Abstract: The relationship between establishment of willow communities and hydrogeomorphic process is investigated from ecological and hydraulic points of view. The study site is a downstream reach of the Yoshino River, Shikoku, Japan, where three alternate bars of similar configuration have formed over the past five decades. Salix chaenomeloides and Salix grasilistyla communities have been successively established since the early 1970s. Their establishment is found clearly restricted in both time and place due to colonization trait and mobility of bed material of the willow communities. With regard to the mobility of the bed material, it is clarified that the colonization niche is restricted to areas where annual maximum dimensionless tractive force of the bed material is around 0.06 during two or three years after their generation. The geomorphologic process at the study site appears to be considerably affected by expansion and growth of the willow communities. The bed level of alternate bars has been aggraded year after year mainly along their outer edges fringed by the growing willow communities. On the contrary, beds of low-water stream have been degraded so much as to potentially damage bank-protection works at some places. Typical changes in flow pattern and distribution of sediment mobility between the early and present stages of establishment of the willow communities are also demonstrated based on two-dimensional shallow open-channel flow simulations.

 

Keywords: willow community, condition of establishment, hydrogeomorphic process,data analysis, shallow-water simulation

In the past several decades, establishment of dense and wide woody communities has been observed in alluvial reaches of major Japanese rivers. The establishment of woody communities has a negative influence on the reliability of river works through reduction of flow capacity and promotion of local erosion. On the other hand, the woody communities, in conjunction with other herbaceous communities, play an important role in the improvement of riverine and riparian ecosystems which have been seriously damaged due to various kinds of human activity carried out throughout the river basins. At present, one of the central issues for Japanese river engineers is how to manage the woody communities reasonably from both hydraulic and ecological points of view.

The purpose of this study is to clarify the correlation between woody communities and hydrogeomorphologic process, in order to contribute to sustainable planning for the management riverine woody communities. The study site is a downstream reach of the Yoshino River, Shikoku, Japan, where willow communities of Salix chaenomeloides and Salix grasilistyla have been successively established since the early 1970s. The succession process and present situation of the willow communities were observed utilizing high-altitude and low-altitude aerial photographs [1], respectively. The geomorphologic process of riverbeds as well as the variation in both flow pattern and spatial distribution of sediment mobility were evaluated using a detailed river-data compilation supplied by the Japanese Ministry of Construction. Following comparison between the vegetative and physical processes, some characteristic features and their interrelationship have been made clear.

The stream length, drainage area and design-flood discharge of the Yoshino River are 198 km, 3,750km² and 19,000m³/s, respectively. The alluvial reach extending about 78 km from the river mouth is almost completely embanked. The bed slope is about 1/2,000 and 1/800 at the river mouth and the upstream end of the alluvial reach, respectively. The sediment supply to the alluvial reach has become much less than before because of dam constructions and elaborate sediment-consolidation works in the upstream basin.

The study site of the present research ranges between cross-sections of 17.0 and 23.0 river-kilometres. Three alternate bars of similar configuration have been appearing for more than 50 years due to spatially periodical meandering of the river course (Fig.2). The averages of bed-slope, channel width and mean diameter of bed material are 1/1,100, 800 m and 25 mm, respectively. Each cross-section of this reach takes the form of a compound channel.

The above-mentioned Salicaseous species have the following colonization trait [2]. They disperse a huge number of light seeds with the aid of wind during spring. Among these seeds, very few of the groups that land on bare areas close to the temporal water edges will germinate soon after. Small seedlings can survive if they are not subjected to strong physical disturbance by subsequent floods and also to serious drought in summer periods. Once they have survived for several years, they begin to facilitate sediment deposition on their floors and thus their tolerance to physical disturbance will be significantly enhanced with their age.

Fig.1 shows the distribution of willow communities observed in 1998. The willow communities are distributed mostly along the outer edges of the bars as well as in their tail parts. However, initial years of their colonization success are neither uniform nor continuous. Inspection of high-altitude photographs and tree-ring examination on sample individuals indicated that the initial years are restricted to the following specific years, namely 1973, 1977, 1986 and 1994. Moreover, it was found that the older communities apt to be distributed in areas closer to the downstream end or the inner edge of the bar. Based on the river-data, it was confirmed that fluctuation of water level in the spring of those years was less than that of the other years and also that no floods big enough to be able to put the bed materials in motion occurred within the subsequent two years [3]. These hydrological conditions seem to well correspond to the above-mentioned germination-survival trait of the Salicaseous species.

Fig. l    Distribution of willow communities observed in 1998.

The geomorphologic process that occurred in the past three decades is outlined in Fig.2. By 1972, the three alternate bars had been formed; however the shape of each was not clear. Subsequently, the surface level of each bar has been aggraded particularly along the outer edges. Referring to Fig.1, it may be seen that the aggraded areas are almost consistent with locations of the willow communities. Such bed-agradation amounts to 2 m in the downstream part of the bar while being less than 1 m in the upstream areas. Thus, at present, each bar looks like a low tableland bordered by patches of willow communities. On the other hand, the bed level has been degraded throughout the region of the low-water stream. However its course has neither been sifted nor remarkably deformed. Bed-degradation amounts to about 1 m on average and up to 3 m at some places along the edges of riverine flood plains.

Fig.2   Geomorphologic process occurred in the period from 1972 to 1995

Referring to the colonization trait, establishment of a willow community is possible at an area where bed material does not move in a few years following its initial year. In order to clarify this point, yearly maximum values of dimensionless tractive force (, with shear velocity, specific gravity of bed material in water, gravitational acceleration, grain diameter) were locally evaluated for the initial years and those following, using a one-dimensional model for non-uniform flow routings.

In Fig.3, contour maps of the yearly maximum values of are shown after averaging them for pairs of adjoining years, whereas the result for the initial year of 1977 is omitted for lack of space. The distribution of the willow communities shown in Fig.1 is superimposed on each contour map to be referred to. Initial year of the willow communities distributing in strip areas in the most upstream part of each bar has been confirmed to be 1994 and it can be seen that, in these areas, the yearly maximum values of were around 0.06, namely the critical value of incipient motion. In other words, the bed materials of those communities, except for the finer parts, were not put into motion for a couple of years after their generation. Moreover, weak motion of some parts of the bed materials seems to have prevented herbaceous plants from invading into the newly established community areas. Such a tendency of the yearly maximum values of , as have been described above, can also be observed with regard to the willow communities that began to establish themselves between 1973 and 1986.

Fig.3    Contour maps of yearly maximum values of dimensionless tractive force, averaged for a couple of years after generation of new willow communities

Fig.4 shows an example of change in flow pattern and sediment mobility affected by the willow communities based on two-dimensional shallow open-channel flow simulations [4]. In this figure, with respect to the years of 1980 and 1995, contour maps of dimensionless tractive force of bed material, positions of cross-sectionally maximum velocity and lines of 50% discharge are compared under the condition that flow discharge is as large as the average of annually maximum discharges at the study site (=9,200m³/s). In the simulation for 1995, hydraulic resistance of trees was estimated as accurately as possible referring to geometrical data of existing trees, such as population density, local averages of tree-height, stem diameter, canopy width and so on, which were collected through a belt-transect survey carried out recently at an interval of about 100m. The positions of cross-sectionally maximum velocity tend to approach considerably closer to the banks of the reverine flood plains at many cross-sections in 1995 than in 1980. The line of 50% discharge for 1995 also shows a comparatively greater shift from the central line of the river. With regard to the contour maps of , it should be noted that its stream-wise variation along the course of the low-water stream is more remarkable in the case of 1995 in comparison with that of 1980. This result suggests that, at present, greater bed-level variation may occur in a period of a big flood than before and thus bank-protection works could be subjected to serious erosion risks. On the other hand, the values of in the regions of alternate bars seem to have been reduced to a level of no or weak sediment transport even in a period of an annual biggest flood. It is predicted that the geomorphologic and vegetative conditions of the bars would be maintained for a few decades in the future if no vegetation management were carried out.

Fig.4 Comparison of flow pattern and distribution of sediment mobility between the years of 1980 and 1995

This paper reports on findings of research investigating the relationship between establishment of willow communities and hydrogeomorphic process at a downstream reach of the Yoshino River, Shikoku, Japan. It has been demonstrated that the willow communities at the study site have established within restricted areas where, during two or three years after their generation, the bed material moved just enough to prevent other herbaceous plants from invading. The expansion and growth of the willow communities are found to have considerably raised the stability of alternate bars and thus to have facilitated their potential of survival. However, the risk of bank-erosion has been significantly raised due to the increase in the mobility of bed materials along the course of the low-water stream. The findings presented here may be useful for a vegetation management plan to be made in the near future.

[1]  Kamada, M. and Okabe, T. (1998): Vegetation mapping with the aid of low-altitude aerial photography, Applied Vegetation Science, No.1, pp.211-218.

[2]  Kamada, M., Kohri, M. and Mihara, S. (1999): Distribution of Salix spp. and Elaeagnus unbellata communities in relation to their stand characteristics on bars in Yoshino River, Shikoku, Japan (in Jpn.), Environmental Systems Research, Vol. 27, pp.331-337.

[3]  Okabe, T.(1999): Correlation between woody-plant communities and river condition (Jpn.), Technical Report on Geomorphologic Processes of Rivers, edited by The Foundation for Management of River Environment, pp.87-105.

[4]  Nagata, N., Hosoda, T. and Muramoto, Y. (1999): Characteristics of river channel processes with bank erosion and development of their numerical models (in Jpn.), Proc. JSCE, No.621/II-47, pp.23-39.