Narmada, the oldest river system in India, originates from Amarkantak in Madhya Pradesh, flows east-west, and joins with the Gulf of Cambay on the Arabian Sea. The river drains 45.64 km3 of annual run-off and a series of dams was proposed to hold some of its water resources for multipurpose use. Currently, three dams have been built in Madhya Pradesh and one is under construction in Gujarat. A comparison of pre- and post-impoundment eco-environment and fisheries revealed changes in water quality, productivity, and aquatic flora and fauna of the river system. Among the fish, species like Tor tor, Labeo fimbriatus and Labeo dyocheilus suffered most. The percentage contributions to total yield of Carp, Catfish, and miscellaneous groups have significantly changed, indicating falls of 17%, 36% and an increase of 410%, respectively. Percentage contributions to catches of Macrobrachium rosenbergii and Tenualosa ilisha have also declined by 46% and about 75% in the estuarine stretch of the river system. Suitable conservation measures for sustenance and development of fishery have been suggested.
The Narmada River, synonymous with the goddess Narmada Mai, has immense aesthetic and religious significance to Indians, especially the people of the central and western regions. The river originates near Amarkantak at about 1050 m above MSL in the Maikaley highlands, flows westward through the hilly terrain and highlands of Madhya Pradesh, and descends to the Potamon plains in Gujarat before merging with the Gulf of Cambay on the West coast. With a total length of 1312 km, the river is wholly fed with run-off discharge from a 98,796 km2 catchment in Madhya Pradesh, Gujarat and Maharashtra.
The entire Narmada basin is being developed under a comprehensive river valley project programme through a series of dams, which should contribute additional fishery resources in reservoirs and enhance inland fish production for India. However, the ecology and fisheries of the river will suffer adversely due to the river modifications. This study evaluates the impact of damming on ecology and fisheries of the Narmada River system.
River resource and valley projects
Hydrographically, the river has several key features: initial hill streams, followed by a highland and ravine stretch, and lastly, a long span of tidal-influenced estuary. The moderate to high range of annual precipitation in the catchment produces a huge volume of run-off through the river course. The 98,796 km2 catchment yields 45.64 km3 of annual run-off, of which 34.50 km3 is utilizable. As early as the late nineteenth century, the British Government considered a proposal for a barrage at Bharuch but did not proceed. After independence, the Government of India took up the issue of damming the river system for development of multipurpose river valley projects. Thirty large dams were proposed under the Sardar Sarovar and Narmada Sagar projects, of which the Tawa, Bargi and Indira Sagar dams have been completed in Madhya Pradesh with the Sardar Sarovar dam currently under construction in Gujarat (Figure 1). After completion of the cascade of dams, 27,421 ha of reservoir area will be available for facilitation of irrigation, hydropower projects, industrial and domestic water supply, and fisheries.
The Narmada River basin has three major physiographic divisions: the Upper Narmada basin, the Central Highlands and the Broach-Baroda plains. The upper Narmada zone of the river runs over black granitic rocks. The Central Highlands are characterized by clay, gravel, boulders and coarse sands. In the Broach-Baroda plains, the river flows through black alluvial soils. Blocking the river course with dams has caused alterations in basin conditions. In the upper valley project areas, a large number of hills and hillocks were submerged, resulting in an uneven depth profile all along the captive river basin.
The basin at the Central Highlands suffered from abnormally low flows of river discharge and subsequently low depth and exposure of a greater part of the riverbed almost year round. Extraction of boulders and pebbles from the riverbed for construction activities has caused damage to the habitats of flora and fauna of the region. The impact of river valley projects has mainly been on river dimensions in the lowermost Broach-Baroda plains of the river system.
Before the river valley projects, the water temperature of the river system varied with altitude and season. At higher altitude, the annual range was milder (15.0–30.5°C) compared to the Central Highlands and lower plains (19.0 to 33.0°C). The mean ambient temperature in the hilly and downstream river section differed by 9.0°C in winter and 7.0°C during summer. Temporal variation in ambient temperature, though not very apparent, has been erratic and higher in the years of drought and low river discharge.
River water was more turbid during the monsoon and turned clearer in the winter months. In the stretches running through Madhya Pradesh, Unni (1996) reported that river water turbidity ranged between a trace and 25 NTU at higher altitude and increased to 260–400 NTU in the downstream courses. The higher values occurred during the monsoon season when the catchment run-off increased the silt contents of the water. Presently, turbidity values have changed in the middle Highlands and the lower plains regions of the river. The values remain higher, between 310 and 480 NTU, with highest turbidity near the drainage points. Catchment denudation has caused high silt contents in run-off waters and the river.
Chemical qualities of the river water have not changed much in the upper hilly region since the river valley projects. In the middle and lower zones, the level of dissolved oxygen recorded fluctuated over a wide range (4.4–9.1 ppm) and the lower values were mostly observed at the wastewater drainage points. Oxygen also dipped to very low levels in the hydrophyte-infested stretches during the night and early morning hours. The free CO2 content of the water has not changed perceptibly during the post-river valley project period. The chloride content of water has stayed at the level of the pre-impoundment period in the greater part of the river (4.6–19.9 ppm). However, the ambient chloride values have increased in the lower plains (615–3248 ppm) because of the decrease in freshwater discharge from upstream and the lower rate of dilution of the incoming tidal salinity from the Arabian Sea.
The texture of the river was mostly sandy pre-impoundment. The percentages of sand, silt, and clay varied within the ranges of 66.7–94.6%, 9.3–26.3% and 6.3–23.5%, respectively. The sediment was acidic to alkaline in situ (pH 6.46 to >9.00), with a moderate organic content (0.094–1.59%) and a C:N ratio of 7.95:26.20. Sediment nutrient levels (Av. N: 3.50–36.90 mg 100 g−1; P: 0.200–1.611 mg 100 g−1) were moderate.
The overall nutrient contents of the river water have changed marginally during the post-river valley project period. Post-impoundment phosphate (P) values were comparatively low in upper reaches of the river (P = 0.002–0.03 mg l−1) and increased in the middle stretch (P = 0.052–0.095 mg l−1), declining again in the lower estuarine zone (P = 0.017–0.033 mg l−1). The nitrate (as N) content was higher than phosphorus and the mean values of the element amounted to 0.172, 0.1315, and 0.1630 mg l−1 in the upper, middle, and lower zones of the system, respectively.
No information on primary productivity in the river was available for the pre-impoundment period. Unni (1996) studied productivity for the post-impoundment period and reported higher gross primary production in the upper hilly range (450–600 mg C m−3h−1) during May and comparatively low values in remaining part of the system. Singh (2009) reported the value of GPP as more in the middle sector (52.9–135.3 mg C m−3h−1) than the upper (67.1–91.3 mg C m−3h−1) and lower (36.3–89.4 mg C m−3h−1) sectors during the period of investigation. The higher primary production in the upper reaches was associated with higher abundance of algal populations in the region. During the post-impoundment period, primary productivity and plankton abundance levels decreased over time throughout the river system (Table 1). The findings suggest that the nutrient levels and the primary productivity of the river were not stable and depended on various extrinsic and intrinsic factors (Welch, 1952).
The Narmada River plankton studied by Unni (1996) covered nearly 550 km of the river course from 19 sampling stations. Mean phytoplankton density varied with time and space. In the upper zone, maximum density was found at Amarkantak (29140 u. l−1) near the origin of the river. The density declined to 2914 u. l−1 at 213 km downstream, and then increased to 8754 u. l−1 at 325 km. Thereafter at Sandia, about 500 km downstream, density decreased to 5890 u. l−1. An increase was recorded near Hoshangabad (7448–10168 u. l−1). The spatial variation in density of phytoplankton indicated that the abundance of organisms was in highest occurrence in high altitude and hilly terrain, and gradually decreased downstream. Among phytoplankton, Bacillariophyceae was dominant, followed by Chlorophyceae and Cyanophyceae. At Amarkantak, Chlorophyceae were dominant, possibly due to their luxuriant growth in semi-stagnant and cold environmental conditions. The percentage contribution of these three major groups of phytoplankton varied from place to place and with the time, and was related to various physicochemical factors like temperature, alkalinity, and nutrients. The maximum growth of phytoplankton occurred in the post-monsoon period, winter and late summer months. Unni (1996) recorded a total of 174 species of phytoplankton from the 500 km stretch between Amarkantak and Sethanighat. Out of the total, 101 species in 27 genera belonged to Bacillariophyceae, while 46 species in 21 genera were Chlorophyceae. Cyanophyceae, Euglenophyceae, Dinophyceae were represented by 19, 4, and 3 species, respectively, and Chrysophyceae was represented by a single species. Among zooplankton, the species diversity was much lower compared to phytoplankton and their density was also quite low compared to the latter. Zooplankton density was higher in the upper reaches (59.0–3132.1 u. l−1) compared to the middle sector (8.3–14.4 u. l−1). Rotifers were much higher in density in the upper reaches (29.5–1695.6 u. l−1). In the middle stretch, the percentages of Rotifers, Cladocerans, Copepods, and Ostracods were not uniform and varied among stations. In total, 111 zooplankton species were recorded in the Narmada River system from the upper and middle zones. No data on plankton population and diversity for the period prior to the implementation of river valley projects were available. Singh (2009) reported very low abundance of plankton in the Narmada River estuary.
River systems can sustain populations of various macrobenthic organisms and the diversity and density of the organisms will be indicative of environmental conditions. The Narmada River, with existing, ongoing, and proposed river valley projects, faces the pressure of severe shortages of river flow and a resultant acute shrinkage of habitat areas for the benthic organisms. Riverbed with mostly gravels, pebbles, and boulders has been gradually replaced by coarse sand bed, which does not support the growth of macrobenthic fauna. Singh (2009) recorded 72 species of macrobenthic organisms from the Narmada River system and found the organisms homogeneously distributed, with dominance of few species varying with time and space. The overall densities fluctuated within 108–13739 nos. m−2. In the middle sector and estuarine zone, the abundance of the organisms varied widely in the range of 69–1493 nos. m−2 and 118–13,739 no. m−2 where Molluscs were dominant. Thiara scabra, Thiara tuberculata, Bellamaya bengalensis, Gyraulus sp., Lymnaea acuminate, Corbicula striatella, Lamellidens marginali and Sphaerium sp. were important Molluscan taxa.
Pre-impoundment, the rich base substrate of the Narmada River course sustained good growth of periphyton. The boulders, pebbles, gravel, and coarse sands provided substrata for a complex group of organisms and provided important feeding grounds for fish and shrimp. Unni (1996) made the detailed studies on these organisms in the late 1980s. He found the growth of periphytic organisms was highly variable in the upper reaches (2700–41,000 u. cm−2) and increased in the downstream stretch between Sandia and Hoshangabad (48,000–89,000 u. cm−2) where the necessary substrates were abundant and the ecological conditions were also favourable for the growth of periphyton. Bacillariophyceae made up the bulk of the periphyton population throughout the upper and middle Highland stretches, while Cyanophyceae and Chlorophyceae were also present in lower percentages.
During the post-impoundment period, the drastic fall in river discharge led to shrinkage of the river spread area and exposure of the periphyton-rich substrata. In addition to periphyton, benthic macrophytes also disappeared to a great extent in the succeeding years.
Macrophyte and associated fauna
Both submerged and floating aquatic plants were common in shallow and low flowing riverine stretches before the river valley projects. Species like Potamogeton crispus, P. pentinatus, Najas urinor, Hydrilla verticillalata, Vallisneria spiralis and Chara sp. were recorded in varying quantities between Manot and Hoshangabad by Unni (1996). Overall, the growth of macrophytes was observed to be faster after winter months when discharge was lower, water remained clear, and the flow was moderately low in the river system. During the investigations carried out by a CIFRI team of scientists in 1999–2001 (Anon, 1999-2001), they observed that the lower stretch of the river below the Sardar Sarovar dam showed varying intensities of macrophyte cover in shallow and low-flowing stretches. At a number of places, the density was high during March to June and the vegetation supported good growth of associated fauna, mainly gastropods, annelids, and insects. With the discharge decreasing to almost negligible levels after completion of the Sardar Sarovar dam, the scenario will be further altered and the shallower parts of the river will gradually be transformed into lacustrine habitat.
Fish and fisheries
The fish fauna of the Narmada River system has been studied by various workers. Hora and Nair (1941) recorded 40 species from Satpura range, and Karamchandani et al. (1967) later listed 77 species from the upper and middle zones. Doria (1990) also mentioned the presence of 76 species from Narmada in Madhya Pradesh. Rao et al. (1991) included the western sector of the river in their fish faunal investigations and prepared a list of 84 species from the system. This may not be considered a complete account of total fish fauna of the system and research should be continued.
Records of annual production of fish from the Narmada River system are insufficient for time scale evaluations and trend analyses. Unni (1996) reported comparative production figures of various Indian rivers and mentioned that Narmada stood at the lower end compared to the East Coast river systems. Annual fish production from the river in Madhya Pradesh was estimated at 269.8 metric tons (Dubey, 1984) between 1958–1959 and 1965–1966, i.e. prior to the development of dams on the river. At that time, Carp alone contributed 176.5 metric tons per year, followed by Catfish (79.5 mt) and miscellaneous groups (13.8 mt). During 1971 and 1972, nearly an 11.1% increase was recorded in total production (300.0 mt) in Madhya Pradesh. The contribution of Carp increased to 186.3 mt, while production of Catfish showed little change (81.0 mt), and miscellaneous species were enhanced by a considerable level. Later, after the commissioning of the valley projects, the riverine production in the same stretches dropped to 100.0 mt (Rao et al., 1991), indicating the adverse impact of the man-made obstructions on the natural river flow and subsequent hydro-ecological changes. The three major groups—Carp, Catfish, and miscellaneous—changed considerably in their percentile contribution to the annual catch (Table 2). The percentage contribution of Carp production has dropped by 17%, while the contribution of Catfish suffered the most with a 36% decrease within two decades. The contribution of the miscellaneous catch, however, increased by 410% during the same period. The changes in population structure of the fish occurred due to the alteration in habitat conditions. Hydro-graphical changes like lowering of depth, disappearance of deep pools, and exposure of suitable breeding and nursery areas were responsible for limiting fair weather sheltering places, free movement, and propagation of large sized Carp and Catfish.
The fisheries in the lower Broach-Baroda plains were different. Hilsa (Tenualosa ilisha) migrated into freshwater breeding areas of the stretch and were an important component in annual fisheries. Freshwater Carps and Catfish also contributed to the fisheries. In addition, the estuarine and neretic species contributed the bulk of the fisheries in the lower stretches of the Narmada River system. Macrobrachium rosenbergii, the giant freshwater prawn (Figure 2), contributed an important fishery in river-estuarine stretches. Installation of the Sardar Sarovar dam at the head region of the lower plains at Vedgam led to restriction in river flow. As a result, the hydro-ecology of the river has significantly altered and the effect has been seen in the fisheries of the estuarine sector. The percentage share of Carp, mainly Tor tor, Labeo fimbriatus, and L. dyocheilus species, markedly declined. Also, the contribution of large-sized Catfish (Wallago attu, Sperata aor, S. seenghala) was reduced and production of the group was offset with medium- and small-sized species. The abundance of Gegra (Rita pavimentata) has conspicuously fallen in the past 10 to 15 years due to the loss of their favoured rocky and pebbled habitats.
Among the important Carp species, Tor tor (Figure 3) had very low production in recent years. Karamchandani et al. (1967) reported the production of the species to fluctuate between 5.7 and 9.6 metric tons, sharing 25.5–29.6% of the total catch during 1958–1966 at a landing station (Shahganj) in Madhya Pradesh, indicating the abundance during the pre-impoundment period. Physical barriers due to the high dam restricted the breeding movements to the hilly terrains and negatively affected the natural recruitment system for the species. Further, the natural feeding grounds downstream have also disappeared, affecting the population growth of Tor tor. Prior to commissioning of the river valley projects, the annual production of Mahseer (Tor tor) from a 48 km stretch in Madhya Pradesh was 7.6 t, contributing 28% of the total and 46% of the Carp fishery (Karamchandani et al., 1967). Though the percentage contribution of Tor tor has not shown much variation between pre- and post-impoundment periods, total production of the fish has declined along with the decline of Carp as a group from the system. Apart from commercial fisheries, the egg production potential of the species has also suffered significantly. Unni (1996) mentioned the river as the only natural source of Mahseer eggs because of high and sustained abundance. However, studies conducted by the Government of Madhya Pradesh between 1987–1988 and 1995–1996 showed decline in fry production potential of the species by nearly 78% within three decades of commissioning of the river valley projects. Based on information available from the Government of Madhya Pradesh, Mahseer production in the Narmada River also dropped to a low of 53 t in 1996–1997 from 330 t recorded in 1992–1993; this is a matter of great concern.
The reproduction and population growth of Foothill Carp, Labeo fimbriatus and Labeo dyocheilus, suffered like T. tor, in the post-impoundment eco-habitats of the river. Giant freshwater Prawn (M. rosenbergii) fisheries have shown a declining trend during the installation of the Sardar Sarovar dam. Loss of habitats due to the controlled flooding of riverbeds and change in depth profiles created major constraints for Prawn fisheries.
Hilsa (Tenualosa ilisha), the anadromous species (Figure 4), migrate from the Arabian Sea to freshwater habitats in the Narmada River for reproduction and population growth. It has not been reported from the highlands and upper hilly terrain of Madhya Pradesh (Dubey, 1984). Karamchandani et al. (1967) were able to collect good numbers of Hilsa eggs (95,000) during the monsoon season for five consecutive years between 1959 and 1966, and the collection sites covered wide stretches of the lower plains in Gujarat. These findings showed that the breeding season commenced in June–July and continued into September. The season and location of Hilsa breeding grounds did not change greatly even after the installation of dams upstream because of the negligible impact of dams on breeding and nursing habitats of the downstream stretches.
The Sardar Sarovar dam was built stage-wise in Gujarat, gradually blocking the normal river flow, restricting the Hilsa migration range, and also causing a shift in their breeding grounds. Instead of moving up to 160 km upstream, the brood stock of Hilsa only ascended to 110 km and bred within 100 km of the Gulf of Cambay. In the recent past, the fishing grounds have moved further downwards to the Gulf of Cambay, where about 90% of the catch is now harvested. The annual fishery for Hilsa fluctuated even during the pre-impoundment period, and higher catches coincided with high floods in the river. During the filling of the Sardar Sarovar dam, the production, though variable, has shown a declining trend (Figure 5). Annual catch of 16,000 t of the species during 1990–1991 reduced to 4000 t in 2007–2008 and indicated a 75% decline in production over a period of one-and-a-half decades.
The control of the river discharge has detrimentally affected the migration of Hilsa and their abundance in the river. Change in salinity patterns and gradual shrinkage in freshwater habitats have cumulatively affected the Hilsa fishery.
The Narmada River system plays a significant role in the socioeconomic development of Madhya Pradesh, Gujarat, and to some extent Maharashtra. Consumptive uses of river water are increasing rapidly and cannot be replaced unless suitable alternative sources are found. Given this situation, it is imperative to develop and implement effective measures for conservation of available riverine resources, as well as the river's productivity and production functions.
Conservation of the resources
There is a need to improve and conserve the river water resources for the protection of the natural flora and fauna through restoration of their habitats. Natural basin characteristics should not be altered, in view of their important role in providing necessary breeding and feeding habitats for a number of commercial species, particularly the foothill Carps. Deep pools play an important role in sheltering fish stocks, especially large-sized fish during fair-weather periods. An extensive survey of deep pools in all river sections of stakeholder states is needed, with mapping of the hydro-ecological status of pools as a basis for regular monitoring of their status. There is a need to declare the deep pools and gorges as sanctuaries or non-fishing areas. Conservation of these deeper areas of the river would help in protecting the brood stocks of the fish and aid in the recovery of their depleted stocks.
The populations of a number of commercially important species have suffered during the post-impoundment period. Species like Tor tor, Labeo fimbriatus and Labeo dyocheilus are rare in the downstream area of the middle highlands and Broach-Baroda plains. All these species need supplementary support to rebuild the dwindling stocks. Accordingly, artificially bred and reared juveniles of the concerned species should be introduced in properly identified and well-protected areas. Also, the spawn and eggs of other Carps (Catla catla, Labeo rohita, Cirrhinus mrigala) need to be released to restore and strengthen populations in depleted zones.
Conservation of fish stocks
Overfishing is one of the major causes of the depletion of the fish stocks and the composition of the catches. Free fishing in the concerned states is an unstated right of fishers and until ad-libitum fishing is controlled, little progress can be made towards conservation and development of fisheries. Given the importance of conservation for fish stocks, measures such as non-fishing seasons and areas, sanctuaries, and size limits for commercial species will have to be strictly enforced. The state authorities need to formulate the necessary measures and mechanisms for effective implementation.
Use of river waters for waste removal is commonly practiced throughout the Narmada River region. Both non-point and point sources have contributed to degrading water quality and habitats for aquatic flora and fauna. Discharges from cities and industrial areas in Madhya Pradesh, Gujarat and Maharashtra are gradually increasing the pollution loads, which are either untreated or semi-treated. Singh (2009) identified some environmental hot-spots in the Gulf of Cambay. To overcome the problem of pollution and related hazards for the ecosystem and fisheries, positive measures like agglomeration of like industries, adoption of common treatment facilities, execution of zero-effluent measures, and safe disposal of different harmful industrial elements must be adopted.
The Narmada is a rain-fed system and the annual run-off is dependent on the rate of water flow in the catchment areas. Therefore, to restore desired and optimum habitat conditions in the dam-affected river stretches, a suitable flow regime needs to be maintained. Research in this regard is needed and the recommended rate of discharge may be maintained, if not throughout the year then at least for the crisis periods of the flora and fauna.
Increased awareness among stakeholders is essential for the conservation and development of fisheries. In the states of Madhya Pradesh, Gujarat and Maharashtra, the fishers are mostly illiterate and need increased awareness for conservation and eco-friendly exploitation of the fisheries' resources. Further, other stakeholders using the river water for different purposes should also be made aware through programmes for the formulation and implementation of wide holistic conservation measures and the implementation of a common platform.