Effects of Forest Type on Nutrient Fluxes in Throughfall, Stemflow, and Litter Leachate within Acid-Polluted Locations in Southwest China

Although new inputs of acidic anions are decreasing, soil acidification still deserves more academic attention because of the effects of historical stores of SO42− already absorbed into soils. Forest canopy has large, species-specific effects on rainwater chemistry, for which the hydrological mechanism remains unclear. We investigated precipitation, throughfall, stemflow, and litter leachate across three forest types in a severely acid-polluted site located in Southwest China. Precipitation monitored over 4 months, representing summer, fall, winter, and spring, indicated neutral precipitation in Tieshanping with pH ranging from 6.58–7.33. Throughfall and litter leachate in Pinus massoniana Lamb. stands were enriched with greater cation and anion fluxes, as well as more dissolved organic carbon (DOC) flux. Rainwater from pure stands of Cinnamomum camphora (Linn) Presl yielded lower N and DOC inputs to soils with higher base saturation, which would reduce soil acidification and, therefore, improve the sustainability of forest ecosystems.


Introduction
Acid deposition is a global environmental problem that has garnered great interest due to its negative impacts on ecosystems [1,2]. Since the early 1980s, the government has made efforts to decrease acidic emissions, especially SO 2 linked to acid deposition, yet nitrogen oxide emissions have been increasing with the rapid increase in the use of private cars for nearly two decades, so it still remains a severe issue in southern China [2][3][4]. Soil acidification caused by acid deposition has adverse effects on ecosystems, particularly forest ecosystems [5]. Numerous studies report the damage of acid deposition on forests, including decreased growth and dieback [6], changing biogeochemical cycles [7], altered aboveground and belowground biomes [8][9][10], and impacts on ecosystem function and sustainability [11,12]. Additionally, despite reductions of acidic anion inputs [13,14], the recovery of acidified soils could be delayed due to historic adsorption of SO 4 2− into soils [1]. Soil properties, including soil acidity, are highly affected by forest types [15,16].
Actually, the process of the effects of forest canopies on soil nutrients is very complex. On the one hand, forest covers can alter microclimatic conditions [17], including the surface albedo, temperature, wind and humidity, which may affect decomposition rates [18], soil properties, and further affect plant growth as a feedback loop [17]. The effects of forests on the surface albedo and temperature are also complicated and result in various this canonically acidic site provides an opportunity to study the effects of precipitation, throughfall, stemflow, and floor leachate on soil acidity in different forest types and fill the knowledge gap on this topic.
In this study, our objectives were to ascertain the impacts of different waterflow types and forest canopy types on nutrient dynamics in southwest China. We hypothesized that (1) compared to decades ago, the pH of precipitation in Tieshanping has increased and transformed from sulfate acid rain to nitrate acid rain; (2) the influence of throughfall on the nutrient cycle is greater than that of stemflow; (3) a plantation of broadleaved species (Ci) is best to improve the quality of acidic soil, followed by mixed deciduous and conifer, and pure conifer stands.

Study Area and Plot Design
The experiment was conducted in Tieshanping Forest Park (106 • 41.24 E, 29 • 37.42 N), which was a heavily polluted "acid rain control zone" located about 25 km northwest of Chongqing in subtropical southwest China. In June 2018, in each of the three different forest stands (Pi, Pi_Ci, Ci) we established four 20 m × 20 m square study plots. The details about the study site can be found in Chen et al. [10]. The soil is classified as Haplic Acrisol in FAO (IUSS Working Group WRB, 2006), with low pH values (3.80-4.54 from the O horizon to the lower A horizon) [10].

Collection and Analysis of Water Samples
Bulk precipitation, throughfall, stemflow, and litter leachate were sampled as per Wang et al. [46] and measured in July 2018, October 2018, January 2019 and April 2019, which represent summer, autumn, winter, and spring, respectively. Bulk precipitation and throughfall were collected using stainless steel funnels (20 cm diameter) placed about 1.5 m above the forest floor. Three collectors were placed in an open field outside of the forest perimeter to measure bulk precipitation. For throughfall measurement (Figure 1a), five collectors were randomly placed in each plot. The funnels drained into opaque 10 L plastic buckets, which were placed below the ground surface. The stemflow (Figure 1b) was collected by 2 cm o.d. Tygon tubing with a length of about 100 cm. Before installing the tubing, the bark was shaved to ensure the seal without damage to the cambium layer. The tubing was split longitudinally and wrapped spirally around the tree stem, fastened with staples, sealed to the bole with acrylic caulk, and inserted into a 10 L plastic bucket. Litter leachate (Figure 1c) was collected using a topless 30 cm × 30 cm × 8 cm (length × width × height) stainless steel cuboid, which was covered with wire netting (1.5 mm mesh size). Each collector was connected with tubing to a 10 L plastic bucket, which was placed below ground level. The collectors were installed directly under the litter layer with five replicates per plot.
All water samples were collected weekly in the four sampling months and lumped into monthly samples for chemical analysis (stored at 4 • C before pooling). In total, there were 12 samples for precipitation and 16 samples, respectively, for throughfall, stemflow, and litter leachate for each forest type spanning July 2018 to April 2019.
All samples were filtered through 0.45 µm member filters before analyses. Water pH and electrical conductivity (EC) were measured by a benchtop pH meter (REX PHS-3C, Shanghai, China) and a benchtop conductivity meter (REX DDS-12A, Shanghai, China), respectively. We used a total organic carbon analyzer (TOC-L, SHIMADZU, Kyoto, Japan) to determine the dissolved organic carbon in all samples. The content of Ca and Mg cations was analyzed by an atomic absorption spectrophotometer (AAS, HITACHI ZA-3300, Tokyo, Japan), and K and Na concentration was measured using a flame photometer (FP640, Shanghai, China). An ion chromatograph (ICP, Dionex ICS-900, USA) was used to determine the anion concentration in all samples. NO 3 -N and NH 4 -N were measured using a flow analyzer (Seal AA3, Germany). All water samples were collected weekly in the four sampling months and lumpe into monthly samples for chemical analysis (stored at 4 °C before pooling). In total, ther were 12 samples for precipitation and 16 samples, respectively, for throughfall, stemflow and litter leachate for each forest type spanning July 2018 to April 2019.
All samples were filtered through 0.45 μm member filters before analyses. Water pH and electrical conductivity (EC) were measured by a benchtop pH meter (REX PHS-3C Shanghai, China) and a benchtop conductivity meter (REX DDS-12A, Shanghai, China respectively. We used a total organic carbon analyzer (TOC-L, SHIMADZU, Kyoto, Japan to determine the dissolved organic carbon in all samples. The content of Ca and Mg cat ons was analyzed by an atomic absorption spectrophotometer (AAS, HITACHI ZA-3300 Tokyo, Japan), and K and Na concentration was measured using a flame photomete (FP640, Shanghai, China). An ion chromatograph (ICP, Dionex ICS-900, USA) was use to determine the anion concentration in all samples. NO3-N and NH4-N were measure using a flow analyzer (Seal AA3, Germany).
Water pH, EC, DOC, and RCA were analyzed by using one-way ANOVA wit Tukey's HSD to determine the different ion concentration among forest types respectivel for stemflow, throughfall, and litter leachate. For each forest type, one-way ANOVA Tukey's HSD was also used to detect the seasonal variation across water properties. Inde pendent t-tests were performed to determine significant differences across water proper ties, between precipitation and stemflow, throughfall, or litter leachate. All figures wer created in OriginPro 9.0.

Data Curation and Statistical Analysis
Ion concentration was represented as ratio of cations to anions (RCA), calculated as ∑ + (sum of K + , Na + , Ca 2+ , Mg 2+ , NH 4 . The ion fluxes were calculated by multiplying ion concentrations by water fluxes. Base cation and anion fluxes (∑ + flux and ∑ − flux) were, respectively, the sum of K + , Na + , Ca 2+ , Mg 2+ , NH 4 + , and F − , Cl − , SO 4 2− , NO 3 − . Water pH, EC, DOC, and RCA were analyzed by using one-way ANOVA with Tukey's HSD to determine the different ion concentration among forest types respectively for stemflow, throughfall, and litter leachate. For each forest type, one-way ANOVA, Tukey's HSD was also used to detect the seasonal variation across water properties. Independent ttests were performed to determine significant differences across water properties, between precipitation and stemflow, throughfall, or litter leachate. All figures were created in OriginPro 9.0.

Precipitation
The pH values of precipitation varied from 6.58 to 7.33 in the four sampling months, which indicated the precipitation was not acid rain in these months. The pH, EC and DOC contents had similar seasonal variation with highest values in January (Table 1).
Ca 2+ was proportionally the most abundant cation, ranging between 24.81-71.62%, and SO 4 2− was the most abundant anion (varying from 55.65-66.05%), followed by NO 3 − (11.02-22.73%). There was no significant seasonal variation for RCA in precipitation (p = 0.862) ( Table 1). Following the peak ion concentration in January, cation and anion flux also peaked in January ( Figure 2). Ca 2+ flux predominated in July and April (respectively 51.16% and 52.53% of the whole cation flux), while K flux was greatest in October with 7.84 kg·hm −2 . In January, Ca and Mg flux combined accounted for about 30% of the whole cation flux. SO 4 2− flux accounted for 59.94-73.34% of all anion flux ( Figure 2). also peaked in January ( Figure 2). Ca 2+ flux predominated in July and April (respectively 51.16% and 52.53% of the whole cation flux), while K flux was greatest in October with 7.84 kg•hm −2 . In January, Ca and Mg flux combined accounted for about 30% of the whole cation flux. SO4 2− flux accounted for 59.94-73.34% of all anion flux ( Figure 2).   The pH values (6.56-7.87) of all throughfall samples were slightly higher than that of precipitation, although not statistically significant (p > 0.05), with the exception of the October samples in Pi_Ci and Ci. The pH values in all stemflow samples were significantly lower than that of precipitation (p < 0.05), with the exception of the January samples. Forest types had significant effects on stemflow pH (p < 0.001), but not on throughfall pH (p = 0.491). Stemflow pH was about 5.6 in Pi and Pi_Ci, except for the January samples, while the pH of stemflow in Ci varied from 5.66-6.62. There was significant seasonal variation in pH values for stemflow (p < 0.001) and throughfall (p < 0.001), with significantly higher values in January. There were no observed seasonal effects on the pH of litter leachate (p = 0.061), whereas forest type had a significant effect (p < 0.001). The pH values of litter leachate in the Pi_Ci and Ci plots were significantly (p < 0.05) higher than that in the Pi plot, except for Pi_Ci sampling in January (Figure 3a). = 0.491). Stemflow pH was about 5.6 in Pi and Pi_Ci, except for the January samples, while the pH of stemflow in Ci varied from 5.66-6.62. There was significant seasonal variation in pH values for stemflow (p < 0.001) and throughfall (p < 0.001), with significantly higher values in January. There were no observed seasonal effects on the pH of litter leachate (p = 0.061), whereas forest type had a significant effect (p < 0.001). The pH values of litter leachate in the Pi_Ci and Ci plots were significantly (p < 0.05) higher than that in the Pi plot, except for Pi_Ci sampling in January (Figure 3a).  Figure 3. pH, electrical conductivity (a), dissolved organic carbon, and ratio of ∑ + to ∑ − (RCA) (b) of throughfall, stemflow, and litter leachate in different forest types during the sampling months. Capital letters indicate significant differences between sampling times for waterflows of each forest type. Lowercase letters indicate significant effects of forest type at the same sampling time. Asterisks indicate significant differences between precipitation and stemflow, throughfall, and litter leachate at the same sampling time. *, ** and *** respectively indicate significance at p < 0.05, 0.01 and 0.001.
Both EC and DOC of stemflow were significantly greater compared to those of precipitation (p < 0.05), except for EC of Ci in July and October. Forest type had opposite effects on EC and DOC concentration compared to pH of stemflow and litter leachate, with unclear impacts on throughfall. In Pi stands, the EC of stemflow (ranging from 147.7 to 439.25 μs) was significantly higher than that in Pi_Ci (117.50-292.88 μs) and Ci (76. .75 μs) during all the sampling months (p < 0.05), except for the October samples in Pi_Ci. Among all water samples, the DOC concentration in precipitation was the lowest, with an average of 24 mg•L −1 , compared to higher DOC levels in throughfall, litter leachate, and especially stemflow. Forest type did not impact DOC of throughfall. DOC concentration in stemflow (mean of 114.18 mg•L −1 ) was significantly higher than in precipitation in all plantations from all sampling months (p < 0.05). DOC concentration in stemflow Capital letters indicate significant differences between sampling times for waterflows of each forest type. Lowercase letters indicate significant effects of forest type at the same sampling time. Asterisks indicate significant differences between precipitation and stemflow, throughfall, and litter leachate at the same sampling time. *, ** and *** respectively indicate significance at p < 0.05, 0.01 and 0.001.
Both EC and DOC of stemflow were significantly greater compared to those of precipitation (p < 0.05), except for EC of Ci in July and October. Forest type had opposite effects on EC and DOC concentration compared to pH of stemflow and litter leachate, with unclear impacts on throughfall. In Pi stands, the EC of stemflow (ranging from 147.7 to 439.25 µs) was significantly higher than that in Pi_Ci (117.50-292.88 µs) and Ci (76.45-297.75 µs) during all the sampling months (p < 0.05), except for the October samples in Pi_Ci. Among all water samples, the DOC concentration in precipitation was the lowest, with an average of 24 mg·L −1 , compared to higher DOC levels in throughfall, litter leachate, and especially stemflow. Forest type did not impact DOC of throughfall. DOC concentration in stemflow (mean of 114.18 mg·L −1 ) was significantly higher than in precipitation in all plantations from all sampling months (p < 0.05). DOC concentration in stemflow from Ci (50.28-99.41 mg·L −1 ) was lower than that from Pi_Ci (64. .57 mg·L −1 ) and Pi (96. 34-166.16), and the difference was statistically significant in October (p = 0.027) and April (p = 0.032). Similarly, DOC concentration in Ci litter leachate was significantly lower than that in the Pi_Ci and Pi (p < 0.001) litter leachate samples (Figure 3b).
DOC flux in throughfall for Pi and Pi_Ci was triple and double that of precipitation in April, respectively. Due to lower volume, DOC flux in stemflow was reduced compared to that in precipitation, despite the higher concentration of DOC in stemflow. DOC flux in litter leachate was lower than that in precipitation, except in July, and was much lower than that in throughfall. Throughfall DOC in Ci was lower by 46.70%, 34.27%, 42.70% and 55.59% in July, October, January, and April, respectively, and litter leachate DOC flux was significantly lower, by 41.58%, compared to Pi in April (p = 0.027) ( Table 2).

Ion Concentration and Flux
Most cation and anion contents were higher in concentration in throughfall, litter leachate, and especially stemflow, compared to precipitation. Ca 2+ and SO 4 2− were, respectively, the most abundant cation and anion ( Figure 4). The mean sum of cation concentration was 605, 535.53, 966.26 and 662.97 µeq·L −1 , and the mean sum of anion concentration was 755.94, 872.13, 1812.54 and 1110.88 µeq·L −1 in precipitation, throughfall, stemflow, and litter leachate, respectively ( Figure 4).
We also found that the mean sum of cation or anion concentration could be ranked in the order Pi > Pi_Ci > Ci. In Ci, the greater difference in anions (lower by 32.87% and 39.08% compared to Pi_Ci and Pi) than cations (lower by 15.37% and 27.28% than Pi_Ci and Pi) resulted in higher RCA in Ci (Figure 3b). RCA in July was significantly higher than that in the other sampling months for all forest types (p < 0.05), except for stemflow sampling in October in Pi. Generally, RCA in Ci was larger than that in Pi, and the difference was statistically significant in July for stemflow (p = 0.025) and litter leachate (p = 0.011), and in April for stemflow (p = 0.005) and throughfall (p = 0.014). The mean value of RCA in Ci was greater by 35.60%, 12.52% and 35.53% compared to Pi, and by 25.26%, 12.29% and 6.97% compared to Pi_Ci for stemflow, throughfall, and litter leachate, respectively (Figure 3b).
Due to the low nutrient flux (0.01875 to 0.3695 and 0.03950 to 0.6302 kg·hm −2 , respectively, for cation and anion flux) in stemflow, in Figure 5 we show the flux as throughfall plus stemflow (TF+SF) and litter leachate (LL). The litter layer intercepted nutrient flux from forest rainwater, with the exception of NO 3 − . The leaf litter in Pi stands leached the most NO 3 − (3.39-9.60 kg·hm −2 ), which was much greater than that from Ci stand litter (0. 34  Most cation and anion contents were higher in concentration in throughfall, litter leachate, and especially stemflow, compared to precipitation. Ca 2+ and SO4 2− were, respectively, the most abundant cation and anion (Figure 4). The mean sum of cation concentration was 605, 535.53, 966. 26   We also found that the mean sum of cation or anion concentration could be ranked in the order Pi > Pi_Ci > Ci. In Ci, the greater difference in anions (lower by 32.87% and 39.08% compared to Pi_Ci and Pi) than cations (lower by 15.37% and 27.28% than Pi_Ci and Pi) resulted in higher RCA in Ci (Figure 3b). RCA in July was significantly higher than that in the other sampling months for all forest types (p < 0.05), except for stemflow sampling in October in Pi. Generally, RCA in Ci was larger than that in Pi, and the difference was statistically significant in July for stemflow (p = 0.025) and litter leachate (p = 0.011), and in April for stemflow (p = 0.005) and throughfall (p = 0.014). The mean value of RCA in Ci was greater by 35.60%, 12.52% and 35.53% compared to Pi, and by 25.26%,   The cation and anion fluxes of TF + SF in Ci were lower than those in Pi_Ci and Pi ( Figure 5), except in October, while the ratio of cation flux to anion flux of throughfall and stemflow in Ci was significantly higher than that in Pi in July and April (Table 3). Compared to Ci stands, the ratio of cation flux to anion flux in litter leachate of Pi_Ci was greater by 17.39-46.03%. In Ci stands, the ratio of cation flux to anion flux was 80.95%, 39.53% and 36.17% greater than that in Pi stands in July, October, and April, respectively (Table 3). Additionally, for stemflow and litter leachate, base saturation in Ci was significantly higher than that in Pi, except for litter leachate in January, and also significantly higher than that in Pi_Ci, except for stemflow in January and litter leachate in July and January (Table 4).

The Change of Precipitation
The data in this study indicated that precipitation in Tieshanping was neutral with pH values varying from 6.58-7.33. To evaluate the change in precipitation quality, we compared our data with previous data from 2001-2004 [47] and 2003 [48] in the same study site ( Table 1). The mean pH values from 2001-2004 [47] and 2003 [48] were about 4.1, which was more acidic than the mean pH observed in this study, from 2018-2019, by over 2 units, and this could be due to decreased acidic emissions in China [1]. Most of the average ion concentrations substantially increased compared to mean values from 2001-2004, especially for K + (3.5×), Ca 2+ (8.6×), Cl − (6.8×), SO 4 2− and NO 3 − (6-7×). In 2003 [48], most ion concentrations were much lower than those of 2001-2004, except for Ca 2+ , SO 4 2− and NO 3 − , which were still lower than the values observed in this study. The ratio of SO 4 2− to NO 3 − in our study was an average of 3.16, which was much lower than that of the other two previous studies, at 4.27 and 5.3, respectively, in Xiang et al. [47] and Wang and Xu [48] in the same study site in Tieshanping. This increasing proportion of NO 3 − in precipitation compared to previous studies [47,48] was also confirmed by Chen [41] in Chongqing. The increased RCA also confirmed the transformation of precipitation from acidic towards neutral. These data partially confirm our first hypothesis that the pH of precipitation and the proportion of NO 3 − have increased. However, due to the large interannual variability observed over 2001-2004 [47], the 1-year monitoring data in this study are not sufficient to assert that precipitation in Tieshanping is presently neutral, not acidic. In order to do so, more observational data over a longer period would be needed.

The Effects of Water Flow
In our study, both cation and anion fluxes in litter leachate were lower than that in throughfall plus stemflow, which indicated an interception of nutrients at the litter layer. To further detect how much throughfall and stemflow influence the nutrient flux of litter leachate, we performed variation portioning analysis (VPA) using cation, anion, and DOC flux. As indicated by VPA, a total of 68% of litter leachate nutrient flux could be explained by throughfall and stemflow. Nutrient flux from throughfall and stemflow respectively contributed 20% and 3% to litter leachate nutrient flux, while their common contribution was 45%. The contribution of throughfall and stemflow to the forest floor has also been confirmed by previous data [29,32]. The relatively small contribution of stemflow observed in this study is supported by Bellot et al. [49], which found that nutrient inputs of stemflow only represented about 2-3% of total dissolved inputs into the soils. Despite the low contribution, stemflow can play an important role in forest ecosystems, as it concentrates water in the area immediately proximal to the tree trunk, providing a supply of water and nutrients to the tree and soil far more so than throughfall [49,50].
This present study also indicates that stemflow is enriched with a greater nutrient concentration than throughfall. We found the highest DOC content, cation and anion concentration, and lowest pH value in stemflow, followed by throughfall and litter leachate. Higher concentrations of nutrients in stemflow can be attributed to both nutrients leaching from stems and very low stemflow volume [51]. When we replaced nutrient flux with concentration in VPA analysis, the contribution of stemflow to litter leachates increased from 3% to 9%, which further illustrates the importance of stemflow in influencing nutrient concentrations in the root zone that affect rhizosphere microbial communities and function [52]. Thus, our second hypothesis is partially correct in that throughfall does contribute a greater nutrient flux to the forest floor than stemflow; however, the transient impacts of stemflow are more intensive on the root zone, specifically.

The Effects of Forest Type
A previous review that analyzed data from 17 stand pairs determined that coniferous forests intercept more pollutants than deciduous stands [31], a conclusion that our findings support as well. In this study, cation and anion flux in Ci were much lower than in Pi and Pi_Ci in throughfall plus stemflow; however, Ci had the highest pH and RCA and the lowest DOC in stemflow across three forest stands. A study in northwest China illustrated that the enrichment capacity of acid anions and exchangeable bases in the canopy of conifers is stronger than that of broadleaved species [27]. While these results were consistent with ours, Han et al. [27] recommended planting coniferous species to improve forest ecosystem function, a measure that may not be suitable in our study site. The study site in Han et al. [27] was located in northwest China, which does not experience acid rain or soil acidification [53]. Our study was conducted in a typical acid-polluted area with severely acidic soil (average pH of 3.98, 4.54, and 4.6, respectively, in O_horizon of Pi, Pi_Ci and Ci, [10]), which would benefit from species that could supply less acidic rain water and greater base saturation. We also calculated the base saturation of throughfall, stemflow, and litter leachate, finding that Ci had the highest base saturation, followed by Pi_Ci (Table 4). We also found that the sum of anion flux in litter leachate from Ci was much lower than that of Pi, which was mainly due to a lower amount of NO 3 − flux in Ci. The lower flux of NO 3 − from litter leachate in Ci than in Pi observed in this study is consistent with results from central Japan, which showed that average NO 3 − flux of litter leachate in evergreen forests was higher than that in deciduous forests [54]. Our data showed that throughfall N input was negatively associated with soil pH, which is supported by a study from an urban-to-rural transect in southern China [55]. In this study, throughfall NO 3 − flux was the highest in Pi stands, and the lowest in Ci stands, where lower N input of throughfall (plus stemflow) and litter leachate contributed to higher pH values of the soils, an observation supported by our previous data as well [10].
We also found that the DOC concentration and flux in throughfall and litter leachate were minimum in Ci and maximum in Pi. Though DOC flux in stemflow of Ci was relatively higher, it was negligible because of its small volume compared to throughfall and litter leachate. In our study, the DOC concentration in throughfall and DOC fluxes in forest floor leachates were found to be higher in coniferous species stands than in broadleaved species stands, which was supported by some previous studies [29,40,56]. DOC could influence water acidity due to the composition of short-chain acids and large molecules such as humic and fulvic acids [57]. Previous studies indicated that tree species with high DOC in their forest floor leachates stimulate base cation leaching, which would cause soil acidification [58].
In summary, Ci is more beneficial for improving soil acidification than Pi based on the acidic status of soils in Tieshanping. However, our findings are only based on measurements from four 1-month periods, which limited our knowledge of the effects of forest types on soil acidity. Longer-term and more-detailed monitoring data on different forest types in this study area will be needed in the future to further ascertain the effects of different water flow and forest types on soil quality.

Conclusions
Our monitoring data indicated that precipitation in Tieshanping was neutral and the proportion of NO 3 − was increased. Despite very low nutrient fluxes, stemflow also played an important role in affecting the nutrient cycling in the area near the stem by increasing nutrient concentrations. According to hydrological chemical characteristics, pure plantations of Ci could supply lower N and DOC inputs to soils with higher base saturation for improving soil acidity, followed by mixed Ci and Pi stands. In stands of Ci, higher RCA and base saturation accompanied by lower N and DOC inputs in throughfall plus stemflow and litter leachate yielded higher pH in water flow entering the soils, which would support recovery from soil acidification and further ensure sustainable development of forest ecosystems in acid-polluted locations. This study underscores the changing rainwater chemistry of forest hydrological processes and the effects on soil acidification across mixed and pure stands of coniferous species and broadleaved species. While these findings are based on four 1-month periods of measurement, to clearly determine the water chemistry of precipitation and the effects of forest type in Tieshanping, more observational data over a longer period are needed.

Data Availability Statement:
The data presented in this study are available from the corresponding author upon reasonable request.