Throughfall and stemflow nutrient flux in deodar and oak forests, Garhwal Himalaya, India

A study to understand the throughfall and stemflow chemistry under deodar and oak forests of Garhwal Himalaya was conducted during 2014–2015. Total rainfall during experimental period was 1473.8 mm and estimated interception loss was 34.018% for deodar forest, 24.85% for oak forest. Stemflow represented the minimum proportion of gross rainfall, i.e. 0.321% in deodar forest and 0.463% in oak forest. pH of throughfall and stemflow in both deodar (6.087 and 6.47 respectively) and oak forests (6.75 and 7.03 respectively) was significantly more acidic than the gross rainfall (7.15). Electrical conductivity was recorded higher in deodar stemflow (231.89 μs/cm) and throughfall (102.75 μs/cm) compared to oak forest (172.92 μs/cm and 83.83 μs/cm respectively). Net nutrient leaching and stemflow were considerably higher from oak forest than deodar forest. Oak forest has better water yield capacity than deodar forest as the interception loss was higher in deodar forest. The idea of sustainable agriculture may be possible surrounding such forests as the need for chemical fertilizer and water can be met by the nutrient-rich soil, available soil moisture and surface water.


GRAPHICAL ABSTRACT INTRODUCTION
The nutrient cycle, similar to the hydrological cycle and energy flow involves the cycling of elements via biotic and geochemical components of ecosystems. The nutrient turnover rate is considered as one of the most important factors in the functioning and stability of ecosystem (Brasell et al. ). Precipitation is a significant nutrient source for forested ecosystems (Parker ). Quality of rainfall after passing through surface of trees is dramatically altered (Eaton et (Chuyong et al. ). There is very close link between nutrient cycle in forests and the hydrological cycle because, being the main solvent, water transports nutrient elements from aboveground tree stands to the soil underneath (Bruijnzeel ). An attempt was made to measure rainfall partitioning as throughfall, stemflow and interception loss under oak and deodar forests in the Garhwal Himalayan region. A specific objective was to determine nutrient concentration of rainfall components including gross incident rainfall, throughfall and stemflow as well as pathway characteristics of nutrient elements (addition by absorption/deduction by leaching) to soil under oak and deodar forests, two most important tree species of the region.

MATERIALS AND METHODS
The study was conducted on deodar (Cedrus deodara) and oak (Quercus leucotrichophora) trees in Garhwal Himalaya, India. As the study area, the Tarakeshwar sacred grove mixed with oak and pine. The area shows a typical monsoon climate, with recorded annual rainfall of 1,500-1,700 mm (Bisht ). The study areas (Tarkeshwar and its adjacent area) were divided into three plots each. Plots were selected in different locations within the study area so that they represented the whole study area. Each plot consisted of a 10 × 10 m area. In each plot three trees, making it nine trees from each forest types, were sampled.

ANALYSIS
Gross rainfall, throughfall and stemflow were studied on a rain event basis. For chemical analysis of samples, 16 rainfall events between June 2014 and May 2015 were considered. Gross rainfall was measured after each rainfall event in four collection flasks and two rain gauges, placed in adjoining large forest openings in the representative forest plots of deodar and oak. Rainfall events with a time gap between previous and following rain of at least 10 h to give time for the canopy to be dried out completely, were defined as individual rain events (Ahmadi et al. ). For throughfall measurement, a PVC pipes with an internal diameter of 120 mm were taken and divided lengthwise and then were connected with plastic tubing to 20 L containers (Cantu Silva & Gonzalez Rodriguez ). Three such collectors were installed under three individual trees of each plot. For stemflow sample collection, a stemflow collector was made (a tube-like structure made from aluminium sheet, which was then spiralled around the trunk of each tree) and fixed firmly with silicon sealant after smoothing the bark surface. The collection collar was installed at 1.37 m above ground level (Ahmadi et al. ) and stemflow was delivered to a 20 L container. In each forest type, the stemflow sample was collected from trees that had been selected for collection of throughfall samples. To estimate both throughfall and stemflow quantity, the crown projection area was calculated for individual selected trees , Cl À ) were measured by phenol di-sulfonic acid (for NO 3 À ), stannous chloride spectrophotometric method (for PO 4 À3 ) and by argentometric titration method (for Cl À ). To find out the existing trends in pH, electrical conductivity and the net and cumulative deposition of nutrient through gross rainfall, throughfall and stemflow for each forest type, a completely randomized design was used that included treatments for gross rainfall and two forest types (deodar and oak). One-way analysis of variance was attempted on all rain events for area weighted gross rainfall, throughfall, stemflow, net deposition, pH, Electrical conductivity was used with nutrient elements (Na þ , Ca þ2 , K þ, NO 3 À , PO 4 À3 , Cl À ) for testing the mean differences among forest plots. Differences among means were determined by Tukey's significant difference All analyses were done using the SPSS statistical software package (16.0 version).

RESULTS
Hydrological flux (volume of precipitation) Gross incident rainfall during study period was averaged as 1,473.8 mm. Throughfall, stemflow and interception loss as percent of gross incident rainfall is listed in Table 1.

Concentration of nutrients (mg L À1 ) in different rainfall components
Under deodar forest in different rainfall components quantities of phosphate (PO 4 À3 ), calcium (Ca þ2 ), chloride (Cl À ) and electrical conductivity were higher compared with oak forest and for nitrate (NO 3 À ), potassium (K þ ), sodium (Na þ ) and pH the scenario was opposite (  (Table 3).
Oak throughfall and stemflow were more nutrient rich than deodar, except phosphate, calcium, chloride and sodium in deodar throughfall. The average pH of rainwater was higher than pH for both throughfall and stemflow under both forests (Table 3).
Nutrient deposition (kg ha À1 year À1 ) by rainfall components Relative contribution of stemflow of each forest to the gross nutrient deposition is presented in Table 4. Mean total nutrient deposition by throughfall and stemflow were higher in oak than deodar forests. Net gain of nutrients (nitrate, calcium, potassium, sodium) from the canopy was higher in oak compared to deodar forests. PO 4 À3 and Cl À were absorbed by both forests (Table 4).

DISCUSSION
In oak-dominated forests, stemflow value has been recorded as <1% of the gross rainfall and was similar to other reported studies of similar sites (Table 5). This may be due to the thick epiphyte growth on the bark that may impede stemflow. Throughfall in oak-dominated forests was estimated as about 75% and was similar to recorded values for similar sites. Even throughfall in similar ranges has been recorded by several researchers for broad leaf forests (Table 5). Our estimated canopy interception was 24.85%, which was similar to the recorded value (17.7-32.4%) of some researchers at similar sites (Table 5). As found in a separate study, our results also revealed that canopy interception was higher in the coniferous canopy than in  These results depict the fact that leaching of nutrient occurs from both or either the bark or leaf tissues. It also indicates that nutrients were transported to the soil much  a Gain of nutrient from forest has been calculated as: (i) the difference between the amount of nutrient in rainfall and throughfall for a given volume of throughfall; and (ii) the difference between the amount of nutrient in rainfall and stemflow for a given volume of stemflow. b The negative sign stands for absorption of nutrient by forest stand.  Pathak et al. (1985). b Pathak & Singh (1984

CONCLUSION
Oak forest has better water yield capacity than deodar forest, as the interception loss was higher in deodar forest than oak forest. Climate and its changing conditions also have an influence on rainfall partitioning. In oak forest, nutrient deposition by means of throughfall on an area basis was more than for deodar forest. Perhaps these are the reasons that, in this part of Himalaya, people consider oak as a water conserving tree and there is no scarcity of water where oak forests are present. These are the ecosystem services provided by oak forest. Precipitation input plays a significant role in nutrient cycling as nutrients are found readily available in precipitation in both forests.
Based on the result it may be proposed that if local people start agricultural practices surrounding the area it may be beneficial from the view point of human health, as well as appropriate use of land. The idea may be successful as the practice will need less or no chemical fertilizer input for crop production. The idea of sustainable agriculture may be possible surrounding such forests, as the need for chemical fertilizer and water can be met by the nutrient-rich soil, available soil moisture and surface water.

ACKNOWLEDGEMENT
Financial assistance in the form of a Research Grant to N.P.
Todaria by the Ministry of Environment, Forest and Climate Change, Government of India is gratefully acknowledged.

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.