Nutrient Input via Incident Rainfall in a Eucalyptus dunnii Stand in the Pampa biome

We evaluated the nutrient input via incident rainfall in a Eucalyptus dunnii stand in the Brazilian Pampa biome. During the two-year study, we analyzed nutrient concentrations in incident rainfall (P), throughfall via drip (Pi), and stemflow (Et). Ion concentrations (NO2 -, NO3 -, PO4 3-, SO4 2-, Cl-, K+, Ca2+, and Mg2+) were determined from the aqueous solution, and based on the water volume, the quantity per hectare (kg ha-1) was estimated for each element. Statistical differences observed by ANOVA and separation of means contrasts were subjected to Tukey’s test (5%), adopting a completely randomized design. Higher concentrations of SO4 2in P; NO3in Pi solution, and K+, Ca2+, Mg2+, and Clin Et were observed. Greater amounts of nutrients such as Cl-, K+, Ca2+, and SO4 2 were incorporated via Pi.


INTRODUCTION
The productivity of commercial tree plantations is influenced by the supply of minerals dissolved in water and those introduced via atmospheric deposition.Nutrient input through rainfall is important to maintain the productive capacity of soils where eucalyptus silviculture is practiced (Schumacher & Viera, 2015).
Water passing through the canopy of trees removes large amounts of atmospheric particles and aerosols, which are deposited on the surface of plant tissues (Waring & Schlesinger, 1985).These nutrient flows ensure forest productivity and can be boosted by nutrient inputs through rainfall, which comprises an important stage of geochemical cycling (Ashagrie & Zech, 2010).
Trees influence the nutrient return to the soil, changing the amount and chemical composition of non-intercepted rainfall (Carnol & Bazgir, 2013).Tree canopies influence atmospheric deposition of nutrients, contributing to the redistribution of rainfall, changing the pH and ion concentrations, and washing the dry deposits on the canopy and leaching elements from the vegetal tissues (Shen et al., 2013).
Vegetation cover in forest plantations has a prominent function in water balance, and the type of vegetation also influences the uptake of water and nutrients (Diniz et al., 2013).Vegetation cover varies over time, in response to fluctuations in precipitation quality, variations in the availability of aerosols and vegetation exudates, and precipitation intensity (Walling, 1980).
As silviculture branches out into new areas, especially in degraded areas with low productive potential, the evaluation and understanding of nutrient inputs due to incident rainfall may aid in the efficient management of fertilization, leading to the conservation of natural resources.The objective of this study was to evaluate the nutrient input through incident rainfall in a Eucalyptus dunnii stand, established in the Pampa biome of Brazil.

Study site characterization
The study was conducted in a Eucalyptus dunnii stand located in Alegrete, Rio Grande do Sul, Brazil, under the central geographic coordinates of 29°47′S and 55°17′W.According to the climatic classification of Köppen, the climate of this region is type Cfa, with high temperatures in the summer season (Alvares et al., 2014).The annual average temperature is 18.6 °C and the average annual rainfall is 1574 mm.The soil, which presents low fertility, is classified as typical Distrophic Red Argisol (EMBRAPA, 2013), with sandy texture, ranging from sandy-loam to sandy-loam.
Soil preparation was carried out with subsoiling at an average depth of 60 cm and seedling planting at 2.0 m × 3.5 m spacing.Planting was carried out manually during January 2009.The following cultivations were carried out in the planting: application of 300 kg ha -1 of N-P 2 O 5 -K 2 O, in the proportion 06-30-06 + 0.6% of B, 30 days before planting.The second application of fertilizer was at 90 days after planting; 140 kg ha -1 of N-P 2 O 5 -K 2 O was applied in the proportion 22-05-20 + 0.2% of B + 0.4% of Zn.The third application of fertilizer was 270 days after planting, with 140 kg ha -1 of N-P 2 O 5 -K 2 O, at the proportion 22-00-18 + 1.0% of S + 0.3% of B. In addition to chemical fertilizers applied mechanically in the interline, chemical weeding, mechanical weeding, ant-fighting cutters and firebreak maintenance were also carried out in the stand.
The trees were 48 months old at the beginning of the experiment.Their mean diameter at breast height (DBH) was 11.9 cm, mean height of 13.8 m, leaf area index (LAI) of 2.78 m 2 , basal area (G) of 14.0 m 2 ha -1 , number of trees (N) of 1143 and volume of 124.3 m 3 ha -1 .

Water sampling
Chemical analysis of rainfall water was initiated in January 2012 and ended in December 2013.Incident rainfall (P), throughfall (Pi), and stemflow (Et) were evaluated (Figure 1).
Incident rainfall was determined using three rain gauges installed in an area adjacent to the eucalyptus stand.Throughfall was determined with the installation of twelve rain gauges.Both the collectors for the quantification of water volume of incident rainfall and for throughfall, had a diameter of 20 cm.However, for water sampling for chemical analysis, collectors with a diameter of 15 cm were used.
The stemflow (Et) was measured using twelve collectors.The Et collectors were made of a 60 L stocking hose and a collecting container, attached to the trunk, at a height of 150 cm from the ground, using polyurethane foam and silicone glue.
Samples were collected biweekly.In each collector, the precipitated volume was measured.During collection, an aliquot was removed for chemical analysis.In the laboratory, the pH values were determined with a pH meter (Metrohm 827 pH LAB).After filtration using a 0.45-μm filter, the NO 2 -, NO 3 -, PO 4 3-, SO 4 2-, Cl -, K + , Ca 2+ , and Mg 2+ ions were determined by means of ion chromatography (Metrohm 861 Advanced Compac IC), according to the methodology proposed by APHA et al. (1998).
The normal distribution and homogeneity of variances were verified by Shapiro-Wilk and Bartlett tests, respectively.To meet these assumptions, there was a need for logarithmic transformations of the individual concentrations.Tukey's test and Student's t-test were applied after ANOVA, both with a 5% probability of error, to separate the means contrasts.
In the statistical analysis, we adopted a completely randomized design with different numbers of repeitions, where the treatments were incident rainfall, throughfall and stemflow, and each monthly composite sample corresponded to one repetition.The analyses were performed with the assistance of the program Assistat version 7.7 (Silva & Azevedo, 2002).

Distribution of incident rainfall
In relation to the evaluated periods, it was observed that there were higher values of throughfall and stemflow in 2013 (Table 1).The difference of 502.52 mm incident rainfall between the two years was due to the influence of the "La Niña" phenomenon, which occurred in Brazil  The incident rainfall in 2012 was distributed in 91.08% via throughfall and 0.91% via stemflow.In 2013, 91.48% of incident rainfall was quantified in throughfall and 1.28% in stemflow.The highest volumes of throughfall were associated with the leaf area index of the stands, as the canopy cover allowed greater passage of rainwater through the canopies.According to Whitehead & Beadle (2004), low leaf area index and seasonal dynamics are two of the mechanisms that facilitate the passage of precipitation through the canopy to alleviate water deficits in the soil.These results also showed that larger volumes of aqueous solution flowing through the trunk are directly influenced by the precipitation intensity.
In the same study area, Corrêa (2011) also evaluated incident rainfall.However, the monitoring period began at 36 months and ended at 48 months.The author observed a higher volume of incident rainfall, registering 1586 mm year -1 .Of this total, 93% was collected inside the stand, where 98% was from throughfall and 2% from stemflow.These results, obtained at the study site over a different time scale, revealed similarities in the distribution of throughfall, with less interception of rainwater by the canopy when the trees are younger (Chang, 2006).
A similar pattern of incident rainfall distribution was also found by Laclau et al. (2003), in Eucalyptus sp. up to 72 months of age, planted in the Congo.During the rainy season, which covers the months from October to May, the authors observed that 91% of the total effective precipitation corresponded to throughfall and 1.9% to stemflow.
In relation to stemflow, in a study in the Atlantic Forest, considering different levels of ecological succession in native forest fragments, Diniz et al. (2013) found that less than 0.3% of the water flowed through tree trunks; the authors found that volumes of incident rainfall, throughfall, and interception were similar between the three stages evaluated.

Nutrient concentration in incident rainfall, throughfall and stemflow
After the water crossed the canopy of the trees, there was an increase in the pH values in 2013 and in the concentration of most of the nutrients, in the two evaluated periods.The contribution of ions contained in the water from throughfall (Pi), relative to the incident rainfall (P), tended to enrich the accumulated amounts of ions, except for PO 4 3-and NO 2 -in 2013 (Table 2).
During the two years, significant differences were observed in the concentrations of cations and anions, mainly for nitrate (NO 3 -) in the internal precipitation and for PO 4 3-, K + , Mg 2+ , and Cl -that are present in higher concentrations in the stemflow solution.It was

5/9
Nutrient Input via Incident Rainfall… Floresta e Ambiente 2018; 25(3): e20160559 possible to verify that the concentrations of all elements increased after the water crossed the canopy and flowed through the trunk.Due to the action of the crown of trees in leaching of metabolites next to the stemflow, the concentration of K + increased by 280% in 2012 and by 625% in 2013.It was also found that, in 2013, the aqueous solution of the stemflow presented higher pH values than that of incident rainfall and throughfall.Corrêa (2011) analyzed a Eucalyptus dunnii stand at 2.5 years, and found increases of 600% of K + , 295% of Cl -, 125% of Mg 2+ , 43.67% of Ca 2+ , and 10% of PO 4 3-, respectively, after passing through the canopy of the eucalyptus stand.
The order of nutrient input was Cl -> Ca 2+ > SO 4 2-> K + > NO 3 -> Mg 2+ > NO 2 -> PO 4 3-for incident rainfall in 2012.In 2013, the order followed the sequence SO 4 2-> Cl -> Ca 2+ > K + > Mg 2+ = PO 4 3-> NO 3 -> NO 2 -.For throughfall in 2012, the sequence Cl -> Ca 2+ > K + > SO 4 -> Mg 2+ > NO 3 -> NO 2 -> PO 4 3-was observed.In 2013, practically the same sequence was maintained, except for the order PO 4 3-> NO 2 -.For stemflow, in 2012, the sequence K + > Cl -> Mg 2+ > Ca 2+ > SO 4 2-> NO 2 -> NO 3 -> PO 4 3-was verified.In 2013, stemflow followed the sequence With regard to the seasonal contribution of ions (Figure 2), it can be observed that the distribution presented a varied fluctuation, where the highest concentrations of elements occurred in the months when there was less rainfall.Paramee et al. (2005) also verified that the pattern and amount of moisture deposition were strongly influenced by the precipitated volume.The concentration of nutrients is inversely proportional to the amount of water; the lower the precipitation volume, the lower the nutrient dilution in water and, consequently, the higher the concentrations (Chang, 2006;Laclau et al., 2003).
Higher concentrations of SO 4 2-, Ca 2+ , and Cl -in the aqueous solution of incident rainfall were also observed by Corrêa (2011) in the same study area.Heartsill-Scalley et al. (2007) in a subtropical forest and Laclau et al. (2003), in a study of Eucalyptus sp., among other studies (Shen et al., 2013;Wang et al., 2004;Bellot & Escarre, 1991) also showed higher concentrations of these ions.
The presence of high SO 4 2-concentrations in the chemical composition of incident rainfall is indicated as one of the consequences of anthropogenic emissions that contribute to the increase in atmospheric pollution resulting from burning, industrial activity, and combustion of petroleum fuels, as reported by Garten et al. (1988), Singer et al. (1996), and Armbruster et al. (2003).The quality of the incident rainfall is also influenced by the deposition of particles, which are derived from vegetation biomass fires and enrich the precipitation with SO 4 2-, Ca 2+ , Cl -, and NO 3 - (Laclau et al., 2003).This practice is common in the field areas of the Pampa biome, used as an alternative for pasture renewal and elimination of invasive plants (Verdum et al., 2014).
The accumulation of dust and ash on the canopy, which are carried after the leaf intercepts water, may have led to the increase in ion concentrations, as can be observed in the present and previous studies (Shen et al., 2013;Heartsill-Scalley et al., 2007;Balestrini et al., 2007;Dezzeo & Chacón, 2006).After interaction with the canopy, all ions in the throughfall solution were enriched, with higher concentrations of Cl -, K + , SO 4 2-, and Ca 2+ .These findings were consistent with those observed in the native forests of Venezuela (Dezzeo & Chacón, 2006), in Quercus stands (Rodà et al., 1990), and in the native forests of India (Prakasa Rao et al., 1995).
After interaction with the canopy, precipitation is susceptible to leaching potassium from plant tissues.Potassium does not participate in the structural constitution of organic compounds, and therefore, it is found in its ionic form inside the cells, acting in the control of stomatal opening and closure (Epstein & Bloom, 2006).Calcium is closely associated with the cell wall of tissues (Taiz & Zeiger, 2013).Therefore, high concentrations of this nutrient in the solution of the precipitation can be traced to the drift of limestone particles, applied in crops near the study area.
In the solution of stemflow, except for SO 4 2-and NO 3 -, higher concentrations of ions, mainly K + , were registered relative to incident rainfall and throughfall.In a Eucalyptus sp.stand, Laclau et al. (2003) also verified a reduction in the nitrate content after stemflow, to the detriment of the cationic changes resulting from the interaction of this element with the leaves.According to Armbruster et al. (2003), the reduction in sulfate content after stemflow occurs as a result of the interaction of this element with the organic acids of the plant tissue.Nitrate can also be volatilized after interaction with stem water, which makes the solution of stemflow less acidic relative to incident rainfall and throughfall.

Nutrient input via incident rainfall, throughfall and stemflow
After measuring the quantities of ions from incident rainfall on the Eucalyptus dunnii stand (Table 3), a greater contribution of Ca 2+ and Cl -was observed in the incident rainfall and throughfall.In stemflow, the highest contribution was from K + during the two evaluation periods.The higher nutrient quantities observed in 2012 are a reflection of their higher concentrations, which occurred due to lower rainfall volume and less dilution of nutrients.Enrichment of rainwater after passing through the canopy was responsible for greater nutrient input into the stand, thus highlighting the importance of the geochemical cycle in nutritional dynamics of eucalyptus culture.
The elements incorporated in the largest quantities were Ca 2+ , Cl -, SO 4 2-, and K + , a situation also observed by Heartsill-Scalley et al. (2007).This results from the accumulation of dry and wet deposition on the canopy, carried through leaching and foliar washing.Potassium was also found to be higher in stemflow solution.
The addition of potassium to the aqueous solution, after interaction with the canopy, also occurred in an Acacia mangium stand in China, where precipitation measures 2172.7 mm year -1 (Shen et al., 2013); the authors

7/9
Nutrient Input via Incident Rainfall… Floresta e Ambiente 2018; 25(3): e20160559 observed higher input of K + compared to the present study, at 6.14 kg ha -1 via incident rainfall, 62.1 kg ha -1 via throughflow, and 2.94 kg ha -1 via stemflow.These results indicate the influence of precipitation volume on nutrient flow, since there is a lower quantity of ions contributed by wet atmospheric deposition in this study, when compared to Bellot & Escarre (1991), Calil et al. (2010), andHeartsill-Scalley et al. (2007).
Due to the contribution of rainfall water to nutrient input, fertilization management strategies can be adopted based on studies that evaluate the chemical quality of water and the interaction with the canopy, branches and trunk.Taken together, our findings indicate that fertilizer dosages can be optimized to avoid excess nutrients and reduce costs in forestry operations.

CONCLUSION
The greatest contribution of nutrients occurred through internal precipitation compared to contribution from incident rainfall and water flow through the trunk.Significant amounts of K + , Ca 2+ and Cl -entered via incident rainfall, and nutrient washing through the canopies (throughfall) is important for the enrichment of the aqueous solution and mineral fertilization of the forest stand.

13 Pi 12 Pi 13 Et 12 Et 13 mm
*Values followed by the same letter in the row for 2012 and 2013, considering the different precipitation flows, do not differ statistically by Student's t-test, at 5% error probability level.
*Values followed by the same capital letter in the row do not differ statistically in 2012 by Tukey's test, at a 5% error probability level; **Values followed by the same lowercase letter in the line did not differ statistically in 2013 by Tukey's test, at 5% error probability.

Table 3 .
Annual mean quantities (kg ha -1 ) of ions present in solution of incident rainfall (P), throughfall (Pi) and stemflow (Et) in 2012 ( 12 ) and 2013 ( 13 ), in a Eucalyptus dunnii stand, Alegrete/RS.Values followed by the same letter in the row for 2012 and 2013 considering the different precipitation flows did not differ statistically by Student's t-test, at 5% error probability level.