and their interaction on the concentration of P, TP (Total Phosphorus), NH₄⁺ - N, NO₃⁻, EC and pH of soil porewater.

3. Results and discussion

3.1. Effect of seawater on phosphorus concentration

Flooding of soils with seawater significantly increased the concentration of SRP in porewater (Figure 2A). Based on the principal component analysis of variance, the concentration of P was clearly affected by the water type, soil, and sampling period, where they represented 37.3, 25.4, and 24.1% of the variation, respectively (Figure 2B). However, the effect was more pronounced and evident in Biscayne soil compared to Krome soil (Figure 2C). This agrees with the GLIMMIX analysis of variance, where there was a significant effect (p < 0.0000) of soil type, water source, depth, sampling event, and their interaction (Table 2).

Figure 2. (a) the effect of freshwater and seawater flooding on the soluble reactive P concentration in porewater (b) principal component analysis (PCA) of soluble reactive P concentration in porewater samples collected from two soils flooded by seawater and freshwater at 4 sampling periods (c) trends of soluble reactive P concentration in porewater in sampling periods. Day-0: baseline P before the start of the flooding.

Table 2. A generalized linear mixed (GLIMMIX) model analysis of variance. Effect of soil, water, depth, sampling event and their interaction on concentration of P, TP, NO₃⁻ - N, NH₄⁺ - N, and EC of soil porewater.

Effect	P	TP	NH4^+ - N	NO3^− - N	EC
	FValue
Soil	109***	145 ***	2307***	95.4 ***	94.4***
Water	9.7***	72.3 ***	826***	7.54 **	32543***
Soil*Water	84.6***	79.3***	90.0***	7.42 **	239***
Depth	46.1***	68.5***	37.6***	32.8 ***	2.04
Soil*Depth	3.1***	4.6***	26.7***	32.9 ***	18.2***
Water*Depth	48.5***	33.8***	18.6***	0.48	6.30**
Soil*Water*Depth	39.5***	28.8***	13.1***	0.49	1.62
SamplingEvent	44.0***	4.7***	248***	39.1 ***	8.76***
Soil*SamplingEvent	36.2***	57.6***	78.0***	38.9 ***	62.3***
Water*SamplingEvent	5.9***	92.3***	15.0***	5.27 **	8.86***
Soil*Water*SamplingEvent	39.4***	51.6***	7.66***	5.22 **	47.4***
Depth*SamplingEvent	2.9***	24.2***	0.56	14.8 ***	0.47
Soil*Depth*SamplingEvent	21.1***	26.4***	0.55	14.8 ***	2.10
Water*Depth*SamplingEvent	14.1***	22.1***	0.95	2.04	0.64
Soil*Water*Dept*SamplingEvent	14.1***	17.7***	0.85	2.04	0.67

Where ***: significant at p < 0.000, **at p < 0.00.

On the 14^th and 28^th day of sampling, Biscayne soil flooded with seawater had a significantly higher average SRP concentration than freshwater, specifically in the middle and bottom sections of the soil column. In contrast, there was little or insignificant difference in SRP concentration at different depths within Krome soil (Figure 2C). In Biscayne soil, at the middle and bottom section of the column, SRP concentration increased with sampling event (Figure 2C). At the 28^th date of sampling, the concentration of SRP at the middle and bottom of the Biscayne soil column flooded with seawater was up to four times higher than that of the freshwater flooded columns of the same depth. Moreover, despite Biscayne soil initially containing lower available SRP than Krome soil (Table 1), there was a higher concentration of SRP in the middle and bottom column than Krome soil when flooded with seawater. The total P (TP) concentration followed a similar trend to that of SRP, where soil, water, depth, sampling events, and their interactions significantly impacted the TP concentration (Table 2).

Seawater flooded Biscayne soil had a significantly higher concentration of TP than freshwater flooding. However, in Krome soil, there was negligible insignificant difference. Furthermore, there was an increase in TP with sampling depth and sampling event, especially at the middle and bottom of the column. Seawater flooding could increase the concentration of P in soil due to one of the following mechanisms or their combination (Figure 3). First, the activity of P-monoesterase from oceanic microbes, which is responsible for the release of phosphate, could release P; second, P binding Fe (III) oxide could be reduced and release P; third, the anionic exchange of phosphate by seawater originated HCO₃⁻ and B(OH)₄⁻, fourth: the metabolism of sulfate could potentially contributes for the mineralization of PO₄^3- [22–24]. Williams et al. [24] reported an increment of PO₄^3- in saline soils by seawater intrusion due to SO₄^2- reduction. Based on their findings, the PO₄^3- increment was less evident in low saline soil. This agrees with our finding where the increment of P was more prominent in Biscayne soil (EC = 457 µS cm⁻¹) which has almost 3.5 times higher EC level than Krome soil (131 µS cm⁻¹).

Figure 3. Schematic diagram of P release mechanisms in seawater flooded soils.

A high concentration of P in seawater flooded marsh soils was also reported [1]. The presence of a higher amount of CaCO₃ in marine soil is evident, and this causes the binding of P with the surplus Ca available [25]. Biscayne soil used in this study is also a marine water affected soil with a higher concentration of Ca (40.9 g kg⁻¹), which is formed by the flooding of coastal land for months and subsequent drying during the winter dry season. Therefore, the seawater originated anions could release the Ca bound P.

Phosphorus availability can also be impacted by microbial metabolism. Fe oxides may be reduced and combined with sulfides from seawater to create Fe sulfides (FeSx) in anaerobic soil states, making them incapable of binding PO₄^3- in porewater and therefore releasing P to porewater [23]. This phenomenon is not true in the case of freshwater flooding as the freshwater is low in SO₄⁻ content [26]. Again, the increment of P from Biscayne soil could also be highly linked to the microbial utilization of C from carbonates tied with P. So that anaerobic microbes could convert the P containing carbonate to organic carbon and in the process releases P [22]. Additionally, the release of phosphate in flooded soils was reported due to the reduction and dissolution of Fe(hydr)oxides [27]. However, in our study, this couldn’t be true as the increment of P was only in Biscayne soil flooded by seawater. The release of P in soils porewater flooded by seawater shows the potential of P export to groundwater causing water quality deterioration and/or nearby surface water system leading to eutrophication. Based on the study of Steinmuller and Chambers [28], there was a potential of P export to the surrounding water system with a potential of 2 to 100% of reactive soluble P being exported to the water system depending on the texture and carbonate content of the specific soil under investigation. Overall, our findings highlight the significant release of P associated with seawater flooding, indicating that coastal soils are more prone to P release compared to inland young gravely calcareous soils.

3.2. Effect of seawater on ammonium nitrogen concentration

Flooding of soil with seawater significantly increased NH₄⁺ - N concentration in soil porewater (Figure 4A), and it was significantly affected by soil type, water source, sampling event, depth, and their interaction (p < 0.000). Among these factors, soil type, water source, and sampling event had the greatest effect (Table 2). Based on the principal component analysis of variance, the concentration of NH₄⁺ - N was clearly affected by the water type, soil, and sampling period, where they represented 47.7, 25.5 and 24.1% of the variation, respectively (Figure 4B). Regardless of soil type, flooding of soils by seawater resulted in the increment of NH₄⁺ - N concentration, where there was much higher NH₄⁺ - N concentration in seawater flooded soils at all depth and sampling events (Figure 4C). Under flooded conditions, the anaerobic microbes could mineralize organic nitrogen to ammonium nitrogen through the process of ammonification [29]. However, this process is slow since the ammonification process is performed by a limited number and/or type of microflora not requiring oxygen for their respiration. Consequently, the initially low NH₄⁺ - N content on the first day of sampling gradually increases towards the 28^th day of sampling (Figure 4C) due to the slow rate of ammonification under anaerobic conditions and its gradual accumulation over time. After the ammonification, the released NH₄⁺ - N can be accumulated unless an aerobic condition is created, which converts NH₄⁺ - N to NO₃⁻ - N. Release of NH₄⁺ - N was higher in seawater flooding due to the coupling of organic nitrogen mineralization with the release of soil exchangeable NH₄⁺ - N by seawater originated cations. The contribution of cation exchange for the release of NH₄⁺ - N agrees with the report of Ardón et al. [30], where they proposed the increment of NH₄⁺ - N due to the exchange of NH₄⁺ - N in exchange with seawater originated cations (Na⁺, Ca²⁺, Mg²⁺, K⁺). As evidence, the concentration of NH₄⁺ - N in seawater flooded columns of Biscayne soil was negatively linked (r > 0.77) with the concentration of Mg²⁺, and in Krome soil negatively correlated with Mg²⁺ (r > 0.59) and K⁺ (r > 0.57).

Figure 4. (a) the effect of freshwater and seawater flooding on the soluble P concentration in porewater, (b) principal component analysis (PCA) of NH4+ - N concentration in porewater samples collected from two soils flooded by seawater and freshwater at 4 sampling periods, (c) trends of NH4+ - N concentration in porewater in sampling periods. Day-0: baseline NH4+ - N before the start of the flooding.

The decline in the concentration of Mg²⁺ and K⁺ opposite to the increment of NH₄⁺ - N clearly illustrates the contribution of cation exchange. However, this was not true in the case of freshwater flooding where the concentration of NH₄⁺ - N is positively correlated with both K⁺ and Mg²⁺. The desorption of NH₄⁺ - N from seawater intrusion is also reported by the study of Baldwin et al. [31] and Steinmuller and Chambers [28]. Moreover, the higher concentration of NH₄⁺ - N in Biscayne soil (up to 113 mg L⁻¹) as compared to Krome soil (up to 55.7 mg L⁻¹) was also evident (Figures 4A, B and C). This is due to the higher initial content of the total Kjeldahl N (total organic nitrogen and ammonium) in Biscayne (TKN = 2298 mg kg⁻¹) than Krome (TKN = 932 mg kg⁻¹), which is potentially mineralized by the anaerobic microbes. In addition to that, the highest concentration of NH₄⁺ - N in Biscayne soil could be due to the higher initial NO₃⁻ - N (79.8 mg kg⁻¹) content than Krome (9.9 mg kg⁻¹). Additionally, the surplus content of NO₃⁻ - N could convert into NH₄⁺ - N by the process of dissimilatory nitrate reduction, which is a microbial conversion of NO₃⁻ - N into NH₄⁺ - N in anaerobic condition [32–34]. Additionally, it is crucial to keep in mind that the organically bound N may eventually be released in the form of NH₄⁺ by the mineralization of OM triggered by the presence of sulfate (SO₄^2-) in saline water. The existence of SO₄^2- and OM in the soil could also facilitate dissimilatory nitrate reduction as the major pathway producing NH₄⁺ [35]. Thus, an increment of NH₄⁺ - N after the flooding of soils by seawater has wider implications. Seawater flooding of agricultural land that has a high content of absorbed or exchangeable NH₄⁺ could potentially release NH₄⁺ - N and the subsequent drying of the field could facilitate the conversion of NH₄⁺ - N to NO₃⁻ - N through nitrification which is easily leachable to groundwater system and/or washing of NO₃⁻ - N to the surface water system. This event affects agricultural production in coastal land, groundwater quality deterioration, and leads to intensified eutrophication [30].

3.3. Effect of seawater on nitrate concentration

NO₃⁻ - N concentration was significantly higher in seawater flooded soils samples than freshwater flooded samples, and the effect was more pronounced only in Biscayne soil. As that of NH₄⁺ - N, soil, water, depth, sampling event, and their interaction significantly affected NO₃⁻ - N concentration (p < 0.001). The highest effect was from soil, sampling event and the interaction between soil and sampling event, respectively (Table 2). Biscayne soil had a higher content of NO₃⁻ - N than Krome soil in both fresh and seawater at all depths (Data for Krome soil not shown). Additionally, more evidently in Biscayne soil NO₃⁻ - N concentration was higher in seawater samples than freshwater at all sampling periods (Figure 5).

Figure 5. The effect of seawater and freshwater flooding for 28 days on the concentration of NO₃⁻ - N in soil porewater.

In Biscayne soil, the concentration of NO₃⁻ on the 7^th, 14^th, and 28^th day of sampling was significantly lower than on the 1^st day of sampling. However, there was inconsistent effect in Krome soil. This could be due to the low initial NO₃⁻ content of Krome which was 9.9 mg kg⁻¹ compared to the 79.8 mg kg⁻¹ in Biscayne soil. Therefore, in Biscayne soil flooding by seawater especially during the 1^st day of the flooding could have resulted in the release of soil bund NO₃⁻ - N, then the decline of NO₃⁻ - N with the length of flooding could have resulted due to denitrification. Flooding of soil for several days could generally result in the decline of NO₃⁻ - N. Based on the study of Alaoui-Sossé et al. [36], flooding of soil with fresh deionized water for 34 days resulted in 7 and 50 times lower content of NO₃⁻ - N than the initial content at the top and bottom of pots used in their study, respectively. The first reason for the decline of NO₃⁻ - N concentration in both fresh and seawater flooding could be the denitrification of nitrate, which is responsible for the conversion of NO₃⁻ to N₂ or N₂O [37]. In anaerobic saturated soil conditions, the oxidization of soil organic carbon utilizes NO₃⁻ as an electron acceptor and releases N₂ or N₂O. The increment in the emission of N₂O due to flooding is reported by Jin et al. [38], which is again ultimately tied with the denitrification of NO₃⁻. The release of N₂O from flooded soils with a considerable level of NO₃⁻ - N has a serious environmental implications as N₂O has almost 265 times higher global warming potential than CO₂ [39]. The coupling of soil flooding with nitrogen input could exacerbate the release/emission of N₂O by 65–94% [40]. The second probable reason is the conversion of NO₃⁻ to NH₄⁺ by the process called dissimilatory nitrate reduction (DNR). The occurrence of DNR in anaerobic saline conditions is evident [33,41].

3.4. Effect of seawater on soil EC and pH

Porewater collected from seawater flooded soils has significantly higher EC than fresh water (Figure 6A). EC of soils porewater was significantly affected by water, soil, sampling event, and the interaction between them (Table 2). Based on the principal component analysis of variance, the concentration of NH₄⁺ - N was clearly affected by the water type, soil, and sampling period, where they represented 49.8, 25.0 and 24.7% of the variation, respectively (Figure 6B).

Figure 6. (a) the effect of fresh and seawater flooding on soil porewater EC level, (b) principal component analysis (PCA) of EC level in porewater collected from two soils flooded by seawater and freshwater at 4 sampling periods.

This indicates the variability or dependability of saltwater effect on the initial property of soils and sampling periods. Thus, an increment of EC due to the seawater is evident [42–45]. The increment is simply due to the high EC level of seawater used for flooding, where the dissolved salts from the seawater resulted in a higher electrical conductivity of soil porewater. Our findings also demonstrated a disparity in the effects of waters between the two soils, where a significantly higher average EC was observed in Krome as compared to Biscayne soil when flooded by seawater, whereas Biscayne soil had a significantly higher EC than Krome soil when flooded with freshwater (Figure 6A).

The EC of porewater taken from Biscayne soil flooded with fresh water is simply higher than Krome soil because of the higher EC of Biscayne soil used in this study (Table 1). Whereas the higher EC in Krome soil porewater than in Biscayne soil when flooded with saltwater is linked with the behavior of Ca in soil porewater. Ca²⁺ behaved opposite to other ions, especially Mg²⁺ and K⁺. The concentration of both Mg²⁺ and K⁺ was higher in Biscayne soil compared to Krome soil when flooded by both fresh and seawater. However, the concentration of Ca²⁺ was higher in Biscayne than in Krome soil when flooded by freshwater, whereas the opposite was true when flooded with seawater. The variation in EC across soils when they are flooded with fresh, and seawater is therefore caused by the release and/or complexation nature of Ca²⁺ in soils having different properties. This indicates the variability or dependability of seawater’s effect on the initial property of soils, sampling depth, and periods. For instance, the EC of porewater obtained at the top of the Biscayne soil column flooded with freshwater was greater than the EC of porewater collected at the bottom, whereas this was not true in the case of porewater taken from Krome soil.

Flooding both soils with seawater decreased porewater pH at the 1^st day of sampling and then it starts to increase toward the initial porewater pH (Figure 7). Whereas freshwater showed a general increment of porewater pH in Krome soil and Biscayne soil it showed a trend of increment at the first two sampling events and then decline. Generally speaking, salty water could result in a decline of soil pH due to the release of H⁺ into soil solution due to exchange with saltwater originated Na⁺ [46]. On the other hand, both increment and decline of soil pH have been reported in some other studies investigating low pH soil. Wong et al. [45] also reported a decline of soil porewater pH and on the contrary Chambers et al. [42] reported an increment of pH after seawater flooding of soils. The return of soil pH to the initial soil pH agrees with the finding of Pan et al. [14] finding, where the pH of soil becomes uniform after the salinization development of soils with higher pH (8.47–8.73). Moreover, the reason for returning pH to the initial pH level could be due to the buffering capacity of the soils used in this study. Both soils used in this study had higher pH values (Krome = 8.37 and Biscayne 8.13). The change in soil pH could be negligible in soils with Ca content higher than 25 g kg⁻¹ [47], whereas our soils are characterized by much more Ca content of 41 g kg⁻¹ in Krome and 166 g kg⁻¹ in Biscayne soil. Therefore, our soil could have resisted pH changes after the initial slight decline of pH.

Figure 7. The effect of freshwater and seawater flooding for 28 days on the pH of soil porewater.

4. Conclusion

Coastal agriculture and water systems are facing a significant threat from the increasing salinity due to flooding and saltwater intrusion caused by sea level rise driven by global warming. The aim of this study was to investigate the effect of seawater on porewater phosphorous and nitrogen dynamics. Seawater flooding of soils increased the concentration of soluble and total P in porewater for agricultural soils. Moreover, the magnitude and degree of change were affected by soil, water, depth, sampling event, and their interaction. However, the release of P was more pronounced in seawater flooded Biscayne soil. Similarly, the dynamics of NH₄⁺ - N concentration was affected by water, soil, depth, sampling event, and their interaction. On the 28th day, NH₄⁺ - N concentration in samples flooded with seawater and freshwater was 2.1 and 3.7 times for Biscayne soil and 3.9 and 13.3 times higher for Krome soil compared to the 1^st day samples, respectively. However, while the effect of seawater flooding on NO₃⁻ - N concentration was more pronounced for Biscayne soil, it decreased as the flooding period was extended. Biscayne soil exhibits a higher content of NO₃⁻ - N compared to Krome soil. Consequently, flooding with seawater may have induced denitrification of NO₃⁻ - N. Overall, our finding suggests that soil flooding with seawater can lead to the release of P and NH₄⁺ - N, posing a risk of leaching into groundwater and agricultural runoff. This, in turn, could lead to reduced agricultural productivity and deteriorating water quality in coastal areas.

Author contributions

Haimanote K. Bayabil: Conceptualization and Research design. Haimanote K. Bayabil and Niguss Solomon Hailegnaw: Methodology. Niguss Solomon Hailegnaw: Writing- Original draft preparation. Haimanote K. Bayabil and Yuncong C. Li: Revision. Haimanote K. Bayabil: Supervision.

Acknowledgments

This publication is based on research funded by the United States Department of Agriculture’s National Institute of Food and Agriculture under award number 2020-67019-31163 . Any opinions, results, conclusions, or recommendations contained in this publication are solely those of the authors and do not necessarily reflect the views of the USDA. The authors would like to express their gratitude to Christian Bartell and Syed Shaham for their help with the experiment and data collection . We would also like to thank Dr. Edzard van Santen for his help with statistical analysis.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Institute of Food and Agriculture [2020-67019-31163].