Wetlands associated with floodplains of the Ganges and Brahmaputra river systems cover an estimated 0.2 million ha and play a vital role in the fisheries, rural economy and environment. In recent years, most wetlands in the states of Bihar, Assam and West Bengal in India have lost connectivity with their parent rivers due to natural and anthropogenic changes. These changes are causing rapid shrinkage of valuable wetlands and threatening their biodiversity and ecological function. The present study was conducted to assess the effect of river connectivity on the hydrochemistry, sediment enzyme activity, and biotic communities of these wetlands. One wetland area having link channels (open) connecting the main river, and another without link channels (closed), were selected from each of the Ganges and Brahmaputra basins for the study. Comparative analysis of open and closed, ecologically distinct wetlands of the Ganges basin revealed greater seasonal reduction in water depth and higher water conductivity (p < 0.04), nutrients (Ca+2, Mg+2, SiO3-Si), sediment microbial activity (p < 0.004), phosphorous cycling enzymes (p < 0.001), carbon cycling enzymes (p < 0.01), organic matter, conductivity, and plankton and macrozoobenthos (p < 0.01) density in the closed wetland. However, the open wetland showed higher diversity of plankton and macrozoobenthos. A profound impact of connectivity on wetland fisheries was observed in the Brahmaputra basin in Assam, where a higher percentage of catch comprised of indigenous fish species with lower fish yield was noted in the open wetland.
The Ganges and Brahmaputra river systems are the lifeline of the North, East and Northeastern parts of the Indian land mass encompassing the vast basin and characterized by an extensive network of floodplains. Wetlands associated with floodplains of these river systems cover an estimated 0.2 million ha and play a vital role in the fisheries, rural economy and environment of these areas. The wetlands in India are mainly concentrated in the states of West Bengal, Assam, Bihar and Uttar Pradesh. Both the Ganges and Brahmaputra river systems are characterized by a wide array of habitats required for various life history stages of fishes, aquatic invertebrates, birds, amphibians and reptiles. A single physical feature that determines the wetland habitat characteristics is its hydrologic connectivity with the parent river. The connectivity regime of the wetlands, which may be designated as perennial, seasonal, or closed, influences the hydrology, ecology, biotic communities, and fisheries to a varied extent (Reddy et al., 1999; Gilchrest and Schmidt, 1998). Seasonal connection and flooding from the parent river regulate the wetland environment for fish food organisms and the growth, survival, and general well being of indigenous fish. Seasonal or perennial flushing of wetlands through connecting channels maintain a natural hydrologic condition that removes excess organic matter (Poff et al., 1997; Amoros et al., 2000), inhibits competitive exclusion of macrophytes (Wilson and Keddy, 1985), and maintains macrophyte diversity (Wilcox, 1995).
Aquatic enzymes including dehydrogenase, alkaline phosphatase (Jansson et al., 1988), acid phosphatase, and β-glucosidase play a significant role in regenerating nutrients from sediment organic matter (Sinsabaugh et al., 1991) for their utilization by rooted plants, microbes, phytobenthos, and planktonic autotrophs.
To assess the impact of various natural and anthropogenic drivers like hydroperiod, drought (Freeman et al., 1997), pollutants (Demanou et al., 2004), macrophyte type and degree of infestation, nutrient impaction (Pregner and Reddy, 2004; Wright and Reddy, 2001), eutrophication (Zhou et al., 2002), and restoration interventions (Bossio et al., 2006) on wetland ecology and fisheries, various authors have used water quality and sediment biochemical parameters and biotic communities as indicators. In India, a similar study in the Assam wetlands by Manna and Aftabuddin (2007) described the effect of seasonal connectivity on the chemical parameters of water and sediment, along with diversity of molluscs and macrophytes. The loss of connectivity between wetland and river can exert an overall adverse impact on ecosystem processes and functions, as well as on fish production and diversity. In recent years, most wetlands in the states of Bihar, Assam, and West Bengal in India have lost their connectivity with their parent river due to low rainfall, siltation, flood control measures, road construction and intensive water withdrawal for irrigation. This has ultimately resulted in rapid shrinkage of valuable wetland resources and threatened wetland ecological function and biodiversity.
The objectives of the present study were to assess the effect of river connectivity on wetland hydrochemistry, sediment enzyme, and chemical properties, and to examine the density and diversity of biotic communities (macrozoobenthos, plankton and fish).
Materials and methods
One open and one closed wetland from each of the Ganges and Brahmaputra river basins were selected for this study. In the Ganges basin, two wetlands of West Bengal—the Chhari Ganga (23°16.625′; 88°22.72′) with a perennial connection to the Ganges River, and the Bhomra (22°58.450′; 88° 37.819′) with a defunct link channel (Figure 1)—were studied to understand the effect of connectivity on various hydrochemical parameters, sediment enzyme activity, and the density and diversity of plankton and benthic communities. In the Brahmaputra basin, two wetlands of Assam—the Samaguri (26°25.408′; 92°51.342′), a cut off river meander with feeble flow throughout the year with river Kolong (a tributary of Brahmaputra), and Haribhanga (26°31. 154′; 92°47.554′), a natural depression without a link channel connecting it to any river (Figure 1)—were selected to study the effect of connectivity on fish production and fish stock structure. Water, sediment, plankton, and macrozoobenthos samples were collected from six stations, including both pelagic and littoral zones of both Gangetic wetlands during monsoon, winter, and summer seasons.
The water quality variables, such as temperature, pH, and specific conductivity, were measured in situ using a multi-probe meter (WTW universal meter, Model; Multiline P4). Titrimetric methods were used to determine dissolved oxygen, carbon dioxide, total alkalinity, total hardness, Ca+2 and Mg+2 following APHA (1998) guidelines. Silicate (Si) was estimated by colourimetric method (APHA, 1998). Depth and transparency were measured using the Secchi disc method.
Duplicate sediment samples were collected by Ekman grab from each site and the top 15 cm of sediment was taken for analysis. The samples were then pooled, mixed, and passed through a 2.0 mm sieve to collect sediment samples which were brought to the laboratory in iced condition for enzyme analyses. The sediment temperature, pH, oxidation-reduction potential, and specific conductivity were measured in the field using a multi-probe meter (WTW universal meter, Model; Multiline P4). A part of the sample was dried in the shade for organic matter estimation, while another part was kept frozen for enzyme estimation. Sediment organic matter was determined in a muffle furnace by incinerating a moisture-free sample at 600°C for 6 h and expressed as % loss of dry matter due to ignition.
Standard procedures were used for estimation of sediment enzymes: dehydrogenase (Casida et. al., 1964), alkaline phosphatase, acid phosphatase (Tabatabai and Bremmer, 1969), and β-Glucosidase (Eivazi and Tabatabai, 1988). Dehydogenase activity was estimated by incubating a calcium carbonate treated 0.2–0.3 g moist sediment slurry with 100 μl of 3% 2,3,5-triphenyltetrazolium chloride in the dark for 18 h at 37°C. The product triphenylformazan (TPF) was extracted with 10–12 ml acetone and read at 485 nm using a UV-visible spectrophotometer. Enzyme activity was expressed as μg formazan produced hr−1g−1 dry soil.
For analysis of alkaline phosphatase, acid phosphatase, and β-glucosidase, modified universal buffer (MUB) adjusted to pH values of 11.0, 6.5 and 6.0, respectively, was used for making a slurry of 0.3–0.5 g moist soil. Paranitrophenyl phosphate (pNPP) was used as the substrate for alkaline and acid phosphatase activity by dissolving 0.84 g pNPP in 100 ml of MUB at pH 11.0 and pH 6.5, respectively. Paranitrophenyl-β-d-glucopyranoside dissolved in MUB at pH 6.0 was used as the substrate for β-glucosidase. The estimation protocol was similar for all three enzymes: approximately 0.3–0.5 g of moist soil was slurried in a test tube with 4.5 ml of MUB at pH 6.0, followed by addition of 0.5 ml of substrate. Substrate, sample, and buffer were mixed and incubated at 37°C for 1 h. 1.0 ml (0.5 M) calcium chloride solution and 4 ml (0.5 N) sodium hydroxide solution were then added to stop the reaction. The samples were swirled, centrifuged at 9000 rpm for 30 min, and read at 420 nm in a UV-visible spectrophotometer. Moisture was determined to express enzyme activity as μg pNP released per gram of dry soil per hour.
In order to analyze the macrozoobenthic community, a duplicate sediment sample from each station was collected using Ekman grab, pooled, and then sieved (60 mesh size). The materials remaining on the sieve were preserved with a 5% formaldehyde solution. Macrozoobenthos were identified up to their species level following Subba Rao (1989). Organisms were counted, and their density was expressed as no. m−2. The species diversity of macrozoobenthic communities was calculated using the equation: H = ∑ (pl) |ln pl|, where H is the Shannon index and (pl) is the relative abundance of species “l” in the community.
Netplankton was collected by sieving 50 litres of water from the subsurface layer (20–30 cm) through a plankton net (25 no. bolting silk), and the samples were preserved in a 5% formaldehyde solution. Individuals were identified to the generic level following Ward and Whipple (1959) and Needham and Needham (1962). All genera were enumerated using a Sedgwick–Rafter cell counter and the numbers were then converted to densities (cells/filaments/colony forming units l−1). The generic diversity of the plankton was calculated using Shannon index (H), = ∑ (pl) |ln pl|, where (pl) is the relative abundance of genus “l” in the community.
Fish catch data of the Samaguri and Haribhanga wetlands were collected using a structured schedule from the respective fishers' society at monthly intervals. In both the wetlands, the bulk of the fish were caught using the brush park fishing practice, which is prevalent in the area and involves encircling the brush park with a large multi-mesh drag net. Fish were also caught by gill netting in the macrophyte-infested area and drag netting was used in deep pools in these wetlands. Fish yield was calculated from monthly catch data and effective water area of the wetlands.
Hydrochemical, sediment enzyme, chemical variables, plankton, and macrozoobenthos density were represented using mean and standard error. Differences in the means of various hydrochemical variables, sediment variables, and density of plankton and macrozoobenthos were compared between the wetlands using one-way analysis of variance (ANOVA) (Sokal and Rolph, 1981).
Results and discussions
The Bhomra wetland, despite having a link channel, remained hydrologically isolated from the parent river during the study period due to insufficient rainfall in the region, whereas the Chhari Ganga wetland remained perennially connected. The wetlands showed a similar mean annual air and water temperature (Table 1) pattern, as these wetlands are situated in the same agroclimatic region. Water depth during the three seasons was significantly greater (p < 0.001) in the open (1.81–9.52 m) than in the closed (0.24–2.74 m) wetland. The water drawdown due to seasonal change from monsoon to summer was greater in the closed wetland (74%) than in the open wetland (40%), and some sectors of the closed wetland even dried out during summer. The pH was alkaline for both wetlands. Significantly lower water transparency (p < 0.04) was observed in closed compared to open wetland, indicating higher water productivity in the closed wetland. Specific conductivity, free CO2, Ca+2, Mg+2 and Silicate (Si) were also higher (p < 0.04) in the closed wetland. Total hardness and total alkalinities were higher in the closed wetland, but dissolved oxygen was higher in the open wetland (Table 1).
Regular inflow of river water into the open, perennial wetland made the wetland dynamic and maintained greater water depth, good physicochemical properties, and diluted nutrient levels. Water inflow increased dissolved oxygen and replenished nutrients, maintaining water quality. In contrast, in the closed wetland, water depth decreased substantially and some sections even became dry in summer. Reduction in water level during summer caused more oxygen entry into sediment, favoring release of more nutrients through aerobic microbial processes in pore and column water (D'Angelo and Reddy, 1994). A different pattern of water quality parameters in terms of higher specific conductivity, free CO2 and lower total hardness, dissolved oxygen, and silicate-Si in seasonally open wetlands of Assam was reported by Manna and Aftabuddin (2007). This may be due to regional variation of soil reaction and the perennial nature of our open wetland.
Sediment enzyme and other physicochemical properties
Sediment of wetlands provides basic nutrients and texture for growth of microbes causing subsequent decomposition of organic matter of allochthonous and autochthonous origin and mineralization of nutrients (McLatchey and Reddy, 1998). The overall microbial activity is generally expressed as dehydrogenase activity. In our study, the mean activity of dehydrogenase was higher (p < 0.004) in the closed wetland (Table 2). Similarly, alkaline phosphatase (p < 0.001), acid phosphatase (p < 0.000), and ß-glucosidase (p < 0.01) activity were also significantly higher in the closed wetland (Table 2). Overall sediment microbial activity, phosphorous cycling, and carbon cycling enzymes were respectively 1.25, 1.65 and 1.1 times higher in the closed wetland.
In the closed, macrophyte dominated, eutrophic wetland, acid phosphatase activity was higher than alkanine phosphatase activity, while both were nearly the same in the open wetland. Sediment moisture was higher and temperature was slightly higher in the open wetland. Sediment pH was nearly equal in both wetlands and was in the neutral range. Notably higher organic matter, specific conductivity and oxidative state occurred in sediments of the closed wetland (Table 2).
Significantly higher activity of all studied enzymes in the closed wetlands may be attributed to low water depth and higher organic matter content. Higher organic matter and specific conductivity of sediments were also reported by Sugunan et al. (2000) and Manna and Aftabuddin (2007) in closed rather than in seasonally open wetlands in Assam and West Bengal. Water depth in closed wetlands fluctuates widely during the year, alternately creating aerobic and anaerobic conditions that may stimulate organic matter decomposition and nutrient release (D'Angelo and Reddy, 1994). Low water depth caused closed wetlands to support more macrophytic and periphytic vegetation due to the higher water column nutrient level and greater light penetration. The higher organic matter in closed wetlands is due to more macrophyte vegetation and the absence of flushing of bottom sediment by floodwaters. All these factors ultimately govern the succession of closed wetlands from eutrophic to a hypereutrophic state.
Plankton and macrozoobenthos
The study was conducted over a period of one year and revealed the presence of 17 and 18 phytoplankton genera from Bhomra and Chhari Ganga wetlands, respectively, of which 10 genera were found common to both wetlands (Table 3). Generic comparison of phytoplankton of the two wetlands revealed a higher magnitude of Volvox and Spirogyra belonging to Chlophyceae, and Oscillatoria belonging to Cyanophyceae in the Bhomra wetland. On the other hand, a total of 11 genera of zooplankton were recorded in the wetlands, of which 8 were common to both. Density of Cyclops was higher in Bhomra, while Ceriodaphnia was denser in Chhari Ganga (Table 3).
Thirteen macrozoobenthic species were recorded from the Chhari Ganga and Bhomra wetlands, of which 11 species were common to both (Table 4). The closed Bhomra wetland is unique in occurrence of Lamellidens marginalis and Chironomoussp., while Brotia costula and Chaoboroussp. were unique in Chhari Ganga. Gabbia orcula, Gyraulus convexicularis, Indoplanorbis exustus and Thiara tuberculata were of higher magnitude in the closed rather than the open wetlands (Table 4).
The effect of connectivity was also reflected in the density and diversity of netplankton and macrozoobenthic communities (Table 5). Annual average density of plankton was marginally higher in the closed (344 units l−1) than in the open (236 units l−1) wetlands and was found to be statistically insignificant (p > 0.4). However, macrozoobenthos density was significantly (p < 0.01) higher in the closed (13,364 no. m−2) than open wetlands (6931 no. m−2).
The mean generic diversity of plankton in the open wetland (1.85) was higher than in the closed wetland (1.28). The range of plankton generic diversity in the open wetland (1.48–2.22) was narrower than in the closed wetland (0.58–2.17). The narrower range of photoplankton generic diversity throughout the year in the open wetland compared to the wider range in the closed wetland, indicated that open wetland is much more stable in maintaining plankton diversity. However, the diversity of the sedentary macrozoobenthic community was almost identical (0.88 and 0.86) in both wetlands, although the closed wetland had a wider range of diversity throughout the year (Table 5).
The influx of water throughout the year in open wetlands maintains ecological integrity by diluting nutrients and lowering the abundance of plankton and macrozoobenthos. In contrast, water level fluctuations favor fluctuations in diversity of plankton and macrozoobenthos of closed wetlands. Nutrient-rich status due to higher nutrient regeneration facilitates higher planktonic and macrophytic growth in closed wetlands. Higher abundance of submerged and emergent macophytes in closed wetlands due to retention of organic matter caused greater density and fluctuating diversity of macrobenthic communities (Manna and Aftabuddin, 2007). Use of macrobenthic communities as an indicator to assess the impact of restoration processes and estuarine connectivity showed promising implications (Talley and Levin, 1999).
Fish yield and catch composition of two ecologically distinct wetlands of the Brahmaputra basin in Assam were studied. The fish yield in the open wetland, Samaguri, was low (102 kg ha−1yr−1) compared to the closed wetland, Haribhanga (292 kg ha−1yr−1).
Considering poor auto-stocking in the closed wetland, regular stocking with Indian Major Carp and Exotic Carp as a measure of enhancement resulted to higher yield. On the other hand, stocking in Samaguri was of a lower magnitude and supplementary in nature due to auto-stocking of some natural fish groups such as Minor Carp and Clupeid, the risk of fish escape, and poor retrieval that resulted in low fish yield (Table 6).
In open wetland, naturally occurring fish species contributed 53% of the total catch as compared to only 37% in closed wetland (Figure 2). Among natural fish species, L. gonius, C. reba and G. chapra could get access to the wetland through the link channel that favoured their recruitment and propagation and contributed substantially (38%) to the fish yield from the open wetland. On the other hand, minnows and other small fish (Table 6) dominated (35%) the natural fish in the closed wetland, Haribhanga, because of low predatory pressure as observed in the catch composition (Table 6 and Figure 2). Shallow depth (80–175 cm) and fishing pressure in the closed wetland lead to the removal of large predators. As a result, minnows and other small fish multiplied profusely in this ecosystem. A fluviatile condition observed in Samaguri during monsoon facilitated the access of small distant migrants G. chapra through the link channel from the Kolong River for breeding in the wetlands. Similarly, the environment created during monsoon also induced the breeding of Minor Carp, which contributed to the fishery in the open wetland.
This study revealed that higher nutrient concentration, greater sediment nutrient regeneration, and enzymes, coupled with shallow depth in the closed wetland, might have resulted in greater column and benthic productivity, while higher sediment organic matter accumulation could be considered as an indicator of hypereutrophication. The need for link channels for maintaining sensitive biodiversity, particularly for fish requiring local migration for their breeding and propagation, has become more obvious from the study. Productivity in terms of fish production may be of lower magnitude, but its significance in respect to biodiversity conservation is of paramount importance. The wetlands with functional link channels were still in pristine condition, with a relatively lower magnitude of macrozoobenthos and higher indigenous fish species. Lack of river connectivity reduced diversity of the biotic communities, clearly demonstrating deleterious impacts on indigenous fish and necessitating external stocking to sustain the fishery of the closed wetlands. Restoration of link channels connecting wetlands to parent rivers has, therefore, become warranted to maintain biodiversity and productivity of closed wetlands. Fishery development plans for wetlands, therefore, should emphasize conservation and establishment of linking channel as a mandatory component, at least during the monsoon season.
The authors are extremely grateful to staff of Reservoir and Wetlands Division, Central Inland Fisheries Research Institute, Barrackpore, especially Dr. B. K. Biswas, Mr. Subroto Das, Mr. B. Naskar and Mrs. S. Saha and M. Ali for rendering help in collecting and analyzing the samples.