Incident Precipitation Partitioning : Throughfall , Stemflow and Canopy Interception in Eucalyptus dunnii Stand

Atmospheric deposition is responsible for the ions input, which may be due to dust and aerosols and rainfall. During rainfall a portion is intercepted by the tree canopy and returned to the atmosphere by evapotranspiration, another part crosses the forest canopy called throughfall and stemflow. The objective of the study was to quantify incident rainfall partitioning into throughfall, stemflow and canopy interception in a Eucalyptus dunnii stand in southern Brazil. Four plots of 20 m × 21 m were demarcated. The rainfall consists 3 rain collectors in an open area. The throughfall consisted 3 collectors per plot in the line, interlining and diagonal positions of the trees. The stemflow consisted in the installation of three systems per plot formed by a hose in the trunk of the tree that leads the solution to a reservoir. The adjustment of the throughfall, stemflow and canopy interception in function of the incident precipitation was of 99%, 90% and 52%. As the volume of rainfall increases, the coefficient of variation decreases. The annual rainfall was 1903 mm, with a canopy interception average of 8.9%.


Introduction
The understanding of rainwater partitioning is important for the hydrological balance, especially when dealing with large areas covered with vegetation (Rodriguez Suarez et al., 2014).One of the consequences of large areas being forested is the fact that a portion of the precipitation does not reach the soil.The fast-growing Eucalyptus sp genus has caused controversy regarding its water consumption and its impact on a microbasin (Jackson et al., 2005).Comparing two paired micro-basins, Reichert et al. (2017) concluded that the interception by canopy is greater in eucalyptus area than in natural grassland area and pasture During rain events, part of the rainwater that reaches the canopy is intercepted and the atmosphere will be returned later (Llorens & Domingo, 2007).The water that crosses the forest canopy and directly reaches the litter is called the throughfall (Navar, 2011).Another part descends through leaves, branches and flows through the trunk supplying with water and nutrients the roots (Zhang et al., 2016;Johnson & Lehmann, 2006).Some studies do not consider intercept by the canopy in their modeling, or often consider a generic value and low representativity (Chaffe et al., 2010).In regions of tropical climate, the percentage that is intercepted and returned to the atmosphere may be even more significant (Savenije, 2004).
The interception of water is influenced by biotic characteristics such as morphophysiological, height, contact surface of different plant species (Keim et al., 2006) and abiotic factors such as intensity, quantity, duration and size of the raindrop (Calder, 1999;Sulínski et al., 2001).Many factors such as vegetation cover, plant density, relative humidity, temperature, solar radiation, precipitation intensity, direction and wind speed, interfere with the evaluation of the interception by the canopy (Gerrits & Savenije, 2011;Bulcock & Jewitt, 2012;Levia & Frost, 2006;Hörmann et al., 1996), besides the seasonal variation (Salehi et al., 2016).As the rainfall and the degree of canopy closure increases, there is a decrease in the coefficient of variation for the throughfall (Carlyle-Moses et al., 2014;A. Zimmermann & B. Zimmermann, 2014;Thomaz & Antonelli, 2015).
Several studies point to the importance of studies of throughfall and stemflow under a biogeochemical aspect (Staelens et al., 2006;Fan et al., 2015), in addition they provide water to the plants (Bouillet et al., 2002).It is known that both the throughfall and the stemflow have great spatial variation depending on the forest type jas.ccsenet.(Llorens & (Levia et a In Brazil t 4.8 years throughfal

Charac
The study coordinate of 340.7 h sp., being implanted the trees w

Measurement of Precipitation, Throughfall, Stemflow a Canopy Interception
The monitoring of rainfall (P) occurred with the installation of 3 collectors with a diameter of 20 cm at 1.5 m from the ground level.They were allocated in a field area, distant about 50 m from the plots of the eucalyptus stand.Neoprene straps with galvanized steel wires were placed in order to avoid that they be used as perches for birds.
For the throughfall (Tf), 4 plots were demarcated inside the stand with dimensions of 20 m × 21 m.Then, 3 collectors with a diameter of 20 cm and height of uptake of 1 m of the soil level were distributed in the positions line, between line and diagonal of the trees in each of the plots.
The determination of the amount of the incident precipitation (P) and throughfall (Tf) in (mm) was obtained by the following equation: where, P (mm) = incident precipitation, Tf (mm) = throughfall, X (liters) = water accumulated 0.0314 (m²) = catchment area.
The quantification of the stemflow occurred with the installation of reservoirs with capacity of 60 liters in 3 trees of average BHD (breast height diameter) per plot.A 1-inch diameter, longitudinally cut hose, spirally mounted near the tree trunk, allowed the water flowing through the trunk to be stored in the reservoirs.
The canopy interception was determined by the equation used by Lima (1976): The canopy interception is determined indirectly.Is the difference of the incident precipitation by the sum of the throughfall with the stemflow.Data were collected biweekly over 12 months: where, Ci (mm) = canopy interception, P (mm) = Incident precipitation, Tf (mm) = throughfall, Sf (mm) = stemflow.

Statistics and Data Analysis
Statistical analysis was performed using IBM SPSS 20.0.The analysis of variance was performed in the data, assuming a completely randomized design with 3 treatments (rain partitions) and 12 repetitions (months), and the Tukey test was performed to compare the means (α = 0.05).
Regression equations were adjusted for throughfall, stemflow and canopy interception as a function of the incident precipitation variable.The distribution of the regression residues was then analyzed in order to validate the homogeneity of variance.The residues were presented in graphic form as a function of the variable analyzed.
The coefficient of variation (CV) is a well accepted statistical parameter (Zimmermann et al., 2007), and it was used to quantify the variability of the monthly rainfall indexes, as well as the influence of rainfall on the CV obtained: where, ∂ is the standard deviation of the collected volume and Φ is the volume average.

Distribution of Incident Precipitation, Throughfall, Stemflow, Canopy Interception
The annual precipitation was 1903 mm, with a monthly average of 159 mm.The month of September recorded the highest volume, 258 mm (p ≤ 0.05).April and August recorded the lowest volumes, 54 and 67 mm respectively, with no statistical difference between them (Table 1).
The measured throughfall was 1718 mm, that is, 90.3% of the incident rain crossed the canopy and reached the forest floor.The month of September likewise differed statistically from the others, and recorded the highest volume, 255 mm.The months of April and August did not differ statistically among themselves and accumulated the lowest indexes, 46 and 59 mm respectively.Evaluating the rainfall partition with Eucalyptus dunnii between 4 and 5 years of age, Dick et al. (2018) found an average of 91% for throughfall and 1% for stemflow.The stemflow recorded much lower volumes, with the accumulated total of 14.5 mm, which represents only 0.8% of the incident precipitation.Unlike the other rain partitions, the highest indexes were for the months of March, September, October and December, and did not differ statistically from one another.
The mean canopy interception was 8.9%.During the warmest 3 months the average interception was 12.3%, however, in the months representing autumn, the average interception by the canopies was only 5.4%.The explanation for this variation is possibly due to the degree of closure of the canopy.Before starting the season with milder temperatures, abscission of some of the leaves occurs, due to the greater shading and lower solar radiation, intercepting less amount of the incident rainfall.
Among the climatic variables, the one that most influenced the canopy interception was the sum of days without rain in the month with a negative correlation of 71% and probability of error of 0.8%.Then, the mean wind speed (m/s) was 66% and probability of error of 2.8%.Table 2 shows the pearson correlation between the intercept by the crown and the climatic variables.Note.ns = not significant; EVT = evapotranspiration.
Climatic variable source: AGRITEMPO (2018).Sadeghi et al. (2018) evaluated the dynamic interception of the incident precipitation and found a variation between 19 and 64% with the species Robinia pseudoacacia during the stage of leafless and full-leaf respectively.The authors also observed variation with the species Platanus orientalis (22 and 47%) for the same foliar stages respectively.

Table 2 .
Pearson correlation between climatic variables and canopy interception