Changes,of,water,chemistry,from,rainfall,to,stream,flow,in,Obagbile,Catchment,,Southwest,Nigeria

Mojisola Hannah OMOGBEHIN, Isaac Ayo OLUWATIMILEHIN＊

Department of Geography, Obafemi Awolowo University, Ile Ife, 220101, Nigeria

Keywords：Water chemistry Rainwater Overland flow Soil water Obagbile Catchment Principal component analysis

ABSTRACT Water chemistry changes when it flows through different pathways.This study aims to characterize the differences of water (including rainwater, overland flow, soil water, groundwater, and stream water) chemistry of five kinds of water in Obagbile Catchment in Southwest Nigeria, determine the changes in water chemistry that occur as the water moves from one pathway to another, and identify the factors responsible for the water chemistry changes.To do these,we collected 50 water samples from 10 heavy storms that received equal to or more than 10 mm of rain within an hour to test the changes of water chemical properties across various pathways in this study.The results show that overland flow had the highest pH and electrical conductivity (EC) and rainwater had the lowest values of the two parameters.Ca2+, Mg2+, Cl¯, and HCOO¯ were found to have their highest concentrations in stream water; meanwhile, NO3–, NH4+,and SO42– were found to have almost the same low concentrations in all the water samples.K+ was only dominant in stream water; while dissolved organic carbon (DOC) was lowest in rainwater, same in overland flow, soil water, and groundwater samples, and highest in stream flow.Principal component analysis (PCA) showed that for all the water samples from different pathways,two factors mainly accounted for the total variances.The two factors were related to the crustal and anthropogenic sources in rainwater, suggesting that the high loadings of major cations (e.g., Ca2+ and Mg2+) in rainwater samples are soil-derived.The PCA for the overland flow and soil water showed strong correlations among pH, EC, and the concentrations of Na+, Mg2+, HCOO−,and CH3COO−, while the high loadings of all the parameters and the strong correlations among each other were evident in the stream water.In conclusion,the chemical constituents found in water are also the components of pathways through which the water flows.The major factors responsible for the change in the chemical properties of water in Obagbile Catchment are weathering and anthropogenic activities.

Different hydro-chemical characteristics are often associated with the water flowing along different pathways.For instance, the overland flow might be relatively high in organic carbon found in shallow soil layers, but low in mineral content, due to the relatively rapid travel time to the stream (Pi et al., 2021).The chemical composition of deep groundwater usually reflects a higher degree of mineralization, due to the longer contact time with bedrock, alayer of lower organic content (Ćuk et al., 2020; Borchardt et al., 2022).Thus, the length of the contact time of water with the soil or rock medium is a factor that determines the signatures of various water types (Mgbenu and Egbueri, 2019).

Meanwhile, rainwater gets its chemistry based on the nature of aerosols that form the hygroscopic nuclei upon which condensation occurs.One kind of rainwater, whose elemental composition is dominated by Al, Si, Fe, and Ca,would suggest that its aerosol has a high degree of dust and a limited marine or anthropogenic contribution (Ventura et al., 2021).One kind of rainwater with high nitrogen and sulfur compositions would indicate atmospheric pollution and emissions from the exhaust of cars and various industrial processes.Another kind of rainwater, whose chemical composition is Na+and Cl−dominated, would imply that the aerosol is of a maritime origin (Qiu and Felix, 2021).

All kinds of water from different pathways have their respective distinct chemistry imbued by the pathways(Eludoyin and Ijisesan, 2020).However, as water moves from one source area (e.g., atmosphere) to another through various pathways, changes begin to occur in its chemistry because the chemistry of water begins to equilibrate with that of the pathway and the receiving environment.As such, the rainwater chemistry changes when it moves from the atmosphere to the soil surface to become overland flow.The rainwater chemistry also changes as it infiltrates into the soil to become soil water.Further changes occur with percolation as the infiltrating water joins the groundwater body.As such, different kinds of water from different pathways will have a modified distinct chemistry after mixing with water from other passages through different pathways (Durgadoo et al., 2017,Borchardt et al., 2022).Therefore, the water chemistry at the ultimate recipient source area will be a comprehensive reflection of the chemistry of the different constituent water and their flow pathways (Sterte et al., 2021).Rainwater,overland flow, soil water, and groundwater mix together in river channel to become stream flow.The magnitude of the contribution from each of these pathways will determine the chemistry of stream flow.

Various studies have shown that all kinds of water have their own chemistry, which can be changed or altered as water moves from one storage area to another (Harvey and Gooseff, 2015; Leibowitz et al., 2018; Allan et al., 2020;Xenopoulos et al., 2021).Rainwater has its own constituents including various substances that fall from atmosphere as wet deposition during rainfall events (Tomasi and Lupi, 2017; Mathieu et al., 2018; Bergmann et al., 2019).However, these previous studies have been conducted in a temperate environment.Hence, it should be of interest to know not only the differences in the chemistry of rainwater, overland flow, soil water, groundwater, and stream flow,but also the changes that occur as the water moves from one pathway to another.This study attempts to evaluate and investigate the dynamism of the water chemistry in a second-order catchment in Ile Ife, Southwest Nigeria.

2.1.Study area

The study area is the Obagbile Catchment (07°26′–7°33′N, 04°30′–4°35′E) that is situated in Obafemi Awolowo University Research Farm, Ile Ife, Osun State, at a mean altitude of roughly 300 (±50) m a.s.l.(Fig.1).The lake that constitutes the study area is a man-made lake formed by the inundation of the dam constructed down the confluence of the Elerin and Omifunfun streams flowing from a relatively higher plain through a man-made forest.The lake is trapezoidal in shape, located in the northern part of the Obafemi Awolowo University campus, and occupies an area of 44,515.42 m2.The catchment is underlain by granite-gneisses of the Precambrian basement complex.It is usually flashy as with the quick rise and fall of water level following a storm event.The study area comprises a halocrystalline rock consisting of quartz, feldspar, biotite, and mica.These rocks are composed of Si,Al, Fe, Mg, Ca, Na, and K (Aduwo and Adeniyi, 2019).The soil found within the study area is an origin of granite-gneiss.

The climate of Obagbile Catchment falls under the Koppen’s tropical monsoon (Am) type characterized by the two pronounced seasons: wet and dry two seasons.The wet season covers the period between March and early November, while the dry season extends from late November to February next year.The rainfall during the wet season is bimodal, with the mean temperature ranging between 26°C and 28°C.The temperature in the dry season ranges from a mean nighttime of 21°C to a daytime mean of 30°C with a mean annual precipitation of approximately 1500 mm characterized by two rainfall peaks in July and September (Aduwo and Adeniyi, 2019).The upstream of the study area is cultivated to various tropical food (e.g., cassava, maize, yam, and cocoyam) and tree (e.g., oil palm, citrus, and plantain) crops on a local and mechanized scale.Pesticides, herbicides, and other agrochemicals are used on the farm.In the downstream of the study area, the water is used for different domestic purposes, such as cleaning and washing of clothes and utensils.

2.2.Methods

2.2.1.Sampling techniques

The data required for this study were primarily sourced.Direct rainfall was collected using bulk precipitation collectors which were placed on ground.The equipment consists of a funnel and a 4 L collector attached to the funnel.We installed piezometers using guar gum as a drilling method to make a 100 mm diameter and 1 m depth hole located above the elevation from which soil water was extracted with a suction pump to obtain soil water samples.Deep piezometers were used to sample the groundwater.Water samples of overland flow were collected along the ground surface by putting sampling bottles across the rill before rill gets into the channel.Stream water samples were taken by dipping the samplers into the stream.No stream gage was used, but a staff gauge wasconstructed in the study area to monitor the stream water level at different sampling periods.A water level recorder was also mounted at the edge of the stream to determine the water level.

Fig.1.Location of the Osun State in Nigeria (a); location of the Ile Ife City in Osun State (b); and overview of the Obagbile Catchment in the Ile Ife City (c).

2.2.2.Sampling method

In 2020, 10 heavy storm events with rainfall above 10 mm per hour were sampled, spreading from the beginning of the rainy season in April to October.The criteria used for water sample collection was that daytime rainfall exceeds 10 mm.Unfiltered water samples were collected during the storm events, and 50 water samples were collected during the course of this research.All samplers were rinsed with distilled water before the collection of each sample.The funnel was not washed or rinsed during the rainfall sample collection to maximize the effects of dry deposition.Samples were preserved at 4°C in a refrigerator prior to the analysis.Adequate arrangement was made to ensure the proper collection, preservation, and delivery of the samples to the laboratory of the International Institute of Tropical Agriculture in Ibadan, Nigeria.

All parameters were determined in the laboratory, except for the pH and electrical conductivity (EC) that were directly determined in the field by using a pH meter and an Aquapro HM Digital device (AP-1 Model from Water Anywhere, Vista, California, USA), respectively.This was done to overcome the effect of degassing (CO2loss) and consequent pH change on the water samples that might occur as samples are being transported from the field to the laboratory for cold storage prior to the analysis.While the EC was measured and reported in µS/m, the other parameters, namely, the concentration of dissolved organic carbon (DOC), ionic concentrations of NH4+, Na+, K+,Ca2+, Mg2+, and determinants like Fe2+, Cl–, SO42–, NO3–, HCOO–, and CH3COO–, were determined in the laboratory in a concentration (mmol/L) unit.

2.3.Data analysis

Each water sample was analyzed for cations, including Na+, K+, Ca2+, Mg2+, Fe2+, and NH4+by an atomic absorption spectrophotometer Buck 210/211 (Agilent Technologies, Santa Clara, California, USA), while the anions which include Cl−, SO42−, CO32−, NO3−and other determinants were determined using the colorimetric method and the Technicon Autoanalyzer II (SEAL Analytical, Hamburg, Germany) (Fig.2).Both descriptive and inferential statistics were used with the aid of Statistical Package for Social Sciences (SPSS) version 25 (IBM Corporation,New York, USA) and XLSTAT version 2019 (Microsoft Corporation, Redmond, Washington, USA).Principal component analysis (PCA) was used to identify the water chemistry changes within the study area.

An inter-variety relationship was investigated through a multivariate statistical analysis to identify the factors responsible for the water chemistry dynamics of Obagbile Catchment.The PCA coupled with varimax rotation was conducted by using SPSS version 25 and XLSTAT version 2019 to determine the preferential association among the chemical constituents in all the water samples, wheren=50 for all the water samples andn=10 for the water samples of each pathway.The PCA can be used to infer the correlations between the observations in terms of the underlying factors that are not directly observable (Mohamed et al., 2015).The PCA has three analytical stages(Zhang and Yang, 2016): (1) for all variables, a correlation matrix is generated; (2) factors are extracted from the correlation matrix based on the correlation coefficients of the variables to maximize the relationship among some of the factors and variables; and (3) the factors are rotated.First, the parameter correlation matrix is determined and used to account for the degree of mutually shared variability between the individual pairs of water quality variables.Subsequently, the eigenvalues and the factor loadings for the correlation matrix are determined.The eigenvalues correspond to the eigenvectors, which identify the groups of variables that are highly correlated among them.Lower eigenvalues may contribute little to the explanatory ability of data.Only the first few factors are needed to account for much of the parameter variability.When the correlation matrix and the eigenvalues are obtained, the factor loadings are used to measure the correlation among the variables and the factors.Factor rotation is employed to facilitate interpretation by providing a simpler factor structure (Watkins, 2018).

Factor loading is a measure of the degree of closeness among variables by using a rotation model factor analysis that provides several positive features allowing the data set interpretation.By examining the factor loadings,communalities, and eigenvalues, the variables belonging to a specific chemical process can be identified, and the importance of the major element can be evaluated in terms of the total data set and each factor.The factor loadings correspond to the correlations of each variable with the factor; hence, when the value approaches 1.00, the variable is more correlated with the factor.The factor loadings can be classified as strong, moderate, and weak corresponding to absolute loading values (strong, 0.75 and above; moderate, 0.50–0.75; and weak, 0.30–0.49)(Jacintha et al., 2017).Factor analysis of pH, EC, cations, anions, DOC, and organic acids in the water samples are performed to identify and associate their possible sources for classification in this study.

Next, we conducted the PCA coupled with varimax rotation using XLSTAT version 2019 to identify the factors responsible for the water chemistry dynamics.The PCA was used to determine the preferential association among the chemical constituents found in the water samples.We also used factor rotation to facilitate interpretation by providing a simpler factor structure (Flora and Flake, 2017; Watkins, 2018).

3.1.Characteristics of the chemical properties of water from different pathways

The mean values of the pH and EC for the water samples from the pathways indicated that overland flow had the highest values both in pH and EC, while rainwater had the lowest values (Fig.2a and b).For the rainwater chemistry,major cations, such as Ca2+and Mg2+, showed a dominance pattern (Fig.2c).The highest values of Ca2+and Mg2+concentrations were recorded in stream water, while the lowest values were recorded in rainwater (Fig.2c).NO3–,NH4+, and SO42–had their lowest concentrations in rainwater samples (Fig.2c and d).Meanwhile, K+had almost a constant concentration in all pathways, except the stream water (Fig.2c), while DOC had a constant concentration in overland flow, soil water, and groundwater (Fig.2e).In general, the highest concentrations of all the ions were found in stream water.Figure 2c reveals that in the cation category of stream water, Ca2+showed high dominance with 0.91 mmol/L concentration, followed by Mg2+with 0.18 mmol/L, K+with 0.16 mmol/L, Na+with 0.14 mmol/L, NH4+with 0.02 mmol/L, and Fe2+with 0.02 mmol/L; and the concentration of the dominant cations was in the order of Ca2+>Mg+>K+>Na+> Fe2+=NH4+in stream water.In the anion category of stream water, Cl−had the highest concentration of 0.84 mmol/L, and the concentrations of NO3–, SO42−, and CO32–were 0.03, 0.04, and 0.23 mmol/L,respectively.The dominance analysis was as follows: Cl−> CO32–>NO3–>SO42–for the other water categories except the stream water under consideration (Fig.2d).Figure 2f shows that HCOO−had a dominant concentration of 0.07 mmol/L while CH3COO−had a dominant total concentration 0.03mmol/L in stream water.

3.2.Changes in water chemistry from rainwater to stream flow

During the course of this study, a change in the concentration of each parameter was observed as water moved from one pathway to another.A one-way analysis of variance (ANOVA) at a 0.05 alpha level showed significant changes in the chemical composition of rainwater as it began contact with the soil surface and flowed into the river channel (Table 1).The Scheffe’s Test for multiple comparison further revealed that all kinds of water differed in chemistry statistically, especially for all the parameters in the stream water samples (Fig.2).These changes could be attributed to the chemistry of the medium through which the water flows.A fluctuation in the solute composition of water was evident as it moved along the pathways.The EC and pH showed a similar variation pattern as their values peaked most in overland flow and groundwater (Fig.2a and b).The major cations (i.e., Ca2+, Mg2+, Na+, Fe2+,K+, and NH4+) and anions (i.e., Cl−, CO32−, NO3−, and SO42−) also showed variations in their concentrations as water traveled through the pathways, having major concentrations in stream water (Fig.2c and d).The DOC and organic acids varied differently from all the other elements.DOC showed a low concentration in rainwater, increased in overland flow and remained constant until water flowed down to the groundwater layer, and increased as waterjoined the stream bodies (Fig.2f).CH3COO−had its lowest concentration in soil water, but gradually increased in groundwater and peaked in stream water (Fig.2e).Meanwhile, HCOO−decreased in overland flow and groundwater, but increased in soil water and peaked in stream water (Fig.2e).

Table 1Multiple comparisons of water chemistry using one-way analysis of variance (ANOVA) andScheffe’s Test among all the five kinds of water with different pathway sources in Obagbile Catchment.

Fig.2.Differences in pH (a), electrical conductivity (EC, b), and concentration of major cations (c), anions (d),dissolved organic carbon (DOC, e) and organic acids (f) in five kinds of water from different pathways in Obagbile Catchment.

3.3.Principal component analysis (PCA) of the water chemistry

The data analysis results showed that for all the water samples from different pathways, two factors mainly accounted for the total variances (Fig.3).For rainwater, the first two factors accounted for 62.30% of the total variance, where factor 1 accounted for 42.84% and factor 2 for 19.46%; for overland flow, soil water, groundwater,and stream water, the two primary factors accounted for 65.25%, 78.41%, 60.86%, and 88.65%, respectively, where the first factor of each kind of water accounted for 36.51%, 63.86%, 46.33%, and 78.45%, respectively.The two factors were related to the crustal and anthropogenic sources in rainwater, suggesting that the high loading of major cations (e.g., Ca2+and Mg2+) in rainwater samples are soil-derived (Fig.3a).The PCA for overland flow and soil water showed strong correlations among pH, EC, and the concentrations of Na+, Mg2+, HCOO−, and CH3COO−(Fig.3b and c), while the high loadings of all the parameters and the strong correlations among each other were evident in stream water (Fig.3e).

Fig.3.Principal component analysis (PCA) of selected parameters of water chemistry for rainwater (a), overland flow (b), soil water (c), groundwater (d), and stream water (e) in Obagbile Catchment.

The investigation results of this study showed that rainwater chemistry within Obagbile Catchment consists of various chemical constituents in varying proportions.The dominant cations are Ca2+and Mg2+with the highest concentrations in stream water, followed by groundwater, then soil water, overland flow, and rainwater.The high concentration of these ions in stream water samples suggested the influence of the particulate matter in the soil particles from different sources that entered the stream channel through the rainout and washout processes, as well as overland flow.The concentrations of these crustal ions in water largely depend on their availability in soil, from where they are transported through the wind action to the atmosphere.The nature of this cationic dominance pattern suggested that the study area is located outside the region where the sea salt particles predominate as condensation nuclei.

4.1.pH of water across the different pathways

The mean pH of the rainwater samples collected from the study area was 6.17, and varied across the different pathways.The pH of overland flow, soil water, groundwater, and stream water was 8.90, 7.40, 8.40, and 6.40,respectively.Studies revealed that natural rainwater usually has a pH between 5.0 and 5.5 due to the equilibrium dissolution of the atmospheric CO2(Mestre et al., 2016; Akram and Rehman, 2018).

The pH of soil water in the study area was 7.40.The soil water pH ranging from 7.30 to 8.11 indicated the presence of CaCO3.The soil water with a pH ranging from 7.30 to 7.80 was also regarded as slightly alkaline.Upon the percolation of soil water to groundwater, we observed an increase in pH from 7.40 to 8.40, which can be attributed to the carbonate dissolution presumably from the underlying rocks, thus resulting in the increase of the calcium and carbonate contents of water and further causing the pH increase of groundwater.The pH of stream water was relatively compared to that of the overland flow, soil water, and groundwater.

The variations in the water properties can be attributed to the dissolution of the ionic components found along the pathways.The pH of overland flow indicated the presence of sodium as salts.The land use of the study area is for agricultural purposes; thus, the application of agriculture fertilizers containing sulfate and phosphate, domestic discharges, photosynthesis, pesticides and insecticides on the farmland can invariably cause an increment in the pH values of soil water (Perera et al., 2013; Boyd, 2015; Reddy and Naidu, 2016; Kominko et al., 2017; Kumar et al.,2020).

For most soils, the pH tends to increase with soil depth (Datta et al., 2015), because the upper horizons receive maximum leaching by rainfall and by dissolved carbonic and organic acids, which consequently remove metal cations (e.g., Ca2+, K+, and Mg2+) and replace them with H+ion (Barman et al., 2017).

The idea that the pH of soil water increases with depth (Datta et al., 2015) held strongly in this research.However,the pH of the majority of water samples collected from all the five pathways fell within the acceptable limit of the World Health Organization (i.e., the pH ranges from 6.00 to 8.50 for domestic use).

4.2.Electrical conductivity (EC) of water across the different pathways

The EC value showed a high dispersion around the mean value according to the standard deviation.Rainwater had an EC of 69.26 µS/m, which increased to 99.45 µS/m in overland flow, 82.60 µS/m in soil water, 94.20 µS/m in groundwater, and eventually decreased to 72.13 µS/m in stream water (Fig.2b).The relatively high EC value of rainwater may be caused by the large amount of pollutants from atmosphere before rains set in.To a certain extent,this observation depicted the impact of the atmospheric particulate matters on rainwater chemistry.The EC measures the capacity of water to conduct electrical current and, as such, is directly related to the concentration of salts dissolved in the water.Salts dissolve into positively and negatively charged ions.The EC analysis for all the water samples from the different pathways evidently showed variations in the salt content of the water samples across the pathways with different impurities, which resulted from the suspended dust particles in the rainwater that were washed down with the rain.The drastic increase of EC in overland flow, soil water, and groundwater may be the result of the salt dissolution from fertilizers, pesticides, and organic wastes.

4.3.Ion and dissolved organic carbon (DOC) concentrations of water across the different pathways

4.3.1.Calcium and magnesium concentrations

The sample analysis showed that the concentration of Mg2+constituent in rainwater was 0.12 mmol/L, which increased to 0.15 mmol/L in soil water, 0.14 mmol/L in overland flow, 0.16 mmol/L in groundwater, and 0.18 mmol/L in stream water.Mg2+is a constituent of most agriculture fertilizers and can be easily transported into the stream channel, since its high concentration in stream water.The high Mg2+concentration in stream water may also be associated with the ion dissociation through weathering activities.While for Ca2+, the highest concentration of 0.91 mmol/L was found in stream water, while it reduced to 0.36 mmol/L in groundwater, 0.33 mmol/L in soil water, 0.30 mmol/L in overland flow, and 0.25 mmol/L in rainwater.The decrease in the concentration of Ca2+observed in soil water, overland flow, and groundwater could be as a result of sequestering due to an ion exchange mechanism (Fageria, 2012).An increment in the Ca2+concentration in stream water was also noted, which could be an indication of hardness (Mestre et al., 2016; Akram and Rehman, 2018).

4.3.2.Sodium concentration

The relative mean contributions of Na+and Cl−in rainwater of the study area were low compared to that of the coastal area, where they are likely to be the most dominant ionic constituents.Previous research has shown that the presence of sea salt in rainwater decreases with the distance from the coast by approximately 80% within the first kilometer, and then decreases slowly, as indicated by the chloride content of rainwater (Ganyaglo et al., 2017).The Na+dominance is so wide spread in rainwater that most of the models used for the rainwater chemistry assume a marine source for the aerosol cations (Tripathee et al., 2017).In this study, the sodium composition in the rainwater samples, which was soil-derived, further increased as it flows through all the pathways.The highest concentration was found in stream water because it is usually found in natural water due to its high solubility.

4.3.3.Chloride concentration

The elevated chloride levels in stream water suggested input from human or animal wastes or fertilizers and input of water from various pathways, many of which contain salts.The rainwater with a high chloride concentration is an indication of proximity to the sea.When ocean evaporates, some anions travel with the water vapor (Adegunwa et al., 2020).The rain generated from lake or stream evaporation will not have large chloride quantities because they are fresh water bodies.This explains the lower chloride concentration in the study area in comparison to the water obtained from the coasts.A huge proportion of Cl−loading is annually cycled through various reservoirs on Earth,and almost all of them are caused by human activities (Ciacci et al., 2017).Given the fact that the study area is notin any way close to the sea, the chloride concentration of rainwater could be of an anthropogenic source due to biomass burning.There could also be the possibility of transport from the sea because the chloride concentration rapidly diminishes with the distance from the ocean (Naranjo et al., 2015).In the inland areas, the corresponding concentration of marine-based Cl−in rainwater largely depends on topography, vegetation, distance from coast,wind direction, and wave energy (Du and Hesp, 2020).

4.3.4.K concentration

Fertilizers are the main anthropogenic sources of K+, and the K+concentration in the rainwater samples is an indication of the use of low-K-based fertilizers (Fig.2c).Thus, the low concentration of this cation in all the water samples, except in stream water, is an indication of the fact that the rocks in this region contain K+in trace proportion.Moreover, the study area is an agricultural-based one and K+is highly soluble in water; hence, the type of fertilizers used does not contain large proportions of K+.This explains its inappreciable composition in all the water samples, except in stream water.The high K+concentration in stream water could be from the direct washing of farm tools used in fertilizer application in the stream.K is an abundant element which can be released naturally through the weathering of feldspar minerals.Although its mobility is limited by the fact that soils and plants can readily absorb it; its concentration in water bodies is usually low, except in water with a highly dissolved solid content from thermal systems (Ahmad et al., 2016; Elemike et al., 2019).

4.3.5.Sulfate and nitrate concentrations

According to the previous research, sulfate in rainwater mainly arises from anthropogenic emission (Xu et al.,2015).Although it can be derived from sea spray, it is usually neutral and does not increase the acidity of rainwater(Wu et al., 2021).However, SO42−from anthropogenic source is considered as a major component that increases the H+ion concentration of rainwater (Bhaskar et al., 2022) and decreases its pH, causing it to be acidic.However, the low mean concentration of SO42−in the rainwater samples within the study area indicated that the study area is relatively free from such pollution.Rainwater usually gets high nitrate loading from pollution of various sorts (e.g.,exhaust of automobiles and industrial pollution).With the low nitrate value in rainwater within the study area, one may consequentially concludes that the area is free from atmospheric pollution of nitrate, which is of an anthropogenic origin.However, the concentration of nitrate in water from different pathway sources may greatly vary depending on the season, rainfall, rate of fertilizer application, tillage method, land use pattern, soil type, and drainage system.Nitrate is highly soluble in water.As a result, it moves readily and exhibits a slight increase in composition of water from all the other pathways sourced from fertilizers and animal manure washed off or leached into water.Note that a high nitrate level in the stream might be caused by eutrophication (Wurtsbaugh et al., 2019).

Sulfate in rainwater is usually dissolved by the impurities and gases in the atmosphere, which results in the low concentration of sulfate in rainwater.This could also be an indication that the atmosphere within one region is relatively free from pollution.Decaying organic matters, such as leaves and trees, could be a source of the high SO42−concentration in stream, although the amount of sulfate contributed from this pathway is generally small.

4.3.6.Ammonium concentration

The study area is also said to be free from pollution of nitrogenous substances that could be anthropogenic origin,because the mean concentration of NH4+was quite low.The high NH4+concentration in rainwater is usually from agricultural activities, animal wastes, decaying organic matters, and dry leaves (Ghaly and Ramakrishnan, 2015).The land use pattern around the study area is mainly agriculture; therefore, the fertilizers used within the study area have become very little, resulting in the release of small quantities of NH4+into the atmosphere and further leading to insignificant concentration of NH4+in rainwater.Ammonium ions can be introduced into rainwater in the atmosphere from various sources.When combined with water, ammonia is converted to ammonium ions.During dry seasons, it can escape back into the atmosphere.Ammonia, which is primarily derived from fertilizers and animal wastes, is almost insignificant in all the water samples.This result implied that agricultural practices could be on a subsistence level; therefore, the application of any sort of fertilizers is low, and the possibility of the introduction of animal wastes to any pathway is minimal.

4.3.7.Carbonate concentration

In general, the concentration of CO32−in rainwater is usually soil-derived.Upon rainfall, atmospheric carbon is usually added to the carbonate particles in soil water and overland flow, leading to an increase in the carbonate concentration.The groundwater increases in mineral content as it moves through the pores and the fracture openings in rocks, which is why deeper and older groundwater can be highly mineralized (Mlipano et al., 2018).At some point, the water reaches an equilibrium or balance, which could prevent it from dissolving additional substances.The higher CO32−concentration in stream water could be attributed to the mixing process due to several inputs from all the pathways.

4.3.8.Concentrations of formic and acetic acids

The mean concentrations of HCOO−and CH3COO−in the rainwater samples collected during this study were 0.07 and 0.01 mmol/L, respectively.The low mean concentration of these organic acids within the study area can be attributed to the lack of anthropogenic activity and necessary sources of these organic acids.The vertical profiles of the two kinds of carboxylic acid suggest that in remote continental areas, the direct CH3COO−emissions from vegetation may represent major atmospheric sources (Chaliyakunnel et al., 2016).High concentrations of formicand acetic acids have been found in rainwater sampled from different regions of both developed and developing countries of the world (Sun et al., 2016; Niu et al., 2018; Franco et al., 2020).

The concentration of CH3COO−increased from rainwater to overland flow and groundwater, with the highest concentration in stream water and the lowest concentration in soil water.The high concentration in stream water is an indication of the acid concentration addition from different pathways.Formic and acetic acids significantly contribute to the acidity of rainwater in both rural and urban areas (Sun et al., 2016; Niu et al., 2018).As put forward by scholars, the acid sources include direct anthropogenic emissions andin situproduction from precursors in the troposphere (Liggio et al., 2017; Cruz et al., 2019).Delort et al.(2017) showed that the presence of formic and acetic acids in rainwater results from the partitioning of acids between gas and aqueous phases, which is dependent on their pH and liquid water content.Organic acids in the atmosphere mainly originate from biogenic and anthropogenic sources.Their direct emissions from vegetation-growing processes are among typical biogenic sources, whereas anthropogenic sources include vehicle emissions and the combustion of wood and agricultural debris (Verma et al., 2017).Commonly, the concentrations of the two kinds of organic acid directly emitted from vegetation and formed secondarily in the atmosphere increase during plant-growing seasons (Tani and Mochizuki,2021).

4.3.9.Dissolved organic carbon (DOC)

The PCA result of rainwater chemistry showed a strong correlation among Ca2+, Mg2+, CO32−, EC, Cl−,HCOO−,and CH3COO−, which can be explained by factor 1 that accounted for 42.84% of the total variance with an eigenvalue of 7.93 (Fig.3a).A high bicarbonate loading was found in factor 1.A byproduct of alkaline carbonates(e.g., dolomite) was found to be strongly with Mg2+and Ca2+(r=0.73 and 0.77, respectively).The wind-blown dust transported from the area is the source of the alkaline soil-derived components.These cations are frequently found in soil and dust, explaining the association of these cations in the same factor.The EC was strongly correlated with most cations (r≥0.99).The contribution of Cl−could be from fertilizers (r=0.79).This is an indication of the influence of the crustal and anthropogenic contributions provided by the soil and road dusts from the unpaved roads and the bare soil surfaces in the study area.Compounds like Ca2+, Mg2+, and K+were added to the atmosphere within the study area in the form of their carbonate salts.These elements, some of which are basic in nature,initiated the acid neutralization in rainwater.Thus, this factor can be termed crustal or soil factor.Factor 2 accounted for 19.46% of the total variance with an eigenvalue of 3.28, including variables such as Fe+, K+, NH4+,HCOO−, CH3COO−, and DOC, which had a strong correlation with each other in the water samples.Their loadings could be from anthropogenic sources, such as manure and fertilizers (for Fe2+, K+, NH4+, and DOC) and biomass burning (for the organic acids, except CH3COO−).This factor is referred to as an anthropogenic factor.Based on this analysis, factor 1 represents the ion contributions from both natural and local anthropogenic activities, while factor 2 represents an almost insignificant ion contribution from an anthropogenic source.

For overland flow chemistry, the PCA showed that only two factors accounted for 65.25% of the total variance within the data set.Factor 1 accounted for 36.51% of the total variance with an eigenvalue of 6.94.Correlations ranging from a slight to a strong level between each other were observed.These parameters included the pH, EC,Fe2+, Cu2+, Cl−, CO32−, NH4+, NO3−, CH3COO−, HCOO−, and DOC.Most of the parameters were loaded with positive values.The pH and EC (r=1.00) were the controlling factors of all the chemical constituents.These factors can be anthropogenic as a result of the salt deposition from fertilizers.Factor 2 accounted for 28.74% of the total variance with an eigenvalue of 5.46, including Ca2+, Mg2+, K+, and Na+(Fig.3b).The factor loadings were strongly correlated with each other and could be of a terrigenous origin, which corresponded to overland flow.

The PCA results for stream water showed that the two factors accounted for 88.65% of the total variance of all the data sets.Factor 1 alone accounted for 78.45% of the total variance, while factor 2 accounted for 10.20%loading of the total variance.Factor 1 showed strong loadings of the pH, EC, Ca2+, Mg2+, Na+, Cl−, CO32−, and HCOO−(Fig.3e).It provides a novel insight into the major ion chemistry of stream water and further reveals the presence of rocks in the stream from which the weathering yielded combinations of dissolved cations and anions.For example, Ca2+and Mg2+originate from the weathering of carbonates, silicates, and evaporites; Na+and K+are from the weathering of evaporites and silicates; CO32−is from carbonates and silicates; and SO42−and Cl−are from evaporites.Conversely, factor 1 can be termed as a weathering factor.Factor 2 showed a strong NH4+and NO3−loading, and a moderate CH3COO−loading.The high concentration of NH4+can be attributed to the dissolution of ammonia fertilizer (e.g., urea and ammonium nitrate) in the stream.

DOC originates from multiple sources, with a significant proportion from fossil fuel and biogenic fraction containing both marine and terrestrial organic carbon (Siudek et al., 2015).The concentration of DOC was highest in stream water and lowest in rainwater of the study area and was equally lower comparing to other parameters as the water moved along the pathways (Fig.2e).

Some other metallic determinants, particularly the heavy metals (e.g., Mn2+, Zn2+, and Cu2+), were discovered in the rainwater samples.The metallic ions were in such low concentrations and didn’t have significant values; hence,they were not documented in the changes of water chemistry along the pathways in Obagbile Catchment.The concentration of observed ions in rainwater within Obagbile Catchment reflected nutrient loadings from terrigenous substances that served as aerosols.The absence of pollutants in rainwater indicates the absence of anthropogenic emissions.Chemical fluxes are found to be associated with rainwater inputs and soil water outputs, finally the chemical fluxes become inputs to stream water.

The chemical and statistical analyses of rainwater within the study area reveal that rainwater is close to neutral based on the pH value (6.78) and high carbonate loadings originated from soil and dust.Several other changes are also observed in the chemical composition of water as it moves from one pathway to another.The major factors responsible for the changes are weathering and anthropogenic activities (i.e., fertilizers from nearby farmlands and chemical precipitates from nearby dumpsite).The findings reveal that the water chemistry changes are highly significant and closely associated with the medium through which water flows, resulting in variations in the water chemistry as water flows from one major pathway to another.

The change generally observed in the chemical properties of water in the study area is largely caused by the influence of the solutes and minerals found along the pathways.The dominant cations are Ca2+, Mg2+, Na+, Fe2+,NH4+, and K+, and the concentration of cations in rainwater follow the order of Ca2+>Mg2+>K+>Na+>Fe2+>NH4+.In the anion category, Cl−has the highest concentration and the order is Cl−>CO32−>NO3−>SO42−, except in stream water.For the two kinds of organic acid, both of them have their dominance in the order of HCOO−>CH3COO−in rainwater, soil water, and stream water.The dominance of Mg2+and Ca2+composition in rainwater is an indication of a large incorporation of soil dust into rainwater, which reflects major crustal influences.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.