CASE STUDIES ON WATER RESOURCE REDUCTION CAUSED BY EXTRAORDINARY INCREASE OF FOREST FOLIAGE

Huaxia Yao *, Michio Hashino * and Tohru Kanda **

*Environmental Hydrology, Department of Civil Engineering,

The University of Tokushima, 2-1 Minami-josanjima,

Tokushima 770-8506, Japan. Tel/Fax: 81-88-6567332,

e-mail: yao@ce. tokushima-u.ac.jp, michio@ce.tokushima-u.ac.jp

** Department of Civil Engineering, Kobe University, 1-1 Rokkodai, Nada,

Kobe 657-8501, Japan.  Tel: 81-78-8036053, Fax: 81-78-8036069,

e-mail: kanda@kobe-u.ac.jp

Abstract: Annual water budgets are analyzed for two mountainous and forested basins in the Shikoku Island of Japan, and their changing trends within 36 years are explained. Similar and notable trends are found: small increase in rainfall, but small decrease in runoff and large increase in evapotranspiration. The remarkably increased evapotranspiration is found to be mainly caused by extraordinary forest plantation and growth, not by rainfall change, temperature change or domestic water usage.

Mean annual runoff within 1984-1993 decreased by 68 mm at the Nagayasuguchi Dam basin, compared with the runoff in 1957-1964, although the rainfall increased by 165 mm. That is because the evapotranspiration increased remarkably by 308 mm due to doubled forest foliage. In another basin of the Hiji River, annual runoff within 1960-1969 decreased by 140 mm while the rainfall increased by 34 mm, because the evapotranspiration increased by 173 mm due to enhanced forest foliage. These results imply that extraordinarily enhanced forest may bring about reduction in water resource in rivers, although it has many beneficial functions in flood control, sediment prevention and environmental improvement. 

Keywords: annual water budget, water resource reduction, forest foliage, rainfall, temperature, evapotranspiration, Shikoku

1    INTRODUCTION

Water management practices including direct measurements such as dams, drainage canals and irrigation system, and indirect measurements such as forest control, affect water quantity and quality in both beneficial and unexpected ways. These beneficial or unexpected effects of management should be evaluated properly. Reduction of water resource in a watershed caused by forest growth and canopy amassment has not been well noticed, although beneficial functions of flood reduction and soil conservation are widely known.

Forest is vitally necessary to maintain the ecosystem in a watershed, to adjust stormwater before forming dangerous flood peaks, to prevent soil erosion on slopes, and to give productive timbers. A watershed without forest is vulnerable to storm rains and debris flow. However, too much forests or extraordinarily planted trees are not the ideal circumstance, because in that case the evaporative water loss from foliage and canopy might be very large, and the runoff in river (water resource available) might get decreased (Hashino et al., 1999a). And any kind of forest can not substitute the functions of dam for cutting flood peak and supplying water in drought seasons. Therefore there should be an optimistic range for forest foliage (or biomass): below this range the watershed is dangerous with floods and debris, above this range the watershed is possibly in shortage of water resource.

This paper does not identify the optimistic range of forest, but aims to demonstrate the minus or negative effects (not yet well noticed) caused by forest increase and growth. Past data of water balance in two watersheds in Japan are analyzed and used to demonstrate the unexpected reduction of water yield, as Japan is typical of rapid plantation. Yearly changes of annual water budgets are found from field data, and the reasons for water reduction are explained.

2    RESULTS AT NAGAYASUGUCHI BASIN

The basin of Nagayasuguchi Dam (Fig. 1) is the upper stream catchment of the Naka River in Tokushima Prefecture, Shikoku Island of Japan, with an area of 494 km², with rich precipitation, and being covered by conifer and broadleaf trees. Based on data of 37 years (1957—1993) of annual rainfall and runoff, annual evapotranspiration amount can be simply estimated by water balance principle (annual evaporation is the residual of rainfall minus runoff). Water budgets of each year and their yearly changing trends are illustrated in Fig. 2. The 10-year moving mean values (for example the mean rainfall for 1961 is the average of rainfalls during 1957-1966) are calculated, and are used to analyze the changing trends.

Two periods of ten years, 1957-1966 and 1984-1993, will be considered for comparisons as shown in Table 1. Annual rainfall had a small increase trend within the 37 years, with the 10-year mean value increased by 165 mm (or 5.1%) from 3,204 mm for 1957-1966 to 3,369 mm for 1984—1993. On contrast, annual water yield (runoff) had a tiny decrease trend, the 10-year mean of runoff decreased by 68 mm (2.5%) from 2,755 mm in 1957-1966 to 2,687 mm in 1984-1993. Between the same periods, evapotranspiration showed a dramatically evident increase trend, increased by 308 mm (60.8%) from 506 mm to 814 mm. As a result, annual runoff got decreased although rainfall was increasing, because the evaporative loss increased extraordinarily.

Table 1    Comparison of water budgets of two periods for the Nagayasuguchi Dam basin

Then why did the evapotranspiration increase so much during the 37 years? Possible reasons should include changes and effects of rainfall, temperature, forest, agricultural activity and domestic water usage. Effects of rainfall on evapotranspiration are complex, depending on intensities of rain events and rainy days in a year. It is roughly supposed that the contributing fraction to evapotranspiration of the increased rainfall (165 mm) be same as the evaporation-rainfall ratio: 814/3369 or 0.242. That means the rainfall increase contributed about 40 mm to evapotranspiration increase, giving only 13 % contribution to the total 308 mm of evapotranspiration increase.

Alteration of temperature causes the potential evaporation to change, and then affects actual evapotranspiration. Mean temperature decreased from 14.88 ℃ in 1957-1966 to 13.19 ℃ in 1984-1993, and potential evaporation decreased by 9.22% if the potential is estimated with the Hamon formula. Its effect to actual evapotranspiration is roughly supposed to be a decrease in this percentage, e.g. a minus contribution of 47 mm (506´9.22%). That means the real or net increase of evapotranspiration is 355 mm (308+47).

Agricultural activities and domestic water usage did not change evidently, because the basin is a natural mountain area with little resident people.

However, there was a rapid growth in area of forests and canopy’s foliage mass during 1960 through 1990, caused by the policy of extending plantation. The tree density has arisen to three times of that in 1950s, canopy foliage has become doubled. This increase of trees and foliage mass produces more canopy interception of rainfall, takes more soil water for transpiration. Therefore among the net evaporation increase of 355 mm, besides the 40 mm due to rainfall increase, the residual 315 mm is most possibly caused by the forest plantation and growth. Forest change is the main reason for evaporation increase.

Some theoretical and experimental researches on the function of forest foliage in evapotranspiration processes were made by the authors and other researchers (e.g. Hashino et al., 1999b; Schellekens et al., 2000). Roughly the interception evaporation and transpiration are positively proportional to the foliage area and volume. When canopy biomass gets larger, more rainfall water can be intercepted and evaporated, more soil water may be transpirated. These phenomena can not be explained by only energy balance theory, but should be explained by complex three-dimensional thermal-dynamic and aerodynamic theories which are not the main topics of this paper.

3    RESULTS AT HIJI BASIN

As for the basin of Hiji River in Ehime Prefecture of Shikoku, with an area of 1,009 km² and averaged rainfall of 1,800mm, a similar analysis of annual water balance is made for 36 years of 1960-1995. A 85% of this basin is covered by forest and has experienced an increase in trees’ foliage. A typical forest stand is shown in Fig. 3 and the high density can be seen.

Basin averaged rainfall is obtained by a spline interpolation algorithm (Yao and Hashino, 2000). Annual evapotranspiration is the subtraction of runoff from rainfall. Annual budgets and 10-year moving means are shown in Fig. 4.

Comparing the 10-year means of water budget items between the period 1960-1969 and 1983-1992, as listed in Table 2, it is found that annual rainfall increased by 34 mm (2.0%) from 1,721 to 1,755 mm, but runoff decreased by 140 mm (11.1%) from 1,258 to 1,118 mm, while evapotranspiration increased by 173 mm (37.3%) from 464 to 637 mm, showing a similar change pattern as in the Nagayasuguchi Dam basin.

Table 2    Comparison of water budgets of two periods for the Hiji River basin

Supposedly, increased rainfall (34 mm) contributed 12 mm to evapotranspiration according to the evaporation-rainfall ratio of 0.363, possessing only 6.9 % in the total 173 mm increase of evapotranspiration. The change of rainfall can not interpret the change of evapotranspiration. On the other hand, mean temperature had a tiny decrease: from 13.2 ℃ in 1960-1969 to 12.8 ℃ in 1983-1992, causing the potential evapotranspiration to decrease by 2.26%, or causing actual evapotranspiration to decrease by 11 mm. As a result, rainfall-caused increase amount of 12 mm is almost deleted by the potential evapotranspiration-caused decrease amount of 11 mm.

Domestic water usage and its alteration are negligible for this basin, because farmland possesses a small part of the basin. Therefore, forest foliage increase is identified to be the main reason for the 173 mm increase of evapotranspiration. In fact, canopy foliage mass has arisen by about 30 % according to forest investigation statistics.

The changes of tree age can be used to interpret the increase in tree’s biomass. Averaged tree age and its changes during 30 years are listed in Table 3. For both the conifers and broadleaf trees, their age was increasing yearly. That means the amount of tree growth and plantation is larger than the amount of tree death and felling. If the older trees had been cut down or felled while the young trees were growing, the averaged tree age would not have increased as the table shows. In fact, plantation was largely implemented during 1960s and 1970s based on the national plantation policy, but matured stands were not properly felled for timber usage because of rapid extension of import of foreign timber and because of decrease in forestry labors. As a result, the forest in the basin was not carefully managed, and tree density and canopy biomass became very large. Then evaporative loss of water has also substantially increased.

Table 3    Changes of tree age in the Hiji River basin

4    CONCLUSION

Annual water budgets and their yearly changes were obtained from field observatory data in two basins of Shikoku, Japan. They showed a similar changing pattern: a tiny increase in rainfall, a contrary decrease in runoff, and a dramatic increase in evapotranspiration loss. This phenomenon is explained to be caused by extraordinary growth and increase of forest foliage, as increased foliage resulted in larger interception and transpiration. Therefore, apart from many expected benefits such as floods reduction and timber productivity, extra forest growth or plantation may cause the unexpected effect of reducing water yield. This phenomenon should be notified especially for those regions with water scarcity.

References

Hashino, M., H. Yao and H. Yoshida (1999a). Beneficial and unexpected effects of forest on water resources: in the management perspective. Bulletin of Faculty of Engineering, The University of Tokushima, Vol. 44, 37-45.

Hashino, M., H. Yao and H. Yoshida (1999b). Separation of rainfall interception and transpiration from annual evapotranspiration in forested watersheds. Annual Journal of Hydraulic Engineering, JSCE, Vol. 43, 73-78.

Schellekens, J., L. A. Bruijnzeel, F. N. Scatena, N. J. Bink and F. Holwerda (2000). Evaporation from a tropical rain forest, Luquillo Experimental Forest, eastern Puerto Rico. Water Resources Research, Vol. 36, No. 8, 2183-2196.

Yao, H. and M. Hashino (2000). Analysis on spatial distribution of annual water budget and daily runoff along river network in a basin. Annual Journal of Hydraulic Engineering, JSCE, Vol. 44, 289-294.

Fig. 1  Locations of the Nagayasuguchi Dam basin and Hiji River basin

Fig. 2  Water budgets and changing trends at the Nagayasuguchi basin

Fig. 3  Typical forest stand

Fig. 4  Water budgets and changing trends at the Hiji basin