• 
• Japanese
• 
• IETC's Focal Area
• Waste Management
• Publications & Tools
• Meetings & Workshops
• Water and Sanitation
• Publications & Tools
• Meetings & Workshops
• Disaster Management
• Publications & Tools
• Meetings & Workshops
• Resources
• Publications & Tools
• About IETC
• Background and History
• Director
• Activity Reports
• Supporting Foundations
• Procurement
• Internship
• Contact Us

<Forum on the Caspian, Aral and Dead Seas-Perspective
of Water Environmental Management and Politics>

<Symposium on the Aral Sea and The Surrounding Region
-Irrigated Agriculture and the Environment>

Ecofunctional Analysis of Dryland Vegetation: Structure,
Mechanism and their Changes

Tatsuaki Kobayashi, Atsuo Morimura,(Chiba University),
Sayat Temirbekov (Kazakh National Academy of Science),
and Yukihiro Morimoto (Osaka Pref. University)

To predict the ecological impacts of large-scale development on dryland vegetation, the interaction between hydrological changes and vegetation changes should be discussed. In our project, the following two approaches with different scales were adopted to analyse the ecosystem function; namely, remote sensing and the ecophysiological method.

1. Monitoring Vegetation Changes with Satellite Data

The LAI (Leaf Area Index) map of the Ili River Basin was produced using NOAA-AVHRR data. The spatial distribution of LAI was primarily correlated to the distance from the streams and the strength of the flow. Though the year-to-year changes of LAI were affected by precipitation in the whole types of vegetation, the coefficient of variation was largest in xeric vegetation. In natural conditions, xeric vegetation may be mostly regulated by local precipitation.

2. Ecophysiology of Dryland Vegetation

Stand structure and ecophysiological properties were studied in the following four types of shrubland: Halimodendron halodendron-dominated scrub along the Ili River, Calligonum aphyllum-dominated shrubland on sand dunes, Haloxylon aphyllum-dominated shrubland on alluvial plains and its woodland near the irrigated farms. Leaf biomass was largest in H.h. scrub, large in H.a. woodland and small in C.a. shrubland and H.a. shrubland. Though the osmotic potentials of H.h. and C.a. leaves were equivalent to mesophytic trees, that of H.a. was quite low especially in shrubland. Transpiration rate was large in H.h., small in woody H.a. and quite small in shrubby C.a. and H.a.. It is considered that the osmotic potential reflects the moisture-binding capacity of soil and the chemical quality of ground-water, while the transpiration rate reflects the amount of soil moisture. Since the transpiration of small-sized plants was affected by the moisture in the shallow zone of soil, the survival and growth of those might be controlled by local precipitation.

Introduction

The serious decline of the surface of the Aral Sea is assumed to be caused by the excessive consumption of water through the large-scale construction of channels and farms in its watershed. In Kazakhstan, there is another large lake, the Balkhash. Though it is not in a critical condition, the ecological impacts on the terrestrial and aquatic ecosystems concern humans. To conserve the regional ecosystems and secure sustainable development, future changes in ecohydrological systems should be assessed.

For the prediction of ecohydrological changes, we aim to produce a functional model based on geographical systems. The lower part of the Ili River Basin was selected as a study area for the test of the model. The Ili River, whose flow is largest in the Balkhash Lake Watershed, originates in the Tienshan Range and flows into the southern end of the Balkhash Lake. Since the Ili River has the Kapchagai Reservoir in its middle reaches, the hydrological changes have been measured most intensively in this region.

In the plain of the Ili River Basin, vegetation shows the gradient changes along the distance from the river. Hydrophytes represented by Phragmites australis occupy the riverside. Fringing forests and meadows feature the seasonally flooded plain. Scrubs dominated by Halimodendron halodendron cover the outer zone of such humid vegetation. Open shrublands dominated by leaf-less shrubs develop on the dry plain. There are two types of shrubland located on different land conditions, namely, Haloxylon-dominated shrubland on alluvial plains and Calligonum-dominated shrubland on sand dunes. In anthropogenic condition, dense woodland of shrub species surrounds irrigated farms.

In this paper, at first, the spatial distribution and past changes of vegetation are analysed using satellite images, NOAA-AVHRR. We then analyse the ecophysiological properties of component species of dryland vegetation and discuss the mechanism of the zonation of vegetation. In conclusion, the causes of vegetation changes are discussed from the ecophysiological point of view and the perspective of the modeling is presented.

Methods

1. Analysis of Satellite Data

Vegetation changes from 1987 to 1993 were studied with NOAA-LAC data. The geographical structure of data was standardized through geometric adjustment by nearest neighbor resampling method. Part of the data was clipped out on two scales as the area of the Ili River Basin and the Delta.

SAVI (Soil Adjusted Vegetation Index) image was produced from clipped data. Since green leaf absorbs visible-red ray and reflects near-infrared ray, the amount of vegetation is assumed to be shown by the ratio of spectral reflectance between the channels of visible-red zone (R) and near-infrared zone (NIR) in satellite data. To cancel the noise caused by the environmental fluctuation like air, sunlight or topography, data must be normalized to derive vegetation index. SAVI, which is one type of such indices, has advantages for sparse vegetation, because it decreases the noise caused by soil. SAVI was calculated as follows:

SAVI=(NIR-R)/(NIR+R+L)x(1+L).

L is the constant depending on the influence of soil. Here, we employed 0.5.

In order to get the changes of leaf biomass, the equation between SAVI and leaf area index (LAI) was derived. LAI was supplied from the ecophysiological measurements mentioned above. SAVI was measured on the same vegetation with a portable spectoral reflectance meter (Abe co. Model 2703). Using the equation, LAI image was painted based on SAVI image.

2. Ecophysiological Analysis

The ecological and ecophysiological properties of the dominant plants of the following 4 types of shrubland were measured: Halimodendron-dominated scrub, Calligonum-dominated shrubland, Haloxylon-dominated shrubland and Haloxylon-dominated woodland. The locations and land conditions of study sites are described in Table 1. The ecophysiological properties were also measured for some species of hydrophytes, hygrophytes and cultivated plants.

Measurements were conducted in August, 1993 and 1994. Quadrats were set on study sites with the following sizes before measurements: 24m² for Halimodendron-dominated scrub, 225m² for Calligonum-dominated shrubland, 100m² for Haloxylon-dominated shrubland and 100m² Haloxylon-dominated woodland. Species name, height and crown width were measured for all plants taller than 30 cm in each quadrat. For understory vegetation smaller than 30 cm, the coverage of component species and the density of dominant species were measured.

Leaf biomass was measured for randomly sampled individuals in each quadrat. Using the relationships between leaf biomass and crown size for each species, leaf biomass of each quadrat was estimated. Leaf area was calculated from leaf biomass with specific leaf area measured for each species. Here, leaf area of leaf-less shrub was regarded as half of the surface area of green shoot.

Ecophysiological measurements were conducted 1 or 2 sunny days for each quadrat. Diurnal changes in leaf conductance and photosynthetic rate of dominant species were measured every hour from dawn for some same leaves with a portable system (Li-Cor Li-6200). Leaf conductance and photosynthetic rate of the other component species were measured continuously from 14:00 to 16:00. Leaf water potential of dominant species was measured every 2 hours from pre-dawn for one leaf with a dew point psychrometer (Wescor HR33T). Air temperature, humidity and short radiation were measured every 5 minutes with a portable meteorological system (Vaisala HMP31UT, Grant 1209). Leaf temperature was measured every hour for three leaves with a thermocouple.

Osmotic potential of leaf cells was measured for component species. Leaves were taken from field-measured plants. Leaves were immediately packed in a plastic tube and were filled in a bottle. After boiling the bottle, cell solution was extracted with a handy squeezer. Osmotic potential of cell solution was measured with a dew point psychrometer.

Before or after field measurements, vertical soil moisture conditions were studied on each stand. One series of soil samples for each stand were taken from the surface to the groundwater or 5m. Volumetric water content was calculated with fresh weight, dry weight and specific gravity of the soil. Water potential was measured for soil sample in the field with a dew point psychrometer.

Results

1. Analysis of Satellite Data

If we look at a SAVI image of the Ili River Basin and Balkhash Lake in the summer, between the Balkhash Lake and the Kapchagai Reservoir a three high-lit areas can be recognized. The largest one is the Ili River Delta. The other two show the Bereke-Bakanas Farm and the Bakbakty Farm. A narrow belt along the Ili River is also recognized. These results confirm that vegetation is principally dependent on river and channel systems in this region.

A clear relationship between SAVI and LAI was observed in actual vegetation as follows (figure 1);

LAI=0.03291e^10.96SAVI
r²=0.9913.

Using this equation, the changes of LAI image in summer were projected with high-lit areas along river and channel systems (value greater than 1): an area.with humid vegetation like meadow, fringing forest or Halimodendron scrub. The LAI of the background was less than 1, corresponding to dry vegetation like shrubland.

LAI fluctuated from year to year. The changes of LAI in humid vegetation might be related to the flow of river and channels. Those in dry vegetation might be dependent on precipitation. Local precipitation might particularly affect the amount of ephemeral vegetation with shallow roots.

Figure 1: The Relationship Between SAVI and LAI of Actual Vegetation.

2. General Properties of Dryland Vegetation

Vegetation coverage and leaf biomass were related to the difference of site moisture conditions (table 1). They were large in leafy scrub and small in open shrublands. The canopies of shrubland were sparse and only the LAI of Halimodendron scrub exceeded 1.

Table 1. Location and Condition of the Study Site.

LAI was calcuiated only from shrub species.

Table 2. Species Composition in Each Quadrant.

Table 2 shows the species composition of the study sites. The canopy of Halimodendron scrub was composed of two kinds of leafy shrub. It was so closed that understory vegetation was not able to develop.

Both in Haloxylon shrubland and woodland, only one shrub species, Haloxylon aphyllum dominated. Though the coverage of understory species was higher in Haloxylon shrubland, species diversity was higher in Haloxylon woodland. It suggests that species are limited in Haloxylon shrubland because of severe soil moisture condition. Almost all component species belonged to Chenopodiaceae.

There were four shrub or half-shrub species in Calligonum shrubland. Component species were members of various families. Bio-diversity of vegetation was highest in Calligonum shrubland.

Ecophysiological properties of component species were summarized for each vegetation type in table 3. As a whole, changes in osmotic potential were related to site moisture status. Osmotic potential was quite low in Haloxylon aphyllum and medium in component species of Calligonum shrubland and Halimodendron scrub. Those of the meadow and fringing forest were high and that of the hydrophyte was highest. On the other hand, there was wide variation in Calligonum shrubland.

The general changes of stomatal conductance and transpiration rate were also concerned with moisture condition. However, those of Calligonum shrubland were quite low, compared to the relative status in osmotic potential. Those values showed wide variation in Haloxylon shrubland and Calligonum shrubland.

Changes of photosynthetic rate showed the same tendency as transpiration rate. Water use efficiency (WUE) was high in broad-leaf C4 plants as Echinochloa crusgalli and Atriplex tatarica. Though Haloxylon aphyllum and Calligonum aphyllum had been reported as C4 plants, the values of WUE were not high. In Calligonum shrubland, WUE showed low values in large parts of component species.

3. Water Relations of Dryland Vegetation

Figure 2 shows the profiles of soil moisture content and soil water potential. Soil was wet in Halimodendron scrub. In Haloxylon woodland, though the soil moisture contents were equivalent to those in Halimodendron scrub, the water potentials were lower. In Calligonum shrubland, soil moisture content was not high. However, water potential was quite high in the deep zone. This is due to the characteristic particle composition of sand dune. Soil moisture content was quite low in Haloxylon shrubland. In fact, the water potentials in the shallow zone could not be measured in 1993 because it was so desiccated. In 1994, since there was much precipitation, the water potentials in the shallow zone were higher than in 1993. The changes in soil moisture condition affected the ecophysiological changes of plants.

The changes in soil water potential well explained the changes in leaf water potential between Calligonum aphyllum and Haloxylon aphyllum in shrubland. Certainly, low soil moisture content was also associated with the low transpiration rate in Calligonum shrubland.

Diurnal pattern of stomatal conductance presented midday decline in Calligonum aphyllum and Haloxylon aphyllum (figure 3-1). In Halimodendron halodendron and Medicago polymorpha, stomatal conductance was much higher than shrubland species and the shape of diurnal pattern was quite different (figure 3-1). High values in the morning for Medicago and Echinochloa are due to evaporation of dew. Stomatal conductance of Echinochloa crusgalli, which is a leafy C4 plant, was stable and equivalent to the highest value of the leaf-less C4 plant, Haloxylon aphyllum in woodland (figures 3-1, -2). These results suggest that the potential of stomatal conductance is similar in same types of species for photosynthetic system in spite of the difference in habitat condition.

The midday level of stomatal conductance was quite different in the same Haloxylon shrubland (figure. 3-2). The changes of stomatal conductance was well correlated to the shrub height in 1993 (figure 4). As shown in figure 2, soil moisture in the shallow zone was very low and increased according to depth. Aboveground plant size should be related to vertical reach of the root system. So the water stress must be severe in small plants.

Figure 5 shows the relationships between leaf water potential and stomatal conductance in Calligonum aphyllum and Haloxylon aphyllum. It clearly illustrates that stomatal conductance closely depends on leaf water potential in Haloxylon aphyllum, thus, stomatal aperture was affected by water stress.

Table 3-1. Comparison of Parameters IIi River Basin
(Comparisons of parameters in water relations and photosynthesis among component species of dryland vegetation in the lli River Basin)

Table 3-2. and 3-3. Comparison of Parameters IIi River Basin
(Comparisons of parameters in water relations and photosynthesis among component species of dryland vegetation in the lli River Basin)

Figure 2-1: Profiles of Soil Moisture Content in Study Sites

Figure 2-2: Profiles of Soil Water Potential in Study Sites

Figure 3-1: Diurnal Changes of Stomatal Conductance of Component Species for Each Vegetation

Figure 3-2: Diurnal Changes of Stomatal Conductance of Haloxylon Individuals in Shrubland and Woodland

Figure 4: The Relationships Between Shrub Height and Stomatal Conductance in Individuals of Haloxylon
Shrubland

Figure 5: The Relationship Between Leaf Water Potential and Stomatal Conductance in Calligonum Aphyllum and Haloxylon Aphyllum.

On the other hand, the relationship between shrub height and stomatal conductance was not so clear in 1994 (figure 4). It is possibly due to the higher soil moisture content in shallow zones in 1994 (figure 2).

Discussion

There were spatial and time-course changes in dryland vegetation. Such changes could be interpreted as follows.

Vegetation types were associated with the water regime as river flow, ground water and physical property of soil. The functional properties were well characterized in leaf water potential and stomatal conductance of component species. In Haloxylon woodland near the Bereke Farm, soil moisture content was similar to Halimodendron scrub. This result suggests that this stand could be replaced by the other humid vegetation such as Halimodendron scrub. The construction of channels and farms not only produces crop fields but also affects the structure and type of surrounding vegetation.

Based on the results of the ecophysiological study, it could be deduced that stomatal conductance in shrubland was controlled by soil moisture in shallow zones. The effect was more distinct in small plants with shallow root systems. Consequently, it may reduce the growth or survival of plants. Since such soil moisture is supplied from precipitation, leaf biomass in dry vegetation would be affected by local precipitation. The fluctuation of LAI observed in the analysis of satellite data supports this consideration.

Not mentioned in this study, salinization should be another controlling factor of vegetation changes. Salinization was observed in a part of the delta where seasonal flooding had been suspended. It is possibly due to the changes of stream or the decrease of river flow affected by the construction of the Kapchagai Reservoir and channels.

For the production of an ecohydrological model, some areas remain to be solved. First, a geographical database is required as the base of such a model. Vegetation, topography and other environmental factors should be compiled on a computer system. For the construction and maintenance of the database, satellite data will be useful.

The second concern is the seasonal parameterization of water relations for each vegetation type based on the ecophysiological data. Mixing the principal property with stand structure derived from satellite data, ecophysiological function on stand scale shall be calculated.

The third one is the geographical extension of meteorological data with satellite data. For estimation of hydrological function, meteorological data is needed on a fine scale.

References:

Huete, A. R. (1988) A soil adjusted vegetation index (SAVI), Remote Sens. Environ. 25: 295-309.

Plisak, R. P. and N. P. Ogar (1992) The effect of reservoir on vegetation in arid zone. Gylym, Almaty, 228pp.

Plisak, R. P., N. P. Ogar and B. M. Sultanova (1989) Productivity and structure of grassland in desert zone. Nauka, Almaty, 186pp.

Walter, H. And E. O. Box (1983) The Karakum Desert, an example of a well-studied eu-biome. In: WEST, N. E. (ed.) Temperate deserts and semi-deserts. Elsevier, Amsterdam, Oxford, New York, 79-104.

Winter, K. (1981) C4 plants of high biomass in arid region of Asia -Occurrence of C4 photosynthesis in Chenopodiaceae and Polygonaceae from the Middle East and USSR. Oecologia, 48: 100-106.

Zalenskii, O. V. And T. Glagoleva (1981) Pathway of carbon metabolism in halophytic desert species from Chenopodiaceae. Photosynthetica, 15(2): 244-255.

Table of Contents

• Brochure
• 
• 
• International Year of Biodiversity
• 
•