Forests & Water/PRSNT/07

Forest Hydrology in Japan and Proposal of Long-Term Forest Hydrological Research Network

Shigeru OGAWA, Kyoichi OTSUKI and Hirokazu HAGA Faculty of Agriculture, Kyushu University, Sasaguri, Fukuoka, 811-2415 JAPAN

ABSTRACT

First, the status of forest and forestry in Japan are overviewed. Secondly, the trend of forest hydrology in Japan is briefly explained and an example of long-term forest hydrological research conducted in Japan is introduced. Lastly establishment of Long-Term Forest Hydrological Research Network is proposed. 1. INTRODUCTION

Japan is one of the most broadly forested countries in the world, where the forest cover runs up to 68%. Almost all the resources had been supplied from forests since ancient times, but most materials from forests have been replaced by imported or artificial ones recently in connection with economic growth and technological development. Then, the public interest and expectations toward forests have become more diverse and articulate. To correspond to the current state, the Forestry Basic Law enacted in 1964 was revised as the Forest and Forestry Basic Law in 2001, in which multi-functional roles of forests become predominant. However, environmental data in forests have been still insufficient, and the integrated data and analysis of forest environment have been strongly required.

In such a background, interdisciplinary studies on forest environment have been gradually spread, and movements toward establishment of interdisciplinary long-term ecological researches have arose. Hydrological process will be a core issue among these studies because water cycle is one of the most influential process in the environment. However, hydrological database has not been established nor utilized as compared with meteorological database. To overcome this situation, establishment of International Long-term Forest Hydrological Research Network is highly recommended. It will enable comparative studies not only in hydrology but also other related fields such as ecology and geochemistry, deepen our environmental knowledge and enhance the environmental management.

In this paper, firstly the status of the forest management along with the trend of forest hydrology in Japan are introduced Then, an example of long-term forest hydrological research conducted in Japan is briefly introduced, and the establishment of Long-term Forest Hydrological Research Network is proposed. 2. FOREST AND FORESTRY IN JAPAN 2.1 Physical features of Japan Japan is a crescent-shaped archipelago extending along the eastern coast of the Asian continent. It stretches about 3,000 km long from 45o33'N, 153o59'E to 20o25'N, 122o56'E. It consists of four main islands of Hokkaido, Honshu, Shikoku, Kyushu and many other small islands with a total area of 377,873 km2.

Almost 75% of the land is mountains and hills covered by forests, and about 24% of the land is flat. In central Japan, there are numbers of high mountains of the 3,000 meter range. Several mountain ranges act as the backbone of the archipelago. They are associated with volcanism, and there are some 265 volcanoes, of which many are still active. Thus, catastrophic earthquakes and eruptions have occurred regularly throughout Japanese history.

Rivers in Japan are mostly short and swift. Most streams rise in the mountains, run across narrow strips between mountains and sea not distant therefrom. Flooding, debris flows and landslides frequently occur in rainy season and typhoon season.

The climate of Japan is diverse in location but featured by dynamic changes in weather and distinct seasons (Fig.1). It is significantly affected by the elements of 1) the wide range of latitude, 2) latitudinal

- 1 -

position receiving numerous fronts, 3) the location on the east edge of the Asian Continent, 4) the mountainous topography throughout the country, and 5) the surroundings with deep seas having various currents.

Southwestern Japan has long, hot summers and short cool winters. Northern Japan experiences short, warm summer and long, cold winter. In the coastal areas of central Japan, winter is cool and summer is hot. The mountains running through the center of the islands protect the lee side from the effects of the monsoons, effectively creating two climatic regions; the Sea of Japan side and the Pacific Ocean side. The Sea of Japan side is featured by heavy snowfall in winter and year-round humidity, and the Pacific Ocean side is featured by low winter precipitation and dry winter. Annual rainfall throughout Japan is high (generally between 1,000 and 25,00 mm). With these pluvial features, Japan has an average annual precipitation of 1,750mm, almost twice as high as the world average. Almost whole the country has annual precipitation more than 1,000mm. Soils in Japan are formed under the influence of the environment of temperate humid climate, islands, volcanoes, high steep mountains and short steep rivers. Forest soils in Japan are classified into seven soil groups (Podzol, Brown Forest Soil, Red-Yellow Soil, Black Soil, Dark Red Soil, Glay, Peat) and immature soil. About 75% of the forest soil is Brown Forest Soil. Brown Forest Soil is distributed throughout the country except Ryukyu chain. Black Soil are mainly distributed in the area covered by volcanic Kuroboku (Ando). Podzol are located in alpine and sub-alpine zones.

A s ahikawa-10

0102030

0100200300400500

Niigata-10

0102030

0100200300400500

Sendai-10

0102030

0100200300400500

M ats ue-10

0102030

0100200300400500

Fukuoka-10

0

10

20

30

0100200300400500

M iyazaki-10

0

10

20

30

0100200300400500

M ats u moto-10

0102030

0100200300400500

Tokyo-10

0102030

0100200300400500

Takamats u-10

0102030

0100200300400500

Kochi-10

0102030

0100200300400500

Kumamoto-10

0

10

20

30

0100200300400500

Fukui-10

0102030

0100200300400500

: Air temperature (oC) : Solar radiation (MJ/m2) : Precipitation (mm/month)

Fig.11 Climate in Japan (Otsuki, 2001)

2.2 Forests in Japan

In Japan, forest covers about 25.1 million hectares in 1999, which is about 68% of the land. Of the entire forested area, natural forests and plantation forests cover 13.4 million hectares (53%) and 10.4 million hectares (41%) respectively. Of the entire forested area, national forests cover about 7.8 million hectares (31%), municipal and prefectural forests make up about 2.7 million hectares (11%) and private forests is 14.6 million hectares (58%). The current growing stock is about 3.5 billion m3 with an average annual growth of about 80 million m3 consisting mainly of planted forest. 2.2.1 Natural forests

The distribution of the forest zones in Japan is determined primarily by temperature because the rainfall is generally sufficient. Thus the distribution corresponds well with latitude and elevation (see Fig.2). There are five major natural forest zones:

- 2 -

Rainforest (sub-tropical) Rainforests in Japan are restricted to the southwest islands of the Ryukyu chain. The dominant species of these forests are similar to the evergreen broadleaf forest of western part of Japan, but floristic elements from the Asian continent, southern China, the Philippines and Indonesia are also found. Evergreen broadleaf forest (warm temperature/hill) Evergreen broadleaf forests are distributed in the northern islands of Ryukyu chain and lowlands of Kyushu, Shikoku and southern Honshu. They are floristically and biogeographically unique. These forests are restricted to Japan, the southern coast of the Korean peninsula, and central and southern China. They differ from European and North American forests of the same climatic zone featured by dry summer and rainy winter. They are dominated by evergreen oaks with thick glossy leaves. Major species include Quercus, Cyclabalonopsis, Castanopsis, Machilus and Cinnamomum. The forests are layered, multi-aged, with the leaves concentrated in the upper crown. Deciduous broadleaf forest (cool temperature/mountain) Deciduous broadleaf forests are distributed in highlands of Kyushu, Shikoku and southern Honshu, lowlands of northern Honshu and south to the central Hokkaido. Major species are beech (Fagus crenata) and oak (Quercus mongolica var. grossesrrata). Although beech forests predominate in Honshu, the species composition differs between the Sea of Japan side and Pacific Ocean side. In the Sea of Japan side, Fagus crenata is predominate in association with Quercus mongolica var. grossesrrata, Acer japonicum and Betula maximowicziana. In the Pacific Ocean side, Abies and Fagus japonica is predominate in association with Quercus serrata, Carpinus laxiflora and Carpinus tschonoskii. Evergreen coniferous forest (sub-alpine) Evergreen coniferous forests are distributed in the sub-alpine zones of northern Honshu and Hokkaido and lowlands of eastern Hokkaido. In some zone, deciduous birch (Betula ermanii) also grows. Major species are Abies and Picea. These forests are similar to other sub-alpine forests in the northern hemisphere. Shrub/tundra forest (alpine) Shrub/tundra forests are distributed in the alpine zones of northern Honshu and Hokkaido. There are only low trees such as Pinus pumila. 2.2.2 Plantation Forests

Plantation forests are a significant and important in Japan. Reforestation has a long history in Japan. Major species are sugi (Cryptomeria japonica), hinoki (Chamaecyparis obtusa) and matsu (Pinus species). Plantations now occupy about 41% of the total land area, the majority of which were planted after World War II. Initially, plantations were established for protection purposes as well as timber production. Recently the roles of forests have expanded to include wind control near coasts, fish breeding habitat, scenic reserves and recreation areas.

Fig.2 Distribution of natural forest in Japan (Tadaki, 1992)

30oN 35oN 40oN

Tundra

Evergreen conifer forest Evergreen broadleaf forest

Deciduous broadleaf forest

45oN

Rain forest

- 3 -

2.3 Current State of Forest Management in Japan 2.3.1 After the Forestry Basic Law Various measures under the Forestry Basic Law enacted in 1964 have increased forest resources as well as wood supply capability. However, the scale of forest management size has not expanded; 94% of forest owners have less than 20 hectares of forest. Moreover, the demand has shifted drastically from domestic timber to imported timber; the self-sufficiency ratio was fallen to 19% in 1999. Wood demand has declined and forest-products prices have been stagnant for a long time. Forestry employees has been decreasing (70,000) and getting older (29% are older than 65 years old). Improvement of forestry came to a standstill and the forestry and wood industry became stagnant.

The public demands for forests have become diversified such as a growing expectation for recreation, the conservation of the natural environment, and the mitigation of global warming. Forests of 8.87 million hectares are designated as Protection Forests for securing water resources, preventing natural disasters, preserving living environment and so on (see Table 1). It should be noted that almost 90% of the Protection Forests are designated for securing water resources and preventing sediment discharge, thus highly related with hydrology and water resource management.

Table 1 Area of Protection Forest in Japan (1,000 ha) Category National forest Private forest Total Securing water resources 3,291 3,096 6,387 Preventing sediment discharge 783 1,320 2,103 Others 15 categories 429 554 983 Total (some are categolyzed more than one) 4,503 4,970 9,473 Actual total 4,209 4,658 8,867

2.3.2 Reform of national forest management to create "forests for the people" in 1998

The national forests, approximately 20% of the land and about 30% of the entire forest area, are mainly located in the backbone mountain ranges or upstream water reservoir areas. In order to manage the national forests as ''common assets of the nation, with the people's participation, for the sake of people", laws concerning the reform of national forests was enacted in 1998, which include a shift to social-benefit-oriented forest management, the establishment of a simple and efficient operation system through the streamlining and cutting back of organizations and personnel.

By this reform, the national forests for social objectives were enlarged from 50% to 80% of total national forest area. These objectives include 1) water resource management, soil conservation, and sheltering, 2) conservation of ecological, cultural, historical, recreational and spiritual values, and 3) sustainable use of wood resources. 2.3.3 Revision of the Forestry Basic Law as the Forest and Forestry Basic Law in 2001

The Forestry Basic Law enacted in 1964 was revised as the Forest and Forestry Basic Law in July, 2001. The backgrounds of the former and new policies are compared in Table 2. The philosophy of the new policy is to fulfill the multi-functional role of forests by the "sustainable forest management". Directions of the new policy are as follows (Forestry Agency, 2001): Fulfillment of the multi-functional role of forests In order to maintain soundness and vitality of forests, and to meet public demands toward forest management, introduction of long-term cyclical forest management, constant implementation of forestry works including thinning of plantation forests must be promoted. Promotion of cyclical use of forest resources In order that forestry will be responsible for the maintenance of forests as well as the sustainable use of forest resources, fostering work force capable of effective and stable forest management, intensifying forest works and management, securing and developing employees must be promoted. Vitalization of mountain villages In order that mountain villages will play an important role in creating safe and prolific national land, mountain villages need to be vitalized by creating and securing diverse job opportunities, by improving of living conditions and by promoting interchange activities with cities.

- 4 -

Setting the objectives of forest management and use of forest resources In order to stabilize the supply of wood products with reliable quality, increasing efficiency of distribution system through information technology and utilizing local timber must be promoted.

Table 2 Comparison of background of the former and new forest policies Background

The Forestry Basic Law Background

The Forest and Forestry Basic Law Demand for wood

Wood supply did not meet the increasing demand by the economic growth.

80% of wood is imported and domestic supply has been decreasing.

Public demand for forests

Immediate action for enlargement of wood supply and stabilization of the price was required.

Demand for multi-functional role of forest has been increasing

Forest management

Development of deep undeveloped forests and enlargement of plantation forests was conducted to increase wood supply.

Implementation of thinning of plantation forests and promotion of long-term cyclical forest management are required.

Forest owner Ambition of forest owners for plantation and wood production was vital in accordance with the vigorous demand for wood

Decreasing profitability of forestry and generation shift has lost interest of forest owners toward forestry.

Social-benefit of forests

Social-benefit of forests was considered to be involuntarily obtained by the forest management to promote forestry.

Possibility of harm to social-benefit of forests by non-thinning plantation forests and non-planted cutovers has emerged.

3. FOREST HYDROLOGY IN JAPAN 3.1 Trend of forest hydrology in Japan Hydrology was established as a modern science based on the measured data in late 17th century. Forest hydrology was initiated in Germany and France when forest observatories were set in 1860s (Fukushima, 1992). In 1899, the measurement of discharges from two watersheds having different vegetation cover started in Switzerland.

Referring to this experiment, experimental watersheds were set in Ohta and Kasama within the Tokyo Forest District in 1904. It was the beginning of forest hydrology in Japan. The University of Tokyo established an experimental watersheds in Ashiyatsu in the Chiba Experimental Forest in 1913 (Kuraji, 2001). The former National Forest Experimental Station initiated a measurement of discharge in the Tatsunokuchi experimental watersheds in 1937. Other organizations have also started forest hydrological measurement since the International Hydrological Decade (IHD) was initiated in 1965. The Japanese Forestry Society and The Japan Society of Erosion Control Engineering were the major societies to conduct forest hydrology at this stage. However, the other societies dealing with hydrology also studied hydrological cycles of forests because most of the land is covered by forests in Japan. Common objectives in these studies were to quantitatively evaluate the hydrological functions of forests; whether forests conserved water resources or not, and whether forests prevented flood or not . Principal interest was put on the relation between rainfall and discharge.

Since 1980s, hydrological elemental process such as evapotranspiration, soil water movement (infiltration, throughflow and subsurface flow), and overland flow have been studied in various hydrological societies. In forest hydrology, water movement within the canopy such as interception, stemflow and throughfall have also became spotlighted. At the same time, studies on water quality in the hydrological cycle emerged. In 1990s, Soil-Plant-Atmosphere-Continuum (SPAC) models were introduced, which required physiological measurements and knowledge. Remote sensing analysis has became widely used in forest hydrology because of its capability of simultaneous and large-scale measurement on the complex terrain of mountains. Flux measurements obtaining fluxes of CO2, sensible heat and latent heat have been spotlighted because of the global warming, which required eco-physiological knowledge.

As mentioned above, forest hydrology has become more diversified, and ecology, physiology, meteorology, soil physics, remote sensing and other environmental sciences become parts of forest hydrology.

- 5 -

SoilSoil

ForestForest

気孔 吸収

RadiationEnergyWaterCO2

AtmosphereAtmosphere

Quantity & Quality

Fig.3 Current contents of forest hydrology in Japan

3.2 An example of long-term forest hydrological research in Japan Most people believe that one of the most important social benefit of forests is alleviation of water shortage. Forests are therefore often called as "Green Reservoir". However, opposition to the concept has been appealed since the commencement of discharge measurement, and the dispute has been still spotlighted. Thus, a number of long-term forest hydrological research over a wide area is required to settle the dispute and precisely understand the mechanism of hydrological cycle in forests. We have conducted a long-term forest hydrological research for such purposes and herein introduce the outline of the research. 3.2.1 Objective The objective of this study is to discuss the long-term trends of stream discharge and evapotranspiration with recovery of vegetation after a forest. 3.2.2 Study sites The study catchments locate in the Etajima Island (Fig.4). The annual precipitation and the annual mean air temperature in this area are 1400mm and 15oC, respectively. The climate is characterized by relatively scarce precipitation in Japan. This area consists of weathered granitic rock and steep slopes with shallow soils.

The Etajima Island lost most of the forest vegetation by great fire occurred in June 1978. After the fire, Hiroshima Prefecture set three small experimental catchments (A, B and C) for hydrological observation. The forest vegetation of the catchment B and C was burned down except near the channels while the catchment A designated as a control happened to escape from the fire (Hiroshima Prefecture, 1980). Rainfall and stream discharge have been observed in each catchment for about 20 years since then. Table 3 shows the geomorphologic characteristics of these catchments.

The vegetation of these catchments was similar before the fire. Recently, the catchment A is covered with more than 50-year-old second growth forests and comprised of mixed stands of evergreen coniferous trees; Pinus densiflora Sieb. et Zucc., and deciduous broad-leaved trees; Lyonia neziki Nakai et Hara, Juniperus rigida Sieb. et Zucc. and Eurya japonica Thunb.. Ferns also commonly appear in the forest floor. However, a part of the forests with an elevation over 150m is somewhat young because the part was damaged by fire about 40 years ago. Pine wilt disease has gradually increased since early 1990s in the catchment A.

In the catchment B and C, the surfaces have already covered by tree canopies because of seed dispersal by aircrafts (1978~1980) and hillside planting works (1978~1988). The overstories of the both catchments are dominated by Pinus densiflora Sieb. et Zucc. and Quercus serrata Thunb. and the

- 6 -

understories are dominated by Rhododendron mucronulatum Turcz. var. cilliatum Nakai and Eurya japonica Thunb.. Shrub trees in catchment B and C are more than in catchment A.

Table 3 The geomorphologic characteristics of gauged catchments

Catchment History Area* Angle** Aspect*** A control 10.7 33o30’ N53oW B fire-damaged 19.2 30o45’ N28oE C fire-damaged 17.4 37o05’ N35oE

* Catchment area. Unit is ha. ** Mean slope angle. *** Mean slope aspect.

Fig.4 Study site 3.2.3 Method The yearly variation and long-term trend of the stream discharge were analyzed using the running average method (period: 30, 100, and 200-day) based on daily discharge data from 1979 to 1999. If the period included some missing data, the average value was estimated using existing data. Additionally, the variation of discharge-duration curves was analyzed. The trend of runoff rate (runoff/rainfall) during the high flow was also analyzed using hourly data of several representative high flow events every year.

Daily evapotranspiration was estimated by short-time period water budget method. In this method, we assume that there is no difference between water storage at time t1 and t2. Then evapotranspiration (E) during the water budged period (t1 ~ t2) is

- 7 -

dttqdttpEt

t

t

t)()( 2

1

2

1∫∫ −= , (1)

where p(t) is the daily rainfall, q(t) is the daily discharge and t1 and t2 are the time of before and after high flow period, respectively.

Normalized Difference Vegetation Index (NDVI) was computed using spectral bands in the visible red (RED) and near-infrared (NIR) range based on satellite remote sensing data:

REDNIRREDNIRNDVI

+−

= . (2)

In this analysis, we used the bands 5 and 6 of Landsat MSS and the bands 3 and 4 of Landsat TM. The grid scales to calculated spatially averaged NDVI were 56×84 (MSS) and 112×168 (TM) pixels. NDVI of Hiroshima city was also calculated as reference of the study catchments. Table 4 shows the types of Landsat data used in the analysis.

Table 4 The types of data used in NDVI analysis. Year 1979 1980 1981 1984 1985 1986 Date 9/11 5/29 6/2 5/8 10/2 9/3 11/6 Type M M M MT T MT T

Year 1987 1989 1992 1993 1994

Date 5/8 11/25 4/27 5/21 3/5 8/8 Type T T T T T T

* M: MSS, T: TM, MT: MSS and TM 3.2.4 Results Long-term trend of stream discharge Fig.5 shows the variation of the running average discharge in each catchment. The discharges in the fire-damaged catchments B and C were larger than the one in the control catchment A from 1981 to 1990. The differences in discharge between these catchments have gradually decreased since 1990 except in 1994. The discharges of the catchment B and C were smaller than the catchment A in 1994 when extraordinary water shortage occurred; annual rainfall was 796mm. Catchment B Catchment C

i)

Catchment AA A

A 100

( 10

Run

ning

ave

rage

dis

char

ge [m

m/d

ay]

1

0.1 100

ii)(10

1

0.1 100

iii)(10

1

0.1 8

Figure Av

1 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00

- 8 -

Year 2. Compariso charge with three

ern of the trend of stream dis

aged periods are (i) 30- , (ii) 100 - , and (iii) 200 - days, respectivFig.5 Comparison of stream discharge

Average periods are (i) 30 days (ii) 100 days and (c) 200days respectively

ely.

Fig.6 shows the discharge-duration curves. The ordinary and scanty water discharges, the 185th and

355th discharges from the maximum discharge respectively, in the catchments B and C were higher than the ones in the catchment A until 1988. However, these in catchments B and C were nearly equal to or lower than the one in the catchment A after 1991. Although the ordinary discharges were almost identical in 1993 when the annual rainfall was 2,100mm, one and a half times as much as mean value, the scanty water discharges differed in this year. In the drought year of 1994, both the ordinary and scanty water discharges greatly differed.

- 9 -

1983 1984 1986

1988 1989 1991

1992 1993 1994

1995 1996 1997

1998 2000

0 60 120 180 240 300 0 60 120 180 240 300

0 60 120 180 240 300 1000

100

10

1

0.1

1000

100

10

1

0.1

1000

100

10

1

0.1

1000

100

10

1

0.1

1000

100

10

1

0.1

Dis

char

ge [m

m/d

ay]

Order

Figure

Catchment A

Catchment B

Catchment C

4. Discharge-duration curves.Fig.6 Discharge-duration curves

Fig.7 shows the trend of runoff rate. The runoff rate in the catchment A has not remarkably changed while the runoff rates in the catchment B and C have reduced since around 1990. The degree of reduction, was remarkable for events with low averaged-rainfall intensity.

Figu dre 3. Trend of the runoff rate in catchment A anFig.7 Trend of the runoff rate in cathmnets A and B

Catchment B Catchment A

0.0

0.5

1.0

1.5

2.0

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 Year

Run

off r

ate

Characteristics of evapotranspiration and NDVI Fig.8 shows the 200-day running average of daily evapotranspiration using the short-time period water budget method. The evapotranspiration in the catchment A was larger than the ones of the catchment B and C until 1991. They have been almost identical since 1992.

81 82 83 84 85 86 87 88 89 90 91 92 93 94

5

4

3

2

1

0

200-

days

run

ning

ave

rage

ev

apot

rans

pira

tion

[mm

/day

]

Figure 5. Comp ree catchments.

Catchment A Catchment B Catchment C

arison of the trends of evapotrans piration with thFig.8 Comparison of the trend of evapotranspiration

Fig.9 shows the trends of NDVI of the urban area and the catchments. The fact that NDVI in the urban

area of Hiroshima city has not changed much except for November 25, 1987 and March 5, 1993 induced that the background of NDVI has not greatly changed. Although NDVI in the catchment B and C tended to

Fig u re a

-0.10

0.10.20.30.40.50.6

7909

11

8005

29

8106

02

8405

08

8609

03

8405

08

8510

02

8609

03

8611

06

8705

08

8711

25

8904

27

9205

21

9303

05

9408

08

C atchment A C atchment B C atchment C Urban area

re 6 . T re nd of N D V I in thre e catchme nts a nd urba n aFig.9 Trend of NDVI in three catchments and urban area

- 10 -

be larger than the one in the catchment A for about five years after the fire, they have exceeded the one in the catchment A since around 1986. As shown in Fig.10, there was positive correlation between NDVI and evapotranspiration.

0

1

2

3

0 0.2 0.4 0.6

Evap

otra

nspi

ratio

n [m

m/d

ay]

NDVI

◇ Catchment A □ Catchment B △ Catchment C

Figure tion. 7. Relationships between NDVI and evapotranspiraFig.10 Relationship between NDVI and evapotranspiration

3.2.5 Conclusion The long-term trend of stream discharge and evapotranspiration after forest fire in 1987 imply that the vegetation recovery influenced the water balance in the fire-damaged catchments. The results are summarized as follows: 1) Stream discharge increased after forest vegetation was burned. 2) Evapotranspiration increased and stream discharge decreased with vegetation recovery. 3) Ordinary, low, and scanty water discharges in the control catchment were greater than those in the

fire-damaged catchments when extraordinary water shortage occurred. It could not easily conclude that the existence of forest vegetation is not effective in increasing stream discharge because stream discharge in the control catchment was kept in higher level than the one in the fire-damaged catchment during water shortage period. We conclude that it takes more than several decades to recover the roles of forests to alleviation of water shortage while vegetation condition is recovered within about ten years. It implies that soil water storage and ground water recharge were important for the roles of forests to alleviation of water shortage. 4. LONG-TERM FOREST HYDROLOGICAL RESEARCH Human may have kept the memories of benefits from forests since ancient times, and reflected how much we have damaged forests in recent centuries. We accordingly tend to or would like to believe that forests give us abundant benefits and respect forests as the land to be fostered and conserved. However, we do not have sufficient data to prove these benefits and lead how to foster forests with producing resources and promoting the social benefits. Although we must not evaluate the benefits of forests from a single point of view, we have responsibility to monitor ecosystems of forests and open them to the public. Among many factors existed in the ecosystems of forests, water circulation is the predominant factor because it is most influential process in the ecosystem. However, forest hydrological monitoring systems are scarcely distributed and forest hydrological database has not yet established.

Recognizing the importance of ecosystem in recent years, ecological research networks have been widely spreading worldwide. Since they could be useful models for forest hydrological research network, we herein introduce the International Long-Term Ecological Research (ILTER), and the activities of

- 11 -

environment monotoring conducted by the Japanese Council of University Experimental Forests. Based on these background, we propose to establish the Long-Term Forest Hydrological Research Network and link it with other hydrological research network. 4.1 International Long-Term Ecological Research (ILTER) The U.S. National Science Foundation established the U.S.LTER program in 1980 to support research on long-term ecological phenomena in the United States. The U.S.LTER now has 24 LTER sites. In 1993, the U.S.LTER Network held a meeting on international network in long-term ecological research to meet the demand in assessing and resolving complex environmental issues. In the meeting, the establishment of International LTER (ILTER) Network was decided. The mission of the ILTER network is as follows (ILTER, 2001) :

Promote and enhance the understanding of long-term ecological phenomena across national and regional boundaries;

Promote comparative analysis and synthesis across sites; Facilitate interaction among participating scientists across disciplines and sites; Promote comparability of observations and experiments, integration of research and monitoring,

and encourage data exchange; Enhance training and education in comparative long-term ecological research and its relevant

technologies; Contribute to the scientific basis for ecosystem management; Facilitate international collaboration among comprehensive, site-based, long-term ecological

research programs; and Facilitate development of such programs where they currently do not exist.

Twenty one countries have established formal national LTER programs and joined the ILTER network in 2000. Japan is still in process of developing the national LTER. The LTER Sub Committee was formed in the Ecological Society of Japan in 1999, and has been preparing to establish the Japanese LTER (JLTER) Network. 4.2 Environmental Monotoring conducted by University Experimental Forests There are 27 universities having university forests, in which the total number of research forests are 81. The Japanese Council of University Experimental Forests has been conducting two joint projects. The one is the Phenology Observation Network Project in which tree phenology has been observed in 22 forests in 12 universities. The other is the Material Dynamics Monitoring Project in Forested Watershed in which water quality of rainfall and stream flow have been observed in 15 forest watersheds in 14 universities.

The later was initiated to continue at shortest 50 years. For the comparative study, observations have been conducted in a same manner under the unified manual. The database of acid rain and the list of experimental watersheds have been opened through internet (in Japanese).

Acid Rain Data : http://pc3.nrs-unet.ocn.ne.jp:591/juef_data/Acidopen/Start.htm Watershed Catalog : http://forester.uf.u-tokyo.ac.jp/~kuraji/univ_for_exp_list.htm

4.3 Proposal to establish Long-Term Forest Hydrological Research Network Although the above mentioned data are useful and valuable, the data in the mountainous area are still too scarce to understand the ecosystem in detail. Thus, establishment of long-term forest hydrological research network is strongly required. In United States, these network has already established and data are freely presented. In the U.S.LTER, such data are freely supplyed through internet http://lternet.edu/data/. The United States Department of Agriculture presents the various hydrological data (precipitation, snow, discharge, soil water, climate, etc.) observed for 34 years in the Reynolds Creek Experimental Watershed through internet http://www.nwrc.ars.usda.gov/databases/index.html.

To meet the demand in establishing the long-term forest hydrological research network and open the database in Japan, the Mountainous Watershed Catalog & Database Research Group in Japan Society of Hydrology and Water Resources initiated the preparation for the catalog and database with other scientific groups related to hydrology in 2000. The Japanese Council of University Experimental Forests is also preparing the fourth stage of the Material Dynamics Monitoring Project in Forested Watershed. To extend the project as the forth stage, the following objectives are formed: 1) clarify the dynamics of water, energy and material in each forested watershed,

- 12 -

2) compare the dynamics with other forested watersheds and clalify the characteristic depend on the botany, geology, geography and climate,

3) construct the environmental database network with other monitoring database downstream, clarify the dynamics in river system, and evaluate the effect of forests,

4) internationally link the database network with other environmental database networks such as ILTER. To accomplish these objectives, a core project of long-term & large area material dynamics in forest

and three sub projects to 1) analyze the conservational function of water quantity and quality of forest, 2) open the database and link with other forest environment monitoring system such as ILTER, and 3) link the monitoring database with other monitoring systems down stream such as Meteorological Database and River Water Level & Quality are formed. This process is for the Japanese Council of University Experimental Forests project, but essential for forest hydrology in general. Consequently, we propose to extend this approach worldwide and establish the international long-term forest hydrological research network.

Fig.11 Concept of the Long-Term Forest Hydrological Research Network

Basin

Basin

City

Environmental Monitoring

Forest

Cooperation with Other Monitoring

Systems

EANET LTER Japan

LTER Asia

Open Databas

Lon

g Te

rm

Coast Farm

Lar

ge A

rea

Interdisciplinary

Basin

5. CONCLUSION Meteorological observation network has long history and wide spread in the world. In case of Japan, there are almost 150 standard meteorological station and about 1,300 Automated Meteorological System (AMeDAS) stations managed by the Japan Meteorological Agency. The facilities, instruments, elements, method and database have been strictly followed by the guidelines, and accuracy and homogeneity of the data have been highly maintained. Their data are sold by the Japan Meteorological Business Support Center through on-line system or various off-line system such as CD-ROM, FD, MO and MT.

Hydrological observation network has not yet standardized compared with the meteorological network. Moreover, hydrological database system has not been established nor utilized as compared with meteorological database. Since hydrological process is the predominant process in ecosystem, it is strongly

- 13 -

recommended to establish the hydrological research sites and construct monitoring and database network. One who relates forests should keep in mind that we have responsibility to clarify the effect of forests on the ecosystems in a basin, region, nation and world, and forest hydrological process is the base of the research. REFERENCE Agata, Y. (2000) : Status of national hydrological database supplied by public and semi-public

organizations, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.2-3

Berris, S. N., and Harr, R. D. (1987) : Comparative snow accumulation and melt during rainfall in forested and clear-cut plots in the western cascades of Oregon, Water Resources Research, 23, pp.135-142.

Conrad Totman (1989) : The green archipelago -forestry in pre-industrial Japan-, University of California Press, Berkley

Forest Agency (2001) : Annual Report on Trend of Forestry Fiscal Year 2000 -(Summary) (Provisional Translation)

Forest Agency, Ministry of Agriculture, Forestry and Fisheries (1988): Report of outline for roles of forest region in water resources (in Japanese), pp.92-103.

Harr, R. D. (1981) : Some characteristics and consequences of snowmelt during rainfall in western Oregon, Journal of Hydrology, 53, pp.277-304

Harr, R. D., Levno, A., and Mersereau, R. (1982) : Streamflow changes after logging 130-year-old Douglas fir in two small watersheds, Water Resources Research, 18, pp.637-644.

Hiroshima Prefecture (1980) : Forest hydrology and erosion control-Annual Report (in Japanese). Jo Sasse (1998) : The Forests of Japan, Japan Forest Technical Association, Tokyo Jones, J.A., and Grant, G. E. (1996) : Comment on “Peak flow responses to clear cutting and roads in small

and large basin, western Cascades, Oregon”, Water Resources Research, 32, pp.959-974. Kodera, K. (2000) : Acivities of the working group on water environmental geography of the Association

of Japanese Geographer and the database of water environment in river watershed, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.12-13 (in Japanese)

Kondo, A. et al. (2000) : Global scale database and the construction of database for comparative hydrological studies, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.14 (in Japanese)

Kubota, J., and Siverpalan, M. (1995) : Towards a catchment-scale model of subsurface runoff generation based on synthesis of small-scale process-cased modelling and field studies, Hydrological Processes, 9, pp.541-554.

Kuraji, K. (2000) : Observation of stream discharge by the of Ministry of International Trade and Industry and its catalog, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.8-9 (in Japanese)

Kuraji, K. and Shibano, H. (2001) : Hydrological observation in experimental watersheds in University Forests, Journal of Japan Soc. Hydrol. & Water Resour., 14(6), pp.489-498 (in Japanese)

Otsuki, K. (2001) : Rice culture in Japan, Rice Culture in Asia, Korean National Committee on Irrigation and Drainage, Seoul, pp.216-235

Otsuki, K., Ogawa, S., Kume, A. and Kumagai, T. (2002) : Toward establishment of Japan-Korea Long-Term Forest Hydrological Research Network, Proc. Annual Meeting of Korea Water Resources Association in 2002, pp.51-58

Tadaki, Y. (1992) : Structure and ecology of forests, A guide of Forest Instructo edited by Forest Agency, Japan Forestry Promotion Association, Tokyo, pp.32-69 (in Japanese)

Uchida, T. (2000) : Hydrological observation sites in mountains -What is necessary information for the mountainous watershed catalog ? -, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.6-7

Ushiyama, M. (2000) : Construction of national experimental watershed survey database, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.4-5 (in Japanese)

Yamanaka, T. (2000) : Approaches of the stream discharge problem working group of the Japanese Association of Hydrological Science -Mountainous Watershed Hydrological Databese-, Proc. 1st Meeting of the Mountainous Watershed Catalog and Database Research Group of JSHWR, pp.10-11 (in Japanese)

- 14 -