Next Article in Journal
Development of Soil Fertility Index Using Machine Learning and Visible-Near-Infrared Spectroscopy
Previous Article in Journal
Carbon Losses from Topsoil in Abandoned Peat Extraction Sites Due to Ground Subsidence and Erosion
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 12
Issue 12
10.3390/land12122154
land-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Determination of Spatial Pattern of Environmental Consequences of Dams in Watersheds

by
Myung Sik Cho
* and
Jiaguo Qi
Center for Global Change and Earth Observations, Michigan State University, 1405 S Harrison Rd, East Lansing, MI 48823, USA
*
Author to whom correspondence should be addressed.
Land 2023, 12(12), 2154; https://doi.org/10.3390/land12122154
Submission received: 26 October 2023 / Revised: 5 December 2023 / Accepted: 11 December 2023 / Published: 12 December 2023
(This article belongs to the Section Land Systems and Global Change)
Browse Figures
Review Reports Versions Notes

Abstract

:
Many hydro-dams have been built for beneficial gains, but they are causing numerous unintended negative effects on the environment. The complexity of dam impacts and the insufficient knowledge of developing countries result in many uncertainties in managing land systems and environmental impact assessments (EIAs). Also, considering the consequences beyond the dam sites proved challenging in EIAs. In this context, this paper aims to determine the spatial pattern of the environmental consequences of dams, quantify the distance of impacts, and identify different patterns in both upstream and downstream areas. We considered the hydrological linkage of dams with wetlands in watersheds and the spatial relationship between watersheds to explicate the spatial pattern. Two hundred and ninety wetlands in the Mekong, Salween, and Irrawaddy Basins were categorized into those linked with dams and those without dams, and the hydrological characteristics were statistically compared for two periods (before and after the dam booms) using the Mann–Whitney U test. The watersheds having significant differences were found, and their spatial relationship in terms of location (i.e., upstream and downstream) and distance was determined by utilizing the HydroBASINS’ Pfafstetter coding system. The results indicate that the impacts of dams on downstream areas extend significantly farther than their effects on upstream regions in HydroBASINS level-7 representation. The quantitatively determined spatial patterns on upstream and downstream areas can provide accurate spatial baseline information in land system management and EIA.

1. Introduction

The complexity of the environmental consequences of hydro-dams brings many uncertainties in managing the land systems in sustainable ways. Globally, more than 38,000 dams are in operation [1], and new dams have been actively built in many developing countries to meet the increased water, food, and energy demands [2,3]. Along with the beneficial gains, dams cause irreversible changes to the environment, such as land degradation [4], ecosystem disturbances [5], and sediment blockage [6]. Even though accurate environmental impact assessments (EIAs) are necessary to respond to the negative environmental consequences of dams [7,8], the lack of measurement and understanding of the consequences of dams in developing countries limits the assessment of the influences of dams on the environment [7,9].
The spatial patterns of the environmental consequences of dams can improve the EIAs of dams by providing the spatial baseline information representing the spatial boundaries of the influence of dams on the environment [10,11]. Remote sensing approaches were used to determine the spatial patterns, based on their time-series observations over extensive space [9,12]. Some studies determined that the distance of the influence of dams on the environment is mostly within 5–10 km around the dam site [11,13] and up to 80 km [4]. While the spatial patterns of influences of dams on near areas were studied, those of dams on distant areas remain poorly characterized. The hydrological influences of dams were determined through watersheds, while the watersheds were studied using in situ measurements because dams disrupt the natural river flow in watershed scales [9,14]. To quantify the consequences, previous studies compared streamflow discharge from gauging stations before and after dam construction [15,16], isolated the dam impacts from climate variability [17], and modeled the influences of dams on the downstream streamflow [18]. However, spatial patterns of influences of dams on watersheds were qualitatively determined. For example, a few distantly located stations were selected to examine the consequences of dams without the consideration of whether the impacts of dams on the hydrology reach the selected stations since the spatial boundaries of dam impacts were not quantitatively determined [19,20]. This approach is problematic in quantifying the environmental results of dams, especially in developing regions like Southeast Asia, because they have insufficient stations to characterize the hydrological alteration caused by dams [21,22]. In this context, providing the spatial pattern of the dam impacts is effective in assessing the environmental results from the dams.
Hydrologically connected wetlands with dams can represent the spatial patterns of the environmental impacts of dams on watersheds due to their hydrological linkage with dams at the watershed level [14]. Wetlands are sensitive to external pressure, such as hydrological changes [23], climate variability [24], and land use/land cover changes [25]. Changes in inundated areas of wetlands are one of the most significant hydrological alterations caused by dams [26,27]. The different hydrological alteration in wetland areas was found by the location (e.g., upstream and downstream) and proximity of the wetlands to dams [14,28,29,30]. For example, dams increased inundated areas of upstream wetlands and closely located downstream wetlands, but the inundation of downstream wetlands decreased in proportion to their distance from the dam [14]. This showed that the characteristics of hydrological alteration in wetlands can be used to determine the spatial pattern of dam impacts on watersheds. Also, the determination of the spatial pattern using the wetland changes and their location in regard to dam placement enables the quantification of the environmental consequences of dams in watershed scales, since the hydrological changes of wetlands affect their ecosystems and local livelihoods [31]. Thus, characterizing the spatial patterns of environmental consequences of dams can provide EIAs and help establish more appropriate environmental management for dams in watershed scales.
The spatial pattern of influences of dams on watersheds in Southeast Asia needs to be studied. The Mekong, Irrawaddy, and Salween Basins in the mainland of Southeast Asia are the 10th largest river in the mean annual flow [32], the fourth highest river in the total dissolved load [33], and the 26th largest river in annual discharge [21], respectively. The three basins are one of the hotspots of dam construction in the 21st century. As of 2017, 186 were already commissioned, with 45 under construction, and another 110 proposed and planned [34]. However, the problem is that dams in the region are mostly planned project-by-project without consideration of their cumulative impacts on river processes [35]. Additionally, foreign investment was the main driver of dam projects in the region [36], so the negative consequences of dams were less considered during the construction [7,37]. In this context, 96% of sediment is expected to be trapped when all proposed dams are built [22], and the massive changes caused by dams would seriously affect the environment as well as society. For example, 2.1 million people in the floodplain of the Mekong Basin would suffer from a decline in protein intake [38], because the fish would decrease due to the hydrological alteration [39]. Thus, it is necessary to suggest the spatial pattern of the consequences of dams in three basins in Southeast Asia to establish comprehensive policies.
The objective of this paper is to determine the spatial patterns of environmental consequences of dams on the watersheds using the statistical analysis of the hydrological alterations in wetland areas and their spatial relationships (i.e., location and distance) with dams in watershed scales. Changes in monthly wetland inundation from 1987 to 2020 were used to characterize the alteration of wetland areas. The Pfafstetter coding system, which was developed to represent spatial relationships between watersheds, was utilized to examine the spatial relationships between watersheds of wetlands and dams and determine the spatial pattern of the influences of dams on watersheds. The structure would be that Section 2 illustrates the study area and materials used in this study. The detailed methods, including the analysis of hydrological characteristics of wetland areas (Section 3.1), the rules of deciding location (Section 3.2) and distance (Section 3.3) of watersheds of wetland and distances, the quantification of consequences of dams on watersheds (Section 3.4), and the contribution of climate variability on wetlands (Section 3.5), were introduced. Then, the findings were summarized in Section 4 and discussed in Section 5, and the overall research was concluded in Section 6.

2. Study Area and Materials

2.1. Study Area

The Mekong, Salween, and Irrawaddy Basins in the mainland of Southeast Asia were selected as study areas because many dams have been constructed recently that impose strong impacts on the local communities and ecosystems (Figure 1). The three basins are the largest in the region, and they cover seven countries: Vietnam, Thailand, Laos, Cambodia, Myanmar, China, and India [34]. The three basins share a similar climate, topography, and societal characteristics. The region is controlled by the Southwest Asian monsoon and East Asian monsoon with wet and dry seasons [40]. Also, El Niño Southern Oscillation (ENSO) causes interannual climatic variabilities. El Niño events are responsible for a decrease in rainfall, while La Niña increases rainfall [41]. The climate system in the region is a major factor affecting hydrology, which plays a significant role in ecosystems and society [23,42,43]. For topography, the Mekong River starts from the Tibetan Plateau at 4500 m and flows out to the South China Seas through the Mekong Delta with 14,500 m3/s of mean annual discharge and 4800 km of length [32,44,45]. The Irrawaddy River starts from the Himalayan Mountains and flows out to the Andaman Sea through the Irrawaddy Delta with 13,000 m3/s of mean annual discharge and 2170 km of length [33,42]. The Salween River originates at the Tibetan Plateau and flows out to the Gulf of Martaban through the Salween Delta with 210 km3/year of average annual discharge and 3200 km of length [21,43]. For the upstream areas of the three basins, snowmelts serve as the major water source [43,46], while precipitation and groundwater are also the major water sources [42,46]. For the Mekong Basin, tributaries contribute to 84% of river flow [44]. The population in the region highly relies on foods from rainfed and irrigation rice, and population growth requires more water for irrigation [27,47]. Also, the population is consuming 47–82% of their protein from wetlands and river fisheries [39,48].

2.2. Materials

2.2.1. Watershed

This study considered watershed levels from 7 to 11 in HydroBASINS to examine the spatial pattern of influences of dams on watersheds. Spatial boundaries of watersheds systematically showed the hydrological systems over the region [49], so the HydroBASINS dataset, which is the most popular dataset representing global watershed boundaries [50], was facilitated to obtain watershed boundaries. The HydroBASINS was derived from the World Wildlife Fund’s HydroSHEDS (hydrological data and maps based on SHuttle Elevation Derivatives at multiple Scales), which provide hydrographic information based on Shuttle Radar Topography Mission (SRTM) [51]. Watershed boundaries in the HydroBASINS were delineated in a hierarchical sub-basin breakdown using the Pfafstetter coding system by considering upstream, downstream, or non-linkage [52]. HydroBASINS have 12 hierarchical levels of watersheds, where level 1 represents the largest watersheds (i.e., highest watershed level) and level 12 represents the smallest watersheds (i.e., lowest watershed level). In other words, level 1 breaks into sub-basins of level 2, level 2 is divided into sub-basins of level 3, and so on. In this study, we used levels 7 to 11 based on the preliminary analysis of the selection of watershed levels. In our preliminary analysis, watershed levels that are larger than those from 1 to 6 had large boundaries, making it difficult to separate watersheds with dams from those without dams. Also, watershed level 12 had small boundaries so it could not provide enough sample sizes.

2.2.2. Dams

This study considered 134 commissioned hydropower dams, which began operation from 1966 to 2017 (Figure 1). Dam information was collected from the “2017 Dataset on the Dams of the Irrawaddy Mekong, Red, and Salween River Basins” generated by the Greater Mekong Consultative Group on International Agricultural Research (CGIAR) Research Program on Water, Land, and Ecosystems [34]. The dam information was collected by official reports and visual checks on earth observations. Since the dataset is the most reliable and popularly used dam data for Southeast Asia, we used the “2017 Dataset on the Dams of the Irrawaddy Mekong, Red, and Salween River Basins” for our study [22].

2.2.3. Wetlands

We selected 290 natural wetlands from HydroLAKES, which is the only available dataset providing wetland information over the region. The reasons for focusing on natural wetlands are that (i) naturally occurring wetlands are closely connected to the hydrological systems so they are sensitive to hydrological alterations within the watersheds [53], and (ii) artificially generated water bodies are affected by more anthropogenic stressors due to their original purposes (e.g., irrigation reservoirs) [54]. The HydroLAKES dataset was derived from the eight water products, including MOD44W and the Global Lakes and Wetlands Database (GLWD). The dataset has 1,427,688 water bodies over 10 ha, so it is widely used for various studies [55,56].
Since HydroLAKES provides insufficient information determining the natural wetlands, we manually selected the natural wetlands through visual inspections of a variety of images (i.e., Google Earth, Sentinel-2 A/B, and Landsat 5–8). In the manual selection for the natural wetlands, we found evidence of dammed or artificially created structures in the satellite images because artificial water bodies were blocked by the structure to inundate areas. Even though our method possibly included artificial water bodies and excluded natural wetlands, this method is the only available approach to identify the natural wetlands. Then, since the wetlands in Southeast Asia have large seasonal variations in the inundations, we updated the water masks by extending the coverage and combining them with other water masks suggested by [14].

2.2.4. Water-Inundated Areas of Wetlands

The water pixels of the European Commission’s Global Water Surface Layer (GWSL) were used to quantify the monthly wetland areas from 1987 to 2020. The GWSL was derived from the three million Landsat images using expert systems based on a procedural sequential decision tree classification [57]. This product provides three classified pixels, including water, land, and cloud, on annual and monthly scales at 30 m spatial resolution globally. Since the GWSL is the only available dataset providing the monthly inundated areas of wetlands over the region over 30 years, it was selected for monitoring the water-inundated areas of wetlands to examine the changes in the trend, intra-annual variability, and amplitude of wetland inundations (details will be discussed in Section 3).

2.2.5. Precipitation Data

Monthly precipitation observations, which were derived from TerraClimate, over wetlands from 1987 to 2020 were used to determine the contribution of precipitation to changes in wetland inundations. TerraClimate is based on climate observations and climate reanalysis datasets [58] and provides climate observations in grids at a 4638.3 m spatial resolution. Since precipitation around the wetlands could not be monitored by field observations due to the insufficiency of weather stations [21], we used a grid-formatted product to examine the precipitation.

3. Methods

3.1. Hydrological Characteristics of Wetlands from Monthly Inundation Areas

Monthly inundation areas from 1987 to 2020 were delineated using 290 water masks defined in Section 2.2.3 by extracting water pixels in the GWSL monthly datasets. Since the GWSL monthly datasets were based on optical sensors, they provided limited water delineations during wet seasons [57]. We removed the inundation areas, which covered over 10% of cloud cover based on cloud labels in the GWSL monthly datasets, to reduce the cloud cover issues. On average, we could get 4.9 monthly observations for a wetland inundation per year and 159 monthly observations per wetland.
Three characteristics of wetland inundations—trend, intra-annual variability, and amplitude of inundation areas—were considered to examine whether dams affect the wetlands. These three characteristics were selected because dams significantly (1) affect the amount of water, which results in changes in the trend of the wetland inundations; (2) control the timing of water release, which leads to the changes in the intra-annual variability of the water inundations; and (3) store and release water depending on the water demands, which cause different maximum and minimum water inundations [17,53]. The three characteristics of wetland inundations were calculated from the monthly inundated areas each year. Then, the trend of the annual characteristics of the inundations from 1987 to 2020 was calculated for each wetland to examine the influence of dams on the wetlands.
First, the trend of inundation refers to the tendency of monthly water inundation areas, and it shows whether the inundated areas are increased or decreased. It was calculated by the seasonal Kendall Trend test, which is a nonparametric test analyzing whether seasonal data are changed in monotonic trends [59] (1)–(4):
S i = ∑ k = 1 n i − 1 ∑ j = k + 1 n i s g n x i j − x i k  
s g n   x i j − x i k = 1 ,         x i j − x i k > 0 0 ,         x i j − x i k = 0 − 1 ,         x i j − x i k < 0    
where x i j is an observation (an inundated area) of year j in month i, and n i is the observation of year n in month i. s g n x i j − x i k decides the value 1, 0, and −1 according to its sign. For example, x i j is larger than the x i k , then the value 1 is assigned. S i indicates whether the positive value (negative value) shows that inundated areas in month i in the later years tend to be larger (smaller) than those in month i in earlier years.
V a r   S i = n i n i − 1 2 n i + 5 − ∑ p = 1 g i t i p ( t i p − 1 ) ( 2 t i p + 5 )  
where g i is the number of tied groups for the month i, and t i p is the number of data in the group p for month i. Then, S ’ and V A R ( S ’ ) are calculated by summing all values of S i and V a r   S i , respectively, from month 1 to 12 (i).
T r = S ’ − 1 ( V a r   S ’ ) 1 / 2 ,     S ’ > 0 0 ,     S ’ = 0 S ’ + 1 ( V a r   S ’ ) 1 / 2 ,     S ’ < 0
where Tr is the result of the seasonal Kendall trend test for a wetland over the observations. The positive value indicates that the inundation areas tend to increase over time, and the negative value shows the decreases in the inundation areas over time. The trend of the inundated areas of wetlands (Tr) was used for the study, but the significance of levels was not considered. The slope was converted to binary variables for the logistic regression model: 1 (positive) and 0 (negative).
Second, the intra-annual variability refers to the degree of fluctuations in monthly inundation areas on an annual scale. Since wetlands in the region have apparent increasing and decreasing patterns of inundations within a year due to dry and wet seasons [40], the intra-annual variability shows whether the fluctuations were affected by dams or not. The measure was based on the method suggested by Feng et al. [60], which compares the distribution of monthly inundated areas for a year by the uniform distribution, which has stable inundated areas over a year (5)–(8):
I ¯ k = ∑ m = 1 12 i k , m  
which is total inundated areas at the monthly scale for a year k, and i k , m is an inundated area for a month m in year k.
P ¯ m , k = i k , m I ¯ k  
which is the probability distribution of the total inundated areas for a month k in year y.
D ¯ k = ∑ m = 1 12 P ¯ m , k · log 2 ⁡ ( 12 · P ¯ m , k )
which measures the distance between the distribution of an observed monthly inundated area for a year k and the uniform distribution of monthly inundated areas (i.e., 1/12).
I V k = D ¯ k · I ¯ k I ¯ m a x  
where I ¯ m a x is the maximum inundated areas for the entire observations. I V k measures the intra-annual variability for a year k by comparing the observed distribution with the uniform distribution. I V k is 0 when the inundated areas are same for a year k, and I V k is maximized (at log 2 ⁡ 12 = 3.585) when a wetland is inundated only for one month.
Third, amplitude refers to differences between the maximum and minimum peaks of inundated areas in a year, and it shows whether there are storing or releasing effects of dams on the wetlands. The measure was based on (9):
A m p = I k ,   m a x − I k , m i n
where I k ,   m a x is the maximum inundated area and I k , m i n is the minimum inundated area for a year k. The amplitude was calculated for each year, and the trend of the amplitude was calculated using Sen’s Slope Estimator for robustness.

3.2. Pfafstetter Codes: Determination of the Location of Wetlands Regarding the Dam Placement

The Pfafstetter codes were used to determine whether wetlands have hydrological linkage with dams and (i.e., linked or not) the location of wetlands regarding the dam placement (i.e., upstream or downstream). The concept of the Pfafstetter coding system was first articulated by Otto Pfafstetter [61], and it was spatially realized by Verdin and Verdin [52]. Watersheds have unique streamflow directions, so their arrangements have network order, such as upstream or downstream. The Pfafstetter code was designed to represent this watershed characteristic using the base-10 numbering system. The Pfafstetter coding system uses digits effectively; its digit can represent the hierarchical relationships of a watershed (e.g., a watershed and its sub-watersheds), and its number can show the topographical upstream and downstream relationship, so Pfafstetter codes are useful in watershed topological analyses and multi-scale hydrological analyses [52,62,63]. HydroBASINS provide watershed boundaries with Pfafstetter codes (the column named “Pfaf_id” in HydroBASINS) based on the Pfafstetter coding system from watershed level 1 (i.e., highest in watershed hierarchical systems) to 12 (i.e., lowest in watershed hierarchical systems).
HydroBASINS used the number of digits in the Pfafstetter coding system to represent watershed levels. For example, a watershed with the Pfafstetter code 7454825, which is seven digits, indicates that its watershed level is 7 (Figure 2). The digits (i.e., 0–9) in the Pfafstetter code can indicate the hydrological relationships of watersheds as follows: (i) a watershed located upstream of another watershed, (ii) a watershed located downstream of another watershed, or (iii) a watershed not hydrologically linked with another watershed. To examine the meaning of the digit, the concepts of “tributary watershed” and “inter-basin watershed” suggested by Verdin and Verdin (1999) should be understood (Figure 2). The “tributary watershed” refers to a watershed separated from the confluence to the upstream of the tributary, which flows to the mainstream. The “inter-basin tributary” refers to a watershed included in the mainstream located between two tributary watersheds. In a selected number of digits, tributary watersheds were labeled in even numbers (2, 4, 6, and 8), and inter-basin watersheds were labeled in odd numbers (1, 3, 5, 7, and 9). For example, a watershed having the Pfafstetter code 7454825 is an inter-basin watershed in watershed level 7 (7454825) and a tributary watershed in watershed level 6 (745482). The digits from 1 to 9 were sequentially labeled from the mouth (i.e., downstream) to the source (i.e., upstream) by adversely tracing the streamflow, and the digit 0 was used for the case that a watershed boundary is the same with the boundary of higher watershed level (Figure 2). When the digits reach 9, labeling the digit for a certain number of digits is completed, and a watershed that is unlabeled and located next to the watershed with the digit 9 is broken into sub-watersheds (Figure 2). Then, the sub-watersheds were labeled again from digits 1 to 9 for the next certain number of digits (i.e., lower watershed level). For instance, a watershed having the Pfafstetter code 745489 (watershed level 6) breaks into sub-watersheds (watershed level 7) with Pfafstetter codes 7454891, 7454892, 7454893, 7454894, 7454895, 7454896, 7454897, 7454898, and 7454899.
Using the principles of labeling the Pfafstetter codes, we interpreted the codes to examine whether a wetland is hydrologically linked with hydro-dams (i.e., located upstream or downstream) or is not hydrologically linked with dams. The relationship of wetlands with dams was determined by checking the last number of digits for a certain watershed level using the following rules:
  • If the digit for the last number of digits is odd,
    • Upstream: the watersheds with Pfafstetter code having the larger number for the last number of digits.
    • Downstream: the watersheds with Pfafstetter code having the smaller odd number for the last number of digits.
  • If the digit for the last number of digits is even,
    • Upstream: all sub-watersheds will be upstream (i.e., watersheds with the lower watershed level will be upstream).
    • Downstream: the watersheds with Pfafstetter code having the smaller odd number for the last number of digits.
  • If the digit for the last number of digits is 0: Skip this.
If the last digits of the Pfafstetter codes are odd numbers, it indicates that the watershed is an inter-basin watershed. Upstream refers to all the watersheds located upstream from a selected watershed (1-a). Given that the tributary watershed lacks inter-basin watersheds for upstream watersheds, the downstream for the inter-basin watersheds (i.e., the digit for the last number of digits is odd) would be inter-basin watersheds located downstream from the selected watershed (1-b). Let us say there is a watershed 7454825. In this example, upstream watersheds would be coded as 7454826, 7454827, 7454828, and 7454829. Meanwhile, downstream watersheds would be coded as 7454821 and 7454823 (Figure 2).
If the last digit of the Pfafstetter code is an even number, it represents a tributary watershed that does not have upstream watersheds in the selected digit. Instead, its upstream watersheds are its sub-watersheds, as they flow into the tributary watershed (2-a). Like the inter-basin watersheds (i.e., the digit for the last number of digits is odd), the downstream for the tributary watersheds would be inter-basin watersheds located downstream from the selected watershed. For example, if the code number is 745482, then its downstream watershed would be coded as 745481 and its upstream sub-watersheds would be 745482, with the codes 7454821, 7454822, 7454823, 7454824, 7454825, 7454826, 7454827, 7454828, and 7454829 respectively.
Using the rules mentioned above, the relationship between a wetland and the closest dam was determined, starting from the last digit of the Pfafstetter code. If the last digit shows that the wetland is located upstream or downstream from a dam, then the location of the wetland in regard to dam placement is decided. If the last digit did not show this, we considered the penultimate digit to determine the location of a wetland to the closest dam. This procedure was carried out until (i) the location of a wetland was determined in a certain number of digits (i.e., 4–11 digits), or (ii) the procedure was at the first third number of digits. Since the first third number of digits does not indicate the topographical relationship (i.e., upstream or downstream), we determined them as wetlands not having hydrological linkage with dams on the watershed scale. Using this method, the location of wetlands in regard to dam placements was investigated using the above rules from watershed level 7 to level 11.
For example, a hydro-dam is in watershed 745482 in watershed level 6. Another is in watershed 7454825 in watershed level 7 (Figure 2). Wetlands are in the watersheds 7454822, 7454823, and 7454829 in watershed level 7, in addition to 745481, 745482 (sub-watersheds have wetlands), and 745486 in watershed level 6. In watershed level 7, the watershed 7454823 is located downstream from the dam, while the watershed 7454829 is located upstream from the dam. This is because their final digits are odd, making them inter-basin watersheds. For the wetland in watershed 7454822, there are no hydrological links with the dam, as the wetland is in the tributary watershed (the last number of digits is even) and the dam is in the inter-basin watershed (the last digit is odd). In watershed level 6, a wetland in watershed 745481 is located downstream from the dam, while watershed 745486 has no hydrological links with the dam.

3.3. Spatial Units for Watersheds

The concept of “spatial units” for watersheds was devised to quantify spatial patterns of the influence of dams on watersheds (Figure 3). Here, a spatial unit for watersheds refers to the distance measurement of watersheds from watersheds with dams. For example, a watershed having a selected dam is coded as spatial unit 0, which means that dams are located in this watershed. Direct upstream and downstream watersheds are coded as spatial unit 1, as there is only one watershed unit separating them from the dam (Figure 3). Subsequent watersheds are coded as spatial units 2 and 3, as they are two and three watershed units away from the watershed with the dam, respectively (Figure 3). Under this concept, the spatial units for watersheds were labeled by considering the distance of watersheds to watersheds with dams.
For instance, in Figure 2, watershed 7454823 (watershed level 7) is downstream from the dam (7454825) in spatial unit 1, while watershed 7454829 is upstream from the dam in spatial unit 2. The watershed 7454822 does not have a spatial unit assigned, because it is not hydrologically linked with the dam. In watershed level 6, the watershed 745481 is downstream from the dam (745482) in spatial unit 1. The watershed 745486 does not have a spatial unit. Finally, watershed 745482 contains the dam, so it is coded as a spatial unit is 0. Using this rule in combination with the rule described in Section 3.2, spatial units for watersheds were labeled for wetlands to represent the location of wetlands in relation to dam placement.

3.4. Quantification of Spatial Pattern of Dam Results on Watersheds

The Mann–Whitney U test was used to quantify the spatial pattern of the consequences of dams on watersheds (Figure 4). The Mann–Whitney U test is a nonparametric test of the null hypothesis of whether two groups are from the same population [64], thereby the Mann–Whitney U test enabled us to statistically determine the influence of dams on watersheds by comparing watersheds having linkage with dams (i.e., spatial units 1, 2, and 3) and having no linkage with dams (i.e., no spatial unit). The hydrological alterations in wetlands (i.e., trend, intra-annual variability, and amplitude of inundated areas) were targeted observations in the Mann–Whitney U test. Since the null hypothesis of the Mann–Whitney U test used in this analysis was that watersheds having dam linkage and ones not having dam linkage have similar hydrological characteristics of wetland inundations, the rejection of the null hypothesis indicates that the hydrological alteration in wetlands was not similar in watersheds with and without dams.
The results of the Mann–Whitney U test for two periods (i.e., before and after the boom of dams) were examined to determine the influence of dams on watersheds. When the results become significant (i.e., the rejection of the null hypothesis of the Mann–Whitney U test) after the boom of dams, it means that dams caused the different hydrological alteration between two watersheds (i.e., one with dam linkage and one without dam linkage). Under this rule, each spatial unit for watersheds was analyzed to find the spatial units having significant influences on dams. Also, since little is known about which watershed level accounts for the impacts of dams on watersheds, this study conducted the Mann–Whitney U test for each watershed level (i.e., 7–11) to select the appropriate watershed level, which elucidates the influences of dams on watersheds in specific spatial units. Additionally, the results of the Mann–Whitney U test in terms of spatial units and watershed level were compared in upstream and downstream watersheds to examine the different consequences of dams on upstream and downstream watersheds.
The entire study period (1987–2020) was divided into two periods: (i) before the boom of dams (P1) and (ii) after the boom of dams (P2) to examine how dams affect the wetlands for quantifying the influences of dams on hydrological alteration in dams (Figure 5). Since dams in Southeast Asia have been built since 1966, we could not select a year showing any dam and after dams. Instead, we selected the year 2008 to divide the periods because dams dramatically have been commissioned since 2008 (Figure 5). The reason for not relating the commissioned year of individual dams to a wetland was that dams have cumulative influences on wetlands [17]; therefore, it was difficult to determine which dam affects a wetland. By dividing the entire period into two by 2008, the P1 (before the boom of dams; 1987–2007) can represent hydrological characteristics with fewer dams in the region and the P2 (after the boom of dams; 2008–2020) can represent hydrological characteristics with more dams in the region. During the P1, 59 dams were commissioned, and 75 dams were commissioned during the P2.
In summary, the Mann–Whitney tests were conducted for every spatial unit in the watershed level from 7 to 11 by comparing hydrological characteristics of wetlands for the watersheds without dam linkage with dams to (i) upstream-located wetlands and (ii) downstream-located wetlands. The spatial unit with the failure to the rejection of the null hypothesis (p ≥ 0.1) for the P1 (before the boom of dams; 1987–2007) and the rejection of the null hypothesis (p < 0.1) for the P2 (after the boom of dams; 2008–2020) was selected for every watershed level. Then, a watershed level having the rejection with consecutive spatial units from spatial unit 1 was considered as the appropriate watershed level because dams should have more effects on closely located watersheds [23,65]. For example, a watershed level having the selection of spatial units 2 and 3, but not having the selection of spatial unit 1, was not considered as the appropriate watershed level. Instead, a watershed level having the selection of spatial units 1 and 2 was selected as the appropriate watershed level. The selected spatial units in the appropriate watershed level were considered as the spatial boundaries of the impacts of dams on the watersheds.

3.5. Investigation of Influence of Climate Variability on the Inundation

The influence of climate variability on the changes in wetland areas was investigated. Climate variability is another main driving force on the changes in the wetland inundations [24,66], so the characterization of the influence of climate variability on wetland areas, which can support our findings in Section 3.4, can be explained by dams. Given that precipitation is a major factor of wetland inundation among other climate factors [67], the Pearson correlation analysis between precipitation and inundation areas was conducted to examine the influences of precipitation on the changes in wetland areas from 1987 to 2020. The monthly inundated areas of wetlands in the spatial boundaries of impacts of dams determined by Section 3.4 and the monthly precipitation of the TerraClimate pixels covering the wetlands were subjective to the correlation analysis. The analysis was conducted for the P1 (before the boom of dams; 1987–2007 due to the data availability of the TerraClimate, see the details in Section 2.2.5) and the P2 (after the boom of dams; 2008–2020), respectively, and the standardization was conducted for precipitation and inundation areas because of the differences in units. If the result of the correlation analysis was not different between the P1 and P2, then the impacts of climate variability on the wetland inundation were not changed between for two periods. This can exclude the influences of climate variability on wetlands and confirm the dam impacts on wetlands [17,68].

4. Results

4.1. Mann–Whitney U Tests for Various Levels and Spatial Units

The watershed level 7 in HydroBASINS was an appropriate watershed level representing the impacts of dams on watersheds (Table 1). In Table 1, at watershed level 7, spatial units 2 and 3 are selected for downstream and spatial unit 1 is chosen for upstream. At watershed level 8, spatial units from 2 to 4 are selected for downstream, but the non-selection of spatial unit 2 for upstream led to an inappropriate representation of water levels at level 8 representing the impacts of dams on watersheds. At watershed level 9, spatial units 3 and 4 are selected for downstream, but there was a failure in selecting spatial units 2 and 4 for upstream. At watershed level 10, spatial units from 4 to 6 are selected for downstream, but all spatial units for upstream failed to be selected. At watershed level 11, spatial units 4 and 5 were selected for downstream but failed to select spatial units 2 and 3 for upstream. In consideration of the results of the Mann–Whitney U test for both upstream and downstream areas, watershed level 7 was selected as an appropriate watershed level.
The selection of spatial units of watershed level 7 showed the spatial boundary of the impacts of dams on watersheds (Table 2). For the downstream watersheds, the spatial unit 3 showed the significance in intra-annual variability and amplitude of the wetland inundation for the P2 (p-values are 0.011 and 0.014, respectively), but the insignificance for the P1 (0.269 and 0.218). For the upstream watersheds, spatial unit 1 showed the significance in trend, intra-annual variability, and amplitude of the wetland inundation for the P2 (p-values are 0.094, 0.023, and 0.067, respectively), but the insignificance for the P1 (0.565, 0.459 and 0.379).
Spatial unit 3 for downstream and spatial unit 1 for upstream can be represented spatially and their spatial boundaries can identify the spatial patterns of the impacts of dams on watersheds (Figure 6 and Figure 7). For the impacts of dams on the upstream watersheds, the distances range from 1.1 km to 655.2 km, with an average distance of 133.4 km and a median distance of 55.8 km. The areas range from 1251 km2 to 39,026 km2 with an average of 12,356 km2 and a median of 11,695 km2. For the impacts of dams on the downstream watersheds, the distances range from 33.7 km to 1577.5 km with an average distance of 641 km and a median distance of 441.4 km. The areas range from 22,588 km2 to 463,544 km2 with an average of 165,361 km2 and a median of 153,452 km2.

4.2. Correlation Analysis between Precipitation and Inundation Areas

The Pearson correlation analysis was conducted for 19 wetlands that are within the spatial boundaries of impacts of dams on wetlands (Figure 6), which are described in Section 3.1. The 15 wetlands did not change the relationship (i.e., significance → significance or insignificance → insignificance) between the precipitation and inundated areas of wetlands from P1 to P2 (Table 3). The change in inundated areas of one wetland became correlated with the precipitation (insignificance → significance), while the changes in three wetlands became not correlated with the precipitation (significance → insignificance). Excluding the one wetland, 18 wetlands showed that precipitation did not relate to the inundated wetlands (Table 3). In other words, we can say that the changes in the characteristics of the inundated areas of wetlands can be explained by the influences of dams on wetlands, as we discussed in Section 3.1.

5. Discussion

The spatial pattern of the influences of dams on watersheds was quantitatively determined using the spatial relationships between watersheds with dams and wetlands. Our watershed-shaped spatial pattern represents a more accurate spatial boundary of consequences of dams compared to the buffer-shaped spatial patterns. Previous studies defined the spatial patterns of the influences of dams on surrounding areas using buffer-shaped boundaries, including circular buffers on 10 km [13], 50 km [2], and 80 km [4], linear buffers along the river 2 km [69], and ellipsoidal buffer of 21 km for the major axis and 16 km for the minor axis [11]. The buffer-shaped patterns can represent the environmental consequences of dams in near surroundings but have limitations in capturing the hydrological effects of dams on distant areas [14]. Hydrological alteration causes significant changes in sediment [19], water quality [70], fishery [71], agriculture [72], and local livelihoods [36] beyond the dam sites, so the buffer-shaped patterns can only partially characterize the consequences of dams [12,17]. Contrarily, watersheds, defined by hydrological relationships over space, are useful spatial units to represent the consequences of dams [9,73]. In this context, our study proposed more accurate spatial patterns to estimate the influences of dams on the environment.
The proposed watershed-shaped spatial pattern has implications for land system management and EIAs. Previously, the spatial pattern was qualitatively determined so the land system management and EIA related to dams were inaccurate. In the Mekong Basin, the EIA and social assessment for the Yali Dam project, which was operated in 2000 in Vietnam, considered the impacts 6 km downstream, although the negative consequences were reported 70 km downstream in Cambodia [10]. Due to EIA’s inaccurate spatial base information, there was no compensation for villages in Cambodia [10]. Since existing EIA only considered near locations due to insufficient definition of the spatial pattern [7,9] and limited the influences within a country [7,74], the suggested quantitative spatial pattern of the influences of dams on watersheds can be used to set the spatial boundaries for EIAs. In particular, the focus on watersheds in this study enables us to consider the distant and transboundary consequences of dams, which cause conflicts among countries and regions due to unequal distribution of benefits and loss from dams.
The different spatial patterns of dam impacts on downstream and upstream watersheds showed the anisotropy pattern of the environmental consequences of dams, which can lead to a more accurate EIA (Table 1 and Figure 5). EIAs did not consider the influences of dams on upstream and downstream differently [7], except for dislocating people living in the potential impounded areas; however, our findings can suggest the different spatial baseline information for upstream and downstream areas. We found that the impacts on downstream areas reach several times more distant than those on upstream areas. Interestingly, this finding can be linked with that the downstream stations showed twice as large hydrological alteration values as the upstream ones [30]. Along with the intensity in hydrological alterations being two times larger downstream, this study confirmed impacts on downstream areas that were several times farther.
The validation of our findings is challenging because there is no quantitative approach to determine the spatial pattern of the consequences of dams, so further studies are necessary to confirm the spatial pattern. Instead, our findings based on the quantitative approach can be compared with the qualitative approaches in the same region. Previous studies in the Mekong Basin considered that the consequences can be shown within approximately 700 km [20,75,76,77]. This qualitatively determined spatial distance (700 km) is close to our quantitative downstream distance (641 km). Also, Xue et al. [68] addressed that upstream dams did not have significant impacts on downstream discharge in 1600 km. This study can support our findings [30].
Given that insufficient spatial baseline information largely contributed to inaccurate EIA [7,78], our finding on spatial patterns of influences of dams on the environment can help estimate an accurate cost of dams. For example, 245 major global dams had 96% higher actual costs than the predicted costs [79], and costs from EIA established by Chinese enforcement were estimated at only 10% of the actual cost [7]. In this context, further investigation of the influences of dams on ecology and society based on our spatial boundary is necessary for a better EIA. Basin-wide research using (1) in situ measurement, including streamflow [15], total suspended sediment [19], water quality [70], and biotic communities [80]; (2) hydrological modeling [18]; and (3) remote sensing measurement [4,14] within watersheds, where potentially impacted by dams (i.e., our proposed spatial boundary), can enable us to better understand the consequences of dams. Ultimately, this can contribute to the accurate EIA and dams as long-term sustainable drivers of economic growth [81].

6. Conclusions

This paper determined the spatial patterns of the influences of dams on watersheds based on the changes in inundation characteristics of wetlands regarding the location of dams in perspectives of watersheds. The Pfafstetter coding system was used to identify the spatial relationship of wetlands with dams (i.e., upstream, downstream, or no hydrological linkage). Also, watershed level and spatial unit of watersheds were considered to quantify the spatial boundaries of impacts of dams. The HydroBASINS level 7’s spatial unit 1 was the spatial boundary for the dam impacts on the upstream, and its spatial unit 3 was that on the downstream. The spatial impacts of dams reach up to 33.7–1577.5 km downstream and 1.1–655.2 km upstream in distance. The dams affect the watershed coverage of up to 22,588–463,544 km2 downstream and 1251–39,026 km2 upstream. This different spatial boundary showed the anisotropy of the spatial pattern of the dams in the watersheds. Since the impacts of dams on watersheds were qualitatively determined, our findings fill the gaps in assessing the environmental impacts of dams by the spatial baseline information in a quantitative way. Further investigation of the consequences of dams using in situ measurement, hydrological models, and remote sensing methods based on our suggested spatial boundaries can contribute to better EIAs for dams.
However, the monthly inundation data that we used in our analysis (i.e., GWSL) had gaps in observations due to cloud cover issues. The GWSL detected water bodies by only using the Landsat series, which are optical sensors, so it failed to detect the water bodies under the cloud. Given the 4-month wet season in Southeast Asia, the gaps can be problematic. The synthetic-aperture radar (SAR) can fill the gaps by penetrating the cloud covers. However, the high-resolution SAR, Sentinel-1, began its mission in 2014 and prior SARs have insufficient spatial resolution to detect the small-sized wetlands. The integration of SAR and optical sensors can improve the monitoring of wetland areas. Also, our method needs to be extended to global watersheds to examine the spatial patterns of the influence of dams on watersheds. Given that our study area was a tropical region with a monsoon climate, the application of this method to other regions, such as arid and temperate climate regions, warrants further exploration.

Author Contributions

Conceptualization, M.S.C. and J.Q.; Methodology, M.S.C.; Writing—original draft, M.S.C.; Writing—review & editing, M.S.C. and J.Q.; Visualization, M.S.C.; Funding acquisition, J.Q. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the NASA Land-Cover and Land-Use Change (LCLUC) Program (grant numbers 80NSSC18K1134).

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to the inclusion of sensitive information.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Mulligan, M.; van Soesbergen, A.; Sáenz, L. GOODD, a Global Dataset of More than 38,000 Georeferenced Dams. Sci. Data 2020, 7, 31. [Google Scholar] [CrossRef] [PubMed]
  2. Fan, P.; Cho, M.S.; Lin, Z.; Ouyang, Z.; Qi, J.; Chen, J.; Moran, E.F. Recently Constructed Hydropower Dams Were Associated with Reduced Economic Production, Population, and Greenness in Nearby Areas. Proc. Natl. Acad. Sci. USA 2022, 119, e2108038119. [Google Scholar] [CrossRef]
  3. Poff, N.L.R.; Schmidt, J.C. How Dams Can Go with the Flow. Science 2016, 353, 1099–1100. [Google Scholar] [CrossRef] [PubMed]
  4. Chen, G.; Powers, R.P.; Carvalho, L.M.T.D.; Mora, B. Spatiotemporal Patterns of Tropical Deforestation and Forest Degradation in Response to the Operation of the Tucuruí Hydroelectricdam in the Amazon Basin. Appl. Geogr. 2015, 63, 1–8. [Google Scholar] [CrossRef]
  5. Han, X.; Feng, L.; Hu, C.; Chen, X. Wetland Changes of China’s Largest Freshwater Lake and Their Linkage with the Three Gorges Dam. Remote Sens. Environ. 2018, 204, 799–811. [Google Scholar] [CrossRef]
  6. Ezcurra, E.; Barrios, E.; Ezcurra, P.; Ezcurra, A.; Vanderplank, S.; Vidal, O.; Villanueva-Almanza, L.; Aburto-Oropeza, O. A Natural Experiment Reveals the Impact of Hydroelectric Dams on the Estuaries of Tropical Rivers. Sci. Adv. 2019, 5, aau9875. [Google Scholar] [CrossRef]
  7. Tullos, D. Assessing the Influence of Environmental Impact Assessments on Science and Policy: An Analysis of the Three Gorges Project. J. Environ. Manag. 2009, 90, S208–S223. [Google Scholar] [CrossRef]
  8. Tilt, B.; Braun, Y.; He, D. Social Impacts of Large Dam Projects: A Comparison of International Case Studies and Implications for Best Practice. J. Environ. Manag. 2009, 90, S249–S257. [Google Scholar] [CrossRef]
  9. Winemiller, K.O.; McIntyre, P.B.; Castello, L.; Fluet-Chouinard, E.; Giarrizzo, T.; Nam, S.; Baird, I.G.; Darwall, W.; Lujan, N.K.; Harrison, I.; et al. Balancing Hydropower and Biodiversity in the Amazon, Congo, and Mekong. Science 2016, 351, 128–129. [Google Scholar] [CrossRef]
  10. Baird, I.G.; Silvano, R.A.M.; Parlee, B.; Poesch, M.; Maclean, B.; Napoleon, A.; Lepine, M.; Hallwass, G. The Downstream Impacts of Hydropower Dams and Indigenous and Local Knowledge: Examples from the Peace–Athabasca, Mekong, and Amazon. Environ. Manag. 2021, 67, 682–696. [Google Scholar] [CrossRef]
  11. Cho, M.S.; Qi, J. Quantifying Spatiotemporal Impacts of Hydro-Dams on Land Use / Land Cover Changes in the Lower Mekong River Basin. Appl. Geogr. 2021, 136, 102588. [Google Scholar] [CrossRef]
  12. Abrial, E.; Lorenzón, R.E.; Rabuffetti, A.P.; Blettler, M.C.M.; Espínola, L.A. Hydroecological Implication of Long-Term Flow Variations in the Middle Paraná River Floodplain. J. Hydrol. 2021, 603, 126957. [Google Scholar] [CrossRef]
  13. Ouyang, W.; Shan, Y.; Hao, F.; Shi, X.; Wang, X. Accumulated Impact Assessment of River Buffer Zone after 30 Years of Dam Disturbance in the Yellow River Basin. Stoch. Environ. Res. Risk Assess. 2013, 27, 1069–1079. [Google Scholar] [CrossRef]
  14. Cho, M.S.; Qi, J. Characterization of the Impacts of Hydro-Dams on Wetland Inundations in Southeast Asia. Sci. Total Environ. 2023, 864, 160941. [Google Scholar] [CrossRef]
  15. Han, Z.; Long, D.; Fang, Y.; Hou, A.; Hong, Y. Impacts of Climate Change and Human Activities on the Flow Regime of the Dammed Lancang River in Southwest China. J. Hydrol. 2019, 570, 96–105. [Google Scholar] [CrossRef]
  16. Lu, X.X.; Li, S.; Kummu, M.; Padawangi, R.; Wang, J.J. Observed Changes in the Water Flow at Chiang Saen in the Lower Mekong: Impacts of Chinese Dams? Quat. Int. 2014, 336, 145–157. [Google Scholar] [CrossRef]
  17. Poff, N.L.R.; Olden, J.D.; Merritt, D.M.; Pepin, D.M. Homogenization of Regional River Dynamics by Dams and Global Biodiversity Implications. Proc. Natl. Acad. Sci. USA 2007, 104, 5732–5737. [Google Scholar] [CrossRef]
  18. Chaudhari, S.; Pokhrel, Y. Alteration of River Flow and Flood Dynamics by Existing and Planned Hydropower Dams in the Amazon River Basin. Water Resour. Res. 2022, 58, e2021WR030555. [Google Scholar] [CrossRef]
  19. Bussi, G.; Darby, S.E.; Whitehead, P.G.; Jin, L.; Dadson, S.J.; Voepel, H.E.; Vasilopoulos, G.; Hackney, C.R.; Hutton, C.; Berchoux, T.; et al. Impact of Dams and Climate Change on Suspended Sediment Flux to the Mekong Delta. Sci. Total Environ. 2021, 755, 142468. [Google Scholar] [CrossRef]
  20. Zheng, Y.; Zhang, G.; Wu, Y.; Xu, Y.J.; Dai, C. Dam Effects on Downstream Riparian Wetlands: The Nenjiang River, Northeast China. Water 2019, 11, 2038. [Google Scholar] [CrossRef]
  21. Salmivaara, A.; Kummu, M.; Keskinen, M. Using Global Datasets to Create Environmental Profiles for Data-Poor Regions: A Case from the Irrawaddy and Salween River Basins. Environ. Manag. 2013, 51, 897–911. [Google Scholar] [CrossRef]
  22. Bonnema, M.; Hossain, F. Assessing the Potential of the Surface Water and Ocean Topography Mission for Reservoir Monitoring in the Mekong River Basin. Water Resour. Res. 2019, 55, 444–461. [Google Scholar] [CrossRef]
  23. Dang, T.D.; Cochrane, T.A.; Arias, M.E.; Van, P.D.T.; de Vries, T.T. Hydrological Alterations from Water Infrastructure Development in the Mekong Floodplains. Hydrol. Process. 2016, 30, 3824–3838. [Google Scholar] [CrossRef]
  24. Erwin, K.L. Wetlands and Global Climate Change: The Role of Wetland Restoration in a Changing World. Wetl. Ecol. Manag. 2009, 17, 71–84. [Google Scholar] [CrossRef]
  25. Ekumah, B.; Armah, F.A.; Afrifa, E.K.A.; Aheto, D.W.; Odoi, J.O.; Afitiri, A.-R. Assessing Land Use and Land Cover Change in Coastal Urban Wetlands of International Importance in Ghana Using Intensity Analysis. Wetl. Ecol. Manag. 2020, 28, 271–284. [Google Scholar] [CrossRef]
  26. Wang, J.; Sheng, Y.; Tong, T.S.D. Monitoring Decadal Lake Dynamics across the Yangtze Basin Downstream of Three Gorges Dam. Remote Sens. Environ. 2014, 152, 251–269. [Google Scholar] [CrossRef]
  27. Arias, M.E.; Cochrane, T.A.; Piman, T.; Kummu, M.; Caruso, B.S.; Killeen, T.J. Quantifying Changes in Flooding and Habitats in the Tonle Sap Lake (Cambodia) Caused by Water Infrastructure Development and Climate Change in the Mekong Basin. J. Environ. Manag. 2012, 112, 53–66. [Google Scholar] [CrossRef]
  28. Middleton, B.A.; Souter, N.J. Functional Integrity of Freshwater Forested Wetlands, Hydrologic Alteration, and Climate Change. Ecosyst. Health Sustain. 2016, 2, e01200. [Google Scholar] [CrossRef]
  29. Kirchherr, J.; Pohlner, H.; Charles, K.J. Cleaning up the Big Muddy: A Meta-Synthesis of the Research on the Social Impact of Dams. Environ. Impact Assess. Rev. 2016, 60, 115–125. [Google Scholar] [CrossRef]
  30. Timpe, K.; Kaplan, D. The Changing Hydrology of a Dammed Amazon. Sci. Adv. 2017, 3, e1700611. [Google Scholar] [CrossRef]
  31. Piman, T.; Cochrane, T.A.; Arias, M.E.; Green, A.; Dat, N.D. Assessment of Flow Changes from Hydropower Development and Operations in Sekong, Sesan, and Srepok Rivers of the Mekong Basin. J. Water Resour. Plan. Manag. 2013, 139, 723–732. [Google Scholar] [CrossRef]
  32. Ziv, G.; Baran, E.; Nam, S.; Rodriguez-Iturbe, I.; Levin, S.A. Trading-off Fish Biodiversity, Food Security, and Hydropower in the Mekong River Basin. Proc. Natl. Acad. Sci. USA 2012, 109, 5609–5614. [Google Scholar] [CrossRef]
  33. Robinson, R.A.J.; Bird, M.I.; Nay, W.O.; Hoey, T.B.; Maung, M.A.; Higgitt, D.L.; Lu, X.X.; Aung, S.; Tin, T.; Swe, L.W. The Irrawaddy River Sediment Flux to the Indian Ocean: The Original Nineteenth-Century Data Revisited. J. Geol. 2007, 115, 629–640. [Google Scholar] [CrossRef]
  34. WLE. Dataset on the Dams of the Irrawaddy, Mekong, Red and Salween River Basins; WLE: Sydney, Australia, 2017. [Google Scholar]
  35. Schmitt, R.J.P.; Bizzi, S.; Castelletti, A.; Kondolf, G.M. Improved Trade-Offs of Hydropower and Sand Connectivity by Strategic Dam Planning in the Mekong. Nat. Sustain. 2018, 1, 96–104. [Google Scholar] [CrossRef]
  36. Soukhaphon, A.; Baird, I.G.; Hogan, Z.S. The Impacts of Hydropower Dams in the Mekong River Basin: A Review. Water 2021, 13, 265. [Google Scholar] [CrossRef]
  37. Fearnside, P.M. Environmental and Social Impacts of Hydroelectric Dams in Brazilian Amazonia: Implications for the Aluminum Industry. World Dev. 2016, 77, 48–65. [Google Scholar] [CrossRef]
  38. Xu, J.; Grumbine, R.E.; Beckschäfer, P. Landscape Transformation through the Use of Ecological and Socioeconomic Indicators in Xishuangbanna, Southwest China, Mekong Region. Ecol. Indic. 2014, 36, 749–756. [Google Scholar] [CrossRef]
  39. Dugan, P.J.; Barlow, C.; Agostinho, A.A.; Baran, E.; Cada, G.F.; Chen, D.; Cowx, I.G.; Ferguson, J.W.; Jutagate, T.; Mallen-cooper, M.; et al. Fish Migration, Dams, and Loss of Ecosystem Services in the Mekong Basin. Ambio 2010, 39, 344–348. [Google Scholar] [CrossRef]
  40. Räsänen, T.A.; Kummu, M. Spatiotemporal Influences of ENSO on Precipitation and Flood Pulse in the Mekong River Basin. J. Hydrol. 2013, 476, 154–168. [Google Scholar] [CrossRef]
  41. Frappart, F.; Biancamaria, S.; Normandin, C.; Blarel, F.; Bourrel, L.; Aumont, M.; Azemar, P.; Vu, P.L.; Le Toan, T.; Lubac, B.; et al. Influence of Recent Climatic Events on the Surface Water Storage of the Tonle Sap Lake. Sci. Total Environ. 2018, 636, 1520–1533. [Google Scholar] [CrossRef]
  42. Shamsudduha, M.; Panda, D.K. Spatio-Temporal Changes in Terrestrial Water Storage in the Himalayan River Basins and Risks to Water Security in the Region: A Review. Int. J. Disaster Risk Reduct. 2019, 35, 101068. [Google Scholar] [CrossRef]
  43. Bridgestock, L.; Henderson, G.M.; Holdship, P.; Khant, W.; Thu, W.M.; Aung, M.; Chi, N.; Jotautas, J.; Tipper, E.; Chapman, H.; et al. Dissolved Trace Element Concentrations and Fluxes in the Irrawaddy, Salween, Sittaung and Kaladan Rivers. Sci. Total Environ. 2022, 841, 156756. [Google Scholar] [CrossRef]
  44. Mekong River Commission (MRC). Overview of the Hydrology of the Mekong Basin; Mekong River Commission: Vientiane, Laos, 2005; p. 73. [Google Scholar]
  45. Wang, W.; Lu, H.; Ruby Leung, L.; Li, H.Y.; Zhao, J.; Tian, F.; Yang, K.; Sothea, K. Dam Construction in Lancang-Mekong River Basin Could Mitigate Future Flood Risk from Warming-Induced Intensified Rainfall. Geophys. Res. Lett. 2017, 44, 10–378. [Google Scholar] [CrossRef]
  46. Pokhrel, Y.; Burbano, M.; Roush, J.; Kang, H.; Sridhar, V.; Hyndman, D.W. A Review of the Integrated Effects of Changing Climate, Land Use, and Dams on Mekong River Hydrology. Water 2018, 10, 266. [Google Scholar] [CrossRef]
  47. Yoshida, Y.; Lee, H.S.; Trung, B.H.; Tran, H.D.; Lall, M.K.; Kakar, K.; Xuan, T.D. Impacts of Mainstream Hydropower Dams on Fisheries and Agriculture in Lower Mekong Basin. Sustainability 2020, 12, 2408. [Google Scholar] [CrossRef]
  48. Intralawan, A.; Wood, D.; Frankel, R.; Costanza, R.; Kubiszewski, I. Tradeoff Analysis between Electricity Generation and Ecosystem Services in the Lower Mekong Basin. Ecosyst. Serv. 2018, 30, 27–35. [Google Scholar] [CrossRef]
  49. Palacios-Cabrera, T.; Valdes-Abellan, J.; Jodar-Abellan, A.; Rodrigo-Comino, J. Land-Use Changes and Precipitation Cycles to Understand Hydrodynamic Responses in Semiarid Mediterranean Karstic Watersheds. Sci. Total Environ. 2022, 819, 153182. [Google Scholar] [CrossRef] [PubMed]
  50. Peters, R.; Berlekamp, J.; Lucía, A.; Stefani, V.; Tockner, K.; Zarfl, C. Integrated Impact Assessment for Sustainable Hydropower Planning in the Vjosa Catchment (Greece, Albania). Sustainability 2021, 13, 1514. [Google Scholar] [CrossRef]
  51. Lehner, B.; Grill, G. Global River Hydrography and Network Routing: Baseline Data and New Approaches to Study the World’s Large River Systems. Hydrol. Process. 2013, 27, 2171–2186. [Google Scholar] [CrossRef]
  52. Verdin, K.L.; Verdin, J.P. A Topological System for Delineation and Codification of the Earth’s River Basins. J. Hydrol. 1999, 218, 1–12. [Google Scholar] [CrossRef]
  53. Richter, B.D.; Baumgartner, J.V.; Powell, J.; Braun, D.P. A Method for Assessing Hydrologic Alteration within Ecosystems. Conserv. Biol. 1996, 10, 1163–1174. [Google Scholar] [CrossRef]
  54. Brock, M.A.; Smith, R.G.B.; Jarman, P.J. Drain It, Dam It: Alteration of Water Regime in Shallow Wetlands on the New England Tableland of New South Wales, Australia. Wetl. Ecol. Manag. 1999, 7, 37–46. [Google Scholar] [CrossRef]
  55. Cooley, S.W.; Ryan, J.C.; Smith, L.C. Human Alteration of Global Surface Water Storage Variability. Nature 2021, 591, 78–81. [Google Scholar] [CrossRef] [PubMed]
  56. Messager, M.L.; Lehner, B.; Grill, G.; Nedeva, I.; Schmitt, O. Estimating the Volume and Age of Water Stored in Global Lakes Using a Geo-Statistical Approach. Nat. Commun. 2016, 7, 13603. [Google Scholar] [CrossRef] [PubMed]
  57. Pekel, J.-F.; Cottam, A.; Gorelick, N.; Belward, A.S. High-Resolution Mapping of Global Surface Water and Its Long-Term Changes. Nature 2016, 540, 418–422. [Google Scholar] [CrossRef]
  58. Abatzoglou, J.T.; Dobrowski, S.Z.; Parks, S.A.; Hegewisch, K.C. TerraClimate, a High-Resolution Global Dataset of Monthly Climate and Climatic Water Balance from 1958–2015. Sci. Data 2018, 5, 170191. [Google Scholar] [CrossRef]
  59. Hirsch, R.M.; Slack, J.R.; Smith, R.A. Techniques of Trend Analysis for Monthly Water Quality Data. Water Resour. Res. 1982, 18, 107–121. [Google Scholar] [CrossRef]
  60. Feng, X.; Porporato, A.; Rodriguez-Iturbe, I. Changes in Rainfall Seasonality in the Tropics. Nat. Clim. Chang. 2013, 3, 811–815. [Google Scholar] [CrossRef]
  61. Pfafstetter, O. Classification of Hydrographic Basins: Coding Methodology; Departamento Nacional de Obras de Saneamento: Brasilia, Brazil, 1989; Volume 18, pp. 1–2. [Google Scholar]
  62. Fürst, J.; Hörhan, T. Coding of Watershed and River Hierarchy to Support GIS-Based Hydrological Analyses at Different Scales. Comput. Geosci. 2009, 35, 688–696. [Google Scholar] [CrossRef]
  63. Lehner, B.; Verdin, K.; Jarvis, A. New Global Hydrography Derived from Spaceborne Elevation Data. Eos Trans. Am. Geophys. Union 2008, 89, 93–94. [Google Scholar] [CrossRef]
  64. Mann, H.B.; Whitney, D.R. On a Test of Whether One of Two Random Variables Is Stochastically Larger than the Other. Ann. Math. Stat. 1947, 18, 50–60. [Google Scholar] [CrossRef]
  65. Zhang, Z.; Jin, G.; Tang, H.; Zhang, S.; Zhu, D.; Xu, J. How Does the Three Gorges Dam Affect the Spatial and Temporal Variation of Water Levels in the Poyang Lake? J. Hydrol. 2022, 605, 127356. [Google Scholar] [CrossRef]
  66. Junk, W.J. Long-Term Environmental Trends and the Future of Tropical Wetlands. Environ. Conserv. 2002, 29, 414–435. [Google Scholar] [CrossRef]
  67. Reis, V.; Hermoso, V.; Hamilton, S.K.; Bunn, S.E.; Fluet-Chouinard, E.; Venables, B.; Linke, S. Characterizing Seasonal Dynamics of Amazonian Wetlands for Conservation and Decision Making. Aquat. Conserv. 2019, 29, 1073–1082. [Google Scholar] [CrossRef]
  68. Xue, Z.; Liu, J.P.; Ge, Q. Changes in Hydrology and Sediment Delivery of the Mekong River in the Last 50 Years: Connection to Damming, Monsoon, and ENSO. Earth Surf. Process. Landf. 2011, 36, 296–308. [Google Scholar] [CrossRef]
  69. Jiang, X.; Lu, D.; Moran, E.; Calvi, M.F.; Dutra, L.V.; Li, G. Examining Impacts of the Belo Monte Hydroelectric Dam Construction on Land-Cover Changes Using Multitemporal Landsat Imagery. Appl. Geogr. 2018, 97, 35–47. [Google Scholar] [CrossRef]
  70. Ahmad, S.K.; Hossain, F.; Holtgrieve, G.W.; Pavelsky, T.; Galelli, S. Predicting the Likely Thermal Impact of Current and Future Dams Around the World. Earths Future 2021, 9, e2020EF001916. [Google Scholar] [CrossRef]
  71. Barbarossa, V.; Schmitt, R.J.P.; Huijbregts, M.A.J.; Zarfl, C.; King, H.; Schipper, A.M. Impacts of Current and Future Large Dams on the Geographic Range Connectivity of Freshwater Fish Worldwide. Proc. Natl. Acad. Sci. USA 2020, 117, 3648–3655. [Google Scholar] [CrossRef]
  72. Rufin, P.; Gollnow, F.; Müller, D.; Hostert, P. Synthesizing Dam-Induced Land System Change. Ambio 2019, 48, 1183–1194. [Google Scholar] [CrossRef]
  73. Grill, G.; Ouellet Dallaire, C.; Fluet Chouinard, E.; Sindorf, N.; Lehner, B. Development of New Indicators to Evaluate River Fragmentation and Flow Regulation at Large Scales: A Case Study for the Mekong River Basin. Ecol. Indic. 2014, 45, 148–159. [Google Scholar] [CrossRef]
  74. Castro, M.C.; Krieger, G.R.; Balge, M.Z.; Tanner, M.; Utzinger, J.; Whittaker, M.; Singer, B.H. Examples of Coupled Human and Environmental Systems from the Extractive Industry and Hydropower Sector Interfaces. Proc. Natl. Acad. Sci. USA 2016, 113, 14528–14535. [Google Scholar] [CrossRef]
  75. Graf, W.L. Downstream Hydrologic and Geomorphic Effects of Large Dams on American Rivers. Geomorphology 2006, 79, 336–360. [Google Scholar] [CrossRef]
  76. Maingi, J.K.; Marsh, S.E. Quantifying Hydrologic Impacts Following Dam Construction along the Tana River, Kenya. J. Arid Environ. 2002, 50, 53–79. [Google Scholar] [CrossRef]
  77. Pal, S.; Saha, T.K. Identifying Dam-Induced Wetland Changes Using an Inundation Frequency Approach: The Case of the Atreyee River Basin of Indo-Bangladesh. Ecohydrol. Hydrobiol. 2018, 18, 66–81. [Google Scholar] [CrossRef]
  78. Vaidyanathan, G. Dam Controversy: Remaking the Mekong. Nature 2011, 478, 305–307. [Google Scholar] [CrossRef]
  79. Latrubesse, E.M.; Arima, E.Y.; Dunne, T.; Park, E.; Baker, V.R.; D’Horta, F.M.; Wight, C.; Wittmann, F.; Zuanon, J.; Baker, P.A.; et al. Damming the Rivers of the Amazon Basin. Nature 2017, 546, 363–369. [Google Scholar] [CrossRef]
  80. Van Looy, K.; Tormos, T.; Souchon, Y. Disentangling Dam Impacts in River Networks. Ecol. Indic. 2014, 37, 10–20. [Google Scholar] [CrossRef]
  81. Faria, F.A.M.D.; Davis, A.; Severnini, E.; Jaramillo, P. The Local Socio-Economic Impacts of Large Hydropower Plant Development in a Developing Country. Energy Econ. 2017, 67, 533–544. [Google Scholar] [CrossRef]
Land 12 02154 g001
Figure 1. A map of the location of dams and wetlands in three basins (Mekong, Salween, and Irrawaddy Basins) in mainland Southeast Asia. Blue lines indicate the Irrawaddy, Salween, and Mekong Rivers from the left of the map, and the black thick line represents the national boundary.
Figure 1. A map of the location of dams and wetlands in three basins (Mekong, Salween, and Irrawaddy Basins) in mainland Southeast Asia. Blue lines indicate the Irrawaddy, Salween, and Mekong Rivers from the left of the map, and the black thick line represents the national boundary.
Land 12 02154 g001
Land 12 02154 g002
Figure 2. Pfafstetter codes for identifying topographical relationships of watersheds with examples of the watersheds with the Pfafstetter codes in the watershed level 6 (left) and sub-watersheds of the watershed with the code 745482 in the watershed level 7 (right). The last number of the digit with the 1 is the most downstream watershed and one with the 9 is the most upstream watershed. The last number of the digit with even numbers are tributary watersheds (black colored font) and the ones with odd numbers were inter-basin, which is a mainstream located between two tributaries (red colored font).
Figure 2. Pfafstetter codes for identifying topographical relationships of watersheds with examples of the watersheds with the Pfafstetter codes in the watershed level 6 (left) and sub-watersheds of the watershed with the code 745482 in the watershed level 7 (right). The last number of the digit with the 1 is the most downstream watershed and one with the 9 is the most upstream watershed. The last number of the digit with even numbers are tributary watersheds (black colored font) and the ones with odd numbers were inter-basin, which is a mainstream located between two tributaries (red colored font).
Land 12 02154 g002
Land 12 02154 g003
Figure 3. A conceptual figure for spatial units for watersheds. Starting from the watershed with a dam (spatial unit 0), the watersheds located both upstream and downstream were labeled from spatial unit 1. The watersheds without hydrological linkage were not labeled.
Figure 3. A conceptual figure for spatial units for watersheds. Starting from the watershed with a dam (spatial unit 0), the watersheds located both upstream and downstream were labeled from spatial unit 1. The watersheds without hydrological linkage were not labeled.
Land 12 02154 g003
Land 12 02154 g004
Figure 4. The workflow of conducting Mann–Whitney U test over HydroBASINS levels, locations (upstream or downstream), and spatial units using three hydrological characteristics of wetlands.
Figure 4. The workflow of conducting Mann–Whitney U test over HydroBASINS levels, locations (upstream or downstream), and spatial units using three hydrological characteristics of wetlands.
Land 12 02154 g004
Land 12 02154 g005
Figure 5. A histogram for the commissioned year of selected dams. The red dotted line is dividing the P1 (1987–2007) and the P2 (2008–2020) by the year 2018. Many dams have been commissioned since 2008, so the year 2008 was selected as a breaking point.
Figure 5. A histogram for the commissioned year of selected dams. The red dotted line is dividing the P1 (1987–2007) and the P2 (2008–2020) by the year 2018. Many dams have been commissioned since 2008, so the year 2008 was selected as a breaking point.
Land 12 02154 g005
Land 12 02154 g006
Figure 6. Spatial boundaries of the distant impacts of dams on upstream watersheds (inclined lines) and downstream watersheds (grey-filled color).
Figure 6. Spatial boundaries of the distant impacts of dams on upstream watersheds (inclined lines) and downstream watersheds (grey-filled color).
Land 12 02154 g006
Land 12 02154 g007
Figure 7. Distance and areas of impact of dams on watersheds. The box shows 25–75% of distribution, and whiskers represent maximum and minimum values. ‘X’ indicates the mean values and the horizontal line in the box indicates the median values of distribution.
Figure 7. Distance and areas of impact of dams on watersheds. The box shows 25–75% of distribution, and whiskers represent maximum and minimum values. ‘X’ indicates the mean values and the horizontal line in the box indicates the median values of distribution.
Land 12 02154 g007
Table 1. The summary of results of comparing Mann–Whitney U tests for P1 (before the boom of dams) and P2 (after the boom of dams). The comparison was conducted over upstream (Up) and downstream (Down), watershed levels (7–11), and spatial units (1–8). o indicates that there was a failure to the rejection of the null hypothesis during the P1 (p ≥ 0.1) and there was the rejection of the null hypothesis during the P2 (p < 0.1). This represents that there are significant consequences of dams over the watersheds. Contrarily, x indicates that there are no significant consequences of dams over the watersheds. NA indicates insufficient sample sizes to conduct the Mann–Whitney U tests (n < 5).
Table 1. The summary of results of comparing Mann–Whitney U tests for P1 (before the boom of dams) and P2 (after the boom of dams). The comparison was conducted over upstream (Up) and downstream (Down), watershed levels (7–11), and spatial units (1–8). o indicates that there was a failure to the rejection of the null hypothesis during the P1 (p ≥ 0.1) and there was the rejection of the null hypothesis during the P2 (p < 0.1). This represents that there are significant consequences of dams over the watersheds. Contrarily, x indicates that there are no significant consequences of dams over the watersheds. NA indicates insufficient sample sizes to conduct the Mann–Whitney U tests (n < 5).
Spatial UnitsLevel 7Level 8Level 9Level 10Level 11
DownUpDownUpDownUpDownUpDownUp
1NAoNANANANANANANANA
2o oxNAxNAxNAx
3o oooxNAxNAx
4x o oooxoo
5 x x o o
6 x o x
7 x x
8 o
Table 2. p-values from Mann–Whitney U tests for the watershed level 7. The Mann–Whitney U tests were conducted to compare the watersheds with wetlands not having the hydrological linkage to (i) downstream-located watersheds and (ii) upstream-located watersheds. Var. indicates the intra-annual variability of the wetland inundations. * indicates the significance (p < 0.1), and † indicates insufficient sample sizes in Mann–Whitney U tests (the minimum sample size for the test is 5).
Table 2. p-values from Mann–Whitney U tests for the watershed level 7. The Mann–Whitney U tests were conducted to compare the watersheds with wetlands not having the hydrological linkage to (i) downstream-located watersheds and (ii) upstream-located watersheds. Var. indicates the intra-annual variability of the wetland inundations. * indicates the significance (p < 0.1), and † indicates insufficient sample sizes in Mann–Whitney U tests (the minimum sample size for the test is 5).
Spatial UnitPeriodDownstream-Located WatershedsUpstream-Located Watersheds
CountsTrendVar.AmplitudeCountsTrendVar.Amplitude
1P14 †0.1460.9930.729100.5650.4590.379
P20.8070.003 *0.017 *0.094 *0.023 *0.067 *
2P160.2430.2930.223
P20.7730.015 *0.020 *
3P1120.049 *0.2690.218
P20.8800.011 *0.014 *
4P1390.074 *0.8190.554
P20.1210.1770.007 *
5P1390.074 *0.8190.554
P20.1210.1770.007 *
Table 3. Results of correlation between precipitation and inundated areas of wetlands.
Table 3. Results of correlation between precipitation and inundated areas of wetlands.
P2
SignificanceInsignificance
P1Significance03
Insignificance115
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Cho, M.S.; Qi, J. Determination of Spatial Pattern of Environmental Consequences of Dams in Watersheds. Land 2023, 12, 2154. https://doi.org/10.3390/land12122154

AMA Style

Cho MS, Qi J. Determination of Spatial Pattern of Environmental Consequences of Dams in Watersheds. Land. 2023; 12(12):2154. https://doi.org/10.3390/land12122154

Chicago/Turabian Style

Cho, Myung Sik, and Jiaguo Qi. 2023. "Determination of Spatial Pattern of Environmental Consequences of Dams in Watersheds" Land 12, no. 12: 2154. https://doi.org/10.3390/land12122154

APA Style

Cho, M. S., & Qi, J. (2023). Determination of Spatial Pattern of Environmental Consequences of Dams in Watersheds. Land, 12(12), 2154. https://doi.org/10.3390/land12122154

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop