Journal of Hydrology X 18 (2023) 100146

Contents lists available at ScienceDirect

Journal of Hydrology X
journal homepage: www.sciencedirect.com/journal/journal-of-hydrology-x

Research papers

Influence of alpine vegetation on water storage and discharge functions in

an alpine headwater of Northern Japan Alps
Mayu Fujino a, 1, Koichi Sakakibara a, *, Maki Tsujimura b, Keisuke Suzuki a
a
Faculty of Science, Shinshu University, Japan
b
Faculty of Life and Environmental Sciences, University of Tsukuba, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: As mountainous areas provide abundant water resources to lower elevations, and alpine zones are major
Alpine zone recharge areas for water resources, it is important to understand water storage and discharge processes in these
Land cover zones. Regarding water storage, sedimentary structures (e.g., talus and moraines) in alpine zones function as
Alpine vegetation
aquifers. However, the functions of vegetation, thought to contribute to water recharge and storage in forested
Water storage function
watersheds, have rarely been investigated. Accordingly, we evaluated the influence of alpine vegetation on water
Component separation
storage processes in alpine zones. Two intensive field surveys were conducted on August 17 and October 5, 2019,
in the alpine headwaters of Mt. Norikura in the Northern Japan Alps. Chemical analyses were conducted of
rainwater, snowmelt water, and runoff water from bare and vegetated catchments. From the results, a two-
component separation was conducted to calculate the contributions of precipitation and groundwater compo­
nents to runoff water. Our results implied that runoff water from vegetated catchments was in contact with the
regolith for longer, with the contribution of groundwater being higher in this runoff water. Moreover, the
groundwater component contribution tended to increase as the ratio of vegetation area to bare area in each
catchment increased, suggesting a higher water storage function for vegetated areas. In other words, the sub­
surface water flow should be slower in vegetated areas due to the presence of vegetated soils compared to bare
areas where coarse-grained sediments are dominant. Accordingly, the alpine vegetated area has a higher water
storage function than the alpine bare area.

1. Introduction predicted an increase in the population using water resources from

mountainous areas to 1.5 billion by 2050. Accordingly, understanding
Mountainous areas are regarded as “water towers” because of their the water environment and hydrological cycle in mountainous areas,
high water-potential and recognized as major water recharge areas, including alpine zones, is a crucial issue. However, relevant interna­
providing water to downstream areas with higher human activity tional studies are lacking owing to the difficulty of access and re­
(Viviroli et al., 2020). In Japan, precipitation generally increases along strictions to protect wilderness areas, leading to a knowledge gap about
with elevation (Suzuki et al., 2008), providing abundant rainwater to the water environment and water cycle in alpine areas (Clow et al.,
mountainous areas. Low temperatures often cause temporary storage of 2003).
precipitation (snow and ice) on the ground surface that delays the In alpine zones, water is stored temporarily in unconsolidated sedi­
conversion of precipitation to runoff, leading to a stable supply of ments unique to such areas (e.g., talus, moraine), afterward reaching the
downstream water resources during the snowmelt season (Cochand bedrock where a part of water infiltrates into the bedrock and the rest
et al., 2019). Precipitation and snowmelt in mountainous areas are discharges as runoff. Investigations of the spatial distribution of
crucial for sustainable livelihoods in arid and semi-arid regions, e.g., the groundwater using ground-penetrating radar, seismic refraction to­
Andean and Himalayan foothills that experience a dry season from mography, and electrical resistivity tomography in the headwaters of
snowmelt to summer (Barnett et al., 2005). Viviroli et al. (2020) the Rocky Mountains (e.g., McClymont et al., 2011; Christensen et al.,

Abbreviations: DEM, digital elevation model; EC, Electrical conductivity; TDS, total dissolved solids.
* Corresponding author at: Faculty of Science, Shinshu University, 3-1-1 Asahi, Matsumoto 390-8621, Japan.
E-mail address: k_sakaki@shinshu-u.ac.jp (K. Sakakibara).
1
Current affiliation: Graduate School of Life and Environmental Sciences, University of Tsukuba, Japan.

https://doi.org/10.1016/j.hydroa.2022.100146
Received 14 July 2022; Received in revised form 11 November 2022; Accepted 21 November 2022
Available online 13 December 2022
2589-9155/© 2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

2020) have revealed thin saturated zones on the surface of the bedrock. maintain base flow during dry (no-rainfall) periods. Forested watersheds
Such studies have confirmed the storage of water on the bedrock surface, have higher water storage functions than grassland watersheds accord­
particularly in its staircase shape and hollows. This information high­ ing to studies using the mixing model concept (Neill et al., 2011; Birch
lights the importance of bedrock surface topography for the ground­ et al. 2021). Sakakibara et al. (2019) have shown that old water stored in
water storage function in alpine zones. These studies confirmed that deeper subsurface areas discharges in response to rainstorm events,
impermeable layers (e.g., fine-grained sand, frozen soil, and rock gla­ suggesting that forests play a significant role in storing water. These
ciers) inside the talus and moraines prevent water infiltration, whereby water recharge and storage functions derive from forest processes, such
a thin, saturated zone forms on their surfaces. Water retention at the as interception, transpiration, and infiltration into soil and rock layers.
surface of fine-grained sand layers has also been observed in the alpine In zero-order basins containing relatively shallow soils, soil depth has a
catchment of the Tibetan Plateau by Hu et al. (2019), who investigated significant effect on water storage and yield of stormflow, and deeper
geologic structures employing core drilling and groundwater level soils delay hydrologic response to storms (Sidle et al., 1995; Sidle et al.,
monitoring. Hayashi (2020) indicated that the surface topography and 2000). The high infiltration capacity of forest soils is maintained by two
fracture structure of bedrock and the presence of impermeable layers are factors, namely a large amount of soil pore space, where water is stored
probably responsible for temporal water storage in sedimentary struc­ by the root system; and well-developed tree trunks and O-horizon that
tures. Such storage occurs despite the rapid infiltration of water in talus disperse raindrop energy at the ground surface, (Neary et al., 2009).
and moraines owing to the predominance of coarse-grained sediments. Several studies have focused on source watersheds in alpine zones,
Cochand et al. (2019) showed that unconsolidated sediments could such as the Rocky Mountains, European Alps, and Andes, where snow
contribute to reducing the water flow rate and increasing recharge to the and ice deposited in winter are the main water sources (Somers and
underlying bedrock groundwater. The combination of sedimentary McKenzie, 2020). The base rock is exposed in the headwaters of these
landforms acting as aquifers is considered to maintain the base flow of areas, with sparse vegetation cover. Therefore, few studies have evalu­
mountain rivers (Gordon et al., 2015; Chang et al., 2018). These studies ated the water cycle in catchments where sufficient rainfall occurs even
have emphasized the importance of assessing water storage and runoff in summer and vegetation is developed well. It is important to under­
processes in alpine zones. stand the hydrological cycle in regions that experience abundant rainfall
The considerable contribution of groundwater in alpine zones to and dynamic hydrological cycles occur even in summer, e.g., the Asian
maintaining the baseflow in downstream rivers has been estimated monsoon climate region (Sakakibara and Suzuki, 2022). Asfaha et al.
quantitatively by applying component separation using the mixing (2015) have shown that peak discharge during precipitation events is
model concept (e.g., Liu et al., 2004; Frisbee et al., 2011; Muir et al., correlated negatively with vegetation coverage, even in mountainous
2011; Engel et al., 2016; Chang et al., 2018; Schmieder et al., 2018; Zhou watersheds with steep slopes. Peña-Arancibia et al. (2019) have noted
et al., 2021). Liu et al. (2004) separated stream water into three com­ that the transformation of the hydrological cycle owing to land cover
ponents, namely surface flow (precipitation), baseflow, and talus water change is greater at smaller scales, such as individual headwater
in the headwater catchment of the Rocky Mountains. The tracer was set catchments. Alpine headwaters are susceptible to climate change, and
based on principal component analysis using EC, dissolved ions (Ca2+, concerns are being raised about changes in water circulation processes
Mg2+, Na+, SO2− 4 ), silicon (Si) concentrations, acid neutralization ca­ owing to upward shifts in vegetation distribution (e.g., Takahashi and
pacity (ANC), and the oxygen stable isotope ratio (δ18O). The contri­ Yoshida, 2009) and changes in snow and ice environments (e.g., Mes­
bution of each end-component to the total discharge during the serli et al., 2004). Such factors illustrate the importance of evaluating
snowmelt season (May–October) was calculated at 36 %, 28 %, and 36 the influence of vegetation on the hydrological cycle in alpine zones.
%, respectively. The contribution of groundwater (baseflow + talus However, little is known about how alpine shrub communities control
water) was high (64 %), emphasizing the importance of understanding water storage and runoff processes in alpine headwaters.
the exchange between surface water and groundwater in the alpine The aim of our study was clarifying whether alpine vegetated areas
zone. Schmieder et al. (2018) conducted component separation of runoff do function as water storage areas and determining which factors
water during snowmelt events in a high-elevation catchment in the contribute to such function. For this purpose, we compared the quality
Austrian Alps using EC and δ18O as tracers. They showed that the con­ of runoff water from bare and vegetated catchments in an alpine
tributions of groundwater, glacial melt water, and precipitation com­ headwater catchment.
ponents to runoff water were 49 %, 35 %, and 10 %, respectively. These
results indicated a higher contribution of groundwater than that of 2. Study site
glacial melt water. Muir et al. (2011) separated runoff water during
precipitation events into two components, namely new water (water The study area is Mt. Norikura, a stratovolcano located at the
from the event) and old water (water stored in the watershed prior to the southern end of the Northern Japan Alps (Fig. 1). The main peak of Mt.
event) in the headwater catchment of the Rocky Mountains, using δ18O Norikura is Kengamine (3026 m a.s.l.), and it has an alpine headwater
as tracer. The results showed that the contribution of old water was area. Mt. Norikura formed during the mid-Pleistocene to Holocene
70–90 %, indicating that groundwater flows out even during precipi­ period, and it is part of the Norikura volcanic belt, which extends
tation events. The high proportion of groundwater components in approximately 5 km from north to south (Nakano et al., 1987). Fallen
mountain rivers implies that groundwater stored in alpine headwaters tephra layers, consisting of clayey volcanic ash and scoriae, assumed to
probably contributes to maintaining the base flow of such rivers. This have been formed by eruptive activity, are distributed in the area. The
conclusion is supported by studies using the water balance (e.g., Gordon occurrence of the latest activity is estimated at approximately 9000
et al., 2015). years ago (Okuno et al., 1994). The mountains tops are covered with
Several studies in low-elevation areas consider differences in land andesite and dacite, mainly comprising plagioclase (55 %), pyroxene
cover as a controlling factor in water storage functions and runoff (15 %), orthopyroxene (10 %), and quartz (10 %) (Anazawa and
characteristics (Eltahir and Bras, 1996). Studies have also focused on Ohmori, 2005). According to Sakakibara and Suzuki (2022), the air
changes in water runoff processes associated with deforestation and the temperature varied between –19.9 ℃ and +20.0 ℃ from October 2018
regeneration and development stages of forests. Asano et al. (2004) to September 2019 and between +2.4 ℃ and +20.0 ℃ from July 2019
showed that the amount of bedrock-derived solute output to the stream to September 2019. In addition, Sakakibara and Suzuki (2022) has
was higher in vegetated catchments than in non-vegetated catchments. shown that the total precipitation over approximately-three months
In addition, employing discharge monitoring (e.g., Lana-Renault et al., (July–September 2019) was 1581.4 mm.
2011) and numerical simulations (e.g., Peña-Arancibia et al., 2019), The main land cover types are bare and vegetated areas. The bare
investigations have suggested that forests reduce peak discharge and area is covered with rock debris produced and transported by periglacial

2
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

Fig. 1. Map of the east slope of Mt. Norikura showing the topographical contours, land cover, and sampling locations (Data source of 1-m mesh DEM and aerial
photographs: the Matsumoto Erosion Control Office; Ministry of Land, Infrastructure, Transport and Tourism; Japan).

action. An overgrown snow valley is visible at the end of winter. The 3.2. Hydrological observation and chemical analysis
snowline retreats from the beginning of June to the end of October to an
altitude of approximately 2700 m. In vegetated areas, an alpine shrub Precipitation and stream water levels were observed at the gauging
community dominated by Pinus pumila is developed in the range station at 2600 m a.s.l. in the catchment with an area of 0.16 km2 from
2500–2850 m a.s.l., and Pinus pumila is distributed in patches from 2850 July 5 to October 5, 2019 (Fig. 1). Please refer to Sakakibara and Suzuki
m a.s.l. to the summit (3026 m a.s.l.) (Takahashi and Yoshida, 2009). (2022) for a detailed description of the methods used to obtain the
Various alpine plants grow here, such as lingonberry (Vaccinium vitis- precipitation and stream-level data and the results.
idaea) and koiwakagami (Schizocodon soldanelloides f. alpinus). In the Two intensive field surveys were conducted immediately after rain­
vegetated areas, the soil is developed well. Shibata et al. (1976) con­ fall events, namely on August 17, 2019 (163.4 mm/day precipitation on
ducted a study close to the study site at an elevation of 2750 m on the August 16) and October 5, 2019 (84.8 mm/day precipitation on October
eastern slope of Mt. Norikura, identifying A-horizon (0.15–0.40 m deep) 4). Water samples were collected at 18 sites (runoff water from a bare
and B-horizon (0.40–0.80 m deep) that consist of volcanic ash and catchment at 10 sites, runoff water from a vegetated catchment at 7
gravel. A thick O-horizon of approximately 0.10–0.20 m had formed sites, and snowmelt water at 1 site; Fig. 1). Runoff water from the talus
owing to low temperatures in the alpine zone beyond the tree line and deposit was sampled as runoff water from a bare catchment, and that
the interception of wind and solar radiation by alpine shrub commu­ from the Pinus pumila forest as runoff water from a vegetated catchment.
nities (Koizumi et al., 2018; Shibata et al., 1976). Several gullies occur in Runoff from both of these catchments persisted during the entire
this basin, and the valley bottoms of the gullies become streams with snowmelt season, while runoff only from the vegetated catchment
intermittent runoff after rainfall events. continued after snowmelt. During the study period, rainwater was
collected six times at the gauging station. However, one of the rain
3. Material and methods samples was eliminated from the analysis because the rain sampler had
been destroyed by strong winds or wild animals.
3.1. Calculation of vegetated area The water temperature was measured in situ using a digital water
temperature meter (ASF-270T; ASONE, Japan), and pH and EC of each
We analyzed data from a 1 m mesh digital elevation model (DEM) water sample were measured with a portable water quality meter (D-
and aerial photographs (Matsumoto Erosion Control Office, Ministry of 200–2; HORIBA, Japan). The samples were filtered using a 0.22 µm pore
Land, Infrastructure, Transport and Tourism). Watershed analysis was size syringe filter, and the concentrations of major inorganic dissolved
conducted using the hydrological analysis tool ArcGIS (ESRI, USA), and ions (F− , Cl− , NO–3, SO2− 2+ 2+
4 , Na , NH4 , K , Mg , Ca ) were analyzed by
+ + +

aerial photographs were extracted into the shape of the catchment area ion chromatography (ICS-1500/ICS-2100; DIONEX, USA). The bicar­
for each water sampling point. ImageJ, a public domain Java image bonate ion concentration (HCO–3) was analyzed by sulfuric acid titration
processing program, was used to measure the vegetation cover area in using the pH 4.8 alkalinity titration method. The ion balance, including
each extracted catchment shape. The extracted shape images were considering hydrogen ions, was less than 4.6 % for all water samples,
converted into grayscale images. Subsequently, one of the multiple in­ implying sufficient analytical accuracy.
flection points in the grayscale intensity histogram was set manually as
the threshold separating the vegetated and bare areas, referring to the
aerial photograph. We defined pixels with luminance lower than the 3.3. Component separation of runoff
threshold as vegetation areas and measured the number of pixels. The
number of pixels in the measured vegetation area was converted to an We assessed the water storage function in the vegetated area of an
area by referring to the overall number of pixels in the catchment area. alpine headwater basin. Because the degree of groundwater contribu­
tion to runoff water could reflect the water storage function, we sepa

3
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

rated the components of the runoff water. A simple runoff model was N6 and N8 are unclear in the watershed analysis, although it was
constructed. We defined water adequately affected by rock weathering confirmed that water flows from the base of vegetation. The vegetation
as “groundwater.” Rainwater falling in the basin and the meltwater of cover rates in the bare and vegetated catchments are 21.6 ± 12.8 % and
ice and snow originating from precipitation were defined as “precipi­ 77.4 ± 22.1 %, respectively. The calculated vegetation cover rate of the
tation.” Runoff water was assumed to consist of groundwater and pre­ vegetated catchments is higher than that of bare catchments, indicating
cipitation components. Then, we separated the runoff water into two that our classification method for runoff water is reasonable. However,
components, namely precipitation and groundwater, using the following the vegetation cover rate (46.7 %) for N10 (bare catchment) is higher
equations (Sklash and Farvolden, 1979). than that (45.7 %) for N15 (vegetated catchment). In the N10 catch­
ment, vegetation is distributed sparsely only on the ridges, and vege­
RGW + RP = 1 (1)
tation soils are likely less developed in this catchment than in other
vegetated catchments. Accordingly, classifying N10 as runoff water from
CGW RGW + CP RP = CD (2)
a bare catchment is not problematic.
where R is the contribution of each component to runoff, C is the tracer
value, and subscripts GW, P, and D represent groundwater, precipita­ 4.2. Hydro-chemical characteristics
tion, and discharge (runoff water), respectively.
We used the ratios of sodium ions to hydrogen ions (Na+/H+) and Fig. 2 shows the temporal variations in precipitation and stream
magnesium ions to hydrogen ions [Mg2+/(H+)2] as tracers in conducting water levels at the gauging station. Throughout the observation period,
the component separation. Anazawa and Ohomori (2005) theoretically the stream water level clearly responds to precipitation, rising imme­
showed that the following relationship holds for the amounts of dis­ diately after precipitation and dropping rapidly afterward. Precipitation
solved ions in water samples affected adequately by rock weathering in of 163.4 mm is observed on August 16, the day before the first intensive
the Mt. Norikura region, which includes the study area: field survey (August 17). The stream water level started rising at 19:50
[ ] on August 15, reaching its highest level (0.27 m) at 9:50 on August 16,
( +)
Mg2+ Na subsequently decreasing to 0 m at 11:50 on August 18. On October 4, the
log : log = 0.84 : 1 (3)
(H + )2
H+ day before the second intensive field survey (October 5), 84.8 mm of
precipitation was observed. The stream water level started rising at 9:40
This equation is useful for separating runoff water into groundwater a.m. on October 4, reaching its highest level (0.10 m) at 16:50 a.m. on
and precipitation components. Generally, Mg2+, Na+, and H+ concen­ the same day, and declining afterward.
trations in groundwater mainly change because of rock weathering. The relationship between pH and EC of the sampled water is shown
However, most minerals in our study area are feldspar and pyroxene, in Fig. 3. Rainwater and snowmelt water pH is 5.1–5.5, i.e., lower than
with extremely small dissolution rates (10− 12–10− 16 mol m− 2 s− 1), that of runoff water. The pH of runoff water from the bare catchments
depending on temperature and other conditions (Kim, 2002; White and (5.4–6.1) tends to be lower than that from vegetated catchments
Brantley, 2003). We collected the runoff water immediately after rain­ (5.6–6.5). Similar to pH, the EC of rainwater and snowmelt water is
fall events and we expected a brief interaction time between the pre­ lower (0.20–0.46 mS/m) than that of runoff water, and the EC of runoff
cipitation components contributing to the runoff and the regolith. In this water from bare catchment (0.20–0.74 mS/m) tends to be lower than
paper, we commonly refer to shallow bedrock, coarse-grained sediment, that of runoff water from vegetated catchments (0.38–1.18 mS/m).
and soil as regolith. The brief time scale of the two-component separa­ Runoff water from vegetated catchments has the highest pH and EC
tion and the conservative tracers used implied that the tracer values of among the watersheds.
the runoff water and each component would likely not change. Fig. 4 shows the water quality of rain-, snowmelt, and runoff water.
The five sites, including snowmelt water (N1, N3, N4, N5, and N8)
4. Results temporally dried up on October 5 and the figure does not show these
results. The water quality at each location remains unchanged between
4.1. Vegetated area August and October. The total dissolved ion concentration is clearly
higher in runoff water (78.0 ± 33.6 µeq/L) than in rainwater (16.0 ± 3.0
Table 1 lists the area and vegetation cover rates calculated by µeq/L) or snowmelt water (17.2 µeq/L), i.e., similar to the EC trend. The
watershed analysis using ArcGIS and image analysis. The catchments for dissolved ion concentrations of the runoff water in bare and vegetated
catchments are 73.2 ± 32.5 µeq/L and 90.7 ± 38.0 µeq/L, respectively,
i.e., runoff water from vegetated catchments has higher amounts of
Table 1 dissolved content than that from bare catchments.
Area and vegetation cover rate of each catchment. “-” indicates dried-up runoff
The water quality composition of runoff water from bare and vege­
water on sampling day.
tated catchments was characterized as types Ca-NO3 and Na(K)–HCO3,
Sampling point Catchment Bare area Vegetated Vegetation respectively. The monovalent cation (Na++K+) concentration relative to
(km2) (km2) area (km2) cover (%)
divalent cations (Ca2+, Mg2+) in runoff water appeared to be higher in
Runoff water N1 0.061 0.047 0.014 22.5 the vegetated catchments. In the bare catchments, HCO–3 had the lowest
from bare N2 0.0021 0.0018 0.00027 12.8
anion content but the highest in the vegetated catchments, except for
catchment N3 0.0065 0.0058 0.00071 11.0
N4 0.0021 0.0019 0.00017 8.4 N14 (October), N15 (August and October), and N16 (August). The
N7 0.15 0.12 0.029 19.5 highest anion content in the runoff (N14, N15, and N16) was SO2− 4 .
N9 0.0020 0.0017 0.00029 14.4
N10 0.15 0.081 0.071 46.7 4.3. Mixing of precipitation and groundwater
N11 0.13 0.083 0.050 37.5

The relationship between log (Na+/H+) and log [Mg2+/(H+)2] for all
Runoff water N6 ー ー ー ー
water samples is shown in Fig. 5. The solid line represents the theoretical
from N8 ー ー ー ー
vegetated N12 0.077 0.033 0.043 56.3 line indicating water quality derived from rock weathering, as clarified
catchment N13 0.044 0.0045 0.039 89.7 by Anazawa and Ohomori (2005). Samples of precipitation origin
N14 0.0010 <0.00001 0.0010 99.8 (rainwater and snowmelt water) plot far from the theoretical line,
N15 0.15 0.080 0.068 45.7 whereas some runoff water (N17 and N18) plot almost on the line. The
N16 0.010 0.00045 0.0096 95.5
other runoff water is distributed on a straight-line connecting water

4
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

Fig. 2. Temporal variations of precipitation and stream water level at the gauging station.

Fig. 4. Spatial distribution of rain-, snowmelt, and runoff water chemistry.

by a mixture of two components, i.e., precipitation and groundwater.

4.4. Contribution of precipitation and groundwater components to runoff

Fig. 3. Relationship between pH and EC of all water samples.
water

deriving from precipitation and some runoff water (N17 and N18). The
Table 2 shows the contribution of the groundwater component to
plots representing runoff water from bare catchments are located closer
runoff water, with Na+/H+ and Mg2+/(H+)2 as tracers. The magnitudes
to the precipitation samples. Most runoff water from vegetated catch­
of the values with Na+/H+ as tracer and Mg2+/(H+)2 as tracer do not
ments plots close to N17 and N18, located on the theoretical line.
match, but the overall trend is roughly consistent. However, comparing
Fig. 6 shows the relationship of tracer values for runoff water and
the standard deviation of each tracer value (Na+/H+ and Mg2+/(H+)2)
precipitation and groundwater components. As described in Section 3.3
in the groundwater component, the deviation of Mg2+/(H+)2 is smaller
(Method), the samples of precipitation origin (rainwater and snowmelt
than that of Na+/H+ (Fig. 6). This finding indicates that the tracer values
water) are defined as the “precipitation component,” and the runoff
of the groundwater component is more stable when Mg2+/(H+)2 is used
water plots on the theoretical line (Fig. 5) as the “groundwater
as tracer, implying Mg2+/(H+)2 is a suitable tracer for separating runoff
component”. The Na+/H+ values for the precipitation and groundwater
water components. The discussion proceeds using the results of the two-
components are 0.2 ± 0.2 and 49.2 ± 15.8, respectively (Fig. 6a). Their
component separation with Mg2+/(H+)2 as tracer.
Mg2+/(H+)2 values are 0.007 ± 0.003 and 18.7 ± 3.5, respectively
Fig. 7 shows the spatial distribution of each component contribution
(Fig. 6b). The Na+/H+ and Mg2+/(H+)2 values of runoff water,
to runoff water. The contribution of the groundwater component varies
excluding the two components, range from 0.8 to 38.9 and 0.06–11.9,
between 0.3 and 63.8 % (2.3 ± 3.8 % for runoff water from bare
respectively. Accordingly, the tracer values of all runoff water are be­
catchments, and 23.2 ± 22.3 % for runoff water from vegetated catch­
tween those of the two components. A hypothesis is formulated from this
ments). The groundwater component largely contributes to runoff water
finding, namely that runoff water in the study area could be explained
from vegetated catchments compared with that from bare catchments.

5
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

Table 2
Contribution of groundwater components to runoff water (percentage), with
Na+/H+ and Mg2+/(H+)2 used as tracers. “-” indicates unknown because runoff
was dried up on sampling day.
Na+/H+ Mg2+/(H+)2

August October August October

Runoff water from bare N1 14.2 ー 3.6 ー

catchment N2 3.2 7.2 0.4 1.3
N3 3.6 ー 0.5 ー
N4 1.1 ー 0.4 ー
N7 15.6 29.1 5.1 14.6
N9 6.4 7.6 1.0 1.0
N10 6.1 3.9 0.8 0.3
N11 5.0 6.7 0.6 0.7

Runoff water from vegetated N6 17.2 31.4 6.1 12.5

catchment N8 13.9 ー 4.6 ー
N12 17.3 25.0 6.0 8.5
N13 47.3 31.4 29.4 15.5
N14 5.9 5.3 0.9 0.8
N15 79.0 65.1 57.9 44.5
N16 65.4 62.6 51.4 63.8

Fig. 5. Relationship between log (Na+/H+) and log [Mg2+/(H+)2] of all sam­
ples with the theoretical line formed by weathering of rocks in the study area
(clarified by Anazawa and Ohomori, 2005).

Fig. 7. Spatial distribution of precipitation and groundwater components

contribution to runoff water.

contribution rate in October being 12.4 % higher than that in August.

5. Discussion

5.1. Water quality formation and water runoff processes in alpine zone

The rapid response of stream levels to rainfall events in the catch­

ment with an area of 0.16 km2 (Fig. 2) indicates that rainwater is dis­
charged quickly after rainfall events. This characteristic can be seen in
other small catchments in Japan (e.g., Sidle et al., 1995; Katsuyama
et al., 2000). Additionally, recent studies have revealed that ground­
water in alpine zones contributes to runoff water (e.g., Liu et al., 2004;
Fig. 6. Relationship of tracer values among runoff water, precipitation Zhou et al., 2021). Accordingly, this study assumed that runoff water
component, and groundwater component when (a) Na+/H+ and (b) Mg2+/ forms from the mixing of precipitation and groundwater components.
(H+)2 are used as tracers. We separated the runoff water into precipitation and groundwater
components. The calculated groundwater component contribution to
The groundwater contribution is particularly high at N15 and N16 (54.4 runoff water was 0.3–63.8 % and higher in vegetated catchments (23.2
± 7.2 %), where SO2− ± 22.3 %) than in bare catchments (2.3 ± 3.8 %). This result indicated
4 is high. In comparison, the contribution of the
groundwater component to the runoff water is higher in October than in that vegetated and bare catchments have different groundwater storage
August. This difference is particularly large at N16, with the functions and/or discharge processes. Fig. 8 explains the effect of the

6
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

in the runoff water from bare and vegetated catchments, respectively.

Generally, divalent cations in groundwater are replaced by monovalent
cations according to the strength of adsorption of basic cations on rocks
and soils (Ca2+ > Mg2+ > K+ > Na+), along with the progress in the
rock–water interaction and increase in the proportion of monovalent
cations. As shown in Fig. 4, the cation composition of runoff water from
bare catchments is characterized by Ca2+, whereas that from vegetated
catchments is characterized by monovalent cations (Na++K+). Runoff
with a higher ratio of monovalent cations (Na++K+) to divalent cations
(Ca2+ and Mg2+) is more common in runoff water from vegetated
catchments. These results suggest that the quality of runoff water from
vegetated catchments could reflect the effects of water quality evolution
resulting from longer contact with the regolith. These interpretations
support the results of the two component separation, which indicates
high contribution from the groundwater component to the runoff water
from vegetated catchments. Even in a small groundwater flow field, such
as the headwaters of an alpine zone, vegetation cover slows the
discharge of precipitation and snowmelt water and increases water-
geologic contact time, and water quality could be affected by its evo­
lution through cation exchange reactions. In other words, faster flows
through coarse-grained sediments may be more dominant in bare areas,
while slower subsurface flows with long times of contact with soil and
bedrock may be more dominant in vegetated areas.

5.2. Role of alpine vegetation in water cycle

As suggested in Section 5.1, the groundwater component stored in

the subsurface zone in vegetated areas contributes to runoff water and,
here, we discuss how alpine vegetation affects the water cycle in alpine
zones. The most obvious difference between bare and vegetated areas is
the presence or absence of vegetation, and the most significant effect on
the water cycle is the difference in the amount of water reaching the
ground surface. In bare areas, almost 100 % of precipitation reaches the
Fig. 8. Relationship between the ratio of vegetated area to bare area and the
contribution of groundwater components to runoff water. ground surface; however, in vegetated areas, part of precipitation
evaporates before reaching the ground surface owing to canopy inter­
ception (Toba and Ohta, 2005). The canopy density of the dominant
presence or absence of vegetation, relationship between the ratio of the
Pinus pumila in the vegetated area of the study site is high at more than 5
vegetated area to the bare area, and the contribution of the groundwater
m2/m− 2(− |-) (Kajimoto, 1989). This is sufficient to effectively trap
component. A trend is detected of the larger the ratio of vegetated area
precipitation in the canopy, causing abundant water evaporation.
to bare area, the higher would be the groundwater component contri­
Uehara and Kume (2012) calculated a canopy interception rate of 47 %
bution, excluding N14 and N15. This finding suggests that land cover
based on precipitation, including fog water and rainfall, in Pinus pumila
conditions affect the magnitude of the groundwater component contri­
forests in the Mt. Tateyama area of the Northern Japan Alps. The canopy
bution to runoff water. Accordingly, such effects could be discussed in
interception rate also depends on meteorological conditions and
terms of water quality.
geographical factors. However, if the canopy interception rate of the
The basic water quality values (e.g., pH, EC, and total dissolved ion
Pinus pumila forest in the study area followed the results of Uehara and
concentrations) of runoff water from vegetated catchments are higher
Kume (2012), the water supply to the ground surface in the vegetated
than those from bare catchments (Figs. 3 and 4). Previous studies have
catchments would be approximately half that of bare catchments. If
shown that the longer the contact time between water and regolith, the
transpiration from Pinus pumila is considered, the total amount of water
higher would be the dissolved ion concentration in water owing to
that could contribute to groundwater recharge and runoff would be even
dissolution from the regolith (Baraer et al., 2014). As the dissolved ion
lower. Therefore, the groundwater component cannot be assumed to be
concentration increases, EC increases. Moreover, Japanese forest soils
stored in the vegetated area by considering only the relationship be­
have a high acid buffering capacity, and basic cations adsorbed on
tween the contribution rate of the groundwater component and the
minerals are replaced with H+ in aqueous solutions, resulting in a lower
presence or absence of vegetation.
H+ concentration in the solution (Ohte and Tokuchi, 1999). Clay min­
Studies have indicated that land cover and land use conditions
erals with cation exchange groups are formed along with the progression
significantly affect the hydrologic cycle, particularly forests that in­
of rock weathering. Anazawa and Ohmori (2005) considered that clay
crease water source recharge and delay runoff (e.g., Lana-Renault et al.,
minerals, such as kaolinite, are formed near the summit of Mt. Norikura.
2011; Peña-Arancibia et al., 2019). Forest functions derive mainly from
The EC and pH, and dissolved ion concentration are expected to increase
forest soils with high permeability and water retention ability (Asfaha
with the progression of water–geologic interactions because of the
et al., 2015). In the Pinus pumila forests of the study area, soil develop­
dissolution of basic cations from the regolith and ion exchange in clay
ment was visually confirmed at the time of the field surveys and we
minerals. Accordingly, the higher pH, EC, and dissolved ion concen­
focused our discussion on the forest soil horizons. Outcrop observations
trations in the runoff water from vegetated catchments compared with
revealed a O-horizon and relatively thick soil horizons in the vegetated
that from bare catchments could be attributed to the higher contribution
areas. The study area is a low-temperature alpine zone beyond the tree
of water with advanced water–geologic interactions in the runoff water
line, and the effects of wind and solar radiation are limited in the Pinus
from vegetated catchments.
pumila forest, making it difficult for the temperature to rise and litter to
We observed Ca-NO3 and Na(K)–HCO3 types of water quality mostly
decay. Consequently, a strongly undecayed O-horizon of approximately

7
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

0.10–0.20 m has formed (Shibata et al., 1976; Koizumi et al., 2018). As 5.3. Uncertainties of this study, and future work
the study area is located within a national park, we were unable to
conduct cross-sectional soil or other surveys that could detrimentally The presence of vegetation results in water quality formation pro­
alter the environment. Therefore, the information on the soil horizons cesses that differ from those of bare areas. The dissolved ion concen­
presented here derives from a previous study (Shibata et al., 1976), tration in runoff water is probably controlled by leaching from soil
which was conducted in the vicinity of the study site (at an elevation of particles, by uptake of plant roots, and by leaching of deposits and nu­
2,750 m on the eastern slope of Mt. Norikura). A-horizon in the Pinus trients accumulated on plant tissue surfaces via precipitation in vege­
pumila forest area is 0.15–0.40 m deep, and B-horizon, comprising vol­ tated areas. Therefore, it is necessary to consider the influence of such
canic ash and gravel, is 0.40–0.80 m deep. An aggregate structure of processes on the values of Na+/H+ and Mg2+/(H+)2, used as conserva­
0.2–2 mm accounts for approximately 30 % in A-horizon, and one of tive tracers in this study. Shibata (1976) showed that the pH of
more than 2 mm accounts for 50 % in B-horizon. The situation 0.80 m throughfall in Pinus pumila forests (pH: 3.81) is approximately 1.0 lower
below the surface is not confirmed yet as surveying has not been con­ than that of rainfall in bare areas (pH: 4.81) during the same period on
ducted. However, the vegetated area is adjacent to the bare area con­ Mt. Norikura. The interpretation is that the pH of water passing through
sisting of bedrock of Mt. Norikura and debris layers; therefore, a the forest canopy is probably lower than before passing through it. This
reasonable assumption is that soil horizons several meters deep has not could be ascribed to H+ leaching from the vegetation, implying that
formed. vegetation added 138.01 µeq/L of H+ to water deriving from precipi­
The thick O-horizon and the aggregate structure of the soil indicate tation. According to Uehara (2015), the Na+ concentration of precipi­
that a highly permeable soil horizons has formed. The thick O-horizon tation is 5.5 µeq/L and that of throughfall in Pinus pumila forests is 3.8
disperses raindrop energy (Neary et al., 2009) and is thought to prevent µeq/L on Mt. Tateyama in the Northern Japan Alps. These results could
crusts forming that could reduce the infiltration capacity. The O-horizon also indicate that Na+ uptake by the aboveground portion of the Pinus
probably also helps prevent soil detachment caused by raindrops and pumila forest is 1.7 µeq/L. Therefore, the Na+/H+ values in precipitation
surface runoff, thereby reducing soil erosion (Miyata et al., 2009). would decrease by 0.00009 before reaching the forest floor. The
Accordingly, it could be assumed that precipitation reaching the forest leaching of Mg2+ from the aboveground portion of the Pinus pumila
floor of the Pinus pumila forest easily infiltrates the soil horizons, despite forest is 5.3 µeq/L (Uehara, 2015), suggesting that the Mg2+/(H+)2
the high canopy interception rate. Infiltrated water flows more slowly value increases by 0.0003 due to vegetation. The tracer values of the
through soil particles than through the debris layers in bare areas, where runoff water samples, excluding the precipitation and groundwater
the grain size is larger. This difference in the rate of subterranean water components, ranged from 0.8 to 38.9 for Na+/H+ and from 0.06 to 11.9
flow is reflected in the water storage function in the alpine vegetation for Mg2+/(H+)2. This finding suggests negligible influence from vege­
area. Flow in several bare catchments ceases in October when the tation on the water quality before rainfall reaches the forest floor.
snowmelt season ends, whereas runoff from vegetated catchments The adsorption strength of basic cations on soil is generally in the
mostly does not (Figs. 4 and 7). This indicates that the soil horizons order Ca2+ > Mg2+ > K+ > Na+. Because Na+ is less easily adsorbed to
enables precipitation to be stored temporarily underground. Rock- soil than other cations, Na+ concentration is considered a strong indi­
derived elements, such as Na+ and Mg2+ are probably added to the cator of dissolution through rock weathering. The results of the two-
stored water, i.e., the groundwater component we defined, which mixes component separation, using the ratio of Na+ to H+ (Na+/H+) as
with rainwater and is discharged. tracer, showed that the contribution of the groundwater component was
Studies have revealed that sedimentary structures, such as talus and higher in the vegetated catchments (Table 2). Comparing these results
moraines, function as aquifers in alpine zones (e.g., Clow et al., 2003; with those obtained when Mg2+/(H+)2 was used as tracer showed
Muir et al., 2011; Christensen et al., 2020). Component separation consistent trends, although the magnitudes of the values differed.
methods have been applied to runoff from alpine headwater basins of Therefore, the Mg2+/(H+)2 value of runoff water is probably not easily
various sizes (2.25–4210 km2) previously (e.g., Liu et al., 2004; Frisbee affected by the water quality formation processes occurring because of
et al, 2011; Engel et al., 2016; Chang et al., 2018; Schmieder et al., 2018; vegetation soil horizons (e.g., ion exchange in clay minerals).
Zhou et al., 2021). The contribution of the groundwater component to Water quality changes occur in the ground owing to rock weathering.
runoff water during snowmelt is calculated at 14–60 %. Compared with The order of stability of minerals to weathering was shown by Goldich
this value, many runoff waters in this study area showed a lower (1938). According to the report, the order of susceptibility to weathering
groundwater component contribution, probably owing to heavy pre­ among colored minerals is olivine, pyroxene, hornblende, and biotite.
cipitation in the watershed. According to Cowie et al. (2017), the Among the colorless minerals, the order is Ca-plagioclase, Na-plagio­
contribution of each component to runoff varies depending on the clase, potassium feldspar, and quartz. The dissolution rates of plagio­
amount and timing of snowmelt and precipitation supplies, as well as clase and pyroxene, common in the study area, have been estimated at
groundwater storage capacity. The annual precipitation range in the 10− 10.9–10− 11.4 mol Si m− 2 s− 1 and 10− 9.9–10− 10.6 mol Si m− 2 s− 1,
watersheds is mainly 100 mm to 1500 mm. However, this study area respectively (Fujimoto, 1991). Plagioclase and pyroxene are most sus­
receives considerable rainfall, even during the short summer season ceptible to weathering, but their interaction with water occurs at low
(1581 mm/3 months during the study period). The catchment area in velocity. Shikazono and Fujimoto (1996) indicated that the dissolution
our study is small (0.0010–0.15 km2) and just below the summit. rate of silicate minerals under natural conditions is two to three orders of
Smaller catchments that receive large rainfall inputs have relatively magnitude smaller than the experimentally estimated dissolution rate.
small storage capacities compared to other studies’ catchments. The According to Kim (2002) and White and Brantley (2003), the dissolution
trend of higher groundwater contribution in October than in August rate of plagioclase under natural conditions is 10− 12 –10− 16 mol Si m− 2
could be attributed to less precipitation in October. s− 1. In our study, the runoff water was sampled one day after the rainfall
The role of vegetation in water storage in alpine zones has not been events, implying we dealt with a short time scale of one rainfall event.
evaluated before. However, this study suggests that soils formed by the Therefore, it is unlikely that water in the catchments deriving from
development of alpine vegetation could temporarily store water in the rainfall could be affected by rock weathering or the water quality
alpine zone and delay rainwater discharge based on water quality, significantly changed, as the water discharged in a short time span.
components of runoff water, and soil horizons. In other words, this study However, we found that runoff water from vegetated catchments had
suggests that the presence of alpine vegetation could enhance storage higher cation concentrations than that from bare catchments, suggesting
capacity in alpine zones and that influence of vegetation should be that water retained in the soil for a long time is affected by water–rock
evaluated in the water cycle of alpine zones. interactions and contributes to runoff in vegetated areas.
In addition to the presence or absence of vegetation, other factors

8
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

may also affect the contribution of groundwater components. Previous Data availability
studies have shown that the characteristics of rainfall and geologic
structures could be such factors in headwater catchments (Onda et al., Data will be made available on request.
2006; Sriwongsitanon and Taesombat, 2011). As Sidle et al. (2000)
revealed, changes in soil moisture contents alter runoff pathways, this is Acknowledgments
possible that two catchments with different sediment structures in this
study have different runoff pathways. In Mt. Norikura, according to Asai This work was supported financially by the Japan Society for the
et al. (2001), groundwater flows mainly along the geological structure Promotion of Science (KAKENHI Grants 19K13433 and 22K13246).
caused by lava flow in the downstream area. Additionally, runoff was
observed even after the end of the snowmelt season despite the narrow
and unclear catchment boundary of N6 and high dissolved ion concen­ Funding
trations, especially sulfate, in N15 and N16. This evidence may indicate
water circulation deeper within the subsurface with water flowing Japan Society for the Promotion of Science (JSPS).
across topographic catchment boundaries. In the future, to improve
understanding of the hydrological cycle in the alpine zone, quantitative References
evaluation based on field surveys (e.g., discharge rate and residence
Anazawa, K., Ohomori, H., 2005. The hydrochemistry of surface waters in andesitic
time of water), as well as an understanding of the geological structure volcanic area, Norikura volcano, central Japan. Chemosphere 59, 605–615. https://
could play important roles. doi.org/10.1016/j.chemosphere.2004.10.018.
Asai, K., Tsujimura, M., Satake, H., 2001. Groundwater flow system and rainfall-runoff
processes in the east slope of Mt. Norikura (in Japanese). Journal of Japanese
6. Conclusions Association of Hydrological Sciences 31 (4), 4_135.
Asano, Y., Ohte, N., Uchida, T., 2004. Sources of weathering-derived solutes in two
We collected and chemically analyzed runoff water from bare and granitic catchments with contrasting forest growth. Hydrological processes 18,
651–666. https://doi.org/10.1002/hyp.1260.
vegetated catchments in alpine headwaters. The effect of vegetation on
Asfaha, T., Frankl, A., Haile, M., Zenebe, A., Nyssen, J., 2015. Determinants of peak
water storage in alpine zones was discussed, using chemical data and discharge in steep mountain catchments–Case of the Rift Valley escarpment of
component separation results. Northern Ethiopia. Journal of Hydrology 529, 1725–1739. https://doi.org/10.1016/
j.jhydrol.2015.08.013.
Runoff water from bare catchments showed mainly Ca-NO3-type
Baraer, M., McKenzie, J., Mark, B., Gordon, R., Bury, J., Condom, T., Gomez, J., Knox, S.,
water quality, whereas that from vegetated catchments showed Na(K)– Fortner, S., 2014. Contribution of groundwater to the outflow from ungauged
HCO3 type. The water quality composition clearly differed between bare glacierized catchments: A multi-site study in the tropical Cordillera Blanca, Peru.
and vegetated catchments, suggesting that runoff water undergoes Hydrological Processes 29, 2561–2581. https://doi.org/10.1002/hyp.10386.
Barnett, T., Adam, J., Lettenmaier, D., 2005. Potential impacts of a warming climate on
different water quality formation processes depending on the presence water availability in snow-dominated regions. Nature 438, 303–309. https://doi.
or absence of vegetation. Runoff water was collected after precipitation org/10.1038/nature04141.
events when the stream levels rose. All parameters (dissolved ion con­ Birch, A., Stallard, R., Barnard, H., 2021. Precipitation characteristics and land cover
control wet season runoff source and rainfall partitioning in three humid tropical
centrations, EC, and pH) were higher in the runoff than in the rainwater. catchments in central Panama. Water Resources Research 57, WR028058. https://
These results suggested that precipitation contributes to runoff water, doi.org/10.1029/2020WR028058.
but leached elements are added due to interaction with the regolith. Chang, Q., Ma, R., Zhou, A., Hu, Y., Liu, Y., 2018. Using isotopic and geochemical tracers
to determine the contribution of glacier-snow meltwater to streamflow in a partly
However, in a short time period, elemental leaching from the regolith glacierized alpine-Gorge catchment in northeastern Qinghai-Tibet plateau. Journal
was most likely not sufficient to alter the quality of precipitation. of Geophysical Research: Atmospheres 123, 10037–10056. https://doi.org/
Accordingly, we believe that groundwater is discharged and mixed with 10.1029/2018JD028683.
Christensen, C.W., Hayashi, M., Bentley, L.R., 2020. Hydrogeological characterization of
the precipitation component, forming the quality of runoff water. We
an alpine aquifer system in the Canadian Rocky mountains. Hydrogeology Journal
separated runoff water into precipitation and groundwater components 28 (5), 1871–1890.
using chemical tracers (e.g., Mg2+/(H+)2), which reflect information on Clow, D., Schrott, L., Webb, R., Campbell, D., Torizzo, A., Dornblaser, M., 2003. Ground
water occurrence and contributions to streamflow in an alpine catchment, Colorado
water that has been affected sufficiently by rock weathering. Our results
Front Range. Groundwater 41, 937–950. https://doi.org/10.1111/j.1745-
indicated that the contribution of groundwater to runoff ranged from 0.3 6584.2003.tb02436.x.
to 63.8 % and was clearly higher in runoff water from vegetated Cochand, M., Christe, P., Ornstein, P., Hunkeler, D., 2019. Groundwater storage in high
catchments (23.2 ± 22.3 %) than from bare catchments (2.3 ± 3.8 %). alpine catchments and its contribution to streamflow. Water Resources Research 55,
2613–2630. https://doi.org/10.1029/2018WR022989.
The relationship between the contribution of the groundwater compo­ Cowie, R., Knowles, J., Dailey, K., Williams, M., Mills, T., Molotch, N., 2017. Sources of
nent and the ratio of vegetated area to bare area in each catchment streamflow along a headwater catchment elevational gradient. Journal of Hydrology
showed a positive correlation. These results suggested that the higher 549, 163–178. https://doi.org/10.1016/j.jhydrol.2017.03.044.
Eltahir, E.A.B., Bras, R.L., 1996. Precipitation recycling. Precipitation recycling. Reviews
the percentage of vegetation cover, the greater would be the ground­ of geophysics 34 (3), 367–378.
water contribution to runoff. Our finding implies that alpine vegetation, Engel, M., Penna, D., Bertoldi, G., Dell’Agnese, A., Soulsby, C., Comiti, F., 2016.
not traditionally recognized for its water storage ability, could tempo­ Identifying run-off contributions during melt-induced runoff events in a glacierized
alpine catchment. Hydrological Processes 30, 343–364. https://doi.org/10.1002/
rarily hold water in the subsurface area. It also implies the need to assess hyp.10577.
the presence of vegetation to understand the hydrological cycle in Frisbee, M., Phillips, F., Campbell, A., Liu, F., Sanchez, S., 2011. Streamflow generation
mountainous areas, especially in alpine zones, which supply water re­ in a large, alpine watershed in the southern Rocky Mountains of Colorado: Is
streamflow generation simply the aggregation of hillslope runoff responses? Water
sources to downstream areas. The discussion was deepened regarding
Resources Research 47, WR009391. https://doi.org/10.1029/2010WR009391.
the delay between water infiltration into the soil horizons and discharge Fujimoto, K., 1991. Rate and mechanism of dissolution of silicate minerals into aqueous
in the alpine vegetated zone. solution (in Japanese). Mining Geology 41, 163–172.
Goldich, S., 1938. A study in rock-weathering. The Journal of Geology 46, 17–58.
https://doi.org/10.1086/624619.
Declaration of Competing Interest Gordon, R., Lautz, L., McKenzie, J., Mark, B., Chavez, D., Baraer, M., 2015. Sources and
pathways of stream generation in tropical proglacial valleys of the Cordillera Blanca,
The authors declare that they have no known competing financial Peru. Journal of Hydrology 522, 628–644. https://doi.org/10.1016/j.
jhydrol.2015.01.013.
interests or personal relationships that could have appeared to influence Hayashi, M., 2020. Alpine hydrogeology: The critical role of groundwater in sourcing the
the work reported in this paper. headwaters of the world. Groundwater 58, 498–510. https://doi.org/10.1111/
gwat.12965.
Hu, Y., Ma, R., Wang, Y., Chang, Q., Wang, S., Ge, M., Bu, J., Sun, Z., 2019. Using
hydrogeochemical data to trace groundwater flow paths in a cold alpine catchment.
Hydrological Processes 33, 1942–1960. https://doi.org/10.1002/hyp.13440.

9
M. Fujino et al. Journal of Hydrology X 18 (2023) 100146

Kajimoto, T., 1989. Aboveground biomass and litterfall of Pinus pumila scrubs growing Sakakibara, K., Suzuki, K., 2022. Controlling factors and characteristics of peak runoff in
on the Kiso mountain range in central Japan. Ecological Research 4, 55–69. https:// an alpine headwater under the Asian monsoon climate. Mountain Research and
doi.org/10.1007/BF02346943. Development 42, R1–R8. https://doi.org/10.1659/MRD-JOURNAL-D-21-00030.1.
Katsuyama, M., Ohte, N., Uchida, T., Asano, Y., Kimoto, A., 2000. Effects of the Sakakibara, K., Tsujimura, M., Iwagami, S., Sato, Y., Nagano, K., Onda, Y., 2019.
differences of hydrological processes on the streamwater chemistry (in Japanese). Effectivity of dissolved SF6 tracer for clarification of rainfall–runoff processes in a
Journal of Japan Society of Hydrology and Water Resources 13, 227–239. https:// forested headwater catchment. Hydrological Processes 33, 892–904. https://doi.
doi.org/10.3178/jjshwr.13.227. org/10.1002/hyp.13398.
Kim, K., 2002. Plagioclase weathering in the groundwater system of a sandy, silicate Schmieder, J., Garvelmann, J., Marke, T., Strasser, U., 2018. Spatio-temporal tracer
aquifer. Hydrological Processes 16, 1793–1806. https://doi.org/10.1002/hyp.1081. variability in the glacier melt end-member—How does it affect hydrograph
Koizumi, T., Hattori, M., Nara, K., 2018. Ectomycorrhizal fungal communities in alpine separation results? Hydrological processes 32, 1828–1843. https://doi.org/10.1002/
relict forests of Pinus pumila on Mt. Norikura. Japan. Mycorrhiza 28, 129–145. hyp.11628.
https://doi.org/10.1007/s00572-017-0817-5. Shibata, O., Kato, K., Arai, M., Kinoshita, T., 1976. On the micro environment in Mt.
Lana-Renault, N., Latron, J., Karssenberg, D., Serrano-Muela, P., Regüés, D., Norikura, the central mountain region of Japan (in Japanese). Journal of the Faculty
Bierkens, M., 2011. Differences in stream flow in relation to changes in land cover: A of Science Shinshu University 11, 25–37.
comparative study in two sub-Mediterranean mountain catchments. Journal of Shikazono, N., Fujimoto, K., 1996. Dissolution rates of silicate minerals in aqueous
Hydrology 41, 366–378. https://doi.org/10.1016/j.jhydrol.2011.10.020. solution − a comparison of experimental and field results and interpretation of
Liu, F., Williams, M., Caine, N., 2004. Source waters and flow paths in an alpine groundwater chemistry by a coupled fluid flow− dissolution kinetics model (in
catchment, Colorado Front Range. United States. Water Resources Research 40, Japanese). Chikyukagaku (Geochemistry) 30, 91–97. https://doi.org/10.14934/
W09401. https://doi.org/10.1029/2004WR003076. chikyukagaku.30.91.
McClymont, A., Roy, J., Hayashi, M., Bentley, L., Maurer, H., Langston, G., 2011. Sidle, R., Tsuboyama, Y., Noguchi, S., Hosoda, I., Fujieda, M., Shimizu, T., 1995.
Investigating groundwater flow paths within proglacial moraineusing multiple Seasonal hydrologic response at various spatial scales in a small forested catchment,
geophysical methods. Journal of Hydrology 399, 57–69. https://doi.org/10.1016/j. Hitachi Ohta, Japan. Journal of Hydrology 168, 227–250. https://doi.org/10.1016/
jhydrol.2010.12.036. 0022-1694(94)02639-S.
Messerli, B., Viviroli, D., Weingartner, R., 2004. Mountains of the world: Vulnerable Sidle, R., Tsuboyama, Y., Noguchi, S., Hosoda, I., Fujieda, M., Shimizu, T., 2000.
water towers for the 21st century. Ambio Special Report 13, 29–34. https://www. Stormflow generation in steep forested headwaters: a linked hydrogeomorphic
jstor.org/stable/25094585. paradigm. Hydrological Processes 14, 369–385. https://doi.org/10.1002/(SICI)
Miyata, S., Kosugi, K., Gomi, T., Mizuyama, T., 2009. Effects of forest floor coverage on 1099-1085(20000228)14:3%3C369::AID-HYP943%3E3.0.CO;2-P.
overland flow and soil erosion on hillslopes in Japanese cypress plantation forests. Sklash, M., Farvolden, R., 1979. The role of groundwater in storm runoff. Journal of
Water Resources Research 45, W06402. https://doi.org/10.1029/2008WR007270. Hydrology 43, 45–65. https://doi.org/10.1016/0022-1694(79)90164-1.
Muir, D., Hayashi, M., McClymont, A., 2011. Hydrological storage and transmission Somers, L., McKenzie, J., 2020. A review of groundwater in high mountain
characteristics of an alpine talus. Hydrological Processes 25, 2954–2966. https:// environments. Wiley Interdisciplinary Reviews: Water 7, e1475.
doi.org/10.1002/hyp.8060. Sriwongsitanon, N., Taesombat, W., 2011. Effects of land cover on runoff coefficient.
Nakano, S., Fukuoka, T., Aramaki, S., 1987. Traace element abundances in the Journal of Hydrology 410, 226–238. https://doi.org/10.1016/j.
Quaternary volcanic rocks of the Norikura volcanic chain, central Honshu, Japan. jhydrol.2011.09.021.
Geochemical Journal 21, 159–172. https://doi.org/10.2343/geochemj.21.159. Suzuki, K., Kuramoto, T., Tanaka, M., Shar, K., 2008. Water balance and mass balance in
Neary, D., Ice, G., Jackson, C., 2009. Linkages between forest soils and water quality and a mountainous river basin, Northern Japan Alps. Bulletin of Glaciological Research
quantity. Forest ecology and management 258, 2269–2281. https://doi.org/ 26, 1–8.
10.1016/j.foreco.2009.05.027. Takahashi, K., Yoshida, S., 2009. How the scrub height of dwarf pine Pinus pumila
Neill, C., Chaves, J., Biggs, T., Deegan, L., Elsenbeer, H., Figueiredo, R., Germer, S., decreases at the treeline. Ecological Research 24, 847–854. https://doi.org/
Jhonson, M., Lehmann, J., Markewitz, D., Piccolo, M., 2011. Runoff sources and land 10.1007/s11284-008-0558-1.
cover change in the Amazon: an end-member mixing analysis from small watersheds. Toba, T., Ohta, T., 2005. An observational study of the factors that influence interception
Biogeochemistry 105, 7–18. https://doi.org/10.1007/s10533-011-9597-8. loss in boreal and temperate forests. Journal of Hydrology 313, 208–220. https://
Ohte, N., Tokuchi, N., 1999. Geographical variation of the acid buffering of vegetated doi.org/10.1016/j.jhydrol.2005.03.003.
catchments: factors determining the bicarbonate leaching. Global Biogeochemical Uehara, Y., Kume, A., 2012. Canopy rainfall interception and fog capture by pinus
Cycles 13, 969–996. https://doi.org/10.1029/1999GB900029. pumila regal at Mt.Tateyama in the Northern Japan Alps, Japan. Arctic, Antarctic,
Okuno, M., Nakamura, T., Moriya, I., Hayakawa, Y., 1994. Radiocarbon ages of wood and Alpine Research 44, 143–150. https://doi.org/10.1657/1938-4246-44.1.143.
charcoal just below the Kuraigahara tephra from Norikura-dake volcano, central Uehara, Y. (2015): Water and nutrient cycle in alpine region at Mt. Tateyama (in
Japan (in Japanese). Bulletin of the Nagoya University Furukawa Museum 10, Japanese). Kyushu University Institutional Repository, 72p. .
71–77. Viviroli, D., Kummu, M., Meybeck, M., Kallio, M., Wada, Y., 2020. Increasing
Onda, Y., Tsujimura, M., Fujihara, J., Ito, J., 2006. Runoff generation mechanisms in dependence of lowland populations on mountain water resources. Nature
high-relief mountainous watersheds with different underlying geology. Journal of Sustainability 3, 917–928. https://doi.org/10.1038/s41893-020-0559-9.
Hydrology 331, 659–673. https://doi.org/10.1016/j.jhydrol.2006.06.009. White, A., Brantley, S., 2003. The effect of time on the weathering of silicate minerals:
Peña-Arancibia, J., Bruijnzeel, L., Mulligan, M., Van Dijk, A., 2019. Forests as ‘sponges’ why do weathering rate differ in the laboratory and field? Chemical Geology 202,
and ‘pumps’: Assessing the impact of deforestation on dry-season flows across the 479–506. https://doi.org/10.1016/j.chemgeo.2003.03.001.
tropics. Journal of Hydrology 574, 946–963. https://doi.org/10.1016/j. Zhou, J., Ding, Y., Wu, J., Liu, F., Wang., S., 2021. Streamflow generation in semi-arid,
jhydrol.2019.04.064. glacier-covered, montane catchment in the upper Shule River, Qilian Mountains,
northeastern Tibetan plateau. Hydrological Processes 35, e14276.

10